[logo]     MINIX 3 - Appendix B [logo]

Index PostScript PDF ASCII text Tar file


_________________________ Page 639    File:  _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/ansi.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00000	/* The <ansi.h> header attempts to decide whether the compiler has enough
 00001	 * conformance to Standard C for Minix to take advantage of.  If so, the
 00002	 * symbol _ANSI is defined (as 31459).  Otherwise _ANSI is not defined
 00003	 * here, but it may be defined by applications that want to bend the rules.
 00004	 * The magic number in the definition is to inhibit unnecessary bending
 00005	 * of the rules.  (For consistency with the new '#ifdef _ANSI" tests in
 00006	 * the headers, _ANSI should really be defined as nothing, but that would
 00007	 * break many library routines that use "#if _ANSI".)
 00008	
 00009	 * If _ANSI ends up being defined, a macro
 00010	 *
 00011	 *      _PROTOTYPE(function, params)
 00012	 *
 00013	 * is defined.  This macro expands in different ways, generating either
 00014	 * ANSI Standard C prototypes or old-style K&R (Kernighan & Ritchie)
 00015	 * prototypes, as needed.  Finally, some programs use _CONST, _VOIDSTAR etc
 00016	 * in such a way that they are portable over both ANSI and K&R compilers.
 00017	 * The appropriate macros are defined here.
 00018	 */
 00019	
 00020	#ifndef _ANSI_H
 00021	#define _ANSI_H
 00022	
 00023	#if __STDC__ == 1
 00024	#define _ANSI           31459   /* compiler claims full ANSI conformance */
 00025	#endif
 00026	
 00027	#ifdef __GNUC__
 00028	#define _ANSI           31459   /* gcc conforms enough even in non-ANSI mode */
 00029	#endif
 00030	
 00031	#ifdef _ANSI
 00032	
 00033	/* Keep everything for ANSI prototypes. */
 00034	#define _PROTOTYPE(function, params)    function params
 00035	#define _ARGS(params)                   params
 00036	
 00037	#define _VOIDSTAR       void *
 00038	#define _VOID           void
 00039	#define _CONST          const
 00040	#define _VOLATILE       volatile
 00041	#define _SIZET          size_t
 00042	
 00043	#else
 00044	
 00045	/* Throw away the parameters for K&R prototypes. */
 00046	#define _PROTOTYPE(function, params)    function()
 00047	#define _ARGS(params)                   ()
 00048	
 00049	#define _VOIDSTAR       void *
 00050	#define _VOID           void
 00051	#define _CONST
 00052	#define _VOLATILE
 00053	#define _SIZET          int
 00054	

_________________________ Page 640    File: include/ansi.h _________________________
 00055	#endif /* _ANSI */
 00056	
 00057	/* This should be defined as restrict when a C99 compiler is used. */
 00058	#define _RESTRICT
 00059	
 00060	/* Setting any of _MINIX, _POSIX_C_SOURCE or _POSIX2_SOURCE implies
 00061	 * _POSIX_SOURCE.  (Seems wrong to put this here in ANSI space.)
 00062	 */
 00063	#if defined(_MINIX) || _POSIX_C_SOURCE > 0 || defined(_POSIX2_SOURCE)
 00064	#undef _POSIX_SOURCE
 00065	#define _POSIX_SOURCE   1
 00066	#endif
 00067	
 00068	#endif /* ANSI_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/limits.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00100	/* The <limits.h> header defines some basic sizes, both of the language types 
 00101	 * (e.g., the number of bits in an integer), and of the operating system (e.g.
 00102	 * the number of characters in a file name.
 00103	 */
 00104	
 00105	#ifndef _LIMITS_H
 00106	#define _LIMITS_H
 00107	
 00108	/* Definitions about chars (8 bits in MINIX, and signed). */
 00109	#define CHAR_BIT           8    /* # bits in a char */
 00110	#define CHAR_MIN        -128    /* minimum value of a char */
 00111	#define CHAR_MAX         127    /* maximum value of a char */
 00112	#define SCHAR_MIN       -128    /* minimum value of a signed char */
 00113	#define SCHAR_MAX        127    /* maximum value of a signed char */
 00114	#define UCHAR_MAX        255    /* maximum value of an unsigned char */
 00115	#define MB_LEN_MAX         1    /* maximum length of a multibyte char */
 00116	
 00117	/* Definitions about shorts (16 bits in MINIX). */
 00118	#define SHRT_MIN  (-32767-1)    /* minimum value of a short */
 00119	#define SHRT_MAX       32767    /* maximum value of a short */
 00120	#define USHRT_MAX     0xFFFF    /* maximum value of unsigned short */
 00121	
 00122	/* _EM_WSIZE is a compiler-generated symbol giving the word size in bytes. */
 00123	#define INT_MIN (-2147483647-1) /* minimum value of a 32-bit int */
 00124	#define INT_MAX   2147483647    /* maximum value of a 32-bit int */
 00125	#define UINT_MAX  0xFFFFFFFF    /* maximum value of an unsigned 32-bit int */
 00126	
 00127	/*Definitions about longs (32 bits in MINIX). */
 00128	#define LONG_MIN (-2147483647L-1)/* minimum value of a long */
 00129	#define LONG_MAX  2147483647L   /* maximum value of a long */
 00130	#define ULONG_MAX 0xFFFFFFFFL   /* maximum value of an unsigned long */
 00131	
 00132	#include <sys/dir.h>
 00133	
 00134	/* Minimum sizes required by the POSIX P1003.1 standard (Table 2-3). */
 00135	#ifdef _POSIX_SOURCE            /* these are only visible for POSIX */
 00136	#define _POSIX_ARG_MAX    4096  /* exec() may have 4K worth of args */
 00137	#define _POSIX_CHILD_MAX     6  /* a process may have 6 children */
 00138	#define _POSIX_LINK_MAX      8  /* a file may have 8 links */
 00139	#define _POSIX_MAX_CANON   255  /* size of the canonical input queue */

_________________________ Page 641    File: include/limits.h _________________________
 00140	#define _POSIX_MAX_INPUT   255  /* you can type 255 chars ahead */
 00141	#define _POSIX_NAME_MAX DIRSIZ  /* a file name may have 14 chars */
 00142	#define _POSIX_NGROUPS_MAX   0  /* supplementary group IDs are optional */
 00143	#define _POSIX_OPEN_MAX     16  /* a process may have 16 files open */
 00144	#define _POSIX_PATH_MAX    255  /* a pathname may contain 255 chars */
 00145	#define _POSIX_PIPE_BUF    512  /* pipes writes of 512 bytes must be atomic */
 00146	#define _POSIX_STREAM_MAX    8  /* at least 8 FILEs can be open at once */
 00147	#define _POSIX_TZNAME_MAX    3  /* time zone names can be at least 3 chars */
 00148	#define _POSIX_SSIZE_MAX 32767  /* read() must support 32767 byte reads */
 00149	
 00150	/* Values actually implemented by MINIX (Tables 2-4, 2-5, 2-6, and 2-7). */
 00151	/* Some of these old names had better be defined when not POSIX. */
 00152	#define _NO_LIMIT          100  /* arbitrary number; limit not enforced */
 00153	
 00154	#define NGROUPS_MAX          0  /* supplemental group IDs not available */
 00155	#define ARG_MAX          16384  /* # bytes of args + environ for exec() */
 00156	#define CHILD_MAX    _NO_LIMIT  /* MINIX does not limit children */
 00157	#define OPEN_MAX            20  /* # open files a process may have */
 00158	#define LINK_MAX      SHRT_MAX  /* # links a file may have */
 00159	#define MAX_CANON          255  /* size of the canonical input queue */
 00160	#define MAX_INPUT          255  /* size of the type-ahead buffer */
 00161	#define NAME_MAX        DIRSIZ  /* # chars in a file name */
 00162	#define PATH_MAX           255  /* # chars in a path name */
 00163	#define PIPE_BUF          7168  /* # bytes in atomic write to a pipe */
 00164	#define STREAM_MAX          20  /* must be the same as FOPEN_MAX in stdio.h */
 00165	#define TZNAME_MAX           3  /* maximum bytes in a time zone name is 3 */
 00166	#define SSIZE_MAX        32767  /* max defined byte count for read() */
 00167	
 00168	#endif /* _POSIX_SOURCE */
 00169	
 00170	#endif /* _LIMITS_H */





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/errno.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00200	/* The <errno.h> header defines the numbers of the various errors that can
 00201	 * occur during program execution.  They are visible to user programs and 
 00202	 * should be small positive integers.  However, they are also used within 
 00203	 * MINIX, where they must be negative.  For example, the READ system call is 
 00204	 * executed internally by calling do_read().  This function returns either a 
 00205	 * (negative) error number or a (positive) number of bytes actually read.
 00206	 *
 00207	 * To solve the problem of having the error numbers be negative inside the
 00208	 * the system and positive outside, the following mechanism is used.  All the
 00209	 * definitions are are the form:
 00210	 *
 00211	 *      #define EPERM           (_SIGN 1)
 00212	 *
 00213	 * If the macro _SYSTEM is defined, then  _SIGN is set to "-", otherwise it is
 00214	 * set to "".  Thus when compiling the operating system, the  macro _SYSTEM
 00215	 * will be defined, setting EPERM to (- 1), whereas when when this
 00216	 * file is included in an ordinary user program, EPERM has the value ( 1).
 00217	 */
 00218	
 00219	#ifndef _ERRNO_H                /* check if <errno.h> is already included */

_________________________ Page 642    File: include/errno.h _________________________
 00220	#define _ERRNO_H                /* it is not included; note that fact */
 00221	
 00222	/* Now define _SIGN as "" or "-" depending on _SYSTEM. */
 00223	#ifdef _SYSTEM
 00224	#   define _SIGN         -
 00225	#   define OK            0
 00226	#else
 00227	#   define _SIGN         
 00228	#endif
 00229	
 00230	extern int errno;                 /* place where the error numbers go */
 00231	
 00232	/* Here are the numerical values of the error numbers. */
 00233	#define _NERROR               70  /* number of errors */  
 00234	
 00235	#define EGENERIC      (_SIGN 99)  /* generic error */
 00236	#define EPERM         (_SIGN  1)  /* operation not permitted */
 00237	#define ENOENT        (_SIGN  2)  /* no such file or directory */
 00238	#define ESRCH         (_SIGN  3)  /* no such process */
 00239	#define EINTR         (_SIGN  4)  /* interrupted function call */
 00240	#define EIO           (_SIGN  5)  /* input/output error */
 00241	#define ENXIO         (_SIGN  6)  /* no such device or address */
 00242	#define E2BIG         (_SIGN  7)  /* arg list too long */
 00243	#define ENOEXEC       (_SIGN  8)  /* exec format error */
 00244	#define EBADF         (_SIGN  9)  /* bad file descriptor */
 00245	#define ECHILD        (_SIGN 10)  /* no child process */
 00246	#define EAGAIN        (_SIGN 11)  /* resource temporarily unavailable */
 00247	#define ENOMEM        (_SIGN 12)  /* not enough space */
 00248	#define EACCES        (_SIGN 13)  /* permission denied */
 00249	#define EFAULT        (_SIGN 14)  /* bad address */
 00250	#define ENOTBLK       (_SIGN 15)  /* Extension: not a block special file */
 00251	#define EBUSY         (_SIGN 16)  /* resource busy */
 00252	#define EEXIST        (_SIGN 17)  /* file exists */
 00253	#define EXDEV         (_SIGN 18)  /* improper link */
 00254	#define ENODEV        (_SIGN 19)  /* no such device */
 00255	#define ENOTDIR       (_SIGN 20)  /* not a directory */
 00256	#define EISDIR        (_SIGN 21)  /* is a directory */
 00257	#define EINVAL        (_SIGN 22)  /* invalid argument */
 00258	#define ENFILE        (_SIGN 23)  /* too many open files in system */
 00259	#define EMFILE        (_SIGN 24)  /* too many open files */
 00260	#define ENOTTY        (_SIGN 25)  /* inappropriate I/O control operation */
 00261	#define ETXTBSY       (_SIGN 26)  /* no longer used */
 00262	#define EFBIG         (_SIGN 27)  /* file too large */
 00263	#define ENOSPC        (_SIGN 28)  /* no space left on device */
 00264	#define ESPIPE        (_SIGN 29)  /* invalid seek */
 00265	#define EROFS         (_SIGN 30)  /* read-only file system */
 00266	#define EMLINK        (_SIGN 31)  /* too many links */
 00267	#define EPIPE         (_SIGN 32)  /* broken pipe */
 00268	#define EDOM          (_SIGN 33)  /* domain error       (from ANSI C std) */
 00269	#define ERANGE        (_SIGN 34)  /* result too large   (from ANSI C std) */
 00270	#define EDEADLK       (_SIGN 35)  /* resource deadlock avoided */
 00271	#define ENAMETOOLONG  (_SIGN 36)  /* file name too long */
 00272	#define ENOLCK        (_SIGN 37)  /* no locks available */
 00273	#define ENOSYS        (_SIGN 38)  /* function not implemented */
 00274	#define ENOTEMPTY     (_SIGN 39)  /* directory not empty */
 00275	
 00276	/* The following errors relate to networking. */
 00277	#define EPACKSIZE     (_SIGN 50)  /* invalid packet size for some protocol */
 00278	#define EOUTOFBUFS    (_SIGN 51)  /* not enough buffers left */
 00279	#define EBADIOCTL     (_SIGN 52)  /* illegal ioctl for device */

_________________________ Page 643    File: include/errno.h _________________________
 00280	#define EBADMODE      (_SIGN 53)  /* badmode in ioctl */
 00281	#define EWOULDBLOCK   (_SIGN 54)
 00282	#define EBADDEST      (_SIGN 55)  /* not a valid destination address */
 00283	#define EDSTNOTRCH    (_SIGN 56)  /* destination not reachable */
 00284	#define EISCONN       (_SIGN 57)  /* all ready connected */
 00285	#define EADDRINUSE    (_SIGN 58)  /* address in use */
 00286	#define ECONNREFUSED  (_SIGN 59)  /* connection refused */
 00287	#define ECONNRESET    (_SIGN 60)  /* connection reset */
 00288	#define ETIMEDOUT     (_SIGN 61)  /* connection timed out */
 00289	#define EURG          (_SIGN 62)  /* urgent data present */
 00290	#define ENOURG        (_SIGN 63)  /* no urgent data present */
 00291	#define ENOTCONN      (_SIGN 64)  /* no connection (yet or anymore) */
 00292	#define ESHUTDOWN     (_SIGN 65)  /* a write call to a shutdown connection */
 00293	#define ENOCONN       (_SIGN 66)  /* no such connection */
 00294	#define EAFNOSUPPORT  (_SIGN 67)  /* address family not supported */
 00295	#define EPROTONOSUPPORT (_SIGN 68) /* protocol not supported by AF */
 00296	#define EPROTOTYPE    (_SIGN 69)  /* Protocol wrong type for socket */
 00297	#define EINPROGRESS   (_SIGN 70)  /* Operation now in progress */
 00298	#define EADDRNOTAVAIL (_SIGN 71)  /* Can't assign requested address */
 00299	#define EALREADY      (_SIGN 72)  /* Connection already in progress */
 00300	#define EMSGSIZE      (_SIGN 73)  /* Message too long */
 00301	
 00302	/* The following are not POSIX errors, but they can still happen. 
 00303	 * All of these are generated by the kernel and relate to message passing.
 00304	 */
 00305	#define ELOCKED      (_SIGN 101)  /* can't send message due to deadlock */
 00306	#define EBADCALL     (_SIGN 102)  /* illegal system call number */
 00307	#define EBADSRCDST   (_SIGN 103)  /* bad source or destination process */
 00308	#define ECALLDENIED  (_SIGN 104)  /* no permission for system call */
 00309	#define EDEADDST     (_SIGN 105)  /* send destination is not alive */
 00310	#define ENOTREADY    (_SIGN 106)  /* source or destination is not ready */
 00311	#define EBADREQUEST  (_SIGN 107)  /* destination cannot handle request */
 00312	#define EDONTREPLY   (_SIGN 201)  /* pseudo-code: don't send a reply */
 00313	
 00314	#endif /* _ERRNO_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/unistd.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00400	/* The <unistd.h> header contains a few miscellaneous manifest constants. */
 00401	
 00402	#ifndef _UNISTD_H
 00403	#define _UNISTD_H
 00404	
 00405	#ifndef _TYPES_H
 00406	#include <sys/types.h>
 00407	#endif
 00408	
 00409	/* Values used by access().  POSIX Table 2-8. */
 00410	#define F_OK               0    /* test if file exists */
 00411	#define X_OK               1    /* test if file is executable */
 00412	#define W_OK               2    /* test if file is writable */
 00413	#define R_OK               4    /* test if file is readable */
 00414	
 00415	/* Values used for whence in lseek(fd, offset, whence).  POSIX Table 2-9. */
 00416	#define SEEK_SET           0    /* offset is absolute  */
 00417	#define SEEK_CUR           1    /* offset is relative to current position */
 00418	#define SEEK_END           2    /* offset is relative to end of file */
 00419	

_________________________ Page 644    File: include/unistd.h _________________________
 00420	/* This value is required by POSIX Table 2-10. */
 00421	#define _POSIX_VERSION 199009L  /* which standard is being conformed to */
 00422	
 00423	/* These three definitions are required by POSIX Sec. 8.2.1.2. */
 00424	#define STDIN_FILENO       0    /* file descriptor for stdin */
 00425	#define STDOUT_FILENO      1    /* file descriptor for stdout */
 00426	#define STDERR_FILENO      2    /* file descriptor for stderr */
 00427	
 00428	#ifdef _MINIX
 00429	/* How to exit the system or stop a server process. */
 00430	#define RBT_HALT           0
 00431	#define RBT_REBOOT         1
 00432	#define RBT_PANIC          2    /* a server panics */
 00433	#define RBT_MONITOR        3    /* let the monitor do this */
 00434	#define RBT_RESET          4    /* hard reset the system */
 00435	#endif
 00436	
 00437	/* What system info to retrieve with sysgetinfo(). */
 00438	#define SI_KINFO           0    /* get kernel info via PM */
 00439	#define SI_PROC_ADDR       1    /* address of process table */
 00440	#define SI_PROC_TAB        2    /* copy of entire process table */
 00441	#define SI_DMAP_TAB        3    /* get device <-> driver mappings */
 00442	
 00443	/* NULL must be defined in <unistd.h> according to POSIX Sec. 2.7.1. */
 00444	#define NULL    ((void *)0)
 00445	
 00446	/* The following relate to configurable system variables. POSIX Table 4-2. */
 00447	#define _SC_ARG_MAX        1
 00448	#define _SC_CHILD_MAX      2
 00449	#define _SC_CLOCKS_PER_SEC 3
 00450	#define _SC_CLK_TCK        3
 00451	#define _SC_NGROUPS_MAX    4
 00452	#define _SC_OPEN_MAX       5
 00453	#define _SC_JOB_CONTROL    6
 00454	#define _SC_SAVED_IDS      7
 00455	#define _SC_VERSION        8
 00456	#define _SC_STREAM_MAX     9
 00457	#define _SC_TZNAME_MAX    10
 00458	
 00459	/* The following relate to configurable pathname variables. POSIX Table 5-2. */
 00460	#define _PC_LINK_MAX       1    /* link count */
 00461	#define _PC_MAX_CANON      2    /* size of the canonical input queue */
 00462	#define _PC_MAX_INPUT      3    /* type-ahead buffer size */
 00463	#define _PC_NAME_MAX       4    /* file name size */
 00464	#define _PC_PATH_MAX       5    /* pathname size */
 00465	#define _PC_PIPE_BUF       6    /* pipe size */
 00466	#define _PC_NO_TRUNC       7    /* treatment of long name components */
 00467	#define _PC_VDISABLE       8    /* tty disable */
 00468	#define _PC_CHOWN_RESTRICTED 9  /* chown restricted or not */
 00469	
 00470	/* POSIX defines several options that may be implemented or not, at the
 00471	 * implementer's whim.  This implementer has made the following choices:
 00472	 *
 00473	 * _POSIX_JOB_CONTROL       not defined:        no job control
 00474	 * _POSIX_SAVED_IDS         not defined:        no saved uid/gid
 00475	 * _POSIX_NO_TRUNC          defined as -1:      long path names are truncated
 00476	 * _POSIX_CHOWN_RESTRICTED  defined:            you can't give away files
 00477	 * _POSIX_VDISABLE          defined:            tty functions can be disabled
 00478	 */
 00479	#define _POSIX_NO_TRUNC       (-1)

_________________________ Page 645    File: include/unistd.h _________________________
 00480	#define _POSIX_CHOWN_RESTRICTED  1
 00481	
 00482	/* Function Prototypes. */
 00483	_PROTOTYPE( void _exit, (int _status)                                   );
 00484	_PROTOTYPE( int access, (const char *_path, int _amode)                 );
 00485	_PROTOTYPE( unsigned int alarm, (unsigned int _seconds)                 );
 00486	_PROTOTYPE( int chdir, (const char *_path)                              );
 00487	_PROTOTYPE( int fchdir, (int fd)                                        );
 00488	_PROTOTYPE( int chown, (const char *_path, _mnx_Uid_t _owner, _mnx_Gid_t _group)   
 00489	_PROTOTYPE( int close, (int _fd)                                        );
 00490	_PROTOTYPE( char *ctermid, (char *_s)                                   );
 00491	_PROTOTYPE( char *cuserid, (char *_s)                                   );
 00492	_PROTOTYPE( int dup, (int _fd)                                          );
 00493	_PROTOTYPE( int dup2, (int _fd, int _fd2)                               );
 00494	_PROTOTYPE( int execl, (const char *_path, const char *_arg, ...)       );
 00495	_PROTOTYPE( int execle, (const char *_path, const char *_arg, ...)      );
 00496	_PROTOTYPE( int execlp, (const char *_file, const char *arg, ...)       );
 00497	_PROTOTYPE( int execv, (const char *_path, char *const _argv[])         );
 00498	_PROTOTYPE( int execve, (const char *_path, char *const _argv[], 
 00499	                                                char *const _envp[])    );
 00500	_PROTOTYPE( int execvp, (const char *_file, char *const _argv[])        );
 00501	_PROTOTYPE( pid_t fork, (void)                                          );
 00502	_PROTOTYPE( long fpathconf, (int _fd, int _name)                        );
 00503	_PROTOTYPE( char *getcwd, (char *_buf, size_t _size)                    );
 00504	_PROTOTYPE( gid_t getegid, (void)                                       );
 00505	_PROTOTYPE( uid_t geteuid, (void)                                       );
 00506	_PROTOTYPE( gid_t getgid, (void)                                        );
 00507	_PROTOTYPE( int getgroups, (int _gidsetsize, gid_t _grouplist[])        );
 00508	_PROTOTYPE( char *getlogin, (void)                                      );
 00509	_PROTOTYPE( pid_t getpgrp, (void)                                       );
 00510	_PROTOTYPE( pid_t getpid, (void)                                        );
 00511	_PROTOTYPE( pid_t getppid, (void)                                       );
 00512	_PROTOTYPE( uid_t getuid, (void)                                        );
 00513	_PROTOTYPE( int isatty, (int _fd)                                       );
 00514	_PROTOTYPE( int link, (const char *_existing, const char *_new)         );
 00515	_PROTOTYPE( off_t lseek, (int _fd, off_t _offset, int _whence)          );
 00516	_PROTOTYPE( long pathconf, (const char *_path, int _name)               );
 00517	_PROTOTYPE( int pause, (void)                                           );
 00518	_PROTOTYPE( int pipe, (int _fildes[2])                                  );
 00519	_PROTOTYPE( ssize_t read, (int _fd, void *_buf, size_t _n)              );
 00520	_PROTOTYPE( int rmdir, (const char *_path)                              );
 00521	_PROTOTYPE( int setgid, (_mnx_Gid_t _gid)                               );
 00522	_PROTOTYPE( int setpgid, (pid_t _pid, pid_t _pgid)                      );
 00523	_PROTOTYPE( pid_t setsid, (void)                                        );
 00524	_PROTOTYPE( int setuid, (_mnx_Uid_t _uid)                               );
 00525	_PROTOTYPE( unsigned int sleep, (unsigned int _seconds)                 );
 00526	_PROTOTYPE( long sysconf, (int _name)                                   );
 00527	_PROTOTYPE( pid_t tcgetpgrp, (int _fd)                                  );
 00528	_PROTOTYPE( int tcsetpgrp, (int _fd, pid_t _pgrp_id)                    );
 00529	_PROTOTYPE( char *ttyname, (int _fd)                                    );
 00530	_PROTOTYPE( int unlink, (const char *_path)                             );
 00531	_PROTOTYPE( ssize_t write, (int _fd, const void *_buf, size_t _n)       );
 00532	
 00533	/* Open Group Base Specifications Issue 6 (not complete) */
 00534	_PROTOTYPE( int symlink, (const char *path1, const char *path2)         );
 00535	_PROTOTYPE( int getopt, (int _argc, char **_argv, char *_opts)          );
 00536	extern char *optarg;
 00537	extern int optind, opterr, optopt;
 00538	_PROTOTYPE( int usleep, (useconds_t _useconds)                          );
 00539	

_________________________ Page 646    File: include/unistd.h _________________________
 00540	#ifdef _MINIX
 00541	#ifndef _TYPE_H
 00542	#include <minix/type.h>
 00543	#endif
 00544	_PROTOTYPE( int brk, (char *_addr)                                      );
 00545	_PROTOTYPE( int chroot, (const char *_name)                             );
 00546	_PROTOTYPE( int mknod, (const char *_name, _mnx_Mode_t _mode, Dev_t _addr)      );
 00547	_PROTOTYPE( int mknod4, (const char *_name, _mnx_Mode_t _mode, Dev_t _addr,
 00548	            long _size)                                                 );
 00549	_PROTOTYPE( char *mktemp, (char *_template)                             );
 00550	_PROTOTYPE( int mount, (char *_spec, char *_name, int _flag)            );
 00551	_PROTOTYPE( long ptrace, (int _req, pid_t _pid, long _addr, long _data) );
 00552	_PROTOTYPE( char *sbrk, (int _incr)                                     );
 00553	_PROTOTYPE( int sync, (void)                                            );
 00554	_PROTOTYPE( int fsync, (int fd)                                         );
 00555	_PROTOTYPE( int umount, (const char *_name)                             );
 00556	_PROTOTYPE( int reboot, (int _how, ...)                                 );
 00557	_PROTOTYPE( int gethostname, (char *_hostname, size_t _len)             );
 00558	_PROTOTYPE( int getdomainname, (char *_domain, size_t _len)             );
 00559	_PROTOTYPE( int ttyslot, (void)                                         );
 00560	_PROTOTYPE( int fttyslot, (int _fd)                                     );
 00561	_PROTOTYPE( char *crypt, (const char *_key, const char *_salt)          );
 00562	_PROTOTYPE( int getsysinfo, (int who, int what, void *where)            );
 00563	_PROTOTYPE( int getprocnr, (void)                                       );
 00564	_PROTOTYPE( int findproc, (char *proc_name, int *proc_nr)               );
 00565	_PROTOTYPE( int allocmem, (phys_bytes size, phys_bytes *base)           );
 00566	_PROTOTYPE( int freemem, (phys_bytes size, phys_bytes base)             );
 00567	#define DEV_MAP 1
 00568	#define DEV_UNMAP 2
 00569	#define mapdriver(driver, device, style) devctl(DEV_MAP, driver, device, style)
 00570	#define unmapdriver(device) devctl(DEV_UNMAP, 0, device, 0)
 00571	_PROTOTYPE( int devctl, (int ctl_req, int driver, int device, int style));
 00572	
 00573	/* For compatibility with other Unix systems */
 00574	_PROTOTYPE( int getpagesize, (void)                                     );
 00575	_PROTOTYPE( int setgroups, (int ngroups, const gid_t *gidset)           );
 00576	
 00577	#endif
 00578	
 00579	_PROTOTYPE( int readlink, (const char *, char *, int));
 00580	_PROTOTYPE( int getopt, (int, char **, char *));
 00581	extern int optind, opterr, optopt;
 00582	
 00583	#endif /* _UNISTD_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/string.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00600	/* The <string.h> header contains prototypes for the string handling 
 00601	 * functions.
 00602	 */
 00603	
 00604	#ifndef _STRING_H
 00605	#define _STRING_H
 00606	
 00607	#define NULL    ((void *)0)
 00608	
 00609	#ifndef _SIZE_T

_________________________ Page 647    File: include/string.h _________________________
 00610	#define _SIZE_T
 00611	typedef unsigned int size_t;    /* type returned by sizeof */
 00612	#endif /*_SIZE_T */
 00613	
 00614	/* Function Prototypes. */
 00615	#ifndef _ANSI_H
 00616	#include <ansi.h>
 00617	#endif
 00618	
 00619	_PROTOTYPE( void *memchr, (const void *_s, int _c, size_t _n)           );
 00620	_PROTOTYPE( int memcmp, (const void *_s1, const void *_s2, size_t _n)   );
 00621	_PROTOTYPE( void *memcpy, (void *_s1, const void *_s2, size_t _n)       );
 00622	_PROTOTYPE( void *memmove, (void *_s1, const void *_s2, size_t _n)      );
 00623	_PROTOTYPE( void *memset, (void *_s, int _c, size_t _n)                 );
 00624	_PROTOTYPE( char *strcat, (char *_s1, const char *_s2)                  );
 00625	_PROTOTYPE( char *strchr, (const char *_s, int _c)                      );
 00626	_PROTOTYPE( int strncmp, (const char *_s1, const char *_s2, size_t _n)  );
 00627	_PROTOTYPE( int strcmp, (const char *_s1, const char *_s2)              );
 00628	_PROTOTYPE( int strcoll, (const char *_s1, const char *_s2)             );
 00629	_PROTOTYPE( char *strcpy, (char *_s1, const char *_s2)                  );
 00630	_PROTOTYPE( size_t strcspn, (const char *_s1, const char *_s2)          );
 00631	_PROTOTYPE( char *strerror, (int _errnum)                               );
 00632	_PROTOTYPE( size_t strlen, (const char *_s)                             );
 00633	_PROTOTYPE( char *strncat, (char *_s1, const char *_s2, size_t _n)      );
 00634	_PROTOTYPE( char *strncpy, (char *_s1, const char *_s2, size_t _n)      );
 00635	_PROTOTYPE( char *strpbrk, (const char *_s1, const char *_s2)           );
 00636	_PROTOTYPE( char *strrchr, (const char *_s, int _c)                     );
 00637	_PROTOTYPE( size_t strspn, (const char *_s1, const char *_s2)           );
 00638	_PROTOTYPE( char *strstr, (const char *_s1, const char *_s2)            );
 00639	_PROTOTYPE( char *strtok, (char *_s1, const char *_s2)                  );
 00640	_PROTOTYPE( size_t strxfrm, (char *_s1, const char *_s2, size_t _n)     );
 00641	
 00642	#ifdef _POSIX_SOURCE
 00643	/* Open Group Base Specifications Issue 6 (not complete) */
 00644	 char *strdup(const char *_s1);
 00645	#endif
 00646	
 00647	#ifdef _MINIX
 00648	/* For backward compatibility. */
 00649	_PROTOTYPE( char *index, (const char *_s, int _charwanted)              );
 00650	_PROTOTYPE( char *rindex, (const char *_s, int _charwanted)             );
 00651	_PROTOTYPE( void bcopy, (const void *_src, void *_dst, size_t _length)  );
 00652	_PROTOTYPE( int bcmp, (const void *_s1, const void *_s2, size_t _length));
 00653	_PROTOTYPE( void bzero, (void *_dst, size_t _length)                    );
 00654	_PROTOTYPE( void *memccpy, (char *_dst, const char *_src, int _ucharstop,
 00655	                                                    size_t _size)       );
 00656	
 00657	/* Misc. extra functions */
 00658	_PROTOTYPE( int strcasecmp, (const char *_s1, const char *_s2)          );
 00659	_PROTOTYPE( int strncasecmp, (const char *_s1, const char *_s2,
 00660	                                                        size_t _len)    );
 00661	_PROTOTYPE( size_t strnlen, (const char *_s, size_t _n)                 );
 00662	#endif
 00663	
 00664	#endif /* _STRING_H */

_________________________ Page 648    File: include/string.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/signal.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00700	/* The <signal.h> header defines all the ANSI and POSIX signals.
 00701	 * MINIX supports all the signals required by POSIX. They are defined below.
 00702	 * Some additional signals are also supported.
 00703	 */
 00704	
 00705	#ifndef _SIGNAL_H
 00706	#define _SIGNAL_H
 00707	
 00708	#ifndef _ANSI_H
 00709	#include <ansi.h>
 00710	#endif
 00711	#ifdef _POSIX_SOURCE
 00712	#ifndef _TYPES_H
 00713	#include <sys/types.h>
 00714	#endif
 00715	#endif
 00716	
 00717	/* Here are types that are closely associated with signal handling. */
 00718	typedef int sig_atomic_t;
 00719	
 00720	#ifdef _POSIX_SOURCE
 00721	#ifndef _SIGSET_T
 00722	#define _SIGSET_T
 00723	typedef unsigned long sigset_t;
 00724	#endif
 00725	#endif
 00726	
 00727	#define _NSIG             20    /* number of signals used */
 00728	
 00729	#define SIGHUP             1    /* hangup */
 00730	#define SIGINT             2    /* interrupt (DEL) */
 00731	#define SIGQUIT            3    /* quit (ASCII FS) */
 00732	#define SIGILL             4    /* illegal instruction */
 00733	#define SIGTRAP            5    /* trace trap (not reset when caught) */
 00734	#define SIGABRT            6    /* IOT instruction */
 00735	#define SIGIOT             6    /* SIGABRT for people who speak PDP-11 */
 00736	#define SIGUNUSED          7    /* spare code */
 00737	#define SIGFPE             8    /* floating point exception */
 00738	#define SIGKILL            9    /* kill (cannot be caught or ignored) */
 00739	#define SIGUSR1           10    /* user defined signal # 1 */
 00740	#define SIGSEGV           11    /* segmentation violation */
 00741	#define SIGUSR2           12    /* user defined signal # 2 */
 00742	#define SIGPIPE           13    /* write on a pipe with no one to read it */
 00743	#define SIGALRM           14    /* alarm clock */
 00744	#define SIGTERM           15    /* software termination signal from kill */
 00745	#define SIGCHLD           17    /* child process terminated or stopped */
 00746	
 00747	#define SIGEMT             7    /* obsolete */
 00748	#define SIGBUS            10    /* obsolete */
 00749	
 00750	/* MINIX specific signals. These signals are not used by user proceses, 
 00751	 * but meant to inform system processes, like the PM, about system events.
 00752	 */
 00753	#define SIGKMESS          18    /* new kernel message */
 00754	#define SIGKSIG           19    /* kernel signal pending */

_________________________ Page 649    File: include/signal.h _________________________
 00755	#define SIGKSTOP          20    /* kernel shutting down */
 00756	
 00757	/* POSIX requires the following signals to be defined, even if they are
 00758	 * not supported.  Here are the definitions, but they are not supported.
 00759	 */
 00760	#define SIGCONT           18    /* continue if stopped */
 00761	#define SIGSTOP           19    /* stop signal */
 00762	#define SIGTSTP           20    /* interactive stop signal */
 00763	#define SIGTTIN           21    /* background process wants to read */
 00764	#define SIGTTOU           22    /* background process wants to write */
 00765	
 00766	/* The sighandler_t type is not allowed unless _POSIX_SOURCE is defined. */
 00767	typedef void _PROTOTYPE( (*__sighandler_t), (int) );
 00768	
 00769	/* Macros used as function pointers. */
 00770	#define SIG_ERR    ((__sighandler_t) -1)        /* error return */
 00771	#define SIG_DFL    ((__sighandler_t)  0)        /* default signal handling */
 00772	#define SIG_IGN    ((__sighandler_t)  1)        /* ignore signal */
 00773	#define SIG_HOLD   ((__sighandler_t)  2)        /* block signal */
 00774	#define SIG_CATCH  ((__sighandler_t)  3)        /* catch signal */
 00775	#define SIG_MESS   ((__sighandler_t)  4)        /* pass as message (MINIX) */
 00776	
 00777	#ifdef _POSIX_SOURCE
 00778	struct sigaction {
 00779	  __sighandler_t sa_handler;    /* SIG_DFL, SIG_IGN, or pointer to function */
 00780	  sigset_t sa_mask;             /* signals to be blocked during handler */
 00781	  int sa_flags;                 /* special flags */
 00782	};
 00783	
 00784	/* Fields for sa_flags. */
 00785	#define SA_ONSTACK   0x0001     /* deliver signal on alternate stack */
 00786	#define SA_RESETHAND 0x0002     /* reset signal handler when signal caught */
 00787	#define SA_NODEFER   0x0004     /* don't block signal while catching it */
 00788	#define SA_RESTART   0x0008     /* automatic system call restart */
 00789	#define SA_SIGINFO   0x0010     /* extended signal handling */
 00790	#define SA_NOCLDWAIT 0x0020     /* don't create zombies */
 00791	#define SA_NOCLDSTOP 0x0040     /* don't receive SIGCHLD when child stops */
 00792	
 00793	/* POSIX requires these values for use with sigprocmask(2). */
 00794	#define SIG_BLOCK          0    /* for blocking signals */
 00795	#define SIG_UNBLOCK        1    /* for unblocking signals */
 00796	#define SIG_SETMASK        2    /* for setting the signal mask */
 00797	#define SIG_INQUIRE        4    /* for internal use only */
 00798	#endif  /* _POSIX_SOURCE */
 00799	
 00800	/* POSIX and ANSI function prototypes. */
 00801	_PROTOTYPE( int raise, (int _sig)                                       );
 00802	_PROTOTYPE( __sighandler_t signal, (int _sig, __sighandler_t _func)     );
 00803	
 00804	#ifdef _POSIX_SOURCE
 00805	_PROTOTYPE( int kill, (pid_t _pid, int _sig)                            );
 00806	_PROTOTYPE( int sigaction,
 00807	    (int _sig, const struct sigaction *_act, struct sigaction *_oact)   );
 00808	_PROTOTYPE( int sigaddset, (sigset_t *_set, int _sig)                   );
 00809	_PROTOTYPE( int sigdelset, (sigset_t *_set, int _sig)                   );
 00810	_PROTOTYPE( int sigemptyset, (sigset_t *_set)                           );
 00811	_PROTOTYPE( int sigfillset, (sigset_t *_set)                            );
 00812	_PROTOTYPE( int sigismember, (const sigset_t *_set, int _sig)           );
 00813	_PROTOTYPE( int sigpending, (sigset_t *_set)                            );
 00814	_PROTOTYPE( int sigprocmask,

_________________________ Page 650    File: include/signal.h _________________________
 00815	            (int _how, const sigset_t *_set, sigset_t *_oset)           );
 00816	_PROTOTYPE( int sigsuspend, (const sigset_t *_sigmask)                  );
 00817	#endif
 00818	
 00819	#endif /* _SIGNAL_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/fcntl.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 00900	/* The <fcntl.h> header is needed by the open() and fcntl() system calls,
 00901	 * which  have a variety of parameters and flags.  They are described here.  
 00902	 * The formats of the calls to each of these are:
 00903	 *
 00904	 *      open(path, oflag [,mode])       open a file
 00905	 *      fcntl(fd, cmd [,arg])           get or set file attributes
 00906	 * 
 00907	 */
 00908	
 00909	#ifndef _FCNTL_H
 00910	#define _FCNTL_H
 00911	
 00912	#ifndef _TYPES_H
 00913	#include <sys/types.h>
 00914	#endif
 00915	
 00916	/* These values are used for cmd in fcntl().  POSIX Table 6-1.  */
 00917	#define F_DUPFD            0    /* duplicate file descriptor */
 00918	#define F_GETFD            1    /* get file descriptor flags */
 00919	#define F_SETFD            2    /* set file descriptor flags */
 00920	#define F_GETFL            3    /* get file status flags */
 00921	#define F_SETFL            4    /* set file status flags */
 00922	#define F_GETLK            5    /* get record locking information */
 00923	#define F_SETLK            6    /* set record locking information */
 00924	#define F_SETLKW           7    /* set record locking info; wait if blocked */
 00925	
 00926	/* File descriptor flags used for fcntl().  POSIX Table 6-2. */
 00927	#define FD_CLOEXEC         1    /* close on exec flag for third arg of fcntl */
 00928	
 00929	/* L_type values for record locking with fcntl().  POSIX Table 6-3. */
 00930	#define F_RDLCK            1    /* shared or read lock */
 00931	#define F_WRLCK            2    /* exclusive or write lock */
 00932	#define F_UNLCK            3    /* unlock */
 00933	
 00934	/* Oflag values for open().  POSIX Table 6-4. */
 00935	#define O_CREAT        00100    /* creat file if it doesn't exist */
 00936	#define O_EXCL         00200    /* exclusive use flag */
 00937	#define O_NOCTTY       00400    /* do not assign a controlling terminal */
 00938	#define O_TRUNC        01000    /* truncate flag */
 00939	
 00940	/* File status flags for open() and fcntl().  POSIX Table 6-5. */
 00941	#define O_APPEND       02000    /* set append mode */
 00942	#define O_NONBLOCK     04000    /* no delay */
 00943	
 00944	/* File access modes for open() and fcntl().  POSIX Table 6-6. */
 00945	#define O_RDONLY           0    /* open(name, O_RDONLY) opens read only */
 00946	#define O_WRONLY           1    /* open(name, O_WRONLY) opens write only */
 00947	#define O_RDWR             2    /* open(name, O_RDWR) opens read/write */
 00948	
 00949	/* Mask for use with file access modes.  POSIX Table 6-7. */

_________________________ Page 651    File: include/fcntl.h _________________________
 00950	#define O_ACCMODE         03    /* mask for file access modes */
 00951	
 00952	/* Struct used for locking.  POSIX Table 6-8. */
 00953	struct flock {
 00954	  short l_type;                 /* type: F_RDLCK, F_WRLCK, or F_UNLCK */
 00955	  short l_whence;               /* flag for starting offset */
 00956	  off_t l_start;                /* relative offset in bytes */
 00957	  off_t l_len;                  /* size; if 0, then until EOF */
 00958	  pid_t l_pid;                  /* process id of the locks' owner */
 00959	};
 00960	
 00961	/* Function Prototypes. */
 00962	_PROTOTYPE( int creat, (const char *_path, _mnx_Mode_t _mode)           );
 00963	_PROTOTYPE( int fcntl, (int _filedes, int _cmd, ...)                    );
 00964	_PROTOTYPE( int open,  (const char *_path, int _oflag, ...)             );
 00965	
 00966	#endif /* _FCNTL_H */




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/termios.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01000	/* The <termios.h> header is used for controlling tty modes. */
 01001	
 01002	#ifndef _TERMIOS_H
 01003	#define _TERMIOS_H
 01004	
 01005	typedef unsigned short tcflag_t;
 01006	typedef unsigned char cc_t;
 01007	typedef unsigned int speed_t;
 01008	
 01009	#define NCCS               20   /* size of cc_c array, some extra space
 01010	                                 * for extensions. */
 01011	
 01012	/* Primary terminal control structure. POSIX Table 7-1. */
 01013	struct termios {
 01014	  tcflag_t c_iflag;             /* input modes */
 01015	  tcflag_t c_oflag;             /* output modes */
 01016	  tcflag_t c_cflag;             /* control modes */
 01017	  tcflag_t c_lflag;             /* local modes */
 01018	  speed_t  c_ispeed;            /* input speed */
 01019	  speed_t  c_ospeed;            /* output speed */
 01020	  cc_t c_cc[NCCS];              /* control characters */
 01021	};
 01022	
 01023	/* Values for termios c_iflag bit map.  POSIX Table 7-2. */
 01024	#define BRKINT          0x0001  /* signal interrupt on break */
 01025	#define ICRNL           0x0002  /* map CR to NL on input */
 01026	#define IGNBRK          0x0004  /* ignore break */
 01027	#define IGNCR           0x0008  /* ignore CR */
 01028	#define IGNPAR          0x0010  /* ignore characters with parity errors */
 01029	#define INLCR           0x0020  /* map NL to CR on input */
 01030	#define INPCK           0x0040  /* enable input parity check */
 01031	#define ISTRIP          0x0080  /* mask off 8th bit */
 01032	#define IXOFF           0x0100  /* enable start/stop input control */
 01033	#define IXON            0x0200  /* enable start/stop output control */
 01034	#define PARMRK          0x0400  /* mark parity errors in the input queue */

_________________________ Page 652    File: include/termios.h _________________________
 01035	
 01036	/* Values for termios c_oflag bit map.  POSIX Sec. 7.1.2.3. */
 01037	#define OPOST           0x0001  /* perform output processing */
 01038	
 01039	/* Values for termios c_cflag bit map.  POSIX Table 7-3. */
 01040	#define CLOCAL          0x0001  /* ignore modem status lines */
 01041	#define CREAD           0x0002  /* enable receiver */
 01042	#define CSIZE           0x000C  /* number of bits per character */
 01043	#define         CS5     0x0000  /* if CSIZE is CS5, characters are 5 bits */
 01044	#define         CS6     0x0004  /* if CSIZE is CS6, characters are 6 bits */
 01045	#define         CS7     0x0008  /* if CSIZE is CS7, characters are 7 bits */
 01046	#define         CS8     0x000C  /* if CSIZE is CS8, characters are 8 bits */
 01047	#define CSTOPB          0x0010  /* send 2 stop bits if set, else 1 */
 01048	#define HUPCL           0x0020  /* hang up on last close */
 01049	#define PARENB          0x0040  /* enable parity on output */
 01050	#define PARODD          0x0080  /* use odd parity if set, else even */
 01051	
 01052	/* Values for termios c_lflag bit map.  POSIX Table 7-4. */
 01053	#define ECHO            0x0001  /* enable echoing of input characters */
 01054	#define ECHOE           0x0002  /* echo ERASE as backspace */
 01055	#define ECHOK           0x0004  /* echo KILL */
 01056	#define ECHONL          0x0008  /* echo NL */
 01057	#define ICANON          0x0010  /* canonical input (erase and kill enabled) */
 01058	#define IEXTEN          0x0020  /* enable extended functions */
 01059	#define ISIG            0x0040  /* enable signals */
 01060	#define NOFLSH          0x0080  /* disable flush after interrupt or quit */
 01061	#define TOSTOP          0x0100  /* send SIGTTOU (job control, not implemented*/
 01062	
 01063	/* Indices into c_cc array.  Default values in parentheses. POSIX Table 7-5. */
 01064	#define VEOF                 0  /* cc_c[VEOF] = EOF char (^D) */
 01065	#define VEOL                 1  /* cc_c[VEOL] = EOL char (undef) */
 01066	#define VERASE               2  /* cc_c[VERASE] = ERASE char (^H) */
 01067	#define VINTR                3  /* cc_c[VINTR] = INTR char (DEL) */
 01068	#define VKILL                4  /* cc_c[VKILL] = KILL char (^U) */
 01069	#define VMIN                 5  /* cc_c[VMIN] = MIN value for timer */
 01070	#define VQUIT                6  /* cc_c[VQUIT] = QUIT char (^\) */
 01071	#define VTIME                7  /* cc_c[VTIME] = TIME value for timer */
 01072	#define VSUSP                8  /* cc_c[VSUSP] = SUSP (^Z, ignored) */
 01073	#define VSTART               9  /* cc_c[VSTART] = START char (^S) */
 01074	#define VSTOP               10  /* cc_c[VSTOP] = STOP char (^Q) */
 01075	
 01076	#define _POSIX_VDISABLE   (cc_t)0xFF    /* You can't even generate this 
 01077	                                         * character with 'normal' keyboards.
 01078	                                         * But some language specific keyboards
 01079	                                         * can generate 0xFF. It seems that all
 01080	                                         * 256 are used, so cc_t should be a
 01081	                                         * short...
 01082	                                         */
 01083	
 01084	/* Values for the baud rate settings.  POSIX Table 7-6. */
 01085	#define B0              0x0000  /* hang up the line */
 01086	#define B50             0x1000  /* 50 baud */
 01087	#define B75             0x2000  /* 75 baud */
 01088	#define B110            0x3000  /* 110 baud */
 01089	#define B134            0x4000  /* 134.5 baud */
 01090	#define B150            0x5000  /* 150 baud */
 01091	#define B200            0x6000  /* 200 baud */
 01092	#define B300            0x7000  /* 300 baud */
 01093	#define B600            0x8000  /* 600 baud */
 01094	#define B1200           0x9000  /* 1200 baud */

_________________________ Page 653    File: include/termios.h _________________________
 01095	#define B1800           0xA000  /* 1800 baud */
 01096	#define B2400           0xB000  /* 2400 baud */
 01097	#define B4800           0xC000  /* 4800 baud */
 01098	#define B9600           0xD000  /* 9600 baud */
 01099	#define B19200          0xE000  /* 19200 baud */
 01100	#define B38400          0xF000  /* 38400 baud */
 01101	
 01102	/* Optional actions for tcsetattr().  POSIX Sec. 7.2.1.2. */
 01103	#define TCSANOW            1    /* changes take effect immediately */
 01104	#define TCSADRAIN          2    /* changes take effect after output is done */
 01105	#define TCSAFLUSH          3    /* wait for output to finish and flush input */
 01106	
 01107	/* Queue_selector values for tcflush().  POSIX Sec. 7.2.2.2. */
 01108	#define TCIFLUSH           1    /* flush accumulated input data */
 01109	#define TCOFLUSH           2    /* flush accumulated output data */
 01110	#define TCIOFLUSH          3    /* flush accumulated input and output data */
 01111	
 01112	/* Action values for tcflow().  POSIX Sec. 7.2.2.2. */
 01113	#define TCOOFF             1    /* suspend output */
 01114	#define TCOON              2    /* restart suspended output */
 01115	#define TCIOFF             3    /* transmit a STOP character on the line */
 01116	#define TCION              4    /* transmit a START character on the line */
 01117	
 01118	/* Function Prototypes. */
 01119	#ifndef _ANSI_H
 01120	#include <ansi.h>
 01121	#endif
 01122	
 01123	_PROTOTYPE( int tcsendbreak, (int _fildes, int _duration)                    );
 01124	_PROTOTYPE( int tcdrain, (int _filedes)                                      );
 01125	_PROTOTYPE( int tcflush, (int _filedes, int _queue_selector)                 );
 01126	_PROTOTYPE( int tcflow, (int _filedes, int _action)                          );
 01127	_PROTOTYPE( speed_t cfgetispeed, (const struct termios *_termios_p)          );
 01128	_PROTOTYPE( speed_t cfgetospeed, (const struct termios *_termios_p)          );
 01129	_PROTOTYPE( int cfsetispeed, (struct termios *_termios_p, speed_t _speed)    );
 01130	_PROTOTYPE( int cfsetospeed, (struct termios *_termios_p, speed_t _speed)    );
 01131	_PROTOTYPE( int tcgetattr, (int _filedes, struct termios *_termios_p)        );
 01132	_PROTOTYPE( int tcsetattr, \
 01133	        (int _filedes, int _opt_actions, const struct termios *_termios_p)   );
 01134	
 01135	#define cfgetispeed(termios_p)          ((termios_p)->c_ispeed)
 01136	#define cfgetospeed(termios_p)          ((termios_p)->c_ospeed)
 01137	#define cfsetispeed(termios_p, speed)   ((termios_p)->c_ispeed = (speed), 0)
 01138	#define cfsetospeed(termios_p, speed)   ((termios_p)->c_ospeed = (speed), 0)
 01139	
 01140	#ifdef _MINIX
 01141	/* Here are the local extensions to the POSIX standard for Minix. Posix
 01142	 * conforming programs are not able to access these, and therefore they are
 01143	 * only defined when a Minix program is compiled.
 01144	 */
 01145	
 01146	/* Extensions to the termios c_iflag bit map.  */
 01147	#define IXANY           0x0800  /* allow any key to continue ouptut */
 01148	
 01149	/* Extensions to the termios c_oflag bit map. They are only active iff
 01150	 * OPOST is enabled. */
 01151	#define ONLCR           0x0002  /* Map NL to CR-NL on output */
 01152	#define XTABS           0x0004  /* Expand tabs to spaces */
 01153	#define ONOEOT          0x0008  /* discard EOT's (^D) on output) */
 01154	

_________________________ Page 654    File: include/termios.h _________________________
 01155	/* Extensions to the termios c_lflag bit map.  */
 01156	#define LFLUSHO         0x0200  /* Flush output. */
 01157	
 01158	/* Extensions to the c_cc array. */
 01159	#define VREPRINT          11    /* cc_c[VREPRINT] (^R) */
 01160	#define VLNEXT            12    /* cc_c[VLNEXT] (^V) */
 01161	#define VDISCARD          13    /* cc_c[VDISCARD] (^O) */
 01162	
 01163	/* Extensions to baud rate settings. */
 01164	#define B57600          0x0100  /* 57600 baud */
 01165	#define B115200         0x0200  /* 115200 baud */
 01166	
 01167	/* These are the default settings used by the kernel and by 'stty sane' */
 01168	
 01169	#define TCTRL_DEF       (CREAD | CS8 | HUPCL)
 01170	#define TINPUT_DEF      (BRKINT | ICRNL | IXON | IXANY)
 01171	#define TOUTPUT_DEF     (OPOST | ONLCR)
 01172	#define TLOCAL_DEF      (ISIG | IEXTEN | ICANON | ECHO | ECHOE)
 01173	#define TSPEED_DEF      B9600
 01174	
 01175	#define TEOF_DEF        '\4'    /* ^D */
 01176	#define TEOL_DEF        _POSIX_VDISABLE
 01177	#define TERASE_DEF      '\10'   /* ^H */
 01178	#define TINTR_DEF       '\3'    /* ^C */
 01179	#define TKILL_DEF       '\25'   /* ^U */
 01180	#define TMIN_DEF        1
 01181	#define TQUIT_DEF       '\34'   /* ^\ */
 01182	#define TSTART_DEF      '\21'   /* ^Q */
 01183	#define TSTOP_DEF       '\23'   /* ^S */
 01184	#define TSUSP_DEF       '\32'   /* ^Z */
 01185	#define TTIME_DEF       0
 01186	#define TREPRINT_DEF    '\22'   /* ^R */
 01187	#define TLNEXT_DEF      '\26'   /* ^V */
 01188	#define TDISCARD_DEF    '\17'   /* ^O */
 01189	
 01190	/* Window size. This information is stored in the TTY driver but not used.
 01191	 * This can be used for screen based applications in a window environment. 
 01192	 * The ioctls TIOCGWINSZ and TIOCSWINSZ can be used to get and set this 
 01193	 * information.
 01194	 */
 01195	
 01196	struct winsize
 01197	{
 01198	        unsigned short  ws_row;         /* rows, in characters */
 01199	        unsigned short  ws_col;         /* columns, in characters */
 01200	        unsigned short  ws_xpixel;      /* horizontal size, pixels */
 01201	        unsigned short  ws_ypixel;      /* vertical size, pixels */
 01202	};
 01203	#endif /* _MINIX */
 01204	
 01205	#endif /* _TERMIOS_H */

_________________________ Page 655    File: include/termios.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/timers.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01300	/* This library provides generic watchdog timer management functionality.
 01301	 * The functions operate on a timer queue provided by the caller. Note that
 01302	 * the timers must use absolute time to allow sorting. The library provides:
 01303	 *
 01304	 *    tmrs_settimer:     (re)set a new watchdog timer in the timers queue 
 01305	 *    tmrs_clrtimer:     remove a timer from both the timers queue 
 01306	 *    tmrs_exptimers:    check for expired timers and run watchdog functions
 01307	 *
 01308	 * Author:
 01309	 *    Jorrit N. Herder <jnherder@cs.vu.nl>
 01310	 *    Adapted from tmr_settimer and tmr_clrtimer in src/kernel/clock.c. 
 01311	 *    Last modified: September 30, 2004.
 01312	 */
 01313	
 01314	#ifndef _TIMERS_H
 01315	#define _TIMERS_H
 01316	
 01317	#include <limits.h>
 01318	#include <sys/types.h>
 01319	
 01320	struct timer;
 01321	typedef void (*tmr_func_t)(struct timer *tp);
 01322	typedef union { int ta_int; long ta_long; void *ta_ptr; } tmr_arg_t;
 01323	
 01324	/* A timer_t variable must be declare for each distinct timer to be used.
 01325	 * The timers watchdog function and expiration time are automatically set
 01326	 * by the library function tmrs_settimer, but its argument is not.
 01327	 */
 01328	typedef struct timer
 01329	{
 01330	  struct timer  *tmr_next;      /* next in a timer chain */
 01331	  clock_t       tmr_exp_time;   /* expiration time */
 01332	  tmr_func_t    tmr_func;       /* function to call when expired */
 01333	  tmr_arg_t     tmr_arg;        /* random argument */
 01334	} timer_t;
 01335	
 01336	/* Used when the timer is not active. */
 01337	#define TMR_NEVER    ((clock_t) -1 < 0) ? ((clock_t) LONG_MAX) : ((clock_t) -1)
 01338	#undef TMR_NEVER
 01339	#define TMR_NEVER       ((clock_t) LONG_MAX)
 01340	
 01341	/* These definitions can be used to set or get data from a timer variable. */ 
 01342	#define tmr_arg(tp) (&(tp)->tmr_arg)
 01343	#define tmr_exp_time(tp) (&(tp)->tmr_exp_time)
 01344	
 01345	/* Timers should be initialized once before they are being used. Be careful
 01346	 * not to reinitialize a timer that is in a list of timers, or the chain
 01347	 * will be broken.
 01348	 */
 01349	#define tmr_inittimer(tp) (void)((tp)->tmr_exp_time = TMR_NEVER, \
 01350	        (tp)->tmr_next = NULL)
 01351	
 01352	/* The following generic timer management functions are available. They
 01353	 * can be used to operate on the lists of timers. Adding a timer to a list 
 01354	 * automatically takes care of removing it.

_________________________ Page 656    File: include/timers.h _________________________
 01355	 */
 01356	_PROTOTYPE( clock_t tmrs_clrtimer, (timer_t **tmrs, timer_t *tp, clock_t *new_head)
 01357	_PROTOTYPE( void tmrs_exptimers, (timer_t **tmrs, clock_t now, clock_t *new_head)  
 01358	_PROTOTYPE( clock_t tmrs_settimer, (timer_t **tmrs, timer_t *tp, 
 01359	        clock_t exp_time, tmr_func_t watchdog, clock_t *new_head)                  
 01360	
 01361	#endif /* _TIMERS_H */
 01362	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/types.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01400	/* The <sys/types.h> header contains important data type definitions.
 01401	 * It is considered good programming practice to use these definitions, 
 01402	 * instead of the underlying base type.  By convention, all type names end 
 01403	 * with _t.
 01404	 */
 01405	
 01406	#ifndef _TYPES_H
 01407	#define _TYPES_H
 01408	
 01409	#ifndef _ANSI_H
 01410	#include <ansi.h>
 01411	#endif
 01412	
 01413	/* The type size_t holds all results of the sizeof operator.  At first glance,
 01414	 * it seems obvious that it should be an unsigned int, but this is not always 
 01415	 * the case. For example, MINIX-ST (68000) has 32-bit pointers and 16-bit
 01416	 * integers. When one asks for the size of a 70K struct or array, the result 
 01417	 * requires 17 bits to express, so size_t must be a long type.  The type 
 01418	 * ssize_t is the signed version of size_t.
 01419	 */
 01420	#ifndef _SIZE_T
 01421	#define _SIZE_T
 01422	typedef unsigned int size_t;
 01423	#endif
 01424	
 01425	#ifndef _SSIZE_T
 01426	#define _SSIZE_T
 01427	typedef int ssize_t;
 01428	#endif
 01429	
 01430	#ifndef _TIME_T
 01431	#define _TIME_T
 01432	typedef long time_t;               /* time in sec since 1 Jan 1970 0000 GMT */
 01433	#endif
 01434	
 01435	#ifndef _CLOCK_T
 01436	#define _CLOCK_T
 01437	typedef long clock_t;              /* unit for system accounting */
 01438	#endif
 01439	
 01440	#ifndef _SIGSET_T
 01441	#define _SIGSET_T
 01442	typedef unsigned long sigset_t;
 01443	#endif
 01444	

_________________________ Page 657    File: include/sys/types.h _________________________
 01445	/* Open Group Base Specifications Issue 6 (not complete) */
 01446	typedef long useconds_t;        /* Time in microseconds */
 01447	
 01448	/* Types used in disk, inode, etc. data structures. */
 01449	typedef short          dev_t;      /* holds (major|minor) device pair */
 01450	typedef char           gid_t;      /* group id */
 01451	typedef unsigned long  ino_t;      /* i-node number (V3 filesystem) */
 01452	typedef unsigned short mode_t;     /* file type and permissions bits */
 01453	typedef short        nlink_t;      /* number of links to a file */
 01454	typedef unsigned long  off_t;      /* offset within a file */
 01455	typedef int            pid_t;      /* process id (must be signed) */
 01456	typedef short          uid_t;      /* user id */
 01457	typedef unsigned long zone_t;      /* zone number */
 01458	typedef unsigned long block_t;     /* block number */
 01459	typedef unsigned long  bit_t;      /* bit number in a bit map */
 01460	typedef unsigned short zone1_t;    /* zone number for V1 file systems */
 01461	typedef unsigned short bitchunk_t; /* collection of bits in a bitmap */
 01462	
 01463	typedef unsigned char   u8_t;      /* 8 bit type */
 01464	typedef unsigned short u16_t;      /* 16 bit type */
 01465	typedef unsigned long  u32_t;      /* 32 bit type */
 01466	
 01467	typedef char            i8_t;      /* 8 bit signed type */
 01468	typedef short          i16_t;      /* 16 bit signed type */
 01469	typedef long           i32_t;      /* 32 bit signed type */
 01470	
 01471	typedef struct { u32_t _[2]; } u64_t;
 01472	
 01473	/* The following types are needed because MINIX uses K&R style function
 01474	 * definitions (for maximum portability).  When a short, such as dev_t, is
 01475	 * passed to a function with a K&R definition, the compiler automatically
 01476	 * promotes it to an int.  The prototype must contain an int as the parameter,
 01477	 * not a short, because an int is what an old-style function definition
 01478	 * expects.  Thus using dev_t in a prototype would be incorrect.  It would be
 01479	 * sufficient to just use int instead of dev_t in the prototypes, but Dev_t
 01480	 * is clearer.
 01481	 */
 01482	typedef int            Dev_t;
 01483	typedef int       _mnx_Gid_t;
 01484	typedef int          Nlink_t;
 01485	typedef int       _mnx_Uid_t;
 01486	typedef int             U8_t;
 01487	typedef unsigned long  U32_t;
 01488	typedef int             I8_t;
 01489	typedef int            I16_t;
 01490	typedef long           I32_t;
 01491	
 01492	/* ANSI C makes writing down the promotion of unsigned types very messy.  When
 01493	 * sizeof(short) == sizeof(int), there is no promotion, so the type stays
 01494	 * unsigned.  When the compiler is not ANSI, there is usually no loss of
 01495	 * unsignedness, and there are usually no prototypes so the promoted type
 01496	 * doesn't matter.  The use of types like Ino_t is an attempt to use ints
 01497	 * (which are not promoted) while providing information to the reader.
 01498	 */
 01499	
 01500	typedef unsigned long  Ino_t;
 01501	
 01502	#if _EM_WSIZE == 2
 01503	/*typedef unsigned int      Ino_t; Ino_t is now 32 bits */
 01504	typedef unsigned int    Zone1_t;

_________________________ Page 658    File: include/sys/types.h _________________________
 01505	typedef unsigned int Bitchunk_t;
 01506	typedef unsigned int      U16_t;
 01507	typedef unsigned int  _mnx_Mode_t;
 01508	
 01509	#else /* _EM_WSIZE == 4, or _EM_WSIZE undefined */
 01510	/*typedef int             Ino_t; Ino_t is now 32 bits */
 01511	typedef int             Zone1_t;
 01512	typedef int          Bitchunk_t;
 01513	typedef int               U16_t;
 01514	typedef int         _mnx_Mode_t;
 01515	
 01516	#endif /* _EM_WSIZE == 2, etc */
 01517	 
 01518	/* Signal handler type, e.g. SIG_IGN */
 01519	typedef void _PROTOTYPE( (*sighandler_t), (int) );
 01520	
 01521	/* Compatibility with other systems */
 01522	typedef unsigned char   u_char;
 01523	typedef unsigned short  u_short;
 01524	typedef unsigned int    u_int;
 01525	typedef unsigned long   u_long;
 01526	typedef char            *caddr_t;
 01527	
 01528	#endif /* _TYPES_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/sigcontext.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01600	#ifndef _SIGCONTEXT_H
 01601	#define _SIGCONTEXT_H
 01602	
 01603	/* The sigcontext structure is used by the sigreturn(2) system call.
 01604	 * sigreturn() is seldom called by user programs, but it is used internally
 01605	 * by the signal catching mechanism.
 01606	 */
 01607	
 01608	#ifndef _ANSI_H
 01609	#include <ansi.h>
 01610	#endif
 01611	
 01612	#ifndef _MINIX_SYS_CONFIG_H
 01613	#include <minix/sys_config.h>
 01614	#endif
 01615	
 01616	#if !defined(_MINIX_CHIP)
 01617	#include "error, configuration is not known"
 01618	#endif
 01619	
 01620	/* The following structure should match the stackframe_s structure used
 01621	 * by the kernel's context switching code.  Floating point registers should
 01622	 * be added in a different struct.
 01623	 */
 01624	struct sigregs {  
 01625	  short sr_gs;
 01626	  short sr_fs;
 01627	  short sr_es;
 01628	  short sr_ds;
 01629	  int sr_di;

_________________________ Page 659    File: include/sys/sigcontext.h _____________________
 01630	  int sr_si;
 01631	  int sr_bp;
 01632	  int sr_st;                    /* stack top -- used in kernel */
 01633	  int sr_bx;
 01634	  int sr_dx;
 01635	  int sr_cx;
 01636	  int sr_retreg;
 01637	  int sr_retadr;                /* return address to caller of save -- used
 01638	                                 * in kernel */
 01639	  int sr_pc;
 01640	  int sr_cs;
 01641	  int sr_psw;
 01642	  int sr_sp;
 01643	  int sr_ss;
 01644	};
 01645	
 01646	struct sigframe {               /* stack frame created for signalled process */
 01647	  _PROTOTYPE( void (*sf_retadr), (void) );
 01648	  int sf_signo;
 01649	  int sf_code;
 01650	  struct sigcontext *sf_scp;
 01651	  int sf_fp;
 01652	  _PROTOTYPE( void (*sf_retadr2), (void) );
 01653	  struct sigcontext *sf_scpcopy;
 01654	};
 01655	
 01656	struct sigcontext {
 01657	  int sc_flags;                 /* sigstack state to restore */
 01658	  long sc_mask;                 /* signal mask to restore */
 01659	  struct sigregs sc_regs;       /* register set to restore */
 01660	};
 01661	
 01662	#define sc_gs sc_regs.sr_gs
 01663	#define sc_fs sc_regs.sr_fs
 01664	#define sc_es sc_regs.sr_es
 01665	#define sc_ds sc_regs.sr_ds
 01666	#define sc_di sc_regs.sr_di
 01667	#define sc_si sc_regs.sr_si 
 01668	#define sc_fp sc_regs.sr_bp
 01669	#define sc_st sc_regs.sr_st             /* stack top -- used in kernel */
 01670	#define sc_bx sc_regs.sr_bx
 01671	#define sc_dx sc_regs.sr_dx
 01672	#define sc_cx sc_regs.sr_cx
 01673	#define sc_retreg sc_regs.sr_retreg
 01674	#define sc_retadr sc_regs.sr_retadr     /* return address to caller of 
 01675	                                        save -- used in kernel */
 01676	#define sc_pc sc_regs.sr_pc
 01677	#define sc_cs sc_regs.sr_cs
 01678	#define sc_psw sc_regs.sr_psw
 01679	#define sc_sp sc_regs.sr_sp
 01680	#define sc_ss sc_regs.sr_ss
 01681	
 01682	/* Values for sc_flags.  Must agree with <minix/jmp_buf.h>. */
 01683	#define SC_SIGCONTEXT   2       /* nonzero when signal context is included */
 01684	#define SC_NOREGLOCALS  4       /* nonzero when registers are not to be
 01685	                                        saved and restored */
 01686	
 01687	_PROTOTYPE( int sigreturn, (struct sigcontext *_scp)                    );
 01688	
 01689	#endif /* _SIGCONTEXT_H */

_________________________ Page 660    File: include/sys/sigcontext.h _____________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/stat.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01700	/* The <sys/stat.h> header defines a struct that is used in the stat() and
 01701	 * fstat functions.  The information in this struct comes from the i-node of
 01702	 * some file.  These calls are the only approved way to inspect i-nodes.
 01703	 */
 01704	
 01705	#ifndef _STAT_H
 01706	#define _STAT_H
 01707	
 01708	#ifndef _TYPES_H
 01709	#include <sys/types.h>
 01710	#endif
 01711	
 01712	struct stat {
 01713	  dev_t st_dev;                 /* major/minor device number */
 01714	  ino_t st_ino;                 /* i-node number */
 01715	  mode_t st_mode;               /* file mode, protection bits, etc. */
 01716	  short int st_nlink;           /* # links; TEMPORARY HACK: should be nlink_t*/
 01717	  uid_t st_uid;                 /* uid of the file's owner */
 01718	  short int st_gid;             /* gid; TEMPORARY HACK: should be gid_t */
 01719	  dev_t st_rdev;
 01720	  off_t st_size;                /* file size */
 01721	  time_t st_atime;              /* time of last access */
 01722	  time_t st_mtime;              /* time of last data modification */
 01723	  time_t st_ctime;              /* time of last file status change */
 01724	};
 01725	
 01726	/* Traditional mask definitions for st_mode. */
 01727	/* The ugly casts on only some of the definitions are to avoid suprising sign
 01728	 * extensions such as S_IFREG != (mode_t) S_IFREG when ints are 32 bits.
 01729	 */
 01730	#define S_IFMT  ((mode_t) 0170000)      /* type of file */
 01731	#define S_IFLNK ((mode_t) 0120000)      /* symbolic link, not implemented */
 01732	#define S_IFREG ((mode_t) 0100000)      /* regular */
 01733	#define S_IFBLK 0060000         /* block special */
 01734	#define S_IFDIR 0040000         /* directory */
 01735	#define S_IFCHR 0020000         /* character special */
 01736	#define S_IFIFO 0010000         /* this is a FIFO */
 01737	#define S_ISUID 0004000         /* set user id on execution */
 01738	#define S_ISGID 0002000         /* set group id on execution */
 01739	                                /* next is reserved for future use */
 01740	#define S_ISVTX   01000         /* save swapped text even after use */
 01741	
 01742	/* POSIX masks for st_mode. */
 01743	#define S_IRWXU   00700         /* owner:  rwx------ */
 01744	#define S_IRUSR   00400         /* owner:  r-------- */
 01745	#define S_IWUSR   00200         /* owner:  -w------- */
 01746	#define S_IXUSR   00100         /* owner:  --x------ */
 01747	
 01748	#define S_IRWXG   00070         /* group:  ---rwx--- */
 01749	#define S_IRGRP   00040         /* group:  ---r----- */
 01750	#define S_IWGRP   00020         /* group:  ----w---- */
 01751	#define S_IXGRP   00010         /* group:  -----x--- */
 01752	
 01753	#define S_IRWXO   00007         /* others: ------rwx */
 01754	#define S_IROTH   00004         /* others: ------r-- */ 

_________________________ Page 661    File: include/sys/stat.h _________________________
 01755	#define S_IWOTH   00002         /* others: -------w- */
 01756	#define S_IXOTH   00001         /* others: --------x */
 01757	
 01758	/* The following macros test st_mode (from POSIX Sec. 5.6.1.1). */
 01759	#define S_ISREG(m)      (((m) & S_IFMT) == S_IFREG)     /* is a reg file */
 01760	#define S_ISDIR(m)      (((m) & S_IFMT) == S_IFDIR)     /* is a directory */
 01761	#define S_ISCHR(m)      (((m) & S_IFMT) == S_IFCHR)     /* is a char spec */
 01762	#define S_ISBLK(m)      (((m) & S_IFMT) == S_IFBLK)     /* is a block spec */
 01763	#define S_ISFIFO(m)     (((m) & S_IFMT) == S_IFIFO)     /* is a pipe/FIFO */
 01764	#define S_ISLNK(m)      (((m) & S_IFMT) == S_IFLNK)     /* is a sym link */
 01765	
 01766	/* Function Prototypes. */
 01767	_PROTOTYPE( int chmod, (const char *_path, _mnx_Mode_t _mode)           );
 01768	_PROTOTYPE( int fstat, (int _fildes, struct stat *_buf)                 );
 01769	_PROTOTYPE( int mkdir, (const char *_path, _mnx_Mode_t _mode)           );
 01770	_PROTOTYPE( int mkfifo, (const char *_path, _mnx_Mode_t _mode)          );
 01771	_PROTOTYPE( int stat, (const char *_path, struct stat *_buf)            );
 01772	_PROTOTYPE( mode_t umask, (_mnx_Mode_t _cmask)                          );
 01773	
 01774	/* Open Group Base Specifications Issue 6 (not complete) */
 01775	_PROTOTYPE( int lstat, (const char *_path, struct stat *_buf)           );
 01776	
 01777	#endif /* _STAT_H */



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/dir.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01800	/* The <dir.h> header gives the layout of a directory. */
 01801	
 01802	#ifndef _DIR_H
 01803	#define _DIR_H
 01804	
 01805	#include <sys/types.h>
 01806	
 01807	#define DIRBLKSIZ       512     /* size of directory block */
 01808	
 01809	#ifndef DIRSIZ
 01810	#define DIRSIZ  60
 01811	#endif
 01812	
 01813	struct direct {
 01814	  ino_t d_ino;
 01815	  char d_name[DIRSIZ];
 01816	};
 01817	
 01818	#endif /* _DIR_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/wait.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 01900	/* The <sys/wait.h> header contains macros related to wait(). The value
 01901	 * returned by wait() and waitpid() depends on whether the process 
 01902	 * terminated by an exit() call, was killed by a signal, or was stopped
 01903	 * due to job control, as follows:
 01904	 *

_________________________ Page 662    File: include/sys/wait.h _________________________
 01905	 *                               High byte   Low byte
 01906	 *                              +---------------------+
 01907	 *      exit(status)            |  status  |    0     |
 01908	 *                              +---------------------+
 01909	 *      killed by signal        |    0     |  signal  |
 01910	 *                              +---------------------+
 01911	 *      stopped (job control)   |  signal  |   0177   |
 01912	 *                              +---------------------+
 01913	 */
 01914	
 01915	#ifndef _WAIT_H
 01916	#define _WAIT_H
 01917	
 01918	#ifndef _TYPES_H
 01919	#include <sys/types.h>
 01920	#endif
 01921	
 01922	#define _LOW(v)         ( (v) & 0377)
 01923	#define _HIGH(v)        ( ((v) >> 8) & 0377)
 01924	
 01925	#define WNOHANG         1       /* do not wait for child to exit */
 01926	#define WUNTRACED       2       /* for job control; not implemented */
 01927	
 01928	#define WIFEXITED(s)    (_LOW(s) == 0)                      /* normal exit */
 01929	#define WEXITSTATUS(s)  (_HIGH(s))                          /* exit status */
 01930	#define WTERMSIG(s)     (_LOW(s) & 0177)                    /* sig value */
 01931	#define WIFSIGNALED(s)  (((unsigned int)(s)-1 & 0xFFFF) < 0xFF) /* signaled */
 01932	#define WIFSTOPPED(s)   (_LOW(s) == 0177)                   /* stopped */
 01933	#define WSTOPSIG(s)     (_HIGH(s) & 0377)                   /* stop signal */
 01934	
 01935	/* Function Prototypes. */
 01936	_PROTOTYPE( pid_t wait, (int *_stat_loc)                                   );
 01937	_PROTOTYPE( pid_t waitpid, (pid_t _pid, int *_stat_loc, int _options)      );
 01938	
 01939	#endif /* _WAIT_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/ioctl.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02000	/*      sys/ioctl.h - All ioctl() command codes.        Author: Kees J. Bot
 02001	 *                                                              23 Nov 2002
 02002	 *
 02003	 * This header file includes all other ioctl command code headers.
 02004	 */
 02005	
 02006	#ifndef _S_IOCTL_H
 02007	#define _S_IOCTL_H
 02008	
 02009	/* A driver that uses ioctls claims a character for its series of commands.
 02010	 * For instance:  #define TCGETS  _IOR('T',  8, struct termios)
 02011	 * This is a terminal ioctl that uses the character 'T'.  The character(s)
 02012	 * used in each header file are shown in the comment following.
 02013	 */
 02014	
 02015	#include <sys/ioc_tty.h>        /* 'T' 't' 'k'          */
 02016	#include <sys/ioc_disk.h>       /* 'd'                  */
 02017	#include <sys/ioc_memory.h>     /* 'm'                  */
 02018	#include <sys/ioc_cmos.h>       /* 'c'                  */
 02019	

_________________________ Page 663    File: include/sys/ioctl.h _________________________
 02020	#endif /* _S_IOCTL_H */





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/sys/ioc_disk.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02100	/*      sys/ioc_disk.h - Disk ioctl() command codes.    Author: Kees J. Bot
 02101	 *                                                              23 Nov 2002
 02102	 *
 02103	 */
 02104	
 02105	#ifndef _S_I_DISK_H
 02106	#define _S_I_DISK_H
 02107	
 02108	#include <minix/ioctl.h>
 02109	
 02110	#define DIOCSETP        _IOW('d', 3, struct partition)
 02111	#define DIOCGETP        _IOR('d', 4, struct partition)
 02112	#define DIOCEJECT       _IO ('d', 5)
 02113	#define DIOCTIMEOUT     _IOW('d', 6, int)
 02114	#define DIOCOPENCT      _IOR('d', 7, int)
 02115	
 02116	#endif /* _S_I_DISK_H */




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/ioctl.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02200	/*      minix/ioctl.h - Ioctl helper definitions.       Author: Kees J. Bot
 02201	 *                                                              23 Nov 2002
 02202	 *
 02203	 * This file is included by every header file that defines ioctl codes.
 02204	 */
 02205	
 02206	#ifndef _M_IOCTL_H
 02207	#define _M_IOCTL_H
 02208	
 02209	#ifndef _TYPES_H
 02210	#include <sys/types.h>
 02211	#endif
 02212	
 02213	#if _EM_WSIZE >= 4
 02214	/* Ioctls have the command encoded in the low-order word, and the size
 02215	 * of the parameter in the high-order word. The 3 high bits of the high-
 02216	 * order word are used to encode the in/out/void status of the parameter.
 02217	 */
 02218	#define _IOCPARM_MASK   0x1FFF
 02219	#define _IOC_VOID       0x20000000
 02220	#define _IOCTYPE_MASK   0xFFFF
 02221	#define _IOC_IN         0x40000000
 02222	#define _IOC_OUT        0x80000000
 02223	#define _IOC_INOUT      (_IOC_IN | _IOC_OUT)
 02224	

_________________________ Page 664    File: include/minix/ioctl.h ________________________
 02225	#define _IO(x,y)        ((x << 8) | y | _IOC_VOID)
 02226	#define _IOR(x,y,t)     ((x << 8) | y | ((sizeof(t) & _IOCPARM_MASK) << 16) |\
 02227	                                _IOC_OUT)
 02228	#define _IOW(x,y,t)     ((x << 8) | y | ((sizeof(t) & _IOCPARM_MASK) << 16) |\
 02229	                                _IOC_IN)
 02230	#define _IORW(x,y,t)    ((x << 8) | y | ((sizeof(t) & _IOCPARM_MASK) <
 02231	                                _IOC_INOUT)
 02232	#else
 02233	/* No fancy encoding on a 16-bit machine. */
 02234	
 02235	#define _IO(x,y)        ((x << 8) | y)
 02236	#define _IOR(x,y,t)     _IO(x,y)
 02237	#define _IOW(x,y,t)     _IO(x,y)
 02238	#define _IORW(x,y,t)    _IO(x,y)
 02239	#endif
 02240	
 02241	int ioctl(int _fd, int _request, void *_data);
 02242	
 02243	#endif /* _M_IOCTL_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/config.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02300	#ifndef _CONFIG_H
 02301	#define _CONFIG_H
 02302	
 02303	/* Minix release and version numbers. */
 02304	#define OS_RELEASE "3"
 02305	#define OS_VERSION "1.0"
 02306	
 02307	/* This file sets configuration parameters for the MINIX kernel, FS, and PM.
 02308	 * It is divided up into two main sections.  The first section contains
 02309	 * user-settable parameters.  In the second section, various internal system
 02310	 * parameters are set based on the user-settable parameters.
 02311	 *
 02312	 * Parts of config.h have been moved to sys_config.h, which can be included
 02313	 * by other include files that wish to get at the configuration data, but
 02314	 * don't want to pollute the users namespace. Some editable values have
 02315	 * gone there.
 02316	 *
 02317	 * This is a modified version of config.h for compiling a small Minix system
 02318	 * with only the options described in the text, Operating Systems Design and
 02319	 * Implementation, 3rd edition. See the version of config.h in the full 
 02320	 * source code directory for information on alternatives omitted here.
 02321	 */
 02322	
 02323	/* The MACHINE (called _MINIX_MACHINE) setting can be done
 02324	 * in <minix/machine.h>.
 02325	 */
 02326	#include <minix/sys_config.h>
 02327	
 02328	#define MACHINE      _MINIX_MACHINE
 02329	
 02330	#define IBM_PC       _MACHINE_IBM_PC
 02331	#define SUN_4        _MACHINE_SUN_4
 02332	#define SUN_4_60     _MACHINE_SUN_4_60
 02333	#define ATARI        _MACHINE_ATARI
 02334	#define MACINTOSH    _MACHINE_MACINTOSH

_________________________ Page 665    File: include/minix/config.h _______________________
 02335	
 02336	/* Number of slots in the process table for non-kernel processes. The number
 02337	 * of system processes defines how many processes with special privileges 
 02338	 * there can be. User processes share the same properties and count for one. 
 02339	 *
 02340	 * These can be changed in sys_config.h.
 02341	 */
 02342	#define NR_PROCS          _NR_PROCS 
 02343	#define NR_SYS_PROCS      _NR_SYS_PROCS
 02344	
 02345	#define NR_BUFS 128
 02346	#define NR_BUF_HASH 128
 02347	
 02348	/* Number of controller tasks (/dev/cN device classes). */
 02349	#define NR_CTRLRS          2
 02350	
 02351	/* Enable or disable the second level file system cache on the RAM disk. */
 02352	#define ENABLE_CACHE2      0
 02353	
 02354	/* Enable or disable swapping processes to disk. */
 02355	#define ENABLE_SWAP        0 
 02356	
 02357	/* Include or exclude an image of /dev/boot in the boot image. 
 02358	 * Please update the makefile in /usr/src/tools/ as well.
 02359	 */
 02360	#define ENABLE_BOOTDEV     0    /* load image of /dev/boot at boot time */
 02361	
 02362	/* DMA_SECTORS may be increased to speed up DMA based drivers. */
 02363	#define DMA_SECTORS        1    /* DMA buffer size (must be >= 1) */
 02364	
 02365	/* Include or exclude backwards compatibility code. */
 02366	#define ENABLE_BINCOMPAT   0    /* for binaries using obsolete calls */
 02367	#define ENABLE_SRCCOMPAT   0    /* for sources using obsolete calls */
 02368	
 02369	/* Which process should receive diagnostics from the kernel and system? 
 02370	 * Directly sending it to TTY only displays the output. Sending it to the
 02371	 * log driver will cause the diagnostics to be buffered and displayed.
 02372	 */
 02373	#define OUTPUT_PROC_NR  LOG_PROC_NR     /* TTY_PROC_NR or LOG_PROC_NR */
 02374	
 02375	/* NR_CONS, NR_RS_LINES, and NR_PTYS determine the number of terminals the
 02376	 * system can handle.
 02377	 */
 02378	#define NR_CONS            4    /* # system consoles (1 to 8) */
 02379	#define NR_RS_LINES        0    /* # rs232 terminals (0 to 4) */
 02380	#define NR_PTYS            0    /* # pseudo terminals (0 to 64) */
 02381	
 02382	/*===========================================================================*
 02383	 *      There are no user-settable parameters after this line                *
 02384	 *===========================================================================*/
 02385	/* Set the CHIP type based on the machine selected. The symbol CHIP is actually
 02386	 * indicative of more than just the CPU.  For example, machines for which
 02387	 * CHIP == INTEL are expected to have 8259A interrrupt controllers and the
 02388	 * other properties of IBM PC/XT/AT/386 types machines in general. */
 02389	#define INTEL             _CHIP_INTEL   /* CHIP type for PC, XT, AT, 386 and clones */
 02390	#define M68000            _CHIP_M68000  /* CHIP type for Atari, Amiga, Macintosh   
 02391	#define SPARC             _CHIP_SPARC   /* CHIP type for SUN-4 (e.g. SPARCstation) 
 02392	
 02393	/* Set the FP_FORMAT type based on the machine selected, either hw or sw    */
 02394	#define FP_NONE  _FP_NONE       /* no floating point support                */

_________________________ Page 666    File: include/minix/config.h _______________________
 02395	#define FP_IEEE  _FP_IEEE       /* conform IEEE floating point standard     */
 02396	
 02397	/* _MINIX_CHIP is defined in sys_config.h. */
 02398	#define CHIP    _MINIX_CHIP
 02399	
 02400	/* _MINIX_FP_FORMAT is defined in sys_config.h. */
 02401	#define FP_FORMAT       _MINIX_FP_FORMAT
 02402	
 02403	/* _ASKDEV and _FASTLOAD are defined in sys_config.h. */
 02404	#define ASKDEV _ASKDEV
 02405	#define FASTLOAD _FASTLOAD
 02406	
 02407	#endif /* _CONFIG_H */



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/sys_config.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02500	#ifndef _MINIX_SYS_CONFIG_H
 02501	#define _MINIX_SYS_CONFIG_H 1
 02502	
 02503	/* This is a modified sys_config.h for compiling a small Minix system 
 02504	 * with only the options described in the text, Operating Systems Design and 
 02505	 * Implementation, 3rd edition. See the sys_config.h in the full 
 02506	 * source code directory for information on alternatives omitted here.
 02507	 */
 02508	
 02509	/*===========================================================================*
 02510	 *              This section contains user-settable parameters               *
 02511	 *===========================================================================*/
 02512	#define _MINIX_MACHINE       _MACHINE_IBM_PC
 02513	
 02514	#define _MACHINE_IBM_PC             1   /* any  8088 or 80x86-based system */
 02515	
 02516	/* Word size in bytes (a constant equal to sizeof(int)). */
 02517	#if __ACK__ || __GNUC__
 02518	#define _WORD_SIZE      _EM_WSIZE
 02519	#define _PTR_SIZE       _EM_WSIZE
 02520	#endif
 02521	
 02522	#define _NR_PROCS       64
 02523	#define _NR_SYS_PROCS   32
 02524	
 02525	/* Set the CHIP type based on the machine selected. The symbol CHIP is actually
 02526	 * indicative of more than just the CPU.  For example, machines for which
 02527	 * CHIP == INTEL are expected to have 8259A interrrupt controllers and the
 02528	 * other properties of IBM PC/XT/AT/386 types machines in general. */
 02529	#define _CHIP_INTEL     1       /* CHIP type for PC, XT, AT, 386 and clones */
 02530	
 02531	/* Set the FP_FORMAT type based on the machine selected, either hw or sw    */
 02532	#define _FP_NONE        0       /* no floating point support                */
 02533	#define _FP_IEEE        1       /* conform IEEE floating point standard     */
 02534	
 02535	#define _MINIX_CHIP          _CHIP_INTEL
 02536	
 02537	#define _MINIX_FP_FORMAT   _FP_NONE
 02538	
 02539	#ifndef _MINIX_MACHINE

_________________________ Page 667    File: include/minix/sys_config.h ___________________
 02540	error "In <minix/sys_config.h> please define _MINIX_MACHINE"
 02541	#endif
 02542	
 02543	#ifndef _MINIX_CHIP
 02544	error "In <minix/sys_config.h> please define _MINIX_MACHINE to have a legal valu
 02545	#endif
 02546	
 02547	#if (_MINIX_MACHINE == 0)
 02548	error "_MINIX_MACHINE has incorrect value (0)"
 02549	#endif
 02550	
 02551	#endif /* _MINIX_SYS_CONFIG_H */
 02552	
 02553	


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/const.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02600	/* Copyright (C) 2001 by Prentice-Hall, Inc.  See the copyright notice in
 02601	 * the file /usr/src/LICENSE.
 02602	 */
 02603	
 02604	#ifndef CHIP
 02605	#error CHIP is not defined
 02606	#endif
 02607	
 02608	#define EXTERN        extern    /* used in *.h files */
 02609	#define PRIVATE       static    /* PRIVATE x limits the scope of x */
 02610	#define PUBLIC                  /* PUBLIC is the opposite of PRIVATE */
 02611	#define FORWARD       static    /* some compilers require this to be 'static'*/
 02612	
 02613	#define TRUE               1    /* used for turning integers into Booleans */
 02614	#define FALSE              0    /* used for turning integers into Booleans */
 02615	
 02616	#define HZ                60    /* clock freq (software settable on IBM-PC) */
 02617	
 02618	#define SUPER_USER (uid_t) 0    /* uid_t of superuser */
 02619	
 02620	/* Devices. */
 02621	#define MAJOR              8    /* major device = (dev>>MAJOR) & 0377 */
 02622	#define MINOR              0    /* minor device = (dev>>MINOR) & 0377 */
 02623	
 02624	#define NULL     ((void *)0)    /* null pointer */
 02625	#define CPVEC_NR          16    /* max # of entries in a SYS_VCOPY request */
 02626	#define CPVVEC_NR         64    /* max # of entries in a SYS_VCOPY request */
 02627	#define NR_IOREQS       MIN(NR_BUFS, 64)
 02628	                                /* maximum number of entries in an iorequest */
 02629	
 02630	/* Message passing constants. */
 02631	#define MESS_SIZE (sizeof(message))     /* might need usizeof from FS here */
 02632	#define NIL_MESS ((message *) 0)        /* null pointer */
 02633	
 02634	/* Memory related constants. */
 02635	#define SEGMENT_TYPE  0xFF00    /* bit mask to get segment type */
 02636	#define SEGMENT_INDEX 0x00FF    /* bit mask to get segment index */
 02637	
 02638	#define LOCAL_SEG     0x0000    /* flags indicating local memory segment */
 02639	#define NR_LOCAL_SEGS      3    /* # local segments per process (fixed) */

_________________________ Page 668    File: include/minix/const.h ________________________
 02640	#define T                  0    /* proc[i].mem_map[T] is for text */
 02641	#define D                  1    /* proc[i].mem_map[D] is for data */
 02642	#define S                  2    /* proc[i].mem_map[S] is for stack */
 02643	
 02644	#define REMOTE_SEG    0x0100    /* flags indicating remote memory segment */
 02645	#define NR_REMOTE_SEGS     3    /* # remote memory regions (variable) */
 02646	
 02647	#define BIOS_SEG      0x0200    /* flags indicating BIOS memory segment */
 02648	#define NR_BIOS_SEGS       3    /* # BIOS memory regions (variable) */
 02649	
 02650	#define PHYS_SEG      0x0400    /* flag indicating entire physical memory */
 02651	
 02652	/* Labels used to disable code sections for different reasons. */
 02653	#define DEAD_CODE          0    /* unused code in normal configuration */
 02654	#define FUTURE_CODE        0    /* new code to be activated + tested later */
 02655	#define TEMP_CODE          1    /* active code to be removed later */
 02656	
 02657	/* Process name length in the PM process table, including '\0'. */
 02658	#define PROC_NAME_LEN   16
 02659	
 02660	/* Miscellaneous */
 02661	#define BYTE            0377    /* mask for 8 bits */
 02662	#define READING            0    /* copy data to user */
 02663	#define WRITING            1    /* copy data from user */
 02664	#define NO_NUM        0x8000    /* used as numerical argument to panic() */
 02665	#define NIL_PTR   (char *) 0    /* generally useful expression */
 02666	#define HAVE_SCATTERED_IO  1    /* scattered I/O is now standard */
 02667	
 02668	/* Macros. */
 02669	#define MAX(a, b)   ((a) > (b) ? (a) : (b))
 02670	#define MIN(a, b)   ((a) < (b) ? (a) : (b))
 02671	
 02672	/* Memory is allocated in clicks. */
 02673	#if (CHIP == INTEL)
 02674	#define CLICK_SIZE      1024    /* unit in which memory is allocated */
 02675	#define CLICK_SHIFT       10    /* log2 of CLICK_SIZE */
 02676	#endif
 02677	
 02678	#if (CHIP == SPARC) || (CHIP == M68000)
 02679	#define CLICK_SIZE      4096    /* unit in which memory is allocated */
 02680	#define CLICK_SHIFT       12    /* log2 of CLICK_SIZE */
 02681	#endif
 02682	
 02683	/* Click to byte conversions (and vice versa). */
 02684	#define HCLICK_SHIFT       4    /* log2 of HCLICK_SIZE */
 02685	#define HCLICK_SIZE       16    /* hardware segment conversion magic */
 02686	#if CLICK_SIZE >= HCLICK_SIZE
 02687	#define click_to_hclick(n) ((n) << (CLICK_SHIFT - HCLICK_SHIFT))
 02688	#else
 02689	#define click_to_hclick(n) ((n) >> (HCLICK_SHIFT - CLICK_SHIFT))
 02690	#endif
 02691	#define hclick_to_physb(n) ((phys_bytes) (n) << HCLICK_SHIFT)
 02692	#define physb_to_hclick(n) ((n) >> HCLICK_SHIFT)
 02693	
 02694	#define ABS             -999    /* this process means absolute memory */
 02695	
 02696	/* Flag bits for i_mode in the inode. */
 02697	#define I_TYPE          0170000 /* this field gives inode type */
 02698	#define I_REGULAR       0100000 /* regular file, not dir or special */
 02699	#define I_BLOCK_SPECIAL 0060000 /* block special file */

_________________________ Page 669    File: include/minix/const.h ________________________
 02700	#define I_DIRECTORY     0040000 /* file is a directory */
 02701	#define I_CHAR_SPECIAL  0020000 /* character special file */
 02702	#define I_NAMED_PIPE    0010000 /* named pipe (FIFO) */
 02703	#define I_SET_UID_BIT   0004000 /* set effective uid_t on exec */
 02704	#define I_SET_GID_BIT   0002000 /* set effective gid_t on exec */
 02705	#define ALL_MODES       0006777 /* all bits for user, group and others */
 02706	#define RWX_MODES       0000777 /* mode bits for RWX only */
 02707	#define R_BIT           0000004 /* Rwx protection bit */
 02708	#define W_BIT           0000002 /* rWx protection bit */
 02709	#define X_BIT           0000001 /* rwX protection bit */
 02710	#define I_NOT_ALLOC     0000000 /* this inode is free */
 02711	
 02712	/* Flag used only in flags argument of dev_open. */
 02713	#define RO_BIT          0200000 /* Open device readonly; fail if writable. */
 02714	
 02715	/* Some limits. */
 02716	#define MAX_BLOCK_NR  ((block_t) 077777777)     /* largest block number */
 02717	#define HIGHEST_ZONE   ((zone_t) 077777777)     /* largest zone number */
 02718	#define MAX_INODE_NR ((ino_t) 037777777777)     /* largest inode number */
 02719	#define MAX_FILE_POS ((off_t) 037777777777)     /* largest legal file offset */
 02720	
 02721	#define NO_BLOCK              ((block_t) 0)     /* absence of a block number */
 02722	#define NO_ENTRY                ((ino_t) 0)     /* absence of a dir entry */
 02723	#define NO_ZONE                ((zone_t) 0)     /* absence of a zone number */
 02724	#define NO_DEV                  ((dev_t) 0)     /* absence of a device numb */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/type.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 02800	#ifndef _TYPE_H
 02801	#define _TYPE_H
 02802	
 02803	#ifndef _MINIX_SYS_CONFIG_H
 02804	#include <minix/sys_config.h>
 02805	#endif
 02806	
 02807	#ifndef _TYPES_H
 02808	#include <sys/types.h>
 02809	#endif
 02810	
 02811	/* Type definitions. */
 02812	typedef unsigned int vir_clicks;        /*  virtual addr/length in clicks */
 02813	typedef unsigned long phys_bytes;       /* physical addr/length in bytes */
 02814	typedef unsigned int phys_clicks;       /* physical addr/length in clicks */
 02815	
 02816	#if (_MINIX_CHIP == _CHIP_INTEL)
 02817	typedef unsigned int vir_bytes; /* virtual addresses and lengths in bytes */
 02818	#endif
 02819	
 02820	#if (_MINIX_CHIP == _CHIP_M68000)
 02821	typedef unsigned long vir_bytes;/* virtual addresses and lengths in bytes */
 02822	#endif
 02823	
 02824	#if (_MINIX_CHIP == _CHIP_SPARC)
 02825	typedef unsigned long vir_bytes;/* virtual addresses and lengths in bytes */
 02826	#endif
 02827	
 02828	/* Memory map for local text, stack, data segments. */
 02829	struct mem_map {

_________________________ Page 670    File: include/minix/type.h _________________________
 02830	  vir_clicks mem_vir;           /* virtual address */
 02831	  phys_clicks mem_phys;         /* physical address */
 02832	  vir_clicks mem_len;           /* length */
 02833	};
 02834	
 02835	/* Memory map for remote memory areas, e.g., for the RAM disk. */
 02836	struct far_mem {
 02837	  int in_use;                   /* entry in use, unless zero */
 02838	  phys_clicks mem_phys;         /* physical address */
 02839	  vir_clicks mem_len;           /* length */
 02840	};
 02841	
 02842	/* Structure for virtual copying by means of a vector with requests. */
 02843	struct vir_addr {
 02844	  int proc_nr;
 02845	  int segment;
 02846	  vir_bytes offset;
 02847	};
 02848	
 02849	#define phys_cp_req vir_cp_req 
 02850	struct vir_cp_req {
 02851	  struct vir_addr src;
 02852	  struct vir_addr dst;
 02853	  phys_bytes count;
 02854	};
 02855	
 02856	typedef struct {
 02857	  vir_bytes iov_addr;           /* address of an I/O buffer */
 02858	  vir_bytes iov_size;           /* sizeof an I/O buffer */
 02859	} iovec_t;
 02860	
 02861	/* PM passes the address of a structure of this type to KERNEL when
 02862	 * sys_sendsig() is invoked as part of the signal catching mechanism.
 02863	 * The structure contain all the information that KERNEL needs to build
 02864	 * the signal stack.
 02865	 */
 02866	struct sigmsg {
 02867	  int sm_signo;                 /* signal number being caught */
 02868	  unsigned long sm_mask;        /* mask to restore when handler returns */
 02869	  vir_bytes sm_sighandler;      /* address of handler */
 02870	  vir_bytes sm_sigreturn;       /* address of _sigreturn in C library */
 02871	  vir_bytes sm_stkptr;          /* user stack pointer */
 02872	};
 02873	
 02874	/* This is used to obtain system information through SYS_GETINFO. */
 02875	struct kinfo {
 02876	  phys_bytes code_base;         /* base of kernel code */
 02877	  phys_bytes code_size;         
 02878	  phys_bytes data_base;         /* base of kernel data */
 02879	  phys_bytes data_size;
 02880	  vir_bytes proc_addr;          /* virtual address of process table */
 02881	  phys_bytes kmem_base;         /* kernel memory layout (/dev/kmem) */
 02882	  phys_bytes kmem_size;
 02883	  phys_bytes bootdev_base;      /* boot device from boot image (/dev/boot) */
 02884	  phys_bytes bootdev_size;
 02885	  phys_bytes bootdev_mem;
 02886	  phys_bytes params_base;       /* parameters passed by boot monitor */
 02887	  phys_bytes params_size;
 02888	  int nr_procs;                 /* number of user processes */
 02889	  int nr_tasks;                 /* number of kernel tasks */

_________________________ Page 671    File: include/minix/type.h _________________________
 02890	  char release[6];              /* kernel release number */
 02891	  char version[6];              /* kernel version number */
 02892	  int relocking;                /* relocking check (for debugging) */
 02893	};
 02894	
 02895	struct machine {
 02896	  int pc_at;
 02897	  int ps_mca;
 02898	  int processor;
 02899	  int protected;
 02900	  int vdu_ega;
 02901	  int vdu_vga;
 02902	};
 02903	
 02904	#endif /* _TYPE_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/ipc.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 03000	#ifndef _IPC_H
 03001	#define _IPC_H
 03002	
 03003	/*==========================================================================* 
 03004	 * Types relating to messages.                                              *
 03005	 *==========================================================================*/ 
 03006	
 03007	#define M1                 1
 03008	#define M3                 3
 03009	#define M4                 4
 03010	#define M3_STRING         14
 03011	
 03012	typedef struct {int m1i1, m1i2, m1i3; char *m1p1, *m1p2, *m1p3;} mess_1;
 03013	typedef struct {int m2i1, m2i2, m2i3; long m2l1, m2l2; char *m2p1;} mess_2;
 03014	typedef struct {int m3i1, m3i2; char *m3p1; char m3ca1[M3_STRING];} mess_3;
 03015	typedef struct {long m4l1, m4l2, m4l3, m4l4, m4l5;} mess_4;
 03016	typedef struct {short m5c1, m5c2; int m5i1, m5i2; long m5l1, m5l2, m5l3;}mess_5;
 03017	typedef struct {int m7i1, m7i2, m7i3, m7i4; char *m7p1, *m7p2;} mess_7;
 03018	typedef struct {int m8i1, m8i2; char *m8p1, *m8p2, *m8p3, *m8p4;} mess_8;
 03019	
 03020	typedef struct {
 03021	  int m_source;                 /* who sent the message */
 03022	  int m_type;                   /* what kind of message is it */
 03023	  union {
 03024	        mess_1 m_m1;
 03025	        mess_2 m_m2;
 03026	        mess_3 m_m3;
 03027	        mess_4 m_m4;
 03028	        mess_5 m_m5;
 03029	        mess_7 m_m7;
 03030	        mess_8 m_m8;
 03031	  } m_u;
 03032	} message;
 03033	
 03034	/* The following defines provide names for useful members. */
 03035	#define m1_i1  m_u.m_m1.m1i1
 03036	#define m1_i2  m_u.m_m1.m1i2
 03037	#define m1_i3  m_u.m_m1.m1i3
 03038	#define m1_p1  m_u.m_m1.m1p1
 03039	#define m1_p2  m_u.m_m1.m1p2

_________________________ Page 672    File: include/minix/ipc.h _________________________
 03040	#define m1_p3  m_u.m_m1.m1p3
 03041	
 03042	#define m2_i1  m_u.m_m2.m2i1
 03043	#define m2_i2  m_u.m_m2.m2i2
 03044	#define m2_i3  m_u.m_m2.m2i3
 03045	#define m2_l1  m_u.m_m2.m2l1
 03046	#define m2_l2  m_u.m_m2.m2l2
 03047	#define m2_p1  m_u.m_m2.m2p1
 03048	
 03049	#define m3_i1  m_u.m_m3.m3i1
 03050	#define m3_i2  m_u.m_m3.m3i2
 03051	#define m3_p1  m_u.m_m3.m3p1
 03052	#define m3_ca1 m_u.m_m3.m3ca1
 03053	
 03054	#define m4_l1  m_u.m_m4.m4l1
 03055	#define m4_l2  m_u.m_m4.m4l2
 03056	#define m4_l3  m_u.m_m4.m4l3
 03057	#define m4_l4  m_u.m_m4.m4l4
 03058	#define m4_l5  m_u.m_m4.m4l5
 03059	
 03060	#define m5_c1  m_u.m_m5.m5c1
 03061	#define m5_c2  m_u.m_m5.m5c2
 03062	#define m5_i1  m_u.m_m5.m5i1
 03063	#define m5_i2  m_u.m_m5.m5i2
 03064	#define m5_l1  m_u.m_m5.m5l1
 03065	#define m5_l2  m_u.m_m5.m5l2
 03066	#define m5_l3  m_u.m_m5.m5l3
 03067	
 03068	#define m7_i1  m_u.m_m7.m7i1
 03069	#define m7_i2  m_u.m_m7.m7i2
 03070	#define m7_i3  m_u.m_m7.m7i3
 03071	#define m7_i4  m_u.m_m7.m7i4
 03072	#define m7_p1  m_u.m_m7.m7p1
 03073	#define m7_p2  m_u.m_m7.m7p2
 03074	
 03075	#define m8_i1  m_u.m_m8.m8i1
 03076	#define m8_i2  m_u.m_m8.m8i2
 03077	#define m8_p1  m_u.m_m8.m8p1
 03078	#define m8_p2  m_u.m_m8.m8p2
 03079	#define m8_p3  m_u.m_m8.m8p3
 03080	#define m8_p4  m_u.m_m8.m8p4
 03081	
 03082	/*==========================================================================* 
 03083	 * Minix run-time system (IPC).                                             *
 03084	 *==========================================================================*/ 
 03085	
 03086	/* Hide names to avoid name space pollution. */
 03087	#define echo            _echo
 03088	#define notify          _notify
 03089	#define sendrec         _sendrec
 03090	#define receive         _receive
 03091	#define send            _send
 03092	#define nb_receive      _nb_receive
 03093	#define nb_send         _nb_send
 03094	
 03095	_PROTOTYPE( int echo, (message *m_ptr)                                  );
 03096	_PROTOTYPE( int notify, (int dest)                                      );
 03097	_PROTOTYPE( int sendrec, (int src_dest, message *m_ptr)                 );
 03098	_PROTOTYPE( int receive, (int src, message *m_ptr)                      );
 03099	_PROTOTYPE( int send, (int dest, message *m_ptr)                        );

_________________________ Page 673    File: include/minix/ipc.h _________________________
 03100	_PROTOTYPE( int nb_receive, (int src, message *m_ptr)                   );
 03101	_PROTOTYPE( int nb_send, (int dest, message *m_ptr)                     );
 03102	
 03103	#endif /* _IPC_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/syslib.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 03200	/* Prototypes for system library functions. */
 03201	
 03202	#ifndef _SYSLIB_H
 03203	#define _SYSLIB_H
 03204	
 03205	#ifndef _TYPES_H
 03206	#include <sys/types.h>
 03207	#endif
 03208	
 03209	#ifndef _IPC_H
 03210	#include <minix/ipc.h>
 03211	#endif
 03212	
 03213	#ifndef _DEVIO_H
 03214	#include <minix/devio.h>
 03215	#endif
 03216	
 03217	/* Forward declaration */
 03218	struct reg86u;
 03219	
 03220	#define SYSTASK SYSTEM
 03221	
 03222	/*==========================================================================* 
 03223	 * Minix system library.                                                    *
 03224	 *==========================================================================*/ 
 03225	_PROTOTYPE( int _taskcall, (int who, int syscallnr, message *msgptr));
 03226	
 03227	_PROTOTYPE( int sys_abort, (int how, ...));
 03228	_PROTOTYPE( int sys_exec, (int proc, char *ptr,  
 03229	                                char *aout, vir_bytes initpc));
 03230	_PROTOTYPE( int sys_fork, (int parent, int child));
 03231	_PROTOTYPE( int sys_newmap, (int proc, struct mem_map *ptr));
 03232	_PROTOTYPE( int sys_exit, (int proc));
 03233	_PROTOTYPE( int sys_trace, (int req, int proc, long addr, long *data_p));
 03234	
 03235	_PROTOTYPE( int sys_svrctl, (int proc, int req, int priv,vir_bytes argp));
 03236	_PROTOTYPE( int sys_nice, (int proc, int priority));
 03237	
 03238	_PROTOTYPE( int sys_int86, (struct reg86u *reg86p));
 03239	
 03240	/* Shorthands for sys_sdevio() system call. */
 03241	#define sys_insb(port, proc_nr, buffer, count) \
 03242	        sys_sdevio(DIO_INPUT, port, DIO_BYTE, proc_nr, buffer, count)
 03243	#define sys_insw(port, proc_nr, buffer, count) \
 03244	        sys_sdevio(DIO_INPUT, port, DIO_WORD, proc_nr, buffer, count)
 03245	#define sys_outsb(port, proc_nr, buffer, count) \
 03246	        sys_sdevio(DIO_OUTPUT, port, DIO_BYTE, proc_nr, buffer, count)
 03247	#define sys_outsw(port, proc_nr, buffer, count) \
 03248	        sys_sdevio(DIO_OUTPUT, port, DIO_WORD, proc_nr, buffer, count)
 03249	_PROTOTYPE( int sys_sdevio, (int req, long port, int type, int proc_nr,

_________________________ Page 674    File: include/minix/syslib.h _______________________
 03250	        void *buffer, int count));
 03251	
 03252	/* Clock functionality: get system times or (un)schedule an alarm call. */
 03253	_PROTOTYPE( int sys_times, (int proc_nr, clock_t *ptr));
 03254	_PROTOTYPE(int sys_setalarm, (clock_t exp_time, int abs_time));
 03255	
 03256	/* Shorthands for sys_irqctl() system call. */
 03257	#define sys_irqdisable(hook_id) \
 03258	    sys_irqctl(IRQ_DISABLE, 0, 0, hook_id) 
 03259	#define sys_irqenable(hook_id) \
 03260	    sys_irqctl(IRQ_ENABLE, 0, 0, hook_id) 
 03261	#define sys_irqsetpolicy(irq_vec, policy, hook_id) \
 03262	    sys_irqctl(IRQ_SETPOLICY, irq_vec, policy, hook_id)
 03263	#define sys_irqrmpolicy(irq_vec, hook_id) \
 03264	    sys_irqctl(IRQ_RMPOLICY, irq_vec, 0, hook_id)
 03265	_PROTOTYPE ( int sys_irqctl, (int request, int irq_vec, int policy,
 03266	    int *irq_hook_id) );
 03267	
 03268	/* Shorthands for sys_vircopy() and sys_physcopy() system calls. */
 03269	#define sys_biosin(bios_vir, dst_vir, bytes) \
 03270	        sys_vircopy(SELF, BIOS_SEG, bios_vir, SELF, D, dst_vir, bytes)
 03271	#define sys_biosout(src_vir, bios_vir, bytes) \
 03272	        sys_vircopy(SELF, D, src_vir, SELF, BIOS_SEG, bios_vir, bytes)
 03273	#define sys_datacopy(src_proc, src_vir, dst_proc, dst_vir, bytes) \
 03274	        sys_vircopy(src_proc, D, src_vir, dst_proc, D, dst_vir, bytes)
 03275	#define sys_textcopy(src_proc, src_vir, dst_proc, dst_vir, bytes) \
 03276	        sys_vircopy(src_proc, T, src_vir, dst_proc, T, dst_vir, bytes)
 03277	#define sys_stackcopy(src_proc, src_vir, dst_proc, dst_vir, bytes) \
 03278	        sys_vircopy(src_proc, S, src_vir, dst_proc, S, dst_vir, bytes)
 03279	_PROTOTYPE(int sys_vircopy, (int src_proc, int src_seg, vir_bytes src_vir,
 03280	        int dst_proc, int dst_seg, vir_bytes dst_vir, phys_bytes bytes));
 03281	
 03282	#define sys_abscopy(src_phys, dst_phys, bytes) \
 03283	        sys_physcopy(NONE, PHYS_SEG, src_phys, NONE, PHYS_SEG, dst_phys, bytes)
 03284	_PROTOTYPE(int sys_physcopy, (int src_proc, int src_seg, vir_bytes src_vir,
 03285	        int dst_proc, int dst_seg, vir_bytes dst_vir, phys_bytes bytes));
 03286	_PROTOTYPE(int sys_memset, (unsigned long pattern, 
 03287	                phys_bytes base, phys_bytes bytes));
 03288	
 03289	/* Vectored virtual / physical copy calls. */
 03290	#if DEAD_CODE           /* library part not yet implemented */
 03291	_PROTOTYPE(int sys_virvcopy, (phys_cp_req *vec_ptr,int vec_size,int *nr_ok));
 03292	_PROTOTYPE(int sys_physvcopy, (phys_cp_req *vec_ptr,int vec_size,int *nr_ok));
 03293	#endif
 03294	
 03295	_PROTOTYPE(int sys_umap, (int proc_nr, int seg, vir_bytes vir_addr,
 03296	         vir_bytes bytes, phys_bytes *phys_addr));
 03297	_PROTOTYPE(int sys_segctl, (int *index, u16_t *seg, vir_bytes *off,
 03298	        phys_bytes phys, vir_bytes size));
 03299	
 03300	/* Shorthands for sys_getinfo() system call. */
 03301	#define sys_getkmessages(dst)   sys_getinfo(GET_KMESSAGES, dst, 0,0,0)
 03302	#define sys_getkinfo(dst)       sys_getinfo(GET_KINFO, dst, 0,0,0)
 03303	#define sys_getmachine(dst)     sys_getinfo(GET_MACHINE, dst, 0,0,0)
 03304	#define sys_getproctab(dst)     sys_getinfo(GET_PROCTAB, dst, 0,0,0)
 03305	#define sys_getprivtab(dst)     sys_getinfo(GET_PRIVTAB, dst, 0,0,0)
 03306	#define sys_getproc(dst,nr)     sys_getinfo(GET_PROC, dst, 0,0, nr)
 03307	#define sys_getrandomness(dst)  sys_getinfo(GET_RANDOMNESS, dst, 0,0,0)
 03308	#define sys_getimage(dst)       sys_getinfo(GET_IMAGE, dst, 0,0,0)
 03309	#define sys_getirqhooks(dst)    sys_getinfo(GET_IRQHOOKS, dst, 0,0,0)

_________________________ Page 675    File: include/minix/syslib.h _______________________
 03310	#define sys_getmonparams(v,vl)  sys_getinfo(GET_MONPARAMS, v,vl, 0,0)
 03311	#define sys_getschedinfo(v1,v2) sys_getinfo(GET_SCHEDINFO, v1,0, v2,0)
 03312	#define sys_getlocktimings(dst) sys_getinfo(GET_LOCKTIMING, dst, 0,0,0)
 03313	#define sys_getbiosbuffer(virp, sizep) sys_getinfo(GET_BIOSBUFFER, virp, \
 03314	        sizeof(*virp), sizep, sizeof(*sizep))
 03315	_PROTOTYPE(int sys_getinfo, (int request, void *val_ptr, int val_len,
 03316	                                 void *val_ptr2, int val_len2)          );
 03317	
 03318	/* Signal control. */
 03319	_PROTOTYPE(int sys_kill, (int proc, int sig) );
 03320	_PROTOTYPE(int sys_sigsend, (int proc_nr, struct sigmsg *sig_ctxt) ); 
 03321	_PROTOTYPE(int sys_sigreturn, (int proc_nr, struct sigmsg *sig_ctxt) );
 03322	_PROTOTYPE(int sys_getksig, (int *k_proc_nr, sigset_t *k_sig_map) ); 
 03323	_PROTOTYPE(int sys_endksig, (int proc_nr) );
 03324	
 03325	/* NOTE: two different approaches were used to distinguish the device I/O
 03326	 * types 'byte', 'word', 'long': the latter uses #define and results in a
 03327	 * smaller implementation, but looses the static type checking.
 03328	 */
 03329	_PROTOTYPE(int sys_voutb, (pvb_pair_t *pvb_pairs, int nr_ports)         );
 03330	_PROTOTYPE(int sys_voutw, (pvw_pair_t *pvw_pairs, int nr_ports)         );
 03331	_PROTOTYPE(int sys_voutl, (pvl_pair_t *pvl_pairs, int nr_ports)         );
 03332	_PROTOTYPE(int sys_vinb, (pvb_pair_t *pvb_pairs, int nr_ports)          );
 03333	_PROTOTYPE(int sys_vinw, (pvw_pair_t *pvw_pairs, int nr_ports)          );
 03334	_PROTOTYPE(int sys_vinl, (pvl_pair_t *pvl_pairs, int nr_ports)          );
 03335	
 03336	/* Shorthands for sys_out() system call. */
 03337	#define sys_outb(p,v)   sys_out((p), (unsigned long) (v), DIO_BYTE)
 03338	#define sys_outw(p,v)   sys_out((p), (unsigned long) (v), DIO_WORD)
 03339	#define sys_outl(p,v)   sys_out((p), (unsigned long) (v), DIO_LONG)
 03340	_PROTOTYPE(int sys_out, (int port, unsigned long value, int type)       ); 
 03341	
 03342	/* Shorthands for sys_in() system call. */
 03343	#define sys_inb(p,v)    sys_in((p), (unsigned long*) (v), DIO_BYTE)
 03344	#define sys_inw(p,v)    sys_in((p), (unsigned long*) (v), DIO_WORD)
 03345	#define sys_inl(p,v)    sys_in((p), (unsigned long*) (v), DIO_LONG)
 03346	_PROTOTYPE(int sys_in, (int port, unsigned long *value, int type)       );
 03347	
 03348	#endif /* _SYSLIB_H */
 03349	

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/sysutil.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 03400	#ifndef _EXTRALIB_H
 03401	#define _EXTRALIB_H
 03402	
 03403	/* Extra system library definitions to support device drivers and servers.
 03404	 *
 03405	 * Created:
 03406	 *      Mar 15, 2004 by Jorrit N. Herder
 03407	 *
 03408	 * Changes:
 03409	 *      May 31, 2005: added printf, kputc (relocated from syslib)
 03410	 *      May 31, 2005: added getuptime
 03411	 *      Mar 18, 2005: added tickdelay
 03412	 *      Oct 01, 2004: added env_parse, env_prefix, env_panic
 03413	 *      Jul 13, 2004: added fkey_ctl
 03414	 *      Apr 28, 2004: added report, panic 

_________________________ Page 676    File: include/minix/sysutil.h ______________________
 03415	 *      Mar 31, 2004: setup like other libraries, such as syslib
 03416	 */
 03417	
 03418	/*==========================================================================* 
 03419	 * Miscellaneous helper functions.
 03420	 *==========================================================================*/ 
 03421	
 03422	/* Environment parsing return values. */
 03423	#define EP_BUF_SIZE   128       /* local buffer for env value */
 03424	#define EP_UNSET        0       /* variable not set */
 03425	#define EP_OFF          1       /* var = off */
 03426	#define EP_ON           2       /* var = on (or field left blank) */
 03427	#define EP_SET          3       /* var = 1:2:3 (nonblank field) */
 03428	#define EP_EGETKENV     4       /* sys_getkenv() failed ... */
 03429	
 03430	_PROTOTYPE( void env_setargs, (int argc, char *argv[])                  );
 03431	_PROTOTYPE( int env_get_param, (char *key, char *value, int max_size)   );
 03432	_PROTOTYPE( int env_prefix, (char *env, char *prefix)                   );
 03433	_PROTOTYPE( void env_panic, (char *key)                                 );
 03434	_PROTOTYPE( int env_parse, (char *env, char *fmt, int field, long *param,
 03435	                                long min, long max)                     );
 03436	
 03437	#define fkey_map(fkeys, sfkeys) fkey_ctl(FKEY_MAP, (fkeys), (sfkeys))
 03438	#define fkey_unmap(fkeys, sfkeys) fkey_ctl(FKEY_UNMAP, (fkeys), (sfkeys))
 03439	#define fkey_events(fkeys, sfkeys) fkey_ctl(FKEY_EVENTS, (fkeys), (sfkeys))
 03440	_PROTOTYPE( int fkey_ctl, (int req, int *fkeys, int *sfkeys)            );
 03441	
 03442	_PROTOTYPE( int printf, (const char *fmt, ...));
 03443	_PROTOTYPE( void kputc, (int c));
 03444	_PROTOTYPE( void report, (char *who, char *mess, int num));
 03445	_PROTOTYPE( void panic, (char *who, char *mess, int num));
 03446	_PROTOTYPE( int getuptime, (clock_t *ticks));
 03447	_PROTOTYPE( int tickdelay, (clock_t ticks));
 03448	
 03449	#endif /* _EXTRALIB_H */
 03450	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/callnr.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 03500	#define NCALLS            91    /* number of system calls allowed */
 03501	
 03502	#define EXIT               1 
 03503	#define FORK               2 
 03504	#define READ               3 
 03505	#define WRITE              4 
 03506	#define OPEN               5 
 03507	#define CLOSE              6 
 03508	#define WAIT               7
 03509	#define CREAT              8 
 03510	#define LINK               9 
 03511	#define UNLINK            10 
 03512	#define WAITPID           11
 03513	#define CHDIR             12 
 03514	#define TIME              13

_________________________ Page 677    File: include/minix/callnr.h _______________________
 03515	#define MKNOD             14 
 03516	#define CHMOD             15 
 03517	#define CHOWN             16 
 03518	#define BRK               17
 03519	#define STAT              18 
 03520	#define LSEEK             19
 03521	#define GETPID            20
 03522	#define MOUNT             21 
 03523	#define UMOUNT            22 
 03524	#define SETUID            23
 03525	#define GETUID            24
 03526	#define STIME             25
 03527	#define PTRACE            26
 03528	#define ALARM             27
 03529	#define FSTAT             28 
 03530	#define PAUSE             29
 03531	#define UTIME             30 
 03532	#define ACCESS            33 
 03533	#define SYNC              36 
 03534	#define KILL              37
 03535	#define RENAME            38
 03536	#define MKDIR             39
 03537	#define RMDIR             40
 03538	#define DUP               41 
 03539	#define PIPE              42 
 03540	#define TIMES             43
 03541	#define SETGID            46
 03542	#define GETGID            47
 03543	#define SIGNAL            48
 03544	#define IOCTL             54
 03545	#define FCNTL             55
 03546	#define EXEC              59
 03547	#define UMASK             60 
 03548	#define CHROOT            61 
 03549	#define SETSID            62
 03550	#define GETPGRP           63
 03551	
 03552	/* The following are not system calls, but are processed like them. */
 03553	#define UNPAUSE           65    /* to MM or FS: check for EINTR */
 03554	#define REVIVE            67    /* to FS: revive a sleeping process */
 03555	#define TASK_REPLY        68    /* to FS: reply code from tty task */
 03556	
 03557	/* Posix signal handling. */
 03558	#define SIGACTION         71
 03559	#define SIGSUSPEND        72
 03560	#define SIGPENDING        73
 03561	#define SIGPROCMASK       74
 03562	#define SIGRETURN         75
 03563	
 03564	#define REBOOT            76    /* to PM */
 03565	
 03566	/* MINIX specific calls, e.g., to support system services. */
 03567	#define SVRCTL            77
 03568	                                /* unused */
 03569	#define GETSYSINFO        79    /* to PM or FS */
 03570	#define GETPROCNR         80    /* to PM */
 03571	#define DEVCTL            81    /* to FS */
 03572	#define FSTATFS           82    /* to FS */
 03573	#define ALLOCMEM          83    /* to PM */
 03574	#define FREEMEM           84    /* to PM */

_________________________ Page 678    File: include/minix/callnr.h _______________________
 03575	#define SELECT            85    /* to FS */
 03576	#define FCHDIR            86    /* to FS */
 03577	#define FSYNC             87    /* to FS */
 03578	#define GETPRIORITY       88    /* to PM */
 03579	#define SETPRIORITY       89    /* to PM */
 03580	#define GETTIMEOFDAY      90    /* to PM */





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/com.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 03600	#ifndef _MINIX_COM_H
 03601	#define _MINIX_COM_H 
 03602	
 03603	/*===========================================================================*
 03604	 *                              Magic process numbers                        *
 03605	 *===========================================================================*/
 03606	
 03607	#define ANY             0x7ace  /* used to indicate 'any process' */
 03608	#define NONE            0x6ace  /* used to indicate 'no process at all' */
 03609	#define SELF            0x8ace  /* used to indicate 'own process' */
 03610	
 03611	/*===========================================================================*
 03612	 *              Process numbers of processes in the system image             *
 03613	 *===========================================================================*/
 03614	
 03615	/* The values of several task numbers depend on whether they or other tasks
 03616	 * are enabled. They are defined as (PREVIOUS_TASK - ENABLE_TASK) in general.
 03617	 * ENABLE_TASK is either 0 or 1, so a task either gets a new number, or gets
 03618	 * the same number as the previous task and is further unused. Note that the
 03619	 * order should correspond to the order in the task table defined in table.c. 
 03620	 */
 03621	
 03622	/* Kernel tasks. These all run in the same address space. */
 03623	#define IDLE             -4     /* runs when no one else can run */
 03624	#define CLOCK            -3     /* alarms and other clock functions */
 03625	#define SYSTEM           -2     /* request system functionality */
 03626	#define KERNEL           -1     /* pseudo-process for IPC and scheduling */
 03627	#define HARDWARE     KERNEL     /* for hardware interrupt handlers */
 03628	
 03629	/* Number of tasks. Note that NR_PROCS is defined in <minix/config.h>. */
 03630	#define NR_TASKS          4 
 03631	
 03632	/* User-space processes, that is, device drivers, servers, and INIT. */
 03633	#define PM_PROC_NR        0     /* process manager */
 03634	#define FS_PROC_NR        1     /* file system */
 03635	#define RS_PROC_NR        2     /* reincarnation server */
 03636	#define MEM_PROC_NR       3     /* memory driver (RAM disk, null, etc.) */
 03637	#define LOG_PROC_NR       4     /* log device driver */
 03638	#define TTY_PROC_NR       5     /* terminal (TTY) driver */
 03639	#define DRVR_PROC_NR      6     /* device driver for boot medium */
 03640	#define INIT_PROC_NR      7     /* init -- goes multiuser */
 03641	
 03642	/* Number of processes contained in the system image. */
 03643	#define NR_BOOT_PROCS   (NR_TASKS + INIT_PROC_NR + 1)
 03644	

_________________________ Page 679    File: include/minix/com.h _________________________
 03645	/*===========================================================================*
 03646	 *                         Kernel notification types                         *
 03647	 *===========================================================================*/
 03648	
 03649	/* Kernel notification types. In principle, these can be sent to any process,
 03650	 * so make sure that these types do not interfere with other message types.
 03651	 * Notifications are prioritized because of the way they are unhold() and
 03652	 * blocking notifications are delivered. The lowest numbers go first. The
 03653	 * offset are used for the per-process notification bit maps. 
 03654	 */
 03655	#define NOTIFY_MESSAGE            0x1000
 03656	#define NOTIFY_FROM(p_nr)        (NOTIFY_MESSAGE | ((p_nr) + NR_TASKS)) 
 03657	#  define SYN_ALARM     NOTIFY_FROM(CLOCK)      /* synchronous alarm */
 03658	#  define SYS_SIG       NOTIFY_FROM(SYSTEM)     /* system signal */
 03659	#  define HARD_INT      NOTIFY_FROM(HARDWARE)   /* hardware interrupt */
 03660	#  define NEW_KSIG      NOTIFY_FROM(HARDWARE)   /* new kernel signal */
 03661	#  define FKEY_PRESSED  NOTIFY_FROM(TTY_PROC_NR)/* function key press */
 03662	
 03663	/* Shorthands for message parameters passed with notifications. */
 03664	#define NOTIFY_SOURCE           m_source
 03665	#define NOTIFY_TYPE             m_type
 03666	#define NOTIFY_ARG              m2_l1
 03667	#define NOTIFY_TIMESTAMP        m2_l2
 03668	#define NOTIFY_FLAGS            m2_i1
 03669	
 03670	/*===========================================================================*
 03671	 *                Messages for BLOCK and CHARACTER device drivers            *
 03672	 *===========================================================================*/
 03673	
 03674	/* Message types for device drivers. */
 03675	#define DEV_RQ_BASE   0x400     /* base for device request types */
 03676	#define DEV_RS_BASE   0x500     /* base for device response types */
 03677	
 03678	#define CANCEL          (DEV_RQ_BASE +  0) /* general req to force a task to cancel */
 03679	#define DEV_READ        (DEV_RQ_BASE +  3) /* read from minor device */
 03680	#define DEV_WRITE       (DEV_RQ_BASE +  4) /* write to minor device */
 03681	#define DEV_IOCTL       (DEV_RQ_BASE +  5) /* I/O control code */
 03682	#define DEV_OPEN        (DEV_RQ_BASE +  6) /* open a minor device */
 03683	#define DEV_CLOSE       (DEV_RQ_BASE +  7) /* close a minor device */
 03684	#define DEV_SCATTER     (DEV_RQ_BASE +  8) /* write from a vector */
 03685	#define DEV_GATHER      (DEV_RQ_BASE +  9) /* read into a vector */
 03686	#define TTY_SETPGRP     (DEV_RQ_BASE + 10) /* set process group */
 03687	#define TTY_EXIT        (DEV_RQ_BASE + 11) /* process group leader exited */    
 03688	#define DEV_SELECT      (DEV_RQ_BASE + 12) /* request select() attention */
 03689	#define DEV_STATUS      (DEV_RQ_BASE + 13) /* request driver status */
 03690	
 03691	#define DEV_REPLY       (DEV_RS_BASE + 0) /* general task reply */
 03692	#define DEV_CLONED      (DEV_RS_BASE + 1) /* return cloned minor */
 03693	#define DEV_REVIVE      (DEV_RS_BASE + 2) /* driver revives process */
 03694	#define DEV_IO_READY    (DEV_RS_BASE + 3) /* selected device ready */
 03695	#define DEV_NO_STATUS   (DEV_RS_BASE + 4) /* empty status reply */
 03696	
 03697	/* Field names for messages to block and character device drivers. */
 03698	#define DEVICE          m2_i1   /* major-minor device */
 03699	#define PROC_NR         m2_i2   /* which (proc) wants I/O? */
 03700	#define COUNT           m2_i3   /* how many bytes to transfer */
 03701	#define REQUEST         m2_i3   /* ioctl request code */
 03702	#define POSITION        m2_l1   /* file offset */
 03703	#define ADDRESS         m2_p1   /* core buffer address */
 03704	

_________________________ Page 680    File: include/minix/com.h _________________________
 03705	/* Field names for DEV_SELECT messages to device drivers. */
 03706	#define DEV_MINOR       m2_i1   /* minor device */
 03707	#define DEV_SEL_OPS     m2_i2   /* which select operations are requested */
 03708	#define DEV_SEL_WATCH   m2_i3   /* request notify if no operations are ready */
 03709	
 03710	/* Field names used in reply messages from tasks. */
 03711	#define REP_PROC_NR     m2_i1   /* # of proc on whose behalf I/O was done */
 03712	#define REP_STATUS      m2_i2   /* bytes transferred or error number */
 03713	#  define SUSPEND        -998   /* status to suspend caller, reply later */
 03714	
 03715	/* Field names for messages to TTY driver. */
 03716	#define TTY_LINE        DEVICE  /* message parameter: terminal line */
 03717	#define TTY_REQUEST     COUNT   /* message parameter: ioctl request code */
 03718	#define TTY_SPEK        POSITION/* message parameter: ioctl speed, erasing */
 03719	#define TTY_FLAGS       m2_l2   /* message parameter: ioctl tty mode */
 03720	#define TTY_PGRP        m2_i3   /* message parameter: process group */  
 03721	
 03722	/* Field names for the QIC 02 status reply from tape driver */
 03723	#define TAPE_STAT0      m2_l1
 03724	#define TAPE_STAT1      m2_l2
 03725	
 03726	/*===========================================================================*
 03727	 *                         Messages for networking layer                     *
 03728	 *===========================================================================*/
 03729	
 03730	/* Message types for network layer requests. This layer acts like a driver. */
 03731	#define NW_OPEN         DEV_OPEN
 03732	#define NW_CLOSE        DEV_CLOSE
 03733	#define NW_READ         DEV_READ
 03734	#define NW_WRITE        DEV_WRITE
 03735	#define NW_IOCTL        DEV_IOCTL
 03736	#define NW_CANCEL       CANCEL
 03737	
 03738	/* Base type for data link layer requests and responses. */
 03739	#define DL_RQ_BASE      0x800           
 03740	#define DL_RS_BASE      0x900           
 03741	
 03742	/* Message types for data link layer requests. */
 03743	#define DL_WRITE        (DL_RQ_BASE + 3)
 03744	#define DL_WRITEV       (DL_RQ_BASE + 4)
 03745	#define DL_READ         (DL_RQ_BASE + 5)
 03746	#define DL_READV        (DL_RQ_BASE + 6)
 03747	#define DL_INIT         (DL_RQ_BASE + 7)
 03748	#define DL_STOP         (DL_RQ_BASE + 8)
 03749	#define DL_GETSTAT      (DL_RQ_BASE + 9)
 03750	
 03751	/* Message type for data link layer replies. */
 03752	#define DL_INIT_REPLY   (DL_RS_BASE + 20)
 03753	#define DL_TASK_REPLY   (DL_RS_BASE + 21)
 03754	
 03755	/* Field names for data link layer messages. */
 03756	#define DL_PORT         m2_i1
 03757	#define DL_PROC         m2_i2
 03758	#define DL_COUNT        m2_i3
 03759	#define DL_MODE         m2_l1
 03760	#define DL_CLCK         m2_l2
 03761	#define DL_ADDR         m2_p1
 03762	#define DL_STAT         m2_l1
 03763	
 03764	/* Bits in 'DL_STAT' field of DL replies. */

_________________________ Page 681    File: include/minix/com.h _________________________
 03765	#  define DL_PACK_SEND          0x01
 03766	#  define DL_PACK_RECV          0x02
 03767	#  define DL_READ_IP            0x04
 03768	
 03769	/* Bits in 'DL_MODE' field of DL requests. */
 03770	#  define DL_NOMODE             0x0
 03771	#  define DL_PROMISC_REQ        0x2
 03772	#  define DL_MULTI_REQ          0x4
 03773	#  define DL_BROAD_REQ          0x8
 03774	
 03775	/*===========================================================================*
 03776	 *                  SYSTASK request types and field names                    *
 03777	 *===========================================================================*/
 03778	
 03779	/* System library calls are dispatched via a call vector, so be careful when 
 03780	 * modifying the system call numbers. The numbers here determine which call
 03781	 * is made from the call vector.
 03782	 */ 
 03783	#define KERNEL_CALL     0x600   /* base for kernel calls to SYSTEM */ 
 03784	
 03785	#  define SYS_FORK       (KERNEL_CALL + 0)      /* sys_fork() */
 03786	#  define SYS_EXEC       (KERNEL_CALL + 1)      /* sys_exec() */
 03787	#  define SYS_EXIT       (KERNEL_CALL + 2)      /* sys_exit() */
 03788	#  define SYS_NICE       (KERNEL_CALL + 3)      /* sys_nice() */
 03789	#  define SYS_PRIVCTL    (KERNEL_CALL + 4)      /* sys_privctl() */
 03790	#  define SYS_TRACE      (KERNEL_CALL + 5)      /* sys_trace() */
 03791	#  define SYS_KILL       (KERNEL_CALL + 6)      /* sys_kill() */
 03792	
 03793	#  define SYS_GETKSIG    (KERNEL_CALL + 7)      /* sys_getsig() */
 03794	#  define SYS_ENDKSIG    (KERNEL_CALL + 8)      /* sys_endsig() */
 03795	#  define SYS_SIGSEND    (KERNEL_CALL + 9)      /* sys_sigsend() */
 03796	#  define SYS_SIGRETURN  (KERNEL_CALL + 10)     /* sys_sigreturn() */
 03797	
 03798	#  define SYS_NEWMAP     (KERNEL_CALL + 11)     /* sys_newmap() */
 03799	#  define SYS_SEGCTL     (KERNEL_CALL + 12)     /* sys_segctl() */
 03800	#  define SYS_MEMSET     (KERNEL_CALL + 13)     /* sys_memset() */
 03801	
 03802	#  define SYS_UMAP       (KERNEL_CALL + 14)     /* sys_umap() */
 03803	#  define SYS_VIRCOPY    (KERNEL_CALL + 15)     /* sys_vircopy() */
 03804	#  define SYS_PHYSCOPY   (KERNEL_CALL + 16)     /* sys_physcopy() */
 03805	#  define SYS_VIRVCOPY   (KERNEL_CALL + 17)     /* sys_virvcopy() */
 03806	#  define SYS_PHYSVCOPY  (KERNEL_CALL + 18)     /* sys_physvcopy() */
 03807	
 03808	#  define SYS_IRQCTL     (KERNEL_CALL + 19)     /* sys_irqctl() */
 03809	#  define SYS_INT86      (KERNEL_CALL + 20)     /* sys_int86() */
 03810	#  define SYS_DEVIO      (KERNEL_CALL + 21)     /* sys_devio() */
 03811	#  define SYS_SDEVIO     (KERNEL_CALL + 22)     /* sys_sdevio() */
 03812	#  define SYS_VDEVIO     (KERNEL_CALL + 23)     /* sys_vdevio() */
 03813	
 03814	#  define SYS_SETALARM   (KERNEL_CALL + 24)     /* sys_setalarm() */
 03815	#  define SYS_TIMES      (KERNEL_CALL + 25)     /* sys_times() */
 03816	#  define SYS_GETINFO    (KERNEL_CALL + 26)     /* sys_getinfo() */
 03817	#  define SYS_ABORT      (KERNEL_CALL + 27)     /* sys_abort() */
 03818	
 03819	#define NR_SYS_CALLS    28      /* number of system calls */ 
 03820	
 03821	/* Field names for SYS_MEMSET, SYS_SEGCTL. */
 03822	#define MEM_PTR         m2_p1   /* base */
 03823	#define MEM_COUNT       m2_l1   /* count */
 03824	#define MEM_PATTERN     m2_l2   /* pattern to write */

_________________________ Page 682    File: include/minix/com.h _________________________
 03825	#define MEM_CHUNK_BASE  m4_l1   /* physical base address */
 03826	#define MEM_CHUNK_SIZE  m4_l2   /* size of mem chunk */
 03827	#define MEM_TOT_SIZE    m4_l3   /* total memory size */
 03828	#define MEM_CHUNK_TAG   m4_l4   /* tag to identify chunk of mem */
 03829	
 03830	/* Field names for SYS_DEVIO, SYS_VDEVIO, SYS_SDEVIO. */
 03831	#define DIO_REQUEST     m2_i3   /* device in or output */
 03832	#   define DIO_INPUT        0   /* input */
 03833	#   define DIO_OUTPUT       1   /* output */
 03834	#define DIO_TYPE        m2_i1   /* flag indicating byte, word, or long */ 
 03835	#   define DIO_BYTE       'b'   /* byte type values */
 03836	#   define DIO_WORD       'w'   /* word type values */
 03837	#   define DIO_LONG       'l'   /* long type values */
 03838	#define DIO_PORT        m2_l1   /* single port address */
 03839	#define DIO_VALUE       m2_l2   /* single I/O value */
 03840	#define DIO_VEC_ADDR    m2_p1   /* address of buffer or (p,v)-pairs */
 03841	#define DIO_VEC_SIZE    m2_l2   /* number of elements in vector */
 03842	#define DIO_VEC_PROC    m2_i2   /* number of process where vector is */
 03843	
 03844	/* Field names for SYS_SIGNARLM, SYS_FLAGARLM, SYS_SYNCALRM. */
 03845	#define ALRM_EXP_TIME   m2_l1   /* expire time for the alarm call */
 03846	#define ALRM_ABS_TIME   m2_i2   /* set to 1 to use absolute alarm time */
 03847	#define ALRM_TIME_LEFT  m2_l1   /* how many ticks were remaining */
 03848	#define ALRM_PROC_NR    m2_i1   /* which process wants the alarm? */
 03849	#define ALRM_FLAG_PTR   m2_p1   /* virtual address of timeout flag */   
 03850	
 03851	/* Field names for SYS_IRQCTL. */
 03852	#define IRQ_REQUEST     m5_c1   /* what to do? */
 03853	#  define IRQ_SETPOLICY     1   /* manage a slot of the IRQ table */
 03854	#  define IRQ_RMPOLICY      2   /* remove a slot of the IRQ table */
 03855	#  define IRQ_ENABLE        3   /* enable interrupts */
 03856	#  define IRQ_DISABLE       4   /* disable interrupts */
 03857	#define IRQ_VECTOR      m5_c2   /* irq vector */
 03858	#define IRQ_POLICY      m5_i1   /* options for IRQCTL request */
 03859	#  define IRQ_REENABLE  0x001   /* reenable IRQ line after interrupt */
 03860	#  define IRQ_BYTE      0x100   /* byte values */      
 03861	#  define IRQ_WORD      0x200   /* word values */
 03862	#  define IRQ_LONG      0x400   /* long values */
 03863	#define IRQ_PROC_NR     m5_i2   /* process number, SELF, NONE */
 03864	#define IRQ_HOOK_ID     m5_l3   /* id of irq hook at kernel */
 03865	
 03866	/* Field names for SYS_SEGCTL. */
 03867	#define SEG_SELECT      m4_l1   /* segment selector returned */ 
 03868	#define SEG_OFFSET      m4_l2   /* offset in segment returned */
 03869	#define SEG_PHYS        m4_l3   /* physical address of segment */
 03870	#define SEG_SIZE        m4_l4   /* segment size */
 03871	#define SEG_INDEX       m4_l5   /* segment index in remote map */
 03872	
 03873	/* Field names for SYS_VIDCOPY. */
 03874	#define VID_REQUEST     m4_l1   /* what to do? */
 03875	#  define VID_VID_COPY     1    /* request vid_vid_copy() */
 03876	#  define MEM_VID_COPY     2    /* request mem_vid_copy() */
 03877	#define VID_SRC_ADDR    m4_l2   /* virtual address in memory */
 03878	#define VID_SRC_OFFSET  m4_l3   /* offset in video memory */
 03879	#define VID_DST_OFFSET  m4_l4   /* offset in video memory */
 03880	#define VID_CP_COUNT    m4_l5   /* number of words to be copied */
 03881	
 03882	/* Field names for SYS_ABORT. */
 03883	#define ABRT_HOW        m1_i1   /* RBT_REBOOT, RBT_HALT, etc. */
 03884	#define ABRT_MON_PROC   m1_i2   /* process where monitor params are */

_________________________ Page 683    File: include/minix/com.h _________________________
 03885	#define ABRT_MON_LEN    m1_i3   /* length of monitor params */
 03886	#define ABRT_MON_ADDR   m1_p1   /* virtual address of monitor params */
 03887	
 03888	/* Field names for _UMAP, _VIRCOPY, _PHYSCOPY. */
 03889	#define CP_SRC_SPACE    m5_c1   /* T or D space (stack is also D) */
 03890	#define CP_SRC_PROC_NR  m5_i1   /* process to copy from */
 03891	#define CP_SRC_ADDR     m5_l1   /* address where data come from */
 03892	#define CP_DST_SPACE    m5_c2   /* T or D space (stack is also D) */
 03893	#define CP_DST_PROC_NR  m5_i2   /* process to copy to */
 03894	#define CP_DST_ADDR     m5_l2   /* address where data go to */
 03895	#define CP_NR_BYTES     m5_l3   /* number of bytes to copy */
 03896	
 03897	/* Field names for SYS_VCOPY and SYS_VVIRCOPY. */
 03898	#define VCP_NR_OK       m1_i2   /* number of successfull copies */
 03899	#define VCP_VEC_SIZE    m1_i3   /* size of copy vector */
 03900	#define VCP_VEC_ADDR    m1_p1   /* pointer to copy vector */
 03901	
 03902	/* Field names for SYS_GETINFO. */
 03903	#define I_REQUEST      m7_i3    /* what info to get */
 03904	#   define GET_KINFO       0    /* get kernel information structure */
 03905	#   define GET_IMAGE       1    /* get system image table */
 03906	#   define GET_PROCTAB     2    /* get kernel process table */
 03907	#   define GET_RANDOMNESS  3    /* get randomness buffer */
 03908	#   define GET_MONPARAMS   4    /* get monitor parameters */
 03909	#   define GET_KENV        5    /* get kernel environment string */
 03910	#   define GET_IRQHOOKS    6    /* get the IRQ table */
 03911	#   define GET_KMESSAGES   7    /* get kernel messages */
 03912	#   define GET_PRIVTAB     8    /* get kernel privileges table */
 03913	#   define GET_KADDRESSES  9    /* get various kernel addresses */
 03914	#   define GET_SCHEDINFO  10    /* get scheduling queues */
 03915	#   define GET_PROC       11    /* get process slot if given process */
 03916	#   define GET_MACHINE    12    /* get machine information */
 03917	#   define GET_LOCKTIMING 13    /* get lock()/unlock() latency timing */
 03918	#   define GET_BIOSBUFFER 14    /* get a buffer for BIOS calls */
 03919	#define I_PROC_NR      m7_i4    /* calling process */
 03920	#define I_VAL_PTR      m7_p1    /* virtual address at caller */ 
 03921	#define I_VAL_LEN      m7_i1    /* max length of value */
 03922	#define I_VAL_PTR2     m7_p2    /* second virtual address */ 
 03923	#define I_VAL_LEN2     m7_i2    /* second length, or proc nr */
 03924	
 03925	/* Field names for SYS_TIMES. */
 03926	#define T_PROC_NR      m4_l1    /* process to request time info for */
 03927	#define T_USER_TIME    m4_l1    /* user time consumed by process */
 03928	#define T_SYSTEM_TIME  m4_l2    /* system time consumed by process */
 03929	#define T_CHILD_UTIME  m4_l3    /* user time consumed by process' children */
 03930	#define T_CHILD_STIME  m4_l4    /* sys time consumed by process' children */
 03931	#define T_BOOT_TICKS   m4_l5    /* number of clock ticks since boot time */
 03932	
 03933	/* Field names for SYS_TRACE, SYS_SVRCTL. */
 03934	#define CTL_PROC_NR    m2_i1    /* process number of the caller */
 03935	#define CTL_REQUEST    m2_i2    /* server control request */
 03936	#define CTL_MM_PRIV    m2_i3    /* privilege as seen by PM */
 03937	#define CTL_ARG_PTR    m2_p1    /* pointer to argument */
 03938	#define CTL_ADDRESS    m2_l1    /* address at traced process' space */
 03939	#define CTL_DATA       m2_l2    /* data field for tracing */
 03940	
 03941	/* Field names for SYS_KILL, SYS_SIGCTL */
 03942	#define SIG_REQUEST    m2_l2    /* PM signal control request */
 03943	#define S_GETSIG           0    /* get pending kernel signal */
 03944	#define S_ENDSIG           1    /* finish a kernel signal */

_________________________ Page 684    File: include/minix/com.h _________________________
 03945	#define S_SENDSIG          2    /* POSIX style signal handling */
 03946	#define S_SIGRETURN        3    /* return from POSIX handling */
 03947	#define S_KILL             4    /* servers kills process with signal */
 03948	#define SIG_PROC       m2_i1    /* process number for inform */
 03949	#define SIG_NUMBER     m2_i2    /* signal number to send */
 03950	#define SIG_FLAGS      m2_i3    /* signal flags field */
 03951	#define SIG_MAP        m2_l1    /* used by kernel to pass signal bit map */
 03952	#define SIG_CTXT_PTR   m2_p1    /* pointer to info to restore signal context */
 03953	
 03954	/* Field names for SYS_FORK, _EXEC, _EXIT, _NEWMAP. */
 03955	#define PR_PROC_NR     m1_i1    /* indicates a (child) process */
 03956	#define PR_PRIORITY    m1_i2    /* process priority */
 03957	#define PR_PPROC_NR    m1_i2    /* indicates a (parent) process */
 03958	#define PR_PID         m1_i3    /* process id at process manager */
 03959	#define PR_STACK_PTR   m1_p1    /* used for stack ptr in sys_exec, sys_getsp */
 03960	#define PR_TRACING     m1_i3    /* flag to indicate tracing is on/ off */
 03961	#define PR_NAME_PTR    m1_p2    /* tells where program name is for dmp */
 03962	#define PR_IP_PTR      m1_p3    /* initial value for ip after exec */
 03963	#define PR_MEM_PTR     m1_p1    /* tells where memory map is for sys_newmap */
 03964	
 03965	/* Field names for SYS_INT86 */
 03966	#define INT86_REG86    m1_p1    /* pointer to registers */
 03967	
 03968	/* Field names for SELECT (FS). */
 03969	#define SEL_NFDS       m8_i1
 03970	#define SEL_READFDS    m8_p1
 03971	#define SEL_WRITEFDS   m8_p2
 03972	#define SEL_ERRORFDS   m8_p3
 03973	#define SEL_TIMEOUT    m8_p4
 03974	
 03975	/*===========================================================================*
 03976	 *                Messages for system management server                      *
 03977	 *===========================================================================*/
 03978	
 03979	#define SRV_RQ_BASE             0x700
 03980	
 03981	#define SRV_UP          (SRV_RQ_BASE + 0)       /* start system service */
 03982	#define SRV_DOWN        (SRV_RQ_BASE + 1)       /* stop system service */
 03983	#define SRV_STATUS      (SRV_RQ_BASE + 2)       /* get service status */
 03984	
 03985	#  define SRV_PATH_ADDR         m1_p1           /* path of binary */
 03986	#  define SRV_PATH_LEN          m1_i1           /* length of binary */
 03987	#  define SRV_ARGS_ADDR         m1_p2           /* arguments to be passed */
 03988	#  define SRV_ARGS_LEN          m1_i2           /* length of arguments */
 03989	#  define SRV_DEV_MAJOR         m1_i3           /* major device number */
 03990	#  define SRV_PRIV_ADDR         m1_p3           /* privileges string */
 03991	#  define SRV_PRIV_LEN          m1_i3           /* length of privileges */
 03992	
 03993	/*===========================================================================*
 03994	 *                Miscellaneous messages used by TTY                         *
 03995	 *===========================================================================*/
 03996	
 03997	/* Miscellaneous request types and field names, e.g. used by IS server. */
 03998	#define PANIC_DUMPS             97      /* debug dumps at the TTY on RBT_PANIC */
 03999	#define FKEY_CONTROL            98      /* control a function key at the TTY */
 04000	#  define FKEY_REQUEST       m2_i1      /* request to perform at TTY */
 04001	#  define    FKEY_MAP           10      /* observe function key */
 04002	#  define    FKEY_UNMAP         11      /* stop observing function key */
 04003	#  define    FKEY_EVENTS        12      /* request open key presses */
 04004	#  define FKEY_FKEYS          m2_l1     /* F1-F12 keys pressed */

_________________________ Page 685    File: include/minix/com.h _________________________
 04005	#  define FKEY_SFKEYS         m2_l2     /* Shift-F1-F12 keys pressed */
 04006	#define DIAGNOSTICS     100     /* output a string without FS in between */
 04007	#  define DIAG_PRINT_BUF      m1_p1
 04008	#  define DIAG_BUF_COUNT      m1_i1
 04009	#  define DIAG_PROC_NR        m1_i2
 04010	
 04011	#endif /* _MINIX_COM_H */ 




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/devio.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04100	/* This file provides basic types and some constants for the 
 04101	 * SYS_DEVIO and SYS_VDEVIO system calls, which allow user-level 
 04102	 * processes to perform device I/O. 
 04103	 *
 04104	 * Created: 
 04105	 *      Apr 08, 2004 by Jorrit N. Herder
 04106	 */
 04107	
 04108	#ifndef _DEVIO_H
 04109	#define _DEVIO_H
 04110	
 04111	#include <minix/sys_config.h>     /* needed to include <minix/type.h> */
 04112	#include <sys/types.h>        /* u8_t, u16_t, u32_t needed */
 04113	
 04114	typedef u16_t port_t;
 04115	typedef U16_t Port_t;
 04116	
 04117	/* We have different granularities of port I/O: 8, 16, 32 bits.
 04118	 * Also see <ibm/portio.h>, which has functions for bytes, words,  
 04119	 * and longs. Hence, we need different (port,value)-pair types. 
 04120	 */
 04121	typedef struct { u16_t port;  u8_t value; } pvb_pair_t;
 04122	typedef struct { u16_t port; u16_t value; } pvw_pair_t;
 04123	typedef struct { u16_t port; u32_t value; } pvl_pair_t;
 04124	
 04125	/* Macro shorthand to set (port,value)-pair. */
 04126	#define pv_set(pv, p, v) ((pv).port = (p), (pv).value = (v))
 04127	#define pv_ptr_set(pv_ptr, p, v) ((pv_ptr)->port = (p), (pv_ptr)->value = (v))
 04128	
 04129	#endif  /* _DEVIO_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/minix/dmap.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04200	#ifndef _DMAP_H
 04201	#define _DMAP_H
 04202	
 04203	#include <minix/sys_config.h>
 04204	#include <minix/ipc.h>
 04205	

_________________________ Page 686    File: include/minix/dmap.h _________________________
 04206	/*===========================================================================*
 04207	 *                       Device <-> Driver Table                             *
 04208	 *===========================================================================*/
 04209	
 04210	/* Device table.  This table is indexed by major device number.  It provides
 04211	 * the link between major device numbers and the routines that process them.
 04212	 * The table can be update dynamically. The field 'dmap_flags' describe an 
 04213	 * entry's current status and determines what control options are possible. 
 04214	 */
 04215	#define DMAP_MUTABLE            0x01    /* mapping can be overtaken */
 04216	#define DMAP_BUSY               0x02    /* driver busy with request */
 04217	
 04218	enum dev_style { STYLE_DEV, STYLE_NDEV, STYLE_TTY, STYLE_CLONE };
 04219	
 04220	extern struct dmap {
 04221	  int _PROTOTYPE ((*dmap_opcl), (int, Dev_t, int, int) );
 04222	  void _PROTOTYPE ((*dmap_io), (int, message *) );
 04223	  int dmap_driver;
 04224	  int dmap_flags;
 04225	} dmap[];
 04226	
 04227	/*===========================================================================*
 04228	 *                       Major and minor device numbers                      *
 04229	 *===========================================================================*/
 04230	
 04231	/* Total number of different devices. */
 04232	#define NR_DEVICES                32                    /* number of (major) device
 04233	
 04234	/* Major and minor device numbers for MEMORY driver. */
 04235	#define MEMORY_MAJOR               1    /* major device for memory devices */
 04236	#  define RAM_DEV                  0    /* minor device for /dev/ram */
 04237	#  define MEM_DEV                  1    /* minor device for /dev/mem */
 04238	#  define KMEM_DEV                 2    /* minor device for /dev/kmem */
 04239	#  define NULL_DEV                 3    /* minor device for /dev/null */
 04240	#  define BOOT_DEV                 4    /* minor device for /dev/boot */
 04241	#  define ZERO_DEV                 5    /* minor device for /dev/zero */
 04242	
 04243	#define CTRLR(n) ((n)==0 ? 3 : (8 + 2*((n)-1))) /* magic formula */
 04244	
 04245	/* Full device numbers that are special to the boot monitor and FS. */
 04246	#  define DEV_RAM             0x0100    /* device number of /dev/ram */
 04247	#  define DEV_BOOT            0x0104    /* device number of /dev/boot */
 04248	
 04249	#define FLOPPY_MAJOR               2    /* major device for floppy disks */
 04250	#define TTY_MAJOR                  4    /* major device for ttys */
 04251	#define CTTY_MAJOR                 5    /* major device for /dev/tty */
 04252	
 04253	#define INET_MAJOR                 7    /* major device for inet */
 04254	
 04255	#define LOG_MAJOR                 15    /* major device for log driver */
 04256	#  define IS_KLOG_DEV              0    /* minor device for /dev/klog */
 04257	
 04258	#endif /* _DMAP_H */

_________________________ Page 687    File: include/minix/dmap.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/ibm/portio.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04300	/*
 04301	ibm/portio.h
 04302	
 04303	Created:        Jan 15, 1992 by Philip Homburg
 04304	*/
 04305	
 04306	#ifndef _PORTIO_H_
 04307	#define _PORTIO_H_
 04308	
 04309	#ifndef _TYPES_H
 04310	#include <sys/types.h>
 04311	#endif
 04312	
 04313	unsigned inb(U16_t _port);
 04314	unsigned inw(U16_t _port);
 04315	unsigned inl(U32_t _port);
 04316	void outb(U16_t _port, U8_t _value);
 04317	void outw(U16_t _port, U16_t _value);
 04318	void outl(U16_t _port, U32_t _value);
 04319	void insb(U16_t _port, void *_buf, size_t _count);
 04320	void insw(U16_t _port, void *_buf, size_t _count);
 04321	void insl(U16_t _port, void *_buf, size_t _count);
 04322	void outsb(U16_t _port, void *_buf, size_t _count);
 04323	void outsw(U16_t _port, void *_buf, size_t _count);
 04324	void outsl(U16_t _port, void *_buf, size_t _count);
 04325	void intr_disable(void);
 04326	void intr_enable(void);
 04327	
 04328	#endif /* _PORTIO_H_ */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/ibm/interrupt.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04400	/* Interrupt numbers and hardware vectors. */
 04401	
 04402	#ifndef _INTERRUPT_H
 04403	#define _INTERRUPT_H
 04404	
 04405	#if (CHIP == INTEL)
 04406	
 04407	/* 8259A interrupt controller ports. */
 04408	#define INT_CTL         0x20    /* I/O port for interrupt controller */
 04409	#define INT_CTLMASK     0x21    /* setting bits in this port disables ints */
 04410	#define INT2_CTL        0xA0    /* I/O port for second interrupt controller */
 04411	#define INT2_CTLMASK    0xA1    /* setting bits in this port disables ints */
 04412	
 04413	/* Magic numbers for interrupt controller. */
 04414	#define END_OF_INT      0x20    /* code used to re-enable after an interrupt */
 04415	
 04416	/* Interrupt vectors defined/reserved by processor. */
 04417	#define DIVIDE_VECTOR      0    /* divide error */
 04418	#define DEBUG_VECTOR       1    /* single step (trace) */
 04419	#define NMI_VECTOR         2    /* non-maskable interrupt */

_________________________ Page 688    File: include/ibm/interrupt.h ______________________
 04420	#define BREAKPOINT_VECTOR  3    /* software breakpoint */
 04421	#define OVERFLOW_VECTOR    4    /* from INTO */
 04422	
 04423	/* Fixed system call vector. */
 04424	#define SYS_VECTOR        32    /* system calls are made with int SYSVEC */
 04425	#define SYS386_VECTOR     33    /* except 386 system calls use this */
 04426	#define LEVEL0_VECTOR     34    /* for execution of a function at level 0 */
 04427	
 04428	/* Suitable irq bases for hardware interrupts.  Reprogram the 8259(s) from
 04429	 * the PC BIOS defaults since the BIOS doesn't respect all the processor's
 04430	 * reserved vectors (0 to 31).
 04431	 */
 04432	#define BIOS_IRQ0_VEC   0x08    /* base of IRQ0-7 vectors used by BIOS */
 04433	#define BIOS_IRQ8_VEC   0x70    /* base of IRQ8-15 vectors used by BIOS */
 04434	#define IRQ0_VECTOR     0x50    /* nice vectors to relocate IRQ0-7 to */
 04435	#define IRQ8_VECTOR     0x70    /* no need to move IRQ8-15 */
 04436	
 04437	/* Hardware interrupt numbers. */
 04438	#define NR_IRQ_VECTORS    16
 04439	#define CLOCK_IRQ          0
 04440	#define KEYBOARD_IRQ       1
 04441	#define CASCADE_IRQ        2    /* cascade enable for 2nd AT controller */
 04442	#define ETHER_IRQ          3    /* default ethernet interrupt vector */
 04443	#define SECONDARY_IRQ      3    /* RS232 interrupt vector for port 2 */
 04444	#define RS232_IRQ          4    /* RS232 interrupt vector for port 1 */
 04445	#define XT_WINI_IRQ        5    /* xt winchester */
 04446	#define FLOPPY_IRQ         6    /* floppy disk */
 04447	#define PRINTER_IRQ        7
 04448	#define AT_WINI_0_IRQ     14    /* at winchester controller 0 */
 04449	#define AT_WINI_1_IRQ     15    /* at winchester controller 1 */
 04450	
 04451	/* Interrupt number to hardware vector. */
 04452	#define BIOS_VECTOR(irq)        \
 04453	        (((irq) < 8 ? BIOS_IRQ0_VEC : BIOS_IRQ8_VEC) + ((irq) & 0x07))
 04454	#define VECTOR(irq)     \
 04455	        (((irq) < 8 ? IRQ0_VECTOR : IRQ8_VECTOR) + ((irq) & 0x07))
 04456	
 04457	#endif /* (CHIP == INTEL) */
 04458	
 04459	#endif /* _INTERRUPT_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      include/ibm/ports.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04500	/* Addresses and magic numbers for miscellaneous ports. */
 04501	
 04502	#ifndef _PORTS_H
 04503	#define _PORTS_H
 04504	
 04505	#if (CHIP == INTEL)
 04506	
 04507	/* Miscellaneous ports. */
 04508	#define PCR             0x65    /* Planar Control Register */
 04509	#define PORT_B          0x61    /* I/O port for 8255 port B (kbd, beeper...) */
 04510	#define TIMER0          0x40    /* I/O port for timer channel 0 */
 04511	#define TIMER2          0x42    /* I/O port for timer channel 2 */
 04512	#define TIMER_MODE      0x43    /* I/O port for timer mode control */
 04513	
 04514	#endif /* (CHIP == INTEL) */

_________________________ Page 689    File: include/ibm/ports.h _________________________
 04515	
 04516	#endif /* _PORTS_H */




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/kernel.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04600	#ifndef KERNEL_H
 04601	#define KERNEL_H
 04602	
 04603	/* This is the master header for the kernel.  It includes some other files
 04604	 * and defines the principal constants.
 04605	 */
 04606	#define _POSIX_SOURCE      1    /* tell headers to include POSIX stuff */
 04607	#define _MINIX             1    /* tell headers to include MINIX stuff */
 04608	#define _SYSTEM            1    /* tell headers that this is the kernel */
 04609	
 04610	/* The following are so basic, all the *.c files get them automatically. */
 04611	#include <minix/config.h>       /* global configuration, MUST be first */
 04612	#include <ansi.h>               /* C style: ANSI or K&R, MUST be second */
 04613	#include <sys/types.h>          /* general system types */
 04614	#include <minix/const.h>        /* MINIX specific constants */
 04615	#include <minix/type.h>         /* MINIX specific types, e.g. message */
 04616	#include <minix/ipc.h>          /* MINIX run-time system */
 04617	#include <timers.h>             /* watchdog timer management */
 04618	#include <errno.h>              /* return codes and error numbers */
 04619	#include <ibm/portio.h>         /* device I/O and toggle interrupts */ 
 04620	
 04621	/* Important kernel header files. */
 04622	#include "config.h"             /* configuration, MUST be first */
 04623	#include "const.h"              /* constants, MUST be second */
 04624	#include "type.h"               /* type definitions, MUST be third */
 04625	#include "proto.h"              /* function prototypes */
 04626	#include "glo.h"                /* global variables */
 04627	#include "ipc.h"                /* IPC constants */
 04628	/* #include "debug.h" */        /* debugging, MUST be last kernel header */
 04629	
 04630	#endif /* KERNEL_H */
 04631	




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/config.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04700	#ifndef CONFIG_H
 04701	#define CONFIG_H
 04702	
 04703	/* This file defines the kernel configuration. It allows to set sizes of some
 04704	 * kernel buffers and to enable or disable debugging code, timing features, 
 04705	 * and individual kernel calls.
 04706	 *
 04707	 * Changes:
 04708	 *   Jul 11, 2005       Created.  (Jorrit N. Herder)
 04709	 */

_________________________ Page 690    File: kernel/config.h _________________________
 04710	
 04711	/* In embedded and sensor applications, not all the kernel calls may be
 04712	 * needed. In this section you can specify which kernel calls are needed
 04713	 * and which are not. The code for unneeded kernel calls is not included in
 04714	 * the system binary, making it smaller. If you are not sure, it is best
 04715	 * to keep all kernel calls enabled.
 04716	 */
 04717	#define USE_FORK           1    /* fork a new process */
 04718	#define USE_NEWMAP         1    /* set a new memory map */
 04719	#define USE_EXEC           1    /* update process after execute */
 04720	#define USE_EXIT           1    /* clean up after process exit */
 04721	#define USE_TRACE          1    /* process information and tracing */
 04722	#define USE_GETKSIG        1    /* retrieve pending kernel signals */
 04723	#define USE_ENDKSIG        1    /* finish pending kernel signals */
 04724	#define USE_KILL           1    /* send a signal to a process */
 04725	#define USE_SIGSEND        1    /* send POSIX-style signal */
 04726	#define USE_SIGRETURN      1    /* sys_sigreturn(proc_nr, ctxt_ptr, flags) */
 04727	#define USE_ABORT          1    /* shut down MINIX */
 04728	#define USE_GETINFO        1    /* retrieve a copy of kernel data */
 04729	#define USE_TIMES          1    /* get process and system time info */
 04730	#define USE_SETALARM       1    /* schedule a synchronous alarm */
 04731	#define USE_DEVIO          1    /* read or write a single I/O port */
 04732	#define USE_VDEVIO         1    /* process vector with I/O requests */
 04733	#define USE_SDEVIO         1    /* perform I/O request on a buffer */
 04734	#define USE_IRQCTL         1    /* set an interrupt policy */
 04735	#define USE_SEGCTL         1    /* set up a remote segment */
 04736	#define USE_PRIVCTL        1    /* system privileges control */
 04737	#define USE_NICE           1    /* change scheduling priority */
 04738	#define USE_UMAP           1    /* map virtual to physical address */
 04739	#define USE_VIRCOPY        1    /* copy using virtual addressing */ 
 04740	#define USE_VIRVCOPY       1    /* vector with virtual copy requests */
 04741	#define USE_PHYSCOPY       1    /* copy using physical addressing */
 04742	#define USE_PHYSVCOPY      1    /* vector with physical copy requests */
 04743	#define USE_MEMSET         1    /* write char to a given memory area */
 04744	
 04745	/* Length of program names stored in the process table. This is only used
 04746	 * for the debugging dumps that can be generated with the IS server. The PM
 04747	 * server keeps its own copy of the program name.  
 04748	 */
 04749	#define P_NAME_LEN         8
 04750	
 04751	/* Kernel diagnostics are written to a circular buffer. After each message, 
 04752	 * a system server is notified and a copy of the buffer can be retrieved to 
 04753	 * display the message. The buffers size can safely be reduced.  
 04754	 */
 04755	#define KMESS_BUF_SIZE   256    
 04756	
 04757	/* Buffer to gather randomness. This is used to generate a random stream by 
 04758	 * the MEMORY driver when reading from /dev/random. 
 04759	 */
 04760	#define RANDOM_ELEMENTS   32
 04761	
 04762	/* This section contains defines for valuable system resources that are used
 04763	 * by device drivers. The number of elements of the vectors is determined by 
 04764	 * the maximum needed by any given driver. The number of interrupt hooks may
 04765	 * be incremented on systems with many device drivers. 
 04766	 */
 04767	#define NR_IRQ_HOOKS      16            /* number of interrupt hooks */
 04768	#define VDEVIO_BUF_SIZE   64            /* max elements per VDEVIO request */
 04769	#define VCOPY_VEC_SIZE    16            /* max elements per VCOPY request */

_________________________ Page 691    File: kernel/config.h _________________________
 04770	
 04771	/* How many bytes for the kernel stack. Space allocated in mpx.s. */
 04772	#define K_STACK_BYTES   1024    
 04773	
 04774	/* This section allows to enable kernel debugging and timing functionality.
 04775	 * For normal operation all options should be disabled.
 04776	 */
 04777	#define DEBUG_SCHED_CHECK  0    /* sanity check of scheduling queues */
 04778	#define DEBUG_LOCK_CHECK   0    /* kernel lock() sanity check */
 04779	#define DEBUG_TIME_LOCKS   0    /* measure time spent in locks */
 04780	
 04781	#endif /* CONFIG_H */
 04782	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/const.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04800	/* General macros and constants used by the kernel. */
 04801	#ifndef CONST_H
 04802	#define CONST_H
 04803	
 04804	#include <ibm/interrupt.h>      /* interrupt numbers and hardware vectors */
 04805	#include <ibm/ports.h>          /* port addresses and magic numbers */
 04806	#include <ibm/bios.h>           /* BIOS addresses, sizes and magic numbers */
 04807	#include <ibm/cpu.h>            /* BIOS addresses, sizes and magic numbers */
 04808	#include <minix/config.h>
 04809	#include "config.h"
 04810	
 04811	/* To translate an address in kernel space to a physical address.  This is
 04812	 * the same as umap_local(proc_ptr, D, vir, sizeof(*vir)), but less costly.
 04813	 */
 04814	#define vir2phys(vir)   (kinfo.data_base + (vir_bytes) (vir))
 04815	
 04816	/* Map a process number to a privilege structure id. */
 04817	#define s_nr_to_id(n)   (NR_TASKS + (n) + 1)
 04818	
 04819	/* Translate a pointer to a field in a structure to a pointer to the structure
 04820	 * itself. So it translates '&struct_ptr->field' back to 'struct_ptr'.
 04821	 */
 04822	#define structof(type, field, ptr) \
 04823	        ((type *) (((char *) (ptr)) - offsetof(type, field)))
 04824	
 04825	/* Constants used in virtual_copy(). Values must be 0 and 1, respectively. */
 04826	#define _SRC_   0
 04827	#define _DST_   1
 04828	
 04829	/* Number of random sources */
 04830	#define RANDOM_SOURCES  16
 04831	
 04832	/* Constants and macros for bit map manipulation. */
 04833	#define BITCHUNK_BITS   (sizeof(bitchunk_t) * CHAR_BIT)
 04834	#define BITMAP_CHUNKS(nr_bits) (((nr_bits)+BITCHUNK_BITS-1)/BITCHUNK_BITS)  
 04835	#define MAP_CHUNK(map,bit) (map)[((bit)/BITCHUNK_BITS)]
 04836	#define CHUNK_OFFSET(bit) ((bit)%BITCHUNK_BITS))
 04837	#define GET_BIT(map,bit) ( MAP_CHUNK(map,bit) & (1 << CHUNK_OFFSET(bit) )
 04838	#define SET_BIT(map,bit) ( MAP_CHUNK(map,bit) |= (1 << CHUNK_OFFSET(bit) )
 04839	#define UNSET_BIT(map,bit) ( MAP_CHUNK(map,bit) &= ~(1 << CHUNK_OFFSET(bit) )

_________________________ Page 692    File: kernel/const.h _________________________
 04840	
 04841	#define get_sys_bit(map,bit) \
 04842	        ( MAP_CHUNK(map.chunk,bit) & (1 << CHUNK_OFFSET(bit) )
 04843	#define set_sys_bit(map,bit) \
 04844	        ( MAP_CHUNK(map.chunk,bit) |= (1 << CHUNK_OFFSET(bit) )
 04845	#define unset_sys_bit(map,bit) \
 04846	        ( MAP_CHUNK(map.chunk,bit) &= ~(1 << CHUNK_OFFSET(bit) )
 04847	#define NR_SYS_CHUNKS   BITMAP_CHUNKS(NR_SYS_PROCS)
 04848	
 04849	/* Program stack words and masks. */
 04850	#define INIT_PSW      0x0200    /* initial psw */
 04851	#define INIT_TASK_PSW 0x1200    /* initial psw for tasks (with IOPL 1) */
 04852	#define TRACEBIT      0x0100    /* OR this with psw in proc[] for tracing */
 04853	#define SETPSW(rp, new)         /* permits only certain bits to be set */ \
 04854	        ((rp)->p_reg.psw = (rp)->p_reg.psw & ~0xCD5 | (new) & 0xCD5)
 04855	#define IF_MASK 0x00000200
 04856	#define IOPL_MASK 0x003000
 04857	
 04858	/* Disable/ enable hardware interrupts. The parameters of lock() and unlock()
 04859	 * are used when debugging is enabled. See debug.h for more information.
 04860	 */
 04861	#define lock(c, v)      intr_disable(); 
 04862	#define unlock(c)       intr_enable(); 
 04863	
 04864	/* Sizes of memory tables. The boot monitor distinguishes three memory areas, 
 04865	 * namely low mem below 1M, 1M-16M, and mem after 16M. More chunks are needed
 04866	 * for DOS MINIX.
 04867	 */
 04868	#define NR_MEMS            8    
 04869	
 04870	#endif /* CONST_H */
 04871	
 04872	
 04873	
 04874	
 04875	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/type.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 04900	#ifndef TYPE_H
 04901	#define TYPE_H
 04902	
 04903	typedef _PROTOTYPE( void task_t, (void) );
 04904	
 04905	/* Process table and system property related types. */ 
 04906	typedef int proc_nr_t;                  /* process table entry number */
 04907	typedef short sys_id_t;                 /* system process index */
 04908	typedef struct {                        /* bitmap for system indexes */
 04909	  bitchunk_t chunk[BITMAP_CHUNKS(NR_SYS_PROCS)];
 04910	} sys_map_t;
 04911	
 04912	struct boot_image {
 04913	  proc_nr_t proc_nr;                    /* process number to use */
 04914	  task_t *initial_pc;                   /* start function for tasks */

_________________________ Page 693    File: kernel/type.h _________________________
 04915	  int flags;                            /* process flags */
 04916	  unsigned char quantum;                /* quantum (tick count) */
 04917	  int priority;                         /* scheduling priority */
 04918	  int stksize;                          /* stack size for tasks */
 04919	  short trap_mask;                      /* allowed system call traps */
 04920	  bitchunk_t ipc_to;                    /* send mask protection */
 04921	  long call_mask;                       /* system call protection */
 04922	  char proc_name[P_NAME_LEN];           /* name in process table */
 04923	};
 04924	
 04925	struct memory {
 04926	  phys_clicks base;                     /* start address of chunk */
 04927	  phys_clicks size;                     /* size of memory chunk */
 04928	};
 04929	
 04930	/* The kernel outputs diagnostic messages in a circular buffer. */
 04931	struct kmessages {
 04932	  int km_next;                          /* next index to write */
 04933	  int km_size;                          /* current size in buffer */
 04934	  char km_buf[KMESS_BUF_SIZE];          /* buffer for messages */
 04935	};
 04936	
 04937	struct randomness {
 04938	  struct {
 04939	        int r_next;                             /* next index to write */
 04940	        int r_size;                             /* number of random elements */
 04941	        unsigned short r_buf[RANDOM_ELEMENTS]; /* buffer for random info */
 04942	  } bin[RANDOM_SOURCES];
 04943	};
 04944	
 04945	#if (CHIP == INTEL)
 04946	typedef unsigned reg_t;         /* machine register */
 04947	
 04948	/* The stack frame layout is determined by the software, but for efficiency
 04949	 * it is laid out so the assembly code to use it is as simple as possible.
 04950	 * 80286 protected mode and all real modes use the same frame, built with
 04951	 * 16-bit registers.  Real mode lacks an automatic stack switch, so little
 04952	 * is lost by using the 286 frame for it.  The 386 frame differs only in
 04953	 * having 32-bit registers and more segment registers.  The same names are
 04954	 * used for the larger registers to avoid differences in the code.
 04955	 */
 04956	struct stackframe_s {           /* proc_ptr points here */
 04957	#if _WORD_SIZE == 4
 04958	  u16_t gs;                     /* last item pushed by save */
 04959	  u16_t fs;                     /*  ^ */
 04960	#endif
 04961	  u16_t es;                     /*  | */
 04962	  u16_t ds;                     /*  | */
 04963	  reg_t di;                     /* di through cx are not accessed in C */
 04964	  reg_t si;                     /* order is to match pusha/popa */
 04965	  reg_t fp;                     /* bp */
 04966	  reg_t st;                     /* hole for another copy of sp */
 04967	  reg_t bx;                     /*  | */
 04968	  reg_t dx;                     /*  | */
 04969	  reg_t cx;                     /*  | */
 04970	  reg_t retreg;                 /* ax and above are all pushed by save */
 04971	  reg_t retadr;                 /* return address for assembly code save() */
 04972	  reg_t pc;                     /*  ^  last item pushed by interrupt */
 04973	  reg_t cs;                     /*  | */
 04974	  reg_t psw;                    /*  | */

_________________________ Page 694    File: kernel/type.h _________________________
 04975	  reg_t sp;                     /*  | */
 04976	  reg_t ss;                     /* these are pushed by CPU during interrupt */
 04977	};
 04978	
 04979	struct segdesc_s {              /* segment descriptor for protected mode */
 04980	  u16_t limit_low;
 04981	  u16_t base_low;
 04982	  u8_t base_middle;
 04983	  u8_t access;                  /* |P|DL|1|X|E|R|A| */
 04984	  u8_t granularity;             /* |G|X|0|A|LIMT| */
 04985	  u8_t base_high;
 04986	};
 04987	
 04988	typedef unsigned long irq_policy_t;     
 04989	typedef unsigned long irq_id_t; 
 04990	
 04991	typedef struct irq_hook {
 04992	  struct irq_hook *next;                /* next hook in chain */
 04993	  int (*handler)(struct irq_hook *);    /* interrupt handler */
 04994	  int irq;                              /* IRQ vector number */ 
 04995	  int id;                               /* id of this hook */
 04996	  int proc_nr;                          /* NONE if not in use */
 04997	  irq_id_t notify_id;                   /* id to return on interrupt */
 04998	  irq_policy_t policy;                  /* bit mask for policy */
 04999	} irq_hook_t;
 05000	
 05001	typedef int (*irq_handler_t)(struct irq_hook *);
 05002	
 05003	#endif /* (CHIP == INTEL) */
 05004	
 05005	#if (CHIP == M68000)
 05006	/* M68000 specific types go here. */
 05007	#endif /* (CHIP == M68000) */
 05008	
 05009	#endif /* TYPE_H */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/proto.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05100	/* Function prototypes. */
 05101	
 05102	#ifndef PROTO_H
 05103	#define PROTO_H
 05104	
 05105	/* Struct declarations. */
 05106	struct proc;
 05107	struct timer;
 05108	
 05109	/* clock.c */
 05110	_PROTOTYPE( void clock_task, (void)                                     );
 05111	_PROTOTYPE( void clock_stop, (void)                                     );
 05112	_PROTOTYPE( clock_t get_uptime, (void)                                  );
 05113	_PROTOTYPE( unsigned long read_clock, (void)                            );
 05114	_PROTOTYPE( void set_timer, (struct timer *tp, clock_t t, tmr_func_t f) );
 05115	_PROTOTYPE( void reset_timer, (struct timer *tp)                        );
 05116	
 05117	/* main.c */
 05118	_PROTOTYPE( void main, (void)                                           );
 05119	_PROTOTYPE( void prepare_shutdown, (int how)                            );

_________________________ Page 695    File: kernel/proto.h _________________________
 05120	
 05121	/* utility.c */
 05122	_PROTOTYPE( void kprintf, (const char *fmt, ...)                        );
 05123	_PROTOTYPE( void panic, (_CONST char *s, int n)                         );
 05124	
 05125	/* proc.c */
 05126	_PROTOTYPE( int sys_call, (int function, int src_dest, message *m_ptr)  );
 05127	_PROTOTYPE( int lock_notify, (int src, int dst)                         );
 05128	_PROTOTYPE( int lock_send, (int dst, message *m_ptr)                    );
 05129	_PROTOTYPE( void lock_enqueue, (struct proc *rp)                        );
 05130	_PROTOTYPE( void lock_dequeue, (struct proc *rp)                        );
 05131	
 05132	/* start.c */
 05133	_PROTOTYPE( void cstart, (U16_t cs, U16_t ds, U16_t mds,
 05134	                                U16_t parmoff, U16_t parmsize)          );
 05135	
 05136	/* system.c */
 05137	_PROTOTYPE( int get_priv, (register struct proc *rc, int proc_type)     );
 05138	_PROTOTYPE( void send_sig, (int proc_nr, int sig_nr)                    );
 05139	_PROTOTYPE( void cause_sig, (int proc_nr, int sig_nr)                   );
 05140	_PROTOTYPE( void sys_task, (void)                                       );
 05141	_PROTOTYPE( void get_randomness, (int source)                           );
 05142	_PROTOTYPE( int virtual_copy, (struct vir_addr *src, struct vir_addr *dst, 
 05143	                                vir_bytes bytes)                        );
 05144	#define numap_local(proc_nr, vir_addr, bytes) \
 05145	        umap_local(proc_addr(proc_nr), D, (vir_addr), (bytes))
 05146	_PROTOTYPE( phys_bytes umap_local, (struct proc *rp, int seg, 
 05147	                vir_bytes vir_addr, vir_bytes bytes)                    );
 05148	_PROTOTYPE( phys_bytes umap_remote, (struct proc *rp, int seg, 
 05149	                vir_bytes vir_addr, vir_bytes bytes)                    );
 05150	_PROTOTYPE( phys_bytes umap_bios, (struct proc *rp, vir_bytes vir_addr,
 05151	                vir_bytes bytes)                                        );
 05152	
 05153	/* exception.c */
 05154	_PROTOTYPE( void exception, (unsigned vec_nr)                           );
 05155	
 05156	/* i8259.c */
 05157	_PROTOTYPE( void intr_init, (int mine)                                  );
 05158	_PROTOTYPE( void intr_handle, (irq_hook_t *hook)                        );
 05159	_PROTOTYPE( void put_irq_handler, (irq_hook_t *hook, int irq,
 05160	                                                irq_handler_t handler)  );
 05161	_PROTOTYPE( void rm_irq_handler, (irq_hook_t *hook)                     );
 05162	
 05163	/* klib*.s */
 05164	_PROTOTYPE( void int86, (void)                                          );
 05165	_PROTOTYPE( void cp_mess, (int src,phys_clicks src_clicks,vir_bytes src_offset,
 05166	                phys_clicks dst_clicks, vir_bytes dst_offset)           );
 05167	_PROTOTYPE( void enable_irq, (irq_hook_t *hook)                         );
 05168	_PROTOTYPE( int disable_irq, (irq_hook_t *hook)                         );
 05169	_PROTOTYPE( u16_t mem_rdw, (U16_t segm, vir_bytes offset)               );
 05170	_PROTOTYPE( void phys_copy, (phys_bytes source, phys_bytes dest,
 05171	                phys_bytes count)                                       );
 05172	_PROTOTYPE( void phys_memset, (phys_bytes source, unsigned long pattern,
 05173	                phys_bytes count)                                       );
 05174	_PROTOTYPE( void phys_insb, (U16_t port, phys_bytes buf, size_t count)  );
 05175	_PROTOTYPE( void phys_insw, (U16_t port, phys_bytes buf, size_t count)  );
 05176	_PROTOTYPE( void phys_outsb, (U16_t port, phys_bytes buf, size_t count) );
 05177	_PROTOTYPE( void phys_outsw, (U16_t port, phys_bytes buf, size_t count) );
 05178	_PROTOTYPE( void reset, (void)                                          );
 05179	_PROTOTYPE( void level0, (void (*func)(void))                           );

_________________________ Page 696    File: kernel/proto.h _________________________
 05180	_PROTOTYPE( void monitor, (void)                                        );
 05181	_PROTOTYPE( void read_tsc, (unsigned long *high, unsigned long *low)    );
 05182	_PROTOTYPE( unsigned long read_cpu_flags, (void)                        );
 05183	
 05184	/* mpx*.s */
 05185	_PROTOTYPE( void idle_task, (void)                                      );
 05186	_PROTOTYPE( void restart, (void)                                        );
 05187	
 05188	/* The following are never called from C (pure asm procs). */
 05189	
 05190	/* Exception handlers (real or protected mode), in numerical order. */
 05191	void _PROTOTYPE( int00, (void) ), _PROTOTYPE( divide_error, (void) );
 05192	void _PROTOTYPE( int01, (void) ), _PROTOTYPE( single_step_exception, (void) );
 05193	void _PROTOTYPE( int02, (void) ), _PROTOTYPE( nmi, (void) );
 05194	void _PROTOTYPE( int03, (void) ), _PROTOTYPE( breakpoint_exception, (void) );
 05195	void _PROTOTYPE( int04, (void) ), _PROTOTYPE( overflow, (void) );
 05196	void _PROTOTYPE( int05, (void) ), _PROTOTYPE( bounds_check, (void) );
 05197	void _PROTOTYPE( int06, (void) ), _PROTOTYPE( inval_opcode, (void) );
 05198	void _PROTOTYPE( int07, (void) ), _PROTOTYPE( copr_not_available, (void) );
 05199	void                              _PROTOTYPE( double_fault, (void) );
 05200	void                              _PROTOTYPE( copr_seg_overrun, (void) );
 05201	void                              _PROTOTYPE( inval_tss, (void) );
 05202	void                              _PROTOTYPE( segment_not_present, (void) );
 05203	void                              _PROTOTYPE( stack_exception, (void) );
 05204	void                              _PROTOTYPE( general_protection, (void) );
 05205	void                              _PROTOTYPE( page_fault, (void) );
 05206	void                              _PROTOTYPE( copr_error, (void) );
 05207	
 05208	/* Hardware interrupt handlers. */
 05209	_PROTOTYPE( void hwint00, (void) );
 05210	_PROTOTYPE( void hwint01, (void) );
 05211	_PROTOTYPE( void hwint02, (void) );
 05212	_PROTOTYPE( void hwint03, (void) );
 05213	_PROTOTYPE( void hwint04, (void) );
 05214	_PROTOTYPE( void hwint05, (void) );
 05215	_PROTOTYPE( void hwint06, (void) );
 05216	_PROTOTYPE( void hwint07, (void) );
 05217	_PROTOTYPE( void hwint08, (void) );
 05218	_PROTOTYPE( void hwint09, (void) );
 05219	_PROTOTYPE( void hwint10, (void) );
 05220	_PROTOTYPE( void hwint11, (void) );
 05221	_PROTOTYPE( void hwint12, (void) );
 05222	_PROTOTYPE( void hwint13, (void) );
 05223	_PROTOTYPE( void hwint14, (void) );
 05224	_PROTOTYPE( void hwint15, (void) );
 05225	
 05226	/* Software interrupt handlers, in numerical order. */
 05227	_PROTOTYPE( void trp, (void) );
 05228	_PROTOTYPE( void s_call, (void) ), _PROTOTYPE( p_s_call, (void) );
 05229	_PROTOTYPE( void level0_call, (void) );
 05230	
 05231	/* protect.c */
 05232	_PROTOTYPE( void prot_init, (void)                                      );
 05233	_PROTOTYPE( void init_codeseg, (struct segdesc_s *segdp, phys_bytes base,
 05234	                vir_bytes size, int privilege)                          );
 05235	_PROTOTYPE( void init_dataseg, (struct segdesc_s *segdp, phys_bytes base,
 05236	                vir_bytes size, int privilege)                          );
 05237	_PROTOTYPE( phys_bytes seg2phys, (U16_t seg)                            );
 05238	_PROTOTYPE( void phys2seg, (u16_t *seg, vir_bytes *off, phys_bytes phys));
 05239	_PROTOTYPE( void enable_iop, (struct proc *pp)                          );

_________________________ Page 697    File: kernel/proto.h _________________________
 05240	_PROTOTYPE( void alloc_segments, (struct proc *rp)                      );
 05241	
 05242	#endif /* PROTO_H */
 05243	
 05244	

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/glo.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05300	#ifndef GLO_H
 05301	#define GLO_H
 05302	
 05303	/* Global variables used in the kernel. This file contains the declarations;
 05304	 * storage space for the variables is allocated in table.c, because EXTERN is
 05305	 * defined as extern unless the _TABLE definition is seen. We rely on the 
 05306	 * compiler's default initialization (0) for several global variables. 
 05307	 */
 05308	#ifdef _TABLE
 05309	#undef EXTERN
 05310	#define EXTERN
 05311	#endif
 05312	
 05313	#include <minix/config.h>
 05314	#include "config.h"
 05315	
 05316	/* Variables relating to shutting down MINIX. */
 05317	EXTERN char kernel_exception;           /* TRUE after system exceptions */
 05318	EXTERN char shutdown_started;           /* TRUE after shutdowns / reboots */
 05319	
 05320	/* Kernel information structures. This groups vital kernel information. */
 05321	EXTERN phys_bytes aout;                 /* address of a.out headers */
 05322	EXTERN struct kinfo kinfo;              /* kernel information for users */
 05323	EXTERN struct machine machine;          /* machine information for users */
 05324	EXTERN struct kmessages kmess;          /* diagnostic messages in kernel */
 05325	EXTERN struct randomness krandom;       /* gather kernel random information */
 05326	
 05327	/* Process scheduling information and the kernel reentry count. */
 05328	EXTERN struct proc *prev_ptr;   /* previously running process */
 05329	EXTERN struct proc *proc_ptr;   /* pointer to currently running process */
 05330	EXTERN struct proc *next_ptr;   /* next process to run after restart() */
 05331	EXTERN struct proc *bill_ptr;   /* process to bill for clock ticks */
 05332	EXTERN char k_reenter;          /* kernel reentry count (entry count less 1) */
 05333	EXTERN unsigned lost_ticks;     /* clock ticks counted outside clock task */
 05334	
 05335	/* Interrupt related variables. */
 05336	EXTERN irq_hook_t irq_hooks[NR_IRQ_HOOKS];      /* hooks for general use */
 05337	EXTERN irq_hook_t *irq_handlers[NR_IRQ_VECTORS];/* list of IRQ handlers */
 05338	EXTERN int irq_actids[NR_IRQ_VECTORS];          /* IRQ ID bits active */
 05339	EXTERN int irq_use;                             /* map of all in-use irq's */
 05340	
 05341	/* Miscellaneous. */
 05342	EXTERN reg_t mon_ss, mon_sp;            /* boot monitor stack */
 05343	EXTERN int mon_return;                  /* true if we can return to monitor */
 05344	
 05345	/* Variables that are initialized elsewhere are just extern here. */
 05346	extern struct boot_image image[];       /* system image processes */
 05347	extern char *t_stack[];                 /* task stack space */
 05348	extern struct segdesc_s gdt[];          /* global descriptor table */
 05349	

_________________________ Page 698    File: kernel/glo.h _________________________
 05350	EXTERN _PROTOTYPE( void (*level0_func), (void) );
 05351	
 05352	#endif /* GLO_H */
 05353	
 05354	
 05355	
 05356	
 05357	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/ipc.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05400	#ifndef IPC_H
 05401	#define IPC_H
 05402	
 05403	/* This header file defines constants for MINIX inter-process communication.
 05404	 * These definitions are used in the file proc.c.
 05405	 */
 05406	#include <minix/com.h>
 05407	
 05408	/* Masks and flags for system calls. */
 05409	#define SYSCALL_FUNC    0x0F    /* mask for system call function */
 05410	#define SYSCALL_FLAGS   0xF0    /* mask for system call flags */
 05411	#define NON_BLOCKING    0x10    /* prevent blocking, return error */
 05412	
 05413	/* System call numbers that are passed when trapping to the kernel. The 
 05414	 * numbers are carefully defined so that it can easily be seen (based on 
 05415	 * the bits that are on) which checks should be done in sys_call().
 05416	 */
 05417	#define SEND               1    /* 0 0 0 1 : blocking send */
 05418	#define RECEIVE            2    /* 0 0 1 0 : blocking receive */
 05419	#define SENDREC            3    /* 0 0 1 1 : SEND + RECEIVE */
 05420	#define NOTIFY             4    /* 0 1 0 0 : nonblocking notify */
 05421	#define ECHO               8    /* 1 0 0 0 : echo a message */
 05422	
 05423	/* The following bit masks determine what checks that should be done. */
 05424	#define CHECK_PTR       0x0B    /* 1 0 1 1 : validate message buffer */
 05425	#define CHECK_DST       0x05    /* 0 1 0 1 : validate message destination */
 05426	#define CHECK_SRC       0x02    /* 0 0 1 0 : validate message source */
 05427	
 05428	#endif /* IPC_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/proc.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05500	#ifndef PROC_H
 05501	#define PROC_H
 05502	
 05503	/* Here is the declaration of the process table.  It contains all process
 05504	 * data, including registers, flags, scheduling priority, memory map, 
 05505	 * accounting, message passing (IPC) information, and so on. 
 05506	 *
 05507	 * Many assembly code routines reference fields in it.  The offsets to these
 05508	 * fields are defined in the assembler include file sconst.h.  When changing
 05509	 * struct proc, be sure to change sconst.h to match.

_________________________ Page 699    File: kernel/proc.h _________________________
 05510	 */
 05511	#include <minix/com.h>
 05512	#include "protect.h"
 05513	#include "const.h"
 05514	#include "priv.h"
 05515	 
 05516	struct proc {
 05517	  struct stackframe_s p_reg;    /* process' registers saved in stack frame */
 05518	  reg_t p_ldt_sel;              /* selector in gdt with ldt base and limit */
 05519	  struct segdesc_s p_ldt[2+NR_REMOTE_SEGS]; /* CS, DS and remote segments */
 05520	
 05521	  proc_nr_t p_nr;               /* number of this process (for fast access) */
 05522	  struct priv *p_priv;          /* system privileges structure */
 05523	  char p_rts_flags;             /* SENDING, RECEIVING, etc. */
 05524	
 05525	  char p_priority;              /* current scheduling priority */
 05526	  char p_max_priority;          /* maximum scheduling priority */
 05527	  char p_ticks_left;            /* number of scheduling ticks left */
 05528	  char p_quantum_size;          /* quantum size in ticks */
 05529	
 05530	  struct mem_map p_memmap[NR_LOCAL_SEGS];   /* memory map (T, D, S) */
 05531	
 05532	  clock_t p_user_time;          /* user time in ticks */
 05533	  clock_t p_sys_time;           /* sys time in ticks */
 05534	
 05535	  struct proc *p_nextready;     /* pointer to next ready process */
 05536	  struct proc *p_caller_q;      /* head of list of procs wishing to send */
 05537	  struct proc *p_q_link;        /* link to next proc wishing to send */
 05538	  message *p_messbuf;           /* pointer to passed message buffer */
 05539	  proc_nr_t p_getfrom;          /* from whom does process want to receive? */
 05540	  proc_nr_t p_sendto;           /* to whom does process want to send? */
 05541	
 05542	  sigset_t p_pending;           /* bit map for pending kernel signals */
 05543	
 05544	  char p_name[P_NAME_LEN];      /* name of the process, including \0 */
 05545	};
 05546	
 05547	/* Bits for the runtime flags. A process is runnable iff p_rts_flags == 0. */
 05548	#define SLOT_FREE       0x01    /* process slot is free */
 05549	#define NO_MAP          0x02    /* keeps unmapped forked child from running */
 05550	#define SENDING         0x04    /* process blocked trying to SEND */
 05551	#define RECEIVING       0x08    /* process blocked trying to RECEIVE */
 05552	#define SIGNALED        0x10    /* set when new kernel signal arrives */
 05553	#define SIG_PENDING     0x20    /* unready while signal being processed */
 05554	#define P_STOP          0x40    /* set when process is being traced */
 05555	#define NO_PRIV         0x80    /* keep forked system process from running */
 05556	
 05557	/* Scheduling priorities for p_priority. Values must start at zero (highest
 05558	 * priority) and increment.  Priorities of the processes in the boot image 
 05559	 * can be set in table.c. IDLE must have a queue for itself, to prevent low 
 05560	 * priority user processes to run round-robin with IDLE. 
 05561	 */
 05562	#define NR_SCHED_QUEUES   16    /* MUST equal minimum priority + 1 */
 05563	#define TASK_Q             0    /* highest, used for kernel tasks */
 05564	#define MAX_USER_Q         0    /* highest priority for user processes */   
 05565	#define USER_Q             7    /* default (should correspond to nice 0) */   
 05566	#define MIN_USER_Q        14    /* minimum priority for user processes */
 05567	#define IDLE_Q            15    /* lowest, only IDLE process goes here */
 05568	
 05569	/* Magic process table addresses. */

_________________________ Page 700    File: kernel/proc.h _________________________
 05570	#define BEG_PROC_ADDR (&proc[0])
 05571	#define BEG_USER_ADDR (&proc[NR_TASKS])
 05572	#define END_PROC_ADDR (&proc[NR_TASKS + NR_PROCS])
 05573	
 05574	#define NIL_PROC          ((struct proc *) 0)           
 05575	#define NIL_SYS_PROC      ((struct proc *) 1)           
 05576	#define cproc_addr(n)     (&(proc + NR_TASKS)[(n)])
 05577	#define proc_addr(n)      (pproc_addr + NR_TASKS)[(n)]
 05578	#define proc_nr(p)        ((p)->p_nr)
 05579	
 05580	#define isokprocn(n)      ((unsigned) ((n) + NR_TASKS) < NR_PROCS + NR_TASKS)
 05581	#define isemptyn(n)       isemptyp(proc_addr(n)) 
 05582	#define isemptyp(p)       ((p)->p_rts_flags == SLOT_FREE)
 05583	#define iskernelp(p)      iskerneln((p)->p_nr)
 05584	#define iskerneln(n)      ((n) < 0)
 05585	#define isuserp(p)        isusern((p)->p_nr)
 05586	#define isusern(n)        ((n) >= 0)
 05587	
 05588	/* The process table and pointers to process table slots. The pointers allow
 05589	 * faster access because now a process entry can be found by indexing the
 05590	 * pproc_addr array, while accessing an element i requires a multiplication
 05591	 * with sizeof(struct proc) to determine the address. 
 05592	 */
 05593	EXTERN struct proc proc[NR_TASKS + NR_PROCS];   /* process table */
 05594	EXTERN struct proc *pproc_addr[NR_TASKS + NR_PROCS];
 05595	EXTERN struct proc *rdy_head[NR_SCHED_QUEUES]; /* ptrs to ready list headers */
 05596	EXTERN struct proc *rdy_tail[NR_SCHED_QUEUES]; /* ptrs to ready list tails */
 05597	
 05598	#endif /* PROC_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/sconst.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05600	! Miscellaneous constants used in assembler code.
 05601	W               =       _WORD_SIZE      ! Machine word size.
 05602	
 05603	! Offsets in struct proc. They MUST match proc.h.
 05604	P_STACKBASE     =       0
 05605	GSREG           =       P_STACKBASE
 05606	FSREG           =       GSREG + 2       ! 386 introduces FS and GS segments
 05607	ESREG           =       FSREG + 2
 05608	DSREG           =       ESREG + 2
 05609	DIREG           =       DSREG + 2
 05610	SIREG           =       DIREG + W
 05611	BPREG           =       SIREG + W
 05612	STREG           =       BPREG + W       ! hole for another SP
 05613	BXREG           =       STREG + W
 05614	DXREG           =       BXREG + W
 05615	CXREG           =       DXREG + W
 05616	AXREG           =       CXREG + W
 05617	RETADR          =       AXREG + W       ! return address for save() call
 05618	PCREG           =       RETADR + W
 05619	CSREG           =       PCREG + W
 05620	PSWREG          =       CSREG + W
 05621	SPREG           =       PSWREG + W
 05622	SSREG           =       SPREG + W
 05623	P_STACKTOP      =       SSREG + W
 05624	P_LDT_SEL       =       P_STACKTOP

_________________________ Page 701    File: kernel/sconst.h _________________________
 05625	P_LDT           =       P_LDT_SEL + W
 05626	
 05627	Msize           =       9               ! size of a message in 32-bit words



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/priv.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05700	#ifndef PRIV_H
 05701	#define PRIV_H
 05702	
 05703	/* Declaration of the system privileges structure. It defines flags, system 
 05704	 * call masks, an synchronous alarm timer, I/O privileges, pending hardware 
 05705	 * interrupts and notifications, and so on.
 05706	 * System processes each get their own structure with properties, whereas all 
 05707	 * user processes share one structure. This setup provides a clear separation
 05708	 * between common and privileged process fields and is very space efficient. 
 05709	 *
 05710	 * Changes:
 05711	 *   Jul 01, 2005       Created.  (Jorrit N. Herder)    
 05712	 */
 05713	#include <minix/com.h>
 05714	#include "protect.h"
 05715	#include "const.h"
 05716	#include "type.h"
 05717	 
 05718	struct priv {
 05719	  proc_nr_t s_proc_nr;          /* number of associated process */
 05720	  sys_id_t s_id;                /* index of this system structure */
 05721	  short s_flags;                /* PREEMTIBLE, BILLABLE, etc. */
 05722	
 05723	  short s_trap_mask;            /* allowed system call traps */
 05724	  sys_map_t s_ipc_from;         /* allowed callers to receive from */
 05725	  sys_map_t s_ipc_to;           /* allowed destination processes */
 05726	  long s_call_mask;             /* allowed kernel calls */
 05727	
 05728	  sys_map_t s_notify_pending;   /* bit map with pending notifications */
 05729	  irq_id_t s_int_pending;       /* pending hardware interrupts */
 05730	  sigset_t s_sig_pending;       /* pending signals */
 05731	
 05732	  timer_t s_alarm_timer;        /* synchronous alarm timer */ 
 05733	  struct far_mem s_farmem[NR_REMOTE_SEGS];  /* remote memory map */
 05734	  reg_t *s_stack_guard;         /* stack guard word for kernel tasks */
 05735	};
 05736	
 05737	/* Guard word for task stacks. */
 05738	#define STACK_GUARD     ((reg_t) (sizeof(reg_t) == 2 ? 0xBEEF : 0xDEADBEEF))
 05739	
 05740	/* Bits for the system property flags. */
 05741	#define PREEMPTIBLE     0x01    /* kernel tasks are not preemptible */
 05742	#define BILLABLE        0x04    /* some processes are not billable */
 05743	#define SYS_PROC        0x10    /* system processes are privileged */
 05744	#define SENDREC_BUSY    0x20    /* sendrec() in progress */
 05745	
 05746	/* Magic system structure table addresses. */
 05747	#define BEG_PRIV_ADDR (&priv[0])
 05748	#define END_PRIV_ADDR (&priv[NR_SYS_PROCS])
 05749	

_________________________ Page 702    File: kernel/priv.h _________________________
 05750	#define priv_addr(i)      (ppriv_addr)[(i)]
 05751	#define priv_id(rp)       ((rp)->p_priv->s_id)
 05752	#define priv(rp)          ((rp)->p_priv)
 05753	
 05754	#define id_to_nr(id)    priv_addr(id)->s_proc_nr
 05755	#define nr_to_id(nr)    priv(proc_addr(nr))->s_id
 05756	
 05757	/* The system structures table and pointers to individual table slots. The 
 05758	 * pointers allow faster access because now a process entry can be found by 
 05759	 * indexing the psys_addr array, while accessing an element i requires a 
 05760	 * multiplication with sizeof(struct sys) to determine the address. 
 05761	 */
 05762	EXTERN struct priv priv[NR_SYS_PROCS];          /* system properties table */
 05763	EXTERN struct priv *ppriv_addr[NR_SYS_PROCS];   /* direct slot pointers */
 05764	
 05765	/* Unprivileged user processes all share the same privilege structure.
 05766	 * This id must be fixed because it is used to check send mask entries.
 05767	 */
 05768	#define USER_PRIV_ID    0
 05769	
 05770	/* Make sure the system can boot. The following sanity check verifies that
 05771	 * the system privileges table is large enough for the number of processes
 05772	 * in the boot image. 
 05773	 */
 05774	#if (NR_BOOT_PROCS > NR_SYS_PROCS)
 05775	#error NR_SYS_PROCS must be larger than NR_BOOT_PROCS
 05776	#endif
 05777	
 05778	#endif /* PRIV_H */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/protect.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 05800	/* Constants for protected mode. */
 05801	
 05802	/* Table sizes. */
 05803	#define GDT_SIZE (FIRST_LDT_INDEX + NR_TASKS + NR_PROCS) 
 05804	                                        /* spec. and LDT's */
 05805	#define IDT_SIZE (IRQ8_VECTOR + 8)      /* only up to the highest vector */
 05806	#define LDT_SIZE (2 + NR_REMOTE_SEGS)   /* CS, DS and remote segments */
 05807	
 05808	/* Fixed global descriptors.  1 to 7 are prescribed by the BIOS. */
 05809	#define GDT_INDEX            1  /* GDT descriptor */
 05810	#define IDT_INDEX            2  /* IDT descriptor */
 05811	#define DS_INDEX             3  /* kernel DS */
 05812	#define ES_INDEX             4  /* kernel ES (386: flag 4 Gb at startup) */
 05813	#define SS_INDEX             5  /* kernel SS (386: monitor SS at startup) */
 05814	#define CS_INDEX             6  /* kernel CS */
 05815	#define MON_CS_INDEX         7  /* temp for BIOS (386: monitor CS at startup) */
 05816	#define TSS_INDEX            8  /* kernel TSS */
 05817	#define DS_286_INDEX         9  /* scratch 16-bit source segment */
 05818	#define ES_286_INDEX        10  /* scratch 16-bit destination segment */
 05819	#define A_INDEX             11  /* 64K memory segment at A0000 */
 05820	#define B_INDEX             12  /* 64K memory segment at B0000 */
 05821	#define C_INDEX             13  /* 64K memory segment at C0000 */
 05822	#define D_INDEX             14  /* 64K memory segment at D0000 */
 05823	#define FIRST_LDT_INDEX     15  /* rest of descriptors are LDT's */
 05824	

_________________________ Page 703    File: kernel/protect.h _________________________
 05825	#define GDT_SELECTOR      0x08  /* (GDT_INDEX * DESC_SIZE) bad for asld */
 05826	#define IDT_SELECTOR      0x10  /* (IDT_INDEX * DESC_SIZE) */
 05827	#define DS_SELECTOR       0x18  /* (DS_INDEX * DESC_SIZE) */
 05828	#define ES_SELECTOR       0x20  /* (ES_INDEX * DESC_SIZE) */
 05829	#define FLAT_DS_SELECTOR  0x21  /* less privileged ES */
 05830	#define SS_SELECTOR       0x28  /* (SS_INDEX * DESC_SIZE) */
 05831	#define CS_SELECTOR       0x30  /* (CS_INDEX * DESC_SIZE) */
 05832	#define MON_CS_SELECTOR   0x38  /* (MON_CS_INDEX * DESC_SIZE) */
 05833	#define TSS_SELECTOR      0x40  /* (TSS_INDEX * DESC_SIZE) */
 05834	#define DS_286_SELECTOR   0x49  /* (DS_286_INDEX*DESC_SIZE+TASK_PRIVILEGE) */
 05835	#define ES_286_SELECTOR   0x51  /* (ES_286_INDEX*DESC_SIZE+TASK_PRIVILEGE) */
 05836	
 05837	/* Fixed local descriptors. */
 05838	#define CS_LDT_INDEX         0  /* process CS */
 05839	#define DS_LDT_INDEX         1  /* process DS=ES=FS=GS=SS */
 05840	#define EXTRA_LDT_INDEX      2  /* first of the extra LDT entries */
 05841	
 05842	/* Privileges. */
 05843	#define INTR_PRIVILEGE       0  /* kernel and interrupt handlers */
 05844	#define TASK_PRIVILEGE       1  /* kernel tasks */
 05845	#define USER_PRIVILEGE       3  /* servers and user processes */
 05846	
 05847	/* 286 hardware constants. */
 05848	
 05849	/* Exception vector numbers. */
 05850	#define BOUNDS_VECTOR        5  /* bounds check failed */
 05851	#define INVAL_OP_VECTOR      6  /* invalid opcode */
 05852	#define COPROC_NOT_VECTOR    7  /* coprocessor not available */
 05853	#define DOUBLE_FAULT_VECTOR  8
 05854	#define COPROC_SEG_VECTOR    9  /* coprocessor segment overrun */
 05855	#define INVAL_TSS_VECTOR    10  /* invalid TSS */
 05856	#define SEG_NOT_VECTOR      11  /* segment not present */
 05857	#define STACK_FAULT_VECTOR  12  /* stack exception */
 05858	#define PROTECTION_VECTOR   13  /* general protection */
 05859	
 05860	/* Selector bits. */
 05861	#define TI                0x04  /* table indicator */
 05862	#define RPL               0x03  /* requester privilege level */
 05863	
 05864	/* Descriptor structure offsets. */
 05865	#define DESC_BASE            2  /* to base_low */
 05866	#define DESC_BASE_MIDDLE     4  /* to base_middle */
 05867	#define DESC_ACCESS          5  /* to access byte */
 05868	#define DESC_SIZE            8  /* sizeof (struct segdesc_s) */
 05869	
 05870	/* Base and limit sizes and shifts. */
 05871	#define BASE_MIDDLE_SHIFT   16  /* shift for base --> base_middle */
 05872	
 05873	/* Access-byte and type-byte bits. */
 05874	#define PRESENT           0x80  /* set for descriptor present */
 05875	#define DPL               0x60  /* descriptor privilege level mask */
 05876	#define DPL_SHIFT            5
 05877	#define SEGMENT           0x10  /* set for segment-type descriptors */
 05878	
 05879	/* Access-byte bits. */
 05880	#define EXECUTABLE        0x08  /* set for executable segment */
 05881	#define CONFORMING        0x04  /* set for conforming segment if executable */
 05882	#define EXPAND_DOWN       0x04  /* set for expand-down segment if !executable*/
 05883	#define READABLE          0x02  /* set for readable segment if executable */
 05884	#define WRITEABLE         0x02  /* set for writeable segment if !executable */

_________________________ Page 704    File: kernel/protect.h _________________________
 05885	#define TSS_BUSY          0x02  /* set if TSS descriptor is busy */
 05886	#define ACCESSED          0x01  /* set if segment accessed */
 05887	
 05888	/* Special descriptor types. */
 05889	#define AVL_286_TSS          1  /* available 286 TSS */
 05890	#define LDT                  2  /* local descriptor table */
 05891	#define BUSY_286_TSS         3  /* set transparently to the software */
 05892	#define CALL_286_GATE        4  /* not used */
 05893	#define TASK_GATE            5  /* only used by debugger */
 05894	#define INT_286_GATE         6  /* interrupt gate, used for all vectors */
 05895	#define TRAP_286_GATE        7  /* not used */
 05896	
 05897	/* Extra 386 hardware constants. */
 05898	
 05899	/* Exception vector numbers. */
 05900	#define PAGE_FAULT_VECTOR   14
 05901	#define COPROC_ERR_VECTOR   16  /* coprocessor error */
 05902	
 05903	/* Descriptor structure offsets. */
 05904	#define DESC_GRANULARITY     6  /* to granularity byte */
 05905	#define DESC_BASE_HIGH       7  /* to base_high */
 05906	
 05907	/* Base and limit sizes and shifts. */
 05908	#define BASE_HIGH_SHIFT     24  /* shift for base --> base_high */
 05909	#define BYTE_GRAN_MAX   0xFFFFFL   /* maximum size for byte granular segment */
 05910	#define GRANULARITY_SHIFT   16  /* shift for limit --> granularity */
 05911	#define OFFSET_HIGH_SHIFT   16  /* shift for (gate) offset --> offset_high */
 05912	#define PAGE_GRAN_SHIFT     12  /* extra shift for page granular limits */
 05913	
 05914	/* Type-byte bits. */
 05915	#define DESC_386_BIT  0x08 /* 386 types are obtained by ORing with this */
 05916	                                /* LDT's and TASK_GATE's don't need it */
 05917	
 05918	/* Granularity byte. */
 05919	#define GRANULAR          0x80  /* set for 4K granularilty */
 05920	#define DEFAULT           0x40  /* set for 32-bit defaults (executable seg) */
 05921	#define BIG               0x40  /* set for "BIG" (expand-down seg) */
 05922	#define AVL               0x10  /* 0 for available */
 05923	#define LIMIT_HIGH        0x0F  /* mask for high bits of limit */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/table.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 06000	/* The object file of "table.c" contains most kernel data. Variables that 
 06001	 * are declared in the *.h files appear with EXTERN in front of them, as in
 06002	 *
 06003	 *    EXTERN int x;
 06004	 *
 06005	 * Normally EXTERN is defined as extern, so when they are included in another
 06006	 * file, no storage is allocated.  If EXTERN were not present, but just say,
 06007	 *
 06008	 *    int x;
 06009	 *
 06010	 * then including this file in several source files would cause 'x' to be
 06011	 * declared several times.  While some linkers accept this, others do not,
 06012	 * so they are declared extern when included normally.  However, it must be
 06013	 * declared for real somewhere.  That is done here, by redefining EXTERN as
 06014	 * the null string, so that inclusion of all *.h files in table.c actually

_________________________ Page 705    File: kernel/table.c _________________________
 06015	 * generates storage for them.  
 06016	 *
 06017	 * Various variables could not be declared EXTERN, but are declared PUBLIC
 06018	 * or PRIVATE. The reason for this is that extern variables cannot have a  
 06019	 * default initialization. If such variables are shared, they must also be
 06020	 * declared in one of the *.h files without the initialization.  Examples 
 06021	 * include 'boot_image' (this file) and 'idt' and 'gdt' (protect.c). 
 06022	 *
 06023	 * Changes:
 06024	 *    Aug 02, 2005   set privileges and minimal boot image (Jorrit N. Herder)
 06025	 *    Oct 17, 2004   updated above and tasktab comments  (Jorrit N. Herder)
 06026	 *    May 01, 2004   changed struct for system image  (Jorrit N. Herder)
 06027	 */
 06028	#define _TABLE
 06029	
 06030	#include "kernel.h"
 06031	#include "proc.h"
 06032	#include "ipc.h"
 06033	#include <minix/com.h>
 06034	#include <ibm/int86.h>
 06035	
 06036	/* Define stack sizes for the kernel tasks included in the system image. */
 06037	#define NO_STACK        0
 06038	#define SMALL_STACK     (128 * sizeof(char *))
 06039	#define IDL_S   SMALL_STACK     /* 3 intr, 3 temps, 4 db for Intel */
 06040	#define HRD_S   NO_STACK        /* dummy task, uses kernel stack */
 06041	#define TSK_S   SMALL_STACK     /* system and clock task */
 06042	
 06043	/* Stack space for all the task stacks.  Declared as (char *) to align it. */
 06044	#define TOT_STACK_SPACE (IDL_S + HRD_S + (2 * TSK_S))
 06045	PUBLIC char *t_stack[TOT_STACK_SPACE / sizeof(char *)];
 06046	        
 06047	/* Define flags for the various process types. */
 06048	#define IDL_F   (SYS_PROC | PREEMPTIBLE | BILLABLE)     /* idle task */
 06049	#define TSK_F   (SYS_PROC)                              /* kernel tasks */
 06050	#define SRV_F   (SYS_PROC | PREEMPTIBLE)                /* system services */
 06051	#define USR_F   (BILLABLE | PREEMPTIBLE)                /* user processes */
 06052	
 06053	/* Define system call traps for the various process types. These call masks
 06054	 * determine what system call traps a process is allowed to make.
 06055	 */
 06056	#define TSK_T   (1 << RECEIVE)                  /* clock and system */
 06057	#define SRV_T   (~0)                            /* system services */
 06058	#define USR_T   ((1 << SENDREC) | (1 << ECHO))  /* user processes */
 06059	
 06060	/* Send masks determine to whom processes can send messages or notifications. 
 06061	 * The values here are used for the processes in the boot image. We rely on 
 06062	 * the initialization code in main() to match the s_nr_to_id() mapping for the
 06063	 * processes in the boot image, so that the send mask that is defined here 
 06064	 * can be directly copied onto map[0] of the actual send mask. Privilege
 06065	 * structure 0 is shared by user processes. 
 06066	 */
 06067	#define s(n)            (1 << s_nr_to_id(n))
 06068	#define SRV_M (~0)
 06069	#define SYS_M (~0)
 06070	#define USR_M (s(PM_PROC_NR) | s(FS_PROC_NR) | s(RS_PROC_NR))
 06071	#define DRV_M (USR_M | s(SYSTEM) | s(CLOCK) | s(LOG_PROC_NR) | s(TTY_PROC_NR))
 06072	
 06073	/* Define kernel calls that processes are allowed to make. This is not looking
 06074	 * very nice, but we need to define the access rights on a per call basis. 

_________________________ Page 706    File: kernel/table.c _________________________
 06075	 * Note that the reincarnation server has all bits on, because it should
 06076	 * be allowed to distribute rights to services that it starts. 
 06077	 */
 06078	#define c(n)    (1 << ((n)-KERNEL_CALL))
 06079	#define RS_C    ~0      
 06080	#define PM_C    ~(c(SYS_DEVIO) | c(SYS_SDEVIO) | c(SYS_VDEVIO) \
 06081	    | c(SYS_IRQCTL) | c(SYS_INT86))
 06082	#define FS_C    (c(SYS_KILL) | c(SYS_VIRCOPY) | c(SYS_VIRVCOPY) | c(SYS_UMAP) \
 06083	    | c(SYS_GETINFO) | c(SYS_EXIT) | c(SYS_TIMES) | c(SYS_SETALARM))
 06084	#define DRV_C   (FS_C | c(SYS_SEGCTL) | c(SYS_IRQCTL) | c(SYS_INT86) \
 06085	    | c(SYS_DEVIO) | c(SYS_VDEVIO) | c(SYS_SDEVIO)) 
 06086	#define MEM_C   (DRV_C | c(SYS_PHYSCOPY) | c(SYS_PHYSVCOPY))
 06087	
 06088	/* The system image table lists all programs that are part of the boot image. 
 06089	 * The order of the entries here MUST agree with the order of the programs
 06090	 * in the boot image and all kernel tasks must come first.
 06091	 * Each entry provides the process number, flags, quantum size (qs), scheduling
 06092	 * queue, allowed traps, ipc mask, and a name for the process table. The 
 06093	 * initial program counter and stack size is also provided for kernel tasks.
 06094	 */
 06095	PUBLIC struct boot_image image[] = {
 06096	/* process nr,   pc, flags, qs,  queue, stack, traps, ipcto, call,  name */ 
 06097	 { IDLE,  idle_task, IDL_F,  8, IDLE_Q, IDL_S,     0,     0,     0, "IDLE"  },
 06098	 { CLOCK,clock_task, TSK_F, 64, TASK_Q, TSK_S, TSK_T,     0,     0, "CLOCK" },
 06099	 { SYSTEM, sys_task, TSK_F, 64, TASK_Q, TSK_S, TSK_T,     0,     0, "SYSTEM"},
 06100	 { HARDWARE,      0, TSK_F, 64, TASK_Q, HRD_S,     0,     0,     0, "KERNEL"},
 06101	 { PM_PROC_NR,    0, SRV_F, 32,      3, 0,     SRV_T, SRV_M,  PM_C, "pm"    },
 06102	 { FS_PROC_NR,    0, SRV_F, 32,      4, 0,     SRV_T, SRV_M,  FS_C, "fs"    },
 06103	 { RS_PROC_NR,    0, SRV_F,  4,      3, 0,     SRV_T, SYS_M,  RS_C, "rs"    },
 06104	 { TTY_PROC_NR,   0, SRV_F,  4,      1, 0,     SRV_T, SYS_M, DRV_C, "tty"   },
 06105	 { MEM_PROC_NR,   0, SRV_F,  4,      2, 0,     SRV_T, DRV_M, MEM_C, "memory"},
 06106	 { LOG_PROC_NR,   0, SRV_F,  4,      2, 0,     SRV_T, SYS_M, DRV_C, "log"   },
 06107	 { DRVR_PROC_NR,  0, SRV_F,  4,      2, 0,     SRV_T, SYS_M, DRV_C, "driver"},
 06108	 { INIT_PROC_NR,  0, USR_F,  8, USER_Q, 0,     USR_T, USR_M,     0, "init"  },
 06109	};
 06110	
 06111	/* Verify the size of the system image table at compile time. Also verify that 
 06112	 * the first chunk of the ipc mask has enough bits to accommodate the processes
 06113	 * in the image.  
 06114	 * If a problem is detected, the size of the 'dummy' array will be negative, 
 06115	 * causing a compile time error. Note that no space is actually allocated 
 06116	 * because 'dummy' is declared extern.
 06117	 */
 06118	extern int dummy[(NR_BOOT_PROCS==sizeof(image)/
 06119	        sizeof(struct boot_image))?1:-1];
 06120	extern int dummy[(BITCHUNK_BITS > NR_BOOT_PROCS - 1) ? 1 : -1];
 06121	




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/mpx.s
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 06200	#
 06201	! Chooses between the 8086 and 386 versions of the Minix startup code.
 06202	
 06203	#include <minix/config.h>
 06204	#if _WORD_SIZE == 2

_________________________ Page 707    File: kernel/mpx.s _________________________
 06205	#include "mpx88.s"
 06206	#else
 06207	#include "mpx386.s"
 06208	#endif


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/mpx386.s
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 06300	# 
 06301	! This file, mpx386.s, is included by mpx.s when Minix is compiled for 
 06302	! 32-bit Intel CPUs. The alternative mpx88.s is compiled for 16-bit CPUs.
 06303	
 06304	! This file is part of the lowest layer of the MINIX kernel.  (The other part
 06305	! is "proc.c".)  The lowest layer does process switching and message handling.
 06306	! Furthermore it contains the assembler startup code for Minix and the 32-bit
 06307	! interrupt handlers.  It cooperates with the code in "start.c" to set up a 
 06308	! good environment for main().
 06309	
 06310	! Every transition to the kernel goes through this file.  Transitions to the 
 06311	! kernel may be nested.  The initial entry may be with a system call (i.e., 
 06312	! send or receive a message), an exception or a hardware interrupt;  kernel 
 06313	! reentries may only be made by hardware interrupts.  The count of reentries 
 06314	! is kept in "k_reenter". It is important for deciding whether to switch to 
 06315	! the kernel stack and for protecting the message passing code in "proc.c".
 06316	
 06317	! For the message passing trap, most of the machine state is saved in the
 06318	! proc table.  (Some of the registers need not be saved.)  Then the stack is
 06319	! switched to "k_stack", and interrupts are reenabled.  Finally, the system
 06320	! call handler (in C) is called.  When it returns, interrupts are disabled
 06321	! again and the code falls into the restart routine, to finish off held-up
 06322	! interrupts and run the process or task whose pointer is in "proc_ptr".
 06323	
 06324	! Hardware interrupt handlers do the same, except  (1) The entire state must
 06325	! be saved.  (2) There are too many handlers to do this inline, so the save
 06326	! routine is called.  A few cycles are saved by pushing the address of the
 06327	! appropiate restart routine for a return later.  (3) A stack switch is
 06328	! avoided when the stack is already switched.  (4) The (master) 8259 interrupt
 06329	! controller is reenabled centrally in save().  (5) Each interrupt handler
 06330	! masks its interrupt line using the 8259 before enabling (other unmasked)
 06331	! interrupts, and unmasks it after servicing the interrupt.  This limits the
 06332	! nest level to the number of lines and protects the handler from itself.
 06333	
 06334	! For communication with the boot monitor at startup time some constant
 06335	! data are compiled into the beginning of the text segment. This facilitates 
 06336	! reading the data at the start of the boot process, since only the first
 06337	! sector of the file needs to be read.
 06338	
 06339	! Some data storage is also allocated at the end of this file. This data 
 06340	! will be at the start of the data segment of the kernel and will be read
 06341	! and modified by the boot monitor before the kernel starts.
 06342	
 06343	! sections
 06344	
 06345	.sect .text
 06346	begtext:
 06347	.sect .rom
 06348	begrom:
 06349	.sect .data

_________________________ Page 708    File: kernel/mpx386.s _________________________
 06350	begdata:
 06351	.sect .bss
 06352	begbss:
 06353	
 06354	#include <minix/config.h>
 06355	#include <minix/const.h>
 06356	#include <minix/com.h>
 06357	#include <ibm/interrupt.h>
 06358	#include "const.h"
 06359	#include "protect.h"
 06360	#include "sconst.h"
 06361	
 06362	/* Selected 386 tss offsets. */
 06363	#define TSS3_S_SP0      4
 06364	
 06365	! Exported functions
 06366	! Note: in assembly language the .define statement applied to a function name 
 06367	! is loosely equivalent to a prototype in C code -- it makes it possible to
 06368	! link to an entity declared in the assembly code but does not create
 06369	! the entity.
 06370	
 06371	.define _restart
 06372	.define save
 06373	
 06374	.define _divide_error
 06375	.define _single_step_exception
 06376	.define _nmi
 06377	.define _breakpoint_exception
 06378	.define _overflow
 06379	.define _bounds_check
 06380	.define _inval_opcode
 06381	.define _copr_not_available
 06382	.define _double_fault
 06383	.define _copr_seg_overrun
 06384	.define _inval_tss
 06385	.define _segment_not_present
 06386	.define _stack_exception
 06387	.define _general_protection
 06388	.define _page_fault
 06389	.define _copr_error
 06390	
 06391	.define _hwint00        ! handlers for hardware interrupts
 06392	.define _hwint01
 06393	.define _hwint02
 06394	.define _hwint03
 06395	.define _hwint04
 06396	.define _hwint05
 06397	.define _hwint06
 06398	.define _hwint07
 06399	.define _hwint08
 06400	.define _hwint09
 06401	.define _hwint10
 06402	.define _hwint11
 06403	.define _hwint12
 06404	.define _hwint13
 06405	.define _hwint14
 06406	.define _hwint15
 06407	
 06408	.define _s_call
 06409	.define _p_s_call

_________________________ Page 709    File: kernel/mpx386.s _________________________
 06410	.define _level0_call
 06411	
 06412	! Exported variables.
 06413	.define begbss
 06414	.define begdata
 06415	
 06416	.sect .text
 06417	!*===========================================================================*
 06418	!*                              MINIX                                        *
 06419	!*===========================================================================*
 06420	MINIX:                          ! this is the entry point for the MINIX kernel
 06421	        jmp     over_flags      ! skip over the next few bytes
 06422	        .data2  CLICK_SHIFT     ! for the monitor: memory granularity
 06423	flags:
 06424	        .data2  0x01FD          ! boot monitor flags:
 06425	                                !       call in 386 mode, make bss, make stack,
 06426	                                !       load high, don't patch, will return,
 06427	                                !       uses generic INT, memory vector,
 06428	                                !       new boot code return
 06429	        nop                     ! extra byte to sync up disassembler
 06430	over_flags:
 06431	
 06432	! Set up a C stack frame on the monitor stack.  (The monitor sets cs and ds
 06433	! right.  The ss descriptor still references the monitor data segment.)
 06434	        movzx   esp, sp         ! monitor stack is a 16 bit stack
 06435	        push    ebp
 06436	        mov     ebp, esp
 06437	        push    esi
 06438	        push    edi
 06439	        cmp     4(ebp), 0       ! monitor return vector is
 06440	        jz      noret           ! nonzero if return possible
 06441	        inc     (_mon_return)
 06442	noret:  mov     (_mon_sp), esp  ! save stack pointer for later return
 06443	
 06444	! Copy the monitor global descriptor table to the address space of kernel and
 06445	! switch over to it.  Prot_init() can then update it with immediate effect.
 06446	
 06447	        sgdt    (_gdt+GDT_SELECTOR)             ! get the monitor gdtr
 06448	        mov     esi, (_gdt+GDT_SELECTOR+2)      ! absolute address of GDT
 06449	        mov     ebx, _gdt                       ! address of kernel GDT
 06450	        mov     ecx, 8*8                        ! copying eight descriptors
 06451	copygdt:
 06452	 eseg   movb    al, (esi)
 06453	        movb    (ebx), al
 06454	        inc     esi
 06455	        inc     ebx
 06456	        loop    copygdt
 06457	        mov     eax, (_gdt+DS_SELECTOR+2)       ! base of kernel data
 06458	        and     eax, 0x00FFFFFF                 ! only 24 bits
 06459	        add     eax, _gdt                       ! eax = vir2phys(gdt)
 06460	        mov     (_gdt+GDT_SELECTOR+2), eax      ! set base of GDT
 06461	        lgdt    (_gdt+GDT_SELECTOR)             ! switch over to kernel GDT
 06462	
 06463	! Locate boot parameters, set up kernel segment registers and stack.
 06464	        mov     ebx, 8(ebp)     ! boot parameters offset
 06465	        mov     edx, 12(ebp)    ! boot parameters length
 06466	        mov     eax, 16(ebp)    ! address of a.out headers
 06467	        mov     (_aout), eax
 06468	        mov     ax, ds          ! kernel data
 06469	        mov     es, ax

_________________________ Page 710    File: kernel/mpx386.s _________________________
 06470	        mov     fs, ax
 06471	        mov     gs, ax
 06472	        mov     ss, ax
 06473	        mov     esp, k_stktop   ! set sp to point to the top of kernel stack
 06474	
 06475	! Call C startup code to set up a proper environment to run main().
 06476	        push    edx
 06477	        push    ebx
 06478	        push    SS_SELECTOR
 06479	        push    DS_SELECTOR
 06480	        push    CS_SELECTOR
 06481	        call    _cstart         ! cstart(cs, ds, mds, parmoff, parmlen)
 06482	        add     esp, 5*4
 06483	
 06484	! Reload gdtr, idtr and the segment registers to global descriptor table set
 06485	! up by prot_init().
 06486	
 06487	        lgdt    (_gdt+GDT_SELECTOR)
 06488	        lidt    (_gdt+IDT_SELECTOR)
 06489	
 06490	        jmpf    CS_SELECTOR:csinit
 06491	csinit:
 06492	    o16 mov     ax, DS_SELECTOR
 06493	        mov     ds, ax
 06494	        mov     es, ax
 06495	        mov     fs, ax
 06496	        mov     gs, ax
 06497	        mov     ss, ax
 06498	    o16 mov     ax, TSS_SELECTOR        ! no other TSS is used
 06499	        ltr     ax
 06500	        push    0                       ! set flags to known good state
 06501	        popf                            ! esp, clear nested task and int enable
 06502	
 06503	        jmp     _main                   ! main()
 06504	
 06505	
 06506	!*===========================================================================*
 06507	!*                              interrupt handlers                           *
 06508	!*              interrupt handlers for 386 32-bit protected mode             *
 06509	!*===========================================================================*
 06510	
 06511	!*===========================================================================*
 06512	!*                              hwint00 - 07                                 *
 06513	!*===========================================================================*
 06514	! Note this is a macro, it just looks like a subroutine.
 06515	#define hwint_master(irq)       \
 06516	        call    save                    /* save interrupted process state */;\
 06517	        push    (_irq_handlers+4*irq)   /* irq_handlers[irq]              */;\
 06518	        call    _intr_handle            /* intr_handle(irq_handlers[irq]) */;\
 06519	        pop     ecx                                                         ;\
 06520	        cmp     (_irq_actids+4*irq), 0  /* interrupt still active?        */;\
 06521	        jz      0f                                                          ;\
 06522	        inb     INT_CTLMASK             /* get current mask */              ;\
 06523	        orb     al, [1<<irq]            /* mask irq */                      ;
 06524	        outb    INT_CTLMASK             /* disable the irq                */;\
 06525	0:      movb    al, END_OF_INT                                              ;\
 06526	        outb    INT_CTL                 /* reenable master 8259           */;\
 06527	        ret                             /* restart (another) process      */
 06528	
 06529	! Each of these entry points is an expansion of the hwint_master macro

_________________________ Page 711    File: kernel/mpx386.s _________________________
 06530	        .align  16
 06531	_hwint00:               ! Interrupt routine for irq 0 (the clock).
 06532	        hwint_master(0)
 06533	
 06534	        .align  16
 06535	_hwint01:               ! Interrupt routine for irq 1 (keyboard)
 06536	        hwint_master(1)
 06537	
 06538	        .align  16
 06539	_hwint02:               ! Interrupt routine for irq 2 (cascade!)
 06540	        hwint_master(2)
 06541	
 06542	        .align  16
 06543	_hwint03:               ! Interrupt routine for irq 3 (second serial)
 06544	        hwint_master(3)
 06545	
 06546	        .align  16
 06547	_hwint04:               ! Interrupt routine for irq 4 (first serial)
 06548	        hwint_master(4)
 06549	
 06550	        .align  16
 06551	_hwint05:               ! Interrupt routine for irq 5 (XT winchester)
 06552	        hwint_master(5)
 06553	
 06554	        .align  16
 06555	_hwint06:               ! Interrupt routine for irq 6 (floppy)
 06556	        hwint_master(6)
 06557	
 06558	        .align  16
 06559	_hwint07:               ! Interrupt routine for irq 7 (printer)
 06560	        hwint_master(7)
 06561	
 06562	!*===========================================================================*
 06563	!*                              hwint08 - 15                                 *
 06564	!*===========================================================================*
 06565	! Note this is a macro, it just looks like a subroutine.
 06566	#define hwint_slave(irq)        \
 06567	        call    save                    /* save interrupted process state */;\
 06568	        push    (_irq_handlers+4*irq)   /* irq_handlers[irq]              */;\
 06569	        call    _intr_handle            /* intr_handle(irq_handlers[irq]) */;\
 06570	        pop     ecx                                                         ;\
 06571	        cmp     (_irq_actids+4*irq), 0  /* interrupt still active?        */;\
 06572	        jz      0f                                                          ;\
 06573	        inb     INT2_CTLMASK                                                ;\
 06574	        orb     al, [1<<[irq-8]]                                            ;
 06575	        outb    INT2_CTLMASK            /* disable the irq                */;\
 06576	0:      movb    al, END_OF_INT                                              ;\
 06577	        outb    INT_CTL                 /* reenable master 8259           */;\
 06578	        outb    INT2_CTL                /* reenable slave 8259            */;\
 06579	        ret                             /* restart (another) process      */
 06580	
 06581	! Each of these entry points is an expansion of the hwint_slave macro
 06582	        .align  16
 06583	_hwint08:               ! Interrupt routine for irq 8 (realtime clock)
 06584	        hwint_slave(8)
 06585	
 06586	        .align  16
 06587	_hwint09:               ! Interrupt routine for irq 9 (irq 2 redirected)
 06588	        hwint_slave(9)
 06589	

_________________________ Page 712    File: kernel/mpx386.s _________________________
 06590	        .align  16
 06591	_hwint10:               ! Interrupt routine for irq 10
 06592	        hwint_slave(10)
 06593	
 06594	        .align  16
 06595	_hwint11:               ! Interrupt routine for irq 11
 06596	        hwint_slave(11)
 06597	
 06598	        .align  16
 06599	_hwint12:               ! Interrupt routine for irq 12
 06600	        hwint_slave(12)
 06601	
 06602	        .align  16
 06603	_hwint13:               ! Interrupt routine for irq 13 (FPU exception)
 06604	        hwint_slave(13)
 06605	
 06606	        .align  16
 06607	_hwint14:               ! Interrupt routine for irq 14 (AT winchester)
 06608	        hwint_slave(14)
 06609	
 06610	        .align  16
 06611	_hwint15:               ! Interrupt routine for irq 15
 06612	        hwint_slave(15)
 06613	
 06614	!*===========================================================================*
 06615	!*                              save                                         *
 06616	!*===========================================================================*
 06617	! Save for protected mode.
 06618	! This is much simpler than for 8086 mode, because the stack already points
 06619	! into the process table, or has already been switched to the kernel stack.
 06620	
 06621	        .align  16
 06622	save:
 06623	        cld                     ! set direction flag to a known value
 06624	        pushad                  ! save "general" registers
 06625	    o16 push    ds              ! save ds
 06626	    o16 push    es              ! save es
 06627	    o16 push    fs              ! save fs
 06628	    o16 push    gs              ! save gs
 06629	        mov     dx, ss          ! ss is kernel data segment
 06630	        mov     ds, dx          ! load rest of kernel segments
 06631	        mov     es, dx          ! kernel does not use fs, gs
 06632	        mov     eax, esp        ! prepare to return
 06633	        incb    (_k_reenter)    ! from -1 if not reentering
 06634	        jnz     set_restart1    ! stack is already kernel stack
 06635	        mov     esp, k_stktop
 06636	        push    _restart        ! build return address for int handler
 06637	        xor     ebp, ebp        ! for stacktrace
 06638	        jmp     RETADR-P_STACKBASE(eax)
 06639	
 06640	        .align  4
 06641	set_restart1:
 06642	        push    restart1
 06643	        jmp     RETADR-P_STACKBASE(eax)
 06644	
 06645	!*===========================================================================*
 06646	!*                              _s_call                                      *
 06647	!*===========================================================================*
 06648	        .align  16
 06649	_s_call:

_________________________ Page 713    File: kernel/mpx386.s _________________________
 06650	_p_s_call:
 06651	        cld                     ! set direction flag to a known value
 06652	        sub     esp, 6*4        ! skip RETADR, eax, ecx, edx, ebx, est
 06653	        push    ebp             ! stack already points into proc table
 06654	        push    esi
 06655	        push    edi
 06656	    o16 push    ds
 06657	    o16 push    es
 06658	    o16 push    fs
 06659	    o16 push    gs
 06660	        mov     dx, ss
 06661	        mov     ds, dx
 06662	        mov     es, dx
 06663	        incb    (_k_reenter)
 06664	        mov     esi, esp        ! assumes P_STACKBASE == 0
 06665	        mov     esp, k_stktop
 06666	        xor     ebp, ebp        ! for stacktrace
 06667	                                ! end of inline save
 06668	                                ! now set up parameters for sys_call()
 06669	        push    ebx             ! pointer to user message
 06670	        push    eax             ! src/dest
 06671	        push    ecx             ! SEND/RECEIVE/BOTH
 06672	        call    _sys_call       ! sys_call(function, src_dest, m_ptr)
 06673	                                ! caller is now explicitly in proc_ptr
 06674	        mov     AXREG(esi), eax ! sys_call MUST PRESERVE si
 06675	
 06676	! Fall into code to restart proc/task running.
 06677	
 06678	!*===========================================================================*
 06679	!*                              restart                                      *
 06680	!*===========================================================================*
 06681	_restart:
 06682	
 06683	! Restart the current process or the next process if it is set. 
 06684	
 06685	        cmp     (_next_ptr), 0          ! see if another process is scheduled
 06686	        jz      0f
 06687	        mov     eax, (_next_ptr)
 06688	        mov     (_proc_ptr), eax        ! schedule new process 
 06689	        mov     (_next_ptr), 0
 06690	0:      mov     esp, (_proc_ptr)        ! will assume P_STACKBASE == 0
 06691	        lldt    P_LDT_SEL(esp)          ! enable process' segment descriptors 
 06692	        lea     eax, P_STACKTOP(esp)    ! arrange for next interrupt
 06693	        mov     (_tss+TSS3_S_SP0), eax  ! to save state in process table
 06694	restart1:
 06695	        decb    (_k_reenter)
 06696	    o16 pop     gs
 06697	    o16 pop     fs
 06698	    o16 pop     es
 06699	    o16 pop     ds
 06700	        popad
 06701	        add     esp, 4          ! skip return adr
 06702	        iretd                   ! continue process
 06703	
 06704	!*===========================================================================*
 06705	!*                              exception handlers                           *
 06706	!*===========================================================================*
 06707	_divide_error:
 06708	        push    DIVIDE_VECTOR
 06709	        jmp     exception

_________________________ Page 714    File: kernel/mpx386.s _________________________
 06710	
 06711	_single_step_exception:
 06712	        push    DEBUG_VECTOR
 06713	        jmp     exception
 06714	
 06715	_nmi:
 06716	        push    NMI_VECTOR
 06717	        jmp     exception
 06718	
 06719	_breakpoint_exception:
 06720	        push    BREAKPOINT_VECTOR
 06721	        jmp     exception
 06722	
 06723	_overflow:
 06724	        push    OVERFLOW_VECTOR
 06725	        jmp     exception
 06726	
 06727	_bounds_check:
 06728	        push    BOUNDS_VECTOR
 06729	        jmp     exception
 06730	
 06731	_inval_opcode:
 06732	        push    INVAL_OP_VECTOR
 06733	        jmp     exception
 06734	
 06735	_copr_not_available:
 06736	        push    COPROC_NOT_VECTOR
 06737	        jmp     exception
 06738	
 06739	_double_fault:
 06740	        push    DOUBLE_FAULT_VECTOR
 06741	        jmp     errexception
 06742	
 06743	_copr_seg_overrun:
 06744	        push    COPROC_SEG_VECTOR
 06745	        jmp     exception
 06746	
 06747	_inval_tss:
 06748	        push    INVAL_TSS_VECTOR
 06749	        jmp     errexception
 06750	
 06751	_segment_not_present:
 06752	        push    SEG_NOT_VECTOR
 06753	        jmp     errexception
 06754	
 06755	_stack_exception:
 06756	        push    STACK_FAULT_VECTOR
 06757	        jmp     errexception
 06758	
 06759	_general_protection:
 06760	        push    PROTECTION_VECTOR
 06761	        jmp     errexception
 06762	
 06763	_page_fault:
 06764	        push    PAGE_FAULT_VECTOR
 06765	        jmp     errexception
 06766	
 06767	_copr_error:
 06768	        push    COPROC_ERR_VECTOR
 06769	        jmp     exception

_________________________ Page 715    File: kernel/mpx386.s _________________________
 06770	
 06771	!*===========================================================================*
 06772	!*                              exception                                    *
 06773	!*===========================================================================*
 06774	! This is called for all exceptions which do not push an error code.
 06775	
 06776	        .align  16
 06777	exception:
 06778	 sseg   mov     (trap_errno), 0         ! clear trap_errno
 06779	 sseg   pop     (ex_number)
 06780	        jmp     exception1
 06781	
 06782	!*===========================================================================*
 06783	!*                              errexception                                 *
 06784	!*===========================================================================*
 06785	! This is called for all exceptions which push an error code.
 06786	
 06787	        .align  16
 06788	errexception:
 06789	 sseg   pop     (ex_number)
 06790	 sseg   pop     (trap_errno)
 06791	exception1:                             ! Common for all exceptions.
 06792	        push    eax                     ! eax is scratch register
 06793	        mov     eax, 0+4(esp)           ! old eip
 06794	 sseg   mov     (old_eip), eax
 06795	        movzx   eax, 4+4(esp)           ! old cs
 06796	 sseg   mov     (old_cs), eax
 06797	        mov     eax, 8+4(esp)           ! old eflags
 06798	 sseg   mov     (old_eflags), eax
 06799	        pop     eax
 06800	        call    save
 06801	        push    (old_eflags)
 06802	        push    (old_cs)
 06803	        push    (old_eip)
 06804	        push    (trap_errno)
 06805	        push    (ex_number)
 06806	        call    _exception              ! (ex_number, trap_errno, old_eip,
 06807	                                        !       old_cs, old_eflags)
 06808	        add     esp, 5*4
 06809	        ret
 06810	
 06811	!*===========================================================================*
 06812	!*                              level0_call                                  *
 06813	!*===========================================================================*
 06814	_level0_call:
 06815	        call    save
 06816	        jmp     (_level0_func)
 06817	
 06818	!*===========================================================================*
 06819	!*                              data                                         *
 06820	!*===========================================================================*
 06821	
 06822	.sect .rom      ! Before the string table please
 06823	        .data2  0x526F          ! this must be the first data entry (magic #)
 06824	
 06825	.sect .bss
 06826	k_stack:
 06827	        .space  K_STACK_BYTES   ! kernel stack
 06828	k_stktop:                       ! top of kernel stack
 06829	        .comm   ex_number, 4

_________________________ Page 716    File: kernel/mpx386.s _________________________
 06830	        .comm   trap_errno, 4
 06831	        .comm   old_eip, 4
 06832	        .comm   old_cs, 4
 06833	        .comm   old_eflags, 4


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/start.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 06900	/* This file contains the C startup code for Minix on Intel processors.
 06901	 * It cooperates with mpx.s to set up a good environment for main().
 06902	 *
 06903	 * This code runs in real mode for a 16 bit kernel and may have to switch
 06904	 * to protected mode for a 286.
 06905	 * For a 32 bit kernel this already runs in protected mode, but the selectors
 06906	 * are still those given by the BIOS with interrupts disabled, so the
 06907	 * descriptors need to be reloaded and interrupt descriptors made.
 06908	 */
 06909	
 06910	#include "kernel.h"
 06911	#include "protect.h"
 06912	#include "proc.h"
 06913	#include <stdlib.h>
 06914	#include <string.h>
 06915	
 06916	FORWARD _PROTOTYPE( char *get_value, (_CONST char *params, _CONST char *key));
 06917	/*===========================================================================*
 06918	 *                              cstart                                       *
 06919	 *===========================================================================*/
 06920	PUBLIC void cstart(cs, ds, mds, parmoff, parmsize)
 06921	U16_t cs, ds;                   /* kernel code and data segment */
 06922	U16_t mds;                      /* monitor data segment */
 06923	U16_t parmoff, parmsize;        /* boot parameters offset and length */
 06924	{
 06925	/* Perform system initializations prior to calling main(). Most settings are
 06926	 * determined with help of the environment strings passed by MINIX' loader.
 06927	 */
 06928	  char params[128*sizeof(char *)];              /* boot monitor parameters */
 06929	  register char *value;                         /* value in key=value pair */
 06930	  extern int etext, end;
 06931	
 06932	  /* Decide if mode is protected; 386 or higher implies protected mode.
 06933	   * This must be done first, because it is needed for, e.g., seg2phys().
 06934	   * For 286 machines we cannot decide on protected mode, yet. This is 
 06935	   * done below. 
 06936	   */
 06937	#if _WORD_SIZE != 2
 06938	  machine.protected = 1;        
 06939	#endif
 06940	
 06941	  /* Record where the kernel and the monitor are. */
 06942	  kinfo.code_base = seg2phys(cs);
 06943	  kinfo.code_size = (phys_bytes) &etext;        /* size of code segment */
 06944	  kinfo.data_base = seg2phys(ds);
 06945	  kinfo.data_size = (phys_bytes) &end;          /* size of data segment */
 06946	
 06947	  /* Initialize protected mode descriptors. */
 06948	  prot_init();
 06949	

_________________________ Page 717    File: kernel/start.c _________________________
 06950	  /* Copy the boot parameters to the local buffer. */
 06951	  kinfo.params_base = seg2phys(mds) + parmoff;
 06952	  kinfo.params_size = MIN(parmsize,sizeof(params)-2);
 06953	  phys_copy(kinfo.params_base, vir2phys(params), kinfo.params_size);
 06954	
 06955	  /* Record miscellaneous information for user-space servers. */
 06956	  kinfo.nr_procs = NR_PROCS;
 06957	  kinfo.nr_tasks = NR_TASKS;
 06958	  strncpy(kinfo.release, OS_RELEASE, sizeof(kinfo.release));
 06959	  kinfo.release[sizeof(kinfo.release)-1] = '\0';
 06960	  strncpy(kinfo.version, OS_VERSION, sizeof(kinfo.version));
 06961	  kinfo.version[sizeof(kinfo.version)-1] = '\0';
 06962	  kinfo.proc_addr = (vir_bytes) proc;
 06963	  kinfo.kmem_base = vir2phys(0);
 06964	  kinfo.kmem_size = (phys_bytes) &end;  
 06965	
 06966	  /* Processor?  86, 186, 286, 386, ... 
 06967	   * Decide if mode is protected for older machines. 
 06968	   */
 06969	  machine.processor=atoi(get_value(params, "processor")); 
 06970	#if _WORD_SIZE == 2
 06971	  machine.protected = machine.processor >= 286;         
 06972	#endif
 06973	  if (! machine.protected) mon_return = 0;
 06974	
 06975	  /* XT, AT or MCA bus? */
 06976	  value = get_value(params, "bus");
 06977	  if (value == NIL_PTR || strcmp(value, "at") == 0) {
 06978	      machine.pc_at = TRUE;                     /* PC-AT compatible hardware */
 06979	  } else if (strcmp(value, "mca") == 0) {
 06980	      machine.pc_at = machine.ps_mca = TRUE;    /* PS/2 with micro channel */
 06981	  }
 06982	
 06983	  /* Type of VDU: */
 06984	  value = get_value(params, "video");           /* EGA or VGA video unit */
 06985	  if (strcmp(value, "ega") == 0) machine.vdu_ega = TRUE;
 06986	  if (strcmp(value, "vga") == 0) machine.vdu_vga = machine.vdu_ega = TRUE;
 06987	
 06988	  /* Return to assembler code to switch to protected mode (if 286), 
 06989	   * reload selectors and call main().
 06990	   */
 06991	}
	
 06993	/*===========================================================================*
 06994	 *                              get_value                                    *
 06995	 *===========================================================================*/
 06996	
 06997	PRIVATE char *get_value(params, name)
 06998	_CONST char *params;                            /* boot monitor parameters */
 06999	_CONST char *name;                              /* key to look up */
 07000	{
 07001	/* Get environment value - kernel version of getenv to avoid setting up the
 07002	 * usual environment array.
 07003	 */
 07004	  register _CONST char *namep;
 07005	  register char *envp;
 07006	
 07007	  for (envp = (char *) params; *envp != 0;) {
 07008	        for (namep = name; *namep != 0 && *namep == *envp; namep++, envp++)
 07009	                ;

_________________________ Page 718    File: kernel/start.c _________________________
 07010	        if (*namep == '\0' && *envp == '=') return(envp + 1);
 07011	        while (*envp++ != 0)
 07012	                ;
 07013	  }
 07014	  return(NIL_PTR);
 07015	}





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/main.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 07100	/* This file contains the main program of MINIX as well as its shutdown code.
 07101	 * The routine main() initializes the system and starts the ball rolling by
 07102	 * setting up the process table, interrupt vectors, and scheduling each task 
 07103	 * to run to initialize itself.
 07104	 * The routine shutdown() does the opposite and brings down MINIX. 
 07105	 *
 07106	 * The entries into this file are:
 07107	 *   main:              MINIX main program
 07108	 *   prepare_shutdown:  prepare to take MINIX down
 07109	 *
 07110	 * Changes:
 07111	 *   Nov 24, 2004   simplified main() with system image  (Jorrit N. Herder)
 07112	 *   Aug 20, 2004   new prepare_shutdown() and shutdown()  (Jorrit N. Herder)
 07113	 */
 07114	#include "kernel.h"
 07115	#include <signal.h>
 07116	#include <string.h>
 07117	#include <unistd.h>
 07118	#include <a.out.h>
 07119	#include <minix/callnr.h>
 07120	#include <minix/com.h>
 07121	#include "proc.h"
 07122	
 07123	/* Prototype declarations for PRIVATE functions. */
 07124	FORWARD _PROTOTYPE( void announce, (void));     
 07125	FORWARD _PROTOTYPE( void shutdown, (timer_t *tp));
 07126	
 07127	/*===========================================================================*
 07128	 *                              main                                         *
 07129	 *===========================================================================*/
 07130	PUBLIC void main()
 07131	{
 07132	/* Start the ball rolling. */
 07133	  struct boot_image *ip;        /* boot image pointer */
 07134	  register struct proc *rp;     /* process pointer */
 07135	  register struct priv *sp;     /* privilege structure pointer */
 07136	  register int i, s;
 07137	  int hdrindex;                 /* index to array of a.out headers */
 07138	  phys_clicks text_base;
 07139	  vir_clicks text_clicks, data_clicks;
 07140	  reg_t ktsb;                   /* kernel task stack base */
 07141	  struct exec e_hdr;            /* for a copy of an a.out header */
 07142	
 07143	  /* Initialize the interrupt controller. */
 07144	  intr_init(1);

_________________________ Page 719    File: kernel/main.c _________________________
 07145	
 07146	  /* Clear the process table. Anounce each slot as empty and set up mappings 
 07147	   * for proc_addr() and proc_nr() macros. Do the same for the table with 
 07148	   * privilege structures for the system processes. 
 07149	   */
 07150	  for (rp = BEG_PROC_ADDR, i = -NR_TASKS; rp < END_PROC_ADDR; ++rp, ++i) {
 07151	        rp->p_rts_flags = SLOT_FREE;            /* initialize free slot */
 07152	        rp->p_nr = i;                           /* proc number from ptr */
 07153	        (pproc_addr + NR_TASKS)[i] = rp;        /* proc ptr from number */
 07154	  }
 07155	  for (sp = BEG_PRIV_ADDR, i = 0; sp < END_PRIV_ADDR; ++sp, ++i) {
 07156	        sp->s_proc_nr = NONE;                   /* initialize as free */
 07157	        sp->s_id = i;                           /* priv structure index */
 07158	        ppriv_addr[i] = sp;                     /* priv ptr from number */
 07159	  }
 07160	
 07161	  /* Set up proc table entries for tasks and servers.  The stacks of the
 07162	   * kernel tasks are initialized to an array in data space.  The stacks
 07163	   * of the servers have been added to the data segment by the monitor, so
 07164	   * the stack pointer is set to the end of the data segment.  All the
 07165	   * processes are in low memory on the 8086.  On the 386 only the kernel
 07166	   * is in low memory, the rest is loaded in extended memory.
 07167	   */
 07168	
 07169	  /* Task stacks. */
 07170	  ktsb = (reg_t) t_stack;
 07171	
 07172	  for (i=0; i < NR_BOOT_PROCS; ++i) {
 07173	        ip = &image[i];                         /* process' attributes */
 07174	        rp = proc_addr(ip->proc_nr);            /* get process pointer */
 07175	        rp->p_max_priority = ip->priority;      /* max scheduling priority */
 07176	        rp->p_priority = ip->priority;          /* current priority */
 07177	        rp->p_quantum_size = ip->quantum;       /* quantum size in ticks */
 07178	        rp->p_ticks_left = ip->quantum;         /* current credit */
 07179	        strncpy(rp->p_name, ip->proc_name, P_NAME_LEN); /* set process name */
 07180	        (void) get_priv(rp, (ip->flags & SYS_PROC));    /* assign structure */
 07181	        priv(rp)->s_flags = ip->flags;                  /* process flags */
 07182	        priv(rp)->s_trap_mask = ip->trap_mask;          /* allowed traps */
 07183	        priv(rp)->s_call_mask = ip->call_mask;          /* kernel call mask */
 07184	        priv(rp)->s_ipc_to.chunk[0] = ip->ipc_to;       /* restrict targets */
 07185	        if (iskerneln(proc_nr(rp))) {           /* part of the kernel? */ 
 07186	                if (ip->stksize > 0) {          /* HARDWARE stack size is 0 */
 07187	                        rp->p_priv->s_stack_guard = (reg_t *) ktsb;
 07188	                        *rp->p_priv->s_stack_guard = STACK_GUARD;
 07189	                }
 07190	                ktsb += ip->stksize;    /* point to high end of stack */
 07191	                rp->p_reg.sp = ktsb;    /* this task's initial stack ptr */
 07192	                text_base = kinfo.code_base >> CLICK_SHIFT;
 07193	                                        /* processes that are in the kernel */
 07194	                hdrindex = 0;           /* all use the first a.out header */
 07195	        } else {
 07196	                hdrindex = 1 + i-NR_TASKS;      /* servers, drivers, INIT */
 07197	        }
 07198	
 07199	        /* The bootstrap loader created an array of the a.out headers at
 07200	         * absolute address 'aout'. Get one element to e_hdr.
 07201	         */
 07202	        phys_copy(aout + hdrindex * A_MINHDR, vir2phys(&e_hdr),
 07203	                                                (phys_bytes) A_MINHDR);
 07204	        /* Convert addresses to clicks and build process memory map */

_________________________ Page 720    File: kernel/main.c _________________________
 07205	        text_base = e_hdr.a_syms >> CLICK_SHIFT;
 07206	        text_clicks = (e_hdr.a_text + CLICK_SIZE-1) >> CLICK_SHIFT;
 07207	        if (!(e_hdr.a_flags & A_SEP)) text_clicks = 0;     /* common I&D */
 07208	        data_clicks = (e_hdr.a_total + CLICK_SIZE-1) >> CLICK_SHIFT;
 07209	        rp->p_memmap[T].mem_phys = text_base;
 07210	        rp->p_memmap[T].mem_len  = text_clicks;
 07211	        rp->p_memmap[D].mem_phys = text_base + text_clicks;
 07212	        rp->p_memmap[D].mem_len  = data_clicks;
 07213	        rp->p_memmap[S].mem_phys = text_base + text_clicks + data_clicks;
 07214	        rp->p_memmap[S].mem_vir  = data_clicks; /* empty - stack is in data */
 07215	
 07216	        /* Set initial register values.  The processor status word for tasks 
 07217	         * is different from that of other processes because tasks can
 07218	         * access I/O; this is not allowed to less-privileged processes 
 07219	         */
 07220	        rp->p_reg.pc = (reg_t) ip->initial_pc;
 07221	        rp->p_reg.psw = (iskernelp(rp)) ? INIT_TASK_PSW : INIT_PSW;
 07222	
 07223	        /* Initialize the server stack pointer. Take it down one word
 07224	         * to give crtso.s something to use as "argc".
 07225	         */
 07226	        if (isusern(proc_nr(rp))) {             /* user-space process? */ 
 07227	                rp->p_reg.sp = (rp->p_memmap[S].mem_vir +
 07228	                                rp->p_memmap[S].mem_len) << CLICK_SHIFT;
 07229	                rp->p_reg.sp -= sizeof(reg_t);
 07230	        }
 07231	        
 07232	        /* Set ready. The HARDWARE task is never ready. */
 07233	        if (rp->p_nr != HARDWARE) {
 07234	                rp->p_rts_flags = 0;            /* runnable if no flags */
 07235	                lock_enqueue(rp);               /* add to scheduling queues */
 07236	        } else {
 07237	                rp->p_rts_flags = NO_MAP;       /* prevent from running */
 07238	        }
 07239	
 07240	        /* Code and data segments must be allocated in protected mode. */
 07241	        alloc_segments(rp);
 07242	  }
 07243	
 07244	  /* We're definitely not shutting down. */
 07245	  shutdown_started = 0;
 07246	
 07247	  /* MINIX is now ready. All boot image processes are on the ready queue.
 07248	   * Return to the assembly code to start running the current process. 
 07249	   */
 07250	  bill_ptr = proc_addr(IDLE);           /* it has to point somewhere */
 07251	  announce();                           /* print MINIX startup banner */
 07252	  restart();
 07253	}
	
 07255	/*===========================================================================*
 07256	 *                              announce                                     *
 07257	 *===========================================================================*/
 07258	PRIVATE void announce(void)
 07259	{
 07260	  /* Display the MINIX startup banner. */
 07261	  kprintf("MINIX %s.%s."  
 07262	      "Copyright 2006, Vrije Universiteit, Amsterdam, The Netherlands\n", 
 07263	      OS_RELEASE, OS_VERSION);
 07264	

_________________________ Page 721    File: kernel/main.c _________________________
 07265	  /* Real mode, or 16/32-bit protected mode? */
 07266	  kprintf("Executing in %s mode.\n\n",
 07267	      machine.protected ? "32-bit protected" : "real");
 07268	}
	
 07270	/*===========================================================================*
 07271	 *                              prepare_shutdown                             *
 07272	 *===========================================================================*/
 07273	PUBLIC void prepare_shutdown(how)
 07274	int how;
 07275	{
 07276	/* This function prepares to shutdown MINIX. */
 07277	  static timer_t shutdown_timer;
 07278	  register struct proc *rp; 
 07279	  message m;
 07280	
 07281	  /* Show debugging dumps on panics. Make sure that the TTY task is still 
 07282	   * available to handle them. This is done with help of a non-blocking send. 
 07283	   * We rely on TTY to call sys_abort() when it is done with the dumps.
 07284	   */
 07285	  if (how == RBT_PANIC) {
 07286	      m.m_type = PANIC_DUMPS;
 07287	      if (nb_send(TTY_PROC_NR,&m)==OK)  /* don't block if TTY isn't ready */
 07288	          return;                       /* await sys_abort() from TTY */
 07289	  }
 07290	
 07291	  /* Send a signal to all system processes that are still alive to inform 
 07292	   * them that the MINIX kernel is shutting down. A proper shutdown sequence
 07293	   * should be implemented by a user-space server. This mechanism is useful
 07294	   * as a backup in case of system panics, so that system processes can still
 07295	   * run their shutdown code, e.g, to synchronize the FS or to let the TTY
 07296	   * switch to the first console. 
 07297	   */
 07298	  kprintf("Sending SIGKSTOP to system processes ...\n"); 
 07299	  for (rp=BEG_PROC_ADDR; rp<END_PROC_ADDR; rp++) {
 07300	      if (!isemptyp(rp) && (priv(rp)->s_flags & SYS_PROC) && !i
 07301	          send_sig(proc_nr(rp), SIGKSTOP);
 07302	  }
 07303	
 07304	  /* We're shutting down. Diagnostics may behave differently now. */
 07305	  shutdown_started = 1;
 07306	
 07307	  /* Notify system processes of the upcoming shutdown and allow them to be 
 07308	   * scheduled by setting a watchog timer that calls shutdown(). The timer 
 07309	   * argument passes the shutdown status. 
 07310	   */
 07311	  kprintf("MINIX will now be shut down ...\n");
 07312	  tmr_arg(&shutdown_timer)->ta_int = how;
 07313	
 07314	  /* Continue after 1 second, to give processes a chance to get
 07315	   * scheduled to do shutdown work.
 07316	   */
 07317	  set_timer(&shutdown_timer, get_uptime() + HZ, shutdown);
 07318	}
	
 07320	/*===========================================================================*
 07321	 *                              shutdown                                     *
 07322	 *===========================================================================*/
 07323	PRIVATE void shutdown(tp)
 07324	timer_t *tp;

_________________________ Page 722    File: kernel/main.c _________________________
 07325	{
 07326	/* This function is called from prepare_shutdown or stop_sequence to bring 
 07327	 * down MINIX. How to shutdown is in the argument: RBT_HALT (return to the
 07328	 * monitor), RBT_MONITOR (execute given code), RBT_RESET (hard reset). 
 07329	 */
 07330	  int how = tmr_arg(tp)->ta_int;
 07331	  u16_t magic; 
 07332	
 07333	  /* Now mask all interrupts, including the clock, and stop the clock. */
 07334	  outb(INT_CTLMASK, ~0); 
 07335	  clock_stop();
 07336	
 07337	  if (mon_return && how != RBT_RESET) {
 07338	        /* Reinitialize the interrupt controllers to the BIOS defaults. */
 07339	        intr_init(0);
 07340	        outb(INT_CTLMASK, 0);
 07341	        outb(INT2_CTLMASK, 0);
 07342	
 07343	        /* Return to the boot monitor. Set the program if not already done. */
 07344	        if (how != RBT_MONITOR) phys_copy(vir2phys(""), kinfo.params_base, 1); 
 07345	        level0(monitor);
 07346	  }
 07347	
 07348	  /* Reset the system by jumping to the reset address (real mode), or by
 07349	   * forcing a processor shutdown (protected mode). First stop the BIOS 
 07350	   * memory test by setting a soft reset flag. 
 07351	   */
 07352	  magic = STOP_MEM_CHECK;
 07353	  phys_copy(vir2phys(&magic), SOFT_RESET_FLAG_ADDR, SOFT_RESET_FLAG_SIZE);
 07354	  level0(reset);
 07355	}
	




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/proc.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 07400	/* This file contains essentially all of the process and message handling.
 07401	 * Together with "mpx.s" it forms the lowest layer of the MINIX kernel.
 07402	 * There is one entry point from the outside:
 07403	 *
 07404	 *   sys_call:        a system call, i.e., the kernel is trapped with an INT
 07405	 *
 07406	 * As well as several entry points used from the interrupt and task level:
 07407	 *
 07408	 *   lock_notify:     notify a process of a system event
 07409	 *   lock_send:       send a message to a process
 07410	 *   lock_enqueue:    put a process on one of the scheduling queues 
 07411	 *   lock_dequeue:    remove a process from the scheduling queues
 07412	 *
 07413	 * Changes:
 07414	 *   Aug 19, 2005     rewrote scheduling code  (Jorrit N. Herder)
 07415	 *   Jul 25, 2005     rewrote system call handling  (Jorrit N. Herder)
 07416	 *   May 26, 2005     rewrote message passing functions  (Jorrit N. Herder)
 07417	 *   May 24, 2005     new notification system call  (Jorrit N. Herder)
 07418	 *   Oct 28, 2004     nonblocking send and receive calls  (Jorrit N. Herder)
 07419	 *

_________________________ Page 723    File: kernel/proc.c _________________________
 07420	 * The code here is critical to make everything work and is important for the
 07421	 * overall performance of the system. A large fraction of the code deals with
 07422	 * list manipulation. To make this both easy to understand and fast to execute 
 07423	 * pointer pointers are used throughout the code. Pointer pointers prevent
 07424	 * exceptions for the head or tail of a linked list. 
 07425	 *
 07426	 *  node_t *queue, *new_node;   // assume these as global variables
 07427	 *  node_t **xpp = &queue;      // get pointer pointer to head of queue 
 07428	 *  while (*xpp != NULL)        // find last pointer of the linked list
 07429	 *      xpp = &(*xpp)->next;    // get pointer to next pointer 
 07430	 *  *xpp = new_node;            // now replace the end (the NULL pointer) 
 07431	 *  new_node->next = NULL;      // and mark the new end of the list
 07432	 * 
 07433	 * For example, when adding a new node to the end of the list, one normally 
 07434	 * makes an exception for an empty list and looks up the end of the list for 
 07435	 * nonempty lists. As shown above, this is not required with pointer pointers.
 07436	 */
 07437	
 07438	#include <minix/com.h>
 07439	#include <minix/callnr.h>
 07440	#include "kernel.h"
 07441	#include "proc.h"
 07442	
 07443	/* Scheduling and message passing functions. The functions are available to 
 07444	 * other parts of the kernel through lock_...(). The lock temporarily disables 
 07445	 * interrupts to prevent race conditions. 
 07446	 */
 07447	FORWARD _PROTOTYPE( int mini_send, (struct proc *caller_ptr, int dst,
 07448	                message *m_ptr, unsigned flags) );
 07449	FORWARD _PROTOTYPE( int mini_receive, (struct proc *caller_ptr, int src,
 07450	                message *m_ptr, unsigned flags) );
 07451	FORWARD _PROTOTYPE( int mini_notify, (struct proc *caller_ptr, int dst) );
 07452	
 07453	FORWARD _PROTOTYPE( void enqueue, (struct proc *rp) );
 07454	FORWARD _PROTOTYPE( void dequeue, (struct proc *rp) );
 07455	FORWARD _PROTOTYPE( void sched, (struct proc *rp, int *queue, int *front) );
 07456	FORWARD _PROTOTYPE( void pick_proc, (void) );
 07457	
 07458	#define BuildMess(m_ptr, src, dst_ptr) \
 07459	        (m_ptr)->m_source = (src);                                      \
 07460	        (m_ptr)->m_type = NOTIFY_FROM(src);                             \
 07461	        (m_ptr)->NOTIFY_TIMESTAMP = get_uptime();                       \
 07462	        switch (src) {                                                  \
 07463	        case HARDWARE:                                                  \
 07464	                (m_ptr)->NOTIFY_ARG = priv(dst_ptr)->s_int_pending;     \
 07465	                priv(dst_ptr)->s_int_pending = 0;                       \
 07466	                break;                                                  \
 07467	        case SYSTEM:                                                    \
 07468	                (m_ptr)->NOTIFY_ARG = priv(dst_ptr)->s_sig_pending;     \
 07469	                priv(dst_ptr)->s_sig_pending = 0;                       \
 07470	                break;                                                  \
 07471	        }
 07472	
 07473	#define CopyMess(s,sp,sm,dp,dm) \
 07474	        cp_mess(s, (sp)->p_memmap[D].mem_phys,  \
 07475	                 (vir_bytes)sm, (dp)->p_memmap[D].mem_phys, (vir_bytes)dm)
 07476	

_________________________ Page 724    File: kernel/proc.c _________________________
 07477	/*===========================================================================*
 07478	 *                              sys_call                                     * 
 07479	 *===========================================================================*/
 07480	PUBLIC int sys_call(call_nr, src_dst, m_ptr)
 07481	int call_nr;                    /* system call number and flags */
 07482	int src_dst;                    /* src to receive from or dst to send to */
 07483	message *m_ptr;                 /* pointer to message in the caller's space */
 07484	{
 07485	/* System calls are done by trapping to the kernel with an INT instruction.
 07486	 * The trap is caught and sys_call() is called to send or receive a message
 07487	 * (or both). The caller is always given by 'proc_ptr'.
 07488	 */
 07489	  register struct proc *caller_ptr = proc_ptr;  /* get pointer to caller */
 07490	  int function = call_nr & SYSCALL_FUNC;        /* get system call function */
 07491	  unsigned flags = call_nr & SYSCALL_FLAGS;     /* get flags */
 07492	  int mask_entry;                               /* bit to check in send mask */
 07493	  int result;                                   /* the system call's result */
 07494	  vir_clicks vlo, vhi;          /* virtual clicks containing message to send */
 07495	
 07496	  /* Check if the process has privileges for the requested call. Calls to the 
 07497	   * kernel may only be SENDREC, because tasks always reply and may not block 
 07498	   * if the caller doesn't do receive(). 
 07499	   */
 07500	  if (! (priv(caller_ptr)->s_trap_mask & (1 << function)) || 
 07501	          (iskerneln(src_dst) && function != SENDREC
 07502	           && function != RECEIVE)) { 
 07503	      kprintf("sys_call: trap %d not allowed, caller %d, src_dst %d\n", 
 07504	          function, proc_nr(caller_ptr), src_dst);
 07505	      return(ECALLDENIED);              /* trap denied by mask or kernel */
 07506	  }
 07507	  
 07508	  /* Require a valid source and/ or destination process, unless echoing. */
 07509	  if (! (isokprocn(src_dst) || src_dst == ANY || function == ECHO)) { 
 07510	      kprintf("sys_call: invalid src_dst, src_dst %d, caller %d\n", 
 07511	          src_dst, proc_nr(caller_ptr));
 07512	      return(EBADSRCDST);               /* invalid process number */
 07513	  }
 07514	
 07515	  /* If the call involves a message buffer, i.e., for SEND, RECEIVE, SENDREC, 
 07516	   * or ECHO, check the message pointer. This check allows a message to be 
 07517	   * anywhere in data or stack or gap. It will have to be made more elaborate 
 07518	   * for machines which don't have the gap mapped. 
 07519	   */
 07520	  if (function & CHECK_PTR) {   
 07521	      vlo = (vir_bytes) m_ptr >> CLICK_SHIFT;           
 07522	      vhi = ((vir_bytes) m_ptr + MESS_SIZE - 1) >> CLICK_SHIFT;
 07523	      if (vlo < caller_ptr->p_memmap[D].mem_vir || vlo > vhi ||
 07524	              vhi >= caller_ptr->p_memmap[S].mem_vir + 
 07525	              caller_ptr->p_memmap[S].mem_len) {
 07526	          kprintf("sys_call: invalid message pointer, trap %d, caller %d\n",
 07527	                function, proc_nr(caller_ptr));
 07528	          return(EFAULT);               /* invalid message pointer */
 07529	      }
 07530	  }
 07531	
 07532	  /* If the call is to send to a process, i.e., for SEND, SENDREC or NOTIFY,
 07533	   * verify that the caller is allowed to send to the given destination and
 07534	   * that the destination is still alive. 
 07535	   */
 07536	  if (function & CHECK_DST) {   

_________________________ Page 725    File: kernel/proc.c _________________________
 07537	      if (! get_sys_bit(priv(caller_ptr)->s_ipc_to, nr_to_id(src_dst))) {
 07538	          kprintf("sys_call: ipc mask denied %d sending to %d\n",
 07539	                proc_nr(caller_ptr), src_dst);
 07540	          return(ECALLDENIED);          /* call denied by ipc mask */
 07541	      }
 07542	
 07543	      if (isemptyn(src_dst) && !shutdown_started) {
 07544	          kprintf("sys_call: dead dest; %d, %d, %d\n", 
 07545	              function, proc_nr(caller_ptr), src_dst);
 07546	          return(EDEADDST);             /* cannot send to the dead */
 07547	      }
 07548	  }
 07549	
 07550	  /* Now check if the call is known and try to perform the request. The only
 07551	   * system calls that exist in MINIX are sending and receiving messages.
 07552	   *   - SENDREC: combines SEND and RECEIVE in a single system call
 07553	   *   - SEND:    sender blocks until its message has been delivered
 07554	   *   - RECEIVE: receiver blocks until an acceptable message has arrived
 07555	   *   - NOTIFY:  nonblocking call; deliver notification or mark pending
 07556	   *   - ECHO:    nonblocking call; directly echo back the message 
 07557	   */
 07558	  switch(function) {
 07559	  case SENDREC:
 07560	      /* A flag is set so that notifications cannot interrupt SENDREC. */
 07561	      priv(caller_ptr)->s_flags |= SENDREC_BUSY;
 07562	      /* fall through */
 07563	  case SEND:                    
 07564	      result = mini_send(caller_ptr, src_dst, m_ptr, flags);
 07565	      if (function == SEND || result != OK) {   
 07566	          break;                                /* done, or SEND failed */
 07567	      }                                         /* fall through for SENDREC */
 07568	  case RECEIVE:                 
 07569	      if (function == RECEIVE)
 07570	          priv(caller_ptr)->s_flags &= ~SENDREC_BUSY;
 07571	      result = mini_receive(caller_ptr, src_dst, m_ptr, flags);
 07572	      break;
 07573	  case NOTIFY:
 07574	      result = mini_notify(caller_ptr, src_dst);
 07575	      break;
 07576	  case ECHO:
 07577	      CopyMess(caller_ptr->p_nr, caller_ptr, m_ptr, caller_ptr, m_ptr);
 07578	      result = OK;
 07579	      break;
 07580	  default:
 07581	      result = EBADCALL;                        /* illegal system call */
 07582	  }
 07583	
 07584	  /* Now, return the result of the system call to the caller. */
 07585	  return(result);
 07586	}
	
 07588	/*===========================================================================*
 07589	 *                              mini_send                                    * 
 07590	 *===========================================================================*/
 07591	PRIVATE int mini_send(caller_ptr, dst, m_ptr, flags)
 07592	register struct proc *caller_ptr;       /* who is trying to send a message? */
 07593	int dst;                                /* to whom is message being sent? */
 07594	message *m_ptr;                         /* pointer to message buffer */
 07595	unsigned flags;                         /* system call flags */
 07596	{

_________________________ Page 726    File: kernel/proc.c _________________________
 07597	/* Send a message from 'caller_ptr' to 'dst'. If 'dst' is blocked waiting
 07598	 * for this message, copy the message to it and unblock 'dst'. If 'dst' is
 07599	 * not waiting at all, or is waiting for another source, queue 'caller_ptr'.
 07600	 */
 07601	  register struct proc *dst_ptr = proc_addr(dst);
 07602	  register struct proc **xpp;
 07603	  register struct proc *xp;
 07604	
 07605	  /* Check for deadlock by 'caller_ptr' and 'dst' sending to each other. */
 07606	  xp = dst_ptr;
 07607	  while (xp->p_rts_flags & SENDING) {           /* check while sending */
 07608	        xp = proc_addr(xp->p_sendto);           /* get xp's destination */
 07609	        if (xp == caller_ptr) return(ELOCKED);  /* deadlock if cyclic */
 07610	  }
 07611	
 07612	  /* Check if 'dst' is blocked waiting for this message. The destination's 
 07613	   * SENDING flag may be set when its SENDREC call blocked while sending.  
 07614	   */
 07615	  if ( (dst_ptr->p_rts_flags & (RECEIVING | SENDING)) == RECEIVING &&
 07616	       (dst_ptr->p_getfrom == ANY || dst_ptr->p_getfrom == caller_ptr->p_nr)) {
 07617	        /* Destination is indeed waiting for this message. */
 07618	        CopyMess(caller_ptr->p_nr, caller_ptr, m_ptr, dst_ptr,
 07619	                 dst_ptr->p_messbuf);
 07620	        if ((dst_ptr->p_rts_flags &= ~RECEIVING) == 0) enqueue(dst_ptr);
 07621	  } else if ( ! (flags & NON_BLOCKING)) {
 07622	        /* Destination is not waiting.  Block and dequeue caller. */
 07623	        caller_ptr->p_messbuf = m_ptr;
 07624	        if (caller_ptr->p_rts_flags == 0) dequeue(caller_ptr);
 07625	        caller_ptr->p_rts_flags |= SENDING;
 07626	        caller_ptr->p_sendto = dst;
 07627	
 07628	        /* Process is now blocked.  Put in on the destination's queue. */
 07629	        xpp = &dst_ptr->p_caller_q;             /* find end of list */
 07630	        while (*xpp != NIL_PROC) xpp = &(*xpp)->p_q_link;       
 07631	        *xpp = caller_ptr;                      /* add caller to end */
 07632	        caller_ptr->p_q_link = NIL_PROC;        /* mark new end of list */
 07633	  } else {
 07634	        return(ENOTREADY);
 07635	  }
 07636	  return(OK);
 07637	}
	
 07639	/*===========================================================================*
 07640	 *                              mini_receive                                 * 
 07641	 *===========================================================================*/
 07642	PRIVATE int mini_receive(caller_ptr, src, m_ptr, flags)
 07643	register struct proc *caller_ptr;       /* process trying to get message */
 07644	int src;                                /* which message source is wanted */
 07645	message *m_ptr;                         /* pointer to message buffer */
 07646	unsigned flags;                         /* system call flags */
 07647	{
 07648	/* A process or task wants to get a message.  If a message is already queued,
 07649	 * acquire it and deblock the sender.  If no message from the desired source
 07650	 * is available block the caller, unless the flags don't allow blocking.  
 07651	 */
 07652	  register struct proc **xpp;
 07653	  register struct notification **ntf_q_pp;
 07654	  message m;
 07655	  int bit_nr;
 07656	  sys_map_t *map;

_________________________ Page 727    File: kernel/proc.c _________________________
 07657	  bitchunk_t *chunk;
 07658	  int i, src_id, src_proc_nr;
 07659	
 07660	  /* Check to see if a message from desired source is already available.
 07661	   * The caller's SENDING flag may be set if SENDREC couldn't send. If it is
 07662	   * set, the process should be blocked.
 07663	   */
 07664	  if (!(caller_ptr->p_rts_flags & SENDING)) {
 07665	
 07666	    /* Check if there are pending notifications, except for SENDREC. */
 07667	    if (! (priv(caller_ptr)->s_flags & SENDREC_BUSY)) {
 07668	
 07669	        map = &priv(caller_ptr)->s_notify_pending;
 07670	        for (chunk=&map->chunk[0]; chunk<&map->chunk[NR_SYS_CHUNKS]; chu
 07671	
 07672	            /* Find a pending notification from the requested source. */ 
 07673	            if (! *chunk) continue;                     /* no bits in chunk */
 07674	            for (i=0; ! (*chunk & (1<<i)); ++i) {}      /* look up the bi
 07675	            src_id = (chunk - &map->chunk[0]) * BITCHUNK_BITS + i;
 07676	            if (src_id >= NR_SYS_PROCS) break;          /* out of range */
 07677	            src_proc_nr = id_to_nr(src_id);             /* get source proc */
 07678	            if (src!=ANY && src!=src_proc_nr) continue; /* source not ok */
 07679	            *chunk &= ~(1 << i);                        /* no longer pend
 07680	
 07681	            /* Found a suitable source, deliver the notification message. */
 07682	            BuildMess(&m, src_proc_nr, caller_ptr);     /* assemble message */
 07683	            CopyMess(src_proc_nr, proc_addr(HARDWARE), &m, caller_ptr, m_ptr);
 07684	            return(OK);                                 /* report success */
 07685	        }
 07686	    }
 07687	
 07688	    /* Check caller queue. Use pointer pointers to keep code simple. */
 07689	    xpp = &caller_ptr->p_caller_q;
 07690	    while (*xpp != NIL_PROC) {
 07691	        if (src == ANY || src == proc_nr(*xpp)) {
 07692	            /* Found acceptable message. Copy it and update status. */
 07693	            CopyMess((*xpp)->p_nr, *xpp, (*xpp)->p_messbuf, caller_ptr, m_ptr);
 07694	            if (((*xpp)->p_rts_flags &= ~SENDING) == 0) enqueue(*xpp);
 07695	            *xpp = (*xpp)->p_q_link;            /* remove from queue */
 07696	            return(OK);                         /* report success */
 07697	        }
 07698	        xpp = &(*xpp)->p_q_link;                /* proceed to next */
 07699	    }
 07700	  }
 07701	
 07702	  /* No suitable message is available or the caller couldn't send in SENDREC. 
 07703	   * Block the process trying to receive, unless the flags tell otherwise.
 07704	   */
 07705	  if ( ! (flags & NON_BLOCKING)) {
 07706	      caller_ptr->p_getfrom = src;              
 07707	      caller_ptr->p_messbuf = m_ptr;
 07708	      if (caller_ptr->p_rts_flags == 0) dequeue(caller_ptr);
 07709	      caller_ptr->p_rts_flags |= RECEIVING;             
 07710	      return(OK);
 07711	  } else {
 07712	      return(ENOTREADY);
 07713	  }
 07714	}
	

_________________________ Page 728    File: kernel/proc.c _________________________
 07716	/*===========================================================================*
 07717	 *                              mini_notify                                  * 
 07718	 *===========================================================================*/
 07719	PRIVATE int mini_notify(caller_ptr, dst)
 07720	register struct proc *caller_ptr;       /* sender of the notification */
 07721	int dst;                                /* which process to notify */
 07722	{
 07723	  register struct proc *dst_ptr = proc_addr(dst);
 07724	  int src_id;                           /* source id for late delivery */
 07725	  message m;                            /* the notification message */
 07726	
 07727	  /* Check to see if target is blocked waiting for this message. A process 
 07728	   * can be both sending and receiving during a SENDREC system call.
 07729	   */
 07730	  if ((dst_ptr->p_rts_flags & (RECEIVING|SENDING)) == RECEIVING &&
 07731	      ! (priv(dst_ptr)->s_flags & SENDREC_BUSY) &&
 07732	      (dst_ptr->p_getfrom == ANY || dst_ptr->p_getfrom == caller_ptr->p_nr)) {
 07733	
 07734	      /* Destination is indeed waiting for a message. Assemble a notification 
 07735	       * message and deliver it. Copy from pseudo-source HARDWARE, since the
 07736	       * message is in the kernel's address space.
 07737	       */ 
 07738	      BuildMess(&m, proc_nr(caller_ptr), dst_ptr);
 07739	      CopyMess(proc_nr(caller_ptr), proc_addr(HARDWARE), &m, 
 07740	          dst_ptr, dst_ptr->p_messbuf);
 07741	      dst_ptr->p_rts_flags &= ~RECEIVING;       /* deblock destination */
 07742	      if (dst_ptr->p_rts_flags == 0) enqueue(dst_ptr);
 07743	      return(OK);
 07744	  } 
 07745	
 07746	  /* Destination is not ready to receive the notification. Add it to the 
 07747	   * bit map with pending notifications. Note the indirectness: the system id 
 07748	   * instead of the process number is used in the pending bit map.
 07749	   */ 
 07750	  src_id = priv(caller_ptr)->s_id;
 07751	  set_sys_bit(priv(dst_ptr)->s_notify_pending, src_id); 
 07752	  return(OK);
 07753	}
	
 07755	/*===========================================================================*
 07756	 *                              lock_notify                                  *
 07757	 *===========================================================================*/
 07758	PUBLIC int lock_notify(src, dst)
 07759	int src;                        /* sender of the notification */
 07760	int dst;                        /* who is to be notified */
 07761	{
 07762	/* Safe gateway to mini_notify() for tasks and interrupt handlers. The sender
 07763	 * is explicitly given to prevent confusion where the call comes from. MINIX 
 07764	 * kernel is not reentrant, which means to interrupts are disabled after 
 07765	 * the first kernel entry (hardware interrupt, trap, or exception). Locking
 07766	 * is done by temporarily disabling interrupts. 
 07767	 */
 07768	  int result;
 07769	
 07770	  /* Exception or interrupt occurred, thus already locked. */
 07771	  if (k_reenter >= 0) {
 07772	      result = mini_notify(proc_addr(src), dst); 
 07773	  }
 07774	
 07775	  /* Call from task level, locking is required. */

_________________________ Page 729    File: kernel/proc.c _________________________
 07776	  else {
 07777	      lock(0, "notify");
 07778	      result = mini_notify(proc_addr(src), dst); 
 07779	      unlock(0);
 07780	  }
 07781	  return(result);
 07782	}
	
 07784	/*===========================================================================*
 07785	 *                              enqueue                                      * 
 07786	 *===========================================================================*/
 07787	PRIVATE void enqueue(rp)
 07788	register struct proc *rp;       /* this process is now runnable */
 07789	{
 07790	/* Add 'rp' to one of the queues of runnable processes.  This function is 
 07791	 * responsible for inserting a process into one of the scheduling queues. 
 07792	 * The mechanism is implemented here.   The actual scheduling policy is
 07793	 * defined in sched() and pick_proc().
 07794	 */
 07795	  int q;                                        /* scheduling queue to use */
 07796	  int front;                                    /* add to front or back */
 07797	
 07798	  /* Determine where to insert to process. */
 07799	  sched(rp, &q, &front);
 07800	
 07801	  /* Now add the process to the queue. */
 07802	  if (rdy_head[q] == NIL_PROC) {                /* add to empty queue */
 07803	      rdy_head[q] = rdy_tail[q] = rp;           /* create a new queue */
 07804	      rp->p_nextready = NIL_PROC;               /* mark new end */
 07805	  } 
 07806	  else if (front) {                             /* add to head of queue */
 07807	      rp->p_nextready = rdy_head[q];            /* chain head of queue */
 07808	      rdy_head[q] = rp;                         /* set new queue head */
 07809	  } 
 07810	  else {                                        /* add to tail of queue */
 07811	      rdy_tail[q]->p_nextready = rp;            /* chain tail of queue */       
 07812	      rdy_tail[q] = rp;                         /* set new queue tail */
 07813	      rp->p_nextready = NIL_PROC;               /* mark new end */
 07814	  }
 07815	
 07816	  /* Now select the next process to run. */
 07817	  pick_proc();                  
 07818	}
	
 07820	/*===========================================================================*
 07821	 *                              dequeue                                      * 
 07822	 *===========================================================================*/
 07823	PRIVATE void dequeue(rp)
 07824	register struct proc *rp;       /* this process is no longer runnable */
 07825	{
 07826	/* A process must be removed from the scheduling queues, for example, because
 07827	 * it has blocked.  If the currently active process is removed, a new process
 07828	 * is picked to run by calling pick_proc().
 07829	 */
 07830	  register int q = rp->p_priority;              /* queue to use */
 07831	  register struct proc **xpp;                   /* iterate over queue */
 07832	  register struct proc *prev_xp;
 07833	
 07834	  /* Side-effect for kernel: check if the task's stack still is ok? */
 07835	  if (iskernelp(rp)) {                          

_________________________ Page 730    File: kernel/proc.c _________________________
 07836	        if (*priv(rp)->s_stack_guard != STACK_GUARD)
 07837	                panic("stack overrun by task", proc_nr(rp));
 07838	  }
 07839	
 07840	  /* Now make sure that the process is not in its ready queue. Remove the 
 07841	   * process if it is found. A process can be made unready even if it is not 
 07842	   * running by being sent a signal that kills it.
 07843	   */
 07844	  prev_xp = NIL_PROC;                           
 07845	  for (xpp = &rdy_head[q]; *xpp != NIL_PROC; xpp = &(*xpp)->p_nextready) {
 07846	
 07847	      if (*xpp == rp) {                         /* found process to remove */
 07848	          *xpp = (*xpp)->p_nextready;           /* replace with next chain */
 07849	          if (rp == rdy_tail[q])                /* queue tail removed */
 07850	              rdy_tail[q] = prev_xp;            /* set new tail */
 07851	          if (rp == proc_ptr || rp == next_ptr) /* active process removed */
 07852	              pick_proc();                      /* pick new process to run */
 07853	          break;
 07854	      }
 07855	      prev_xp = *xpp;                           /* save previous in chain */
 07856	  }
 07857	}
	
 07859	/*===========================================================================*
 07860	 *                              sched                                        * 
 07861	 *===========================================================================*/
 07862	PRIVATE void sched(rp, queue, front)
 07863	register struct proc *rp;                       /* process to be scheduled */
 07864	int *queue;                                     /* return: queue to use */
 07865	int *front;                                     /* return: front or back */
 07866	{
 07867	/* This function determines the scheduling policy.  It is called whenever a
 07868	 * process must be added to one of the scheduling queues to decide where to
 07869	 * insert it.  As a side-effect the process' priority may be updated.  
 07870	 */
 07871	  static struct proc *prev_ptr = NIL_PROC;      /* previous without time */
 07872	  int time_left = (rp->p_ticks_left > 0);       /* quantum fully consumed */
 07873	  int penalty = 0;                              /* change in priority */
 07874	
 07875	  /* Check whether the process has time left. Otherwise give a new quantum 
 07876	   * and possibly raise the priority.  Processes using multiple quantums 
 07877	   * in a row get a lower priority to catch infinite loops in high priority
 07878	   * processes (system servers and drivers). 
 07879	   */
 07880	  if ( ! time_left) {                           /* quantum consumed ? */
 07881	      rp->p_ticks_left = rp->p_quantum_size;    /* give new quantum */
 07882	      if (prev_ptr == rp) penalty ++;           /* catch infinite loops */
 07883	      else penalty --;                          /* give slow way back */
 07884	      prev_ptr = rp;                            /* store ptr for next */
 07885	  }
 07886	
 07887	  /* Determine the new priority of this process. The bounds are determined
 07888	   * by IDLE's queue and the maximum priority of this process. Kernel tasks 
 07889	   * and the idle process are never changed in priority.
 07890	   */
 07891	  if (penalty != 0 && ! iskernelp(rp)) {
 07892	      rp->p_priority += penalty;                /* update with penalty */
 07893	      if (rp->p_priority < rp->p_max_priority)  /* check upper bound */ 
 07894	          rp->p_priority=rp->p_max_priority;
 07895	      else if (rp->p_priority > IDLE_Q-1)       /* check lower bound */

_________________________ Page 731    File: kernel/proc.c _________________________
 07896	          rp->p_priority = IDLE_Q-1;
 07897	  }
 07898	
 07899	  /* If there is time left, the process is added to the front of its queue, 
 07900	   * so that it can immediately run. The queue to use simply is always the
 07901	   * process' current priority. 
 07902	   */
 07903	  *queue = rp->p_priority;
 07904	  *front = time_left;
 07905	}
	
 07907	/*===========================================================================*
 07908	 *                              pick_proc                                    * 
 07909	 *===========================================================================*/
 07910	PRIVATE void pick_proc()
 07911	{
 07912	/* Decide who to run now.  A new process is selected by setting 'next_ptr'.
 07913	 * When a billable process is selected, record it in 'bill_ptr', so that the 
 07914	 * clock task can tell who to bill for system time.
 07915	 */
 07916	  register struct proc *rp;                     /* process to run */
 07917	  int q;                                        /* iterate over queues */
 07918	
 07919	  /* Check each of the scheduling queues for ready processes. The number of
 07920	   * queues is defined in proc.h, and priorities are set in the image table.
 07921	   * The lowest queue contains IDLE, which is always ready.
 07922	   */
 07923	  for (q=0; q < NR_SCHED_QUEUES; q++) { 
 07924	      if ( (rp = rdy_head[q]) != NIL_PROC) {
 07925	          next_ptr = rp;                        /* run process 'rp' next */
 07926	          if (priv(rp)->s_flags & BILLABLE)             
 07927	              bill_ptr = rp;                    /* bill for system time */
 07928	          return;                                
 07929	      }
 07930	  }
 07931	}
	
 07933	/*===========================================================================*
 07934	 *                              lock_send                                    *
 07935	 *===========================================================================*/
 07936	PUBLIC int lock_send(dst, m_ptr)
 07937	int dst;                        /* to whom is message being sent? */
 07938	message *m_ptr;                 /* pointer to message buffer */
 07939	{
 07940	/* Safe gateway to mini_send() for tasks. */
 07941	  int result;
 07942	  lock(2, "send");
 07943	  result = mini_send(proc_ptr, dst, m_ptr, NON_BLOCKING);
 07944	  unlock(2);
 07945	  return(result);
 07946	}
	
 07948	/*===========================================================================*
 07949	 *                              lock_enqueue                                 *
 07950	 *===========================================================================*/
 07951	PUBLIC void lock_enqueue(rp)
 07952	struct proc *rp;                /* this process is now runnable */
 07953	{
 07954	/* Safe gateway to enqueue() for tasks. */
 07955	  lock(3, "enqueue");

_________________________ Page 732    File: kernel/proc.c _________________________
 07956	  enqueue(rp);
 07957	  unlock(3);
 07958	}
	
 07960	/*===========================================================================*
 07961	 *                              lock_dequeue                                 *
 07962	 *===========================================================================*/
 07963	PUBLIC void lock_dequeue(rp)
 07964	struct proc *rp;                /* this process is no longer runnable */
 07965	{
 07966	/* Safe gateway to dequeue() for tasks. */
 07967	  lock(4, "dequeue");
 07968	  dequeue(rp);
 07969	  unlock(4);
 07970	}
	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/exception.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 08000	/* This file contains a simple exception handler.  Exceptions in user
 08001	 * processes are converted to signals. Exceptions in a kernel task cause
 08002	 * a panic.
 08003	 */
 08004	
 08005	#include "kernel.h"
 08006	#include <signal.h>
 08007	#include "proc.h"
 08008	
 08009	/*===========================================================================*
 08010	 *                              exception                                    *
 08011	 *===========================================================================*/
 08012	PUBLIC void exception(vec_nr)
 08013	unsigned vec_nr;
 08014	{
 08015	/* An exception or unexpected interrupt has occurred. */
 08016	
 08017	  struct ex_s {
 08018	        char *msg;
 08019	        int signum;
 08020	        int minprocessor;
 08021	  };
 08022	  static struct ex_s ex_data[] = {
 08023	        { "Divide error", SIGFPE, 86 },
 08024	        { "Debug exception", SIGTRAP, 86 },
 08025	        { "Nonmaskable interrupt", SIGBUS, 86 },
 08026	        { "Breakpoint", SIGEMT, 86 },
 08027	        { "Overflow", SIGFPE, 86 },
 08028	        { "Bounds check", SIGFPE, 186 },
 08029	        { "Invalid opcode", SIGILL, 186 },
 08030	        { "Coprocessor not available", SIGFPE, 186 },
 08031	        { "Double fault", SIGBUS, 286 },
 08032	        { "Copressor segment overrun", SIGSEGV, 286 },
 08033	        { "Invalid TSS", SIGSEGV, 286 },
 08034	        { "Segment not present", SIGSEGV, 286 },

_________________________ Page 733    File: kernel/exception.c _________________________
 08035	        { "Stack exception", SIGSEGV, 286 },    /* STACK_FAULT already used */
 08036	        { "General protection", SIGSEGV, 286 },
 08037	        { "Page fault", SIGSEGV, 386 },         /* not close */
 08038	        { NIL_PTR, SIGILL, 0 },                 /* probably software trap */
 08039	        { "Coprocessor error", SIGFPE, 386 },
 08040	  };
 08041	  register struct ex_s *ep;
 08042	  struct proc *saved_proc;
 08043	
 08044	  /* Save proc_ptr, because it may be changed by debug statements. */
 08045	  saved_proc = proc_ptr;        
 08046	
 08047	  ep = &ex_data[vec_nr];
 08048	
 08049	  if (vec_nr == 2) {            /* spurious NMI on some machines */
 08050	        kprintf("got spurious NMI\n");
 08051	        return;
 08052	  }
 08053	
 08054	  /* If an exception occurs while running a process, the k_reenter variable 
 08055	   * will be zero. Exceptions in interrupt handlers or system traps will make 
 08056	   * k_reenter larger than zero.
 08057	   */
 08058	  if (k_reenter == 0 && ! iskernelp(saved_proc)) {
 08059	        cause_sig(proc_nr(saved_proc), ep->signum);
 08060	        return;
 08061	  }
 08062	
 08063	  /* Exception in system code. This is not supposed to happen. */
 08064	  if (ep->msg == NIL_PTR || machine.processor < ep->minprocessor)
 08065	        kprintf("\nIntel-reserved exception %d\n", vec_nr);
 08066	  else
 08067	        kprintf("\n%s\n", ep->msg);
 08068	  kprintf("k_reenter = %d ", k_reenter);
 08069	  kprintf("process %d (%s), ", proc_nr(saved_proc), saved_proc->p_name);
 08070	  kprintf("pc = %u:0x%x", (unsigned) saved_proc->p_reg.cs,
 08071	  (unsigned) saved_proc->p_reg.pc);
 08072	
 08073	  panic("exception in a kernel task", NO_NUM);
 08074	}
	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/i8259.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 08100	/* This file contains routines for initializing the 8259 interrupt controller:
 08101	 *      put_irq_handler: register an interrupt handler
 08102	 *      rm_irq_handler: deregister an interrupt handler
 08103	 *      intr_handle:    handle a hardware interrupt
 08104	 *      intr_init:      initialize the interrupt controller(s)
 08105	 */
 08106	
 08107	#include "kernel.h"
 08108	#include "proc.h"
 08109	#include <minix/com.h>

_________________________ Page 734    File: kernel/i8259.c _________________________
 08110	
 08111	#define ICW1_AT         0x11    /* edge triggered, cascade, need ICW4 */
 08112	#define ICW1_PC         0x13    /* edge triggered, no cascade, need ICW4 */
 08113	#define ICW1_PS         0x19    /* level triggered, cascade, need ICW4 */
 08114	#define ICW4_AT_SLAVE   0x01    /* not SFNM, not buffered, normal EOI, 8086 */
 08115	#define ICW4_AT_MASTER  0x05    /* not SFNM, not buffered, normal EOI, 8086 */
 08116	#define ICW4_PC_SLAVE   0x09    /* not SFNM, buffered, normal EOI, 8086 */
 08117	#define ICW4_PC_MASTER  0x0D    /* not SFNM, buffered, normal EOI, 8086 */
 08118	
 08119	#define set_vec(nr, addr)       ((void)0)
 08120	
 08121	/*===========================================================================*
 08122	 *                              intr_init                                    *
 08123	 *===========================================================================*/
 08124	PUBLIC void intr_init(mine)
 08125	int mine;
 08126	{
 08127	/* Initialize the 8259s, finishing with all interrupts disabled.  This is
 08128	 * only done in protected mode, in real mode we don't touch the 8259s, but
 08129	 * use the BIOS locations instead.  The flag "mine" is set if the 8259s are
 08130	 * to be programmed for MINIX, or to be reset to what the BIOS expects.
 08131	 */
 08132	  int i;
 08133	
 08134	  intr_disable();
 08135	
 08136	      /* The AT and newer PS/2 have two interrupt controllers, one master,
 08137	       * one slaved at IRQ 2.  (We don't have to deal with the PC that
 08138	       * has just one controller, because it must run in real mode.)
 08139	       */
 08140	      outb(INT_CTL, machine.ps_mca ? ICW1_PS : ICW1_AT);
 08141	      outb(INT_CTLMASK, mine ? IRQ0_VECTOR : BIOS_IRQ0_VEC);
 08142	                                                        /* ICW2 for master */
 08143	      outb(INT_CTLMASK, (1 << CASCADE_IRQ));            /* ICW3 tells slaves 
 08144	      outb(INT_CTLMASK, ICW4_AT_MASTER);
 08145	      outb(INT_CTLMASK, ~(1 << CASCADE_IRQ));           /* IRQ 0-7 mask */
 08146	      outb(INT2_CTL, machine.ps_mca ? ICW1_PS : ICW1_AT);
 08147	      outb(INT2_CTLMASK, mine ? IRQ8_VECTOR : BIOS_IRQ8_VEC);
 08148	                                                        /* ICW2 for slave */
 08149	      outb(INT2_CTLMASK, CASCADE_IRQ);          /* ICW3 is slave nr */
 08150	      outb(INT2_CTLMASK, ICW4_AT_SLAVE);
 08151	      outb(INT2_CTLMASK, ~0);                           /* IRQ 8-15 mask */
 08152	
 08153	      /* Copy the BIOS vectors from the BIOS to the Minix location, so we
 08154	       * can still make BIOS calls without reprogramming the i8259s.
 08155	       */
 08156	      phys_copy(BIOS_VECTOR(0) * 4L, VECTOR(0) * 4L, 8 * 4L);
 08157	}
	
 08159	/*===========================================================================*
 08160	 *                              put_irq_handler                              *
 08161	 *===========================================================================*/
 08162	PUBLIC void put_irq_handler(hook, irq, handler)
 08163	irq_hook_t *hook;
 08164	int irq;
 08165	irq_handler_t handler;
 08166	{
 08167	/* Register an interrupt handler. */
 08168	  int id;
 08169	  irq_hook_t **line;

_________________________ Page 735    File: kernel/i8259.c _________________________
 08170	
 08171	  if (irq < 0 || irq >= NR_IRQ_VECTORS)
 08172	      panic("invalid call to put_irq_handler", irq);
 08173	
 08174	  line = &irq_handlers[irq];
 08175	  id = 1;
 08176	  while (*line != NULL) {
 08177	      if (hook == *line) return;        /* extra initialization */
 08178	      line = &(*line)->next;
 08179	      id <<= 1;
 08180	  }
 08181	  if (id == 0) panic("Too many handlers for irq", irq);
 08182	
 08183	  hook->next = NULL;
 08184	  hook->handler = handler;
 08185	  hook->irq = irq;
 08186	  hook->id = id;
 08187	  *line = hook;
 08188	
 08189	  irq_use |= 1 << irq;
 08190	}
	
 08192	/*===========================================================================*
 08193	 *                              rm_irq_handler                               *
 08194	 *===========================================================================*/
 08195	PUBLIC void rm_irq_handler(hook)
 08196	irq_hook_t *hook;
 08197	{
 08198	/* Unregister an interrupt handler. */
 08199	  int irq = hook->irq; 
 08200	  int id = hook->id;
 08201	  irq_hook_t **line;
 08202	
 08203	  if (irq < 0 || irq >= NR_IRQ_VECTORS) 
 08204	      panic("invalid call to rm_irq_handler", irq);
 08205	
 08206	  line = &irq_handlers[irq];
 08207	  while (*line != NULL) {
 08208	      if ((*line)->id == id) {
 08209	          (*line) = (*line)->next;
 08210	          if (! irq_handlers[irq]) irq_use &= ~(1 << irq);
 08211	          return;
 08212	      }
 08213	      line = &(*line)->next;
 08214	  }
 08215	  /* When the handler is not found, normally return here. */
 08216	}
	
 08218	/*===========================================================================*
 08219	 *                              intr_handle                                  *
 08220	 *===========================================================================*/
 08221	PUBLIC void intr_handle(hook)
 08222	irq_hook_t *hook;
 08223	{
 08224	/* Call the interrupt handlers for an interrupt with the given hook list.
 08225	 * The assembly part of the handler has already masked the IRQ, reenabled the
 08226	 * controller(s) and enabled interrupts.
 08227	 */
 08228	
 08229	  /* Call list of handlers for an IRQ. */

_________________________ Page 736    File: kernel/i8259.c _________________________
 08230	  while (hook != NULL) {
 08231	      /* For each handler in the list, mark it active by setting its ID bit,
 08232	       * call the function, and unmark it if the function returns true.
 08233	       */
 08234	      irq_actids[hook->irq] |= hook->id;
 08235	      if ((*hook->handler)(hook)) irq_actids[hook->irq] &= ~hook->id;
 08236	      hook = hook->next;
 08237	  }
 08238	
 08239	  /* The assembly code will now disable interrupts, unmask the IRQ if and only
 08240	   * if all active ID bits are cleared, and restart a process.
 08241	   */
 08242	}



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/protect.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 08300	/* This file contains code for initialization of protected mode, to initialize
 08301	 * code and data segment descriptors, and to initialize global descriptors
 08302	 * for local descriptors in the process table.
 08303	 */
 08304	
 08305	#include "kernel.h"
 08306	#include "proc.h"
 08307	#include "protect.h"
 08308	
 08309	#define INT_GATE_TYPE   (INT_286_GATE | DESC_386_BIT)
 08310	#define TSS_TYPE        (AVL_286_TSS  | DESC_386_BIT)
 08311	
 08312	struct desctableptr_s {
 08313	  char limit[sizeof(u16_t)];
 08314	  char base[sizeof(u32_t)];             /* really u24_t + pad for 286 */
 08315	};
 08316	
 08317	struct gatedesc_s {
 08318	  u16_t offset_low;
 08319	  u16_t selector;
 08320	  u8_t pad;                     /* |000|XXXXX| ig & trpg, |XXXXXXXX| task g */
 08321	  u8_t p_dpl_type;              /* |P|DL|0|TYPE| */
 08322	  u16_t offset_high;
 08323	};
 08324	
 08325	struct tss_s {
 08326	  reg_t backlink;
 08327	  reg_t sp0;                    /* stack pointer to use during interrupt */
 08328	  reg_t ss0;                    /*   "   segment  "  "    "        "     */
 08329	  reg_t sp1;
 08330	  reg_t ss1;
 08331	  reg_t sp2;
 08332	  reg_t ss2;
 08333	  reg_t cr3;
 08334	  reg_t ip;
 08335	  reg_t flags;
 08336	  reg_t ax;
 08337	  reg_t cx;
 08338	  reg_t dx;
 08339	  reg_t bx;

_________________________ Page 737    File: kernel/protect.c _________________________
 08340	  reg_t sp;
 08341	  reg_t bp;
 08342	  reg_t si;
 08343	  reg_t di;
 08344	  reg_t es;
 08345	  reg_t cs;
 08346	  reg_t ss;
 08347	  reg_t ds;
 08348	  reg_t fs;
 08349	  reg_t gs;
 08350	  reg_t ldt;
 08351	  u16_t trap;
 08352	  u16_t iobase;
 08353	/* u8_t iomap[0]; */
 08354	};
 08355	
 08356	PUBLIC struct segdesc_s gdt[GDT_SIZE];          /* used in klib.s and mpx.s */
 08357	PRIVATE struct gatedesc_s idt[IDT_SIZE];        /* zero-init so none present */
 08358	PUBLIC struct tss_s tss;                        /* zero init */
 08359	
 08360	FORWARD _PROTOTYPE( void int_gate, (unsigned vec_nr, vir_bytes offset,
 08361	                unsigned dpl_type) );
 08362	FORWARD _PROTOTYPE( void sdesc, (struct segdesc_s *segdp, phys_bytes base,
 08363	                vir_bytes size) );
 08364	
 08365	/*===========================================================================*
 08366	 *                              prot_init                                    *
 08367	 *===========================================================================*/
 08368	PUBLIC void prot_init()
 08369	{
 08370	/* Set up tables for protected mode.
 08371	 * All GDT slots are allocated at compile time.
 08372	 */
 08373	  struct gate_table_s *gtp;
 08374	  struct desctableptr_s *dtp;
 08375	  unsigned ldt_index;
 08376	  register struct proc *rp;
 08377	
 08378	  static struct gate_table_s {
 08379	        _PROTOTYPE( void (*gate), (void) );
 08380	        unsigned char vec_nr;
 08381	        unsigned char privilege;
 08382	  }
 08383	  gate_table[] = {
 08384	        { divide_error, DIVIDE_VECTOR, INTR_PRIVILEGE },
 08385	        { single_step_exception, DEBUG_VECTOR, INTR_PRIVILEGE },
 08386	        { nmi, NMI_VECTOR, INTR_PRIVILEGE },
 08387	        { breakpoint_exception, BREAKPOINT_VECTOR, USER_PRIVILEGE },
 08388	        { overflow, OVERFLOW_VECTOR, USER_PRIVILEGE },
 08389	        { bounds_check, BOUNDS_VECTOR, INTR_PRIVILEGE },
 08390	        { inval_opcode, INVAL_OP_VECTOR, INTR_PRIVILEGE },
 08391	        { copr_not_available, COPROC_NOT_VECTOR, INTR_PRIVILEGE },
 08392	        { double_fault, DOUBLE_FAULT_VECTOR, INTR_PRIVILEGE },
 08393	        { copr_seg_overrun, COPROC_SEG_VECTOR, INTR_PRIVILEGE },
 08394	        { inval_tss, INVAL_TSS_VECTOR, INTR_PRIVILEGE },
 08395	        { segment_not_present, SEG_NOT_VECTOR, INTR_PRIVILEGE },
 08396	        { stack_exception, STACK_FAULT_VECTOR, INTR_PRIVILEGE },
 08397	        { general_protection, PROTECTION_VECTOR, INTR_PRIVILEGE },
 08398	        { page_fault, PAGE_FAULT_VECTOR, INTR_PRIVILEGE },
 08399	        { copr_error, COPROC_ERR_VECTOR, INTR_PRIVILEGE },

_________________________ Page 738    File: kernel/protect.c _________________________
 08400	        { hwint00, VECTOR( 0), INTR_PRIVILEGE },
 08401	        { hwint01, VECTOR( 1), INTR_PRIVILEGE },
 08402	        { hwint02, VECTOR( 2), INTR_PRIVILEGE },
 08403	        { hwint03, VECTOR( 3), INTR_PRIVILEGE },
 08404	        { hwint04, VECTOR( 4), INTR_PRIVILEGE },
 08405	        { hwint05, VECTOR( 5), INTR_PRIVILEGE },
 08406	        { hwint06, VECTOR( 6), INTR_PRIVILEGE },
 08407	        { hwint07, VECTOR( 7), INTR_PRIVILEGE },
 08408	        { hwint08, VECTOR( 8), INTR_PRIVILEGE },
 08409	        { hwint09, VECTOR( 9), INTR_PRIVILEGE },
 08410	        { hwint10, VECTOR(10), INTR_PRIVILEGE },
 08411	        { hwint11, VECTOR(11), INTR_PRIVILEGE },
 08412	        { hwint12, VECTOR(12), INTR_PRIVILEGE },
 08413	        { hwint13, VECTOR(13), INTR_PRIVILEGE },
 08414	        { hwint14, VECTOR(14), INTR_PRIVILEGE },
 08415	        { hwint15, VECTOR(15), INTR_PRIVILEGE },
 08416	        { s_call, SYS386_VECTOR, USER_PRIVILEGE },      /* 386 system call */
 08417	        { level0_call, LEVEL0_VECTOR, TASK_PRIVILEGE },
 08418	  };
 08419	
 08420	  /* Build gdt and idt pointers in GDT where the BIOS expects them. */
 08421	  dtp= (struct desctableptr_s *) &gdt[GDT_INDEX];
 08422	  * (u16_t *) dtp->limit = (sizeof gdt) - 1;
 08423	  * (u32_t *) dtp->base = vir2phys(gdt);
 08424	
 08425	  dtp= (struct desctableptr_s *) &gdt[IDT_INDEX];
 08426	  * (u16_t *) dtp->limit = (sizeof idt) - 1;
 08427	  * (u32_t *) dtp->base = vir2phys(idt);
 08428	
 08429	  /* Build segment descriptors for tasks and interrupt handlers. */
 08430	  init_codeseg(&gdt[CS_INDEX],
 08431	         kinfo.code_base, kinfo.code_size, INTR_PRIVILEGE);
 08432	  init_dataseg(&gdt[DS_INDEX],
 08433	         kinfo.data_base, kinfo.data_size, INTR_PRIVILEGE);
 08434	  init_dataseg(&gdt[ES_INDEX], 0L, 0, TASK_PRIVILEGE);
 08435	
 08436	  /* Build scratch descriptors for functions in klib88. */
 08437	  init_dataseg(&gdt[DS_286_INDEX], 0L, 0, TASK_PRIVILEGE);
 08438	  init_dataseg(&gdt[ES_286_INDEX], 0L, 0, TASK_PRIVILEGE);
 08439	
 08440	  /* Build local descriptors in GDT for LDT's in process table.
 08441	   * The LDT's are allocated at compile time in the process table, and
 08442	   * initialized whenever a process' map is initialized or changed.
 08443	   */
 08444	  for (rp = BEG_PROC_ADDR, ldt_index = FIRST_LDT_INDEX;
 08445	       rp < END_PROC_ADDR; ++rp, ldt_index++) {
 08446	        init_dataseg(&gdt[ldt_index], vir2phys(rp->p_ldt),
 08447	                                     sizeof(rp->p_ldt), INTR_PRIVILEGE);
 08448	        gdt[ldt_index].access = PRESENT | LDT;
 08449	        rp->p_ldt_sel = ldt_index * DESC_SIZE;
 08450	  }
 08451	
 08452	  /* Build main TSS.
 08453	   * This is used only to record the stack pointer to be used after an
 08454	   * interrupt.
 08455	   * The pointer is set up so that an interrupt automatically saves the
 08456	   * current process's registers ip:cs:f:sp:ss in the correct slots in the
 08457	   * process table.
 08458	   */
 08459	  tss.ss0 = DS_SELECTOR;

_________________________ Page 739    File: kernel/protect.c _________________________
 08460	  init_dataseg(&gdt[TSS_INDEX], vir2phys(&tss), sizeof(tss), INTR_PRIVILEGE
 08461	  gdt[TSS_INDEX].access = PRESENT | (INTR_PRIVILEGE << DPL_SHIFT) | TSS_TYPE;
 08462	
 08463	  /* Build descriptors for interrupt gates in IDT. */
 08464	  for (gtp = &gate_table[0];
 08465	       gtp < &gate_table[sizeof gate_table / sizeof gate_table[0]]; ++gtp) {
 08466	        int_gate(gtp->vec_nr, (vir_bytes) gtp->gate,
 08467	                 PRESENT | INT_GATE_TYPE | (gtp->privilege << DPL_SHIFT));
 08468	  }
 08469	
 08470	  /* Complete building of main TSS. */
 08471	  tss.iobase = sizeof tss;      /* empty i/o permissions map */
 08472	}
	
 08474	/*===========================================================================*
 08475	 *                              init_codeseg                                 *
 08476	 *===========================================================================*/
 08477	PUBLIC void init_codeseg(segdp, base, size, privilege)
 08478	register struct segdesc_s *segdp;
 08479	phys_bytes base;
 08480	vir_bytes size;
 08481	int privilege;
 08482	{
 08483	/* Build descriptor for a code segment. */
 08484	  sdesc(segdp, base, size);
 08485	  segdp->access = (privilege << DPL_SHIFT)
 08486	                | (PRESENT | SEGMENT | EXECUTABLE | READABLE);
 08487	                /* CONFORMING = 0, ACCESSED = 0 */
 08488	}
	
 08490	/*===========================================================================*
 08491	 *                              init_dataseg                                 *
 08492	 *===========================================================================*/
 08493	PUBLIC void init_dataseg(segdp, base, size, privilege)
 08494	register struct segdesc_s *segdp;
 08495	phys_bytes base;
 08496	vir_bytes size;
 08497	int privilege;
 08498	{
 08499	/* Build descriptor for a data segment. */
 08500	  sdesc(segdp, base, size);
 08501	  segdp->access = (privilege << DPL_SHIFT) | (PRESENT | SEGMENT | WRITEABLE);
 08502	                /* EXECUTABLE = 0, EXPAND_DOWN = 0, ACCESSED = 0 */
 08503	}
	
 08505	/*===========================================================================*
 08506	 *                              sdesc                                        *
 08507	 *===========================================================================*/
 08508	PRIVATE void sdesc(segdp, base, size)
 08509	register struct segdesc_s *segdp;
 08510	phys_bytes base;
 08511	vir_bytes size;
 08512	{
 08513	/* Fill in the size fields (base, limit and granularity) of a descriptor. */
 08514	  segdp->base_low = base;
 08515	  segdp->base_middle = base >> BASE_MIDDLE_SHIFT;
 08516	  segdp->base_high = base >> BASE_HIGH_SHIFT;
 08517	
 08518	  --size;                       /* convert to a limit, 0 size means 4G */
 08519	  if (size > BYTE_GRAN_MAX) {

_________________________ Page 740    File: kernel/protect.c _________________________
 08520	        segdp->limit_low = size >> PAGE_GRAN_SHIFT;
 08521	        segdp->granularity = GRANULAR | (size >>
 08522	                                     (PAGE_GRAN_SHIFT + GRANULARITY_SHIFT));
 08523	  } else {
 08524	        segdp->limit_low = size;
 08525	        segdp->granularity = size >> GRANULARITY_SHIFT;
 08526	  }
 08527	  segdp->granularity |= DEFAULT;        /* means BIG for data seg */
 08528	}
	
 08530	/*===========================================================================*
 08531	 *                              seg2phys                                     *
 08532	 *===========================================================================*/
 08533	PUBLIC phys_bytes seg2phys(seg)
 08534	U16_t seg;
 08535	{
 08536	/* Return the base address of a segment, with seg being either a 8086 segment
 08537	 * register, or a 286/386 segment selector.
 08538	 */
 08539	  phys_bytes base;
 08540	  struct segdesc_s *segdp;
 08541	
 08542	  if (! machine.protected) {
 08543	        base = hclick_to_physb(seg);
 08544	  } else {
 08545	        segdp = &gdt[seg >> 3];
 08546	        base =    ((u32_t) segdp->base_low << 0)
 08547	                | ((u32_t) segdp->base_middle << 16)
 08548	                | ((u32_t) segdp->base_high << 24);
 08549	  }
 08550	  return base;
 08551	}
	
 08553	/*===========================================================================*
 08554	 *                              phys2seg                                     *
 08555	 *===========================================================================*/
 08556	PUBLIC void phys2seg(seg, off, phys)
 08557	u16_t *seg;
 08558	vir_bytes *off;
 08559	phys_bytes phys;
 08560	{
 08561	/* Return a segment selector and offset that can be used to reach a physical
 08562	 * address, for use by a driver doing memory I/O in the A0000 - DFFFF range.
 08563	 */
 08564	  *seg = FLAT_DS_SELECTOR;
 08565	  *off = phys;
 08566	}
	
 08568	/*===========================================================================*
 08569	 *                              int_gate                                     *
 08570	 *===========================================================================*/
 08571	PRIVATE void int_gate(vec_nr, offset, dpl_type)
 08572	unsigned vec_nr;
 08573	vir_bytes offset;
 08574	unsigned dpl_type;
 08575	{
 08576	/* Build descriptor for an interrupt gate. */
 08577	  register struct gatedesc_s *idp;
 08578	
 08579	  idp = &idt[vec_nr];

_________________________ Page 741    File: kernel/protect.c _________________________
 08580	  idp->offset_low = offset;
 08581	  idp->selector = CS_SELECTOR;
 08582	  idp->p_dpl_type = dpl_type;
 08583	  idp->offset_high = offset >> OFFSET_HIGH_SHIFT;
 08584	}
	
 08586	/*===========================================================================*
 08587	 *                              enable_iop                                   * 
 08588	 *===========================================================================*/
 08589	PUBLIC void enable_iop(pp)
 08590	struct proc *pp;
 08591	{
 08592	/* Allow a user process to use I/O instructions.  Change the I/O Permission
 08593	 * Level bits in the psw. These specify least-privileged Current Permission
 08594	 * Level allowed to execute I/O instructions. Users and servers have CPL 3. 
 08595	 * You can't have less privilege than that. Kernel has CPL 0, tasks CPL 1.
 08596	 */
 08597	  pp->p_reg.psw |= 0x3000;
 08598	}
	
 08600	/*===========================================================================*
 08601	 *                              alloc_segments                               *
 08602	 *===========================================================================*/
 08603	PUBLIC void alloc_segments(rp)
 08604	register struct proc *rp;
 08605	{
 08606	/* This is called at system initialization from main() and by do_newmap(). 
 08607	 * The code has a separate function because of all hardware-dependencies.
 08608	 * Note that IDLE is part of the kernel and gets TASK_PRIVILEGE here.
 08609	 */
 08610	  phys_bytes code_bytes;
 08611	  phys_bytes data_bytes;
 08612	  int privilege;
 08613	
 08614	  if (machine.protected) {
 08615	      data_bytes = (phys_bytes) (rp->p_memmap[S].mem_vir + 
 08616	          rp->p_memmap[S].mem_len) << CLICK_SHIFT;
 08617	      if (rp->p_memmap[T].mem_len == 0)
 08618	          code_bytes = data_bytes;      /* common I&D, poor protect */
 08619	      else
 08620	          code_bytes = (phys_bytes) rp->p_memmap[T].mem_len << CLICK_SHIFT;
 08621	      privilege = (iskernelp(rp)) ? TASK_PRIVILEGE : USER_PRIVILEGE;
 08622	      init_codeseg(&rp->p_ldt[CS_LDT_INDEX],
 08623	          (phys_bytes) rp->p_memmap[T].mem_phys << CLICK_SHIFT,
 08624	          code_bytes, privilege);
 08625	      init_dataseg(&rp->p_ldt[DS_LDT_INDEX],
 08626	          (phys_bytes) rp->p_memmap[D].mem_phys << CLICK_SHIFT,
 08627	          data_bytes, privilege);
 08628	      rp->p_reg.cs = (CS_LDT_INDEX * DESC_SIZE) | TI | privilege;
 08629	      rp->p_reg.gs =
 08630	      rp->p_reg.fs =
 08631	      rp->p_reg.ss =
 08632	      rp->p_reg.es =
 08633	      rp->p_reg.ds = (DS_LDT_INDEX*DESC_SIZE) | TI | privilege;
 08634	  } else {
 08635	      rp->p_reg.cs = click_to_hclick(rp->p_memmap[T].mem_phys);
 08636	      rp->p_reg.ss =
 08637	      rp->p_reg.es =
 08638	      rp->p_reg.ds = click_to_hclick(rp->p_memmap[D].mem_phys);
 08639	  }

_________________________ Page 742    File: kernel/protect.c _________________________
 08640	}
	




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/klib.s
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 08700	#
 08701	! Chooses between the 8086 and 386 versions of the low level kernel code.
 08702	
 08703	#include <minix/config.h>
 08704	#if _WORD_SIZE == 2
 08705	#include "klib88.s"
 08706	#else
 08707	#include "klib386.s"
 08708	#endif


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/klib386.s
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 08800	# 
 08801	! sections
 08802	
 08803	.sect .text; .sect .rom; .sect .data; .sect .bss
 08804	
 08805	#include <minix/config.h>
 08806	#include <minix/const.h>
 08807	#include "const.h"
 08808	#include "sconst.h"
 08809	#include "protect.h"
 08810	
 08811	! This file contains a number of assembly code utility routines needed by the
 08812	! kernel.  They are:
 08813	
 08814	.define _monitor        ! exit Minix and return to the monitor
 08815	.define _int86          ! let the monitor make an 8086 interrupt call
 08816	.define _cp_mess        ! copies messages from source to destination
 08817	.define _exit           ! dummy for library routines
 08818	.define __exit          ! dummy for library routines
 08819	.define ___exit         ! dummy for library routines
 08820	.define ___main         ! dummy for GCC
 08821	.define _phys_insw      ! transfer data from (disk controller) port to memory
 08822	.define _phys_insb      ! likewise byte by byte
 08823	.define _phys_outsw     ! transfer data from memory to (disk controller) port
 08824	.define _phys_outsb     ! likewise byte by byte
 08825	.define _enable_irq     ! enable an irq at the 8259 controller
 08826	.define _disable_irq    ! disable an irq
 08827	.define _phys_copy      ! copy data from anywhere to anywhere in memory
 08828	.define _phys_memset    ! write pattern anywhere in memory
 08829	.define _mem_rdw        ! copy one word from [segment:offset]
 08830	.define _reset          ! reset the system
 08831	.define _idle_task      ! task executed when there is no work
 08832	.define _level0         ! call a function at level 0
 08833	.define _read_tsc       ! read the cycle counter (Pentium and up)
 08834	.define _read_cpu_flags ! read the cpu flags

_________________________ Page 743    File: kernel/klib386.s _________________________
 08835	
 08836	! The routines only guarantee to preserve the registers the C compiler
 08837	! expects to be preserved (ebx, esi, edi, ebp, esp, segment registers, and
 08838	! direction bit in the flags).
 08839	
 08840	.sect .text
 08841	!*===========================================================================*
 08842	!*                              monitor                                      *
 08843	!*===========================================================================*
 08844	! PUBLIC void monitor();
 08845	! Return to the monitor.
 08846	
 08847	_monitor:
 08848	        mov     esp, (_mon_sp)          ! restore monitor stack pointer
 08849	    o16 mov     dx, SS_SELECTOR         ! monitor data segment
 08850	        mov     ds, dx
 08851	        mov     es, dx
 08852	        mov     fs, dx
 08853	        mov     gs, dx
 08854	        mov     ss, dx
 08855	        pop     edi
 08856	        pop     esi
 08857	        pop     ebp
 08858	    o16 retf                            ! return to the monitor
 08859	
 08860	
 08861	!*===========================================================================*
 08862	!*                              int86                                        *
 08863	!*===========================================================================*
 08864	! PUBLIC void int86();
 08865	_int86:
 08866	        cmpb    (_mon_return), 0        ! is the monitor there?
 08867	        jnz     0f
 08868	        movb    ah, 0x01                ! an int 13 error seems appropriate
 08869	        movb    (_reg86+ 0), ah         ! reg86.w.f = 1 (set carry flag)
 08870	        movb    (_reg86+13), ah         ! reg86.b.ah = 0x01 = "invalid command"
 08871	        ret
 08872	0:      push    ebp                     ! save C registers
 08873	        push    esi
 08874	        push    edi
 08875	        push    ebx
 08876	        pushf                           ! save flags
 08877	        cli                             ! no interruptions
 08878	
 08879	        inb     INT2_CTLMASK
 08880	        movb    ah, al
 08881	        inb     INT_CTLMASK
 08882	        push    eax                     ! save interrupt masks
 08883	        mov     eax, (_irq_use)         ! map of in-use IRQ's
 08884	        and     eax, ~[1<<CLOCK_IRQ]    ! keep the clock ticking
 08885	        outb    INT_CTLMASK             ! enable all unused IRQ's and vv.
 08886	        movb    al, ah
 08887	        outb    INT2_CTLMASK
 08888	
 08889	        mov     eax, SS_SELECTOR        ! monitor data segment
 08890	        mov     ss, ax
 08891	        xchg    esp, (_mon_sp)          ! switch stacks
 08892	        push    (_reg86+36)             ! parameters used in INT call
 08893	        push    (_reg86+32)
 08894	        push    (_reg86+28)

_________________________ Page 744    File: kernel/klib386.s _________________________
 08895	        push    (_reg86+24)
 08896	        push    (_reg86+20)
 08897	        push    (_reg86+16)
 08898	        push    (_reg86+12)
 08899	        push    (_reg86+ 8)
 08900	        push    (_reg86+ 4)
 08901	        push    (_reg86+ 0)
 08902	        mov     ds, ax                  ! remaining data selectors
 08903	        mov     es, ax
 08904	        mov     fs, ax
 08905	        mov     gs, ax
 08906	        push    cs
 08907	        push    return                  ! kernel return address and selector
 08908	    o16 jmpf    20+2*4+10*4+2*4(esp)    ! make the call
 08909	return:
 08910	        pop     (_reg86+ 0)
 08911	        pop     (_reg86+ 4)
 08912	        pop     (_reg86+ 8)
 08913	        pop     (_reg86+12)
 08914	        pop     (_reg86+16)
 08915	        pop     (_reg86+20)
 08916	        pop     (_reg86+24)
 08917	        pop     (_reg86+28)
 08918	        pop     (_reg86+32)
 08919	        pop     (_reg86+36)
 08920	        lgdt    (_gdt+GDT_SELECTOR)     ! reload global descriptor table
 08921	        jmpf    CS_SELECTOR:csinit      ! restore everything
 08922	csinit: mov     eax, DS_SELECTOR
 08923	        mov     ds, ax
 08924	        mov     es, ax
 08925	        mov     fs, ax
 08926	        mov     gs, ax
 08927	        mov     ss, ax
 08928	        xchg    esp, (_mon_sp)          ! unswitch stacks
 08929	        lidt    (_gdt+IDT_SELECTOR)     ! reload interrupt descriptor table
 08930	        andb    (_gdt+TSS_SELECTOR+DESC_ACCESS), ~0x02  ! clear TSS busy bit
 08931	        mov     eax, TSS_SELECTOR
 08932	        ltr     ax                      ! set TSS register
 08933	
 08934	        pop     eax
 08935	        outb    INT_CTLMASK             ! restore interrupt masks
 08936	        movb    al, ah
 08937	        outb    INT2_CTLMASK
 08938	
 08939	        add     (_lost_ticks), ecx      ! record lost clock ticks
 08940	
 08941	        popf                            ! restore flags
 08942	        pop     ebx                     ! restore C registers
 08943	        pop     edi
 08944	        pop     esi
 08945	        pop     ebp
 08946	        ret
 08947	
 08948	
 08949	!*===========================================================================*
 08950	!*                              cp_mess                                      *
 08951	!*===========================================================================*
 08952	! PUBLIC void cp_mess(int src, phys_clicks src_clicks, vir_bytes src_offset,
 08953	!                     phys_clicks dst_clicks, vir_bytes dst_offset);
 08954	! This routine makes a fast copy of a message from anywhere in the address

_________________________ Page 745    File: kernel/klib386.s _________________________
 08955	! space to anywhere else.  It also copies the source address provided as a
 08956	! parameter to the call into the first word of the destination message.
 08957	!
 08958	! Note that the message size, "Msize" is in DWORDS (not bytes) and must be set
 08959	! correctly.  Changing the definition of message in the type file and not
 08960	! changing it here will lead to total disaster.
 08961	
 08962	CM_ARGS =       4 + 4 + 4 + 4 + 4       ! 4 + 4 + 4 + 4 + 4
 08963	!               es  ds edi esi eip      proc scl sof dcl dof
 08964	
 08965	        .align  16
 08966	_cp_mess:
 08967	        cld
 08968	        push    esi
 08969	        push    edi
 08970	        push    ds
 08971	        push    es
 08972	
 08973	        mov     eax, FLAT_DS_SELECTOR
 08974	        mov     ds, ax
 08975	        mov     es, ax
 08976	
 08977	        mov     esi, CM_ARGS+4(esp)             ! src clicks
 08978	        shl     esi, CLICK_SHIFT
 08979	        add     esi, CM_ARGS+4+4(esp)           ! src offset
 08980	        mov     edi, CM_ARGS+4+4+4(esp)         ! dst clicks
 08981	        shl     edi, CLICK_SHIFT
 08982	        add     edi, CM_ARGS+4+4+4+4(esp)       ! dst offset
 08983	
 08984	        mov     eax, CM_ARGS(esp)       ! process number of sender
 08985	        stos                            ! copy number of sender to dest message
 08986	        add     esi, 4                  ! do not copy first word
 08987	        mov     ecx, Msize - 1          ! remember, first word does not count
 08988	        rep
 08989	        movs                            ! copy the message
 08990	
 08991	        pop     es
 08992	        pop     ds
 08993	        pop     edi
 08994	        pop     esi
 08995	        ret                             ! that is all folks!
 08996	
 08997	
 08998	!*===========================================================================*
 08999	!*                              exit                                         *
 09000	!*===========================================================================*
 09001	! PUBLIC void exit();
 09002	! Some library routines use exit, so provide a dummy version.
 09003	! Actual calls to exit cannot occur in the kernel.
 09004	! GNU CC likes to call ___main from main() for nonobvious reasons.
 09005	
 09006	_exit:
 09007	__exit:
 09008	___exit:
 09009	        sti
 09010	        jmp     ___exit
 09011	
 09012	___main:
 09013	        ret
 09014	

_________________________ Page 746    File: kernel/klib386.s _________________________
 09015	
 09016	!*===========================================================================*
 09017	!*                              phys_insw                                    *
 09018	!*===========================================================================*
 09019	! PUBLIC void phys_insw(Port_t port, phys_bytes buf, size_t count);
 09020	! Input an array from an I/O port.  Absolute address version of insw().
 09021	
 09022	_phys_insw:
 09023	        push    ebp
 09024	        mov     ebp, esp
 09025	        cld
 09026	        push    edi
 09027	        push    es
 09028	        mov     ecx, FLAT_DS_SELECTOR
 09029	        mov     es, cx
 09030	        mov     edx, 8(ebp)             ! port to read from
 09031	        mov     edi, 12(ebp)            ! destination addr
 09032	        mov     ecx, 16(ebp)            ! byte count
 09033	        shr     ecx, 1                  ! word count
 09034	rep o16 ins                             ! input many words
 09035	        pop     es
 09036	        pop     edi
 09037	        pop     ebp
 09038	        ret
 09039	
 09040	
 09041	!*===========================================================================*
 09042	!*                              phys_insb                                    *
 09043	!*===========================================================================*
 09044	! PUBLIC void phys_insb(Port_t port, phys_bytes buf, size_t count);
 09045	! Input an array from an I/O port.  Absolute address version of insb().
 09046	
 09047	_phys_insb:
 09048	        push    ebp
 09049	        mov     ebp, esp
 09050	        cld
 09051	        push    edi
 09052	        push    es
 09053	        mov     ecx, FLAT_DS_SELECTOR
 09054	        mov     es, cx
 09055	        mov     edx, 8(ebp)             ! port to read from
 09056	        mov     edi, 12(ebp)            ! destination addr
 09057	        mov     ecx, 16(ebp)            ! byte count
 09058	!       shr     ecx, 1                  ! word count
 09059	   rep  insb                            ! input many bytes
 09060	        pop     es
 09061	        pop     edi
 09062	        pop     ebp
 09063	        ret
 09064	
 09065	
 09066	!*===========================================================================*
 09067	!*                              phys_outsw                                   *
 09068	!*===========================================================================*
 09069	! PUBLIC void phys_outsw(Port_t port, phys_bytes buf, size_t count);
 09070	! Output an array to an I/O port.  Absolute address version of outsw().
 09071	
 09072	        .align  16
 09073	_phys_outsw:
 09074	        push    ebp

_________________________ Page 747    File: kernel/klib386.s _________________________
 09075	        mov     ebp, esp
 09076	        cld
 09077	        push    esi
 09078	        push    ds
 09079	        mov     ecx, FLAT_DS_SELECTOR
 09080	        mov     ds, cx
 09081	        mov     edx, 8(ebp)             ! port to write to
 09082	        mov     esi, 12(ebp)            ! source addr
 09083	        mov     ecx, 16(ebp)            ! byte count
 09084	        shr     ecx, 1                  ! word count
 09085	rep o16 outs                            ! output many words
 09086	        pop     ds
 09087	        pop     esi
 09088	        pop     ebp
 09089	        ret
 09090	
 09091	
 09092	!*===========================================================================*
 09093	!*                              phys_outsb                                   *
 09094	!*===========================================================================*
 09095	! PUBLIC void phys_outsb(Port_t port, phys_bytes buf, size_t count);
 09096	! Output an array to an I/O port.  Absolute address version of outsb().
 09097	
 09098	        .align  16
 09099	_phys_outsb:
 09100	        push    ebp
 09101	        mov     ebp, esp
 09102	        cld
 09103	        push    esi
 09104	        push    ds
 09105	        mov     ecx, FLAT_DS_SELECTOR
 09106	        mov     ds, cx
 09107	        mov     edx, 8(ebp)             ! port to write to
 09108	        mov     esi, 12(ebp)            ! source addr
 09109	        mov     ecx, 16(ebp)            ! byte count
 09110	   rep  outsb                           ! output many bytes
 09111	        pop     ds
 09112	        pop     esi
 09113	        pop     ebp
 09114	        ret
 09115	
 09116	
 09117	!*==========================================================================*
 09118	!*                              enable_irq                                  *
 09119	!*==========================================================================*/
 09120	! PUBLIC void enable_irq(irq_hook_t *hook)
 09121	! Enable an interrupt request line by clearing an 8259 bit.
 09122	! Equivalent C code for hook->irq < 8:
 09123	!   if ((irq_actids[hook->irq] &= ~hook->id) == 0)
 09124	!       outb(INT_CTLMASK, inb(INT_CTLMASK) & ~(1 << irq));
 09125	
 09126	        .align  16
 09127	_enable_irq:
 09128	        push    ebp
 09129	        mov     ebp, esp
 09130	        pushf
 09131	        cli
 09132	        mov     eax, 8(ebp)             ! hook
 09133	        mov     ecx, 8(eax)             ! irq
 09134	        mov     eax, 12(eax)            ! id bit

_________________________ Page 748    File: kernel/klib386.s _________________________
 09135	        not     eax
 09136	        and     _irq_actids(ecx*4), eax ! clear this id bit
 09137	        jnz     en_done                 ! still masked by other handlers?
 09138	        movb    ah, ~1
 09139	        rolb    ah, cl                  ! ah = ~(1 << (irq % 8))
 09140	        mov     edx, INT_CTLMASK        ! enable irq < 8 at the master 8259
 09141	        cmpb    cl, 8
 09142	        jb      0f
 09143	        mov     edx, INT2_CTLMASK       ! enable irq >= 8 at the slave 8259
 09144	0:      inb     dx
 09145	        andb    al, ah
 09146	        outb    dx                      ! clear bit at the 8259
 09147	en_done:popf
 09148	        leave
 09149	        ret
 09150	
 09151	
 09152	!*==========================================================================*
 09153	!*                              disable_irq                                 *
 09154	!*==========================================================================*/
 09155	! PUBLIC int disable_irq(irq_hook_t *hook)
 09156	! Disable an interrupt request line by setting an 8259 bit.
 09157	! Equivalent C code for irq < 8:
 09158	!   irq_actids[hook->irq] |= hook->id;
 09159	!   outb(INT_CTLMASK, inb(INT_CTLMASK) | (1 << irq));
 09160	! Returns true iff the interrupt was not already disabled.
 09161	
 09162	        .align  16
 09163	_disable_irq:
 09164	        push    ebp
 09165	        mov     ebp, esp
 09166	        pushf
 09167	        cli
 09168	        mov     eax, 8(ebp)             ! hook
 09169	        mov     ecx, 8(eax)             ! irq
 09170	        mov     eax, 12(eax)            ! id bit
 09171	        or      _irq_actids(ecx*4), eax ! set this id bit
 09172	        movb    ah, 1
 09173	        rolb    ah, cl                  ! ah = (1 << (irq % 8))
 09174	        mov     edx, INT_CTLMASK        ! disable irq < 8 at the master 8259
 09175	        cmpb    cl, 8
 09176	        jb      0f
 09177	        mov     edx, INT2_CTLMASK       ! disable irq >= 8 at the slave 8259
 09178	0:      inb     dx
 09179	        testb   al, ah
 09180	        jnz     dis_already             ! already disabled?
 09181	        orb     al, ah
 09182	        outb    dx                      ! set bit at the 8259
 09183	        mov     eax, 1                  ! disabled by this function
 09184	        popf
 09185	        leave
 09186	        ret
 09187	dis_already:
 09188	        xor     eax, eax                ! already disabled
 09189	        popf
 09190	        leave
 09191	        ret
 09192	
 09193	

_________________________ Page 749    File: kernel/klib386.s _________________________
 09194	!*===========================================================================*
 09195	!*                              phys_copy                                    *
 09196	!*===========================================================================*
 09197	! PUBLIC void phys_copy(phys_bytes source, phys_bytes destination,
 09198	!                       phys_bytes bytecount);
 09199	! Copy a block of physical memory.
 09200	
 09201	PC_ARGS =       4 + 4 + 4 + 4   ! 4 + 4 + 4
 09202	!               es edi esi eip   src dst len
 09203	
 09204	        .align  16
 09205	_phys_copy:
 09206	        cld
 09207	        push    esi
 09208	        push    edi
 09209	        push    es
 09210	
 09211	        mov     eax, FLAT_DS_SELECTOR
 09212	        mov     es, ax
 09213	
 09214	        mov     esi, PC_ARGS(esp)
 09215	        mov     edi, PC_ARGS+4(esp)
 09216	        mov     eax, PC_ARGS+4+4(esp)
 09217	
 09218	        cmp     eax, 10                 ! avoid align overhead for small counts
 09219	        jb      pc_small
 09220	        mov     ecx, esi                ! align source, hope target is too
 09221	        neg     ecx
 09222	        and     ecx, 3                  ! count for alignment
 09223	        sub     eax, ecx
 09224	        rep
 09225	   eseg movsb
 09226	        mov     ecx, eax
 09227	        shr     ecx, 2                  ! count of dwords
 09228	        rep
 09229	   eseg movs
 09230	        and     eax, 3
 09231	pc_small:
 09232	        xchg    ecx, eax                ! remainder
 09233	        rep
 09234	   eseg movsb
 09235	
 09236	        pop     es
 09237	        pop     edi
 09238	        pop     esi
 09239	        ret
 09240	
 09241	!*===========================================================================*
 09242	!*                              phys_memset                                  *
 09243	!*===========================================================================*
 09244	! PUBLIC void phys_memset(phys_bytes source, unsigned long pattern,
 09245	!       phys_bytes bytecount);
 09246	! Fill a block of physical memory with pattern.
 09247	
 09248	        .align  16
 09249	_phys_memset:
 09250	        push    ebp
 09251	        mov     ebp, esp
 09252	        push    esi
 09253	        push    ebx

_________________________ Page 750    File: kernel/klib386.s _________________________
 09254	        push    ds
 09255	        mov     esi, 8(ebp)
 09256	        mov     eax, 16(ebp)
 09257	        mov     ebx, FLAT_DS_SELECTOR
 09258	        mov     ds, bx
 09259	        mov     ebx, 12(ebp)
 09260	        shr     eax, 2
 09261	fill_start:
 09262	        mov     (esi), ebx
 09263	        add     esi, 4
 09264	        dec     eax
 09265	        jnz     fill_start
 09266	        ! Any remaining bytes?
 09267	        mov     eax, 16(ebp)
 09268	        and     eax, 3
 09269	remain_fill:
 09270	        cmp     eax, 0
 09271	        jz      fill_done
 09272	        movb    bl, 12(ebp)
 09273	        movb    (esi), bl
 09274	        add     esi, 1
 09275	        inc     ebp
 09276	        dec     eax
 09277	        jmp     remain_fill
 09278	fill_done:
 09279	        pop     ds
 09280	        pop     ebx
 09281	        pop     esi
 09282	        pop     ebp
 09283	        ret
 09284	
 09285	!*===========================================================================*
 09286	!*                              mem_rdw                                      *
 09287	!*===========================================================================*
 09288	! PUBLIC u16_t mem_rdw(U16_t segment, u16_t *offset);
 09289	! Load and return word at far pointer segment:offset.
 09290	
 09291	        .align  16
 09292	_mem_rdw:
 09293	        mov     cx, ds
 09294	        mov     ds, 4(esp)              ! segment
 09295	        mov     eax, 4+4(esp)           ! offset
 09296	        movzx   eax, (eax)              ! word to return
 09297	        mov     ds, cx
 09298	        ret
 09299	
 09300	
 09301	!*===========================================================================*
 09302	!*                              reset                                        *
 09303	!*===========================================================================*
 09304	! PUBLIC void reset();
 09305	! Reset the system by loading IDT with offset 0 and interrupting.
 09306	
 09307	_reset:
 09308	        lidt    (idt_zero)
 09309	        int     3               ! anything goes, the 386 will not like it
 09310	.sect .data
 09311	idt_zero:       .data4  0, 0
 09312	.sect .text
 09313	

_________________________ Page 751    File: kernel/klib386.s _________________________
 09314	
 09315	!*===========================================================================*
 09316	!*                              idle_task                                    *
 09317	!*===========================================================================*
 09318	_idle_task:
 09319	! This task is called when the system has nothing else to do.  The HLT
 09320	! instruction puts the processor in a state where it draws minimum power.
 09321	        push    halt
 09322	        call    _level0         ! level0(halt)
 09323	        pop     eax
 09324	        jmp     _idle_task
 09325	halt:
 09326	        sti
 09327	        hlt
 09328	        cli
 09329	        ret
 09330	
 09331	!*===========================================================================*
 09332	!*                            level0                                         *
 09333	!*===========================================================================*
 09334	! PUBLIC void level0(void (*func)(void))
 09335	! Call a function at permission level 0.  This allows kernel tasks to do
 09336	! things that are only possible at the most privileged CPU level.
 09337	!
 09338	_level0:
 09339	        mov     eax, 4(esp)
 09340	        mov     (_level0_func), eax
 09341	        int     LEVEL0_VECTOR
 09342	        ret
 09343	
 09344	
 09345	!*===========================================================================*
 09346	!*                            read_tsc                                       *
 09347	!*===========================================================================*
 09348	! PUBLIC void read_tsc(unsigned long *high, unsigned long *low);
 09349	! Read the cycle counter of the CPU. Pentium and up. 
 09350	.align 16
 09351	_read_tsc:
 09352	.data1 0x0f             ! this is the RDTSC instruction 
 09353	.data1 0x31             ! it places the TSC in EDX:EAX
 09354	        push ebp
 09355	        mov ebp, 8(esp)
 09356	        mov (ebp), edx
 09357	        mov ebp, 12(esp)
 09358	        mov (ebp), eax
 09359	        pop ebp
 09360	        ret
 09361	
 09362	!*===========================================================================*
 09363	!*                            read_flags                                           
 09364	!*===========================================================================*
 09365	! PUBLIC unsigned long read_cpu_flags(void);
 09366	! Read CPU status flags from C.
 09367	.align 16
 09368	_read_cpu_flags:
 09369	        pushf
 09370	        mov eax, (esp)
 09371	        popf
 09372	        ret
 09373	

_________________________ Page 752    File: kernel/klib386.s _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/utility.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 09400	/* This file contains a collection of miscellaneous procedures:
 09401	 *   panic:         abort MINIX due to a fatal error
 09402	 *   kprintf:       diagnostic output for the kernel 
 09403	 *
 09404	 * Changes:
 09405	 *   Dec 10, 2004   kernel printing to circular buffer  (Jorrit N. Herder)
 09406	 * 
 09407	 * This file contains the routines that take care of kernel messages, i.e.,
 09408	 * diagnostic output within the kernel. Kernel messages are not directly
 09409	 * displayed on the console, because this must be done by the output driver. 
 09410	 * Instead, the kernel accumulates characters in a buffer and notifies the
 09411	 * output driver when a new message is ready. 
 09412	 */
 09413	
 09414	#include <minix/com.h>
 09415	#include "kernel.h"
 09416	#include <stdarg.h>
 09417	#include <unistd.h>
 09418	#include <stddef.h>
 09419	#include <stdlib.h>
 09420	#include <signal.h>
 09421	#include "proc.h"
 09422	
 09423	#define END_OF_KMESS    -1
 09424	FORWARD _PROTOTYPE(void kputc, (int c));
 09425	
 09426	/*===========================================================================*
 09427	 *                              panic                                        *
 09428	 *===========================================================================*/
 09429	PUBLIC void panic(mess,nr)
 09430	_CONST char *mess;
 09431	int nr;
 09432	{
 09433	/* The system has run aground of a fatal kernel error. Terminate execution. */
 09434	  static int panicking = 0;
 09435	  if (panicking ++) return;             /* prevent recursive panics */
 09436	
 09437	  if (mess != NULL) {
 09438	        kprintf("\nKernel panic: %s", mess);
 09439	        if (nr != NO_NUM) kprintf(" %d", nr);
 09440	        kprintf("\n",NO_NUM);
 09441	  }
 09442	
 09443	  /* Abort MINIX. */
 09444	  prepare_shutdown(RBT_PANIC);
 09445	}
	
 09447	/*===========================================================================*
 09448	 *                              kprintf                                      *
 09449	 *===========================================================================*/
 09450	PUBLIC void kprintf(const char *fmt, ...)       /* format to be printed */
 09451	{
 09452	  int c;                                        /* next character in fmt */
 09453	  int d;
 09454	  unsigned long u;                              /* hold number argument */

_________________________ Page 753    File: kernel/utility.c _________________________
 09455	  int base;                                     /* base of number arg */
 09456	  int negative = 0;                             /* print minus sign */
 09457	  static char x2c[] = "0123456789ABCDEF";       /* nr conversion table */
 09458	  char ascii[8 * sizeof(long) / 3 + 2];         /* string for ascii number */
 09459	  char *s = NULL;                               /* string to be printed */
 09460	  va_list argp;                                 /* optional arguments */
 09461	  
 09462	  va_start(argp, fmt);                          /* init variable arguments */
 09463	
 09464	  while((c=*fmt++) != 0) {
 09465	
 09466	      if (c == '%') {                           /* expect format '%key' */
 09467	          switch(c = *fmt++) {                  /* determine what to do */
 09468	
 09469	          /* Known keys are %d, %u, %x, %s, and %%. This is easily extended 
 09470	           * with number types like %b and %o by providing a different base.
 09471	           * Number type keys don't set a string to 's', but use the general
 09472	           * conversion after the switch statement.
 09473	           */ 
 09474	          case 'd':                             /* output decimal */
 09475	              d = va_arg(argp, signed int);
 09476	              if (d < 0) { negative = 1; u = -d; }  else { u = d; }
 09477	              base = 10;
 09478	              break;
 09479	          case 'u':                             /* output unsigned long */
 09480	              u = va_arg(argp, unsigned long);
 09481	              base = 10;
 09482	              break;
 09483	          case 'x':                             /* output hexadecimal */
 09484	              u = va_arg(argp, unsigned long);
 09485	              base = 0x10;
 09486	              break;
 09487	          case 's':                             /* output string */
 09488	              s = va_arg(argp, char *);
 09489	              if (s == NULL) s = "(null)";
 09490	              break;
 09491	          case '%':                             /* output percent */
 09492	              s = "%";                           
 09493	              break;                    
 09494	
 09495	          /* Unrecognized key. */
 09496	          default:                              /* echo back %key */
 09497	              s = "%?";                         
 09498	              s[1] = c;                         /* set unknown key */
 09499	          }
 09500	
 09501	          /* Assume a number if no string is set. Convert to ascii. */
 09502	          if (s == NULL) {
 09503	              s = ascii + sizeof(ascii)-1;
 09504	              *s = 0;                   
 09505	              do {  *--s = x2c[(u % base)]; }   /* work backwards */
 09506	              while ((u /= base) > 0); 
 09507	          }
 09508	
 09509	          /* This is where the actual output for format "%key" is done. */
 09510	          if (negative) kputc('-');             /* print sign if negative */
 09511	          while(*s != 0) { kputc(*s++); }       /* print string/ number */
 09512	          s = NULL;                             /* reset for next round */
 09513	      }
 09514	      else {

_________________________ Page 754    File: kernel/utility.c _________________________
 09515	          kputc(c);                             /* print and continue */
 09516	      }
 09517	  }
 09518	  kputc(END_OF_KMESS);                          /* terminate output */
 09519	  va_end(argp);                                 /* end variable arguments */
 09520	}
	
 09522	/*===========================================================================*
 09523	 *                              kputc                                        *
 09524	 *===========================================================================*/
 09525	PRIVATE void kputc(c)
 09526	int c;                                  /* character to append */
 09527	{
 09528	/* Accumulate a single character for a kernel message. Send a notification
 09529	 * to the output driver if an END_OF_KMESS is encountered. 
 09530	 */
 09531	  if (c != END_OF_KMESS) {
 09532	      kmess.km_buf[kmess.km_next] = c;  /* put normal char in buffer */
 09533	      if (kmess.km_size < KMESS_BUF_SIZE)
 09534	          kmess.km_size += 1;           
 09535	      kmess.km_next = (kmess.km_next + 1) % KMESS_BUF_SIZE;
 09536	  } else {
 09537	      send_sig(OUTPUT_PROC_NR, SIGKMESS);
 09538	  }
 09539	}
	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/system.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 09600	/* Function prototypes for the system library.  
 09601	 * The implementation is contained in src/kernel/system/.  
 09602	 *
 09603	 * The system library allows access to system services by doing a kernel call.
 09604	 * Kernel calls are transformed into request messages to the SYS task that is 
 09605	 * responsible for handling the call. By convention, sys_call() is transformed 
 09606	 * into a message with type SYS_CALL that is handled in a function do_call(). 
 09607	 */ 
 09608	
 09609	#ifndef SYSTEM_H
 09610	#define SYSTEM_H
 09611	
 09612	/* Common includes for the system library. */
 09613	#include "kernel.h"
 09614	#include "proto.h"
 09615	#include "proc.h"
 09616	
 09617	/* Default handler for unused kernel calls. */
 09618	_PROTOTYPE( int do_unused, (message *m_ptr) );
 09619	_PROTOTYPE( int do_exec, (message *m_ptr) );            
 09620	_PROTOTYPE( int do_fork, (message *m_ptr) );
 09621	_PROTOTYPE( int do_newmap, (message *m_ptr) );
 09622	_PROTOTYPE( int do_exit, (message *m_ptr) );
 09623	_PROTOTYPE( int do_trace, (message *m_ptr) );   
 09624	_PROTOTYPE( int do_nice, (message *m_ptr) );

_________________________ Page 755    File: kernel/system.h _________________________
 09625	_PROTOTYPE( int do_copy, (message *m_ptr) );    
 09626	#define do_vircopy      do_copy
 09627	#define do_physcopy     do_copy
 09628	_PROTOTYPE( int do_vcopy, (message *m_ptr) );           
 09629	#define do_virvcopy     do_vcopy
 09630	#define do_physvcopy    do_vcopy
 09631	_PROTOTYPE( int do_umap, (message *m_ptr) );
 09632	_PROTOTYPE( int do_memset, (message *m_ptr) );
 09633	_PROTOTYPE( int do_abort, (message *m_ptr) );
 09634	_PROTOTYPE( int do_getinfo, (message *m_ptr) );
 09635	_PROTOTYPE( int do_privctl, (message *m_ptr) ); 
 09636	_PROTOTYPE( int do_segctl, (message *m_ptr) );
 09637	_PROTOTYPE( int do_irqctl, (message *m_ptr) );
 09638	_PROTOTYPE( int do_devio, (message *m_ptr) );
 09639	_PROTOTYPE( int do_vdevio, (message *m_ptr) );
 09640	_PROTOTYPE( int do_int86, (message *m_ptr) );
 09641	_PROTOTYPE( int do_sdevio, (message *m_ptr) );
 09642	_PROTOTYPE( int do_kill, (message *m_ptr) );
 09643	_PROTOTYPE( int do_getksig, (message *m_ptr) );
 09644	_PROTOTYPE( int do_endksig, (message *m_ptr) );
 09645	_PROTOTYPE( int do_sigsend, (message *m_ptr) );
 09646	_PROTOTYPE( int do_sigreturn, (message *m_ptr) );
 09647	_PROTOTYPE( int do_times, (message *m_ptr) );           
 09648	_PROTOTYPE( int do_setalarm, (message *m_ptr) );        
 09649	
 09650	#endif  /* SYSTEM_H */
 09651	
 09652	
 09653	


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/system.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 09700	/* This task provides an interface between the kernel and user-space system
 09701	 * processes. System services can be accessed by doing a kernel call. Kernel
 09702	 * calls are  transformed into request messages, which are handled by this
 09703	 * task. By convention, a sys_call() is transformed in a SYS_CALL request
 09704	 * message that is handled in a function named do_call(). 
 09705	 *
 09706	 * A private call vector is used to map all kernel calls to the functions that
 09707	 * handle them. The actual handler functions are contained in separate files
 09708	 * to keep this file clean. The call vector is used in the system task's main
 09709	 * loop to handle all incoming requests.  
 09710	 *
 09711	 * In addition to the main sys_task() entry point, which starts the main loop,
 09712	 * there are several other minor entry points:
 09713	 *   get_priv:          assign privilege structure to user or system process
 09714	 *   send_sig:          send a signal directly to a system process
 09715	 *   cause_sig:         take action to cause a signal to occur via PM
 09716	 *   umap_local:        map virtual address in LOCAL_SEG to physical 
 09717	 *   umap_remote:       map virtual address in REMOTE_SEG to physical 
 09718	 *   umap_bios:         map virtual address in BIOS_SEG to physical 
 09719	 *   virtual_copy:      copy bytes from one virtual address to another 
 09720	 *   get_randomness:    accumulate randomness in a buffer
 09721	 *
 09722	 * Changes:
 09723	 *   Aug 04, 2005   check if kernel call is allowed  (Jorrit N. Herder)
 09724	 *   Jul 20, 2005   send signal to services with message  (Jorrit N. Herder) 

_________________________ Page 756    File: kernel/system.c _________________________
 09725	 *   Jan 15, 2005   new, generalized virtual copy function  (Jorrit N. Herder)
 09726	 *   Oct 10, 2004   dispatch system calls from call vector  (Jorrit N. Herder)
 09727	 *   Sep 30, 2004   source code documentation updated  (Jorrit N. Herder)
 09728	 */
 09729	
 09730	#include "kernel.h"
 09731	#include "system.h"
 09732	#include <stdlib.h>
 09733	#include <signal.h>
 09734	#include <unistd.h>
 09735	#include <sys/sigcontext.h>
 09736	#include <ibm/memory.h>
 09737	#include "protect.h"
 09738	
 09739	/* Declaration of the call vector that defines the mapping of kernel calls 
 09740	 * to handler functions. The vector is initialized in sys_init() with map(), 
 09741	 * which makes sure the kernel call numbers are ok. No space is allocated, 
 09742	 * because the dummy is declared extern. If an illegal call is given, the 
 09743	 * array size will be negative and this won't compile. 
 09744	 */
 09745	PUBLIC int (*call_vec[NR_SYS_CALLS])(message *m_ptr);
 09746	
 09747	#define map(call_nr, handler) \
 09748	    {extern int dummy[NR_SYS_CALLS>(unsigned)(call_nr-KERNEL_CALL) ? 1:-1];} \
 09749	    call_vec[(call_nr-KERNEL_CALL)] = (handler)  
 09750	
 09751	FORWARD _PROTOTYPE( void initialize, (void));
 09752	
 09753	/*===========================================================================*
 09754	 *                              sys_task                                     *
 09755	 *===========================================================================*/
 09756	PUBLIC void sys_task()
 09757	{
 09758	/* Main entry point of sys_task.  Get the message and dispatch on type. */
 09759	  static message m;
 09760	  register int result;
 09761	  register struct proc *caller_ptr;
 09762	  unsigned int call_nr;
 09763	  int s;
 09764	
 09765	  /* Initialize the system task. */
 09766	  initialize();
 09767	
 09768	  while (TRUE) {
 09769	      /* Get work. Block and wait until a request message arrives. */
 09770	      receive(ANY, &m);                 
 09771	      call_nr = (unsigned) m.m_type - KERNEL_CALL;      
 09772	      caller_ptr = proc_addr(m.m_source);       
 09773	
 09774	      /* See if the caller made a valid request and try to handle it. */
 09775	      if (! (priv(caller_ptr)->s_call_mask & (1<<call_nr))) {
 09776	          kprintf("SYSTEM: request %d from %d denied.\n", call_nr,m.m_source);
 09777	          result = ECALLDENIED;                 /* illegal message type */
 09778	      } else if (call_nr >= NR_SYS_CALLS) {             /* check call number */
 09779	          kprintf("SYSTEM: illegal request %d from %d.\n", call_nr,m.m_source);
 09780	          result = EBADREQUEST;                 /* illegal message type */
 09781	      } 
 09782	      else {
 09783	          result = (*call_vec[call_nr])(&m);    /* handle the kernel call */
 09784	      }

_________________________ Page 757    File: kernel/system.c _________________________
 09785	
 09786	      /* Send a reply, unless inhibited by a handler function. Use the kernel
 09787	       * function lock_send() to prevent a system call trap. The destination
 09788	       * is known to be blocked waiting for a message.
 09789	       */
 09790	      if (result != EDONTREPLY) {
 09791	          m.m_type = result;                    /* report status of call */
 09792	          if (OK != (s=lock_send(m.m_source, &m))) {
 09793	              kprintf("SYSTEM, reply to %d failed: %d\n", m.m_source, s);
 09794	          }
 09795	      }
 09796	  }
 09797	}
	
 09799	/*===========================================================================*
 09800	 *                              initialize                                   *
 09801	 *===========================================================================*/
 09802	PRIVATE void initialize(void)
 09803	{
 09804	  register struct priv *sp;
 09805	  int i;
 09806	
 09807	  /* Initialize IRQ handler hooks. Mark all hooks available. */
 09808	  for (i=0; i<NR_IRQ_HOOKS; i++) {
 09809	      irq_hooks[i].proc_nr = NONE;
 09810	  }
 09811	
 09812	  /* Initialize all alarm timers for all processes. */
 09813	  for (sp=BEG_PRIV_ADDR; sp < END_PRIV_ADDR; sp++) {
 09814	    tmr_inittimer(&(sp->s_alarm_timer));
 09815	  }
 09816	
 09817	  /* Initialize the call vector to a safe default handler. Some kernel calls 
 09818	   * may be disabled or nonexistant. Then explicitly map known calls to their
 09819	   * handler functions. This is done with a macro that gives a compile error
 09820	   * if an illegal call number is used. The ordering is not important here.
 09821	   */
 09822	  for (i=0; i<NR_SYS_CALLS; i++) {
 09823	      call_vec[i] = do_unused;
 09824	  }
 09825	
 09826	  /* Process management. */
 09827	  map(SYS_FORK, do_fork);               /* a process forked a new process */
 09828	  map(SYS_EXEC, do_exec);               /* update process after execute */
 09829	  map(SYS_EXIT, do_exit);               /* clean up after process exit */
 09830	  map(SYS_NICE, do_nice);               /* set scheduling priority */
 09831	  map(SYS_PRIVCTL, do_privctl);         /* system privileges control */
 09832	  map(SYS_TRACE, do_trace);             /* request a trace operation */
 09833	
 09834	  /* Signal handling. */
 09835	  map(SYS_KILL, do_kill);               /* cause a process to be signaled */
 09836	  map(SYS_GETKSIG, do_getksig);         /* PM checks for pending signals */
 09837	  map(SYS_ENDKSIG, do_endksig);         /* PM finished processing signal */
 09838	  map(SYS_SIGSEND, do_sigsend);         /* start POSIX-style signal */
 09839	  map(SYS_SIGRETURN, do_sigreturn);     /* return from POSIX-style signal */
 09840	
 09841	  /* Device I/O. */
 09842	  map(SYS_IRQCTL, do_irqctl);           /* interrupt control operations */ 
 09843	  map(SYS_DEVIO, do_devio);             /* inb, inw, inl, outb, outw, outl */ 
 09844	  map(SYS_SDEVIO, do_sdevio);           /* phys_insb, _insw, _outsb, _outsw */

_________________________ Page 758    File: kernel/system.c _________________________
 09845	  map(SYS_VDEVIO, do_vdevio);           /* vector with devio requests */ 
 09846	  map(SYS_INT86, do_int86);             /* real-mode BIOS calls */ 
 09847	
 09848	  /* Memory management. */
 09849	  map(SYS_NEWMAP, do_newmap);           /* set up a process memory map */
 09850	  map(SYS_SEGCTL, do_segctl);           /* add segment and get selector */
 09851	  map(SYS_MEMSET, do_memset);           /* write char to memory area */
 09852	
 09853	  /* Copying. */
 09854	  map(SYS_UMAP, do_umap);               /* map virtual to physical address */
 09855	  map(SYS_VIRCOPY, do_vircopy);         /* use pure virtual addressing */
 09856	  map(SYS_PHYSCOPY, do_physcopy);       /* use physical addressing */
 09857	  map(SYS_VIRVCOPY, do_virvcopy);       /* vector with copy requests */
 09858	  map(SYS_PHYSVCOPY, do_physvcopy);     /* vector with copy requests */
 09859	
 09860	  /* Clock functionality. */
 09861	  map(SYS_TIMES, do_times);             /* get uptime and process times */
 09862	  map(SYS_SETALARM, do_setalarm);       /* schedule a synchronous alarm */
 09863	
 09864	  /* System control. */
 09865	  map(SYS_ABORT, do_abort);             /* abort MINIX */
 09866	  map(SYS_GETINFO, do_getinfo);         /* request system information */ 
 09867	}
	
 09869	/*===========================================================================*
 09870	 *                              get_priv                                     *
 09871	 *===========================================================================*/
 09872	PUBLIC int get_priv(rc, proc_type)
 09873	register struct proc *rc;               /* new (child) process pointer */
 09874	int proc_type;                          /* system or user process flag */
 09875	{
 09876	/* Get a privilege structure. All user processes share the same privilege 
 09877	 * structure. System processes get their own privilege structure. 
 09878	 */
 09879	  register struct priv *sp;                     /* privilege structure */
 09880	
 09881	  if (proc_type == SYS_PROC) {                  /* find a new slot */
 09882	      for (sp = BEG_PRIV_ADDR; sp < END_PRIV_ADDR; ++sp) 
 09883	          if (sp->s_proc_nr == NONE && sp->s_id != USER_PRIV_ID) break; 
 09884	      if (sp->s_proc_nr != NONE) return(ENOSPC);
 09885	      rc->p_priv = sp;                          /* assign new slot */
 09886	      rc->p_priv->s_proc_nr = proc_nr(rc);      /* set association */
 09887	      rc->p_priv->s_flags = SYS_PROC;           /* mark as privileged */
 09888	  } else {
 09889	      rc->p_priv = &priv[USER_PRIV_ID];         /* use shared slot */
 09890	      rc->p_priv->s_proc_nr = INIT_PROC_NR;     /* set association */
 09891	      rc->p_priv->s_flags = 0;                  /* no initial flags */
 09892	  }
 09893	  return(OK);
 09894	}
	
 09896	/*===========================================================================*
 09897	 *                              get_randomness                               *
 09898	 *===========================================================================*/
 09899	PUBLIC void get_randomness(source)
 09900	int source;
 09901	{
 09902	/* On machines with the RDTSC (cycle counter read instruction - pentium
 09903	 * and up), use that for high-resolution raw entropy gathering. Otherwise,
 09904	 * use the realtime clock (tick resolution).

_________________________ Page 759    File: kernel/system.c _________________________
 09905	 *
 09906	 * Unfortunately this test is run-time - we don't want to bother with
 09907	 * compiling different kernels for different machines.
 09908	 *
 09909	 * On machines without RDTSC, we use read_clock().
 09910	 */
 09911	  int r_next;
 09912	  unsigned long tsc_high, tsc_low;
 09913	
 09914	  source %= RANDOM_SOURCES;
 09915	  r_next= krandom.bin[source].r_next;
 09916	  if (machine.processor > 486) {
 09917	      read_tsc(&tsc_high, &tsc_low);
 09918	      krandom.bin[source].r_buf[r_next] = tsc_low;
 09919	  } else {
 09920	      krandom.bin[source].r_buf[r_next] = read_clock();
 09921	  }
 09922	  if (krandom.bin[source].r_size < RANDOM_ELEMENTS) {
 09923	        krandom.bin[source].r_size ++;
 09924	  }
 09925	  krandom.bin[source].r_next = (r_next + 1 ) % RANDOM_ELEMENTS;
 09926	}
	
 09928	/*===========================================================================*
 09929	 *                              send_sig                                     *
 09930	 *===========================================================================*/
 09931	PUBLIC void send_sig(proc_nr, sig_nr)
 09932	int proc_nr;                    /* system process to be signalled */
 09933	int sig_nr;                     /* signal to be sent, 1 to _NSIG */
 09934	{
 09935	/* Notify a system process about a signal. This is straightforward. Simply
 09936	 * set the signal that is to be delivered in the pending signals map and 
 09937	 * send a notification with source SYSTEM.
 09938	 */ 
 09939	  register struct proc *rp;
 09940	
 09941	  rp = proc_addr(proc_nr);
 09942	  sigaddset(&priv(rp)->s_sig_pending, sig_nr);
 09943	  lock_notify(SYSTEM, proc_nr); 
 09944	}
	
 09946	/*===========================================================================*
 09947	 *                              cause_sig                                    *
 09948	 *===========================================================================*/
 09949	PUBLIC void cause_sig(proc_nr, sig_nr)
 09950	int proc_nr;                    /* process to be signalled */
 09951	int sig_nr;                     /* signal to be sent, 1 to _NSIG */
 09952	{
 09953	/* A system process wants to send a signal to a process.  Examples are:
 09954	 *  - HARDWARE wanting to cause a SIGSEGV after a CPU exception
 09955	 *  - TTY wanting to cause SIGINT upon getting a DEL
 09956	 *  - FS wanting to cause SIGPIPE for a broken pipe 
 09957	 * Signals are handled by sending a message to PM.  This function handles the 
 09958	 * signals and makes sure the PM gets them by sending a notification. The 
 09959	 * process being signaled is blocked while PM has not finished all signals 
 09960	 * for it. 
 09961	 * Race conditions between calls to this function and the system calls that
 09962	 * process pending kernel signals cannot exist. Signal related functions are
 09963	 * only called when a user process causes a CPU exception and from the kernel 
 09964	 * process level, which runs to completion.

_________________________ Page 760    File: kernel/system.c _________________________
 09965	 */
 09966	  register struct proc *rp;
 09967	
 09968	  /* Check if the signal is already pending. Process it otherwise. */
 09969	  rp = proc_addr(proc_nr);
 09970	  if (! sigismember(&rp->p_pending, sig_nr)) {
 09971	      sigaddset(&rp->p_pending, sig_nr);
 09972	      if (! (rp->p_rts_flags & SIGNALED)) {             /* other pending */
 09973	          if (rp->p_rts_flags == 0) lock_dequeue(rp);   /* make not ready */
 09974	          rp->p_rts_flags |= SIGNALED | SIG_PENDING;    /* update flags */
 09975	          send_sig(PM_PROC_NR, SIGKSIG);
 09976	      }
 09977	  }
 09978	}
	
 09980	/*===========================================================================*
 09981	 *                              umap_local                                   *
 09982	 *===========================================================================*/
 09983	PUBLIC phys_bytes umap_local(rp, seg, vir_addr, bytes)
 09984	register struct proc *rp;       /* pointer to proc table entry for process */
 09985	int seg;                        /* T, D, or S segment */
 09986	vir_bytes vir_addr;             /* virtual address in bytes within the seg */
 09987	vir_bytes bytes;                /* # of bytes to be copied */
 09988	{
 09989	/* Calculate the physical memory address for a given virtual address. */
 09990	  vir_clicks vc;                /* the virtual address in clicks */
 09991	  phys_bytes pa;                /* intermediate variables as phys_bytes */
 09992	  phys_bytes seg_base;
 09993	
 09994	  /* If 'seg' is D it could really be S and vice versa.  T really means T.
 09995	   * If the virtual address falls in the gap,  it causes a problem. On the
 09996	   * 8088 it is probably a legal stack reference, since "stackfaults" are
 09997	   * not detected by the hardware.  On 8088s, the gap is called S and
 09998	   * accepted, but on other machines it is called D and rejected.
 09999	   * The Atari ST behaves like the 8088 in this respect.
 10000	   */
 10001	
 10002	  if (bytes <= 0) return( (phys_bytes) 0);
 10003	  if (vir_addr + bytes <= vir_addr) return 0;   /* overflow */
 10004	  vc = (vir_addr + bytes - 1) >> CLICK_SHIFT;   /* last click of data */
 10005	
 10006	  if (seg != T)
 10007	        seg = (vc < rp->p_memmap[D].mem_vir + rp->p_memmap[D].mem_len ? D : S);
 10008	
 10009	  if ((vir_addr>>CLICK_SHIFT) >= rp->p_memmap[seg].mem_vir + 
 10010	        rp->p_memmap[seg].mem_len) return( (phys_bytes) 0 );
 10011	
 10012	  if (vc >= rp->p_memmap[seg].mem_vir + 
 10013	        rp->p_memmap[seg].mem_len) return( (phys_bytes) 0 );
 10014	
 10015	  seg_base = (phys_bytes) rp->p_memmap[seg].mem_phys;
 10016	  seg_base = seg_base << CLICK_SHIFT;   /* segment origin in bytes */
 10017	  pa = (phys_bytes) vir_addr;
 10018	  pa -= rp->p_memmap[seg].mem_vir << CLICK_SHIFT;
 10019	  return(seg_base + pa);
 10020	}
	

_________________________ Page 761    File: kernel/system.c _________________________
 10022	/*===========================================================================*
 10023	 *                              umap_remote                                  *
 10024	 *===========================================================================*/
 10025	PUBLIC phys_bytes umap_remote(rp, seg, vir_addr, bytes)
 10026	register struct proc *rp;       /* pointer to proc table entry for process */
 10027	int seg;                        /* index of remote segment */
 10028	vir_bytes vir_addr;             /* virtual address in bytes within the seg */
 10029	vir_bytes bytes;                /* # of bytes to be copied */
 10030	{
 10031	/* Calculate the physical memory address for a given virtual address. */
 10032	  struct far_mem *fm;
 10033	
 10034	  if (bytes <= 0) return( (phys_bytes) 0);
 10035	  if (seg < 0 || seg >= NR_REMOTE_SEGS) return( (phys_bytes) 0);
 10036	
 10037	  fm = &rp->p_priv->s_farmem[seg];
 10038	  if (! fm->in_use) return( (phys_bytes) 0);
 10039	  if (vir_addr + bytes > fm->mem_len) return( (phys_bytes) 0);
 10040	
 10041	  return(fm->mem_phys + (phys_bytes) vir_addr); 
 10042	}
	
 10044	/*===========================================================================*
 10045	 *                              umap_bios                                    *
 10046	 *===========================================================================*/
 10047	PUBLIC phys_bytes umap_bios(rp, vir_addr, bytes)
 10048	register struct proc *rp;       /* pointer to proc table entry for process */
 10049	vir_bytes vir_addr;             /* virtual address in BIOS segment */
 10050	vir_bytes bytes;                /* # of bytes to be copied */
 10051	{
 10052	/* Calculate the physical memory address at the BIOS. Note: currently, BIOS
 10053	 * address zero (the first BIOS interrupt vector) is not considered as an 
 10054	 * error here, but since the physical address will be zero as well, the 
 10055	 * calling function will think an error occurred. This is not a problem,
 10056	 * since no one uses the first BIOS interrupt vector.  
 10057	 */
 10058	
 10059	  /* Check all acceptable ranges. */
 10060	  if (vir_addr >= BIOS_MEM_BEGIN && vir_addr + bytes <= BIOS_MEM_END)
 10061	        return (phys_bytes) vir_addr;
 10062	  else if (vir_addr >= BASE_MEM_TOP && vir_addr + bytes <= UPPER_MEM_END
 10063	        return (phys_bytes) vir_addr;
 10064	  kprintf("Warning, error in umap_bios, virtual address 0x%x\n", vir_addr);
 10065	  return 0;
 10066	}
	
 10068	/*===========================================================================*
 10069	 *                              virtual_copy                                 *
 10070	 *===========================================================================*/
 10071	PUBLIC int virtual_copy(src_addr, dst_addr, bytes)
 10072	struct vir_addr *src_addr;      /* source virtual address */
 10073	struct vir_addr *dst_addr;      /* destination virtual address */
 10074	vir_bytes bytes;                /* # of bytes to copy  */
 10075	{
 10076	/* Copy bytes from virtual address src_addr to virtual address dst_addr. 
 10077	 * Virtual addresses can be in ABS, LOCAL_SEG, REMOTE_SEG, or BIOS_SEG.
 10078	 */
 10079	  struct vir_addr *vir_addr[2]; /* virtual source and destination address */
 10080	  phys_bytes phys_addr[2];      /* absolute source and destination */ 
 10081	  int seg_index;

_________________________ Page 762    File: kernel/system.c _________________________
 10082	  int i;
 10083	
 10084	  /* Check copy count. */
 10085	  if (bytes <= 0) return(EDOM);
 10086	
 10087	  /* Do some more checks and map virtual addresses to physical addresses. */
 10088	  vir_addr[_SRC_] = src_addr;
 10089	  vir_addr[_DST_] = dst_addr;
 10090	  for (i=_SRC_; i<=_DST_; i++) {
 10091	
 10092	      /* Get physical address. */
 10093	      switch((vir_addr[i]->segment & SEGMENT_TYPE)) {
 10094	      case LOCAL_SEG:
 10095	          seg_index = vir_addr[i]->segment & SEGMENT_INDEX;
 10096	          phys_addr[i] = umap_local( proc_addr(vir_addr[i]->proc_nr), 
 10097	              seg_index, vir_addr[i]->offset, bytes );
 10098	          break;
 10099	      case REMOTE_SEG:
 10100	          seg_index = vir_addr[i]->segment & SEGMENT_INDEX;
 10101	          phys_addr[i] = umap_remote( proc_addr(vir_addr[i]->proc_nr), 
 10102	              seg_index, vir_addr[i]->offset, bytes );
 10103	          break;
 10104	      case BIOS_SEG:
 10105	          phys_addr[i] = umap_bios( proc_addr(vir_addr[i]->proc_nr),
 10106	              vir_addr[i]->offset, bytes );
 10107	          break;
 10108	      case PHYS_SEG:
 10109	          phys_addr[i] = vir_addr[i]->offset;
 10110	          break;
 10111	      default:
 10112	          return(EINVAL);
 10113	      }
 10114	
 10115	      /* Check if mapping succeeded. */
 10116	      if (phys_addr[i] <= 0 && vir_addr[i]->segment != PHYS_SEG) 
 10117	          return(EFAULT);
 10118	  }
 10119	
 10120	  /* Now copy bytes between physical addresseses. */
 10121	  phys_copy(phys_addr[_SRC_], phys_addr[_DST_], (phys_bytes) bytes);
 10122	  return(OK);
 10123	}
	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/system/do_setalarm.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 10200	/* The kernel call implemented in this file:
 10201	 *   m_type:    SYS_SETALARM 
 10202	 *
 10203	 * The parameters for this kernel call are:
 10204	 *    m2_l1:    ALRM_EXP_TIME           (alarm's expiration time)
 10205	 *    m2_i2:    ALRM_ABS_TIME           (expiration time is absolute?)
 10206	 *    m2_l1:    ALRM_TIME_LEFT          (return seconds left of previous)
 10207	 */
 10208	
 10209	#include "../system.h"

_________________________ Page 763    File: kernel/system/do_setalarm.c __________________
 10210	
 10211	#if USE_SETALARM
 10212	
 10213	FORWARD _PROTOTYPE( void cause_alarm, (timer_t *tp) );
 10214	
 10215	/*===========================================================================*
 10216	 *                              do_setalarm                                  *
 10217	 *===========================================================================*/
 10218	PUBLIC int do_setalarm(m_ptr)
 10219	message *m_ptr;                 /* pointer to request message */
 10220	{
 10221	/* A process requests a synchronous alarm, or wants to cancel its alarm. */
 10222	  register struct proc *rp;     /* pointer to requesting process */
 10223	  int proc_nr;                  /* which process wants the alarm */
 10224	  long exp_time;                /* expiration time for this alarm */
 10225	  int use_abs_time;             /* use absolute or relative time */
 10226	  timer_t *tp;                  /* the process' timer structure */
 10227	  clock_t uptime;               /* placeholder for current uptime */
 10228	
 10229	  /* Extract shared parameters from the request message. */
 10230	  exp_time = m_ptr->ALRM_EXP_TIME;      /* alarm's expiration time */
 10231	  use_abs_time = m_ptr->ALRM_ABS_TIME;  /* flag for absolute time */
 10232	  proc_nr = m_ptr->m_source;            /* process to interrupt later */
 10233	  rp = proc_addr(proc_nr);
 10234	  if (! (priv(rp)->s_flags & SYS_PROC)) return(EPERM);
 10235	
 10236	  /* Get the timer structure and set the parameters for this alarm. */
 10237	  tp = &(priv(rp)->s_alarm_timer);      
 10238	  tmr_arg(tp)->ta_int = proc_nr;        
 10239	  tp->tmr_func = cause_alarm; 
 10240	
 10241	  /* Return the ticks left on the previous alarm. */
 10242	  uptime = get_uptime(); 
 10243	  if ((tp->tmr_exp_time != TMR_NEVER) && (uptime < tp->tmr_exp_time) ) {
 10244	      m_ptr->ALRM_TIME_LEFT = (tp->tmr_exp_time - uptime);
 10245	  } else {
 10246	      m_ptr->ALRM_TIME_LEFT = 0;
 10247	  }
 10248	
 10249	  /* Finally, (re)set the timer depending on the expiration time. */
 10250	  if (exp_time == 0) {
 10251	      reset_timer(tp);
 10252	  } else {
 10253	      tp->tmr_exp_time = (use_abs_time) ? exp_time : exp_time + get_uptime();
 10254	      set_timer(tp, tp->tmr_exp_time, tp->tmr_func);
 10255	  }
 10256	  return(OK);
 10257	}
	
 10259	/*===========================================================================*
 10260	 *                              cause_alarm                                  *
 10261	 *===========================================================================*/
 10262	PRIVATE void cause_alarm(tp)
 10263	timer_t *tp;
 10264	{
 10265	/* Routine called if a timer goes off and the process requested a synchronous
 10266	 * alarm. The process number is stored in timer argument 'ta_int'. Notify that
 10267	 * process with a notification message from CLOCK.
 10268	 */
 10269	  int proc_nr = tmr_arg(tp)->ta_int;            /* get process number */

_________________________ Page 764    File: kernel/system/do_setalarm.c __________________
 10270	  lock_notify(CLOCK, proc_nr);                  /* notify process */
 10271	}
	
 10273	#endif /* USE_SETALARM */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/system/do_exec.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 10300	/* The kernel call implemented in this file:
 10301	 *   m_type:    SYS_EXEC
 10302	 *
 10303	 * The parameters for this kernel call are:
 10304	 *    m1_i1:    PR_PROC_NR              (process that did exec call)
 10305	 *    m1_p1:    PR_STACK_PTR            (new stack pointer)
 10306	 *    m1_p2:    PR_NAME_PTR             (pointer to program name)
 10307	 *    m1_p3:    PR_IP_PTR               (new instruction pointer)
 10308	 */
 10309	#include "../system.h"
 10310	#include <string.h>
 10311	#include <signal.h>
 10312	
 10313	#if USE_EXEC
 10314	
 10315	/*===========================================================================*
 10316	 *                              do_exec                                      *
 10317	 *===========================================================================*/
 10318	PUBLIC int do_exec(m_ptr)
 10319	register message *m_ptr;        /* pointer to request message */
 10320	{
 10321	/* Handle sys_exec().  A process has done a successful EXEC. Patch it up. */
 10322	  register struct proc *rp;
 10323	  reg_t sp;                     /* new sp */
 10324	  phys_bytes phys_name;
 10325	  char *np;
 10326	
 10327	  rp = proc_addr(m_ptr->PR_PROC_NR);
 10328	  sp = (reg_t) m_ptr->PR_STACK_PTR;
 10329	  rp->p_reg.sp = sp;            /* set the stack pointer */
 10330	  phys_memset(vir2phys(&rp->p_ldt[EXTRA_LDT_INDEX]), 0,
 10331	        (LDT_SIZE - EXTRA_LDT_INDEX) * sizeof(rp->p_ldt[0]));
 10332	  rp->p_reg.pc = (reg_t) m_ptr->PR_IP_PTR;      /* set pc */
 10333	  rp->p_rts_flags &= ~RECEIVING;        /* PM does not reply to EXEC call */
 10334	  if (rp->p_rts_flags == 0) lock_enqueue(rp);
 10335	
 10336	  /* Save command name for debugging, ps(1) output, etc. */
 10337	  phys_name = numap_local(m_ptr->m_source, (vir_bytes) m_ptr->PR_NAME_PTR,
 10338	                                        (vir_bytes) P_NAME_LEN - 1);
 10339	  if (phys_name != 0) {
 10340	        phys_copy(phys_name, vir2phys(rp->p_name), (phys_bytes) P_NAME_LEN - 1);
 10341	        for (np = rp->p_name; (*np & BYTE) >= ' '; np++) {}
 10342	        *np = 0;                                        /* mark end */
 10343	  } else {
 10344	        strncpy(rp->p_name, "<unset>", P_NAME_LEN);
 10345	  }
 10346	  return(OK);
 10347	}
 10348	#endif /* USE_EXEC */

_________________________ Page 765    File: kernel/system/do_exec.c ______________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      kernel/clock.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 10400	/* This file contains the clock task, which handles time related functions.
 10401	 * Important events that are handled by the CLOCK include setting and 
 10402	 * monitoring alarm timers and deciding when to (re)schedule processes. 
 10403	 * The CLOCK offers a direct interface to kernel processes. System services 
 10404	 * can access its services through system calls, such as sys_setalarm(). The
 10405	 * CLOCK task thus is hidden from the outside world.  
 10406	 *
 10407	 * Changes:
 10408	 *   Oct 08, 2005   reordering and comment editing (A. S. Woodhull)
 10409	 *   Mar 18, 2004   clock interface moved to SYSTEM task (Jorrit N. Herder) 
 10410	 *   Sep 30, 2004   source code documentation updated  (Jorrit N. Herder)
 10411	 *   Sep 24, 2004   redesigned alarm timers  (Jorrit N. Herder)
 10412	 *
 10413	 * The function do_clocktick() is triggered by the clock's interrupt 
 10414	 * handler when a watchdog timer has expired or a process must be scheduled. 
 10415	 *
 10416	 * In addition to the main clock_task() entry point, which starts the main 
 10417	 * loop, there are several other minor entry points:
 10418	 *   clock_stop:        called just before MINIX shutdown
 10419	 *   get_uptime:        get realtime since boot in clock ticks
 10420	 *   set_timer:         set a watchdog timer (+)
 10421	 *   reset_timer:       reset a watchdog timer (+)
 10422	 *   read_clock:        read the counter of channel 0 of the 8253A timer
 10423	 *
 10424	 * (+) The CLOCK task keeps tracks of watchdog timers for the entire kernel.
 10425	 * The watchdog functions of expired timers are executed in do_clocktick(). 
 10426	 * It is crucial that watchdog functions not block, or the CLOCK task may
 10427	 * be blocked. Do not send() a message when the receiver is not expecting it.
 10428	 * Instead, notify(), which always returns, should be used. 
 10429	 */
 10430	
 10431	#include "kernel.h"
 10432	#include "proc.h"
 10433	#include <signal.h>
 10434	#include <minix/com.h>
 10435	
 10436	/* Function prototype for PRIVATE functions. */ 
 10437	FORWARD _PROTOTYPE( void init_clock, (void) );
 10438	FORWARD _PROTOTYPE( int clock_handler, (irq_hook_t *hook) );
 10439	FORWARD _PROTOTYPE( int do_clocktick, (message *m_ptr) );
 10440	
 10441	/* Clock parameters. */
 10442	#define COUNTER_FREQ (2*TIMER_FREQ) /* counter frequency using square wave */
 10443	#define LATCH_COUNT     0x00    /* cc00xxxx, c = channel, x = any */
 10444	#define SQUARE_WAVE     0x36    /* ccaammmb, a = access, m = mode, b = BCD */
 10445	                                /*   11x11, 11 = LSB then MSB, x11 = sq wave */
 10446	#define TIMER_COUNT ((unsigned) (TIMER_FREQ/HZ)) /* initial value for counter*/
 10447	#define TIMER_FREQ  1193182L    /* clock frequency for timer in PC and AT */
 10448	
 10449	#define CLOCK_ACK_BIT   0x80    /* PS/2 clock interrupt acknowledge bit */
 10450	
 10451	/* The CLOCK's timers queue. The functions in <timers.h> operate on this. 
 10452	 * Each system process possesses a single synchronous alarm timer. If other 
 10453	 * kernel parts want to use additional timers, they must declare their own 
 10454	 * persistent (static) timer structure, which can be passed to the clock

_________________________ Page 766    File: kernel/clock.c _________________________
 10455	 * via (re)set_timer().
 10456	 * When a timer expires its watchdog function is run by the CLOCK task. 
 10457	 */
 10458	PRIVATE timer_t *clock_timers;          /* queue of CLOCK timers */
 10459	PRIVATE clock_t next_timeout;           /* realtime that next timer expires */
 10460	
 10461	/* The time is incremented by the interrupt handler on each clock tick. */
 10462	PRIVATE clock_t realtime;               /* real time clock */
 10463	PRIVATE irq_hook_t clock_hook;          /* interrupt handler hook */
 10464	
 10465	/*===========================================================================*
 10466	 *                              clock_task                                   *
 10467	 *===========================================================================*/
 10468	PUBLIC void clock_task()
 10469	{
 10470	/* Main program of clock task. If the call is not HARD_INT it is an error.
 10471	 */
 10472	  message m;                    /* message buffer for both input and output */
 10473	  int result;                   /* result returned by the handler */
 10474	
 10475	  init_clock();                 /* initialize clock task */
 10476	
 10477	  /* Main loop of the clock task.  Get work, process it. Never reply. */
 10478	  while (TRUE) {
 10479	
 10480	      /* Go get a message. */
 10481	      receive(ANY, &m); 
 10482	
 10483	      /* Handle the request. Only clock ticks are expected. */
 10484	      switch (m.m_type) {
 10485	      case HARD_INT:
 10486	          result = do_clocktick(&m);    /* handle clock tick */
 10487	          break;
 10488	      default:                          /* illegal request type */
 10489	          kprintf("CLOCK: illegal request %d from %d.\n", m.m_type,m.m_source);
 10490	      }
 10491	  }
 10492	}
	
 10494	/*===========================================================================*
 10495	 *                              do_clocktick                                 *
 10496	 *===========================================================================*/
 10497	PRIVATE int do_clocktick(m_ptr)
 10498	message *m_ptr;                         /* pointer to request message */
 10499	{
 10500	/* Despite its name, this routine is not called on every clock tick. It
 10501	 * is called on those clock ticks when a lot of work needs to be done.
 10502	 */
 10503	
 10504	  /* A process used up a full quantum. The interrupt handler stored this
 10505	   * process in 'prev_ptr'.  First make sure that the process is not on the 
 10506	   * scheduling queues.  Then announce the process ready again. Since it has 
 10507	   * no more time left, it gets a new quantum and is inserted at the right 
 10508	   * place in the queues.  As a side-effect a new process will be scheduled.
 10509	   */ 
 10510	  if (prev_ptr->p_ticks_left <= 0 && priv(prev_ptr)->s_flags & PREEM
 10511	      lock_dequeue(prev_ptr);           /* take it off the queues */
 10512	      lock_enqueue(prev_ptr);           /* and reinsert it again */ 
 10513	  }
 10514	

_________________________ Page 767    File: kernel/clock.c _________________________
 10515	  /* Check if a clock timer expired and run its watchdog function. */
 10516	  if (next_timeout <= realtime) { 
 10517	        tmrs_exptimers(&clock_timers, realtime, NULL);
 10518	        next_timeout = clock_timers == NULL ? 
 10519	                TMR_NEVER : clock_timers->tmr_exp_time;
 10520	  }
 10521	
 10522	  /* Inhibit sending a reply. */
 10523	  return(EDONTREPLY);
 10524	}
	
 10526	/*===========================================================================*
 10527	 *                              init_clock                                   *
 10528	 *===========================================================================*/
 10529	PRIVATE void init_clock()
 10530	{
 10531	  /* Initialize the CLOCK's interrupt hook. */
 10532	  clock_hook.proc_nr = CLOCK;
 10533	
 10534	  /* Initialize channel 0 of the 8253A timer to, e.g., 60 Hz. */
 10535	  outb(TIMER_MODE, SQUARE_WAVE);        /* set timer to run continuously */
 10536	  outb(TIMER0, TIMER_COUNT);            /* load timer low byte */
 10537	  outb(TIMER0, TIMER_COUNT >> 8);       /* load timer high byte */
 10538	  put_irq_handler(&clock_hook, CLOCK_IRQ, clock_handler);/* register handler */
 10539	  enable_irq(&clock_hook);              /* ready for clock interrupts */
 10540	}
	
 10542	/*===========================================================================*
 10543	 *                              clock_stop                                   *
 10544	 *===========================================================================*/
 10545	PUBLIC void clock_stop()
 10546	{
 10547	/* Reset the clock to the BIOS rate. (For rebooting) */
 10548	  outb(TIMER_MODE, 0x36);
 10549	  outb(TIMER0, 0);
 10550	  outb(TIMER0, 0);
 10551	}
	
 10553	/*===========================================================================*
 10554	 *                              clock_handler                                *
 10555	 *===========================================================================*/
 10556	PRIVATE int clock_handler(hook)
 10557	irq_hook_t *hook;
 10558	{
 10559	/* This executes on each clock tick (i.e., every time the timer chip generates 
 10560	 * an interrupt). It does a little bit of work so the clock task does not have 
 10561	 * to be called on every tick.  The clock task is called when:
 10562	 *
 10563	 *      (1) the scheduling quantum of the running process has expired, or
 10564	 *      (2) a timer has expired and the watchdog function should be run.
 10565	 *
 10566	 * Many global global and static variables are accessed here.  The safety of
 10567	 * this must be justified. All scheduling and message passing code acquires a 
 10568	 * lock by temporarily disabling interrupts, so no conflicts with calls from 
 10569	 * the task level can occur. Furthermore, interrupts are not reentrant, the 
 10570	 * interrupt handler cannot be bothered by other interrupts.
 10571	 * 
 10572	 * Variables that are updated in the clock's interrupt handler:
 10573	 *      lost_ticks:
 10574	 *              Clock ticks counted outside the clock task. This for example

_________________________ Page 768    File: kernel/clock.c _________________________
 10575	 *              is used when the boot monitor processes a real mode interrupt.
 10576	 *      realtime:
 10577	 *              The current uptime is incremented with all outstanding ticks.
 10578	 *      proc_ptr, bill_ptr:
 10579	 *              These are used for accounting.  It does not matter if proc.c
 10580	 *              is changing them, provided they are always valid pointers,
 10581	 *              since at worst the previous process would be billed.
 10582	 */
 10583	  register unsigned ticks;
 10584	
 10585	  /* Acknowledge the PS/2 clock interrupt. */
 10586	  if (machine.ps_mca) outb(PORT_B, inb(PORT_B) | CLOCK_ACK_BIT);
 10587	
 10588	  /* Get number of ticks and update realtime. */
 10589	  ticks = lost_ticks + 1;
 10590	  lost_ticks = 0;
 10591	  realtime += ticks;
 10592	
 10593	  /* Update user and system accounting times. Charge the current process for
 10594	   * user time. If the current process is not billable, that is, if a non-user
 10595	   * process is running, charge the billable process for system time as well.
 10596	   * Thus the unbillable process' user time is the billable user's system time.
 10597	   */
 10598	  proc_ptr->p_user_time += ticks;
 10599	  if (priv(proc_ptr)->s_flags & PREEMPTIBLE) {
 10600	      proc_ptr->p_ticks_left -= ticks;
 10601	  }
 10602	  if (! (priv(proc_ptr)->s_flags & BILLABLE)) {
 10603	      bill_ptr->p_sys_time += ticks;
 10604	      bill_ptr->p_ticks_left -= ticks;
 10605	  }
 10606	
 10607	  /* Check if do_clocktick() must be called. Done for alarms and scheduling.
 10608	   * Some processes, such as the kernel tasks, cannot be preempted. 
 10609	   */ 
 10610	  if ((next_timeout <= realtime) || (proc_ptr->p_ticks_left <= 0)) {
 10611	      prev_ptr = proc_ptr;                      /* store running process */
 10612	      lock_notify(HARDWARE, CLOCK);             /* send notification */
 10613	  } 
 10614	  return(1);                                    /* reenable interrupts */
 10615	}
	
 10617	/*===========================================================================*
 10618	 *                              get_uptime                                   *
 10619	 *===========================================================================*/
 10620	PUBLIC clock_t get_uptime()
 10621	{
 10622	/* Get and return the current clock uptime in ticks. */
 10623	  return(realtime);
 10624	}
	
 10626	/*===========================================================================*
 10627	 *                              set_timer                                    *
 10628	 *===========================================================================*/
 10629	PUBLIC void set_timer(tp, exp_time, watchdog)
 10630	struct timer *tp;               /* pointer to timer structure */
 10631	clock_t exp_time;               /* expiration realtime */
 10632	tmr_func_t watchdog;            /* watchdog to be called */
 10633	{
 10634	/* Insert the new timer in the active timers list. Always update the 

_________________________ Page 769    File: kernel/clock.c _________________________
 10635	 * next timeout time by setting it to the front of the active list.
 10636	 */
 10637	  tmrs_settimer(&clock_timers, tp, exp_time, watchdog, NULL);
 10638	  next_timeout = clock_timers->tmr_exp_time;
 10639	}
	
 10641	/*===========================================================================*
 10642	 *                              reset_timer                                  *
 10643	 *===========================================================================*/
 10644	PUBLIC void reset_timer(tp)
 10645	struct timer *tp;               /* pointer to timer structure */
 10646	{
 10647	/* The timer pointed to by 'tp' is no longer needed. Remove it from both the
 10648	 * active and expired lists. Always update the next timeout time by setting
 10649	 * it to the front of the active list.
 10650	 */
 10651	  tmrs_clrtimer(&clock_timers, tp, NULL);
 10652	  next_timeout = (clock_timers == NULL) ? 
 10653	        TMR_NEVER : clock_timers->tmr_exp_time;
 10654	}
	
 10656	/*===========================================================================*
 10657	 *                              read_clock                                   *
 10658	 *===========================================================================*/
 10659	PUBLIC unsigned long read_clock()
 10660	{
 10661	/* Read the counter of channel 0 of the 8253A timer.  This counter counts
 10662	 * down at a rate of TIMER_FREQ and restarts at TIMER_COUNT-1 when it
 10663	 * reaches zero. A hardware interrupt (clock tick) occurs when the counter
 10664	 * gets to zero and restarts its cycle.  
 10665	 */
 10666	  unsigned count;
 10667	
 10668	  outb(TIMER_MODE, LATCH_COUNT);
 10669	  count = inb(TIMER0);
 10670	  count |= (inb(TIMER0) << 8);
 10671	  
 10672	  return count;
 10673	}


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/drivers.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 10700	/* This is the master header for all device drivers. It includes some other
 10701	 * files and defines the principal constants.
 10702	 */
 10703	#define _POSIX_SOURCE      1    /* tell headers to include POSIX stuff */
 10704	#define _MINIX             1    /* tell headers to include MINIX stuff */
 10705	#define _SYSTEM            1    /* get negative error number in <errno.h> */
 10706	
 10707	/* The following are so basic, all the *.c files get them automatically. */
 10708	#include <minix/config.h>       /* MUST be first */
 10709	#include <ansi.h>               /* MUST be second */
 10710	#include <minix/type.h>
 10711	#include <minix/com.h>
 10712	#include <minix/dmap.h>
 10713	#include <minix/callnr.h>
 10714	#include <sys/types.h>

_________________________ Page 770    File: drivers/drivers.h _________________________
 10715	#include <minix/const.h>
 10716	#include <minix/devio.h>
 10717	#include <minix/syslib.h>
 10718	#include <minix/sysutil.h>
 10719	#include <minix/bitmap.h>
 10720	
 10721	#include <ibm/interrupt.h>      /* IRQ vectors and miscellaneous ports */
 10722	#include <ibm/bios.h>           /* BIOS index numbers */
 10723	#include <ibm/ports.h>          /* Well-known ports */
 10724	
 10725	#include <string.h>
 10726	#include <signal.h>
 10727	#include <stdlib.h>
 10728	#include <limits.h>
 10729	#include <stddef.h>
 10730	#include <errno.h>
 10731	#include <unistd.h>
 10732	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/libdriver/driver.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 10800	/* Types and constants shared between the generic and device dependent
 10801	 * device driver code.
 10802	 */
 10803	
 10804	#define _POSIX_SOURCE      1    /* tell headers to include POSIX stuff */
 10805	#define _MINIX             1    /* tell headers to include MINIX stuff */
 10806	#define _SYSTEM            1    /* get negative error number in <errno.h> */
 10807	
 10808	/* The following are so basic, all the *.c files get them automatically. */
 10809	#include <minix/config.h>       /* MUST be first */
 10810	#include <ansi.h>               /* MUST be second */
 10811	#include <minix/type.h>
 10812	#include <minix/ipc.h>
 10813	#include <minix/com.h>
 10814	#include <minix/callnr.h>
 10815	#include <sys/types.h>
 10816	#include <minix/const.h>
 10817	#include <minix/syslib.h>
 10818	#include <minix/sysutil.h>
 10819	
 10820	#include <string.h>
 10821	#include <limits.h>
 10822	#include <stddef.h>
 10823	#include <errno.h>
 10824	
 10825	#include <minix/partition.h>
 10826	#include <minix/u64.h>
 10827	
 10828	/* Info about and entry points into the device dependent code. */
 10829	struct driver {
 10830	  _PROTOTYPE( char *(*dr_name), (void) );
 10831	  _PROTOTYPE( int (*dr_open), (struct driver *dp, message *m_ptr) );
 10832	  _PROTOTYPE( int (*dr_close), (struct driver *dp, message *m_ptr) );
 10833	  _PROTOTYPE( int (*dr_ioctl), (struct driver *dp, message *m_ptr) );
 10834	  _PROTOTYPE( struct device *(*dr_prepare), (int device) );

_________________________ Page 771    File: drivers/libdriver/driver.h ___________________
 10835	  _PROTOTYPE( int (*dr_transfer), (int proc_nr, int opcode, off_t position,
 10836	                                        iovec_t *iov, unsigned nr_req) );
 10837	  _PROTOTYPE( void (*dr_cleanup), (void) );
 10838	  _PROTOTYPE( void (*dr_geometry), (struct partition *entry) );
 10839	  _PROTOTYPE( void (*dr_signal), (struct driver *dp, message *m_ptr) );
 10840	  _PROTOTYPE( void (*dr_alarm), (struct driver *dp, message *m_ptr) );
 10841	  _PROTOTYPE( int (*dr_cancel), (struct driver *dp, message *m_ptr) );
 10842	  _PROTOTYPE( int (*dr_select), (struct driver *dp, message *m_ptr) );
 10843	  _PROTOTYPE( int (*dr_other), (struct driver *dp, message *m_ptr) );
 10844	  _PROTOTYPE( int (*dr_hw_int), (struct driver *dp, message *m_ptr) );
 10845	};
 10846	
 10847	#if (CHIP == INTEL)
 10848	
 10849	/* Number of bytes you can DMA before hitting a 64K boundary: */
 10850	#define dma_bytes_left(phys)    \
 10851	   ((unsigned) (sizeof(int) == 2 ? 0 : 0x10000) - (unsigned) ((phys) & 0xFFFF))
 10852	
 10853	#endif /* CHIP == INTEL */
 10854	
 10855	/* Base and size of a partition in bytes. */
 10856	struct device {
 10857	  u64_t dv_base;
 10858	  u64_t dv_size;
 10859	};
 10860	
 10861	#define NIL_DEV         ((struct device *) 0)
 10862	
 10863	/* Functions defined by driver.c: */
 10864	_PROTOTYPE( void driver_task, (struct driver *dr) );
 10865	_PROTOTYPE( char *no_name, (void) );
 10866	_PROTOTYPE( int do_nop, (struct driver *dp, message *m_ptr) );
 10867	_PROTOTYPE( struct device *nop_prepare, (int device) );
 10868	_PROTOTYPE( void nop_cleanup, (void) );
 10869	_PROTOTYPE( void nop_task, (void) );
 10870	_PROTOTYPE( void nop_signal, (struct driver *dp, message *m_ptr) );
 10871	_PROTOTYPE( void nop_alarm, (struct driver *dp, message *m_ptr) );
 10872	_PROTOTYPE( int nop_cancel, (struct driver *dp, message *m_ptr) );
 10873	_PROTOTYPE( int nop_select, (struct driver *dp, message *m_ptr) );
 10874	_PROTOTYPE( int do_diocntl, (struct driver *dp, message *m_ptr) );
 10875	
 10876	/* Parameters for the disk drive. */
 10877	#define SECTOR_SIZE      512    /* physical sector size in bytes */
 10878	#define SECTOR_SHIFT       9    /* for division */
 10879	#define SECTOR_MASK      511    /* and remainder */
 10880	
 10881	/* Size of the DMA buffer buffer in bytes. */
 10882	#define USE_EXTRA_DMA_BUF  0    /* usually not needed */
 10883	#define DMA_BUF_SIZE    (DMA_SECTORS * SECTOR_SIZE)
 10884	
 10885	#if (CHIP == INTEL)
 10886	extern u8_t *tmp_buf;                   /* the DMA buffer */
 10887	#else
 10888	extern u8_t tmp_buf[];                  /* the DMA buffer */
 10889	#endif
 10890	extern phys_bytes tmp_phys;             /* phys address of DMA buffer */

_________________________ Page 772    File: drivers/libdriver/driver.h ___________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/libdriver/drvlib.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 10900	/* IBM device driver definitions                        Author: Kees J. Bot
 10901	 *                                                              7 Dec 1995
 10902	 */
 10903	
 10904	#include <ibm/partition.h>
 10905	
 10906	_PROTOTYPE( void partition, (struct driver *dr, int device, int style, int atapi) )
 10907	
 10908	/* BIOS parameter table layout. */
 10909	#define bp_cylinders(t)         (* (u16_t *) (&(t)[0]))
 10910	#define bp_heads(t)             (* (u8_t *)  (&(t)[2]))
 10911	#define bp_reduced_wr(t)        (* (u16_t *) (&(t)[3]))
 10912	#define bp_precomp(t)           (* (u16_t *) (&(t)[5]))
 10913	#define bp_max_ecc(t)           (* (u8_t *)  (&(t)[7]))
 10914	#define bp_ctlbyte(t)           (* (u8_t *)  (&(t)[8]))
 10915	#define bp_landingzone(t)       (* (u16_t *) (&(t)[12]))
 10916	#define bp_sectors(t)           (* (u8_t *)  (&(t)[14]))
 10917	
 10918	/* Miscellaneous. */
 10919	#define DEV_PER_DRIVE   (1 + NR_PARTITIONS)
 10920	#define MINOR_t0        64
 10921	#define MINOR_r0        120
 10922	#define MINOR_d0p0s0    128
 10923	#define MINOR_fd0p0     (28<<2)
 10924	#define P_FLOPPY        0
 10925	#define P_PRIMARY       1
 10926	#define P_SUB           2




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/libdriver/driver.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 11000	/* This file contains device independent device driver interface.
 11001	 *
 11002	 * Changes:
 11003	 *   Jul 25, 2005   added SYS_SIG type for signals  (Jorrit N. Herder)
 11004	 *   Sep 15, 2004   added SYN_ALARM type for timeouts  (Jorrit N. Herder)
 11005	 *   Jul 23, 2004   removed kernel dependencies  (Jorrit N. Herder)
 11006	 *   Apr 02, 1992   constructed from AT wini and floppy driver  (Kees J. Bot)
 11007	 *
 11008	 *
 11009	 * The drivers support the following operations (using message format m2):
 11010	 *
 11011	 *    m_type      DEVICE    PROC_NR     COUNT    POSITION  ADRRESS
 11012	 * ----------------------------------------------------------------
 11013	 * |  DEV_OPEN  | device  | proc nr |         |         |         |
 11014	 * |------------+---------+---------+---------+---------+---------|
 11015	 * |  DEV_CLOSE | device  | proc nr |         |         |         |
 11016	 * |------------+---------+---------+---------+---------+---------|
 11017	 * |  DEV_READ  | device  | proc nr |  bytes  |  offset | buf ptr |
 11018	 * |------------+---------+---------+---------+---------+---------|
 11019	 * |  DEV_WRITE | device  | proc nr |  bytes  |  offset | buf ptr |

_________________________ Page 773    File: drivers/libdriver/driver.c ___________________
 11020	 * |------------+---------+---------+---------+---------+---------|
 11021	 * | DEV_GATHER | device  | proc nr | iov len |  offset | iov ptr |
 11022	 * |------------+---------+---------+---------+---------+---------|
 11023	 * | DEV_SCATTER| device  | proc nr | iov len |  offset | iov ptr |
 11024	 * |------------+---------+---------+---------+---------+---------|
 11025	 * |  DEV_IOCTL | device  | proc nr |func code|         | buf ptr |
 11026	 * |------------+---------+---------+---------+---------+---------|
 11027	 * |  CANCEL    | device  | proc nr | r/w     |         |         |
 11028	 * |------------+---------+---------+---------+---------+---------|
 11029	 * |  HARD_STOP |         |         |         |         |         |
 11030	 * ----------------------------------------------------------------
 11031	 *
 11032	 * The file contains one entry point:
 11033	 *
 11034	 *   driver_task:       called by the device dependent task entry
 11035	 */
 11036	
 11037	#include "../drivers.h"
 11038	#include <sys/ioc_disk.h>
 11039	#include "driver.h"
 11040	
 11041	#define BUF_EXTRA       0
 11042	
 11043	/* Claim space for variables. */
 11044	PRIVATE u8_t buffer[(unsigned) 2 * DMA_BUF_SIZE + BUF_EXTRA];
 11045	u8_t *tmp_buf;                  /* the DMA buffer eventually */
 11046	phys_bytes tmp_phys;            /* phys address of DMA buffer */
 11047	
 11048	FORWARD _PROTOTYPE( void init_buffer, (void) );
 11049	FORWARD _PROTOTYPE( int do_rdwt, (struct driver *dr, message *mp) );
 11050	FORWARD _PROTOTYPE( int do_vrdwt, (struct driver *dr, message *mp) );
 11051	
 11052	int device_caller;
 11053	
 11054	/*===========================================================================*
 11055	 *                              driver_task                                  *
 11056	 *===========================================================================*/
 11057	PUBLIC void driver_task(dp)
 11058	struct driver *dp;      /* Device dependent entry points. */
 11059	{
 11060	/* Main program of any device driver task. */
 11061	
 11062	  int r, proc_nr;
 11063	  message mess;
 11064	
 11065	  /* Get a DMA buffer. */
 11066	  init_buffer();
 11067	
 11068	  /* Here is the main loop of the disk task.  It waits for a message, carries
 11069	   * it out, and sends a reply.
 11070	   */
 11071	  while (TRUE) {
 11072	
 11073	        /* Wait for a request to read or write a disk block. */
 11074	        if(receive(ANY, &mess) != OK) continue;
 11075	
 11076	        device_caller = mess.m_source;
 11077	        proc_nr = mess.PROC_NR;
 11078	
 11079	        /* Now carry out the work. */

_________________________ Page 774    File: drivers/libdriver/driver.c ___________________
 11080	        switch(mess.m_type) {
 11081	        case DEV_OPEN:          r = (*dp->dr_open)(dp, &mess);  break;
 11082	        case DEV_CLOSE:         r = (*dp->dr_close)(dp, &mess); break;
 11083	        case DEV_IOCTL:         r = (*dp->dr_ioctl)(dp, &mess); break;
 11084	        case CANCEL:            r = (*dp->dr_cancel)(dp, &mess);break;
 11085	        case DEV_SELECT:        r = (*dp->dr_select)(dp, &mess);break;
 11086	
 11087	        case DEV_READ:  
 11088	        case DEV_WRITE:   r = do_rdwt(dp, &mess);       break;
 11089	        case DEV_GATHER: 
 11090	        case DEV_SCATTER: r = do_vrdwt(dp, &mess);      break;
 11091	
 11092	        case HARD_INT:          /* leftover interrupt or expired timer. */
 11093	                                if(dp->dr_hw_int) {
 11094	                                        (*dp->dr_hw_int)(dp, &mess);
 11095	                                }
 11096	                                continue;
 11097	        case SYS_SIG:           (*dp->dr_signal)(dp, &mess);
 11098	                                continue;       /* don't reply */
 11099	        case SYN_ALARM:         (*dp->dr_alarm)(dp, &mess);     
 11100	                                continue;       /* don't reply */
 11101	        default:                
 11102	                if(dp->dr_other)
 11103	                        r = (*dp->dr_other)(dp, &mess);
 11104	                else    
 11105	                        r = EINVAL;
 11106	                break;
 11107	        }
 11108	
 11109	        /* Clean up leftover state. */
 11110	        (*dp->dr_cleanup)();
 11111	
 11112	        /* Finally, prepare and send the reply message. */
 11113	        if (r != EDONTREPLY) {
 11114	                mess.m_type = TASK_REPLY;
 11115	                mess.REP_PROC_NR = proc_nr;
 11116	                /* Status is # of bytes transferred or error code. */
 11117	                mess.REP_STATUS = r;    
 11118	                send(device_caller, &mess);
 11119	        }
 11120	  }
 11121	}
	
 11123	/*===========================================================================*
 11124	 *                              init_buffer                                  *
 11125	 *===========================================================================*/
 11126	PRIVATE void init_buffer()
 11127	{
 11128	/* Select a buffer that can safely be used for DMA transfers.  It may also
 11129	 * be used to read partition tables and such.  Its absolute address is
 11130	 * 'tmp_phys', the normal address is 'tmp_buf'.
 11131	 */
 11132	
 11133	  unsigned left;
 11134	
 11135	  tmp_buf = buffer;
 11136	  sys_umap(SELF, D, (vir_bytes)buffer, (phys_bytes)sizeof(buffer), &tmp_phys);
 11137	
 11138	  if ((left = dma_bytes_left(tmp_phys)) < DMA_BUF_SIZE) {
 11139	        /* First half of buffer crosses a 64K boundary, can't DMA into that */

_________________________ Page 775    File: drivers/libdriver/driver.c ___________________
 11140	        tmp_buf += left;
 11141	        tmp_phys += left;
 11142	  }
 11143	}
	
 11145	/*===========================================================================*
 11146	 *                              do_rdwt                                      *
 11147	 *===========================================================================*/
 11148	PRIVATE int do_rdwt(dp, mp)
 11149	struct driver *dp;              /* device dependent entry points */
 11150	message *mp;                    /* pointer to read or write message */
 11151	{
 11152	/* Carry out a single read or write request. */
 11153	  iovec_t iovec1;
 11154	  int r, opcode;
 11155	  phys_bytes phys_addr;
 11156	
 11157	  /* Disk address?  Address and length of the user buffer? */
 11158	  if (mp->COUNT < 0) return(EINVAL);
 11159	
 11160	  /* Check the user buffer. */
 11161	  sys_umap(mp->PROC_NR, D, (vir_bytes) mp->ADDRESS, mp->COUNT, &phys_addr);
 11162	  if (phys_addr == 0) return(EFAULT);
 11163	
 11164	  /* Prepare for I/O. */
 11165	  if ((*dp->dr_prepare)(mp->DEVICE) == NIL_DEV) return(ENXIO);
 11166	
 11167	  /* Create a one element scatter/gather vector for the buffer. */
 11168	  opcode = mp->m_type == DEV_READ ? DEV_GATHER : DEV_SCATTER;
 11169	  iovec1.iov_addr = (vir_bytes) mp->ADDRESS;
 11170	  iovec1.iov_size = mp->COUNT;
 11171	
 11172	  /* Transfer bytes from/to the device. */
 11173	  r = (*dp->dr_transfer)(mp->PROC_NR, opcode, mp->POSITION, &iovec1, 1);
 11174	
 11175	  /* Return the number of bytes transferred or an error code. */
 11176	  return(r == OK ? (mp->COUNT - iovec1.iov_size) : r);
 11177	}
	
 11179	/*==========================================================================*
 11180	 *                              do_vrdwt                                    *
 11181	 *==========================================================================*/
 11182	PRIVATE int do_vrdwt(dp, mp)
 11183	struct driver *dp;      /* device dependent entry points */
 11184	message *mp;            /* pointer to read or write message */
 11185	{
 11186	/* Carry out an device read or write to/from a vector of user addresses.
 11187	 * The "user addresses" are assumed to be safe, i.e. FS transferring to/from
 11188	 * its own buffers, so they are not checked.
 11189	 */
 11190	  static iovec_t iovec[NR_IOREQS];
 11191	  iovec_t *iov;
 11192	  phys_bytes iovec_size;
 11193	  unsigned nr_req;
 11194	  int r;
 11195	
 11196	  nr_req = mp->COUNT;   /* Length of I/O vector */
 11197	
 11198	  if (mp->m_source < 0) {
 11199	    /* Called by a task, no need to copy vector. */

_________________________ Page 776    File: drivers/libdriver/driver.c ___________________
 11200	    iov = (iovec_t *) mp->ADDRESS;
 11201	  } else {
 11202	    /* Copy the vector from the caller to kernel space. */
 11203	    if (nr_req > NR_IOREQS) nr_req = NR_IOREQS;
 11204	    iovec_size = (phys_bytes) (nr_req * sizeof(iovec[0]));
 11205	
 11206	    if (OK != sys_datacopy(mp->m_source, (vir_bytes) mp->ADDRESS, 
 11207	                SELF, (vir_bytes) iovec, iovec_size))
 11208	        panic((*dp->dr_name)(),"bad I/O vector by", mp->m_source);
 11209	    iov = iovec;
 11210	  }
 11211	
 11212	  /* Prepare for I/O. */
 11213	  if ((*dp->dr_prepare)(mp->DEVICE) == NIL_DEV) return(ENXIO);
 11214	
 11215	  /* Transfer bytes from/to the device. */
 11216	  r = (*dp->dr_transfer)(mp->PROC_NR, mp->m_type, mp->POSITION, iov, nr_req);
 11217	
 11218	  /* Copy the I/O vector back to the caller. */
 11219	  if (mp->m_source >= 0) {
 11220	    sys_datacopy(SELF, (vir_bytes) iovec, 
 11221	        mp->m_source, (vir_bytes) mp->ADDRESS, iovec_size);
 11222	  }
 11223	  return(r);
 11224	}
	
 11226	/*===========================================================================*
 11227	 *                              no_name                                      *
 11228	 *===========================================================================*/
 11229	PUBLIC char *no_name()
 11230	{
 11231	/* Use this default name if there is no specific name for the device. This was
 11232	 * originally done by fetching the name from the task table for this process: 
 11233	 * "return(tasktab[proc_number(proc_ptr) + NR_TASKS].name);", but currently a
 11234	 * real "noname" is returned. Perhaps, some system information service can be
 11235	 * queried for a name at a later time.
 11236	 */
 11237	  static char name[] = "noname";
 11238	  return name;
 11239	}
	
 11241	/*============================================================================*
 11242	 *                              do_nop                                        *
 11243	 *============================================================================*/
 11244	PUBLIC int do_nop(dp, mp)
 11245	struct driver *dp;
 11246	message *mp;
 11247	{
 11248	/* Nothing there, or nothing to do. */
 11249	
 11250	  switch (mp->m_type) {
 11251	  case DEV_OPEN:        return(ENODEV);
 11252	  case DEV_CLOSE:       return(OK);
 11253	  case DEV_IOCTL:       return(ENOTTY);
 11254	  default:              return(EIO);
 11255	  }
 11256	}
	

_________________________ Page 777    File: drivers/libdriver/driver.c ___________________
 11258	/*============================================================================*
 11259	 *                              nop_signal                                    *
 11260	 *============================================================================*/
 11261	PUBLIC void nop_signal(dp, mp)
 11262	struct driver *dp;
 11263	message *mp;
 11264	{
 11265	/* Default action for signal is to ignore. */
 11266	}
	
 11268	/*============================================================================*
 11269	 *                              nop_alarm                                     *
 11270	 *============================================================================*/
 11271	PUBLIC void nop_alarm(dp, mp)
 11272	struct driver *dp;
 11273	message *mp;
 11274	{
 11275	/* Ignore the leftover alarm. */
 11276	}
	
 11278	/*===========================================================================*
 11279	 *                              nop_prepare                                  *
 11280	 *===========================================================================*/
 11281	PUBLIC struct device *nop_prepare(device)
 11282	{
 11283	/* Nothing to prepare for. */
 11284	  return(NIL_DEV);
 11285	}
	
 11287	/*===========================================================================*
 11288	 *                              nop_cleanup                                  *
 11289	 *===========================================================================*/
 11290	PUBLIC void nop_cleanup()
 11291	{
 11292	/* Nothing to clean up. */
 11293	}
	
 11295	/*===========================================================================*
 11296	 *                              nop_cancel                                   *
 11297	 *===========================================================================*/
 11298	PUBLIC int nop_cancel(struct driver *dr, message *m)
 11299	{
 11300	/* Nothing to do for cancel. */
 11301	   return(OK);
 11302	}
	
 11304	/*===========================================================================*
 11305	 *                              nop_select                                   *
 11306	 *===========================================================================*/
 11307	PUBLIC int nop_select(struct driver *dr, message *m)
 11308	{
 11309	/* Nothing to do for select. */
 11310	   return(OK);
 11311	}
	
 11313	/*============================================================================*
 11314	 *                              do_diocntl                                    *
 11315	 *============================================================================*/
 11316	PUBLIC int do_diocntl(dp, mp)
 11317	struct driver *dp;

_________________________ Page 778    File: drivers/libdriver/driver.c ___________________
 11318	message *mp;                    /* pointer to ioctl request */
 11319	{
 11320	/* Carry out a partition setting/getting request. */
 11321	  struct device *dv;
 11322	  struct partition entry;
 11323	  int s;
 11324	
 11325	  if (mp->REQUEST != DIOCSETP && mp->REQUEST != DIOCGETP) {
 11326	        if(dp->dr_other) {
 11327	                return dp->dr_other(dp, mp);
 11328	        } else return(ENOTTY);
 11329	  }
 11330	
 11331	  /* Decode the message parameters. */
 11332	  if ((dv = (*dp->dr_prepare)(mp->DEVICE)) == NIL_DEV) return(ENXIO);
 11333	
 11334	  if (mp->REQUEST == DIOCSETP) {
 11335	        /* Copy just this one partition table entry. */
 11336	        if (OK != (s=sys_datacopy(mp->PROC_NR, (vir_bytes) mp->ADDRESS,
 11337	                SELF, (vir_bytes) &entry, sizeof(entry))))
 11338	            return s;
 11339	        dv->dv_base = entry.base;
 11340	        dv->dv_size = entry.size;
 11341	  } else {
 11342	        /* Return a partition table entry and the geometry of the drive. */
 11343	        entry.base = dv->dv_base;
 11344	        entry.size = dv->dv_size;
 11345	        (*dp->dr_geometry)(&entry);
 11346	        if (OK != (s=sys_datacopy(SELF, (vir_bytes) &entry,
 11347	                mp->PROC_NR, (vir_bytes) mp->ADDRESS, sizeof(entry))))
 11348	            return s;
 11349	  }
 11350	  return(OK);
 11351	}


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/libdriver/drvlib.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 11400	/* IBM device driver utility functions.                 Author: Kees J. Bot
 11401	 *                                                              7 Dec 1995
 11402	 * Entry point:
 11403	 *   partition: partition a disk to the partition table(s) on it.
 11404	 */
 11405	
 11406	#include "driver.h"
 11407	#include "drvlib.h"
 11408	#include <unistd.h>
 11409	
 11410	/* Extended partition? */
 11411	#define ext_part(s)     ((s) == 0x05 || (s) == 0x0F)
 11412	
 11413	FORWARD _PROTOTYPE( void extpartition, (struct driver *dp, int extdev,
 11414	                                                unsigned long extbase) );
 11415	FORWARD _PROTOTYPE( int get_part_table, (struct driver *dp, int device,
 11416	                        unsigned long offset, struct part_entry *table));
 11417	FORWARD _PROTOTYPE( void sort, (struct part_entry *table) );
 11418	
 11419	#ifndef CD_SECTOR_SIZE

_________________________ Page 779    File: drivers/libdriver/drvlib.c ___________________
 11420	#define CD_SECTOR_SIZE 2048
 11421	#endif 
 11422	
 11423	/*============================================================================*
 11424	 *                              partition                                     *
 11425	 *============================================================================*/
 11426	PUBLIC void partition(dp, device, style, atapi)
 11427	struct driver *dp;      /* device dependent entry points */
 11428	int device;             /* device to partition */
 11429	int style;              /* partitioning style: floppy, primary, sub. */
 11430	int atapi;              /* atapi device */
 11431	{
 11432	/* This routine is called on first open to initialize the partition tables
 11433	 * of a device.  It makes sure that each partition falls safely within the
 11434	 * device's limits.  Depending on the partition style we are either making
 11435	 * floppy partitions, primary partitions or subpartitions.  Only primary
 11436	 * partitions are sorted, because they are shared with other operating
 11437	 * systems that expect this.
 11438	 */
 11439	  struct part_entry table[NR_PARTITIONS], *pe;
 11440	  int disk, par;
 11441	  struct device *dv;
 11442	  unsigned long base, limit, part_limit;
 11443	
 11444	  /* Get the geometry of the device to partition */
 11445	  if ((dv = (*dp->dr_prepare)(device)) == NIL_DEV
 11446	                                || cmp64u(dv->dv_size, 0) == 0) return;
 11447	  base = div64u(dv->dv_base, SECTOR_SIZE);
 11448	  limit = base + div64u(dv->dv_size, SECTOR_SIZE);
 11449	
 11450	  /* Read the partition table for the device. */
 11451	  if(!get_part_table(dp, device, 0L, table)) {
 11452	          return;
 11453	  }
 11454	
 11455	  /* Compute the device number of the first partition. */
 11456	  switch (style) {
 11457	  case P_FLOPPY:
 11458	        device += MINOR_fd0p0;
 11459	        break;
 11460	  case P_PRIMARY:
 11461	        sort(table);            /* sort a primary partition table */
 11462	        device += 1;
 11463	        break;
 11464	  case P_SUB:
 11465	        disk = device / DEV_PER_DRIVE;
 11466	        par = device % DEV_PER_DRIVE - 1;
 11467	        device = MINOR_d0p0s0 + (disk * NR_PARTITIONS + par) * NR_PARTITIONS;
 11468	  }
 11469	
 11470	  /* Find an array of devices. */
 11471	  if ((dv = (*dp->dr_prepare)(device)) == NIL_DEV) return;
 11472	
 11473	  /* Set the geometry of the partitions from the partition table. */
 11474	  for (par = 0; par < NR_PARTITIONS; par++, dv++) {
 11475	        /* Shrink the partition to fit within the device. */
 11476	        pe = &table[par];
 11477	        part_limit = pe->lowsec + pe->size;
 11478	        if (part_limit < pe->lowsec) part_limit = limit;
 11479	        if (part_limit > limit) part_limit = limit;

_________________________ Page 780    File: drivers/libdriver/drvlib.c ___________________
 11480	        if (pe->lowsec < base) pe->lowsec = base;
 11481	        if (part_limit < pe->lowsec) part_limit = pe->lowsec;
 11482	
 11483	        dv->dv_base = mul64u(pe->lowsec, SECTOR_SIZE);
 11484	        dv->dv_size = mul64u(part_limit - pe->lowsec, SECTOR_SIZE);
 11485	
 11486	        if (style == P_PRIMARY) {
 11487	                /* Each Minix primary partition can be subpartitioned. */
 11488	                if (pe->sysind == MINIX_PART)
 11489	                        partition(dp, device + par, P_SUB, atapi);
 11490	
 11491	                /* An extended partition has logical partitions. */
 11492	                if (ext_part(pe->sysind))
 11493	                        extpartition(dp, device + par, pe->lowsec);
 11494	        }
 11495	  }
 11496	}
	
 11498	/*============================================================================*
 11499	 *                              extpartition                                  *
 11500	 *============================================================================*/
 11501	PRIVATE void extpartition(dp, extdev, extbase)
 11502	struct driver *dp;      /* device dependent entry points */
 11503	int extdev;             /* extended partition to scan */
 11504	unsigned long extbase;  /* sector offset of the base extended partition */
 11505	{
 11506	/* Extended partitions cannot be ignored alas, because people like to move
 11507	 * files to and from DOS partitions.  Avoid reading this code, it's no fun.
 11508	 */
 11509	  struct part_entry table[NR_PARTITIONS], *pe;
 11510	  int subdev, disk, par;
 11511	  struct device *dv;
 11512	  unsigned long offset, nextoffset;
 11513	
 11514	  disk = extdev / DEV_PER_DRIVE;
 11515	  par = extdev % DEV_PER_DRIVE - 1;
 11516	  subdev = MINOR_d0p0s0 + (disk * NR_PARTITIONS + par) * NR_PARTITIONS;
 11517	
 11518	  offset = 0;
 11519	  do {
 11520	        if (!get_part_table(dp, extdev, offset, table)) return;
 11521	        sort(table);
 11522	
 11523	        /* The table should contain one logical partition and optionally
 11524	         * another extended partition.  (It's a linked list.)
 11525	         */
 11526	        nextoffset = 0;
 11527	        for (par = 0; par < NR_PARTITIONS; par++) {
 11528	                pe = &table[par];
 11529	                if (ext_part(pe->sysind)) {
 11530	                        nextoffset = pe->lowsec;
 11531	                } else
 11532	                if (pe->sysind != NO_PART) {
 11533	                        if ((dv = (*dp->dr_prepare)(subdev)) == NIL_DEV) return;
 11534	
 11535	                        dv->dv_base = mul64u(extbase + offset + pe->lowsec,
 11536	                                                                SECTOR_SIZE);
 11537	                        dv->dv_size = mul64u(pe->size, SECTOR_SIZE);
 11538	
 11539	                        /* Out of devices? */

_________________________ Page 781    File: drivers/libdriver/drvlib.c ___________________
 11540	                        if (++subdev % NR_PARTITIONS == 0) return;
 11541	                }
 11542	        }
 11543	  } while ((offset = nextoffset) != 0);
 11544	}
	
 11546	/*============================================================================*
 11547	 *                              get_part_table                                *
 11548	 *============================================================================*/
 11549	PRIVATE int get_part_table(dp, device, offset, table)
 11550	struct driver *dp;
 11551	int device;
 11552	unsigned long offset;           /* sector offset to the table */
 11553	struct part_entry *table;       /* four entries */
 11554	{
 11555	/* Read the partition table for the device, return true iff there were no
 11556	 * errors.
 11557	 */
 11558	  iovec_t iovec1;
 11559	  off_t position;
 11560	  static unsigned char partbuf[CD_SECTOR_SIZE];
 11561	
 11562	  position = offset << SECTOR_SHIFT;
 11563	  iovec1.iov_addr = (vir_bytes) partbuf;
 11564	  iovec1.iov_size = CD_SECTOR_SIZE;
 11565	  if ((*dp->dr_prepare)(device) != NIL_DEV) {
 11566	        (void) (*dp->dr_transfer)(SELF, DEV_GATHER, position, &iovec1, 1);
 11567	  }
 11568	  if (iovec1.iov_size != 0) {
 11569	        return 0;
 11570	  }
 11571	  if (partbuf[510] != 0x55 || partbuf[511] != 0xAA) {
 11572	        /* Invalid partition table. */
 11573	        return 0;
 11574	  }
 11575	  memcpy(table, (partbuf + PART_TABLE_OFF), NR_PARTITIONS * sizeof(table[0]));
 11576	  return 1;
 11577	}
	
 11579	/*===========================================================================*
 11580	 *                              sort                                         *
 11581	 *===========================================================================*/
 11582	PRIVATE void sort(table)
 11583	struct part_entry *table;
 11584	{
 11585	/* Sort a partition table. */
 11586	  struct part_entry *pe, tmp;
 11587	  int n = NR_PARTITIONS;
 11588	
 11589	  do {
 11590	        for (pe = table; pe < table + NR_PARTITIONS-1; pe++) {
 11591	                if (pe[0].sysind == NO_PART
 11592	                        || (pe[0].lowsec > pe[1].lowsec
 11593	                                        && pe[1].sysind != NO_PART)) {
 11594	                        tmp = pe[0]; pe[0] = pe[1]; pe[1] = tmp;
 11595	                }
 11596	        }
 11597	  } while (--n > 0);
 11598	}

_________________________ Page 782    File: drivers/libdriver/drvlib.c ___________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/memory/memory.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 11600	/* This file contains the device dependent part of the drivers for the
 11601	 * following special files:
 11602	 *     /dev/ram         - RAM disk 
 11603	 *     /dev/mem         - absolute memory
 11604	 *     /dev/kmem        - kernel virtual memory
 11605	 *     /dev/null        - null device (data sink)
 11606	 *     /dev/boot        - boot device loaded from boot image 
 11607	 *     /dev/zero        - null byte stream generator
 11608	 *
 11609	 *  Changes:
 11610	 *      Apr 29, 2005    added null byte generator  (Jorrit N. Herder)
 11611	 *      Apr 09, 2005    added support for boot device  (Jorrit N. Herder)
 11612	 *      Jul 26, 2004    moved RAM driver to user-space  (Jorrit N. Herder)
 11613	 *      Apr 20, 1992    device dependent/independent split  (Kees J. Bot)
 11614	 */
 11615	
 11616	#include "../drivers.h"
 11617	#include "../libdriver/driver.h"
 11618	#include <sys/ioc_memory.h>
 11619	#include "../../kernel/const.h"
 11620	#include "../../kernel/config.h"
 11621	#include "../../kernel/type.h"
 11622	
 11623	#include "assert.h"
 11624	
 11625	#define NR_DEVS            6            /* number of minor devices */
 11626	
 11627	PRIVATE struct device m_geom[NR_DEVS];  /* base and size of each device */
 11628	PRIVATE int m_seg[NR_DEVS];             /* segment index of each device */
 11629	PRIVATE int m_device;                   /* current device */
 11630	PRIVATE struct kinfo kinfo;             /* kernel information */ 
 11631	PRIVATE struct machine machine;         /* machine information */ 
 11632	
 11633	extern int errno;                       /* error number for PM calls */
 11634	
 11635	FORWARD _PROTOTYPE( char *m_name, (void)                                );
 11636	FORWARD _PROTOTYPE( struct device *m_prepare, (int device)              );
 11637	FORWARD _PROTOTYPE( int m_transfer, (int proc_nr, int opcode, off_t position,
 11638	                                        iovec_t *iov, unsigned nr_req)  );
 11639	FORWARD _PROTOTYPE( int m_do_open, (struct driver *dp, message *m_ptr)  );
 11640	FORWARD _PROTOTYPE( void m_init, (void) );
 11641	FORWARD _PROTOTYPE( int m_ioctl, (struct driver *dp, message *m_ptr)    );
 11642	FORWARD _PROTOTYPE( void m_geometry, (struct partition *entry)          );
 11643	
 11644	/* Entry points to this driver. */
 11645	PRIVATE struct driver m_dtab = {
 11646	  m_name,       /* current device's name */
 11647	  m_do_open,    /* open or mount */
 11648	  do_nop,       /* nothing on a close */
 11649	  m_ioctl,      /* specify ram disk geometry */
 11650	  m_prepare,    /* prepare for I/O on a given minor device */
 11651	  m_transfer,   /* do the I/O */
 11652	  nop_cleanup,  /* no need to clean up */
 11653	  m_geometry,   /* memory device "geometry" */
 11654	  nop_signal,   /* system signals */

_________________________ Page 783    File: drivers/memory/memory.c ______________________
 11655	  nop_alarm,
 11656	  nop_cancel,
 11657	  nop_select,
 11658	  NULL,
 11659	  NULL
 11660	};
 11661	
 11662	/* Buffer for the /dev/zero null byte feed. */
 11663	#define ZERO_BUF_SIZE                   1024
 11664	PRIVATE char dev_zero[ZERO_BUF_SIZE];
 11665	
 11666	#define click_to_round_k(n) \
 11667	        ((unsigned) ((((unsigned long) (n) << CLICK_SHIFT) + 512) / 1024))
 11668	
 11669	/*===========================================================================*
 11670	 *                                 main                                      *
 11671	 *===========================================================================*/
 11672	PUBLIC int main(void)
 11673	{
 11674	/* Main program. Initialize the memory driver and start the main loop. */
 11675	  m_init();                     
 11676	  driver_task(&m_dtab);         
 11677	  return(OK);                           
 11678	}
	
 11680	/*===========================================================================*
 11681	 *                               m_name                                      *
 11682	 *===========================================================================*/
 11683	PRIVATE char *m_name()
 11684	{
 11685	/* Return a name for the current device. */
 11686	  static char name[] = "memory";
 11687	  return name;  
 11688	}
	
 11690	/*===========================================================================*
 11691	 *                              m_prepare                                    *
 11692	 *===========================================================================*/
 11693	PRIVATE struct device *m_prepare(device)
 11694	int device;
 11695	{
 11696	/* Prepare for I/O on a device: check if the minor device number is ok. */
 11697	  if (device < 0 || device >= NR_DEVS) return(NIL_DEV);
 11698	  m_device = device;
 11699	
 11700	  return(&m_geom[device]);
 11701	}
	
 11703	/*===========================================================================*
 11704	 *                              m_transfer                                   *
 11705	 *===========================================================================*/
 11706	PRIVATE int m_transfer(proc_nr, opcode, position, iov, nr_req)
 11707	int proc_nr;                    /* process doing the request */
 11708	int opcode;                     /* DEV_GATHER or DEV_SCATTER */
 11709	off_t position;                 /* offset on device to read or write */
 11710	iovec_t *iov;                   /* pointer to read or write request vector */
 11711	unsigned nr_req;                /* length of request vector */
 11712	{
 11713	/* Read or write one the driver's minor devices. */
 11714	  phys_bytes mem_phys;

_________________________ Page 784    File: drivers/memory/memory.c ______________________
 11715	  int seg;
 11716	  unsigned count, left, chunk;
 11717	  vir_bytes user_vir;
 11718	  struct device *dv;
 11719	  unsigned long dv_size;
 11720	  int s;
 11721	
 11722	  /* Get minor device number and check for /dev/null. */
 11723	  dv = &m_geom[m_device];
 11724	  dv_size = cv64ul(dv->dv_size);
 11725	
 11726	  while (nr_req > 0) {
 11727	
 11728	        /* How much to transfer and where to / from. */
 11729	        count = iov->iov_size;
 11730	        user_vir = iov->iov_addr;
 11731	
 11732	        switch (m_device) {
 11733	
 11734	        /* No copying; ignore request. */
 11735	        case NULL_DEV:
 11736	            if (opcode == DEV_GATHER) return(OK);       /* always at EOF */
 11737	            break;
 11738	
 11739	        /* Virtual copying. For RAM disk, kernel memory and boot device. */
 11740	        case RAM_DEV:
 11741	        case KMEM_DEV:
 11742	        case BOOT_DEV:
 11743	            if (position >= dv_size) return(OK);        /* check for EOF */
 11744	            if (position + count > dv_size) count = dv_size - position;
 11745	            seg = m_seg[m_device];
 11746	
 11747	            if (opcode == DEV_GATHER) {                 /* copy actual data */
 11748	                sys_vircopy(SELF,seg,position, proc_nr,D,user_vir, count);
 11749	            } else {
 11750	                sys_vircopy(proc_nr,D,user_vir, SELF,seg,position, count);
 11751	            }
 11752	            break;
 11753	
 11754	        /* Physical copying. Only used to access entire memory. */
 11755	        case MEM_DEV:
 11756	            if (position >= dv_size) return(OK);        /* check for EOF */
 11757	            if (position + count > dv_size) count = dv_size - position;
 11758	            mem_phys = cv64ul(dv->dv_base) + position;
 11759	
 11760	            if (opcode == DEV_GATHER) {                 /* copy data */
 11761	                sys_physcopy(NONE, PHYS_SEG, mem_phys, 
 11762	                        proc_nr, D, user_vir, count);
 11763	            } else {
 11764	                sys_physcopy(proc_nr, D, user_vir, 
 11765	                        NONE, PHYS_SEG, mem_phys, count);
 11766	            }
 11767	            break;
 11768	
 11769	        /* Null byte stream generator. */
 11770	        case ZERO_DEV:
 11771	            if (opcode == DEV_GATHER) {
 11772	                left = count;
 11773	                while (left > 0) {
 11774	                    chunk = (left > ZERO_BUF_SIZE) ? ZERO_BUF_SIZE : left;

_________________________ Page 785    File: drivers/memory/memory.c ______________________
 11775	                    if (OK != (s=sys_vircopy(SELF, D, (vir_bytes) dev_zero, 
 11776	                            proc_nr, D, user_vir, chunk)))
 11777	                        report("MEM","sys_vircopy failed", s);
 11778	                    left -= chunk;
 11779	                    user_vir += chunk;
 11780	                }
 11781	            }
 11782	            break;
 11783	
 11784	        /* Unknown (illegal) minor device. */
 11785	        default:
 11786	            return(EINVAL);
 11787	        }
 11788	
 11789	        /* Book the number of bytes transferred. */
 11790	        position += count;
 11791	        iov->iov_addr += count;
 11792	        if ((iov->iov_size -= count) == 0) { iov++; nr_req--; }
 11793	
 11794	  }
 11795	  return(OK);
 11796	}
	
 11798	/*===========================================================================*
 11799	 *                              m_do_open                                    *
 11800	 *===========================================================================*/
 11801	PRIVATE int m_do_open(dp, m_ptr)
 11802	struct driver *dp;
 11803	message *m_ptr;
 11804	{
 11805	/* Check device number on open.  (This used to give I/O privileges to a 
 11806	 * process opening /dev/mem or /dev/kmem. This may be needed in case of 
 11807	 * memory mapped I/O. With system calls to do I/O this is no longer needed.)
 11808	 */
 11809	  if (m_prepare(m_ptr->DEVICE) == NIL_DEV) return(ENXIO);
 11810	
 11811	  return(OK);
 11812	}
	
 11814	/*===========================================================================*
 11815	 *                              m_init                                       *
 11816	 *===========================================================================*/
 11817	PRIVATE void m_init()
 11818	{
 11819	  /* Initialize this task. All minor devices are initialized one by one. */
 11820	  int i, s;
 11821	
 11822	  if (OK != (s=sys_getkinfo(&kinfo))) {
 11823	      panic("MEM","Couldn't get kernel information.",s);
 11824	  }
 11825	
 11826	  /* Install remote segment for /dev/kmem memory. */
 11827	  m_geom[KMEM_DEV].dv_base = cvul64(kinfo.kmem_base);
 11828	  m_geom[KMEM_DEV].dv_size = cvul64(kinfo.kmem_size);
 11829	  if (OK != (s=sys_segctl(&m_seg[KMEM_DEV], (u16_t *) &s, (vir_bytes *) & ...
 11830	                kinfo.kmem_base, kinfo.kmem_size))) {
 11831	      panic("MEM","Couldn't install remote segment.",s);
 11832	  }
 11833	
 11834	  /* Install remote segment for /dev/boot memory, if enabled. */

_________________________ Page 786    File: drivers/memory/memory.c ______________________
 11835	  m_geom[BOOT_DEV].dv_base = cvul64(kinfo.bootdev_base);
 11836	  m_geom[BOOT_DEV].dv_size = cvul64(kinfo.bootdev_size);
 11837	  if (kinfo.bootdev_base > 0) {
 11838	      if (OK != (s=sys_segctl(&m_seg[BOOT_DEV], (u16_t *) &s, (vir_bytes *)
 11839	              kinfo.bootdev_base, kinfo.bootdev_size))) {
 11840	          panic("MEM","Couldn't install remote segment.",s);
 11841	      }
 11842	  }
 11843	
 11844	  /* Initialize /dev/zero. Simply write zeros into the buffer. */
 11845	  for (i=0; i<ZERO_BUF_SIZE; i++) {
 11846	       dev_zero[i] = '\0';
 11847	  }
 11848	
 11849	  /* Set up memory ranges for /dev/mem. */
 11850	  if (OK != (s=sys_getmachine(&machine))) {
 11851	      panic("MEM","Couldn't get machine information.",s);
 11852	  }
 11853	  if (! machine.protected) {
 11854	        m_geom[MEM_DEV].dv_size =   cvul64(0x100000); /* 1M for 8086 systems */
 11855	  } else {
 11856	        m_geom[MEM_DEV].dv_size = cvul64(0xFFFFFFFF); /* 4G-1 for 386 systems */
 11857	  }
 11858	}
	
 11860	/*===========================================================================*
 11861	 *                              m_ioctl                                      *
 11862	 *===========================================================================*/
 11863	PRIVATE int m_ioctl(dp, m_ptr)
 11864	struct driver *dp;                      /* pointer to driver structure */
 11865	message *m_ptr;                         /* pointer to control message */
 11866	{
 11867	/* I/O controls for the memory driver. Currently there is one I/O control:
 11868	 * - MIOCRAMSIZE: to set the size of the RAM disk.
 11869	 */
 11870	  struct device *dv;
 11871	  if ((dv = m_prepare(m_ptr->DEVICE)) == NIL_DEV) return(ENXIO);
 11872	
 11873	  switch (m_ptr->REQUEST) {
 11874	    case MIOCRAMSIZE: {
 11875	        /* FS wants to create a new RAM disk with the given size. */
 11876	        phys_bytes ramdev_size;
 11877	        phys_bytes ramdev_base;
 11878	        int s;
 11879	
 11880	        if (m_ptr->PROC_NR != FS_PROC_NR) {
 11881	            report("MEM", "warning, MIOCRAMSIZE called by", m_ptr->PROC_NR);
 11882	            return(EPERM);
 11883	        }
 11884	
 11885	        /* Try to allocate a piece of memory for the RAM disk. */
 11886	        ramdev_size = m_ptr->POSITION;
 11887	        if (allocmem(ramdev_size, &ramdev_base) < 0) {
 11888	            report("MEM", "warning, allocmem failed", errno);
 11889	            return(ENOMEM);
 11890	        }
 11891	        dv->dv_base = cvul64(ramdev_base);
 11892	        dv->dv_size = cvul64(ramdev_size);
 11893	
 11894	        if (OK != (s=sys_segctl(&m_seg[RAM_DEV], (u16_t *) &s, (vir_bytes *

_________________________ Page 787    File: drivers/memory/memory.c ______________________
 11895	                ramdev_base, ramdev_size))) {
 11896	                panic("MEM","Couldn't install remote segment.",s);
 11897	        }
 11898	        break;
 11899	    }
 11900	
 11901	    default:
 11902	        return(do_diocntl(&m_dtab, m_ptr));
 11903	  }
 11904	  return(OK);
 11905	}
	
 11907	/*===========================================================================*
 11908	 *                              m_geometry                                   *
 11909	 *===========================================================================*/
 11910	PRIVATE void m_geometry(entry)
 11911	struct partition *entry;
 11912	{
 11913	  /* Memory devices don't have a geometry, but the outside world insists. */
 11914	  entry->cylinders = div64u(m_geom[m_device].dv_size, SECTOR_SIZE) / (64 * 32);
 11915	  entry->heads = 64;
 11916	  entry->sectors = 32;
 11917	}



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/at_wini/at_wini.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 12000	#include "../drivers.h"
 12001	#include "../libdriver/driver.h"
 12002	#include "../libdriver/drvlib.h"
 12003	
 12004	_PROTOTYPE(int main, (void));
 12005	
 12006	#define VERBOSE            0    /* display identify messages during boot */
 12007	#define ENABLE_ATAPI    0       /* add ATAPI cd-rom support to driver */



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/at_wini/at_wini.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 12100	/* This file contains the device dependent part of a driver for the IBM-AT
 12101	 * winchester controller.  Written by Adri Koppes.
 12102	 *
 12103	 * The file contains one entry point:
 12104	 *
 12105	 *   at_winchester_task:        main entry when system is brought up
 12106	 *
 12107	 * Changes:
 12108	 *   Aug 19, 2005   ata pci support, supports SATA  (Ben Gras)
 12109	 *   Nov 18, 2004   moved AT disk driver to user-space  (Jorrit N. Herder)
 12110	 *   Aug 20, 2004   watchdogs replaced by sync alarms  (Jorrit N. Herder)
 12111	 *   Mar 23, 2000   added ATAPI CDROM support  (Michael Temari)
 12112	 *   May 14, 2000   d-d/i rewrite  (Kees J. Bot)
 12113	 *   Apr 13, 1992   device dependent/independent split  (Kees J. Bot)
 12114	 */

_________________________ Page 788    File: drivers/at_wini/at_wini.c ____________________
 12115	
 12116	#include "at_wini.h"
 12117	#include "../libpci/pci.h"
 12118	
 12119	#include <minix/sysutil.h>
 12120	#include <minix/keymap.h>
 12121	#include <sys/ioc_disk.h>
 12122	
 12123	#define ATAPI_DEBUG         0   /* To debug ATAPI code. */
 12124	
 12125	/* I/O Ports used by winchester disk controllers. */
 12126	
 12127	/* Read and write registers */
 12128	#define REG_CMD_BASE0   0x1F0   /* command base register of controller 0 */
 12129	#define REG_CMD_BASE1   0x170   /* command base register of controller 1 */
 12130	#define REG_CTL_BASE0   0x3F6   /* control base register of controller 0 */
 12131	#define REG_CTL_BASE1   0x376   /* control base register of controller 1 */
 12132	
 12133	#define REG_DATA            0   /* data register (offset from the base reg.) */
 12134	#define REG_PRECOMP         1   /* start of write precompensation */
 12135	#define REG_COUNT           2   /* sectors to transfer */
 12136	#define REG_SECTOR          3   /* sector number */
 12137	#define REG_CYL_LO          4   /* low byte of cylinder number */
 12138	#define REG_CYL_HI          5   /* high byte of cylinder number */
 12139	#define REG_LDH             6   /* lba, drive and head */
 12140	#define   LDH_DEFAULT           0xA0    /* ECC enable, 512 bytes per sector */
 12141	#define   LDH_LBA               0x40    /* Use LBA addressing */
 12142	#define   ldh_init(drive)       (LDH_DEFAULT | ((drive) << 4))
 12143	
 12144	/* Read only registers */
 12145	#define REG_STATUS          7   /* status */
 12146	#define   STATUS_BSY            0x80    /* controller busy */
 12147	#define   STATUS_RDY            0x40    /* drive ready */
 12148	#define   STATUS_WF             0x20    /* write fault */
 12149	#define   STATUS_SC             0x10    /* seek complete (obsolete) */
 12150	#define   STATUS_DRQ            0x08    /* data transfer request */
 12151	#define   STATUS_CRD            0x04    /* corrected data */
 12152	#define   STATUS_IDX            0x02    /* index pulse */
 12153	#define   STATUS_ERR            0x01    /* error */
 12154	#define   STATUS_ADMBSY        0x100    /* administratively busy (software) */
 12155	#define REG_ERROR           1   /* error code */
 12156	#define   ERROR_BB              0x80    /* bad block */
 12157	#define   ERROR_ECC             0x40    /* bad ecc bytes */
 12158	#define   ERROR_ID              0x10    /* id not found */
 12159	#define   ERROR_AC              0x04    /* aborted command */
 12160	#define   ERROR_TK              0x02    /* track zero error */
 12161	#define   ERROR_DM              0x01    /* no data address mark */
 12162	
 12163	/* Write only registers */
 12164	#define REG_COMMAND         7   /* command */
 12165	#define   CMD_IDLE              0x00    /* for w_command: drive idle */
 12166	#define   CMD_RECALIBRATE       0x10    /* recalibrate drive */
 12167	#define   CMD_READ              0x20    /* read data */
 12168	#define   CMD_READ_EXT          0x24    /* read data (LBA48 addressed) */
 12169	#define   CMD_WRITE             0x30    /* write data */
 12170	#define   CMD_WRITE_EXT         0x34    /* write data (LBA48 addressed) */
 12171	#define   CMD_READVERIFY        0x40    /* read verify */
 12172	#define   CMD_FORMAT            0x50    /* format track */
 12173	#define   CMD_SEEK              0x70    /* seek cylinder */
 12174	#define   CMD_DIAG              0x90    /* execute device diagnostics */

_________________________ Page 789    File: drivers/at_wini/at_wini.c ____________________
 12175	#define   CMD_SPECIFY           0x91    /* specify parameters */
 12176	#define   ATA_IDENTIFY          0xEC    /* identify drive */
 12177	/* #define REG_CTL              0x206   */ /* control register */
 12178	#define REG_CTL         0       /* control register */
 12179	#define   CTL_NORETRY           0x80    /* disable access retry */
 12180	#define   CTL_NOECC             0x40    /* disable ecc retry */
 12181	#define   CTL_EIGHTHEADS        0x08    /* more than eight heads */
 12182	#define   CTL_RESET             0x04    /* reset controller */
 12183	#define   CTL_INTDISABLE        0x02    /* disable interrupts */
 12184	
 12185	#define REG_STATUS          7   /* status */
 12186	#define   STATUS_BSY            0x80    /* controller busy */
 12187	#define   STATUS_DRDY           0x40    /* drive ready */
 12188	#define   STATUS_DMADF          0x20    /* dma ready/drive fault */
 12189	#define   STATUS_SRVCDSC        0x10    /* service or dsc */
 12190	#define   STATUS_DRQ            0x08    /* data transfer request */
 12191	#define   STATUS_CORR           0x04    /* correctable error occurred */
 12192	#define   STATUS_CHECK          0x01    /* check error */
 12193	
 12194	/* Interrupt request lines. */
 12195	#define NO_IRQ           0      /* no IRQ set yet */
 12196	
 12197	#define ATAPI_PACKETSIZE        12
 12198	#define SENSE_PACKETSIZE        18
 12199	
 12200	/* Common command block */
 12201	struct command {
 12202	  u8_t  precomp;        /* REG_PRECOMP, etc. */
 12203	  u8_t  count;
 12204	  u8_t  sector;
 12205	  u8_t  cyl_lo;
 12206	  u8_t  cyl_hi;
 12207	  u8_t  ldh;
 12208	  u8_t  command;
 12209	};
 12210	
 12211	/* Error codes */
 12212	#define ERR              (-1)   /* general error */
 12213	#define ERR_BAD_SECTOR   (-2)   /* block marked bad detected */
 12214	
 12215	/* Some controllers don't interrupt, the clock will wake us up. */
 12216	#define WAKEUP          (32*HZ) /* drive may be out for 31 seconds max */
 12217	
 12218	/* Miscellaneous. */
 12219	#define MAX_DRIVES         8
 12220	#define COMPAT_DRIVES      4
 12221	#define MAX_SECS         256    /* controller can transfer this many sectors */
 12222	#define MAX_ERRORS         4    /* how often to try rd/wt before quitting */
 12223	#define NR_MINORS       (MAX_DRIVES * DEV_PER_DRIVE)
 12224	#define SUB_PER_DRIVE   (NR_PARTITIONS * NR_PARTITIONS)
 12225	#define NR_SUBDEVS      (MAX_DRIVES * SUB_PER_DRIVE)
 12226	#define DELAY_USECS     1000    /* controller timeout in microseconds */
 12227	#define DELAY_TICKS        1    /* controller timeout in ticks */
 12228	#define DEF_TIMEOUT_TICKS       300     /* controller timeout in ticks */
 12229	#define RECOVERY_USECS 500000   /* controller recovery time in microseconds */
 12230	#define RECOVERY_TICKS    30    /* controller recovery time in ticks */
 12231	#define INITIALIZED     0x01    /* drive is initialized */
 12232	#define DEAF            0x02    /* controller must be reset */
 12233	#define SMART           0x04    /* drive supports ATA commands */
 12234	#define ATAPI              0    /* don't bother with ATAPI; optimise out */

_________________________ Page 790    File: drivers/at_wini/at_wini.c ____________________
 12235	#define IDENTIFIED      0x10    /* w_identify done successfully */
 12236	#define IGNORING        0x20    /* w_identify failed once */
 12237	
 12238	/* Timeouts and max retries. */
 12239	int timeout_ticks = DEF_TIMEOUT_TICKS, max_errors = MAX_ERRORS;
 12240	int wakeup_ticks = WAKEUP;
 12241	long w_standard_timeouts = 0, w_pci_debug = 0, w_instance = 0,
 12242	 w_lba48 = 0, atapi_debug = 0;
 12243	
 12244	int w_testing = 0, w_silent = 0;
 12245	
 12246	int w_next_drive = 0;
 12247	
 12248	/* Variables. */
 12249	
 12250	/* wini is indexed by controller first, then drive (0-3).
 12251	 * controller 0 is always the 'compatability' ide controller, at
 12252	 * the fixed locations, whether present or not.
 12253	 */
 12254	PRIVATE struct wini {           /* main drive struct, one entry per drive */
 12255	  unsigned state;               /* drive state: deaf, initialized, dead */
 12256	  unsigned w_status;            /* device status register */
 12257	  unsigned base_cmd;            /* command base register */
 12258	  unsigned base_ctl;            /* control base register */
 12259	  unsigned irq;                 /* interrupt request line */
 12260	  unsigned irq_mask;            /* 1 << irq */
 12261	  unsigned irq_need_ack;        /* irq needs to be acknowledged */
 12262	  int irq_hook_id;              /* id of irq hook at the kernel */
 12263	  int lba48;                    /* supports lba48 */
 12264	  unsigned lcylinders;          /* logical number of cylinders (BIOS) */
 12265	  unsigned lheads;              /* logical number of heads */
 12266	  unsigned lsectors;            /* logical number of sectors per track */
 12267	  unsigned pcylinders;          /* physical number of cylinders (translated) */
 12268	  unsigned pheads;              /* physical number of heads */
 12269	  unsigned psectors;            /* physical number of sectors per track */
 12270	  unsigned ldhpref;             /* top four bytes of the LDH (head) register */
 12271	  unsigned precomp;             /* write precompensation cylinder / 4 */
 12272	  unsigned max_count;           /* max request for this drive */
 12273	  unsigned open_ct;             /* in-use count */
 12274	  struct device part[DEV_PER_DRIVE];    /* disks and partitions */
 12275	  struct device subpart[SUB_PER_DRIVE]; /* subpartitions */
 12276	} wini[MAX_DRIVES], *w_wn;
 12277	
 12278	PRIVATE int w_device = -1;
 12279	PRIVATE int w_controller = -1;
 12280	PRIVATE int w_major = -1;
 12281	PRIVATE char w_id_string[40];
 12282	
 12283	PRIVATE int win_tasknr;                 /* my task number */
 12284	PRIVATE int w_command;                  /* current command in execution */
 12285	PRIVATE u8_t w_byteval;                 /* used for SYS_IRQCTL */
 12286	PRIVATE int w_drive;                    /* selected drive */
 12287	PRIVATE int w_controller;               /* selected controller */
 12288	PRIVATE struct device *w_dv;            /* device's base and size */
 12289	
 12290	FORWARD _PROTOTYPE( void init_params, (void)                            );
 12291	FORWARD _PROTOTYPE( void init_drive, (struct wini *, int, int, int, int, int, int))
 12292	FORWARD _PROTOTYPE( void init_params_pci, (int)                         );
 12293	FORWARD _PROTOTYPE( int w_do_open, (struct driver *dp, message *m_ptr)  );
 12294	FORWARD _PROTOTYPE( struct device *w_prepare, (int dev)                 );

_________________________ Page 791    File: drivers/at_wini/at_wini.c ____________________
 12295	FORWARD _PROTOTYPE( int w_identify, (void)                              );
 12296	FORWARD _PROTOTYPE( char *w_name, (void)                                );
 12297	FORWARD _PROTOTYPE( int w_specify, (void)                               );
 12298	FORWARD _PROTOTYPE( int w_io_test, (void)                               );
 12299	FORWARD _PROTOTYPE( int w_transfer, (int proc_nr, int opcode, off_t position,
 12300	                                        iovec_t *iov, unsigned nr_req)  );
 12301	FORWARD _PROTOTYPE( int com_out, (struct command *cmd)                  );
 12302	FORWARD _PROTOTYPE( void w_need_reset, (void)                           );
 12303	FORWARD _PROTOTYPE( void ack_irqs, (unsigned int)                       );
 12304	FORWARD _PROTOTYPE( int w_do_close, (struct driver *dp, message *m_ptr) );
 12305	FORWARD _PROTOTYPE( int w_other, (struct driver *dp, message *m_ptr)    );
 12306	FORWARD _PROTOTYPE( int w_hw_int, (struct driver *dp, message *m_ptr)   );
 12307	FORWARD _PROTOTYPE( int com_simple, (struct command *cmd)               );
 12308	FORWARD _PROTOTYPE( void w_timeout, (void)                              );
 12309	FORWARD _PROTOTYPE( int w_reset, (void)                                 );
 12310	FORWARD _PROTOTYPE( void w_intr_wait, (void)                            );
 12311	FORWARD _PROTOTYPE( int at_intr_wait, (void)                            );
 12312	FORWARD _PROTOTYPE( int w_waitfor, (int mask, int value)                );
 12313	FORWARD _PROTOTYPE( void w_geometry, (struct partition *entry)          );
 12314	
 12315	/* Entry points to this driver. */
 12316	PRIVATE struct driver w_dtab = {
 12317	  w_name,               /* current device's name */
 12318	  w_do_open,            /* open or mount request, initialize device */
 12319	  w_do_close,           /* release device */
 12320	  do_diocntl,           /* get or set a partition's geometry */
 12321	  w_prepare,            /* prepare for I/O on a given minor device */
 12322	  w_transfer,           /* do the I/O */
 12323	  nop_cleanup,          /* nothing to clean up */
 12324	  w_geometry,           /* tell the geometry of the disk */
 12325	  nop_signal,           /* no cleanup needed on shutdown */
 12326	  nop_alarm,            /* ignore leftover alarms */
 12327	  nop_cancel,           /* ignore CANCELs */
 12328	  nop_select,           /* ignore selects */
 12329	  w_other,              /* catch-all for unrecognized commands and ioctls */
 12330	  w_hw_int              /* leftover hardware interrupts */
 12331	};
 12332	
 12333	/*===========================================================================*
 12334	 *                              at_winchester_task                           *
 12335	 *===========================================================================*/
 12336	PUBLIC int main()
 12337	{
 12338	/* Set special disk parameters then call the generic main loop. */
 12339	  init_params();
 12340	  driver_task(&w_dtab);
 12341	  return(OK);
 12342	}
	
 12344	/*===========================================================================*
 12345	 *                              init_params                                  *
 12346	 *===========================================================================*/
 12347	PRIVATE void init_params()
 12348	{
 12349	/* This routine is called at startup to initialize the drive parameters. */
 12350	
 12351	  u16_t parv[2];
 12352	  unsigned int vector, size;
 12353	  int drive, nr_drives;
 12354	  struct wini *wn;

_________________________ Page 792    File: drivers/at_wini/at_wini.c ____________________
 12355	  u8_t params[16];
 12356	  int s;
 12357	
 12358	  /* Boot variables. */
 12359	  env_parse("ata_std_timeout", "d", 0, &w_standard_timeouts, 0, 1);
 12360	  env_parse("ata_pci_debug", "d", 0, &w_pci_debug, 0, 1);
 12361	  env_parse("ata_instance", "d", 0, &w_instance, 0, 8);
 12362	  env_parse("ata_lba48", "d", 0, &w_lba48, 0, 1);
 12363	  env_parse("atapi_debug", "d", 0, &atapi_debug, 0, 1);
 12364	
 12365	  if (w_instance == 0) {
 12366	          /* Get the number of drives from the BIOS data area */
 12367	          if ((s=sys_vircopy(SELF, BIOS_SEG, NR_HD_DRIVES_ADDR, 
 12368	                        SELF, D, (vir_bytes) params, NR_HD_DRIVES_SIZE)) != OK)
 12369	                panic(w_name(), "Couldn't read BIOS", s);
 12370	          if ((nr_drives = params[0]) > 2) nr_drives = 2;
 12371	
 12372	          for (drive = 0, wn = wini; drive < COMPAT_DRIVES; drive++, wn++) {
 12373	                if (drive < nr_drives) {
 12374	                    /* Copy the BIOS parameter vector */
 12375	                    vector = (drive == 0) ? BIOS_HD0_PARAMS_ADDR:BIOS_HD1_PARAMS_AD
 12376	                    size = (drive == 0) ? BIOS_HD0_PARAMS_SIZE:BIOS_HD1_PARAMS_SIZE
 12377	                    if ((s=sys_vircopy(SELF, BIOS_SEG, vector,
 12378	                                        SELF, D, (vir_bytes) parv, size)) != OK)
 12379	                                panic(w_name(), "Couldn't read BIOS", s);
 12380	        
 12381	                        /* Calculate the address of the parameters and copy them */
 12382	                        if ((s=sys_vircopy(
 12383	                                SELF, BIOS_SEG, hclick_to_physb(parv[1]) + parv[0],
 12384	                                SELF, D, (phys_bytes) params, 16L))!=OK)
 12385	                            panic(w_name(),"Couldn't copy parameters", s);
 12386	        
 12387	                        /* Copy the parameters to the structures of the drive */
 12388	                        wn->lcylinders = bp_cylinders(params);
 12389	                        wn->lheads = bp_heads(params);
 12390	                        wn->lsectors = bp_sectors(params);
 12391	                        wn->precomp = bp_precomp(params) >> 2;
 12392	                }
 12393	
 12394	                /* Fill in non-BIOS parameters. */
 12395	                init_drive(wn,
 12396	                        drive < 2 ? REG_CMD_BASE0 : REG_CMD_BASE1,
 12397	                        drive < 2 ? REG_CTL_BASE0 : REG_CTL_BASE1,
 12398	                        NO_IRQ, 0, 0, drive);
 12399	                w_next_drive++;
 12400	        }
 12401	  }
 12402	
 12403	  /* Look for controllers on the pci bus. Skip none the first instance,
 12404	   * skip one and then 2 for every instance, for every next instance.
 12405	   */
 12406	  if (w_instance == 0)
 12407	        init_params_pci(0);
 12408	  else
 12409	        init_params_pci(w_instance*2-1);
 12410	
 12411	}
	
 12413	#define ATA_IF_NOTCOMPAT1 (1L << 0)
 12414	#define ATA_IF_NOTCOMPAT2 (1L << 2)

_________________________ Page 793    File: drivers/at_wini/at_wini.c ____________________
 12415	
 12416	/*===========================================================================*
 12417	 *                              init_drive                                   *
 12418	 *===========================================================================*/
 12419	PRIVATE void init_drive(struct wini *w, int base_cmd, int base_ctl, int irq, int ac
 12420	{
 12421	        w->state = 0;
 12422	        w->w_status = 0;
 12423	        w->base_cmd = base_cmd;
 12424	        w->base_ctl = base_ctl;
 12425	        w->irq = irq;
 12426	        w->irq_mask = 1 << irq;
 12427	        w->irq_need_ack = ack;
 12428	        w->irq_hook_id = hook;
 12429	        w->ldhpref = ldh_init(drive);
 12430	        w->max_count = MAX_SECS << SECTOR_SHIFT;
 12431	        w->lba48 = 0;
 12432	}
	
 12434	/*===========================================================================*
 12435	 *                              init_params_pci                              *
 12436	 *===========================================================================*/
 12437	PRIVATE void init_params_pci(int skip)
 12438	{
 12439	  int r, devind, drive;
 12440	  u16_t vid, did;
 12441	  pci_init();
 12442	  for(drive = w_next_drive; drive < MAX_DRIVES; drive++)
 12443	        wini[drive].state = IGNORING;
 12444	  for(r = pci_first_dev(&devind, &vid, &did);
 12445	        r != 0 && w_next_drive < MAX_DRIVES; r = pci_next_dev(&devin
 12446	        int interface, irq, irq_hook;
 12447	        /* Base class must be 01h (mass storage), subclass must
 12448	         * be 01h (ATA).
 12449	         */
 12450	        if (pci_attr_r8(devind, PCI_BCR) != 0x01 ||
 12451	           pci_attr_r8(devind, PCI_SCR) != 0x01) {
 12452	           continue;
 12453	        }
 12454	        /* Found a controller.
 12455	         * Programming interface register tells us more.
 12456	         */
 12457	        interface = pci_attr_r8(devind, PCI_PIFR);
 12458	        irq = pci_attr_r8(devind, PCI_ILR);
 12459	
 12460	        /* Any non-compat drives? */
 12461	        if (interface & (ATA_IF_NOTCOMPAT1 | ATA_IF_NOTCOMPAT2)) {
 12462	                int s;
 12463	                irq_hook = irq;
 12464	                if (skip > 0) {
 12465	                        if (w_pci_debug) printf("atapci skipping controller (remain
 12466	                        skip--;
 12467	                        continue;
 12468	                }
 12469	                if ((s=sys_irqsetpolicy(irq, 0, &irq_hook)) != OK) {
 12470	                        printf("atapci: couldn't set IRQ policy %d\n", irq);
 12471	                        continue;
 12472	                }
 12473	                if ((s=sys_irqenable(&irq_hook)) != OK) {
 12474	                        printf("atapci: couldn't enable IRQ line %d\n", irq);

_________________________ Page 794    File: drivers/at_wini/at_wini.c ____________________
 12475	                        continue;
 12476	                }
 12477	        } else {
 12478	                /* If not.. this is not the ata-pci controller we're
 12479	                 * looking for.
 12480	                 */
 12481	                if (w_pci_debug) printf("atapci skipping compatability controller\n
 12482	                continue;
 12483	        }
 12484	
 12485	        /* Primary channel not in compatability mode? */
 12486	        if (interface & ATA_IF_NOTCOMPAT1) {
 12487	                u32_t base_cmd, base_ctl;
 12488	                base_cmd = pci_attr_r32(devind, PCI_BAR) & 0xffffffe0;
 12489	                base_ctl = pci_attr_r32(devind, PCI_BAR_2) & 0xffffffe0;
 12490	                if (base_cmd != REG_CMD_BASE0 && base_cmd != REG_CMD_BASE1)
 12491	                        init_drive(&wini[w_next_drive],
 12492	                                base_cmd, base_ctl, irq, 1, irq_hook, 0);
 12493	                        init_drive(&wini[w_next_drive+1],
 12494	                                base_cmd, base_ctl, irq, 1, irq_hook, 1);
 12495	                        if (w_pci_debug)
 12496	                                printf("atapci %d: 0x%x 0x%x irq %d\n", devind, bas
 12497	                } else printf("atapci: ignored drives on primary channel, base %x\n
 12498	        }
 12499	
 12500	        /* Secondary channel not in compatability mode? */
 12501	        if (interface & ATA_IF_NOTCOMPAT2) {
 12502	                u32_t base_cmd, base_ctl;
 12503	                base_cmd = pci_attr_r32(devind, PCI_BAR_3) & 0xffffffe0;
 12504	                base_ctl = pci_attr_r32(devind, PCI_BAR_4) & 0xffffffe0;
 12505	                if (base_cmd != REG_CMD_BASE0 && base_cmd != REG_CMD_BASE1)
 12506	                        init_drive(&wini[w_next_drive+2],
 12507	                                base_cmd, base_ctl, irq, 1, irq_hook, 2);
 12508	                        init_drive(&wini[w_next_drive+3],
 12509	                                base_cmd, base_ctl, irq, 1, irq_hook, 3);
 12510	                        if (w_pci_debug)
 12511	                                printf("atapci %d: 0x%x 0x%x irq %d\n", devind, bas
 12512	                } else printf("atapci: ignored drives on secondary channel, base %x
 12513	        }
 12514	        w_next_drive += 4;
 12515	  }
 12516	}
	
 12518	/*===========================================================================*
 12519	 *                              w_do_open                                    *
 12520	 *===========================================================================*/
 12521	PRIVATE int w_do_open(dp, m_ptr)
 12522	struct driver *dp;
 12523	message *m_ptr;
 12524	{
 12525	/* Device open: Initialize the controller and read the partition table. */
 12526	
 12527	  struct wini *wn;
 12528	
 12529	  if (w_prepare(m_ptr->DEVICE) == NIL_DEV) return(ENXIO);
 12530	
 12531	  wn = w_wn;
 12532	
 12533	  /* If we've probed it before and it failed, don't probe it again. */
 12534	  if (wn->state & IGNORING) return ENXIO;

_________________________ Page 795    File: drivers/at_wini/at_wini.c ____________________
 12535	
 12536	  /* If we haven't identified it yet, or it's gone deaf, 
 12537	   * (re-)identify it.
 12538	   */
 12539	  if (!(wn->state & IDENTIFIED) || (wn->state & DEAF)) {
 12540	        /* Try to identify the device. */
 12541	        if (w_identify() != OK) {
 12542	                if (wn->state & DEAF) w_reset();
 12543	                wn->state = IGNORING;
 12544	                return(ENXIO);
 12545	        }
 12546	          /* Do a test transaction unless it's a CD drive (then
 12547	           * we can believe the controller, and a test may fail
 12548	           * due to no CD being in the drive). If it fails, ignore
 12549	           * the device forever.
 12550	           */
 12551	          if (!(wn->state & ATAPI) && w_io_test() != OK) {
 12552	                wn->state |= IGNORING;
 12553	                return(ENXIO);
 12554	          }
 12555	  }
 12556	
 12557	   /* If it's not an ATAPI device, then don't open with RO_BIT. */
 12558	   if (!(wn->state & ATAPI) && (m_ptr->COUNT & RO_BIT)) return EACC
 12559	
 12560	  /* Partition the drive if it's being opened for the first time,
 12561	   * or being opened after being closed.
 12562	   */
 12563	  if (wn->open_ct == 0) {
 12564	
 12565	        /* Partition the disk. */
 12566	        memset(wn->part, sizeof(wn->part), 0);
 12567	        memset(wn->subpart, sizeof(wn->subpart), 0);
 12568	        partition(&w_dtab, w_drive * DEV_PER_DRIVE, P_PRIMARY, wn->state & 
 12569	  }
 12570	  wn->open_ct++;
 12571	  return(OK);
 12572	}
	
 12574	/*===========================================================================*
 12575	 *                              w_prepare                                    *
 12576	 *===========================================================================*/
 12577	PRIVATE struct device *w_prepare(int device)
 12578	{
 12579	/* Prepare for I/O on a device. */
 12580	struct wini *prev_wn;
 12581	prev_wn = w_wn;
 12582	  w_device = device;
 12583	
 12584	  if (device < NR_MINORS) {                     /* d0, d0p[0-3], d1, ... */
 12585	        w_drive = device / DEV_PER_DRIVE;       /* save drive number */
 12586	        w_wn = &wini[w_drive];
 12587	        w_dv = &w_wn->part[device % DEV_PER_DRIVE];
 12588	  } else
 12589	  if ((unsigned) (device -= MINOR_d0p0s0) < NR_SUBDEVS) {/*d[0-7]p[0-3]s[0-3]*/
 12590	        w_drive = device / SUB_PER_DRIVE;
 12591	        w_wn = &wini[w_drive];
 12592	        w_dv = &w_wn->subpart[device % SUB_PER_DRIVE];
 12593	  } else {
 12594	        w_device = -1;

_________________________ Page 796    File: drivers/at_wini/at_wini.c ____________________
 12595	        return(NIL_DEV);
 12596	  }
 12597	  return(w_dv);
 12598	}
	
 12600	/*===========================================================================*
 12601	 *                              w_identify                                   *
 12602	 *===========================================================================*/
 12603	PRIVATE int w_identify()
 12604	{
 12605	/* Find out if a device exists, if it is an old AT disk, or a newer ATA
 12606	 * drive, a removable media device, etc.
 12607	 */
 12608	
 12609	  struct wini *wn = w_wn;
 12610	  struct command cmd;
 12611	  int i, s;
 12612	  unsigned long size;
 12613	#define id_byte(n)      (&tmp_buf[2 * (n)])
 12614	#define id_word(n)      (((u16_t) id_byte(n)[0] <<  0) \
 12615	                        |((u16_t) id_byte(n)[1] <<  8))
 12616	#define id_longword(n)  (((u32_t) id_byte(n)[0] <<  0) \
 12617	                        |((u32_t) id_byte(n)[1] <<  8) \
 12618	                        |((u32_t) id_byte(n)[2] << 16) \
 12619	                        |((u32_t) id_byte(n)[3] << 24))
 12620	
 12621	  /* Try to identify the device. */
 12622	  cmd.ldh     = wn->ldhpref;
 12623	  cmd.command = ATA_IDENTIFY;
 12624	  if (com_simple(&cmd) == OK) {
 12625	        /* This is an ATA device. */
 12626	        wn->state |= SMART;
 12627	
 12628	        /* Device information. */
 12629	        if ((s=sys_insw(wn->base_cmd + REG_DATA, SELF, tmp_buf, SECTOR_SIZE)) != OK
 12630	                panic(w_name(),"Call to sys_insw() failed", s);
 12631	
 12632	        /* Why are the strings byte swapped??? */
 12633	        for (i = 0; i < 40; i++) w_id_string[i] = id_byte(27)[i^1];
 12634	
 12635	        /* Preferred CHS translation mode. */
 12636	        wn->pcylinders = id_word(1);
 12637	        wn->pheads = id_word(3);
 12638	        wn->psectors = id_word(6);
 12639	        size = (u32_t) wn->pcylinders * wn->pheads * wn->psectors;
 12640	
 12641	        if ((id_byte(49)[1] & 0x02) && size > 512L*1024*2) {
 12642	                /* Drive is LBA capable and is big enough to trust it to
 12643	                 * not make a mess of it.
 12644	                 */
 12645	                wn->ldhpref |= LDH_LBA;
 12646	                size = id_longword(60);
 12647	
 12648	                if (w_lba48 && ((id_word(83)) & (1L << 10))) {
 12649	                        /* Drive is LBA48 capable (and LBA48 is turned on). */
 12650	                        if (id_word(102) || id_word(103)) {
 12651	                                /* If no. of sectors doesn't fit in 32 bits,
 12652	                                 * trunacte to this. So it's LBA32 for now.
 12653	                                 * This can still address devices up to 2TB
 12654	                                 * though.

_________________________ Page 797    File: drivers/at_wini/at_wini.c ____________________
 12655	                                 */
 12656	                                size = ULONG_MAX;
 12657	                        } else {
 12658	                                /* Actual number of sectors fits in 32 bits. */
 12659	                                size = id_longword(100);
 12660	                        }
 12661	
 12662	                        wn->lba48 = 1;
 12663	                }
 12664	        }
 12665	
 12666	        if (wn->lcylinders == 0) {
 12667	                /* No BIOS parameters?  Then make some up. */
 12668	                wn->lcylinders = wn->pcylinders;
 12669	                wn->lheads = wn->pheads;
 12670	                wn->lsectors = wn->psectors;
 12671	                while (wn->lcylinders > 1024) {
 12672	                        wn->lheads *= 2;
 12673	                        wn->lcylinders /= 2;
 12674	                }
 12675	        }
 12676	  } else {
 12677	        /* Not an ATA device; no translations, no special features.  Don't
 12678	         * touch it unless the BIOS knows about it.
 12679	         */
 12680	        if (wn->lcylinders == 0) { return(ERR); }       /* no BIOS parameters */
 12681	        wn->pcylinders = wn->lcylinders;
 12682	        wn->pheads = wn->lheads;
 12683	        wn->psectors = wn->lsectors;
 12684	        size = (u32_t) wn->pcylinders * wn->pheads * wn->psectors;
 12685	  }
 12686	
 12687	  /* Size of the whole drive */
 12688	  wn->part[0].dv_size = mul64u(size, SECTOR_SIZE);
 12689	
 12690	  /* Reset/calibrate (where necessary) */
 12691	  if (w_specify() != OK && w_specify() != OK) {
 12692	        return(ERR);
 12693	  }
 12694	
 12695	  if (wn->irq == NO_IRQ) {
 12696	          /* Everything looks OK; register IRQ so we can stop polling. */
 12697	          wn->irq = w_drive < 2 ? AT_WINI_0_IRQ : AT_WINI_1_IRQ;
 12698	          wn->irq_hook_id = wn->irq;    /* id to be returned if interrupt occurs */
 12699	          if ((s=sys_irqsetpolicy(wn->irq, IRQ_REENABLE, &wn->irq_hook_id)) != 
 12700	                panic(w_name(), "couldn't set IRQ policy", s);
 12701	          if ((s=sys_irqenable(&wn->irq_hook_id)) != OK)
 12702	                panic(w_name(), "couldn't enable IRQ line", s);
 12703	  }
 12704	  wn->state |= IDENTIFIED;
 12705	  return(OK);
 12706	}
	
 12708	/*===========================================================================*
 12709	 *                              w_name                                       *
 12710	 *===========================================================================*/
 12711	PRIVATE char *w_name()
 12712	{
 12713	/* Return a name for the current device. */
 12714	  static char name[] = "AT-D0";

_________________________ Page 798    File: drivers/at_wini/at_wini.c ____________________
 12715	
 12716	  name[4] = '0' + w_drive;
 12717	  return name;
 12718	}
	
 12720	/*===========================================================================*
 12721	 *                              w_io_test                                    *
 12722	 *===========================================================================*/
 12723	PRIVATE int w_io_test(void)
 12724	{
 12725	        int r, save_dev;
 12726	        int save_timeout, save_errors, save_wakeup;
 12727	        iovec_t iov;
 12728	        static char buf[SECTOR_SIZE];
 12729	        iov.iov_addr = (vir_bytes) buf;
 12730	        iov.iov_size = sizeof(buf);
 12731	        save_dev = w_device;
 12732	
 12733	        /* Reduce timeout values for this test transaction. */
 12734	        save_timeout = timeout_ticks;
 12735	        save_errors = max_errors;
 12736	        save_wakeup = wakeup_ticks;
 12737	
 12738	        if (!w_standard_timeouts) {
 12739	                timeout_ticks = HZ * 4;
 12740	                wakeup_ticks = HZ * 6;
 12741	                max_errors = 3;
 12742	        }
 12743	
 12744	        w_testing = 1;
 12745	
 12746	        /* Try I/O on the actual drive (not any (sub)partition). */
 12747	        if (w_prepare(w_drive * DEV_PER_DRIVE) == NIL_DEV)
 12748	                panic(w_name(), "Couldn't switch devices", NO_NUM);
 12749	
 12750	        r = w_transfer(SELF, DEV_GATHER, 0, &iov, 1);
 12751	
 12752	        /* Switch back. */
 12753	        if (w_prepare(save_dev) == NIL_DEV)
 12754	                panic(w_name(), "Couldn't switch back devices", NO_NUM);
 12755	
 12756	        /* Restore parameters. */
 12757	        timeout_ticks = save_timeout;
 12758	        max_errors = save_errors;
 12759	        wakeup_ticks = save_wakeup;
 12760	        w_testing = 0;
 12761	
 12762	        /* Test if everything worked. */
 12763	        if (r != OK || iov.iov_size != 0) {
 12764	                return ERR;
 12765	        }
 12766	
 12767	        /* Everything worked. */
 12768	
 12769	        return OK;
 12770	}
	

_________________________ Page 799    File: drivers/at_wini/at_wini.c ____________________
 12772	/*===========================================================================*
 12773	 *                              w_specify                                    *
 12774	 *===========================================================================*/
 12775	PRIVATE int w_specify()
 12776	{
 12777	/* Routine to initialize the drive after boot or when a reset is needed. */
 12778	
 12779	  struct wini *wn = w_wn;
 12780	  struct command cmd;
 12781	
 12782	  if ((wn->state & DEAF) && w_reset() != OK) {
 12783	        return(ERR);
 12784	  }
 12785	
 12786	  if (!(wn->state & ATAPI)) {
 12787	        /* Specify parameters: precompensation, number of heads and sectors. */
 12788	        cmd.precomp = wn->precomp;
 12789	        cmd.count   = wn->psectors;
 12790	        cmd.ldh     = w_wn->ldhpref | (wn->pheads - 1);
 12791	        cmd.command = CMD_SPECIFY;              /* Specify some parameters */
 12792	
 12793	        /* Output command block and see if controller accepts the parameters. */
 12794	        if (com_simple(&cmd) != OK) return(ERR);
 12795	
 12796	        if (!(wn->state & SMART)) {
 12797	                /* Calibrate an old disk. */
 12798	                cmd.sector  = 0;
 12799	                cmd.cyl_lo  = 0;
 12800	                cmd.cyl_hi  = 0;
 12801	                cmd.ldh     = w_wn->ldhpref;
 12802	                cmd.command = CMD_RECALIBRATE;
 12803	
 12804	                if (com_simple(&cmd) != OK) return(ERR);
 12805	        }
 12806	  }
 12807	  wn->state |= INITIALIZED;
 12808	  return(OK);
 12809	}
	
 12811	/*===========================================================================*
 12812	 *                              do_transfer                                  *
 12813	 *===========================================================================*/
 12814	PRIVATE int do_transfer(struct wini *wn, unsigned int precomp, unsigned int count,
 12815	        unsigned int sector, unsigned int opcode)
 12816	{
 12817	        struct command cmd;
 12818	        unsigned secspcyl = wn->pheads * wn->psectors;
 12819	
 12820	        cmd.precomp = precomp;
 12821	        cmd.count   = count;
 12822	        cmd.command = opcode == DEV_SCATTER ? CMD_WRITE : CMD_READ;
 12823	        /* 
 12824	        if (w_lba48 && wn->lba48) {
 12825	        } else  */
 12826	        if (wn->ldhpref & LDH_LBA) {
 12827	                cmd.sector  = (sector >>  0) & 0xFF;
 12828	                cmd.cyl_lo  = (sector >>  8) & 0xFF;
 12829	                cmd.cyl_hi  = (sector >> 16) & 0xFF;
 12830	                cmd.ldh     = wn->ldhpref | ((sector >> 24) & 0xF);
 12831	        } else {

_________________________ Page 800    File: drivers/at_wini/at_wini.c ____________________
 12832	                int cylinder, head, sec;
 12833	                cylinder = sector / secspcyl;
 12834	                head = (sector % secspcyl) / wn->psectors;
 12835	                sec = sector % wn->psectors;
 12836	                cmd.sector  = sec + 1;
 12837	                cmd.cyl_lo  = cylinder & BYTE;
 12838	                cmd.cyl_hi  = (cylinder >> 8) & BYTE;
 12839	                cmd.ldh     = wn->ldhpref | head;
 12840	        }
 12841	
 12842	        return com_out(&cmd);
 12843	}
	
 12845	/*===========================================================================*
 12846	 *                              w_transfer                                   *
 12847	 *===========================================================================*/
 12848	PRIVATE int w_transfer(proc_nr, opcode, position, iov, nr_req)
 12849	int proc_nr;                    /* process doing the request */
 12850	int opcode;                     /* DEV_GATHER or DEV_SCATTER */
 12851	off_t position;                 /* offset on device to read or write */
 12852	iovec_t *iov;                   /* pointer to read or write request vector */
 12853	unsigned nr_req;                /* length of request vector */
 12854	{
 12855	  struct wini *wn = w_wn;
 12856	  iovec_t *iop, *iov_end = iov + nr_req;
 12857	  int r, s, errors;
 12858	  unsigned long block;
 12859	  unsigned long dv_size = cv64ul(w_dv->dv_size);
 12860	  unsigned cylinder, head, sector, nbytes;
 12861	
 12862	  /* Check disk address. */
 12863	  if ((position & SECTOR_MASK) != 0) return(EINVAL);
 12864	
 12865	  errors = 0;
 12866	
 12867	  while (nr_req > 0) {
 12868	        /* How many bytes to transfer? */
 12869	        nbytes = 0;
 12870	        for (iop = iov; iop < iov_end; iop++) nbytes += iop->iov_size;
 12871	        if ((nbytes & SECTOR_MASK) != 0) return(EINVAL);
 12872	
 12873	        /* Which block on disk and how close to EOF? */
 12874	        if (position >= dv_size) return(OK);            /* At EOF */
 12875	        if (position + nbytes > dv_size) nbytes = dv_size - position;
 12876	        block = div64u(add64ul(w_dv->dv_base, position), SECTOR_SIZE);
 12877	
 12878	        if (nbytes >= wn->max_count) {
 12879	                /* The drive can't do more then max_count at once. */
 12880	                nbytes = wn->max_count;
 12881	        }
 12882	
 12883	        /* First check to see if a reinitialization is needed. */
 12884	        if (!(wn->state & INITIALIZED) && w_specify() != OK) return(EIO
 12885	
 12886	        /* Tell the controller to transfer nbytes bytes. */
 12887	        r = do_transfer(wn, wn->precomp, ((nbytes >> SECTOR_SHIFT) & BYTE),
 12888	                block, opcode);
 12889	
 12890	        while (r == OK && nbytes > 0) {
 12891	                /* For each sector, wait for an interrupt and fetch the data

_________________________ Page 801    File: drivers/at_wini/at_wini.c ____________________
 12892	                 * (read), or supply data to the controller and wait for an
 12893	                 * interrupt (write).
 12894	                 */
 12895	
 12896	                if (opcode == DEV_GATHER) {
 12897	                        /* First an interrupt, then data. */
 12898	                        if ((r = at_intr_wait()) != OK) {
 12899	                                /* An error, send data to the bit bucket. */
 12900	                                if (w_wn->w_status & STATUS_DRQ) {
 12901	        if ((s=sys_insw(wn->base_cmd + REG_DATA, SELF, tmp_buf, SECTOR_SIZE)) != OK
 12902	                panic(w_name(),"Call to sys_insw() failed", s);
 12903	                                }
 12904	                                break;
 12905	                        }
 12906	                }
 12907	
 12908	                /* Wait for data transfer requested. */
 12909	                if (!w_waitfor(STATUS_DRQ, STATUS_DRQ)) { r = ERR; break; }
 12910	
 12911	                /* Copy bytes to or from the device's buffer. */
 12912	                if (opcode == DEV_GATHER) {
 12913	        if ((s=sys_insw(wn->base_cmd + REG_DATA, proc_nr, (void *) iov->iov_addr, S
 12914	                panic(w_name(),"Call to sys_insw() failed", s);
 12915	                } else {
 12916	        if ((s=sys_outsw(wn->base_cmd + REG_DATA, proc_nr, (void *) iov->iov_addr, 
 12917	                panic(w_name(),"Call to sys_insw() failed", s);
 12918	
 12919	                        /* Data sent, wait for an interrupt. */
 12920	                        if ((r = at_intr_wait()) != OK) break;
 12921	                }
 12922	
 12923	                /* Book the bytes successfully transferred. */
 12924	                nbytes -= SECTOR_SIZE;
 12925	                position += SECTOR_SIZE;
 12926	                iov->iov_addr += SECTOR_SIZE;
 12927	                if ((iov->iov_size -= SECTOR_SIZE) == 0) { iov++; nr_req--; }
 12928	        }
 12929	
 12930	        /* Any errors? */
 12931	        if (r != OK) {
 12932	                /* Don't retry if sector marked bad or too many errors. */
 12933	                if (r == ERR_BAD_SECTOR || ++errors == max_errors) {
 12934	                        w_command = CMD_IDLE;
 12935	                        return(EIO);
 12936	                }
 12937	        }
 12938	  }
 12939	
 12940	  w_command = CMD_IDLE;
 12941	  return(OK);
 12942	}
	
 12944	/*===========================================================================*
 12945	 *                              com_out                                      *
 12946	 *===========================================================================*/
 12947	PRIVATE int com_out(cmd)
 12948	struct command *cmd;            /* Command block */
 12949	{
 12950	/* Output the command block to the winchester controller and return status */
 12951	

_________________________ Page 802    File: drivers/at_wini/at_wini.c ____________________
 12952	  struct wini *wn = w_wn;
 12953	  unsigned base_cmd = wn->base_cmd;
 12954	  unsigned base_ctl = wn->base_ctl;
 12955	  pvb_pair_t outbyte[7];                /* vector for sys_voutb() */
 12956	  int s;                                /* status for sys_(v)outb() */
 12957	
 12958	  if (w_wn->state & IGNORING) return ERR;
 12959	
 12960	  if (!w_waitfor(STATUS_BSY, 0)) {
 12961	        printf("%s: controller not ready\n", w_name());
 12962	        return(ERR);
 12963	  }
 12964	
 12965	  /* Select drive. */
 12966	  if ((s=sys_outb(base_cmd + REG_LDH, cmd->ldh)) != OK)
 12967	        panic(w_name(),"Couldn't write register to select drive",s);
 12968	
 12969	  if (!w_waitfor(STATUS_BSY, 0)) {
 12970	        printf("%s: com_out: drive not ready\n", w_name());
 12971	        return(ERR);
 12972	  }
 12973	
 12974	  /* Schedule a wakeup call, some controllers are flaky. This is done with
 12975	   * a synchronous alarm. If a timeout occurs a SYN_ALARM message is sent
 12976	   * from HARDWARE, so that w_intr_wait() can call w_timeout() in case the
 12977	   * controller was not able to execute the command. Leftover timeouts are
 12978	   * simply ignored by the main loop. 
 12979	   */
 12980	  sys_setalarm(wakeup_ticks, 0);
 12981	
 12982	  wn->w_status = STATUS_ADMBSY;
 12983	  w_command = cmd->command;
 12984	  pv_set(outbyte[0], base_ctl + REG_CTL, wn->pheads >= 8 ? CTL_EIGHTHEADS : 0);
 12985	  pv_set(outbyte[1], base_cmd + REG_PRECOMP, cmd->precomp);
 12986	  pv_set(outbyte[2], base_cmd + REG_COUNT, cmd->count);
 12987	  pv_set(outbyte[3], base_cmd + REG_SECTOR, cmd->sector);
 12988	  pv_set(outbyte[4], base_cmd + REG_CYL_LO, cmd->cyl_lo);
 12989	  pv_set(outbyte[5], base_cmd + REG_CYL_HI, cmd->cyl_hi);
 12990	  pv_set(outbyte[6], base_cmd + REG_COMMAND, cmd->command);
 12991	  if ((s=sys_voutb(outbyte,7)) != OK)
 12992	        panic(w_name(),"Couldn't write registers with sys_voutb()",s);
 12993	  return(OK);
 12994	}
	
 12996	/*===========================================================================*
 12997	 *                              w_need_reset                                 *
 12998	 *===========================================================================*/
 12999	PRIVATE void w_need_reset()
 13000	{
 13001	/* The controller needs to be reset. */
 13002	  struct wini *wn;
 13003	  int dr = 0;
 13004	
 13005	  for (wn = wini; wn < &wini[MAX_DRIVES]; wn++, dr++) {
 13006	        if (wn->base_cmd == w_wn->base_cmd) {
 13007	                wn->state |= DEAF;
 13008	                wn->state &= ~INITIALIZED;
 13009	        }
 13010	  }
 13011	}

_________________________ Page 803    File: drivers/at_wini/at_wini.c ____________________
	
 13013	/*===========================================================================*
 13014	 *                              w_do_close                                   *
 13015	 *===========================================================================*/
 13016	PRIVATE int w_do_close(dp, m_ptr)
 13017	struct driver *dp;
 13018	message *m_ptr;
 13019	{
 13020	/* Device close: Release a device. */
 13021	  if (w_prepare(m_ptr->DEVICE) == NIL_DEV)
 13022	        return(ENXIO);
 13023	  w_wn->open_ct--;
 13024	  return(OK);
 13025	}
	
 13027	/*===========================================================================*
 13028	 *                              com_simple                                   *
 13029	 *===========================================================================*/
 13030	PRIVATE int com_simple(cmd)
 13031	struct command *cmd;            /* Command block */
 13032	{
 13033	/* A simple controller command, only one interrupt and no data-out phase. */
 13034	  int r;
 13035	
 13036	  if (w_wn->state & IGNORING) return ERR;
 13037	
 13038	  if ((r = com_out(cmd)) == OK) r = at_intr_wait();
 13039	  w_command = CMD_IDLE;
 13040	  return(r);
 13041	}
	
 13043	/*===========================================================================*
 13044	 *                              w_timeout                                    *
 13045	 *===========================================================================*/
 13046	PRIVATE void w_timeout(void)
 13047	{
 13048	  struct wini *wn = w_wn;
 13049	
 13050	  switch (w_command) {
 13051	  case CMD_IDLE:
 13052	        break;          /* fine */
 13053	  case CMD_READ:
 13054	  case CMD_WRITE:
 13055	        /* Impossible, but not on PC's:  The controller does not respond. */
 13056	
 13057	        /* Limiting multisector I/O seems to help. */
 13058	        if (wn->max_count > 8 * SECTOR_SIZE) {
 13059	                wn->max_count = 8 * SECTOR_SIZE;
 13060	        } else {
 13061	                wn->max_count = SECTOR_SIZE;
 13062	        }
 13063	        /*FALL THROUGH*/
 13064	  default:
 13065	        /* Some other command. */
 13066	        if (w_testing)  wn->state |= IGNORING;  /* Kick out this drive. */
 13067	        else if (!w_silent) printf("%s: timeout on command %02x\n", w_name(), w_com
 13068	        w_need_reset();
 13069	        wn->w_status = 0;
 13070	  }
 13071	}

_________________________ Page 804    File: drivers/at_wini/at_wini.c ____________________
	
 13073	/*===========================================================================*
 13074	 *                              w_reset                                      *
 13075	 *===========================================================================*/
 13076	PRIVATE int w_reset()
 13077	{
 13078	/* Issue a reset to the controller.  This is done after any catastrophe,
 13079	 * like the controller refusing to respond.
 13080	 */
 13081	  int s;
 13082	  struct wini *wn = w_wn;
 13083	
 13084	  /* Don't bother if this drive is forgotten. */
 13085	  if (w_wn->state & IGNORING) return ERR;
 13086	
 13087	  /* Wait for any internal drive recovery. */
 13088	  tickdelay(RECOVERY_TICKS);
 13089	
 13090	  /* Strobe reset bit */
 13091	  if ((s=sys_outb(wn->base_ctl + REG_CTL, CTL_RESET)) != OK)
 13092	        panic(w_name(),"Couldn't strobe reset bit",s);
 13093	  tickdelay(DELAY_TICKS);
 13094	  if ((s=sys_outb(wn->base_ctl + REG_CTL, 0)) != OK)
 13095	        panic(w_name(),"Couldn't strobe reset bit",s);
 13096	  tickdelay(DELAY_TICKS);
 13097	
 13098	  /* Wait for controller ready */
 13099	  if (!w_waitfor(STATUS_BSY, 0)) {
 13100	        printf("%s: reset failed, drive busy\n", w_name());
 13101	        return(ERR);
 13102	  }
 13103	
 13104	  /* The error register should be checked now, but some drives mess it up. */
 13105	
 13106	  for (wn = wini; wn < &wini[MAX_DRIVES]; wn++) {
 13107	        if (wn->base_cmd == w_wn->base_cmd) {
 13108	                wn->state &= ~DEAF;
 13109	                if (w_wn->irq_need_ack) {
 13110	                        /* Make sure irq is actually enabled.. */
 13111	                        sys_irqenable(&w_wn->irq_hook_id);
 13112	                }
 13113	        }
 13114	  }
 13115	                
 13116	
 13117	  return(OK);
 13118	}
	
 13120	/*===========================================================================*
 13121	 *                              w_intr_wait                                  *
 13122	 *===========================================================================*/
 13123	PRIVATE void w_intr_wait()
 13124	{
 13125	/* Wait for a task completion interrupt. */
 13126	
 13127	  message m;
 13128	
 13129	  if (w_wn->irq != NO_IRQ) {
 13130	        /* Wait for an interrupt that sets w_status to "not busy". */
 13131	        while (w_wn->w_status & (STATUS_ADMBSY|STATUS_BSY)) {

_________________________ Page 805    File: drivers/at_wini/at_wini.c ____________________
 13132	                receive(ANY, &m);               /* expect HARD_INT message */
 13133	                if (m.m_type == SYN_ALARM) {    /* but check for timeout */
 13134	                    w_timeout();                /* a.o. set w_status */
 13135	                } else if (m.m_type == HARD_INT) {
 13136	                    sys_inb(w_wn->base_cmd + REG_STATUS, &w_wn->w_status);
 13137	                    ack_irqs(m.NOTIFY_ARG);
 13138	                } else {
 13139	                        printf("AT_WINI got unexpected message %d from %d\n",
 13140	                                m.m_type, m.m_source);
 13141	                }
 13142	        }
 13143	  } else {
 13144	        /* Interrupt not yet allocated; use polling. */
 13145	        (void) w_waitfor(STATUS_BSY, 0);
 13146	  }
 13147	}
	
 13149	/*===========================================================================*
 13150	 *                              at_intr_wait                                 *
 13151	 *===========================================================================*/
 13152	PRIVATE int at_intr_wait()
 13153	{
 13154	/* Wait for an interrupt, study the status bits and return error/success. */
 13155	  int r;
 13156	  int s,inbval;         /* read value with sys_inb */ 
 13157	
 13158	  w_intr_wait();
 13159	  if ((w_wn->w_status & (STATUS_BSY | STATUS_WF | STATUS_ERR)) == 0) {
 13160	        r = OK;
 13161	  } else {
 13162	        if ((s=sys_inb(w_wn->base_cmd + REG_ERROR, &inbval)) != OK)
 13163	                panic(w_name(),"Couldn't read register",s);
 13164	        if ((w_wn->w_status & STATUS_ERR) && (inbval & ERROR_BB)) {
 13165	                r = ERR_BAD_SECTOR;     /* sector marked bad, retries won't help */
 13166	        } else {
 13167	                r = ERR;                /* any other error */
 13168	        }
 13169	  }
 13170	  w_wn->w_status |= STATUS_ADMBSY;      /* assume still busy with I/O */
 13171	  return(r);
 13172	}
	
 13174	/*===========================================================================*
 13175	 *                              w_waitfor                                    *
 13176	 *===========================================================================*/
 13177	PRIVATE int w_waitfor(mask, value)
 13178	int mask;                       /* status mask */
 13179	int value;                      /* required status */
 13180	{
 13181	/* Wait until controller is in the required state.  Return zero on timeout.
 13182	 * An alarm that set a timeout flag is used. TIMEOUT is in micros, we need
 13183	 * ticks. Disabling the alarm is not needed, because a static flag is used
 13184	 * and a leftover timeout cannot do any harm.
 13185	 */
 13186	  clock_t t0, t1;
 13187	  int s;
 13188	  getuptime(&t0);
 13189	  do {
 13190	        if ((s=sys_inb(w_wn->base_cmd + REG_STATUS, &w_wn->w_status)) != OK)
 13191	                panic(w_name(),"Couldn't read register",s);

_________________________ Page 806    File: drivers/at_wini/at_wini.c ____________________
 13192	        if ((w_wn->w_status & mask) == value) {
 13193	                return 1;
 13194	        }
 13195	  } while ((s=getuptime(&t1)) == OK && (t1-t0) < timeout_ticks );
 13196	  if (OK != s) printf("AT_WINI: warning, get_uptime failed: %d\n",s);
 13197	
 13198	  w_need_reset();                       /* controller gone deaf */
 13199	  return(0);
 13200	}
	
 13202	/*===========================================================================*
 13203	 *                              w_geometry                                   *
 13204	 *===========================================================================*/
 13205	PRIVATE void w_geometry(entry)
 13206	struct partition *entry;
 13207	{
 13208	  struct wini *wn = w_wn;
 13209	
 13210	  if (wn->state & ATAPI) {              /* Make up some numbers. */
 13211	        entry->cylinders = div64u(wn->part[0].dv_size, SECTOR_SIZE) / (64*32);
 13212	        entry->heads = 64;
 13213	        entry->sectors = 32;
 13214	  } else {                              /* Return logical geometry. */
 13215	        entry->cylinders = wn->lcylinders;
 13216	        entry->heads = wn->lheads;
 13217	        entry->sectors = wn->lsectors;
 13218	  }
 13219	}
	
 13221	/*===========================================================================*
 13222	 *                              w_other                                      *
 13223	 *===========================================================================*/
 13224	PRIVATE int w_other(dr, m)
 13225	struct driver *dr;
 13226	message *m;
 13227	{
 13228	        int r, timeout, prev;
 13229	
 13230	        if (m->m_type != DEV_IOCTL ) {
 13231	                return EINVAL;
 13232	        }
 13233	
 13234	        if (m->REQUEST == DIOCTIMEOUT) {
 13235	                if ((r=sys_datacopy(m->PROC_NR, (vir_bytes)m->ADDRESS,
 13236	                        SELF, (vir_bytes)&timeout, sizeof(timeout))) != OK)
 13237	                        return r;
 13238	        
 13239	                if (timeout == 0) {
 13240	                        /* Restore defaults. */
 13241	                        timeout_ticks = DEF_TIMEOUT_TICKS;
 13242	                        max_errors = MAX_ERRORS;
 13243	                        wakeup_ticks = WAKEUP;
 13244	                        w_silent = 0;
 13245	                } else if (timeout < 0) {
 13246	                        return EINVAL;
 13247	                } else  {
 13248	                        prev = wakeup_ticks;
 13249	        
 13250	                        if (!w_standard_timeouts) {
 13251	                                /* Set (lower) timeout, lower error

_________________________ Page 807    File: drivers/at_wini/at_wini.c ____________________
 13252	                                 * tolerance and set silent mode.
 13253	                                 */
 13254	                                wakeup_ticks = timeout;
 13255	                                max_errors = 3;
 13256	                                w_silent = 1;
 13257	        
 13258	                                if (timeout_ticks > timeout)
 13259	                                        timeout_ticks = timeout;
 13260	                        }
 13261	        
 13262	                        if ((r=sys_datacopy(SELF, (vir_bytes)&prev, 
 13263	                                m->PROC_NR, (vir_bytes)m->ADDRESS, sizeof(prev))) !
 13264	                                return r;
 13265	                }
 13266	        
 13267	                return OK;
 13268	        } else  if (m->REQUEST == DIOCOPENCT) {
 13269	                int count;
 13270	                if (w_prepare(m->DEVICE) == NIL_DEV) return ENXIO;
 13271	                count = w_wn->open_ct;
 13272	                if ((r=sys_datacopy(SELF, (vir_bytes)&count, 
 13273	                        m->PROC_NR, (vir_bytes)m->ADDRESS, sizeof(count))) != OK)
 13274	                        return r;
 13275	                return OK;
 13276	        }
 13277	        return EINVAL;
 13278	}
	
 13280	/*===========================================================================*
 13281	 *                              w_hw_int                                     *
 13282	 *===========================================================================*/
 13283	PRIVATE int w_hw_int(dr, m)
 13284	struct driver *dr;
 13285	message *m;
 13286	{
 13287	  /* Leftover interrupt(s) received; ack it/them. */
 13288	  ack_irqs(m->NOTIFY_ARG);
 13289	
 13290	  return OK;
 13291	}
	
	
 13294	/*===========================================================================*
 13295	 *                              ack_irqs                                     *
 13296	 *===========================================================================*/
 13297	PRIVATE void ack_irqs(unsigned int irqs)
 13298	{
 13299	  unsigned int drive;
 13300	  for (drive = 0; drive < MAX_DRIVES && irqs; drive++) {
 13301	        if (!(wini[drive].state & IGNORING) && wini[drive].irq_need_ack
 13302	                (wini[drive].irq_mask & irqs)) {
 13303	                if (sys_inb((wini[drive].base_cmd + REG_STATUS), &wini[drive].w
 13304	                        printf("couldn't ack irq on drive %d\n", drive);
 13305	                if (sys_irqenable(&wini[drive].irq_hook_id) != OK)
 13306	                        printf("couldn't re-enable drive %d\n", drive);
 13307	                irqs &= ~wini[drive].irq_mask;
 13308	        }
 13309	  }
 13310	}
	

_________________________ Page 808    File: drivers/at_wini/at_wini.c ____________________
	
 13313	#define STSTR(a) if (status & STATUS_ ## a) { strcat(str, #a); strcat(str, " ")
 13314	#define ERRSTR(a) if (e & ERROR_ ## a) { strcat(str, #a); strcat(str, " "); }
 13315	char *strstatus(int status)
 13316	{
 13317	        static char str[200];
 13318	        str[0] = '\0';
 13319	
 13320	        STSTR(BSY);
 13321	        STSTR(DRDY);
 13322	        STSTR(DMADF);
 13323	        STSTR(SRVCDSC);
 13324	        STSTR(DRQ);
 13325	        STSTR(CORR);
 13326	        STSTR(CHECK);
 13327	        return str;
 13328	}
	
 13330	char *strerr(int e)
 13331	{
 13332	        static char str[200];
 13333	        str[0] = '\0';
 13334	
 13335	        ERRSTR(BB);
 13336	        ERRSTR(ECC);
 13337	        ERRSTR(ID);
 13338	        ERRSTR(AC);
 13339	        ERRSTR(TK);
 13340	        ERRSTR(DM);
 13341	
 13342	        return str;
 13343	}




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/tty/tty.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 13400	/*      tty.h - Terminals       */
 13401	
 13402	#include <timers.h>
 13403	
 13404	/* First minor numbers for the various classes of TTY devices. */
 13405	#define CONS_MINOR         0
 13406	#define LOG_MINOR         15
 13407	#define RS232_MINOR       16
 13408	#define TTYPX_MINOR      128
 13409	#define PTYPX_MINOR      192
 13410	
 13411	#define LINEWRAP           1    /* console.c - wrap lines at column 80 */
 13412	
 13413	#define TTY_IN_BYTES     256    /* tty input queue size */
 13414	#define TAB_SIZE           8    /* distance between tab stops */
 13415	#define TAB_MASK           7    /* mask to compute a tab stop position */
 13416	
 13417	#define ESC             '\33'   /* escape */
 13418	
 13419	#define O_NOCTTY       00400    /* from <fcntl.h>, or cc will choke */

_________________________ Page 809    File: drivers/tty/tty.h _________________________
 13420	#define O_NONBLOCK     04000
 13421	
 13422	struct tty;
 13423	typedef _PROTOTYPE( int (*devfun_t), (struct tty *tp, int try_only) );
 13424	typedef _PROTOTYPE( void (*devfunarg_t), (struct tty *tp, int c) );
 13425	
 13426	typedef struct tty {
 13427	  int tty_events;               /* set when TTY should inspect this line */
 13428	  int tty_index;                /* index into TTY table */
 13429	  int tty_minor;                /* device minor number */
 13430	
 13431	  /* Input queue.  Typed characters are stored here until read by a program. */
 13432	  u16_t *tty_inhead;            /* pointer to place where next char goes */
 13433	  u16_t *tty_intail;            /* pointer to next char to be given to prog */
 13434	  int tty_incount;              /* # chars in the input queue */
 13435	  int tty_eotct;                /* number of "line breaks" in input queue */
 13436	  devfun_t tty_devread;         /* routine to read from low level buffers */
 13437	  devfun_t tty_icancel;         /* cancel any device input */
 13438	  int tty_min;                  /* minimum requested #chars in input queue */
 13439	  timer_t tty_tmr;              /* the timer for this tty */
 13440	
 13441	  /* Output section. */
 13442	  devfun_t tty_devwrite;        /* routine to start actual device output */
 13443	  devfunarg_t tty_echo;         /* routine to echo characters input */
 13444	  devfun_t tty_ocancel;         /* cancel any ongoing device output */
 13445	  devfun_t tty_break;           /* let the device send a break */
 13446	
 13447	  /* Terminal parameters and status. */
 13448	  int tty_position;             /* current position on the screen for echoing */
 13449	  char tty_reprint;             /* 1 when echoed input messed up, else 0 */
 13450	  char tty_escaped;             /* 1 when LNEXT (^V) just seen, else 0 */
 13451	  char tty_inhibited;           /* 1 when STOP (^S) just seen (stops output) */
 13452	  char tty_pgrp;                /* slot number of controlling process */
 13453	  char tty_openct;              /* count of number of opens of this tty */
 13454	
 13455	  /* Information about incomplete I/O requests is stored here. */
 13456	  char tty_inrepcode;           /* reply code, TASK_REPLY or REVIVE */
 13457	  char tty_inrevived;           /* set to 1 if revive callback is pending */
 13458	  char tty_incaller;            /* process that made the call (usually FS) */
 13459	  char tty_inproc;              /* process that wants to read from tty */
 13460	  vir_bytes tty_in_vir;         /* virtual address where data is to go */
 13461	  int tty_inleft;               /* how many chars are still needed */
 13462	  int tty_incum;                /* # chars input so far */
 13463	  char tty_outrepcode;          /* reply code, TASK_REPLY or REVIVE */
 13464	  char tty_outrevived;          /* set to 1 if revive callback is pending */
 13465	  char tty_outcaller;           /* process that made the call (usually FS) */
 13466	  char tty_outproc;             /* process that wants to write to tty */
 13467	  vir_bytes tty_out_vir;        /* virtual address where data comes from */
 13468	  int tty_outleft;              /* # chars yet to be output */
 13469	  int tty_outcum;               /* # chars output so far */
 13470	  char tty_iocaller;            /* process that made the call (usually FS) */
 13471	  char tty_ioproc;              /* process that wants to do an ioctl */
 13472	  int tty_ioreq;                /* ioctl request code */
 13473	  vir_bytes tty_iovir;          /* virtual address of ioctl buffer */
 13474	
 13475	  /* select() data */
 13476	  int tty_select_ops;           /* which operations are interesting */
 13477	  int tty_select_proc;          /* which process wants notification */
 13478	
 13479	  /* Miscellaneous. */

_________________________ Page 810    File: drivers/tty/tty.h _________________________
 13480	  devfun_t tty_ioctl;           /* set line speed, etc. at the device level */
 13481	  devfun_t tty_close;           /* tell the device that the tty is closed */
 13482	  void *tty_priv;               /* pointer to per device private data */
 13483	  struct termios tty_termios;   /* terminal attributes */
 13484	  struct winsize tty_winsize;   /* window size (#lines and #columns) */
 13485	
 13486	  u16_t tty_inbuf[TTY_IN_BYTES];/* tty input buffer */
 13487	
 13488	} tty_t;
 13489	
 13490	/* Memory allocated in tty.c, so extern here. */
 13491	extern tty_t tty_table[NR_CONS+NR_RS_LINES+NR_PTYS];
 13492	extern int ccurrent;            /* currently visible console */
 13493	extern int irq_hook_id;         /* hook id for keyboard irq */
 13494	
 13495	extern unsigned long kbd_irq_set;
 13496	extern unsigned long rs_irq_set;
 13497	
 13498	/* Values for the fields. */
 13499	#define NOT_ESCAPED        0    /* previous character is not LNEXT (^V) */
 13500	#define ESCAPED            1    /* previous character was LNEXT (^V) */
 13501	#define RUNNING            0    /* no STOP (^S) has been typed to stop output */
 13502	#define STOPPED            1    /* STOP (^S) has been typed to stop output */
 13503	
 13504	/* Fields and flags on characters in the input queue. */
 13505	#define IN_CHAR       0x00FF    /* low 8 bits are the character itself */
 13506	#define IN_LEN        0x0F00    /* length of char if it has been echoed */
 13507	#define IN_LSHIFT          8    /* length = (c & IN_LEN) >> IN_LSHIFT */
 13508	#define IN_EOT        0x1000    /* char is a line break (^D, LF) */
 13509	#define IN_EOF        0x2000    /* char is EOF (^D), do not return to user */
 13510	#define IN_ESC        0x4000    /* escaped by LNEXT (^V), no interpretation */
 13511	
 13512	/* Times and timeouts. */
 13513	#define force_timeout() ((void) (0))
 13514	
 13515	/* Memory allocated in tty.c, so extern here. */
 13516	extern timer_t *tty_timers;             /* queue of TTY timers */
 13517	extern clock_t tty_next_timeout;        /* next TTY timeout */
 13518	
 13519	/* Number of elements and limit of a buffer. */
 13520	#define buflen(buf)     (sizeof(buf) / sizeof((buf)[0]))
 13521	#define bufend(buf)     ((buf) + buflen(buf))
 13522	
 13523	/* Memory allocated in tty.c, so extern here. */
 13524	extern struct machine machine;  /* machine information (a.o.: pc_at, ega) */
 13525	
 13526	/* Function prototypes for TTY driver. */
 13527	/* tty.c */
 13528	_PROTOTYPE( void handle_events, (struct tty *tp)                        );
 13529	_PROTOTYPE( void sigchar, (struct tty *tp, int sig)                     );
 13530	_PROTOTYPE( void tty_task, (void)                                       );
 13531	_PROTOTYPE( int in_process, (struct tty *tp, char *buf, int count)      );
 13532	_PROTOTYPE( void out_process, (struct tty *tp, char *bstart, char *bpos,
 13533	                                char *bend, int *icount, int *ocount)   );
 13534	_PROTOTYPE( void tty_wakeup, (clock_t now)                              );
 13535	_PROTOTYPE( void tty_reply, (int code, int replyee, int proc_nr,
 13536	                                                        int status)     );
 13537	_PROTOTYPE( int tty_devnop, (struct tty *tp, int try)                   );
 13538	_PROTOTYPE( int select_try, (struct tty *tp, int ops)                   );
 13539	_PROTOTYPE( int select_retry, (struct tty *tp)                          );

_________________________ Page 811    File: drivers/tty/tty.h _________________________
 13540	
 13541	/* console.c */
 13542	_PROTOTYPE( void kputc, (int c)                                         );
 13543	_PROTOTYPE( void cons_stop, (void)                                      );
 13544	_PROTOTYPE( void do_new_kmess, (message *m)                             );
 13545	_PROTOTYPE( void do_diagnostics, (message *m)                           );
 13546	_PROTOTYPE( void scr_init, (struct tty *tp)                             );
 13547	_PROTOTYPE( void toggle_scroll, (void)                                  );
 13548	_PROTOTYPE( int con_loadfont, (message *m)                              );
 13549	_PROTOTYPE( void select_console, (int cons_line)                        );
 13550	
 13551	/* keyboard.c */
 13552	_PROTOTYPE( void kb_init, (struct tty *tp)                              );
 13553	_PROTOTYPE( void kb_init_once, (void)                                   );
 13554	_PROTOTYPE( int kbd_loadmap, (message *m)                               );
 13555	_PROTOTYPE( void do_panic_dumps, (message *m)                           );
 13556	_PROTOTYPE( void do_fkey_ctl, (message *m)                              );
 13557	_PROTOTYPE( void kbd_interrupt, (message *m)                            );
 13558	
 13559	/* vidcopy.s */
 13560	_PROTOTYPE( void vid_vid_copy, (unsigned src, unsigned dst, unsigned count));
 13561	_PROTOTYPE( void mem_vid_copy, (u16_t *src, unsigned dst, unsigned count));




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/tty/tty.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 13600	/* This file contains the terminal driver, both for the IBM console and regular
 13601	 * ASCII terminals.  It handles only the device-independent part of a TTY, the
 13602	 * device dependent parts are in console.c, rs232.c, etc.  This file contains
 13603	 * two main entry points, tty_task() and tty_wakeup(), and several minor entry
 13604	 * points for use by the device-dependent code.
 13605	 *
 13606	 * The device-independent part accepts "keyboard" input from the device-
 13607	 * dependent part, performs input processing (special key interpretation),
 13608	 * and sends the input to a process reading from the TTY.  Output to a TTY
 13609	 * is sent to the device-dependent code for output processing and "screen"
 13610	 * display.  Input processing is done by the device by calling 'in_process'
 13611	 * on the input characters, output processing may be done by the device itself
 13612	 * or by calling 'out_process'.  The TTY takes care of input queuing, the
 13613	 * device does the output queuing.  If a device receives an external signal,
 13614	 * like an interrupt, then it causes tty_wakeup() to be run by the CLOCK task
 13615	 * to, you guessed it, wake up the TTY to check if input or output can
 13616	 * continue.
 13617	 *
 13618	 * The valid messages and their parameters are:
 13619	 *
 13620	 *   HARD_INT:       output has been completed or input has arrived
 13621	 *   SYS_SIG:      e.g., MINIX wants to shutdown; run code to cleanly stop
 13622	 *   DEV_READ:       a process wants to read from a terminal
 13623	 *   DEV_WRITE:      a process wants to write on a terminal
 13624	 *   DEV_IOCTL:      a process wants to change a terminal's parameters
 13625	 *   DEV_OPEN:       a tty line has been opened
 13626	 *   DEV_CLOSE:      a tty line has been closed
 13627	 *   DEV_SELECT:     start select notification request
 13628	 *   DEV_STATUS:     FS wants to know status for SELECT or REVIVE
 13629	 *   CANCEL:         terminate a previous incomplete system call immediately

_________________________ Page 812    File: drivers/tty/tty.c _________________________
 13630	 *
 13631	 *    m_type      TTY_LINE   PROC_NR    COUNT   TTY_SPEK  TTY_FLAGS  ADDRESS
 13632	 * ---------------------------------------------------------------------------
 13633	 * | HARD_INT    |         |         |         |         |         |         |
 13634	 * |-------------+---------+---------+---------+---------+---------+---------|
 13635	 * | SYS_SIG     | sig set |         |         |         |         |         |
 13636	 * |-------------+---------+---------+---------+---------+---------+---------|
 13637	 * | DEV_READ    |minor dev| proc nr |  count  |         O_NONBLOCK| buf ptr |
 13638	 * |-------------+---------+---------+---------+---------+---------+---------|
 13639	 * | DEV_WRITE   |minor dev| proc nr |  count  |         |         | buf ptr |
 13640	 * |-------------+---------+---------+---------+---------+---------+---------|
 13641	 * | DEV_IOCTL   |minor dev| proc nr |func code|erase etc|  flags  |         |
 13642	 * |-------------+---------+---------+---------+---------+---------+---------|
 13643	 * | DEV_OPEN    |minor dev| proc nr | O_NOCTTY|         |         |         |
 13644	 * |-------------+---------+---------+---------+---------+---------+---------|
 13645	 * | DEV_CLOSE   |minor dev| proc nr |         |         |         |         |
 13646	 * |-------------+---------+---------+---------+---------+---------+---------|
 13647	 * | DEV_STATUS  |         |         |         |         |         |         |
 13648	 * |-------------+---------+---------+---------+---------+---------+---------|
 13649	 * | CANCEL      |minor dev| proc nr |         |         |         |         |
 13650	 * ---------------------------------------------------------------------------
 13651	 *
 13652	 * Changes:
 13653	 *   Jan 20, 2004   moved TTY driver to user-space  (Jorrit N. Herder)
 13654	 *   Sep 20, 2004   local timer management/ sync alarms  (Jorrit N. Herder)
 13655	 *   Jul 13, 2004   support for function key observers  (Jorrit N. Herder)  
 13656	 */
 13657	
 13658	#include "../drivers.h"
 13659	#include "../drivers.h"
 13660	#include <termios.h>
 13661	#include <sys/ioc_tty.h>
 13662	#include <signal.h>
 13663	#include <minix/callnr.h>
 13664	#include <minix/keymap.h>
 13665	#include "tty.h"
 13666	
 13667	#include <sys/time.h>
 13668	#include <sys/select.h>
 13669	
 13670	extern int irq_hook_id;
 13671	
 13672	unsigned long kbd_irq_set = 0;
 13673	unsigned long rs_irq_set = 0;
 13674	
 13675	/* Address of a tty structure. */
 13676	#define tty_addr(line)  (&tty_table[line])
 13677	
 13678	/* Macros for magic tty types. */
 13679	#define isconsole(tp)   ((tp) < tty_addr(NR_CONS))
 13680	#define ispty(tp)       ((tp) >= tty_addr(NR_CONS+NR_RS_LINES))
 13681	
 13682	/* Macros for magic tty structure pointers. */
 13683	#define FIRST_TTY       tty_addr(0)
 13684	#define END_TTY         tty_addr(sizeof(tty_table) / sizeof(tty_table[0]))
 13685	
 13686	/* A device exists if at least its 'devread' function is defined. */
 13687	#define tty_active(tp)  ((tp)->tty_devread != NULL)
 13688	
 13689	/* RS232 lines or pseudo terminals can be completely configured out. */

_________________________ Page 813    File: drivers/tty/tty.c _________________________
 13690	#if NR_RS_LINES == 0
 13691	#define rs_init(tp)     ((void) 0)
 13692	#endif
 13693	#if NR_PTYS == 0
 13694	#define pty_init(tp)    ((void) 0)
 13695	#define do_pty(tp, mp)  ((void) 0)
 13696	#endif
 13697	
 13698	FORWARD _PROTOTYPE( void tty_timed_out, (timer_t *tp)                   );
 13699	FORWARD _PROTOTYPE( void expire_timers, (void)                          );
 13700	FORWARD _PROTOTYPE( void settimer, (tty_t *tty_ptr, int enable)         );
 13701	FORWARD _PROTOTYPE( void do_cancel, (tty_t *tp, message *m_ptr)         );
 13702	FORWARD _PROTOTYPE( void do_ioctl, (tty_t *tp, message *m_ptr)          );
 13703	FORWARD _PROTOTYPE( void do_open, (tty_t *tp, message *m_ptr)           );
 13704	FORWARD _PROTOTYPE( void do_close, (tty_t *tp, message *m_ptr)          );
 13705	FORWARD _PROTOTYPE( void do_read, (tty_t *tp, message *m_ptr)           );
 13706	FORWARD _PROTOTYPE( void do_write, (tty_t *tp, message *m_ptr)          );
 13707	FORWARD _PROTOTYPE( void do_select, (tty_t *tp, message *m_ptr)         );
 13708	FORWARD _PROTOTYPE( void do_status, (message *m_ptr)                    );
 13709	FORWARD _PROTOTYPE( void in_transfer, (tty_t *tp)                       );
 13710	FORWARD _PROTOTYPE( int tty_echo, (tty_t *tp, int ch)                   );
 13711	FORWARD _PROTOTYPE( void rawecho, (tty_t *tp, int ch)                   );
 13712	FORWARD _PROTOTYPE( int back_over, (tty_t *tp)                          );
 13713	FORWARD _PROTOTYPE( void reprint, (tty_t *tp)                           );
 13714	FORWARD _PROTOTYPE( void dev_ioctl, (tty_t *tp)                         );
 13715	FORWARD _PROTOTYPE( void setattr, (tty_t *tp)                           );
 13716	FORWARD _PROTOTYPE( void tty_icancel, (tty_t *tp)                       );
 13717	FORWARD _PROTOTYPE( void tty_init, (void)                               );
 13718	
 13719	/* Default attributes. */
 13720	PRIVATE struct termios termios_defaults = {
 13721	  TINPUT_DEF, TOUTPUT_DEF, TCTRL_DEF, TLOCAL_DEF, TSPEED_DEF, TSPEED_DEF,
 13722	  {
 13723	        TEOF_DEF, TEOL_DEF, TERASE_DEF, TINTR_DEF, TKILL_DEF, TMIN_DEF,
 13724	        TQUIT_DEF, TTIME_DEF, TSUSP_DEF, TSTART_DEF, TSTOP_DEF,
 13725	        TREPRINT_DEF, TLNEXT_DEF, TDISCARD_DEF,
 13726	  },
 13727	};
 13728	PRIVATE struct winsize winsize_defaults;        /* = all zeroes */
 13729	
 13730	/* Global variables for the TTY task (declared extern in tty.h). */
 13731	PUBLIC tty_t tty_table[NR_CONS+NR_RS_LINES+NR_PTYS];
 13732	PUBLIC int ccurrent;                    /* currently active console */
 13733	PUBLIC timer_t *tty_timers;             /* queue of TTY timers */
 13734	PUBLIC clock_t tty_next_timeout;        /* time that the next alarm is due */
 13735	PUBLIC struct machine machine;          /* kernel environment variables */
 13736	
 13737	/*===========================================================================*
 13738	 *                              tty_task                                     *
 13739	 *===========================================================================*/
 13740	PUBLIC void main(void)
 13741	{
 13742	/* Main routine of the terminal task. */
 13743	
 13744	  message tty_mess;             /* buffer for all incoming messages */
 13745	  unsigned line;
 13746	  int s;
 13747	  char *types[] = {"task","driver","server", "user"};
 13748	  register struct proc *rp;
 13749	  register tty_t *tp;

_________________________ Page 814    File: drivers/tty/tty.c _________________________
 13750	
 13751	  /* Initialize the TTY driver. */
 13752	  tty_init();
 13753	
 13754	  /* Get kernel environment (protected_mode, pc_at and ega are needed). */ 
 13755	  if (OK != (s=sys_getmachine(&machine))) {
 13756	    panic("TTY","Couldn't obtain kernel environment.", s);
 13757	  }
 13758	
 13759	  /* Final one-time keyboard initialization. */
 13760	  kb_init_once();
 13761	
 13762	  printf("\n");
 13763	
 13764	  while (TRUE) {
 13765	
 13766	        /* Check for and handle any events on any of the ttys. */
 13767	        for (tp = FIRST_TTY; tp < END_TTY; tp++) {
 13768	                if (tp->tty_events) handle_events(tp);
 13769	        }
 13770	
 13771	        /* Get a request message. */
 13772	        receive(ANY, &tty_mess);
 13773	
 13774	        /* First handle all kernel notification types that the TTY supports. 
 13775	         *  - An alarm went off, expire all timers and handle the events. 
 13776	         *  - A hardware interrupt also is an invitation to check for events. 
 13777	         *  - A new kernel message is available for printing.
 13778	         *  - Reset the console on system shutdown. 
 13779	         * Then see if this message is different from a normal device driver
 13780	         * request and should be handled separately. These extra functions
 13781	         * do not operate on a device, in constrast to the driver requests. 
 13782	         */
 13783	        switch (tty_mess.m_type) { 
 13784	        case SYN_ALARM:                 /* fall through */
 13785	                expire_timers();        /* run watchdogs of expired timers */
 13786	                continue;               /* contine to check for events */
 13787	        case HARD_INT: {                /* hardware interrupt notification */
 13788	                if (tty_mess.NOTIFY_ARG & kbd_irq_set)
 13789	                        kbd_interrupt(&tty_mess);/* fetch chars from keyboard *
 13790	#if NR_RS_LINES > 0
 13791	                if (tty_mess.NOTIFY_ARG & rs_irq_set)
 13792	                        rs_interrupt(&tty_mess);/* serial I/O */
 13793	#endif
 13794	                expire_timers();        /* run watchdogs of expired timers */
 13795	                continue;               /* contine to check for events */
 13796	        }
 13797	        case SYS_SIG: {                 /* system signal */
 13798	                sigset_t sigset = (sigset_t) tty_mess.NOTIFY_ARG;
 13799	
 13800	                if (sigismember(&sigset, SIGKSTOP)) {
 13801	                        cons_stop();            /* switch to primary console */
 13802	                        if (irq_hook_id != -1) {
 13803	                                sys_irqdisable(&irq_hook_id);
 13804	                                sys_irqrmpolicy(KEYBOARD_IRQ, &irq_hook_id);
 13805	                        }
 13806	                } 
 13807	                if (sigismember(&sigset, SIGTERM)) cons_stop(); 
 13808	                if (sigismember(&sigset, SIGKMESS)) do_new_kmess(&tty_mess)
 13809	                continue;

_________________________ Page 815    File: drivers/tty/tty.c _________________________
 13810	        }
 13811	        case PANIC_DUMPS:               /* allow panic dumps */
 13812	                cons_stop();            /* switch to primary console */
 13813	                do_panic_dumps(&tty_mess);      
 13814	                continue;
 13815	        case DIAGNOSTICS:               /* a server wants to print some */
 13816	                do_diagnostics(&tty_mess);
 13817	                continue;
 13818	        case FKEY_CONTROL:              /* (un)register a fkey observer */
 13819	                do_fkey_ctl(&tty_mess);
 13820	                continue;
 13821	        default:                        /* should be a driver request */
 13822	                ;                       /* do nothing; end switch */
 13823	        }
 13824	
 13825	        /* Only device requests should get to this point. All requests, 
 13826	         * except DEV_STATUS, have a minor device number. Check this
 13827	         * exception and get the minor device number otherwise.
 13828	         */
 13829	        if (tty_mess.m_type == DEV_STATUS) {
 13830	                do_status(&tty_mess);
 13831	                continue;
 13832	        }
 13833	        line = tty_mess.TTY_LINE;
 13834	        if ((line - CONS_MINOR) < NR_CONS) {
 13835	                tp = tty_addr(line - CONS_MINOR);
 13836	        } else if (line == LOG_MINOR) {
 13837	                tp = tty_addr(0);
 13838	        } else if ((line - RS232_MINOR) < NR_RS_LINES) {
 13839	                tp = tty_addr(line - RS232_MINOR + NR_CONS);
 13840	        } else if ((line - TTYPX_MINOR) < NR_PTYS) {
 13841	                tp = tty_addr(line - TTYPX_MINOR + NR_CONS + NR_RS_LINES);
 13842	        } else if ((line - PTYPX_MINOR) < NR_PTYS) {
 13843	                tp = tty_addr(line - PTYPX_MINOR + NR_CONS + NR_RS_LINES);
 13844	                if (tty_mess.m_type != DEV_IOCTL) {
 13845	                        do_pty(tp, &tty_mess);
 13846	                        continue;
 13847	                }
 13848	        } else {
 13849	                tp = NULL;
 13850	        }
 13851	
 13852	        /* If the device doesn't exist or is not configured return ENXIO. */
 13853	        if (tp == NULL || ! tty_active(tp)) {
 13854	                printf("Warning, TTY got illegal request %d from %d\n",
 13855	                        tty_mess.m_type, tty_mess.m_source);
 13856	                tty_reply(TASK_REPLY, tty_mess.m_source,
 13857	                                                tty_mess.PROC_NR, ENXIO);
 13858	                continue;
 13859	        }
 13860	
 13861	        /* Execute the requested device driver function. */
 13862	        switch (tty_mess.m_type) {
 13863	            case DEV_READ:       do_read(tp, &tty_mess);          break;
 13864	            case DEV_WRITE:      do_write(tp, &tty_mess);         break;
 13865	            case DEV_IOCTL:      do_ioctl(tp, &tty_mess);         break;
 13866	            case DEV_OPEN:       do_open(tp, &tty_mess);          break;
 13867	            case DEV_CLOSE:      do_close(tp, &tty_mess);         break;
 13868	            case DEV_SELECT:     do_select(tp, &tty_mess);        break;
 13869	            case CANCEL:         do_cancel(tp, &tty_mess);        break;

_________________________ Page 816    File: drivers/tty/tty.c _________________________
 13870	            default:            
 13871	                printf("Warning, TTY got unexpected request %d from %d\n",
 13872	                        tty_mess.m_type, tty_mess.m_source);
 13873	            tty_reply(TASK_REPLY, tty_mess.m_source,
 13874	                                                tty_mess.PROC_NR, EINVAL);
 13875	        }
 13876	  }
 13877	}
	
 13879	/*===========================================================================*
 13880	 *                              do_status                                    *
 13881	 *===========================================================================*/
 13882	PRIVATE void do_status(m_ptr)
 13883	message *m_ptr;
 13884	{
 13885	  register struct tty *tp;
 13886	  int event_found;
 13887	  int status;
 13888	  int ops;
 13889	  
 13890	  /* Check for select or revive events on any of the ttys. If we found an, 
 13891	   * event return a single status message for it. The FS will make another 
 13892	   * call to see if there is more.
 13893	   */
 13894	  event_found = 0;
 13895	  for (tp = FIRST_TTY; tp < END_TTY; tp++) {
 13896	        if ((ops = select_try(tp, tp->tty_select_ops)) && 
 13897	                        tp->tty_select_proc == m_ptr->m_source) {
 13898	
 13899	                /* I/O for a selected minor device is ready. */
 13900	                m_ptr->m_type = DEV_IO_READY;
 13901	                m_ptr->DEV_MINOR = tp->tty_index;
 13902	                m_ptr->DEV_SEL_OPS = ops;
 13903	
 13904	                tp->tty_select_ops &= ~ops;     /* unmark select event */
 13905	                event_found = 1;
 13906	                break;
 13907	        }
 13908	        else if (tp->tty_inrevived && tp->tty_incaller == m_ptr->m_source) 
 13909	                
 13910	                /* Suspended request finished. Send a REVIVE. */
 13911	                m_ptr->m_type = DEV_REVIVE;
 13912	                m_ptr->REP_PROC_NR = tp->tty_inproc;
 13913	                m_ptr->REP_STATUS = tp->tty_incum;
 13914	
 13915	                tp->tty_inleft = tp->tty_incum = 0;
 13916	                tp->tty_inrevived = 0;          /* unmark revive event */
 13917	                event_found = 1;
 13918	                break;
 13919	        }
 13920	        else if (tp->tty_outrevived && tp->tty_outcaller == m_ptr->m_source
 13921	                
 13922	                /* Suspended request finished. Send a REVIVE. */
 13923	                m_ptr->m_type = DEV_REVIVE;
 13924	                m_ptr->REP_PROC_NR = tp->tty_outproc;
 13925	                m_ptr->REP_STATUS = tp->tty_outcum;
 13926	
 13927	                tp->tty_outcum = 0;
 13928	                tp->tty_outrevived = 0;         /* unmark revive event */
 13929	                event_found = 1;

_________________________ Page 817    File: drivers/tty/tty.c _________________________
 13930	                break;
 13931	        }
 13932	  }
 13933	
 13934	#if NR_PTYS > 0
 13935	  if (!event_found)
 13936	        event_found = pty_status(m_ptr);
 13937	#endif
 13938	
 13939	  if (! event_found) {
 13940	        /* No events of interest were found. Return an empty message. */
 13941	        m_ptr->m_type = DEV_NO_STATUS;
 13942	  }
 13943	
 13944	  /* Almost done. Send back the reply message to the caller. */
 13945	  if ((status = send(m_ptr->m_source, m_ptr)) != OK) {
 13946	        panic("TTY","send in do_status failed, status\n", status);
 13947	  }
 13948	}
	
 13950	/*===========================================================================*
 13951	 *                              do_read                                      *
 13952	 *===========================================================================*/
 13953	PRIVATE void do_read(tp, m_ptr)
 13954	register tty_t *tp;             /* pointer to tty struct */
 13955	register message *m_ptr;        /* pointer to message sent to the task */
 13956	{
 13957	/* A process wants to read from a terminal. */
 13958	  int r, status;
 13959	  phys_bytes phys_addr;
 13960	
 13961	  /* Check if there is already a process hanging in a read, check if the
 13962	   * parameters are correct, do I/O.
 13963	   */
 13964	  if (tp->tty_inleft > 0) {
 13965	        r = EIO;
 13966	  } else
 13967	  if (m_ptr->COUNT <= 0) {
 13968	        r = EINVAL;
 13969	  } else
 13970	  if (sys_umap(m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS, m_ptr->COUNT,
 13971	                &phys_addr) != OK) {
 13972	        r = EFAULT;
 13973	  } else {
 13974	        /* Copy information from the message to the tty struct. */
 13975	        tp->tty_inrepcode = TASK_REPLY;
 13976	        tp->tty_incaller = m_ptr->m_source;
 13977	        tp->tty_inproc = m_ptr->PROC_NR;
 13978	        tp->tty_in_vir = (vir_bytes) m_ptr->ADDRESS;
 13979	        tp->tty_inleft = m_ptr->COUNT;
 13980	
 13981	        if (!(tp->tty_termios.c_lflag & ICANON)
 13982	                                        && tp->tty_termios.c_cc[VTIME] > 0)
 13983	                if (tp->tty_termios.c_cc[VMIN] == 0) {
 13984	                        /* MIN & TIME specify a read timer that finishes the
 13985	                         * read in TIME/10 seconds if no bytes are available.
 13986	                         */
 13987	                        settimer(tp, TRUE);
 13988	                        tp->tty_min = 1;
 13989	                } else {

_________________________ Page 818    File: drivers/tty/tty.c _________________________
 13990	                        /* MIN & TIME specify an inter-byte timer that may
 13991	                         * have to be cancelled if there are no bytes yet.
 13992	                         */
 13993	                        if (tp->tty_eotct == 0) {
 13994	                                settimer(tp, FALSE);
 13995	                                tp->tty_min = tp->tty_termios.c_cc[VMIN];
 13996	                        }
 13997	                }
 13998	        }
 13999	
 14000	        /* Anything waiting in the input buffer? Clear it out... */
 14001	        in_transfer(tp);
 14002	        /* ...then go back for more. */
 14003	        handle_events(tp);
 14004	        if (tp->tty_inleft == 0)  {
 14005	                if (tp->tty_select_ops)
 14006	                        select_retry(tp);
 14007	                return;                 /* already done */
 14008	        }
 14009	
 14010	        /* There were no bytes in the input queue available, so either suspend
 14011	         * the caller or break off the read if nonblocking.
 14012	         */
 14013	        if (m_ptr->TTY_FLAGS & O_NONBLOCK) {
 14014	                r = EAGAIN;                             /* cancel the read */
 14015	                tp->tty_inleft = tp->tty_incum = 0;
 14016	        } else {
 14017	                r = SUSPEND;                            /* suspend the caller */
 14018	                tp->tty_inrepcode = REVIVE;
 14019	        }
 14020	  }
 14021	  tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, r);
 14022	  if (tp->tty_select_ops)
 14023	        select_retry(tp);
 14024	}
	
 14026	/*===========================================================================*
 14027	 *                              do_write                                     *
 14028	 *===========================================================================*/
 14029	PRIVATE void do_write(tp, m_ptr)
 14030	register tty_t *tp;
 14031	register message *m_ptr;        /* pointer to message sent to the task */
 14032	{
 14033	/* A process wants to write on a terminal. */
 14034	  int r;
 14035	  phys_bytes phys_addr;
 14036	
 14037	  /* Check if there is already a process hanging in a write, check if the
 14038	   * parameters are correct, do I/O.
 14039	   */
 14040	  if (tp->tty_outleft > 0) {
 14041	        r = EIO;
 14042	  } else
 14043	  if (m_ptr->COUNT <= 0) {
 14044	        r = EINVAL;
 14045	  } else
 14046	  if (sys_umap(m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS, m_ptr->COUNT,
 14047	                &phys_addr) != OK) {
 14048	        r = EFAULT;
 14049	  } else {

_________________________ Page 819    File: drivers/tty/tty.c _________________________
 14050	        /* Copy message parameters to the tty structure. */
 14051	        tp->tty_outrepcode = TASK_REPLY;
 14052	        tp->tty_outcaller = m_ptr->m_source;
 14053	        tp->tty_outproc = m_ptr->PROC_NR;
 14054	        tp->tty_out_vir = (vir_bytes) m_ptr->ADDRESS;
 14055	        tp->tty_outleft = m_ptr->COUNT;
 14056	
 14057	        /* Try to write. */
 14058	        handle_events(tp);
 14059	        if (tp->tty_outleft == 0) 
 14060	                return; /* already done */
 14061	
 14062	        /* None or not all the bytes could be written, so either suspend the
 14063	         * caller or break off the write if nonblocking.
 14064	         */
 14065	        if (m_ptr->TTY_FLAGS & O_NONBLOCK) {            /* cancel the write */
 14066	                r = tp->tty_outcum > 0 ? tp->tty_outcum : EAGAIN;
 14067	                tp->tty_outleft = tp->tty_outcum = 0;
 14068	        } else {
 14069	                r = SUSPEND;                            /* suspend the caller */
 14070	                tp->tty_outrepcode = REVIVE;
 14071	        }
 14072	  }
 14073	  tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, r);
 14074	}
	
 14076	/*===========================================================================*
 14077	 *                              do_ioctl                                     *
 14078	 *===========================================================================*/
 14079	PRIVATE void do_ioctl(tp, m_ptr)
 14080	register tty_t *tp;
 14081	message *m_ptr;                 /* pointer to message sent to task */
 14082	{
 14083	/* Perform an IOCTL on this terminal. Posix termios calls are handled
 14084	 * by the IOCTL system call
 14085	 */
 14086	
 14087	  int r;
 14088	  union {
 14089	        int i;
 14090	  } param;
 14091	  size_t size;
 14092	
 14093	  /* Size of the ioctl parameter. */
 14094	  switch (m_ptr->TTY_REQUEST) {
 14095	    case TCGETS:        /* Posix tcgetattr function */
 14096	    case TCSETS:        /* Posix tcsetattr function, TCSANOW option */ 
 14097	    case TCSETSW:       /* Posix tcsetattr function, TCSADRAIN option */
 14098	    case TCSETSF:       /* Posix tcsetattr function, TCSAFLUSH option */
 14099	        size = sizeof(struct termios);
 14100	        break;
 14101	
 14102	    case TCSBRK:        /* Posix tcsendbreak function */
 14103	    case TCFLOW:        /* Posix tcflow function */
 14104	    case TCFLSH:        /* Posix tcflush function */
 14105	    case TIOCGPGRP:     /* Posix tcgetpgrp function */
 14106	    case TIOCSPGRP:     /* Posix tcsetpgrp function */
 14107	        size = sizeof(int);
 14108	        break;
 14109	

_________________________ Page 820    File: drivers/tty/tty.c _________________________
 14110	    case TIOCGWINSZ:    /* get window size (not Posix) */
 14111	    case TIOCSWINSZ:    /* set window size (not Posix) */
 14112	        size = sizeof(struct winsize);
 14113	        break;
 14114	
 14115	    case KIOCSMAP:      /* load keymap (Minix extension) */
 14116	        size = sizeof(keymap_t);
 14117	        break;
 14118	
 14119	    case TIOCSFON:      /* load font (Minix extension) */
 14120	        size = sizeof(u8_t [8192]);
 14121	        break;
 14122	
 14123	    case TCDRAIN:       /* Posix tcdrain function -- no parameter */
 14124	    default:            size = 0;
 14125	  }
 14126	
 14127	  r = OK;
 14128	  switch (m_ptr->TTY_REQUEST) {
 14129	    case TCGETS:
 14130	        /* Get the termios attributes. */
 14131	        r = sys_vircopy(SELF, D, (vir_bytes) &tp->tty_termios,
 14132	                m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS, 
 14133	                (vir_bytes) size);
 14134	        break;
 14135	
 14136	    case TCSETSW:
 14137	    case TCSETSF:
 14138	    case TCDRAIN:
 14139	        if (tp->tty_outleft > 0) {
 14140	                /* Wait for all ongoing output processing to finish. */
 14141	                tp->tty_iocaller = m_ptr->m_source;
 14142	                tp->tty_ioproc = m_ptr->PROC_NR;
 14143	                tp->tty_ioreq = m_ptr->REQUEST;
 14144	                tp->tty_iovir = (vir_bytes) m_ptr->ADDRESS;
 14145	                r = SUSPEND;
 14146	                break;
 14147	        }
 14148	        if (m_ptr->TTY_REQUEST == TCDRAIN) break;
 14149	        if (m_ptr->TTY_REQUEST == TCSETSF) tty_icancel(tp);
 14150	        /*FALL THROUGH*/
 14151	    case TCSETS:
 14152	        /* Set the termios attributes. */
 14153	        r = sys_vircopy( m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS,
 14154	                SELF, D, (vir_bytes) &tp->tty_termios, (vir_bytes) size);
 14155	        if (r != OK) break;
 14156	        setattr(tp);
 14157	        break;
 14158	
 14159	    case TCFLSH:
 14160	        r = sys_vircopy( m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS,
 14161	                SELF, D, (vir_bytes) &param.i, (vir_bytes) size);
 14162	        if (r != OK) break;
 14163	        switch (param.i) {
 14164	            case TCIFLUSH:      tty_icancel(tp);                            break;
 14165	            case TCOFLUSH:      (*tp->tty_ocancel)(tp, 0);                  break;
 14166	            case TCIOFLUSH:     tty_icancel(tp); (*tp->tty_ocancel)(tp, 0); break;
 14167	            default:            r = EINVAL;
 14168	        }
 14169	        break;

_________________________ Page 821    File: drivers/tty/tty.c _________________________
 14170	
 14171	    case TCFLOW:
 14172	        r = sys_vircopy( m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS,
 14173	                SELF, D, (vir_bytes) &param.i, (vir_bytes) size);
 14174	        if (r != OK) break;
 14175	        switch (param.i) {
 14176	            case TCOOFF:
 14177	            case TCOON:
 14178	                tp->tty_inhibited = (param.i == TCOOFF);
 14179	                tp->tty_events = 1;
 14180	                break;
 14181	            case TCIOFF:
 14182	                (*tp->tty_echo)(tp, tp->tty_termios.c_cc[VSTOP]);
 14183	                break;
 14184	            case TCION:
 14185	                (*tp->tty_echo)(tp, tp->tty_termios.c_cc[VSTART]);
 14186	                break;
 14187	            default:
 14188	                r = EINVAL;
 14189	        }
 14190	        break;
 14191	
 14192	    case TCSBRK:
 14193	        if (tp->tty_break != NULL) (*tp->tty_break)(tp,0);
 14194	        break;
 14195	
 14196	    case TIOCGWINSZ:
 14197	        r = sys_vircopy(SELF, D, (vir_bytes) &tp->tty_winsize,
 14198	                m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS, 
 14199	                (vir_bytes) size);
 14200	        break;
 14201	
 14202	    case TIOCSWINSZ:
 14203	        r = sys_vircopy( m_ptr->PROC_NR, D, (vir_bytes) m_ptr->ADDRESS,
 14204	                SELF, D, (vir_bytes) &tp->tty_winsize, (vir_bytes) size);
 14205	        /* SIGWINCH... */
 14206	        break;
 14207	
 14208	    case KIOCSMAP:
 14209	        /* Load a new keymap (only /dev/console). */
 14210	        if (isconsole(tp)) r = kbd_loadmap(m_ptr);
 14211	        break;
 14212	
 14213	    case TIOCSFON:
 14214	        /* Load a font into an EGA or VGA card (hs@hck.hr) */
 14215	        if (isconsole(tp)) r = con_loadfont(m_ptr);
 14216	        break;
 14217	
 14218	/* These Posix functions are allowed to fail if _POSIX_JOB_CONTROL is 
 14219	 * not defined.
 14220	 */
 14221	    case TIOCGPGRP:     
 14222	    case TIOCSPGRP:     
 14223	    default:
 14224	        r = ENOTTY;
 14225	  }
 14226	
 14227	  /* Send the reply. */
 14228	  tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, r);
 14229	}

_________________________ Page 822    File: drivers/tty/tty.c _________________________
	
 14231	/*===========================================================================*
 14232	 *                              do_open                                      *
 14233	 *===========================================================================*/
 14234	PRIVATE void do_open(tp, m_ptr)
 14235	register tty_t *tp;
 14236	message *m_ptr;                 /* pointer to message sent to task */
 14237	{
 14238	/* A tty line has been opened.  Make it the callers controlling tty if
 14239	 * O_NOCTTY is *not* set and it is not the log device.  1 is returned if
 14240	 * the tty is made the controlling tty, otherwise OK or an error code.
 14241	 */
 14242	  int r = OK;
 14243	
 14244	  if (m_ptr->TTY_LINE == LOG_MINOR) {
 14245	        /* The log device is a write-only diagnostics device. */
 14246	        if (m_ptr->COUNT & R_BIT) r = EACCES;
 14247	  } else {
 14248	        if (!(m_ptr->COUNT & O_NOCTTY)) {
 14249	                tp->tty_pgrp = m_ptr->PROC_NR;
 14250	                r = 1;
 14251	        }
 14252	        tp->tty_openct++;
 14253	  }
 14254	  tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, r);
 14255	}
	
 14257	/*===========================================================================*
 14258	 *                              do_close                                     *
 14259	 *===========================================================================*/
 14260	PRIVATE void do_close(tp, m_ptr)
 14261	register tty_t *tp;
 14262	message *m_ptr;                 /* pointer to message sent to task */
 14263	{
 14264	/* A tty line has been closed.  Clean up the line if it is the last close. */
 14265	
 14266	  if (m_ptr->TTY_LINE != LOG_MINOR && --tp->tty_openct == 0) {
 14267	        tp->tty_pgrp = 0;
 14268	        tty_icancel(tp);
 14269	        (*tp->tty_ocancel)(tp, 0);
 14270	        (*tp->tty_close)(tp, 0);
 14271	        tp->tty_termios = termios_defaults;
 14272	        tp->tty_winsize = winsize_defaults;
 14273	        setattr(tp);
 14274	  }
 14275	  tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, OK);
 14276	}
	
 14278	/*===========================================================================*
 14279	 *                              do_cancel                                    *
 14280	 *===========================================================================*/
 14281	PRIVATE void do_cancel(tp, m_ptr)
 14282	register tty_t *tp;
 14283	message *m_ptr;                 /* pointer to message sent to task */
 14284	{
 14285	/* A signal has been sent to a process that is hanging trying to read or write.
 14286	 * The pending read or write must be finished off immediately.
 14287	 */
 14288	
 14289	  int proc_nr;

_________________________ Page 823    File: drivers/tty/tty.c _________________________
 14290	  int mode;
 14291	
 14292	  /* Check the parameters carefully, to avoid cancelling twice. */
 14293	  proc_nr = m_ptr->PROC_NR;
 14294	  mode = m_ptr->COUNT;
 14295	  if ((mode & R_BIT) && tp->tty_inleft != 0 && proc_nr == tp->t
 14296	        /* Process was reading when killed.  Clean up input. */
 14297	        tty_icancel(tp);
 14298	        tp->tty_inleft = tp->tty_incum = 0;
 14299	  }
 14300	  if ((mode & W_BIT) && tp->tty_outleft != 0 && proc_nr == tp->
 14301	        /* Process was writing when killed.  Clean up output. */
 14302	        (*tp->tty_ocancel)(tp, 0);
 14303	        tp->tty_outleft = tp->tty_outcum = 0;
 14304	  }
 14305	  if (tp->tty_ioreq != 0 && proc_nr == tp->tty_ioproc) {
 14306	        /* Process was waiting for output to drain. */
 14307	        tp->tty_ioreq = 0;
 14308	  }
 14309	  tp->tty_events = 1;
 14310	  tty_reply(TASK_REPLY, m_ptr->m_source, proc_nr, EINTR);
 14311	}
	
 14313	PUBLIC int select_try(struct tty *tp, int ops)
 14314	{
 14315	        int ready_ops = 0;
 14316	
 14317	        /* Special case. If line is hung up, no operations will block.
 14318	         * (and it can be seen as an exceptional condition.)
 14319	         */
 14320	        if (tp->tty_termios.c_ospeed == B0) {
 14321	                ready_ops |= ops;
 14322	        }
 14323	
 14324	        if (ops & SEL_RD) {
 14325	                /* will i/o not block on read? */
 14326	                if (tp->tty_inleft > 0) {
 14327	                        ready_ops |= SEL_RD;    /* EIO - no blocking */
 14328	                } else if (tp->tty_incount > 0) {
 14329	                        /* Is a regular read possible? tty_incount
 14330	                         * says there is data. But a read will only succeed
 14331	                         * in canonical mode if a newline has been seen.
 14332	                         */
 14333	                        if (!(tp->tty_termios.c_lflag & ICANON) ||
 14334	                                tp->tty_eotct > 0) {
 14335	                                ready_ops |= SEL_RD;
 14336	                        }
 14337	                }
 14338	        }
 14339	
 14340	        if (ops & SEL_WR)  {
 14341	                if (tp->tty_outleft > 0)  ready_ops |= SEL_WR;
 14342	                else if ((*tp->tty_devwrite)(tp, 1)) ready_ops |= SEL_WR;
 14343	        }
 14344	
 14345	        return ready_ops;
 14346	}
	
 14348	PUBLIC int select_retry(struct tty *tp)
 14349	{

_________________________ Page 824    File: drivers/tty/tty.c _________________________
 14350	        if (select_try(tp, tp->tty_select_ops))
 14351	                notify(tp->tty_select_proc);
 14352	        return OK;
 14353	}
	
 14355	/*===========================================================================*
 14356	 *                              handle_events                                *
 14357	 *===========================================================================*/
 14358	PUBLIC void handle_events(tp)
 14359	tty_t *tp;                      /* TTY to check for events. */
 14360	{
 14361	/* Handle any events pending on a TTY.  These events are usually device
 14362	 * interrupts.
 14363	 *
 14364	 * Two kinds of events are prominent:
 14365	 *      - a character has been received from the console or an RS232 line.
 14366	 *      - an RS232 line has completed a write request (on behalf of a user).
 14367	 * The interrupt handler may delay the interrupt message at its discretion
 14368	 * to avoid swamping the TTY task.  Messages may be overwritten when the
 14369	 * lines are fast or when there are races between different lines, input
 14370	 * and output, because MINIX only provides single buffering for interrupt
 14371	 * messages (in proc.c).  This is handled by explicitly checking each line
 14372	 * for fresh input and completed output on each interrupt.
 14373	 */
 14374	  char *buf;
 14375	  unsigned count;
 14376	  int status;
 14377	
 14378	  do {
 14379	        tp->tty_events = 0;
 14380	
 14381	        /* Read input and perform input processing. */
 14382	        (*tp->tty_devread)(tp, 0);
 14383	
 14384	        /* Perform output processing and write output. */
 14385	        (*tp->tty_devwrite)(tp, 0);
 14386	
 14387	        /* Ioctl waiting for some event? */
 14388	        if (tp->tty_ioreq != 0) dev_ioctl(tp);
 14389	  } while (tp->tty_events);
 14390	
 14391	  /* Transfer characters from the input queue to a waiting process. */
 14392	  in_transfer(tp);
 14393	
 14394	  /* Reply if enough bytes are available. */
 14395	  if (tp->tty_incum >= tp->tty_min && tp->tty_inleft > 0) {
 14396	        if (tp->tty_inrepcode == REVIVE) {
 14397	                notify(tp->tty_incaller);
 14398	                tp->tty_inrevived = 1;
 14399	        } else {
 14400	                tty_reply(tp->tty_inrepcode, tp->tty_incaller, 
 14401	                        tp->tty_inproc, tp->tty_incum);
 14402	                tp->tty_inleft = tp->tty_incum = 0;
 14403	        }
 14404	  }
 14405	  if (tp->tty_select_ops)
 14406	        select_retry(tp);
 14407	#if NR_PTYS > 0
 14408	  if (ispty(tp))
 14409	        select_retry_pty(tp);

_________________________ Page 825    File: drivers/tty/tty.c _________________________
 14410	#endif
 14411	}
	
 14413	/*===========================================================================*
 14414	 *                              in_transfer                                  *
 14415	 *===========================================================================*/
 14416	PRIVATE void in_transfer(tp)
 14417	register tty_t *tp;             /* pointer to terminal to read from */
 14418	{
 14419	/* Transfer bytes from the input queue to a process reading from a terminal. */
 14420	
 14421	  int ch;
 14422	  int count;
 14423	  char buf[64], *bp;
 14424	
 14425	  /* Force read to succeed if the line is hung up, looks like EOF to reader. */
 14426	  if (tp->tty_termios.c_ospeed == B0) tp->tty_min = 0;
 14427	
 14428	  /* Anything to do? */
 14429	  if (tp->tty_inleft == 0 || tp->tty_eotct < tp->tty_min) return;
 14430	
 14431	  bp = buf;
 14432	  while (tp->tty_inleft > 0 && tp->tty_eotct > 0) {
 14433	        ch = *tp->tty_intail;
 14434	
 14435	        if (!(ch & IN_EOF)) {
 14436	                /* One character to be delivered to the user. */
 14437	                *bp = ch & IN_CHAR;
 14438	                tp->tty_inleft--;
 14439	                if (++bp == bufend(buf)) {
 14440	                        /* Temp buffer full, copy to user space. */
 14441	                        sys_vircopy(SELF, D, (vir_bytes) buf, 
 14442	                                tp->tty_inproc, D, tp->tty_in_vir,
 14443	                                (vir_bytes) buflen(buf));
 14444	                        tp->tty_in_vir += buflen(buf);
 14445	                        tp->tty_incum += buflen(buf);
 14446	                        bp = buf;
 14447	                }
 14448	        }
 14449	
 14450	        /* Remove the character from the input queue. */
 14451	        if (++tp->tty_intail == bufend(tp->tty_inbuf))
 14452	                tp->tty_intail = tp->tty_inbuf;
 14453	        tp->tty_incount--;
 14454	        if (ch & IN_EOT) {
 14455	                tp->tty_eotct--;
 14456	                /* Don't read past a line break in canonical mode. */
 14457	                if (tp->tty_termios.c_lflag & ICANON) tp->tty_inleft = 0;
 14458	        }
 14459	  }
 14460	
 14461	  if (bp > buf) {
 14462	        /* Leftover characters in the buffer. */
 14463	        count = bp - buf;
 14464	        sys_vircopy(SELF, D, (vir_bytes) buf, 
 14465	                tp->tty_inproc, D, tp->tty_in_vir, (vir_bytes) count);
 14466	        tp->tty_in_vir += count;
 14467	        tp->tty_incum += count;
 14468	  }
 14469	

_________________________ Page 826    File: drivers/tty/tty.c _________________________
 14470	  /* Usually reply to the reader, possibly even if incum == 0 (EOF). */
 14471	  if (tp->tty_inleft == 0) {
 14472	        if (tp->tty_inrepcode == REVIVE) {
 14473	                notify(tp->tty_incaller);
 14474	                tp->tty_inrevived = 1;
 14475	        } else {
 14476	                tty_reply(tp->tty_inrepcode, tp->tty_incaller, 
 14477	                        tp->tty_inproc, tp->tty_incum);
 14478	                tp->tty_inleft = tp->tty_incum = 0;
 14479	        }
 14480	  }
 14481	}
	
 14483	/*===========================================================================*
 14484	 *                              in_process                                   *
 14485	 *===========================================================================*/
 14486	PUBLIC int in_process(tp, buf, count)
 14487	register tty_t *tp;             /* terminal on which character has arrived */
 14488	char *buf;                      /* buffer with input characters */
 14489	int count;                      /* number of input characters */
 14490	{
 14491	/* Characters have just been typed in.  Process, save, and echo them.  Return
 14492	 * the number of characters processed.
 14493	 */
 14494	
 14495	  int ch, sig, ct;
 14496	  int timeset = FALSE;
 14497	  static unsigned char csize_mask[] = { 0x1F, 0x3F, 0x7F, 0xFF };
 14498	
 14499	  for (ct = 0; ct < count; ct++) {
 14500	        /* Take one character. */
 14501	        ch = *buf++ & BYTE;
 14502	
 14503	        /* Strip to seven bits? */
 14504	        if (tp->tty_termios.c_iflag & ISTRIP) ch &= 0x7F;
 14505	
 14506	        /* Input extensions? */
 14507	        if (tp->tty_termios.c_lflag & IEXTEN) {
 14508	
 14509	                /* Previous character was a character escape? */
 14510	                if (tp->tty_escaped) {
 14511	                        tp->tty_escaped = NOT_ESCAPED;
 14512	                        ch |= IN_ESC;   /* protect character */
 14513	                }
 14514	
 14515	                /* LNEXT (^V) to escape the next character? */
 14516	                if (ch == tp->tty_termios.c_cc[VLNEXT]) {
 14517	                        tp->tty_escaped = ESCAPED;
 14518	                        rawecho(tp, '^');
 14519	                        rawecho(tp, '\b');
 14520	                        continue;       /* do not store the escape */
 14521	                }
 14522	
 14523	                /* REPRINT (^R) to reprint echoed characters? */
 14524	                if (ch == tp->tty_termios.c_cc[VREPRINT]) {
 14525	                        reprint(tp);
 14526	                        continue;
 14527	                }
 14528	        }
 14529	

_________________________ Page 827    File: drivers/tty/tty.c _________________________
 14530	        /* _POSIX_VDISABLE is a normal character value, so better escape it. */
 14531	        if (ch == _POSIX_VDISABLE) ch |= IN_ESC;
 14532	
 14533	        /* Map CR to LF, ignore CR, or map LF to CR. */
 14534	        if (ch == '\r') {
 14535	                if (tp->tty_termios.c_iflag & IGNCR) continue;
 14536	                if (tp->tty_termios.c_iflag & ICRNL) ch = '\n';
 14537	        } else
 14538	        if (ch == '\n') {
 14539	                if (tp->tty_termios.c_iflag & INLCR) ch = '\r';
 14540	        }
 14541	
 14542	        /* Canonical mode? */
 14543	        if (tp->tty_termios.c_lflag & ICANON) {
 14544	
 14545	                /* Erase processing (rub out of last character). */
 14546	                if (ch == tp->tty_termios.c_cc[VERASE]) {
 14547	                        (void) back_over(tp);
 14548	                        if (!(tp->tty_termios.c_lflag & ECHOE)) {
 14549	                                (void) tty_echo(tp, ch);
 14550	                        }
 14551	                        continue;
 14552	                }
 14553	
 14554	                /* Kill processing (remove current line). */
 14555	                if (ch == tp->tty_termios.c_cc[VKILL]) {
 14556	                        while (back_over(tp)) {}
 14557	                        if (!(tp->tty_termios.c_lflag & ECHOE)) {
 14558	                                (void) tty_echo(tp, ch);
 14559	                                if (tp->tty_termios.c_lflag & ECHOK)
 14560	                                        rawecho(tp, '\n');
 14561	                        }
 14562	                        continue;
 14563	                }
 14564	
 14565	                /* EOF (^D) means end-of-file, an invisible "line break". */
 14566	                if (ch == tp->tty_termios.c_cc[VEOF]) ch |= IN_EOT | IN_EOF;
 14567	
 14568	                /* The line may be returned to the user after an LF. */
 14569	                if (ch == '\n') ch |= IN_EOT;
 14570	
 14571	                /* Same thing with EOL, whatever it may be. */
 14572	                if (ch == tp->tty_termios.c_cc[VEOL]) ch |= IN_EOT;
 14573	        }
 14574	
 14575	        /* Start/stop input control? */
 14576	        if (tp->tty_termios.c_iflag & IXON) {
 14577	
 14578	                /* Output stops on STOP (^S). */
 14579	                if (ch == tp->tty_termios.c_cc[VSTOP]) {
 14580	                        tp->tty_inhibited = STOPPED;
 14581	                        tp->tty_events = 1;
 14582	                        continue;
 14583	                }
 14584	
 14585	                /* Output restarts on START (^Q) or any character if IXANY. */
 14586	                if (tp->tty_inhibited) {
 14587	                        if (ch == tp->tty_termios.c_cc[VSTART]
 14588	                                        || (tp->tty_termios.c_iflag & IXANY)) {
 14589	                                tp->tty_inhibited = RUNNING;

_________________________ Page 828    File: drivers/tty/tty.c _________________________
 14590	                                tp->tty_events = 1;
 14591	                                if (ch == tp->tty_termios.c_cc[VSTART])
 14592	                                        continue;
 14593	                        }
 14594	                }
 14595	        }
 14596	
 14597	        if (tp->tty_termios.c_lflag & ISIG) {
 14598	                /* Check for INTR (^?) and QUIT (^\) characters. */
 14599	                if (ch == tp->tty_termios.c_cc[VINTR]
 14600	                                        || ch == tp->tty_termios.c_cc[VQUIT]) {
 14601	                        sig = SIGINT;
 14602	                        if (ch == tp->tty_termios.c_cc[VQUIT]) sig = SIGQUIT;
 14603	                        sigchar(tp, sig);
 14604	                        (void) tty_echo(tp, ch);
 14605	                        continue;
 14606	                }
 14607	        }
 14608	
 14609	        /* Is there space in the input buffer? */
 14610	        if (tp->tty_incount == buflen(tp->tty_inbuf)) {
 14611	                /* No space; discard in canonical mode, keep in raw mode. */
 14612	                if (tp->tty_termios.c_lflag & ICANON) continue;
 14613	                break;
 14614	        }
 14615	
 14616	        if (!(tp->tty_termios.c_lflag & ICANON)) {
 14617	                /* In raw mode all characters are "line breaks". */
 14618	                ch |= IN_EOT;
 14619	
 14620	                /* Start an inter-byte timer? */
 14621	                if (!timeset && tp->tty_termios.c_cc[VMIN] > 0
 14622	                                && tp->tty_termios.c_cc[VTIME] > 0) {
 14623	                        settimer(tp, TRUE);
 14624	                        timeset = TRUE;
 14625	                }
 14626	        }
 14627	
 14628	        /* Perform the intricate function of echoing. */
 14629	        if (tp->tty_termios.c_lflag & (ECHO|ECHONL)) ch = tty_echo(tp, ch);
 14630	
 14631	        /* Save the character in the input queue. */
 14632	        *tp->tty_inhead++ = ch;
 14633	        if (tp->tty_inhead == bufend(tp->tty_inbuf))
 14634	                tp->tty_inhead = tp->tty_inbuf;
 14635	        tp->tty_incount++;
 14636	        if (ch & IN_EOT) tp->tty_eotct++;
 14637	
 14638	        /* Try to finish input if the queue threatens to overflow. */
 14639	        if (tp->tty_incount == buflen(tp->tty_inbuf)) in_transfer(tp);
 14640	  }
 14641	  return ct;
 14642	}
	
 14644	/*===========================================================================*
 14645	 *                              echo                                         *
 14646	 *===========================================================================*/
 14647	PRIVATE int tty_echo(tp, ch)
 14648	register tty_t *tp;             /* terminal on which to echo */
 14649	register int ch;                /* pointer to character to echo */

_________________________ Page 829    File: drivers/tty/tty.c _________________________
 14650	{
 14651	/* Echo the character if echoing is on.  Some control characters are echoed
 14652	 * with their normal effect, other control characters are echoed as "^X",
 14653	 * normal characters are echoed normally.  EOF (^D) is echoed, but immediately
 14654	 * backspaced over.  Return the character with the echoed length added to its
 14655	 * attributes.
 14656	 */
 14657	  int len, rp;
 14658	
 14659	  ch &= ~IN_LEN;
 14660	  if (!(tp->tty_termios.c_lflag & ECHO)) {
 14661	        if (ch == ('\n' | IN_EOT) && (tp->tty_termios.c_lflag
 14662	                                        & (ICANON|ECHONL)) == (ICANON|ECHONL))
 14663	                (*tp->tty_echo)(tp, '\n');
 14664	        return(ch);
 14665	  }
 14666	
 14667	  /* "Reprint" tells if the echo output has been messed up by other output. */
 14668	  rp = tp->tty_incount == 0 ? FALSE : tp->tty_reprint;
 14669	
 14670	  if ((ch & IN_CHAR) < ' ') {
 14671	        switch (ch & (IN_ESC|IN_EOF|IN_EOT|IN_CHAR)) {
 14672	            case '\t':
 14673	                len = 0;
 14674	                do {
 14675	                        (*tp->tty_echo)(tp, ' ');
 14676	                        len++;
 14677	                } while (len < TAB_SIZE && (tp->tty_position & TAB_M
 14678	                break;
 14679	            case '\r' | IN_EOT:
 14680	            case '\n' | IN_EOT:
 14681	                (*tp->tty_echo)(tp, ch & IN_CHAR);
 14682	                len = 0;
 14683	                break;
 14684	            default:
 14685	                (*tp->tty_echo)(tp, '^');
 14686	                (*tp->tty_echo)(tp, '@' + (ch & IN_CHAR));
 14687	                len = 2;
 14688	        }
 14689	  } else
 14690	  if ((ch & IN_CHAR) == '\177') {
 14691	        /* A DEL prints as "^?". */
 14692	        (*tp->tty_echo)(tp, '^');
 14693	        (*tp->tty_echo)(tp, '?');
 14694	        len = 2;
 14695	  } else {
 14696	        (*tp->tty_echo)(tp, ch & IN_CHAR);
 14697	        len = 1;
 14698	  }
 14699	  if (ch & IN_EOF) while (len > 0) { (*tp->tty_echo)(tp, '\b'); len--; }
 14700	
 14701	  tp->tty_reprint = rp;
 14702	  return(ch | (len << IN_LSHIFT));
 14703	}
	
 14705	/*===========================================================================*
 14706	 *                              rawecho                                      *
 14707	 *===========================================================================*/
 14708	PRIVATE void rawecho(tp, ch)
 14709	register tty_t *tp;

_________________________ Page 830    File: drivers/tty/tty.c _________________________
 14710	int ch;
 14711	{
 14712	/* Echo without interpretation if ECHO is set. */
 14713	  int rp = tp->tty_reprint;
 14714	  if (tp->tty_termios.c_lflag & ECHO) (*tp->tty_echo)(tp, ch);
 14715	  tp->tty_reprint = rp;
 14716	}
	
 14718	/*===========================================================================*
 14719	 *                              back_over                                    *
 14720	 *===========================================================================*/
 14721	PRIVATE int back_over(tp)
 14722	register tty_t *tp;
 14723	{
 14724	/* Backspace to previous character on screen and erase it. */
 14725	  u16_t *head;
 14726	  int len;
 14727	
 14728	  if (tp->tty_incount == 0) return(0);  /* queue empty */
 14729	  head = tp->tty_inhead;
 14730	  if (head == tp->tty_inbuf) head = bufend(tp->tty_inbuf);
 14731	  if (*--head & IN_EOT) return(0);              /* can't erase "line breaks" */
 14732	  if (tp->tty_reprint) reprint(tp);             /* reprint if messed up */
 14733	  tp->tty_inhead = head;
 14734	  tp->tty_incount--;
 14735	  if (tp->tty_termios.c_lflag & ECHOE) {
 14736	        len = (*head & IN_LEN) >> IN_LSHIFT;
 14737	        while (len > 0) {
 14738	                rawecho(tp, '\b');
 14739	                rawecho(tp, ' ');
 14740	                rawecho(tp, '\b');
 14741	                len--;
 14742	        }
 14743	  }
 14744	  return(1);                            /* one character erased */
 14745	}
	
 14747	/*===========================================================================*
 14748	 *                              reprint                                      *
 14749	 *===========================================================================*/
 14750	PRIVATE void reprint(tp)
 14751	register tty_t *tp;             /* pointer to tty struct */
 14752	{
 14753	/* Restore what has been echoed to screen before if the user input has been
 14754	 * messed up by output, or if REPRINT (^R) is typed.
 14755	 */
 14756	  int count;
 14757	  u16_t *head;
 14758	
 14759	  tp->tty_reprint = FALSE;
 14760	
 14761	  /* Find the last line break in the input. */
 14762	  head = tp->tty_inhead;
 14763	  count = tp->tty_incount;
 14764	  while (count > 0) {
 14765	        if (head == tp->tty_inbuf) head = bufend(tp->tty_inbuf);
 14766	        if (head[-1] & IN_EOT) break;
 14767	        head--;
 14768	        count--;
 14769	  }

_________________________ Page 831    File: drivers/tty/tty.c _________________________
 14770	  if (count == tp->tty_incount) return;         /* no reason to reprint */
 14771	
 14772	  /* Show REPRINT (^R) and move to a new line. */
 14773	  (void) tty_echo(tp, tp->tty_termios.c_cc[VREPRINT] | IN_ESC);
 14774	  rawecho(tp, '\r');
 14775	  rawecho(tp, '\n');
 14776	
 14777	  /* Reprint from the last break onwards. */
 14778	  do {
 14779	        if (head == bufend(tp->tty_inbuf)) head = tp->tty_inbuf;
 14780	        *head = tty_echo(tp, *head);
 14781	        head++;
 14782	        count++;
 14783	  } while (count < tp->tty_incount);
 14784	}
	
 14786	/*===========================================================================*
 14787	 *                              out_process                                  *
 14788	 *===========================================================================*/
 14789	PUBLIC void out_process(tp, bstart, bpos, bend, icount, ocount)
 14790	tty_t *tp;
 14791	char *bstart, *bpos, *bend;     /* start/pos/end of circular buffer */
 14792	int *icount;                    /* # input chars / input chars used */
 14793	int *ocount;                    /* max output chars / output chars used */
 14794	{
 14795	/* Perform output processing on a circular buffer.  *icount is the number of
 14796	 * bytes to process, and the number of bytes actually processed on return.
 14797	 * *ocount is the space available on input and the space used on output.
 14798	 * (Naturally *icount < *ocount.)  The column position is updated modulo
 14799	 * the TAB size, because we really only need it for tabs.
 14800	 */
 14801	
 14802	  int tablen;
 14803	  int ict = *icount;
 14804	  int oct = *ocount;
 14805	  int pos = tp->tty_position;
 14806	
 14807	  while (ict > 0) {
 14808	        switch (*bpos) {
 14809	        case '\7':
 14810	                break;
 14811	        case '\b':
 14812	                pos--;
 14813	                break;
 14814	        case '\r':
 14815	                pos = 0;
 14816	                break;
 14817	        case '\n':
 14818	                if ((tp->tty_termios.c_oflag & (OPOST|ONLCR))
 14819	                                                        == (OPOST|ONLCR)) {
 14820	                        /* Map LF to CR+LF if there is space.  Note that the
 14821	                         * next character in the buffer is overwritten, so
 14822	                         * we stop at this point.
 14823	                         */
 14824	                        if (oct >= 2) {
 14825	                                *bpos = '\r';
 14826	                                if (++bpos == bend) bpos = bstart;
 14827	                                *bpos = '\n';
 14828	                                pos = 0;
 14829	                                ict--;

_________________________ Page 832    File: drivers/tty/tty.c _________________________
 14830	                                oct -= 2;
 14831	                        }
 14832	                        goto out_done;  /* no space or buffer got changed */
 14833	                }
 14834	                break;
 14835	        case '\t':
 14836	                /* Best guess for the tab length. */
 14837	                tablen = TAB_SIZE - (pos & TAB_MASK);
 14838	
 14839	                if ((tp->tty_termios.c_oflag & (OPOST|XTABS))
 14840	                                                        == (OPOST|XTABS)) {
 14841	                        /* Tabs must be expanded. */
 14842	                        if (oct >= tablen) {
 14843	                                pos += tablen;
 14844	                                ict--;
 14845	                                oct -= tablen;
 14846	                                do {
 14847	                                        *bpos = ' ';
 14848	                                        if (++bpos == bend) bpos = bstart;
 14849	                                } while (--tablen != 0);
 14850	                        }
 14851	                        goto out_done;
 14852	                }
 14853	                /* Tabs are output directly. */
 14854	                pos += tablen;
 14855	                break;
 14856	        default:
 14857	                /* Assume any other character prints as one character. */
 14858	                pos++;
 14859	        }
 14860	        if (++bpos == bend) bpos = bstart;
 14861	        ict--;
 14862	        oct--;
 14863	  }
 14864	out_done:
 14865	  tp->tty_position = pos & TAB_MASK;
 14866	
 14867	  *icount -= ict;       /* [io]ct are the number of chars not used */
 14868	  *ocount -= oct;       /* *[io]count are the number of chars that are used */
 14869	}
	
 14871	/*===========================================================================*
 14872	 *                              dev_ioctl                                    *
 14873	 *===========================================================================*/
 14874	PRIVATE void dev_ioctl(tp)
 14875	tty_t *tp;
 14876	{
 14877	/* The ioctl's TCSETSW, TCSETSF and TCDRAIN wait for output to finish to make
 14878	 * sure that an attribute change doesn't affect the processing of current
 14879	 * output.  Once output finishes the ioctl is executed as in do_ioctl().
 14880	 */
 14881	  int result;
 14882	
 14883	  if (tp->tty_outleft > 0) return;              /* output not finished */
 14884	
 14885	  if (tp->tty_ioreq != TCDRAIN) {
 14886	        if (tp->tty_ioreq == TCSETSF) tty_icancel(tp);
 14887	        result = sys_vircopy(tp->tty_ioproc, D, tp->tty_iovir,
 14888	                        SELF, D, (vir_bytes) &tp->tty_termios,
 14889	                        (vir_bytes) sizeof(tp->tty_termios));

_________________________ Page 833    File: drivers/tty/tty.c _________________________
 14890	        setattr(tp);
 14891	  }
 14892	  tp->tty_ioreq = 0;
 14893	  tty_reply(REVIVE, tp->tty_iocaller, tp->tty_ioproc, result);
 14894	}
	
 14896	/*===========================================================================*
 14897	 *                              setattr                                      *
 14898	 *===========================================================================*/
 14899	PRIVATE void setattr(tp)
 14900	tty_t *tp;
 14901	{
 14902	/* Apply the new line attributes (raw/canonical, line speed, etc.) */
 14903	  u16_t *inp;
 14904	  int count;
 14905	
 14906	  if (!(tp->tty_termios.c_lflag & ICANON)) {
 14907	        /* Raw mode; put a "line break" on all characters in the input queue.
 14908	         * It is undefined what happens to the input queue when ICANON is
 14909	         * switched off, a process should use TCSAFLUSH to flush the queue.
 14910	         * Keeping the queue to preserve typeahead is the Right Thing, however
 14911	         * when a process does use TCSANOW to switch to raw mode.
 14912	         */
 14913	        count = tp->tty_eotct = tp->tty_incount;
 14914	        inp = tp->tty_intail;
 14915	        while (count > 0) {
 14916	                *inp |= IN_EOT;
 14917	                if (++inp == bufend(tp->tty_inbuf)) inp = tp->tty_inbuf;
 14918	                --count;
 14919	        }
 14920	  }
 14921	
 14922	  /* Inspect MIN and TIME. */
 14923	  settimer(tp, FALSE);
 14924	  if (tp->tty_termios.c_lflag & ICANON) {
 14925	        /* No MIN & TIME in canonical mode. */
 14926	        tp->tty_min = 1;
 14927	  } else {
 14928	        /* In raw mode MIN is the number of chars wanted, and TIME how long
 14929	         * to wait for them.  With interesting exceptions if either is zero.
 14930	         */
 14931	        tp->tty_min = tp->tty_termios.c_cc[VMIN];
 14932	        if (tp->tty_min == 0 && tp->tty_termios.c_cc[VTIME] > 0)
 14933	                tp->tty_min = 1;
 14934	  }
 14935	
 14936	  if (!(tp->tty_termios.c_iflag & IXON)) {
 14937	        /* No start/stop output control, so don't leave output inhibited. */
 14938	        tp->tty_inhibited = RUNNING;
 14939	        tp->tty_events = 1;
 14940	  }
 14941	
 14942	  /* Setting the output speed to zero hangs up the phone. */
 14943	  if (tp->tty_termios.c_ospeed == B0) sigchar(tp, SIGHUP);
 14944	
 14945	  /* Set new line speed, character size, etc at the device level. */
 14946	  (*tp->tty_ioctl)(tp, 0);
 14947	}
	

_________________________ Page 834    File: drivers/tty/tty.c _________________________
 14949	/*===========================================================================*
 14950	 *                              tty_reply                                    *
 14951	 *===========================================================================*/
 14952	PUBLIC void tty_reply(code, replyee, proc_nr, status)
 14953	int code;                       /* TASK_REPLY or REVIVE */
 14954	int replyee;                    /* destination address for the reply */
 14955	int proc_nr;                    /* to whom should the reply go? */
 14956	int status;                     /* reply code */
 14957	{
 14958	/* Send a reply to a process that wanted to read or write data. */
 14959	  message tty_mess;
 14960	
 14961	  tty_mess.m_type = code;
 14962	  tty_mess.REP_PROC_NR = proc_nr;
 14963	  tty_mess.REP_STATUS = status;
 14964	
 14965	  if ((status = send(replyee, &tty_mess)) != OK) {
 14966	        panic("TTY","tty_reply failed, status\n", status);
 14967	  }
 14968	}
	
 14970	/*===========================================================================*
 14971	 *                              sigchar                                      *
 14972	 *===========================================================================*/
 14973	PUBLIC void sigchar(tp, sig)
 14974	register tty_t *tp;
 14975	int sig;                        /* SIGINT, SIGQUIT, SIGKILL or SIGHUP */
 14976	{
 14977	/* Process a SIGINT, SIGQUIT or SIGKILL char from the keyboard or SIGHUP from
 14978	 * a tty close, "stty 0", or a real RS-232 hangup.  MM will send the signal to
 14979	 * the process group (INT, QUIT), all processes (KILL), or the session leader
 14980	 * (HUP).
 14981	 */
 14982	  int status;
 14983	
 14984	  if (tp->tty_pgrp != 0) 
 14985	      if (OK != (status = sys_kill(tp->tty_pgrp, sig)))
 14986	        panic("TTY","Error, call to sys_kill failed", status);
 14987	
 14988	  if (!(tp->tty_termios.c_lflag & NOFLSH)) {
 14989	        tp->tty_incount = tp->tty_eotct = 0;    /* kill earlier input */
 14990	        tp->tty_intail = tp->tty_inhead;
 14991	        (*tp->tty_ocancel)(tp, 0);                      /* kill all output */
 14992	        tp->tty_inhibited = RUNNING;
 14993	        tp->tty_events = 1;
 14994	  }
 14995	}
	
 14997	/*===========================================================================*
 14998	 *                              tty_icancel                                  *
 14999	 *===========================================================================*/
 15000	PRIVATE void tty_icancel(tp)
 15001	register tty_t *tp;
 15002	{
 15003	/* Discard all pending input, tty buffer or device. */
 15004	
 15005	  tp->tty_incount = tp->tty_eotct = 0;
 15006	  tp->tty_intail = tp->tty_inhead;
 15007	  (*tp->tty_icancel)(tp, 0);
 15008	}

_________________________ Page 835    File: drivers/tty/tty.c _________________________
	
 15010	/*===========================================================================*
 15011	 *                              tty_init                                     *
 15012	 *===========================================================================*/
 15013	PRIVATE void tty_init()
 15014	{
 15015	/* Initialize tty structure and call device initialization routines. */
 15016	
 15017	  register tty_t *tp;
 15018	  int s;
 15019	  struct sigaction sigact;
 15020	
 15021	  /* Initialize the terminal lines. */
 15022	  for (tp = FIRST_TTY,s=0; tp < END_TTY; tp++,s++) {
 15023	
 15024	        tp->tty_index = s;
 15025	
 15026	        tmr_inittimer(&tp->tty_tmr);
 15027	
 15028	        tp->tty_intail = tp->tty_inhead = tp->tty_inbuf;
 15029	        tp->tty_min = 1;
 15030	        tp->tty_termios = termios_defaults;
 15031	        tp->tty_icancel = tp->tty_ocancel = tp->tty_ioctl = tp->tty_close =
 15032	                                                                tty_devnop;
 15033	        if (tp < tty_addr(NR_CONS)) {
 15034	                scr_init(tp);
 15035	                tp->tty_minor = CONS_MINOR + s;
 15036	        } else
 15037	        if (tp < tty_addr(NR_CONS+NR_RS_LINES)) {
 15038	                rs_init(tp);
 15039	                tp->tty_minor = RS232_MINOR + s-NR_CONS;
 15040	        } else {
 15041	                pty_init(tp);
 15042	                tp->tty_minor = s - (NR_CONS+NR_RS_LINES) + TTYPX_MINOR;
 15043	        }
 15044	  }
 15045	}
	
 15047	/*===========================================================================*
 15048	 *                              tty_timed_out                                *
 15049	 *===========================================================================*/
 15050	PRIVATE void tty_timed_out(timer_t *tp)
 15051	{
 15052	/* This timer has expired. Set the events flag, to force processing. */
 15053	  tty_t *tty_ptr;
 15054	  tty_ptr = &tty_table[tmr_arg(tp)->ta_int];
 15055	  tty_ptr->tty_min = 0;                 /* force read to succeed */
 15056	  tty_ptr->tty_events = 1;              
 15057	}
	
 15059	/*===========================================================================*
 15060	 *                              expire_timers                                *
 15061	 *===========================================================================*/
 15062	PRIVATE void expire_timers(void)
 15063	{
 15064	/* A synchronous alarm message was received. Check if there are any expired 
 15065	 * timers. Possibly set the event flag and reschedule another alarm.  
 15066	 */
 15067	  clock_t now;                          /* current time */
 15068	  int s;

_________________________ Page 836    File: drivers/tty/tty.c _________________________
 15069	
 15070	  /* Get the current time to compare the timers against. */
 15071	  if ((s=getuptime(&now)) != OK)
 15072	        panic("TTY","Couldn't get uptime from clock.", s);
 15073	
 15074	  /* Scan the queue of timers for expired timers. This dispatch the watchdog
 15075	   * functions of expired timers. Possibly a new alarm call must be scheduled.
 15076	   */
 15077	  tmrs_exptimers(&tty_timers, now, NULL);
 15078	  if (tty_timers == NULL) tty_next_timeout = TMR_NEVER;
 15079	  else {                                          /* set new sync alarm */
 15080	        tty_next_timeout = tty_timers->tmr_exp_time;
 15081	        if ((s=sys_setalarm(tty_next_timeout, 1)) != OK)
 15082	                panic("TTY","Couldn't set synchronous alarm.", s);
 15083	  }
 15084	}
	
 15086	/*===========================================================================*
 15087	 *                              settimer                                     *
 15088	 *===========================================================================*/
 15089	PRIVATE void settimer(tty_ptr, enable)
 15090	tty_t *tty_ptr;                 /* line to set or unset a timer on */
 15091	int enable;                     /* set timer if true, otherwise unset */
 15092	{
 15093	  clock_t now;                          /* current time */
 15094	  clock_t exp_time;
 15095	  int s;
 15096	
 15097	  /* Get the current time to calculate the timeout time. */
 15098	  if ((s=getuptime(&now)) != OK)
 15099	        panic("TTY","Couldn't get uptime from clock.", s);
 15100	  if (enable) {
 15101	        exp_time = now + tty_ptr->tty_termios.c_cc[VTIME] * (HZ/10);
 15102	        /* Set a new timer for enabling the TTY events flags. */
 15103	        tmrs_settimer(&tty_timers, &tty_ptr->tty_tmr, 
 15104	                exp_time, tty_timed_out, NULL);  
 15105	  } else {
 15106	        /* Remove the timer from the active and expired lists. */
 15107	        tmrs_clrtimer(&tty_timers, &tty_ptr->tty_tmr, NULL);
 15108	  }
 15109	  
 15110	  /* Now check if a new alarm must be scheduled. This happens when the front
 15111	   * of the timers queue was disabled or reinserted at another position, or
 15112	   * when a new timer was added to the front.
 15113	   */
 15114	  if (tty_timers == NULL) tty_next_timeout = TMR_NEVER;
 15115	  else if (tty_timers->tmr_exp_time != tty_next_timeout) { 
 15116	        tty_next_timeout = tty_timers->tmr_exp_time;
 15117	        if ((s=sys_setalarm(tty_next_timeout, 1)) != OK)
 15118	                panic("TTY","Couldn't set synchronous alarm.", s);
 15119	  }
 15120	}
	
 15122	/*===========================================================================*
 15123	 *                              tty_devnop                                   *
 15124	 *===========================================================================*/
 15125	PUBLIC int tty_devnop(tp, try)
 15126	tty_t *tp;
 15127	int try;
 15128	{

_________________________ Page 837    File: drivers/tty/tty.c _________________________
 15129	  /* Some functions need not be implemented at the device level. */
 15130	}
	
 15132	/*===========================================================================*
 15133	 *                              do_select                                    *
 15134	 *===========================================================================*/
 15135	PRIVATE void do_select(tp, m_ptr)
 15136	register tty_t *tp;             /* pointer to tty struct */
 15137	register message *m_ptr;        /* pointer to message sent to the task */
 15138	{
 15139	        int ops, ready_ops = 0, watch;
 15140	
 15141	        ops = m_ptr->PROC_NR & (SEL_RD|SEL_WR|SEL_ERR);
 15142	        watch = (m_ptr->PROC_NR & SEL_NOTIFY) ? 1 : 0;
 15143	
 15144	        ready_ops = select_try(tp, ops);
 15145	
 15146	        if (!ready_ops && ops && watch) {
 15147	                tp->tty_select_ops |= ops;
 15148	                tp->tty_select_proc = m_ptr->m_source;
 15149	        }
 15150	
 15151	        tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, ready_ops);
 15152	
 15153	        return;
 15154	}





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/tty/keyboard.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 15200	/* Keyboard driver for PC's and AT's.
 15201	 *
 15202	 * Changes: 
 15203	 *   Jul 13, 2004   processes can observe function keys  (Jorrit N. Herder)
 15204	 *   Jun 15, 2004   removed wreboot(), except panic dumps (Jorrit N. Herder)
 15205	 *   Feb 04, 1994   loadable keymaps  (Marcus Hampel)
 15206	 */
 15207	
 15208	#include "../drivers.h"
 15209	#include <sys/time.h>
 15210	#include <sys/select.h>
 15211	#include <termios.h>
 15212	#include <signal.h>
 15213	#include <unistd.h>
 15214	#include <minix/callnr.h>
 15215	#include <minix/com.h>
 15216	#include <minix/keymap.h>
 15217	#include "tty.h"
 15218	#include "keymaps/us-std.src"
 15219	#include "../../kernel/const.h"
 15220	#include "../../kernel/config.h"
 15221	#include "../../kernel/type.h"
 15222	#include "../../kernel/proc.h"
 15223	
 15224	int irq_hook_id = -1;

_________________________ Page 838    File: drivers/tty/keyboard.c _______________________
 15225	
 15226	/* Standard and AT keyboard.  (PS/2 MCA implies AT throughout.) */
 15227	#define KEYBD           0x60    /* I/O port for keyboard data */
 15228	
 15229	/* AT keyboard. */
 15230	#define KB_COMMAND      0x64    /* I/O port for commands on AT */
 15231	#define KB_STATUS       0x64    /* I/O port for status on AT */
 15232	#define KB_ACK          0xFA    /* keyboard ack response */
 15233	#define KB_OUT_FULL     0x01    /* status bit set when keypress char pending */
 15234	#define KB_IN_FULL      0x02    /* status bit set when not ready to receive */
 15235	#define LED_CODE        0xED    /* command to keyboard to set LEDs */
 15236	#define MAX_KB_ACK_RETRIES 0x1000       /* max #times to wait for kb ack */
 15237	#define MAX_KB_BUSY_RETRIES 0x1000      /* max #times to loop while kb busy */
 15238	#define KBIT            0x80    /* bit used to ack characters to keyboard */
 15239	
 15240	/* Miscellaneous. */
 15241	#define ESC_SCAN        0x01    /* reboot key when panicking */
 15242	#define SLASH_SCAN      0x35    /* to recognize numeric slash */
 15243	#define RSHIFT_SCAN     0x36    /* to distinguish left and right shift */
 15244	#define HOME_SCAN       0x47    /* first key on the numeric keypad */
 15245	#define INS_SCAN        0x52    /* INS for use in CTRL-ALT-INS reboot */
 15246	#define DEL_SCAN        0x53    /* DEL for use in CTRL-ALT-DEL reboot */
 15247	
 15248	#define CONSOLE            0    /* line number for console */
 15249	#define KB_IN_BYTES       32    /* size of keyboard input buffer */
 15250	PRIVATE char ibuf[KB_IN_BYTES]; /* input buffer */
 15251	PRIVATE char *ihead = ibuf;     /* next free spot in input buffer */
 15252	PRIVATE char *itail = ibuf;     /* scan code to return to TTY */
 15253	PRIVATE int icount;             /* # codes in buffer */
 15254	
 15255	PRIVATE int esc;                /* escape scan code detected? */
 15256	PRIVATE int alt_l;              /* left alt key state */
 15257	PRIVATE int alt_r;              /* right alt key state */
 15258	PRIVATE int alt;                /* either alt key */
 15259	PRIVATE int ctrl_l;             /* left control key state */
 15260	PRIVATE int ctrl_r;             /* right control key state */
 15261	PRIVATE int ctrl;               /* either control key */
 15262	PRIVATE int shift_l;            /* left shift key state */
 15263	PRIVATE int shift_r;            /* right shift key state */
 15264	PRIVATE int shift;              /* either shift key */
 15265	PRIVATE int num_down;           /* num lock key depressed */
 15266	PRIVATE int caps_down;          /* caps lock key depressed */
 15267	PRIVATE int scroll_down;        /* scroll lock key depressed */
 15268	PRIVATE int locks[NR_CONS];     /* per console lock keys state */
 15269	
 15270	/* Lock key active bits.  Chosen to be equal to the keyboard LED bits. */
 15271	#define SCROLL_LOCK     0x01
 15272	#define NUM_LOCK        0x02
 15273	#define CAPS_LOCK       0x04
 15274	
 15275	PRIVATE char numpad_map[] =
 15276	                {'H', 'Y', 'A', 'B', 'D', 'C', 'V', 'U', 'G', 'S', 'T', '@'};
 15277	
 15278	/* Variables and definition for observed function keys. */
 15279	typedef struct observer { int proc_nr; int events; } obs_t;
 15280	PRIVATE obs_t  fkey_obs[12];    /* observers for F1-F12 */
 15281	PRIVATE obs_t sfkey_obs[12];    /* observers for SHIFT F1-F12 */
 15282	
 15283	FORWARD _PROTOTYPE( int kb_ack, (void)                                  );
 15284	FORWARD _PROTOTYPE( int kb_wait, (void)                                 );

_________________________ Page 839    File: drivers/tty/keyboard.c _______________________
 15285	FORWARD _PROTOTYPE( int func_key, (int scode)                           );
 15286	FORWARD _PROTOTYPE( int scan_keyboard, (void)                           );
 15287	FORWARD _PROTOTYPE( unsigned make_break, (int scode)                    );
 15288	FORWARD _PROTOTYPE( void set_leds, (void)                               );
 15289	FORWARD _PROTOTYPE( void show_key_mappings, (void)                      );
 15290	FORWARD _PROTOTYPE( int kb_read, (struct tty *tp, int try)              );
 15291	FORWARD _PROTOTYPE( unsigned map_key, (int scode)                       );
 15292	
 15293	/*===========================================================================*
 15294	 *                              map_key0                                     *
 15295	 *===========================================================================*/
 15296	/* Map a scan code to an ASCII code ignoring modifiers. */
 15297	#define map_key0(scode)  \
 15298	        ((unsigned) keymap[(scode) * MAP_COLS])
 15299	
 15300	/*===========================================================================*
 15301	 *                              map_key                                      *
 15302	 *===========================================================================*/
 15303	PRIVATE unsigned map_key(scode)
 15304	int scode;
 15305	{
 15306	/* Map a scan code to an ASCII code. */
 15307	
 15308	  int caps, column, lk;
 15309	  u16_t *keyrow;
 15310	
 15311	  if (scode == SLASH_SCAN && esc) return '/';   /* don't map numeric slash 
 15312	
 15313	  keyrow = &keymap[scode * MAP_COLS];
 15314	
 15315	  caps = shift;
 15316	  lk = locks[ccurrent];
 15317	  if ((lk & NUM_LOCK) && HOME_SCAN <= scode && scode <= D
 15318	  if ((lk & CAPS_LOCK) && (keyrow[0] & HASCAPS)) caps = !caps;
 15319	
 15320	  if (alt) {
 15321	        column = 2;
 15322	        if (ctrl || alt_r) column = 3;  /* Ctrl + Alt == AltGr */
 15323	        if (caps) column = 4;
 15324	  } else {
 15325	        column = 0;
 15326	        if (caps) column = 1;
 15327	        if (ctrl) column = 5;
 15328	  }
 15329	  return keyrow[column] & ~HASCAPS;
 15330	}
	
 15332	/*===========================================================================*
 15333	 *                              kbd_interrupt                                *
 15334	 *===========================================================================*/
 15335	PUBLIC void kbd_interrupt(m_ptr)
 15336	message *m_ptr;
 15337	{
 15338	/* A keyboard interrupt has occurred.  Process it. */
 15339	  int scode;
 15340	  static timer_t timer;         /* timer must be static! */
 15341	
 15342	  /* Fetch the character from the keyboard hardware and acknowledge it. */
 15343	  scode = scan_keyboard();
 15344	

_________________________ Page 840    File: drivers/tty/keyboard.c _______________________
 15345	  /* Store the scancode in memory so the task can get at it later. */
 15346	  if (icount < KB_IN_BYTES) {
 15347	        *ihead++ = scode;
 15348	        if (ihead == ibuf + KB_IN_BYTES) ihead = ibuf;
 15349	        icount++;
 15350	        tty_table[ccurrent].tty_events = 1;
 15351	        if (tty_table[ccurrent].tty_select_ops & SEL_RD) {
 15352	                select_retry(&tty_table[ccurrent]);
 15353	        }
 15354	  }
 15355	}
	
 15357	/*===========================================================================*
 15358	 *                              kb_read                                      *
 15359	 *===========================================================================*/
 15360	PRIVATE int kb_read(tp, try)
 15361	tty_t *tp;
 15362	int try;
 15363	{
 15364	/* Process characters from the circular keyboard buffer. */
 15365	  char buf[3];
 15366	  int scode;
 15367	  unsigned ch;
 15368	
 15369	  tp = &tty_table[ccurrent];            /* always use the current console */
 15370	
 15371	  if (try) {
 15372	        if (icount > 0) return 1;
 15373	        return 0;
 15374	  }
 15375	
 15376	  while (icount > 0) {
 15377	        scode = *itail++;                       /* take one key scan code */
 15378	        if (itail == ibuf + KB_IN_BYTES) itail = ibuf;
 15379	        icount--;
 15380	
 15381	        /* Function keys are being used for debug dumps. */
 15382	        if (func_key(scode)) continue;
 15383	
 15384	        /* Perform make/break processing. */
 15385	        ch = make_break(scode);
 15386	
 15387	        if (ch <= 0xFF) {
 15388	                /* A normal character. */
 15389	                buf[0] = ch;
 15390	                (void) in_process(tp, buf, 1);
 15391	        } else
 15392	        if (HOME <= ch && ch <= INSRT) {
 15393	                /* An ASCII escape sequence generated by the numeric pad. */
 15394	                buf[0] = ESC;
 15395	                buf[1] = '[';
 15396	                buf[2] = numpad_map[ch - HOME];
 15397	                (void) in_process(tp, buf, 3);
 15398	        } else
 15399	        if (ch == ALEFT) {
 15400	                /* Choose lower numbered console as current console. */
 15401	                select_console(ccurrent - 1);
 15402	                set_leds();
 15403	        } else
 15404	        if (ch == ARIGHT) {

_________________________ Page 841    File: drivers/tty/keyboard.c _______________________
 15405	                /* Choose higher numbered console as current console. */
 15406	                select_console(ccurrent + 1);
 15407	                set_leds();
 15408	        } else
 15409	        if (AF1 <= ch && ch <= AF12) {
 15410	                /* Alt-F1 is console, Alt-F2 is ttyc1, etc. */
 15411	                select_console(ch - AF1);
 15412	                set_leds();
 15413	        } else
 15414	        if (CF1 <= ch && ch <= CF12) {
 15415	            switch(ch) {
 15416	                case CF1: show_key_mappings(); break; 
 15417	                case CF3: toggle_scroll(); break; /* hardware <-> software */   
 15418	                case CF7: sigchar(&tty_table[CONSOLE], SIGQUIT); break;
 15419	                case CF8: sigchar(&tty_table[CONSOLE], SIGINT); break;
 15420	                case CF9: sigchar(&tty_table[CONSOLE], SIGKILL); break;
 15421	            }
 15422	        }
 15423	  }
 15424	
 15425	  return 1;
 15426	}
	
 15428	/*===========================================================================*
 15429	 *                              make_break                                   *
 15430	 *===========================================================================*/
 15431	PRIVATE unsigned make_break(scode)
 15432	int scode;                      /* scan code of key just struck or released */
 15433	{
 15434	/* This routine can handle keyboards that interrupt only on key depression,
 15435	 * as well as keyboards that interrupt on key depression and key release.
 15436	 * For efficiency, the interrupt routine filters out most key releases.
 15437	 */
 15438	  int ch, make, escape;
 15439	  static int CAD_count = 0;
 15440	
 15441	  /* Check for CTRL-ALT-DEL, and if found, halt the computer. This would
 15442	   * be better done in keyboard() in case TTY is hung, except control and
 15443	   * alt are set in the high level code.
 15444	   */
 15445	  if (ctrl && alt && (scode == DEL_SCAN || scode == INS_SCAN))
 15446	  {
 15447	        if (++CAD_count == 3) sys_abort(RBT_HALT);
 15448	        sys_kill(INIT_PROC_NR, SIGABRT);
 15449	        return -1;
 15450	  }
 15451	
 15452	  /* High-order bit set on key release. */
 15453	  make = (scode & KEY_RELEASE) == 0;            /* true if pressed */
 15454	
 15455	  ch = map_key(scode &= ASCII_MASK);            /* map to ASCII */
 15456	
 15457	  escape = esc;         /* Key is escaped?  (true if added since the XT) */
 15458	  esc = 0;
 15459	
 15460	  switch (ch) {
 15461	        case CTRL:              /* Left or right control key */
 15462	                *(escape ? &ctrl_r : &ctrl_l) = make;
 15463	                ctrl = ctrl_l | ctrl_r;
 15464	                break;

_________________________ Page 842    File: drivers/tty/keyboard.c _______________________
 15465	        case SHIFT:             /* Left or right shift key */
 15466	                *(scode == RSHIFT_SCAN ? &shift_r : &shift_l) = make;
 15467	                shift = shift_l | shift_r;
 15468	                break;
 15469	        case ALT:               /* Left or right alt key */
 15470	                *(escape ? &alt_r : &alt_l) = make;
 15471	                alt = alt_l | alt_r;
 15472	                break;
 15473	        case CALOCK:            /* Caps lock - toggle on 0 -> 1 transition */
 15474	                if (caps_down < make) {
 15475	                        locks[ccurrent] ^= CAPS_LOCK;
 15476	                        set_leds();
 15477	                }
 15478	                caps_down = make;
 15479	                break;
 15480	        case NLOCK:             /* Num lock */
 15481	                if (num_down < make) {
 15482	                        locks[ccurrent] ^= NUM_LOCK;
 15483	                        set_leds();
 15484	                }
 15485	                num_down = make;
 15486	                break;
 15487	        case SLOCK:             /* Scroll lock */
 15488	                if (scroll_down < make) {
 15489	                        locks[ccurrent] ^= SCROLL_LOCK;
 15490	                        set_leds();
 15491	                }
 15492	                scroll_down = make;
 15493	                break;
 15494	        case EXTKEY:            /* Escape keycode */
 15495	                esc = 1;                /* Next key is escaped */
 15496	                return(-1);
 15497	        default:                /* A normal key */
 15498	                if (make) return(ch);
 15499	  }
 15500	
 15501	  /* Key release, or a shift type key. */
 15502	  return(-1);
 15503	}
	
 15505	/*===========================================================================*
 15506	 *                              set_leds                                     *
 15507	 *===========================================================================*/
 15508	PRIVATE void set_leds()
 15509	{
 15510	/* Set the LEDs on the caps, num, and scroll lock keys */
 15511	  int s;
 15512	  if (! machine.pc_at) return;  /* PC/XT doesn't have LEDs */
 15513	
 15514	  kb_wait();                    /* wait for buffer empty  */
 15515	  if ((s=sys_outb(KEYBD, LED_CODE)) != OK)
 15516	      printf("Warning, sys_outb couldn't prepare for LED values: %d\n", s);
 15517	                                /* prepare keyboard to accept LED values */
 15518	  kb_ack();                     /* wait for ack response  */
 15519	
 15520	  kb_wait();                    /* wait for buffer empty  */
 15521	  if ((s=sys_outb(KEYBD, locks[ccurrent])) != OK)
 15522	      printf("Warning, sys_outb couldn't give LED values: %d\n", s);
 15523	                                /* give keyboard LED values */
 15524	  kb_ack();                     /* wait for ack response  */

_________________________ Page 843    File: drivers/tty/keyboard.c _______________________
 15525	}
	
 15527	/*===========================================================================*
 15528	 *                              kb_wait                                      *
 15529	 *===========================================================================*/
 15530	PRIVATE int kb_wait()
 15531	{
 15532	/* Wait until the controller is ready; return zero if this times out. */
 15533	
 15534	  int retries, status, temp;
 15535	  int s;
 15536	
 15537	  retries = MAX_KB_BUSY_RETRIES + 1;    /* wait until not busy */
 15538	  do {
 15539	      s = sys_inb(KB_STATUS, &status);
 15540	      if (status & KB_OUT_FULL) {
 15541	          s = sys_inb(KEYBD, &temp);    /* discard value */
 15542	      }
 15543	      if (! (status & (KB_IN_FULL|KB_OUT_FULL)) )
 15544	          break;                        /* wait until ready */
 15545	  } while (--retries != 0);             /* continue unless timeout */ 
 15546	  return(retries);              /* zero on timeout, positive if ready */
 15547	}
	
 15549	/*===========================================================================*
 15550	 *                              kb_ack                                       *
 15551	 *===========================================================================*/
 15552	PRIVATE int kb_ack()
 15553	{
 15554	/* Wait until kbd acknowledges last command; return zero if this times out. */
 15555	
 15556	  int retries, s;
 15557	  u8_t u8val;
 15558	
 15559	  retries = MAX_KB_ACK_RETRIES + 1;
 15560	  do {
 15561	      s = sys_inb(KEYBD, &u8val);
 15562	      if (u8val == KB_ACK)      
 15563	          break;                /* wait for ack */
 15564	  } while(--retries != 0);      /* continue unless timeout */
 15565	
 15566	  return(retries);              /* nonzero if ack received */
 15567	}
	
 15569	/*===========================================================================*
 15570	 *                              kb_init                                      *
 15571	 *===========================================================================*/
 15572	PUBLIC void kb_init(tp)
 15573	tty_t *tp;
 15574	{
 15575	/* Initialize the keyboard driver. */
 15576	
 15577	  tp->tty_devread = kb_read;    /* input function */
 15578	}
	
 15580	/*===========================================================================*
 15581	 *                              kb_init_once                                 *
 15582	 *===========================================================================*/
 15583	PUBLIC void kb_init_once(void)
 15584	{

_________________________ Page 844    File: drivers/tty/keyboard.c _______________________
 15585	  int i;
 15586	
 15587	  set_leds();                   /* turn off numlock led */
 15588	  scan_keyboard();              /* discard leftover keystroke */
 15589	
 15590	      /* Clear the function key observers array. Also see func_key(). */
 15591	      for (i=0; i<12; i++) {
 15592	          fkey_obs[i].proc_nr = NONE;   /* F1-F12 observers */
 15593	          fkey_obs[i].events = 0;       /* F1-F12 observers */
 15594	          sfkey_obs[i].proc_nr = NONE;  /* Shift F1-F12 observers */
 15595	          sfkey_obs[i].events = 0;      /* Shift F1-F12 observers */
 15596	      }
 15597	
 15598	      /* Set interrupt handler and enable keyboard IRQ. */
 15599	      irq_hook_id = KEYBOARD_IRQ;       /* id to be returned on interrupt */
 15600	      if ((i=sys_irqsetpolicy(KEYBOARD_IRQ, IRQ_REENABLE, &irq_hook_id)) != OK)
 15601	          panic("TTY",  "Couldn't set keyboard IRQ policy", i);
 15602	      if ((i=sys_irqenable(&irq_hook_id)) != OK)
 15603	          panic("TTY", "Couldn't enable keyboard IRQs", i);
 15604	      kbd_irq_set |= (1 << KEYBOARD_IRQ);
 15605	}
	
 15607	/*===========================================================================*
 15608	 *                              kbd_loadmap                                  *
 15609	 *===========================================================================*/
 15610	PUBLIC int kbd_loadmap(m)
 15611	message *m;
 15612	{
 15613	/* Load a new keymap. */
 15614	  int result;
 15615	  result = sys_vircopy(m->PROC_NR, D, (vir_bytes) m->ADDRESS,
 15616	        SELF, D, (vir_bytes) keymap, 
 15617	        (vir_bytes) sizeof(keymap));
 15618	  return(result);
 15619	}
	
 15621	/*===========================================================================*
 15622	 *                              do_fkey_ctl                                  *
 15623	 *===========================================================================*/
 15624	PUBLIC void do_fkey_ctl(m_ptr)
 15625	message *m_ptr;                 /* pointer to the request message */
 15626	{
 15627	/* This procedure allows processes to register a function key to receive
 15628	 * notifications if it is pressed. At most one binding per key can exist.
 15629	 */
 15630	  int i; 
 15631	  int result;
 15632	
 15633	  switch (m_ptr->FKEY_REQUEST) {        /* see what we must do */
 15634	  case FKEY_MAP:                        /* request for new mapping */
 15635	      result = OK;                      /* assume everything will be ok*/
 15636	      for (i=0; i < 12; i++) {          /* check F1-F12 keys */
 15637	          if (bit_isset(m_ptr->FKEY_FKEYS, i+1) ) {
 15638	              if (fkey_obs[i].proc_nr == NONE) { 
 15639	                  fkey_obs[i].proc_nr = m_ptr->m_source;
 15640	                  fkey_obs[i].events = 0;
 15641	                  bit_unset(m_ptr->FKEY_FKEYS, i+1);
 15642	              } else {
 15643	                  printf("WARNING, fkey_map failed F%d\n", i+1);
 15644	                  result = EBUSY;       /* report failure, but try rest */

_________________________ Page 845    File: drivers/tty/keyboard.c _______________________
 15645	              }
 15646	          }
 15647	      }
 15648	      for (i=0; i < 12; i++) {          /* check Shift+F1-F12 keys */
 15649	          if (bit_isset(m_ptr->FKEY_SFKEYS, i+1) ) {
 15650	              if (sfkey_obs[i].proc_nr == NONE) { 
 15651	                  sfkey_obs[i].proc_nr = m_ptr->m_source;
 15652	                  sfkey_obs[i].events = 0;
 15653	                  bit_unset(m_ptr->FKEY_SFKEYS, i+1);
 15654	              } else {
 15655	                  printf("WARNING, fkey_map failed Shift F%d\n", i+1);
 15656	                  result = EBUSY;       /* report failure but try rest */
 15657	              }
 15658	          }
 15659	      }
 15660	      break;
 15661	  case FKEY_UNMAP:
 15662	      result = OK;                      /* assume everything will be ok*/
 15663	      for (i=0; i < 12; i++) {          /* check F1-F12 keys */
 15664	          if (bit_isset(m_ptr->FKEY_FKEYS, i+1) ) {
 15665	              if (fkey_obs[i].proc_nr == m_ptr->m_source) { 
 15666	                  fkey_obs[i].proc_nr = NONE;
 15667	                  fkey_obs[i].events = 0;
 15668	                  bit_unset(m_ptr->FKEY_FKEYS, i+1);
 15669	              } else {
 15670	                  result = EPERM;       /* report failure, but try rest */
 15671	              }
 15672	          }
 15673	      }
 15674	      for (i=0; i < 12; i++) {          /* check Shift+F1-F12 keys */
 15675	          if (bit_isset(m_ptr->FKEY_SFKEYS, i+1) ) {
 15676	              if (sfkey_obs[i].proc_nr == m_ptr->m_source) { 
 15677	                  sfkey_obs[i].proc_nr = NONE;
 15678	                  sfkey_obs[i].events = 0;
 15679	                  bit_unset(m_ptr->FKEY_SFKEYS, i+1);
 15680	              } else {
 15681	                  result = EPERM;       /* report failure, but try rest */
 15682	              }
 15683	          }
 15684	      }
 15685	      break;
 15686	  case FKEY_EVENTS:
 15687	      m_ptr->FKEY_FKEYS = m_ptr->FKEY_SFKEYS = 0;
 15688	      for (i=0; i < 12; i++) {          /* check (Shift+) F1-F12 keys */
 15689	          if (fkey_obs[i].proc_nr == m_ptr->m_source) {
 15690	              if (fkey_obs[i].events) { 
 15691	                  bit_set(m_ptr->FKEY_FKEYS, i+1);
 15692	                  fkey_obs[i].events = 0;
 15693	              }
 15694	          }
 15695	          if (sfkey_obs[i].proc_nr == m_ptr->m_source) {
 15696	              if (sfkey_obs[i].events) { 
 15697	                  bit_set(m_ptr->FKEY_SFKEYS, i+1);
 15698	                  sfkey_obs[i].events = 0;
 15699	              }
 15700	          }
 15701	      }
 15702	      break;
 15703	  default:
 15704	          result =  EINVAL;             /* key cannot be observed */

_________________________ Page 846    File: drivers/tty/keyboard.c _______________________
 15705	  }
 15706	
 15707	  /* Almost done, return result to caller. */
 15708	  m_ptr->m_type = result;
 15709	  send(m_ptr->m_source, m_ptr);
 15710	}
	
 15712	/*===========================================================================*
 15713	 *                              func_key                                     *
 15714	 *===========================================================================*/
 15715	PRIVATE int func_key(scode)
 15716	int scode;                      /* scan code for a function key */
 15717	{
 15718	/* This procedure traps function keys for debugging purposes. Observers of 
 15719	 * function keys are kept in a global array. If a subject (a key) is pressed
 15720	 * the observer is notified of the event. Initialization of the arrays is done
 15721	 * in kb_init, where NONE is set to indicate there is no interest in the key.
 15722	 * Returns FALSE on a key release or if the key is not observable.
 15723	 */
 15724	  message m;
 15725	  int key;
 15726	  int proc_nr;
 15727	  int i,s;
 15728	
 15729	  /* Ignore key releases. If this is a key press, get full key code. */
 15730	  if (scode & KEY_RELEASE) return(FALSE);       /* key release */
 15731	  key = map_key(scode);                         /* include modifiers */
 15732	
 15733	  /* Key pressed, now see if there is an observer for the pressed key.
 15734	   *           F1-F12   observers are in fkey_obs array. 
 15735	   *    SHIFT  F1-F12   observers are in sfkey_req array. 
 15736	   *    CTRL   F1-F12   reserved (see kb_read)
 15737	   *    ALT    F1-F12   reserved (see kb_read)
 15738	   * Other combinations are not in use. Note that Alt+Shift+F1-F12 is yet
 15739	   * defined in <minix/keymap.h>, and thus is easy for future extensions.
 15740	   */
 15741	  if (F1 <= key && key <= F12) {                /* F1-F12 */
 15742	      proc_nr = fkey_obs[key - F1].proc_nr;     
 15743	      fkey_obs[key - F1].events ++ ;    
 15744	  } else if (SF1 <= key && key <= SF12) {       /* Shift F2-F12 */
 15745	      proc_nr = sfkey_obs[key - SF1].proc_nr;   
 15746	      sfkey_obs[key - SF1].events ++;   
 15747	  }
 15748	  else {
 15749	      return(FALSE);                            /* not observable */
 15750	  }
 15751	
 15752	  /* See if an observer is registered and send it a message. */
 15753	  if (proc_nr != NONE) { 
 15754	      m.NOTIFY_TYPE = FKEY_PRESSED;
 15755	      notify(proc_nr);
 15756	  }
 15757	  return(TRUE);
 15758	}
	
 15760	/*===========================================================================*
 15761	 *                              show_key_mappings                            *
 15762	 *===========================================================================*/
 15763	PRIVATE void show_key_mappings()
 15764	{

_________________________ Page 847    File: drivers/tty/keyboard.c _______________________
 15765	    int i,s;
 15766	    struct proc proc;
 15767	
 15768	    printf("\n");
 15769	    printf("System information.   Known function key mappings to request debug dump
 15770	    printf("-----------------------------------------------------------------------
 15771	    for (i=0; i<12; i++) {
 15772	
 15773	      printf(" %sF%d: ", i+1<10? " ":"", i+1);
 15774	      if (fkey_obs[i].proc_nr != NONE) {
 15775	          if ((s=sys_getproc(&proc, fkey_obs[i].proc_nr))!=OK)
 15776	              printf("sys_getproc: %d\n", s);
 15777	          printf("%-14.14s", proc.p_name);
 15778	      } else {
 15779	          printf("%-14.14s", "<none>");
 15780	      }
 15781	
 15782	      printf("    %sShift-F%d: ", i+1<10? " ":"", i+1);
 15783	      if (sfkey_obs[i].proc_nr != NONE) {
 15784	          if ((s=sys_getproc(&proc, sfkey_obs[i].proc_nr))!=OK)
 15785	              printf("sys_getproc: %d\n", s);
 15786	          printf("%-14.14s", proc.p_name);
 15787	      } else {
 15788	          printf("%-14.14s", "<none>");
 15789	      }
 15790	      printf("\n");
 15791	    }
 15792	    printf("\n");
 15793	    printf("Press one of the registered function keys to trigger a debug dump.\n");
 15794	    printf("\n");
 15795	}
	
 15797	/*===========================================================================*
 15798	 *                              scan_keyboard                                *
 15799	 *===========================================================================*/
 15800	PRIVATE int scan_keyboard()
 15801	{
 15802	/* Fetch the character from the keyboard hardware and acknowledge it. */
 15803	  pvb_pair_t byte_in[2], byte_out[2];
 15804	  
 15805	  byte_in[0].port = KEYBD;      /* get the scan code for the key struck */
 15806	  byte_in[1].port = PORT_B;     /* strobe the keyboard to ack the char */
 15807	  sys_vinb(byte_in, 2);         /* request actual input */
 15808	
 15809	  pv_set(byte_out[0], PORT_B, byte_in[1].value | KBIT); /* strobe bit high */
 15810	  pv_set(byte_out[1], PORT_B, byte_in[1].value);        /* then strobe low */
 15811	  sys_voutb(byte_out, 2);       /* request actual output */
 15812	
 15813	  return(byte_in[0].value);             /* return scan code */
 15814	}
	
 15816	/*===========================================================================*
 15817	 *                              do_panic_dumps                               *
 15818	 *===========================================================================*/
 15819	PUBLIC void do_panic_dumps(m)
 15820	message *m;                     /* request message to TTY */
 15821	{
 15822	/* Wait for keystrokes for printing debugging info and reboot. */
 15823	  int quiet, code;
 15824	

_________________________ Page 848    File: drivers/tty/keyboard.c _______________________
 15825	  /* A panic! Allow debug dumps until user wants to shutdown. */
 15826	  printf("\nHit ESC to reboot, DEL to shutdown, F-keys for debug dumps\n");
 15827	
 15828	  (void) scan_keyboard();       /* ack any old input */
 15829	  quiet = scan_keyboard();/* quiescent value (0 on PC, last code on AT)*/
 15830	  for (;;) {
 15831	        tickdelay(10);
 15832	        /* See if there are pending request for output, but don't block. 
 15833	         * Diagnostics can span multiple printf()s, so do it in a loop.
 15834	         */
 15835	        while (nb_receive(ANY, m) == OK) {
 15836	                switch(m->m_type) {
 15837	                case FKEY_CONTROL: do_fkey_ctl(m);      break;
 15838	                case SYS_SIG:      do_new_kmess(m);     break;
 15839	                case DIAGNOSTICS:  do_diagnostics(m);   break;
 15840	                default:        ;       /* do nothing */ 
 15841	                }
 15842	                tickdelay(1);           /* allow more */
 15843	        }
 15844	        code = scan_keyboard();
 15845	        if (code != quiet) {
 15846	                /* A key has been pressed. */
 15847	                switch (code) {                 /* possibly abort MINIX */
 15848	                case ESC_SCAN:  sys_abort(RBT_REBOOT);  return; 
 15849	                case DEL_SCAN:  sys_abort(RBT_HALT);    return; 
 15850	                }
 15851	                (void) func_key(code);          /* check for function key */
 15852	                quiet = scan_keyboard();
 15853	        }
 15854	  }
 15855	}
	




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      drivers/tty/console.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 15900	/* Code and data for the IBM console driver.
 15901	 *
 15902	 * The 6845 video controller used by the IBM PC shares its video memory with
 15903	 * the CPU somewhere in the 0xB0000 memory bank.  To the 6845 this memory
 15904	 * consists of 16-bit words.  Each word has a character code in the low byte
 15905	 * and a so-called attribute byte in the high byte.  The CPU directly modifies
 15906	 * video memory to display characters, and sets two registers on the 6845 that
 15907	 * specify the video origin and the cursor position.  The video origin is the
 15908	 * place in video memory where the first character (upper left corner) can
 15909	 * be found.  Moving the origin is a fast way to scroll the screen.  Some
 15910	 * video adapters wrap around the top of video memory, so the origin can
 15911	 * move without bounds.  For other adapters screen memory must sometimes be
 15912	 * moved to reset the origin.  All computations on video memory use character
 15913	 * (word) addresses for simplicity and assume there is no wrapping.  The
 15914	 * assembly support functions translate the word addresses to byte addresses
 15915	 * and the scrolling function worries about wrapping.
 15916	 */
 15917	
 15918	#include "../drivers.h"
 15919	#include <termios.h>

_________________________ Page 849    File: drivers/tty/console.c ________________________
 15920	#include <minix/callnr.h>
 15921	#include <minix/com.h>
 15922	#include "tty.h"
 15923	
 15924	#include "../../kernel/const.h"
 15925	#include "../../kernel/config.h"
 15926	#include "../../kernel/type.h"
 15927	
 15928	/* Definitions used by the console driver. */
 15929	#define MONO_BASE    0xB0000L   /* base of mono video memory */
 15930	#define COLOR_BASE   0xB8000L   /* base of color video memory */
 15931	#define MONO_SIZE     0x1000    /* 4K mono video memory */
 15932	#define COLOR_SIZE    0x4000    /* 16K color video memory */
 15933	#define EGA_SIZE      0x8000    /* EGA & VGA have at least 32K */
 15934	#define BLANK_COLOR   0x0700    /* determines cursor color on blank screen */
 15935	#define SCROLL_UP          0    /* scroll forward */
 15936	#define SCROLL_DOWN        1    /* scroll backward */
 15937	#define BLANK_MEM ((u16_t *) 0) /* tells mem_vid_copy() to blank the screen */
 15938	#define CONS_RAM_WORDS    80    /* video ram buffer size */
 15939	#define MAX_ESC_PARMS      4    /* number of escape sequence params allowed */
 15940	
 15941	/* Constants relating to the controller chips. */
 15942	#define M_6845         0x3B4    /* port for 6845 mono */
 15943	#define C_6845         0x3D4    /* port for 6845 color */
 15944	#define INDEX              0    /* 6845's index register */
 15945	#define DATA               1    /* 6845's data register */
 15946	#define STATUS             6    /* 6845's status register */
 15947	#define VID_ORG           12    /* 6845's origin register */
 15948	#define CURSOR            14    /* 6845's cursor register */
 15949	
 15950	/* Beeper. */
 15951	#define BEEP_FREQ     0x0533    /* value to put into timer to set beep freq */
 15952	#define B_TIME             3    /* length of CTRL-G beep is ticks */
 15953	
 15954	/* definitions used for font management */
 15955	#define GA_SEQUENCER_INDEX      0x3C4
 15956	#define GA_SEQUENCER_DATA       0x3C5
 15957	#define GA_GRAPHICS_INDEX       0x3CE
 15958	#define GA_GRAPHICS_DATA        0x3CF
 15959	#define GA_VIDEO_ADDRESS        0xA0000L
 15960	#define GA_FONT_SIZE            8192
 15961	
 15962	/* Global variables used by the console driver and assembly support. */
 15963	PUBLIC int vid_index;           /* index of video segment in remote mem map */
 15964	PUBLIC u16_t vid_seg;
 15965	PUBLIC vir_bytes vid_off;       /* video ram is found at vid_seg:vid_off */
 15966	PUBLIC unsigned vid_size;       /* 0x2000 for color or 0x0800 for mono */
 15967	PUBLIC unsigned vid_mask;       /* 0x1FFF for color or 0x07FF for mono */
 15968	PUBLIC unsigned blank_color = BLANK_COLOR; /* display code for blank */
 15969	
 15970	/* Private variables used by the console driver. */
 15971	PRIVATE int vid_port;           /* I/O port for accessing 6845 */
 15972	PRIVATE int wrap;               /* hardware can wrap? */
 15973	PRIVATE int softscroll;         /* 1 = software scrolling, 0 = hardware */
 15974	PRIVATE int beeping;            /* speaker is beeping? */
 15975	PRIVATE unsigned font_lines;    /* font lines per character */
 15976	PRIVATE unsigned scr_width;     /* # characters on a line */
 15977	PRIVATE unsigned scr_lines;     /* # lines on the screen */
 15978	PRIVATE unsigned scr_size;      /* # characters on the screen */
 15979	

_________________________ Page 850    File: drivers/tty/console.c ________________________
 15980	/* Per console data. */
 15981	typedef struct console {
 15982	  tty_t *c_tty;                 /* associated TTY struct */
 15983	  int c_column;                 /* current column number (0-origin) */
 15984	  int c_row;                    /* current row (0 at top of screen) */
 15985	  int c_rwords;                 /* number of WORDS (not bytes) in outqueue */
 15986	  unsigned c_start;             /* start of video memory of this console */
 15987	  unsigned c_limit;             /* limit of this console's video memory */
 15988	  unsigned c_org;               /* location in RAM where 6845 base points */
 15989	  unsigned c_cur;               /* current position of cursor in video RAM */
 15990	  unsigned c_attr;              /* character attribute */
 15991	  unsigned c_blank;             /* blank attribute */
 15992	  char c_reverse;               /* reverse video */
 15993	  char c_esc_state;             /* 0=normal, 1=ESC, 2=ESC[ */
 15994	  char c_esc_intro;             /* Distinguishing character following ESC */
 15995	  int *c_esc_parmp;             /* pointer to current escape parameter */
 15996	  int c_esc_parmv[MAX_ESC_PARMS];       /* list of escape parameters */
 15997	  u16_t c_ramqueue[CONS_RAM_WORDS];     /* buffer for video RAM */
 15998	} console_t;
 15999	
 16000	PRIVATE int nr_cons= 1;         /* actual number of consoles */
 16001	PRIVATE console_t cons_table[NR_CONS];
 16002	PRIVATE console_t *curcons;     /* currently visible */
 16003	
 16004	/* Color if using a color controller. */
 16005	#define color   (vid_port == C_6845)
 16006	
 16007	/* Map from ANSI colors to the attributes used by the PC */
 16008	PRIVATE int ansi_colors[8] = {0, 4, 2, 6, 1, 5, 3, 7};
 16009	
 16010	/* Structure used for font management */
 16011	struct sequence {
 16012	        unsigned short index;
 16013	        unsigned char port;
 16014	        unsigned char value;
 16015	};
 16016	
 16017	FORWARD _PROTOTYPE( int cons_write, (struct tty *tp, int try)           );
 16018	FORWARD _PROTOTYPE( void cons_echo, (tty_t *tp, int c)                  );
 16019	FORWARD _PROTOTYPE( void out_char, (console_t *cons, int c)             );
 16020	FORWARD _PROTOTYPE( void putk, (int c)                                  );
 16021	FORWARD _PROTOTYPE( void beep, (void)                                   );
 16022	FORWARD _PROTOTYPE( void do_escape, (console_t *cons, int c)            );
 16023	FORWARD _PROTOTYPE( void flush, (console_t *cons)                       );
 16024	FORWARD _PROTOTYPE( void parse_escape, (console_t *cons, int c)         );
 16025	FORWARD _PROTOTYPE( void scroll_screen, (console_t *cons, int dir)      );
 16026	FORWARD _PROTOTYPE( void set_6845, (int reg, unsigned val)              );
 16027	FORWARD _PROTOTYPE( void get_6845, (int reg, unsigned *val)             );
 16028	FORWARD _PROTOTYPE( void stop_beep, (timer_t *tmrp)                     );
 16029	FORWARD _PROTOTYPE( void cons_org0, (void)                              );
 16030	FORWARD _PROTOTYPE( int ga_program, (struct sequence *seq)              );
 16031	FORWARD _PROTOTYPE( int cons_ioctl, (tty_t *tp, int)                    );
 16032	
 16033	/*===========================================================================*
 16034	 *                              cons_write                                   *
 16035	 *===========================================================================*/
 16036	PRIVATE int cons_write(tp, try)
 16037	register struct tty *tp;        /* tells which terminal is to be used */
 16038	int try;
 16039	{

_________________________ Page 851    File: drivers/tty/console.c ________________________
 16040	/* Copy as much data as possible to the output queue, then start I/O.  On
 16041	 * memory-mapped terminals, such as the IBM console, the I/O will also be
 16042	 * finished, and the counts updated.  Keep repeating until all I/O done.
 16043	 */
 16044	
 16045	  int count;
 16046	  int result;
 16047	  register char *tbuf;
 16048	  char buf[64];
 16049	  console_t *cons = tp->tty_priv;
 16050	
 16051	  if (try) return 1;    /* we can always write to console */
 16052	
 16053	  /* Check quickly for nothing to do, so this can be called often without
 16054	   * unmodular tests elsewhere.
 16055	   */
 16056	  if ((count = tp->tty_outleft) == 0 || tp->tty_inhibited) return;
 16057	
 16058	  /* Copy the user bytes to buf[] for decent addressing. Loop over the
 16059	   * copies, since the user buffer may be much larger than buf[].
 16060	   */
 16061	  do {
 16062	        if (count > sizeof(buf)) count = sizeof(buf);
 16063	        if ((result = sys_vircopy(tp->tty_outproc, D, tp->tty_out_vir, 
 16064	                        SELF, D, (vir_bytes) buf, (vir_bytes) count)) != OK)
 16065	                break;
 16066	        tbuf = buf;
 16067	
 16068	        /* Update terminal data structure. */
 16069	        tp->tty_out_vir += count;
 16070	        tp->tty_outcum += count;
 16071	        tp->tty_outleft -= count;
 16072	
 16073	        /* Output each byte of the copy to the screen.  Avoid calling
 16074	         * out_char() for the "easy" characters, put them into the buffer
 16075	         * directly.
 16076	         */
 16077	        do {
 16078	                if ((unsigned) *tbuf < ' ' || cons->c_esc_state > 0
 16079	                        || cons->c_column >= scr_width
 16080	                        || cons->c_rwords >= buflen(cons->c_ramqueue))
 16081	                {
 16082	                        out_char(cons, *tbuf++);
 16083	                } else {
 16084	                        cons->c_ramqueue[cons->c_rwords++] =
 16085	                                        cons->c_attr | (*tbuf++ & BYTE);
 16086	                        cons->c_column++;
 16087	                }
 16088	        } while (--count != 0);
 16089	  } while ((count = tp->tty_outleft) != 0 && !tp->tty_inhibited);
 16090	
 16091	  flush(cons);                  /* transfer anything buffered to the screen */
 16092	
 16093	  /* Reply to the writer if all output is finished or if an error occured. */
 16094	  if (tp->tty_outleft == 0 || result != OK) {
 16095	        /* REVIVE is not possible. I/O on memory mapped consoles finishes. */
 16096	        tty_reply(tp->tty_outrepcode, tp->tty_outcaller, tp->tty_outproc,
 16097	                                                        tp->tty_outcum);
 16098	        tp->tty_outcum = 0;
 16099	  }

_________________________ Page 852    File: drivers/tty/console.c ________________________
 16100	}
	
 16102	/*===========================================================================*
 16103	 *                              cons_echo                                    *
 16104	 *===========================================================================*/
 16105	PRIVATE void cons_echo(tp, c)
 16106	register tty_t *tp;             /* pointer to tty struct */
 16107	int c;                          /* character to be echoed */
 16108	{
 16109	/* Echo keyboard input (print & flush). */
 16110	  console_t *cons = tp->tty_priv;
 16111	
 16112	  out_char(cons, c);
 16113	  flush(cons);
 16114	}
	
 16116	/*===========================================================================*
 16117	 *                              out_char                                     *
 16118	 *===========================================================================*/
 16119	PRIVATE void out_char(cons, c)
 16120	register console_t *cons;       /* pointer to console struct */
 16121	int c;                          /* character to be output */
 16122	{
 16123	/* Output a character on the console.  Check for escape sequences first. */
 16124	  if (cons->c_esc_state > 0) {
 16125	        parse_escape(cons, c);
 16126	        return;
 16127	  }
 16128	
 16129	  switch(c) {
 16130	        case 000:               /* null is typically used for padding */
 16131	                return;         /* better not do anything */
 16132	
 16133	        case 007:               /* ring the bell */
 16134	                flush(cons);    /* print any chars queued for output */
 16135	                beep();
 16136	                return;
 16137	
 16138	        case '\b':              /* backspace */
 16139	                if (--cons->c_column < 0) {
 16140	                        if (--cons->c_row >= 0) cons->c_column += scr_width;
 16141	                }
 16142	                flush(cons);
 16143	                return;
 16144	
 16145	        case '\n':              /* line feed */
 16146	                if ((cons->c_tty->tty_termios.c_oflag & (OPOST|ONLCR))
 16147	                                                == (OPOST|ONLCR)) {
 16148	                        cons->c_column = 0;
 16149	                }
 16150	                /*FALL THROUGH*/
 16151	        case 013:               /* CTRL-K */
 16152	        case 014:               /* CTRL-L */
 16153	                if (cons->c_row == scr_lines-1) {
 16154	                        scroll_screen(cons, SCROLL_UP);
 16155	                } else {
 16156	                        cons->c_row++;
 16157	                }
 16158	                flush(cons);
 16159	                return;

_________________________ Page 853    File: drivers/tty/console.c ________________________
 16160	
 16161	        case '\r':              /* carriage return */
 16162	                cons->c_column = 0;
 16163	                flush(cons);
 16164	                return;
 16165	
 16166	        case '\t':              /* tab */
 16167	                cons->c_column = (cons->c_column + TAB_SIZE) & ~TAB_MASK;
 16168	                if (cons->c_column > scr_width) {
 16169	                        cons->c_column -= scr_width;
 16170	                        if (cons->c_row == scr_lines-1) {
 16171	                                scroll_screen(cons, SCROLL_UP);
 16172	                        } else {
 16173	                                cons->c_row++;
 16174	                        }
 16175	                }
 16176	                flush(cons);
 16177	                return;
 16178	
 16179	        case 033:               /* ESC - start of an escape sequence */
 16180	                flush(cons);    /* print any chars queued for output */
 16181	                cons->c_esc_state = 1;  /* mark ESC as seen */
 16182	                return;
 16183	
 16184	        default:                /* printable chars are stored in ramqueue */
 16185	                if (cons->c_column >= scr_width) {
 16186	                        if (!LINEWRAP) return;
 16187	                        if (cons->c_row == scr_lines-1) {
 16188	                                scroll_screen(cons, SCROLL_UP);
 16189	                        } else {
 16190	                                cons->c_row++;
 16191	                        }
 16192	                        cons->c_column = 0;
 16193	                        flush(cons);
 16194	                }
 16195	                if (cons->c_rwords == buflen(cons->c_ramqueue)) flush(cons);
 16196	                cons->c_ramqueue[cons->c_rwords++] = cons->c_attr | (c & BYTE);
 16197	                cons->c_column++;                       /* next column */
 16198	                return;
 16199	  }
 16200	}
	
 16202	/*===========================================================================*
 16203	 *                              scroll_screen                                *
 16204	 *===========================================================================*/
 16205	PRIVATE void scroll_screen(cons, dir)
 16206	register console_t *cons;       /* pointer to console struct */
 16207	int dir;                        /* SCROLL_UP or SCROLL_DOWN */
 16208	{
 16209	  unsigned new_line, new_org, chars;
 16210	
 16211	  flush(cons);
 16212	  chars = scr_size - scr_width;         /* one screen minus one line */
 16213	
 16214	  /* Scrolling the screen is a real nuisance due to the various incompatible
 16215	   * video cards.  This driver supports software scrolling (Hercules?),
 16216	   * hardware scrolling (mono and CGA cards) and hardware scrolling without
 16217	   * wrapping (EGA cards).  In the latter case we must make sure that
 16218	   *            c_start <= c_org && c_org + scr_size <= c_limit
 16219	   * holds, because EGA doesn't wrap around the end of video memory.

_________________________ Page 854    File: drivers/tty/console.c ________________________
 16220	   */
 16221	  if (dir == SCROLL_UP) {
 16222	        /* Scroll one line up in 3 ways: soft, avoid wrap, use origin. */
 16223	        if (softscroll) {
 16224	                vid_vid_copy(cons->c_start + scr_width, cons->c_start, chars);
 16225	        } else
 16226	        if (!wrap && cons->c_org + scr_size + scr_width >= cons->c_limit) {
 16227	                vid_vid_copy(cons->c_org + scr_width, cons->c_start, chars);
 16228	                cons->c_org = cons->c_start;
 16229	        } else {
 16230	                cons->c_org = (cons->c_org + scr_width) & vid_mask;
 16231	        }
 16232	        new_line = (cons->c_org + chars) & vid_mask;
 16233	  } else {
 16234	        /* Scroll one line down in 3 ways: soft, avoid wrap, use origin. */
 16235	        if (softscroll) {
 16236	                vid_vid_copy(cons->c_start, cons->c_start + scr_width, chars);
 16237	        } else
 16238	        if (!wrap && cons->c_org < cons->c_start + scr_width) {
 16239	                new_org = cons->c_limit - scr_size;
 16240	                vid_vid_copy(cons->c_org, new_org + scr_width, chars);
 16241	                cons->c_org = new_org;
 16242	        } else {
 16243	                cons->c_org = (cons->c_org - scr_width) & vid_mask;
 16244	        }
 16245	        new_line = cons->c_org;
 16246	  }
 16247	  /* Blank the new line at top or bottom. */
 16248	  blank_color = cons->c_blank;
 16249	  mem_vid_copy(BLANK_MEM, new_line, scr_width);
 16250	
 16251	  /* Set the new video origin. */
 16252	  if (cons == curcons) set_6845(VID_ORG, cons->c_org);
 16253	  flush(cons);
 16254	}
	
 16256	/*===========================================================================*
 16257	 *                              flush                                        *
 16258	 *===========================================================================*/
 16259	PRIVATE void flush(cons)
 16260	register console_t *cons;       /* pointer to console struct */
 16261	{
 16262	/* Send characters buffered in 'ramqueue' to screen memory, check the new
 16263	 * cursor position, compute the new hardware cursor position and set it.
 16264	 */
 16265	  unsigned cur;
 16266	  tty_t *tp = cons->c_tty;
 16267	
 16268	  /* Have the characters in 'ramqueue' transferred to the screen. */
 16269	  if (cons->c_rwords > 0) {
 16270	        mem_vid_copy(cons->c_ramqueue, cons->c_cur, cons->c_rwords);
 16271	        cons->c_rwords = 0;
 16272	
 16273	        /* TTY likes to know the current column and if echoing messed up. */
 16274	        tp->tty_position = cons->c_column;
 16275	        tp->tty_reprint = TRUE;
 16276	  }
 16277	
 16278	  /* Check and update the cursor position. */
 16279	  if (cons->c_column < 0) cons->c_column = 0;

_________________________ Page 855    File: drivers/tty/console.c ________________________
 16280	  if (cons->c_column > scr_width) cons->c_column = scr_width;
 16281	  if (cons->c_row < 0) cons->c_row = 0;
 16282	  if (cons->c_row >= scr_lines) cons->c_row = scr_lines - 1;
 16283	  cur = cons->c_org + cons->c_row * scr_width + cons->c_column;
 16284	  if (cur != cons->c_cur) {
 16285	        if (cons == curcons) set_6845(CURSOR, cur);
 16286	        cons->c_cur = cur;
 16287	  }
 16288	}
	
 16290	/*===========================================================================*
 16291	 *                              parse_escape                                 *
 16292	 *===========================================================================*/
 16293	PRIVATE void parse_escape(cons, c)
 16294	register console_t *cons;       /* pointer to console struct */
 16295	char c;                         /* next character in escape sequence */
 16296	{
 16297	/* The following ANSI escape sequences are currently supported.
 16298	 * If n and/or m are omitted, they default to 1.
 16299	 *   ESC [nA moves up n lines
 16300	 *   ESC [nB moves down n lines
 16301	 *   ESC [nC moves right n spaces
 16302	 *   ESC [nD moves left n spaces
 16303	 *   ESC [m;nH" moves cursor to (m,n)
 16304	 *   ESC [J clears screen from cursor
 16305	 *   ESC [K clears line from cursor
 16306	 *   ESC [nL inserts n lines ar cursor
 16307	 *   ESC [nM deletes n lines at cursor
 16308	 *   ESC [nP deletes n chars at cursor
 16309	 *   ESC [n@ inserts n chars at cursor
 16310	 *   ESC [nm enables rendition n (0=normal, 4=bold, 5=blinking, 7=reverse)
 16311	 *   ESC M scrolls the screen backwards if the cursor is on the top line
 16312	 */
 16313	
 16314	  switch (cons->c_esc_state) {
 16315	    case 1:                     /* ESC seen */
 16316	        cons->c_esc_intro = '\0';
 16317	        cons->c_esc_parmp = bufend(cons->c_esc_parmv);
 16318	        do {
 16319	                *--cons->c_esc_parmp = 0;
 16320	        } while (cons->c_esc_parmp > cons->c_esc_parmv);
 16321	        switch (c) {
 16322	            case '[':   /* Control Sequence Introducer */
 16323	                cons->c_esc_intro = c;
 16324	                cons->c_esc_state = 2;
 16325	                break;
 16326	            case 'M':   /* Reverse Index */
 16327	                do_escape(cons, c);
 16328	                break;
 16329	            default:
 16330	                cons->c_esc_state = 0;
 16331	        }
 16332	        break;
 16333	
 16334	    case 2:                     /* ESC [ seen */
 16335	        if (c >= '0' && c <= '9') {
 16336	                if (cons->c_esc_parmp < bufend(cons->c_esc_parmv))
 16337	                        *cons->c_esc_parmp = *cons->c_esc_parmp * 10 + (c-'0');
 16338	        } else
 16339	        if (c == ';') {

_________________________ Page 856    File: drivers/tty/console.c ________________________
 16340	                if (cons->c_esc_parmp < bufend(cons->c_esc_parmv))
 16341	                        cons->c_esc_parmp++;
 16342	        } else {
 16343	                do_escape(cons, c);
 16344	        }
 16345	        break;
 16346	  }
 16347	}
	
 16349	/*===========================================================================*
 16350	 *                              do_escape                                    *
 16351	 *===========================================================================*/
 16352	PRIVATE void do_escape(cons, c)
 16353	register console_t *cons;       /* pointer to console struct */
 16354	char c;                         /* next character in escape sequence */
 16355	{
 16356	  int value, n;
 16357	  unsigned src, dst, count;
 16358	  int *parmp;
 16359	
 16360	  /* Some of these things hack on screen RAM, so it had better be up to date */
 16361	  flush(cons);
 16362	
 16363	  if (cons->c_esc_intro == '\0') {
 16364	        /* Handle a sequence beginning with just ESC */
 16365	        switch (c) {
 16366	            case 'M':           /* Reverse Index */
 16367	                if (cons->c_row == 0) {
 16368	                        scroll_screen(cons, SCROLL_DOWN);
 16369	                } else {
 16370	                        cons->c_row--;
 16371	                }
 16372	                flush(cons);
 16373	                break;
 16374	
 16375	            default: break;
 16376	        }
 16377	  } else
 16378	  if (cons->c_esc_intro == '[') {
 16379	        /* Handle a sequence beginning with ESC [ and parameters */
 16380	        value = cons->c_esc_parmv[0];
 16381	        switch (c) {
 16382	            case 'A':           /* ESC [nA moves up n lines */
 16383	                n = (value == 0 ? 1 : value);
 16384	                cons->c_row -= n;
 16385	                flush(cons);
 16386	                break;
 16387	
 16388	            case 'B':           /* ESC [nB moves down n lines */
 16389	                n = (value == 0 ? 1 : value);
 16390	                cons->c_row += n;
 16391	                flush(cons);
 16392	                break;
 16393	
 16394	            case 'C':           /* ESC [nC moves right n spaces */
 16395	                n = (value == 0 ? 1 : value);
 16396	                cons->c_column += n;
 16397	                flush(cons);
 16398	                break;
 16399	

_________________________ Page 857    File: drivers/tty/console.c ________________________
 16400	            case 'D':           /* ESC [nD moves left n spaces */
 16401	                n = (value == 0 ? 1 : value);
 16402	                cons->c_column -= n;
 16403	                flush(cons);
 16404	                break;
 16405	
 16406	            case 'H':           /* ESC [m;nH" moves cursor to (m,n) */
 16407	                cons->c_row = cons->c_esc_parmv[0] - 1;
 16408	                cons->c_column = cons->c_esc_parmv[1] - 1;
 16409	                flush(cons);
 16410	                break;
 16411	
 16412	            case 'J':           /* ESC [sJ clears in display */
 16413	                switch (value) {
 16414	                    case 0:     /* Clear from cursor to end of screen */
 16415	                        count = scr_size - (cons->c_cur - cons->c_org);
 16416	                        dst = cons->c_cur;
 16417	                        break;
 16418	                    case 1:     /* Clear from start of screen to cursor */
 16419	                        count = cons->c_cur - cons->c_org;
 16420	                        dst = cons->c_org;
 16421	                        break;
 16422	                    case 2:     /* Clear entire screen */
 16423	                        count = scr_size;
 16424	                        dst = cons->c_org;
 16425	                        break;
 16426	                    default:    /* Do nothing */
 16427	                        count = 0;
 16428	                        dst = cons->c_org;
 16429	                }
 16430	                blank_color = cons->c_blank;
 16431	                mem_vid_copy(BLANK_MEM, dst, count);
 16432	                break;
 16433	
 16434	            case 'K':           /* ESC [sK clears line from cursor */
 16435	                switch (value) {
 16436	                    case 0:     /* Clear from cursor to end of line */
 16437	                        count = scr_width - cons->c_column;
 16438	                        dst = cons->c_cur;
 16439	                        break;
 16440	                    case 1:     /* Clear from beginning of line to cursor */
 16441	                        count = cons->c_column;
 16442	                        dst = cons->c_cur - cons->c_column;
 16443	                        break;
 16444	                    case 2:     /* Clear entire line */
 16445	                        count = scr_width;
 16446	                        dst = cons->c_cur - cons->c_column;
 16447	                        break;
 16448	                    default:    /* Do nothing */
 16449	                        count = 0;
 16450	                        dst = cons->c_cur;
 16451	                }
 16452	                blank_color = cons->c_blank;
 16453	                mem_vid_copy(BLANK_MEM, dst, count);
 16454	                break;
 16455	
 16456	            case 'L':           /* ESC [nL inserts n lines at cursor */
 16457	                n = value;
 16458	                if (n < 1) n = 1;
 16459	                if (n > (scr_lines - cons->c_row))

_________________________ Page 858    File: drivers/tty/console.c ________________________
 16460	                        n = scr_lines - cons->c_row;
 16461	
 16462	                src = cons->c_org + cons->c_row * scr_width;
 16463	                dst = src + n * scr_width;
 16464	                count = (scr_lines - cons->c_row - n) * scr_width;
 16465	                vid_vid_copy(src, dst, count);
 16466	                blank_color = cons->c_blank;
 16467	                mem_vid_copy(BLANK_MEM, src, n * scr_width);
 16468	                break;
 16469	
 16470	            case 'M':           /* ESC [nM deletes n lines at cursor */
 16471	                n = value;
 16472	                if (n < 1) n = 1;
 16473	                if (n > (scr_lines - cons->c_row))
 16474	                        n = scr_lines - cons->c_row;
 16475	
 16476	                dst = cons->c_org + cons->c_row * scr_width;
 16477	                src = dst + n * scr_width;
 16478	                count = (scr_lines - cons->c_row - n) * scr_width;
 16479	                vid_vid_copy(src, dst, count);
 16480	                blank_color = cons->c_blank;
 16481	                mem_vid_copy(BLANK_MEM, dst + count, n * scr_width);
 16482	                break;
 16483	
 16484	            case '@':           /* ESC [n@ inserts n chars at cursor */
 16485	                n = value;
 16486	                if (n < 1) n = 1;
 16487	                if (n > (scr_width - cons->c_column))
 16488	                        n = scr_width - cons->c_column;
 16489	
 16490	                src = cons->c_cur;
 16491	                dst = src + n;
 16492	                count = scr_width - cons->c_column - n;
 16493	                vid_vid_copy(src, dst, count);
 16494	                blank_color = cons->c_blank;
 16495	                mem_vid_copy(BLANK_MEM, src, n);
 16496	                break;
 16497	
 16498	            case 'P':           /* ESC [nP deletes n chars at cursor */
 16499	                n = value;
 16500	                if (n < 1) n = 1;
 16501	                if (n > (scr_width - cons->c_column))
 16502	                        n = scr_width - cons->c_column;
 16503	
 16504	                dst = cons->c_cur;
 16505	                src = dst + n;
 16506	                count = scr_width - cons->c_column - n;
 16507	                vid_vid_copy(src, dst, count);
 16508	                blank_color = cons->c_blank;
 16509	                mem_vid_copy(BLANK_MEM, dst + count, n);
 16510	                break;
 16511	
 16512	            case 'm':           /* ESC [nm enables rendition n */
 16513	                for (parmp = cons->c_esc_parmv; parmp <= cons->c_esc_parmp
 16514	                                && parmp < bufend(cons->c_esc_parmv); pa
 16515	                        if (cons->c_reverse) {
 16516	                                /* Unswap fg and bg colors */
 16517	                                cons->c_attr =  ((cons->c_attr & 0x7000) >> 4) 
 16518	                                                ((cons->c_attr & 0x0700) << ...
 16519	                                                ((cons->c_attr & 0x8800));

_________________________ Page 859    File: drivers/tty/console.c ________________________
 16520	                        }
 16521	                        switch (n = *parmp) {
 16522	                            case 0:     /* NORMAL */
 16523	                                cons->c_attr = cons->c_blank = BLANK_COLOR;
 16524	                                cons->c_reverse = FALSE;
 16525	                                break;
 16526	
 16527	                            case 1:     /* BOLD  */
 16528	                                /* Set intensity bit */
 16529	                                cons->c_attr |= 0x0800;
 16530	                                break;
 16531	
 16532	                            case 4:     /* UNDERLINE */
 16533	                                if (color) {
 16534	                                        /* Change white to cyan, i.e. lose red
 16535	                                         */
 16536	                                        cons->c_attr = (cons->c_attr & 0xBBFF);
 16537	                                } else {
 16538	                                        /* Set underline attribute */
 16539	                                        cons->c_attr = (cons->c_attr & 0x99FF);
 16540	                                }
 16541	                                break;
 16542	
 16543	                            case 5:     /* BLINKING */
 16544	                                /* Set the blink bit */
 16545	                                cons->c_attr |= 0x8000;
 16546	                                break;
 16547	
 16548	                            case 7:     /* REVERSE */
 16549	                                cons->c_reverse = TRUE;
 16550	                                break;
 16551	
 16552	                            default:    /* COLOR */
 16553	                                if (n == 39) n = 37;    /* set default color */
 16554	                                if (n == 49) n = 40;
 16555	
 16556	                                if (!color) {
 16557	                                        /* Don't mess up a monochrome screen */
 16558	                                } else
 16559	                                if (30 <= n && n <= 37) {
 16560	                                        /* Foreground color */
 16561	                                        cons->c_attr =
 16562	                                                (cons->c_attr & 0xF8FF) |
 16563	                                                (ansi_colors[(n - 30)] << 8);
 16564	                                        cons->c_blank =
 16565	                                                (cons->c_blank & 0xF8FF) |
 16566	                                                (ansi_colors[(n - 30)] << 8);
 16567	                                } else
 16568	                                if (40 <= n && n <= 47) {
 16569	                                        /* Background color */
 16570	                                        cons->c_attr =
 16571	                                                (cons->c_attr & 0x8FFF) |
 16572	                                                (ansi_colors[(n - 40)] << 12)
 16573	                                        cons->c_blank =
 16574	                                                (cons->c_blank & 0x8FFF) |
 16575	                                                (ansi_colors[(n - 40)] << 12)
 16576	                                }
 16577	                        }
 16578	                        if (cons->c_reverse) {
 16579	                                /* Swap fg and bg colors */

_________________________ Page 860    File: drivers/tty/console.c ________________________
 16580	                                cons->c_attr =  ((cons->c_attr & 0x7000) >> 4) 
 16581	                                                ((cons->c_attr & 0x0700) << ...
 16582	                                                ((cons->c_attr & 0x8800));
 16583	                        }
 16584	                }
 16585	                break;
 16586	        }
 16587	  }
 16588	  cons->c_esc_state = 0;
 16589	}
	
 16591	/*===========================================================================*
 16592	 *                              set_6845                                     *
 16593	 *===========================================================================*/
 16594	PRIVATE void set_6845(reg, val)
 16595	int reg;                        /* which register pair to set */
 16596	unsigned val;                   /* 16-bit value to set it to */
 16597	{
 16598	/* Set a register pair inside the 6845.
 16599	 * Registers 12-13 tell the 6845 where in video ram to start
 16600	 * Registers 14-15 tell the 6845 where to put the cursor
 16601	 */
 16602	  pvb_pair_t char_out[4];
 16603	  pv_set(char_out[0], vid_port + INDEX, reg);   /* set index register */
 16604	  pv_set(char_out[1], vid_port + DATA, (val>>8) & BYTE);    /* high byte */
 16605	  pv_set(char_out[2], vid_port + INDEX, reg + 1);           /* again */
 16606	  pv_set(char_out[3], vid_port + DATA, val&BYTE);           /* low byte */
 16607	  sys_voutb(char_out, 4);                       /* do actual output */
 16608	}
	
 16610	/*===========================================================================*
 16611	 *                              get_6845                                     *
 16612	 *===========================================================================*/
 16613	PRIVATE void get_6845(reg, val)
 16614	int reg;                        /* which register pair to set */
 16615	unsigned *val;                  /* 16-bit value to set it to */
 16616	{
 16617	  char v1, v2;
 16618	/* Get a register pair inside the 6845.  */
 16619	  sys_outb(vid_port + INDEX, reg); 
 16620	  sys_inb(vid_port + DATA, &v1); 
 16621	  sys_outb(vid_port + INDEX, reg+1); 
 16622	  sys_inb(vid_port + DATA, &v2); 
 16623	  *val = (v1 << 8) | v2;
 16624	}
	
 16626	/*===========================================================================*
 16627	 *                              beep                                         *
 16628	 *===========================================================================*/
 16629	PRIVATE void beep()
 16630	{
 16631	/* Making a beeping sound on the speaker (output for CRTL-G).
 16632	 * This routine works by turning on the bits 0 and 1 in port B of the 8255
 16633	 * chip that drive the speaker.
 16634	 */
 16635	  static timer_t tmr_stop_beep;
 16636	  pvb_pair_t char_out[3];
 16637	  clock_t now;
 16638	  int port_b_val, s;
 16639	  

_________________________ Page 861    File: drivers/tty/console.c ________________________
 16640	  /* Fetch current time in advance to prevent beeping delay. */
 16641	  if ((s=getuptime(&now)) != OK)
 16642	        panic("TTY","Console couldn't get clock's uptime.", s);
 16643	  if (!beeping) {
 16644	        /* Set timer channel 2, square wave, with given frequency. */
 16645	        pv_set(char_out[0], TIMER_MODE, 0xB6);  
 16646	        pv_set(char_out[1], TIMER2, (BEEP_FREQ >> 0) & BYTE);
 16647	        pv_set(char_out[2], TIMER2, (BEEP_FREQ >> 8) & BYTE);
 16648	        if (sys_voutb(char_out, 3)==OK) {
 16649	                if (sys_inb(PORT_B, &port_b_val)==OK &&
 16650	                    sys_outb(PORT_B, (port_b_val|3))==OK)
 16651	                        beeping = TRUE;
 16652	        }
 16653	  }
 16654	  /* Add a timer to the timers list. Possibly reschedule the alarm. */
 16655	  tmrs_settimer(&tty_timers, &tmr_stop_beep, now+B_TIME, stop_beep, NULL);
 16656	  if (tty_timers->tmr_exp_time != tty_next_timeout) {
 16657	        tty_next_timeout = tty_timers->tmr_exp_time;
 16658	        if ((s=sys_setalarm(tty_next_timeout, 1)) != OK)
 16659	                panic("TTY","Console couldn't set alarm.", s);
 16660	  }
 16661	}
	
 16663	/*===========================================================================*
 16664	 *                              stop_beep                                    *
 16665	 *===========================================================================*/
 16666	PRIVATE void stop_beep(tmrp)
 16667	timer_t *tmrp;
 16668	{
 16669	/* Turn off the beeper by turning off bits 0 and 1 in PORT_B. */
 16670	  int port_b_val;
 16671	  if (sys_inb(PORT_B, &port_b_val)==OK && 
 16672	        sys_outb(PORT_B, (port_b_val & ~3))==OK)
 16673	                beeping = FALSE;
 16674	}
	
 16676	/*===========================================================================*
 16677	 *                              scr_init                                     *
 16678	 *===========================================================================*/
 16679	PUBLIC void scr_init(tp)
 16680	tty_t *tp;
 16681	{
 16682	/* Initialize the screen driver. */
 16683	  console_t *cons;
 16684	  phys_bytes vid_base;
 16685	  u16_t bios_columns, bios_crtbase, bios_fontlines;
 16686	  u8_t bios_rows;
 16687	  int line;
 16688	  int s;
 16689	  static int vdu_initialized = 0;
 16690	  unsigned page_size;
 16691	
 16692	  /* Associate console and TTY. */
 16693	  line = tp - &tty_table[0];
 16694	  if (line >= nr_cons) return;
 16695	  cons = &cons_table[line];
 16696	  cons->c_tty = tp;
 16697	  tp->tty_priv = cons;
 16698	
 16699	  /* Initialize the keyboard driver. */

_________________________ Page 862    File: drivers/tty/console.c ________________________
 16700	  kb_init(tp);
 16701	
 16702	  /* Fill in TTY function hooks. */
 16703	  tp->tty_devwrite = cons_write;
 16704	  tp->tty_echo = cons_echo;
 16705	  tp->tty_ioctl = cons_ioctl;
 16706	
 16707	  /* Get the BIOS parameters that describe the VDU. */
 16708	  if (! vdu_initialized++) {
 16709	
 16710	        /* How about error checking? What to do on failure??? */
 16711	        s=sys_vircopy(SELF, BIOS_SEG, (vir_bytes) VDU_SCREEN_COLS_ADDR,
 16712	                SELF, D, (vir_bytes) &bios_columns, VDU_SCREEN_COLS_SIZE);
 16713	        s=sys_vircopy(SELF, BIOS_SEG, (vir_bytes) VDU_CRT_BASE_ADDR, 
 16714	                SELF, D, (vir_bytes) &bios_crtbase, VDU_CRT_BASE_SIZE);
 16715	        s=sys_vircopy(SELF, BIOS_SEG, (vir_bytes) VDU_SCREEN_ROWS_ADDR, 
 16716	                SELF, D, (vir_bytes) &bios_rows, VDU_SCREEN_ROWS_SIZE);
 16717	        s=sys_vircopy(SELF, BIOS_SEG, (vir_bytes) VDU_FONTLINES_ADDR, 
 16718	                SELF, D, (vir_bytes) &bios_fontlines, VDU_FONTLINES_SIZE);
 16719	
 16720	        vid_port = bios_crtbase;
 16721	        scr_width = bios_columns;
 16722	        font_lines = bios_fontlines;
 16723	        scr_lines = machine.vdu_ega ? bios_rows+1 : 25;
 16724	
 16725	        if (color) {
 16726	                vid_base = COLOR_BASE;
 16727	                vid_size = COLOR_SIZE;
 16728	        } else {
 16729	                vid_base = MONO_BASE;
 16730	                vid_size = MONO_SIZE;
 16731	        }
 16732	        if (machine.vdu_ega) vid_size = EGA_SIZE;
 16733	        wrap = ! machine.vdu_ega;
 16734	
 16735	        s = sys_segctl(&vid_index, &vid_seg, &vid_off, vid_base, vid_si
 16736	
 16737	        vid_size >>= 1;         /* word count */
 16738	        vid_mask = vid_size - 1;
 16739	
 16740	        /* Size of the screen (number of displayed characters.) */
 16741	        scr_size = scr_lines * scr_width;
 16742	
 16743	        /* There can be as many consoles as video memory allows. */
 16744	        nr_cons = vid_size / scr_size;
 16745	        if (nr_cons > NR_CONS) nr_cons = NR_CONS;
 16746	        if (nr_cons > 1) wrap = 0;
 16747	        page_size = vid_size / nr_cons;
 16748	  }
 16749	
 16750	  cons->c_start = line * page_size;
 16751	  cons->c_limit = cons->c_start + page_size;
 16752	  cons->c_cur = cons->c_org = cons->c_start;
 16753	  cons->c_attr = cons->c_blank = BLANK_COLOR;
 16754	
 16755	  if (line != 0) {
 16756	        /* Clear the non-console vtys. */
 16757	        blank_color = BLANK_COLOR;
 16758	        mem_vid_copy(BLANK_MEM, cons->c_start, scr_size);
 16759	  } else {

_________________________ Page 863    File: drivers/tty/console.c ________________________
 16760	        int i, n;
 16761	        /* Set the cursor of the console vty at the bottom. c_cur
 16762	         * is updated automatically later.
 16763	         */
 16764	        scroll_screen(cons, SCROLL_UP);
 16765	        cons->c_row = scr_lines - 1;
 16766	        cons->c_column = 0;
 16767	  }
 16768	  select_console(0);
 16769	  cons_ioctl(tp, 0);
 16770	}
	
 16772	/*===========================================================================*
 16773	 *                              kputc                                        *
 16774	 *===========================================================================*/
 16775	PUBLIC void kputc(c)
 16776	int c;
 16777	{
 16778	        putk(c);
 16779	}
	
 16781	/*===========================================================================*
 16782	 *                              do_new_kmess                                 *
 16783	 *===========================================================================*/
 16784	PUBLIC void do_new_kmess(m)
 16785	message *m;
 16786	{
 16787	/* Notification for a new kernel message. */
 16788	  struct kmessages kmess;                       /* kmessages structure */
 16789	  static int prev_next = 0;                     /* previous next seen */
 16790	  int size, next;
 16791	  int bytes;
 16792	  int r;
 16793	
 16794	  /* Try to get a fresh copy of the buffer with kernel messages. */
 16795	  sys_getkmessages(&kmess);
 16796	
 16797	  /* Print only the new part. Determine how many new bytes there are with 
 16798	   * help of the current and previous 'next' index. Note that the kernel
 16799	   * buffer is circular. This works fine if less then KMESS_BUF_SIZE bytes
 16800	   * is new data; else we miss % KMESS_BUF_SIZE here.  
 16801	   * Check for size being positive, the buffer might as well be emptied!
 16802	   */
 16803	  if (kmess.km_size > 0) {
 16804	      bytes = ((kmess.km_next + KMESS_BUF_SIZE) - prev_next) % KMESS_BUF_SIZE;
 16805	      r=prev_next;                              /* start at previous old */ 
 16806	      while (bytes > 0) {                       
 16807	          putk( kmess.km_buf[(r%KMESS_BUF_SIZE)] );
 16808	          bytes --;
 16809	          r ++;
 16810	      }
 16811	      putk(0);                  /* terminate to flush output */
 16812	  }
 16813	
 16814	  /* Almost done, store 'next' so that we can determine what part of the
 16815	   * kernel messages buffer to print next time a notification arrives.
 16816	   */
 16817	  prev_next = kmess.km_next;
 16818	}
	

_________________________ Page 864    File: drivers/tty/console.c ________________________
 16820	/*===========================================================================*
 16821	 *                              do_diagnostics                               *
 16822	 *===========================================================================*/
 16823	PUBLIC void do_diagnostics(m_ptr)
 16824	message *m_ptr;                 /* pointer to request message */
 16825	{
 16826	/* Print a string for a server. */
 16827	  char c;
 16828	  vir_bytes src;
 16829	  int count;
 16830	  int result = OK;
 16831	  int proc_nr = m_ptr->DIAG_PROC_NR;
 16832	  if (proc_nr == SELF) proc_nr = m_ptr->m_source;
 16833	
 16834	  src = (vir_bytes) m_ptr->DIAG_PRINT_BUF;
 16835	  for (count = m_ptr->DIAG_BUF_COUNT; count > 0; count--) {
 16836	        if (sys_vircopy(proc_nr, D, src++, SELF, D, (vir_bytes) &c, 1) != OK) {
 16837	                result = EFAULT;
 16838	                break;
 16839	        }
 16840	        putk(c);
 16841	  }
 16842	  putk(0);                      /* always terminate, even with EFAULT */
 16843	  m_ptr->m_type = result;
 16844	  send(m_ptr->m_source, m_ptr);
 16845	}
	
 16847	/*===========================================================================*
 16848	 *                              putk                                         *
 16849	 *===========================================================================*/
 16850	PRIVATE void putk(c)
 16851	int c;                          /* character to print */
 16852	{
 16853	/* This procedure is used by the version of printf() that is linked with
 16854	 * the TTY driver.  The one in the library sends a message to FS, which is
 16855	 * not what is needed for printing within the TTY. This version just queues
 16856	 * the character and starts the output.
 16857	 */
 16858	  if (c != 0) {
 16859	        if (c == '\n') putk('\r');
 16860	        out_char(&cons_table[0], (int) c);
 16861	  } else {
 16862	        flush(&cons_table[0]);
 16863	  }
 16864	}
	
 16866	/*===========================================================================*
 16867	 *                              toggle_scroll                                *
 16868	 *===========================================================================*/
 16869	PUBLIC void toggle_scroll()
 16870	{
 16871	/* Toggle between hardware and software scroll. */
 16872	
 16873	  cons_org0();
 16874	  softscroll = !softscroll;
 16875	  printf("%sware scrolling enabled.\n", softscroll ? "Soft" : "Hard");
 16876	}
	

_________________________ Page 865    File: drivers/tty/console.c ________________________
 16878	/*===========================================================================*
 16879	 *                              cons_stop                                    *
 16880	 *===========================================================================*/
 16881	PUBLIC void cons_stop()
 16882	{
 16883	/* Prepare for halt or reboot. */
 16884	  cons_org0();
 16885	  softscroll = 1;
 16886	  select_console(0);
 16887	  cons_table[0].c_attr = cons_table[0].c_blank = BLANK_COLOR;
 16888	}
	
 16890	/*===========================================================================*
 16891	 *                              cons_org0                                    *
 16892	 *===========================================================================*/
 16893	PRIVATE void cons_org0()
 16894	{
 16895	/* Scroll video memory back to put the origin at 0. */
 16896	  int cons_line;
 16897	  console_t *cons;
 16898	  unsigned n;
 16899	
 16900	  for (cons_line = 0; cons_line < nr_cons; cons_line++) {
 16901	        cons = &cons_table[cons_line];
 16902	        while (cons->c_org > cons->c_start) {
 16903	                n = vid_size - scr_size;        /* amount of unused memory */
 16904	                if (n > cons->c_org - cons->c_start)
 16905	                        n = cons->c_org - cons->c_start;
 16906	                vid_vid_copy(cons->c_org, cons->c_org - n, scr_size);
 16907	                cons->c_org -= n;
 16908	        }
 16909	        flush(cons);
 16910	  }
 16911	  select_console(ccurrent);
 16912	}
	
 16914	/*===========================================================================*
 16915	 *                              select_console                               *
 16916	 *===========================================================================*/
 16917	PUBLIC void select_console(int cons_line)
 16918	{
 16919	/* Set the current console to console number 'cons_line'. */
 16920	
 16921	  if (cons_line < 0 || cons_line >= nr_cons) return;
 16922	  ccurrent = cons_line;
 16923	  curcons = &cons_table[cons_line];
 16924	  set_6845(VID_ORG, curcons->c_org);
 16925	  set_6845(CURSOR, curcons->c_cur);
 16926	}
	
 16928	/*===========================================================================*
 16929	 *                              con_loadfont                                 *
 16930	 *===========================================================================*/
 16931	PUBLIC int con_loadfont(m)
 16932	message *m;
 16933	{
 16934	/* Load a font into the EGA or VGA adapter. */
 16935	  int result;
 16936	  static struct sequence seq1[7] = {
 16937	        { GA_SEQUENCER_INDEX, 0x00, 0x01 },

_________________________ Page 866    File: drivers/tty/console.c ________________________
 16938	        { GA_SEQUENCER_INDEX, 0x02, 0x04 },
 16939	        { GA_SEQUENCER_INDEX, 0x04, 0x07 },
 16940	        { GA_SEQUENCER_INDEX, 0x00, 0x03 },
 16941	        { GA_GRAPHICS_INDEX, 0x04, 0x02 },
 16942	        { GA_GRAPHICS_INDEX, 0x05, 0x00 },
 16943	        { GA_GRAPHICS_INDEX, 0x06, 0x00 },
 16944	  };
 16945	  static struct sequence seq2[7] = {
 16946	        { GA_SEQUENCER_INDEX, 0x00, 0x01 },
 16947	        { GA_SEQUENCER_INDEX, 0x02, 0x03 },
 16948	        { GA_SEQUENCER_INDEX, 0x04, 0x03 },
 16949	        { GA_SEQUENCER_INDEX, 0x00, 0x03 },
 16950	        { GA_GRAPHICS_INDEX, 0x04, 0x00 },
 16951	        { GA_GRAPHICS_INDEX, 0x05, 0x10 },
 16952	        { GA_GRAPHICS_INDEX, 0x06,    0 },
 16953	  };
 16954	
 16955	  seq2[6].value= color ? 0x0E : 0x0A;
 16956	
 16957	  if (!machine.vdu_ega) return(ENOTTY);
 16958	  result = ga_program(seq1);    /* bring font memory into view */
 16959	
 16960	  result = sys_physcopy(m->PROC_NR, D, (vir_bytes) m->ADDRESS, 
 16961	        NONE, PHYS_SEG, (phys_bytes) GA_VIDEO_ADDRESS, (phys_bytes)GA_FONT_SIZE);
 16962	
 16963	  result = ga_program(seq2);    /* restore */
 16964	
 16965	  return(result);
 16966	}
	
 16968	/*===========================================================================*
 16969	 *                              ga_program                                   *
 16970	 *===========================================================================*/
 16971	PRIVATE int ga_program(seq)
 16972	struct sequence *seq;
 16973	{
 16974	  pvb_pair_t char_out[14];
 16975	  int i;
 16976	  for (i=0; i<7; i++) {
 16977	      pv_set(char_out[2*i], seq->index, seq->port);
 16978	      pv_set(char_out[2*i+1], seq->index+1, seq->value);
 16979	      seq++;
 16980	  } 
 16981	  return sys_voutb(char_out, 14);
 16982	}
	
 16984	/*===========================================================================*
 16985	 *                              cons_ioctl                                   *
 16986	 *===========================================================================*/
 16987	PRIVATE int cons_ioctl(tp, try)
 16988	tty_t *tp;
 16989	int try;
 16990	{
 16991	/* Set the screen dimensions. */
 16992	
 16993	  tp->tty_winsize.ws_row= scr_lines;
 16994	  tp->tty_winsize.ws_col= scr_width;
 16995	  tp->tty_winsize.ws_xpixel= scr_width * 8;
 16996	  tp->tty_winsize.ws_ypixel= scr_lines * font_lines;
 16997	}

_________________________ Page 867    File: drivers/tty/console.c ________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/pm.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17000	/* This is the master header for PM.  It includes some other files
 17001	 * and defines the principal constants.
 17002	 */
 17003	#define _POSIX_SOURCE      1    /* tell headers to include POSIX stuff */
 17004	#define _MINIX             1    /* tell headers to include MINIX stuff */
 17005	#define _SYSTEM            1    /* tell headers that this is the kernel */
 17006	
 17007	/* The following are so basic, all the *.c files get them automatically. */
 17008	#include <minix/config.h>       /* MUST be first */
 17009	#include <ansi.h>               /* MUST be second */
 17010	#include <sys/types.h>
 17011	#include <minix/const.h>
 17012	#include <minix/type.h>
 17013	
 17014	#include <fcntl.h>
 17015	#include <unistd.h>
 17016	#include <minix/syslib.h>
 17017	#include <minix/sysutil.h>
 17018	
 17019	#include <limits.h>
 17020	#include <errno.h>
 17021	
 17022	#include "const.h"
 17023	#include "type.h"
 17024	#include "proto.h"
 17025	#include "glo.h"





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/const.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17100	/* Constants used by the Process Manager. */
 17101	
 17102	#define NO_MEM ((phys_clicks) 0)  /* returned by alloc_mem() with mem is up */
 17103	
 17104	#define NR_PIDS        30000    /* process ids range from 0 to NR_PIDS-1.
 17105	                                 * (magic constant: some old applications use
 17106	                                 * a 'short' instead of pid_t.)
 17107	                                 */
 17108	
 17109	#define PM_PID             0    /* PM's process id number */
 17110	#define INIT_PID           1    /* INIT's process id number */
 17111	

_________________________ Page 868    File: servers/pm/const.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/type.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17200	/* If there were any type definitions local to the Process Manager, they would
 17201	 * be here.  This file is included only for symmetry with the kernel and File
 17202	 * System, which do have some local type definitions.
 17203	 */
 17204	

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/proto.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17300	/* Function prototypes. */
 17301	
 17302	struct mproc;
 17303	struct stat;
 17304	struct mem_map;
 17305	struct memory;
 17306	
 17307	#include <timers.h>
 17308	
 17309	/* alloc.c */
 17310	_PROTOTYPE( phys_clicks alloc_mem, (phys_clicks clicks)                 );
 17311	_PROTOTYPE( void free_mem, (phys_clicks base, phys_clicks clicks)       );
 17312	_PROTOTYPE( void mem_init, (struct memory *chunks, phys_clicks *free)   );
 17313	#define swap_in()                       ((void)0)
 17314	#define swap_inqueue(rmp)               ((void)0)
 17315	
 17316	/* break.c */
 17317	_PROTOTYPE( int adjust, (struct mproc *rmp,
 17318	                        vir_clicks data_clicks, vir_bytes sp)           );
 17319	_PROTOTYPE( int do_brk, (void)                                          );
 17320	_PROTOTYPE( int size_ok, (int file_type, vir_clicks tc, vir_clicks dc,
 17321	                        vir_clicks sc, vir_clicks dvir, vir_clicks s_vir) );
 17322	
 17323	/* devio.c */
 17324	_PROTOTYPE( int do_dev_io, (void) );
 17325	_PROTOTYPE( int do_dev_io, (void) );
 17326	
 17327	/* dmp.c */
 17328	_PROTOTYPE( int do_fkey_pressed, (void)                                         );
 17329	
 17330	/* exec.c */
 17331	_PROTOTYPE( int do_exec, (void)                                         );
 17332	_PROTOTYPE( void rw_seg, (int rw, int fd, int proc, int seg,
 17333	                                                phys_bytes seg_bytes)   );
 17334	_PROTOTYPE( struct mproc *find_share, (struct mproc *mp_ign, Ino_t ino,
 17335	                        Dev_t dev, time_t ctime)                        );
 17336	
 17337	/* forkexit.c */
 17338	_PROTOTYPE( int do_fork, (void)                                         );
 17339	_PROTOTYPE( int do_pm_exit, (void)                                      );
 17340	_PROTOTYPE( int do_waitpid, (void)                                      );
 17341	_PROTOTYPE( void pm_exit, (struct mproc *rmp, int exit_status)          );
 17342	
 17343	/* getset.c */
 17344	_PROTOTYPE( int do_getset, (void)                                       );

_________________________ Page 869    File: servers/pm/proto.h _________________________
 17345	
 17346	/* main.c */
 17347	_PROTOTYPE( int main, (void)                                            );
 17348	
 17349	/* misc.c */
 17350	_PROTOTYPE( int do_reboot, (void)                                       );
 17351	_PROTOTYPE( int do_getsysinfo, (void)                                   );
 17352	_PROTOTYPE( int do_getprocnr, (void)                                    );
 17353	_PROTOTYPE( int do_svrctl, (void)                                       );
 17354	_PROTOTYPE( int do_allocmem, (void)                                     );
 17355	_PROTOTYPE( int do_freemem, (void)                                      );
 17356	_PROTOTYPE( int do_getsetpriority, (void)                                       );
 17357	
 17358	_PROTOTYPE( void setreply, (int proc_nr, int result)                    );
 17359	
 17360	/* signal.c */
 17361	_PROTOTYPE( int do_alarm, (void)                                        );
 17362	_PROTOTYPE( int do_kill, (void)                                         );
 17363	_PROTOTYPE( int ksig_pending, (void)                                    );
 17364	_PROTOTYPE( int do_pause, (void)                                        );
 17365	_PROTOTYPE( int set_alarm, (int proc_nr, int sec)                       );
 17366	_PROTOTYPE( int check_sig, (pid_t proc_id, int signo)                   );
 17367	_PROTOTYPE( void sig_proc, (struct mproc *rmp, int sig_nr)              );
 17368	_PROTOTYPE( int do_sigaction, (void)                                    );
 17369	_PROTOTYPE( int do_sigpending, (void)                                   );
 17370	_PROTOTYPE( int do_sigprocmask, (void)                                  );
 17371	_PROTOTYPE( int do_sigreturn, (void)                                    );
 17372	_PROTOTYPE( int do_sigsuspend, (void)                                   );
 17373	_PROTOTYPE( void check_pending, (struct mproc *rmp)                     );
 17374	
 17375	/* time.c */
 17376	_PROTOTYPE( int do_stime, (void)                                        );
 17377	_PROTOTYPE( int do_time, (void)                                         );
 17378	_PROTOTYPE( int do_times, (void)                                        );
 17379	_PROTOTYPE( int do_gettimeofday, (void)                                 );
 17380	
 17381	/* timers.c */
 17382	_PROTOTYPE( void pm_set_timer, (timer_t *tp, int delta, 
 17383	        tmr_func_t watchdog, int arg));
 17384	_PROTOTYPE( void pm_expire_timers, (clock_t now));
 17385	_PROTOTYPE( void pm_cancel_timer, (timer_t *tp));
 17386	
 17387	/* trace.c */
 17388	_PROTOTYPE( int do_trace, (void)                                        );
 17389	_PROTOTYPE( void stop_proc, (struct mproc *rmp, int sig_nr)             );
 17390	
 17391	/* utility.c */
 17392	_PROTOTYPE( pid_t get_free_pid, (void)                                  );
 17393	_PROTOTYPE( int allowed, (char *name_buf, struct stat *s_buf, int mask) );
 17394	_PROTOTYPE( int no_sys, (void)                                          );
 17395	_PROTOTYPE( void panic, (char *who, char *mess, int num)                );
 17396	_PROTOTYPE( void tell_fs, (int what, int p1, int p2, int p3)            );
 17397	_PROTOTYPE( int get_stack_ptr, (int proc_nr, vir_bytes *sp)             );
 17398	_PROTOTYPE( int get_mem_map, (int proc_nr, struct mem_map *mem_map)     );
 17399	_PROTOTYPE( char *find_param, (const char *key));
 17400	_PROTOTYPE( int proc_from_pid, (pid_t p));

_________________________ Page 870    File: servers/pm/proto.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/glo.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17500	/* EXTERN should be extern except in table.c */
 17501	#ifdef _TABLE
 17502	#undef EXTERN
 17503	#define EXTERN
 17504	#endif
 17505	
 17506	/* Global variables. */
 17507	EXTERN struct mproc *mp;        /* ptr to 'mproc' slot of current process */
 17508	EXTERN int procs_in_use;        /* how many processes are marked as IN_USE */
 17509	EXTERN char monitor_params[128*sizeof(char *)]; /* boot monitor parameters */
 17510	EXTERN struct kinfo kinfo;                      /* kernel information */
 17511	
 17512	/* The parameters of the call are kept here. */
 17513	EXTERN message m_in;            /* the incoming message itself is kept here. */
 17514	EXTERN int who;                 /* caller's proc number */
 17515	EXTERN int call_nr;             /* system call number */
 17516	
 17517	extern _PROTOTYPE (int (*call_vec[]), (void) ); /* system call handlers */
 17518	extern char core_name[];        /* file name where core images are produced */
 17519	EXTERN sigset_t core_sset;      /* which signals cause core images */
 17520	EXTERN sigset_t ign_sset;       /* which signals are by default ignored */
 17521	




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/mproc.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17600	/* This table has one slot per process.  It contains all the process management
 17601	 * information for each process.  Among other things, it defines the text, data
 17602	 * and stack segments, uids and gids, and various flags.  The kernel and file
 17603	 * systems have tables that are also indexed by process, with the contents
 17604	 * of corresponding slots referring to the same process in all three.
 17605	 */
 17606	#include <timers.h>
 17607	
 17608	EXTERN struct mproc {
 17609	  struct mem_map mp_seg[NR_LOCAL_SEGS]; /* points to text, data, stack */
 17610	  char mp_exitstatus;           /* storage for status when process exits */
 17611	  char mp_sigstatus;            /* storage for signal # for killed procs */
 17612	  pid_t mp_pid;                 /* process id */
 17613	  pid_t mp_procgrp;             /* pid of process group (used for signals) */
 17614	  pid_t mp_wpid;                /* pid this process is waiting for */
 17615	  int mp_parent;                /* index of parent process */
 17616	
 17617	  /* Child user and system times. Accounting done on child exit. */
 17618	  clock_t mp_child_utime;       /* cumulative user time of children */
 17619	  clock_t mp_child_stime;       /* cumulative sys time of children */
 17620	
 17621	  /* Real and effective uids and gids. */
 17622	  uid_t mp_realuid;             /* process' real uid */
 17623	  uid_t mp_effuid;              /* process' effective uid */
 17624	  gid_t mp_realgid;             /* process' real gid */

_________________________ Page 871    File: servers/pm/mproc.h _________________________
 17625	  gid_t mp_effgid;              /* process' effective gid */
 17626	
 17627	  /* File identification for sharing. */
 17628	  ino_t mp_ino;                 /* inode number of file */
 17629	  dev_t mp_dev;                 /* device number of file system */
 17630	  time_t mp_ctime;              /* inode changed time */
 17631	
 17632	  /* Signal handling information. */
 17633	  sigset_t mp_ignore;           /* 1 means ignore the signal, 0 means don't */
 17634	  sigset_t mp_catch;            /* 1 means catch the signal, 0 means don't */
 17635	  sigset_t mp_sig2mess;         /* 1 means transform into notify message */
 17636	  sigset_t mp_sigmask;          /* signals to be blocked */
 17637	  sigset_t mp_sigmask2;         /* saved copy of mp_sigmask */
 17638	  sigset_t mp_sigpending;       /* pending signals to be handled */
 17639	  struct sigaction mp_sigact[_NSIG + 1]; /* as in sigaction(2) */
 17640	  vir_bytes mp_sigreturn;       /* address of C library __sigreturn function */
 17641	  struct timer mp_timer;        /* watchdog timer for alarm(2) */
 17642	
 17643	  /* Backwards compatibility for signals. */
 17644	  sighandler_t mp_func;         /* all sigs vectored to a single user fcn */
 17645	
 17646	  unsigned mp_flags;            /* flag bits */
 17647	  vir_bytes mp_procargs;        /* ptr to proc's initial stack arguments */
 17648	  struct mproc *mp_swapq;       /* queue of procs waiting to be swapped in */
 17649	  message mp_reply;             /* reply message to be sent to one */
 17650	
 17651	  /* Scheduling priority. */
 17652	  signed int mp_nice;           /* nice is PRIO_MIN..PRIO_MAX, standard 0. */
 17653	
 17654	  char mp_name[PROC_NAME_LEN];  /* process name */
 17655	} mproc[NR_PROCS];
 17656	
 17657	/* Flag values */
 17658	#define IN_USE          0x001   /* set when 'mproc' slot in use */
 17659	#define WAITING         0x002   /* set by WAIT system call */
 17660	#define ZOMBIE          0x004   /* set by EXIT, cleared by WAIT */
 17661	#define PAUSED          0x008   /* set by PAUSE system call */
 17662	#define ALARM_ON        0x010   /* set when SIGALRM timer started */
 17663	#define SEPARATE        0x020   /* set if file is separate I & D space */
 17664	#define TRACED          0x040   /* set if process is to be traced */
 17665	#define STOPPED         0x080   /* set if process stopped for tracing */
 17666	#define SIGSUSPENDED    0x100   /* set by SIGSUSPEND system call */
 17667	#define REPLY           0x200   /* set if a reply message is pending */
 17668	#define ONSWAP          0x400   /* set if data segment is swapped out */
 17669	#define SWAPIN          0x800   /* set if on the "swap this in" queue */
 17670	#define DONT_SWAP      0x1000   /* never swap out this process */
 17671	#define PRIV_PROC      0x2000   /* system process, special privileges */
 17672	
 17673	#define NIL_MPROC ((struct mproc *) 0)
 17674	

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/param.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17700	/* The following names are synonyms for the variables in the input message. */
 17701	#define addr            m1_p1
 17702	#define exec_name       m1_p1
 17703	#define exec_len        m1_i1
 17704	#define func            m6_f1

_________________________ Page 872    File: servers/pm/param.h _________________________
 17705	#define grp_id          m1_i1
 17706	#define namelen         m1_i2
 17707	#define pid             m1_i1
 17708	#define procnr          m1_i1
 17709	#define seconds         m1_i1
 17710	#define sig             m6_i1
 17711	#define stack_bytes     m1_i2
 17712	#define stack_ptr       m1_p2
 17713	#define status          m1_i1
 17714	#define usr_id          m1_i1
 17715	#define request         m2_i2
 17716	#define taddr           m2_l1
 17717	#define data            m2_l2
 17718	#define sig_nr          m1_i2
 17719	#define sig_nsa         m1_p1
 17720	#define sig_osa         m1_p2
 17721	#define sig_ret         m1_p3
 17722	#define sig_set         m2_l1
 17723	#define sig_how         m2_i1
 17724	#define sig_flags       m2_i2
 17725	#define sig_context     m2_p1
 17726	#define info_what       m1_i1
 17727	#define info_where      m1_p1
 17728	#define reboot_flag     m1_i1
 17729	#define reboot_code     m1_p1
 17730	#define reboot_strlen   m1_i2
 17731	#define svrctl_req      m2_i1
 17732	#define svrctl_argp     m2_p1
 17733	#define stime           m2_l1
 17734	#define memsize         m4_l1
 17735	#define membase         m4_l2
 17736	
 17737	/* The following names are synonyms for the variables in a reply message. */
 17738	#define reply_res       m_type
 17739	#define reply_res2      m2_i1
 17740	#define reply_ptr       m2_p1
 17741	#define reply_mask      m2_l1   
 17742	#define reply_trace     m2_l2   
 17743	#define reply_time      m2_l1
 17744	#define reply_utime     m2_l2
 17745	#define reply_t1        m4_l1
 17746	#define reply_t2        m4_l2
 17747	#define reply_t3        m4_l3
 17748	#define reply_t4        m4_l4
 17749	#define reply_t5        m4_l5
 17750	
 17751	/* The following names are used to inform the FS about certain events. */
 17752	#define tell_fs_arg1    m1_i1
 17753	#define tell_fs_arg2    m1_i2
 17754	#define tell_fs_arg3    m1_i3
 17755	

_________________________ Page 873    File: servers/pm/param.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/table.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 17800	/* This file contains the table used to map system call numbers onto the
 17801	 * routines that perform them.
 17802	 */
 17803	
 17804	#define _TABLE
 17805	
 17806	#include "pm.h"
 17807	#include <minix/callnr.h>
 17808	#include <signal.h>
 17809	#include "mproc.h"
 17810	#include "param.h"
 17811	
 17812	/* Miscellaneous */
 17813	char core_name[] = "core";      /* file name where core images are produced */
 17814	
 17815	_PROTOTYPE (int (*call_vec[NCALLS]), (void) ) = {
 17816	        no_sys,         /*  0 = unused  */
 17817	        do_pm_exit,     /*  1 = exit    */
 17818	        do_fork,        /*  2 = fork    */
 17819	        no_sys,         /*  3 = read    */
 17820	        no_sys,         /*  4 = write   */
 17821	        no_sys,         /*  5 = open    */
 17822	        no_sys,         /*  6 = close   */
 17823	        do_waitpid,     /*  7 = wait    */
 17824	        no_sys,         /*  8 = creat   */
 17825	        no_sys,         /*  9 = link    */
 17826	        no_sys,         /* 10 = unlink  */
 17827	        do_waitpid,     /* 11 = waitpid */
 17828	        no_sys,         /* 12 = chdir   */
 17829	        do_time,        /* 13 = time    */
 17830	        no_sys,         /* 14 = mknod   */
 17831	        no_sys,         /* 15 = chmod   */
 17832	        no_sys,         /* 16 = chown   */
 17833	        do_brk,         /* 17 = break   */
 17834	        no_sys,         /* 18 = stat    */
 17835	        no_sys,         /* 19 = lseek   */
 17836	        do_getset,      /* 20 = getpid  */
 17837	        no_sys,         /* 21 = mount   */
 17838	        no_sys,         /* 22 = umount  */
 17839	        do_getset,      /* 23 = setuid  */
 17840	        do_getset,      /* 24 = getuid  */
 17841	        do_stime,       /* 25 = stime   */
 17842	        do_trace,       /* 26 = ptrace  */
 17843	        do_alarm,       /* 27 = alarm   */
 17844	        no_sys,         /* 28 = fstat   */
 17845	        do_pause,       /* 29 = pause   */
 17846	        no_sys,         /* 30 = utime   */
 17847	        no_sys,         /* 31 = (stty)  */
 17848	        no_sys,         /* 32 = (gtty)  */
 17849	        no_sys,         /* 33 = access  */
 17850	        no_sys,         /* 34 = (nice)  */
 17851	        no_sys,         /* 35 = (ftime) */
 17852	        no_sys,         /* 36 = sync    */
 17853	        do_kill,        /* 37 = kill    */
 17854	        no_sys,         /* 38 = rename  */

_________________________ Page 874    File: servers/pm/table.c _________________________
 17855	        no_sys,         /* 39 = mkdir   */
 17856	        no_sys,         /* 40 = rmdir   */
 17857	        no_sys,         /* 41 = dup     */
 17858	        no_sys,         /* 42 = pipe    */
 17859	        do_times,       /* 43 = times   */
 17860	        no_sys,         /* 44 = (prof)  */
 17861	        no_sys,         /* 45 = unused  */
 17862	        do_getset,      /* 46 = setgid  */
 17863	        do_getset,      /* 47 = getgid  */
 17864	        no_sys,         /* 48 = (signal)*/
 17865	        no_sys,         /* 49 = unused  */
 17866	        no_sys,         /* 50 = unused  */
 17867	        no_sys,         /* 51 = (acct)  */
 17868	        no_sys,         /* 52 = (phys)  */
 17869	        no_sys,         /* 53 = (lock)  */
 17870	        no_sys,         /* 54 = ioctl   */
 17871	        no_sys,         /* 55 = fcntl   */
 17872	        no_sys,         /* 56 = (mpx)   */
 17873	        no_sys,         /* 57 = unused  */
 17874	        no_sys,         /* 58 = unused  */
 17875	        do_exec,        /* 59 = execve  */
 17876	        no_sys,         /* 60 = umask   */
 17877	        no_sys,         /* 61 = chroot  */
 17878	        do_getset,      /* 62 = setsid  */
 17879	        do_getset,      /* 63 = getpgrp */
 17880	
 17881	        no_sys,         /* 64 = unused */
 17882	        no_sys,         /* 65 = UNPAUSE */
 17883	        no_sys,         /* 66 = unused  */
 17884	        no_sys,         /* 67 = REVIVE  */
 17885	        no_sys,         /* 68 = TASK_REPLY  */
 17886	        no_sys,         /* 69 = unused  */
 17887	        no_sys,         /* 70 = unused  */
 17888	        do_sigaction,   /* 71 = sigaction   */
 17889	        do_sigsuspend,  /* 72 = sigsuspend  */
 17890	        do_sigpending,  /* 73 = sigpending  */
 17891	        do_sigprocmask, /* 74 = sigprocmask */
 17892	        do_sigreturn,   /* 75 = sigreturn   */
 17893	        do_reboot,      /* 76 = reboot  */
 17894	        do_svrctl,      /* 77 = svrctl  */
 17895	
 17896	        no_sys,         /* 78 = unused */
 17897	        do_getsysinfo,  /* 79 = getsysinfo */
 17898	        do_getprocnr,   /* 80 = getprocnr */
 17899	        no_sys,         /* 81 = unused */
 17900	        no_sys,         /* 82 = fstatfs */
 17901	        do_allocmem,    /* 83 = memalloc */
 17902	        do_freemem,     /* 84 = memfree */
 17903	        no_sys,         /* 85 = select */
 17904	        no_sys,         /* 86 = fchdir */
 17905	        no_sys,         /* 87 = fsync */
 17906	        do_getsetpriority,      /* 88 = getpriority */
 17907	        do_getsetpriority,      /* 89 = setpriority */
 17908	        do_time,        /* 90 = gettimeofday */
 17909	};
 17910	/* This should not fail with "array size is negative": */
 17911	extern int dummy[sizeof(call_vec) == NCALLS * sizeof(call_vec[0]) ? 1 : -1];

_________________________ Page 875    File: servers/pm/table.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/main.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 18000	/* This file contains the main program of the process manager and some related
 18001	 * procedures.  When MINIX starts up, the kernel runs for a little while,
 18002	 * initializing itself and its tasks, and then it runs PM and FS.  Both PM
 18003	 * and FS initialize themselves as far as they can. PM asks the kernel for
 18004	 * all free memory and starts serving requests.
 18005	 *
 18006	 * The entry points into this file are:
 18007	 *   main:      starts PM running
 18008	 *   setreply:  set the reply to be sent to process making an PM system call
 18009	 */
 18010	
 18011	#include "pm.h"
 18012	#include <minix/keymap.h>
 18013	#include <minix/callnr.h>
 18014	#include <minix/com.h>
 18015	#include <signal.h>
 18016	#include <stdlib.h>
 18017	#include <fcntl.h>
 18018	#include <sys/resource.h>
 18019	#include <string.h>
 18020	#include "mproc.h"
 18021	#include "param.h"
 18022	
 18023	#include "../../kernel/const.h"
 18024	#include "../../kernel/config.h"
 18025	#include "../../kernel/type.h"
 18026	#include "../../kernel/proc.h"
 18027	
 18028	FORWARD _PROTOTYPE( void get_work, (void)                               );
 18029	FORWARD _PROTOTYPE( void pm_init, (void)                                );
 18030	FORWARD _PROTOTYPE( int get_nice_value, (int queue)                     );
 18031	FORWARD _PROTOTYPE( void get_mem_chunks, (struct memory *mem_chunks)    );
 18032	FORWARD _PROTOTYPE( void patch_mem_chunks, (struct memory *mem_chunks, 
 18033	        struct mem_map *map_ptr)        );
 18034	
 18035	#define click_to_round_k(n) \
 18036	        ((unsigned) ((((unsigned long) (n) << CLICK_SHIFT) + 512) / 1024))
 18037	
 18038	/*===========================================================================*
 18039	 *                              main                                         *
 18040	 *===========================================================================*/
 18041	PUBLIC int main()
 18042	{
 18043	/* Main routine of the process manager. */
 18044	  int result, s, proc_nr;
 18045	  struct mproc *rmp;
 18046	  sigset_t sigset;
 18047	
 18048	  pm_init();                    /* initialize process manager tables */
 18049	
 18050	  /* This is PM's main loop-  get work and do it, forever and forever. */
 18051	  while (TRUE) {
 18052	        get_work();             /* wait for an PM system call */
 18053	
 18054	        /* Check for system notifications first. Special cases. */

_________________________ Page 876    File: servers/pm/main.c _________________________
 18055	        if (call_nr == SYN_ALARM) {
 18056	                pm_expire_timers(m_in.NOTIFY_TIMESTAMP);
 18057	                result = SUSPEND;               /* don't reply */
 18058	        } else if (call_nr == SYS_SIG) {        /* signals pending */
 18059	                sigset = m_in.NOTIFY_ARG;
 18060	                if (sigismember(&sigset, SIGKSIG))  (void) ksig_pending();
 18061	                result = SUSPEND;               /* don't reply */
 18062	        }
 18063	        /* Else, if the system call number is valid, perform the call. */
 18064	        else if ((unsigned) call_nr >= NCALLS) {
 18065	                result = ENOSYS;
 18066	        } else {
 18067	                result = (*call_vec[call_nr])();
 18068	        }
 18069	
 18070	        /* Send the results back to the user to indicate completion. */
 18071	        if (result != SUSPEND) setreply(who, result);
 18072	
 18073	        swap_in();              /* maybe a process can be swapped in? */
 18074	
 18075	        /* Send out all pending reply messages, including the answer to
 18076	         * the call just made above.  The processes must not be swapped out.
 18077	         */
 18078	        for (proc_nr=0, rmp=mproc; proc_nr < NR_PROCS; proc_nr++, rmp++) {
 18079	                /* In the meantime, the process may have been killed by a
 18080	                 * signal (e.g. if a lethal pending signal was unblocked)
 18081	                 * without the PM realizing it. If the slot is no longer in
 18082	                 * use or just a zombie, don't try to reply.
 18083	                 */
 18084	                if ((rmp->mp_flags & (REPLY | ONSWAP | IN_USE | ZOMBIE)) ==
 18085	                   (REPLY | IN_USE)) {
 18086	                        if ((s=send(proc_nr, &rmp->mp_reply)) != OK) {
 18087	                                panic(__FILE__,"PM can't reply to", proc_nr);
 18088	                        }
 18089	                        rmp->mp_flags &= ~REPLY;
 18090	                }
 18091	        }
 18092	  }
 18093	  return(OK);
 18094	}
	
 18096	/*===========================================================================*
 18097	 *                              get_work                                     *
 18098	 *===========================================================================*/
 18099	PRIVATE void get_work()
 18100	{
 18101	/* Wait for the next message and extract useful information from it. */
 18102	  if (receive(ANY, &m_in) != OK) panic(__FILE__,"PM receive error", NO_NUM);
 18103	  who = m_in.m_source;          /* who sent the message */
 18104	  call_nr = m_in.m_type;        /* system call number */
 18105	
 18106	  /* Process slot of caller. Misuse PM's own process slot if the kernel is
 18107	   * calling. This can happen in case of synchronous alarms (CLOCK) or or 
 18108	   * event like pending kernel signals (SYSTEM).
 18109	   */
 18110	  mp = &mproc[who < 0 ? PM_PROC_NR : who];
 18111	}
	

_________________________ Page 877    File: servers/pm/main.c _________________________
 18113	/*===========================================================================*
 18114	 *                              setreply                                     *
 18115	 *===========================================================================*/
 18116	PUBLIC void setreply(proc_nr, result)
 18117	int proc_nr;                    /* process to reply to */
 18118	int result;                     /* result of call (usually OK or error #) */
 18119	{
 18120	/* Fill in a reply message to be sent later to a user process.  System calls
 18121	 * may occasionally fill in other fields, this is only for the main return
 18122	 * value, and for setting the "must send reply" flag.
 18123	 */
 18124	  register struct mproc *rmp = &mproc[proc_nr];
 18125	
 18126	  rmp->mp_reply.reply_res = result;
 18127	  rmp->mp_flags |= REPLY;       /* reply pending */
 18128	
 18129	  if (rmp->mp_flags & ONSWAP)
 18130	        swap_inqueue(rmp);      /* must swap this process back in */
 18131	}
	
 18133	/*===========================================================================*
 18134	 *                              pm_init                                      *
 18135	 *===========================================================================*/
 18136	PRIVATE void pm_init()
 18137	{
 18138	/* Initialize the process manager. 
 18139	 * Memory use info is collected from the boot monitor, the kernel, and
 18140	 * all processes compiled into the system image. Initially this information
 18141	 * is put into an array mem_chunks. Elements of mem_chunks are struct memory,
 18142	 * and hold base, size pairs in units of clicks. This array is small, there
 18143	 * should be no more than 8 chunks. After the array of chunks has been built
 18144	 * the contents are used to initialize the hole list. Space for the hole list
 18145	 * is reserved as an array with twice as many elements as the maximum number
 18146	 * of processes allowed. It is managed as a linked list, and elements of the
 18147	 * array are struct hole, which, in addition to storage for a base and size in 
 18148	 * click units also contain space for a link, a pointer to another element.
 18149	*/
 18150	  int s;
 18151	  static struct boot_image image[NR_BOOT_PROCS];
 18152	  register struct boot_image *ip;
 18153	  static char core_sigs[] = { SIGQUIT, SIGILL, SIGTRAP, SIGABRT,
 18154	                        SIGEMT, SIGFPE, SIGUSR1, SIGSEGV, SIGUSR2 };
 18155	  static char ign_sigs[] = { SIGCHLD };
 18156	  register struct mproc *rmp;
 18157	  register char *sig_ptr;
 18158	  phys_clicks total_clicks, minix_clicks, free_clicks;
 18159	  message mess;
 18160	  struct mem_map mem_map[NR_LOCAL_SEGS];
 18161	  struct memory mem_chunks[NR_MEMS];
 18162	
 18163	  /* Initialize process table, including timers. */
 18164	  for (rmp=&mproc[0]; rmp<&mproc[NR_PROCS]; rmp++) {
 18165	        tmr_inittimer(&rmp->mp_timer);
 18166	  }
 18167	
 18168	  /* Build the set of signals which cause core dumps, and the set of signals
 18169	   * that are by default ignored.
 18170	   */
 18171	  sigemptyset(&core_sset);
 18172	  for (sig_ptr = core_sigs; sig_ptr < core_sigs+sizeof(core_sigs); sig_ptr++)

_________________________ Page 878    File: servers/pm/main.c _________________________
 18173	        sigaddset(&core_sset, *sig_ptr);
 18174	  sigemptyset(&ign_sset);
 18175	  for (sig_ptr = ign_sigs; sig_ptr < ign_sigs+sizeof(ign_sigs); sig_ptr++)
 18176	        sigaddset(&ign_sset, *sig_ptr);
 18177	
 18178	  /* Obtain a copy of the boot monitor parameters and the kernel info struct.  
 18179	   * Parse the list of free memory chunks. This list is what the boot monitor 
 18180	   * reported, but it must be corrected for the kernel and system processes.
 18181	   */
 18182	  if ((s=sys_getmonparams(monitor_params, sizeof(monitor_params))) != OK)
 18183	      panic(__FILE__,"get monitor params failed",s);
 18184	  get_mem_chunks(mem_chunks);
 18185	  if ((s=sys_getkinfo(&kinfo)) != OK)
 18186	      panic(__FILE__,"get kernel info failed",s);
 18187	
 18188	  /* Get the memory map of the kernel to see how much memory it uses. */
 18189	  if ((s=get_mem_map(SYSTASK, mem_map)) != OK)
 18190	        panic(__FILE__,"couldn't get memory map of SYSTASK",s);
 18191	  minix_clicks = (mem_map[S].mem_phys+mem_map[S].mem_len)-mem_map[T].mem_phys;
 18192	  patch_mem_chunks(mem_chunks, mem_map);
 18193	
 18194	  /* Initialize PM's process table. Request a copy of the system image table 
 18195	   * that is defined at the kernel level to see which slots to fill in.
 18196	   */
 18197	  if (OK != (s=sys_getimage(image))) 
 18198	        panic(__FILE__,"couldn't get image table: %d\n", s);
 18199	  procs_in_use = 0;                             /* start populating table */
 18200	  printf("Building process table:");            /* show what's happening */
 18201	  for (ip = &image[0]; ip < &image[NR_BOOT_PROCS]; ip++) {              
 18202	        if (ip->proc_nr >= 0) {                 /* task have negative nrs */
 18203	                procs_in_use += 1;              /* found user process */
 18204	
 18205	                /* Set process details found in the image table. */
 18206	                rmp = &mproc[ip->proc_nr];      
 18207	                strncpy(rmp->mp_name, ip->proc_name, PROC_NAME_LEN); 
 18208	                rmp->mp_parent = RS_PROC_NR;
 18209	                rmp->mp_nice = get_nice_value(ip->priority);
 18210	                if (ip->proc_nr == INIT_PROC_NR) {      /* user process */
 18211	                        rmp->mp_pid = INIT_PID;
 18212	                        rmp->mp_flags |= IN_USE; 
 18213	                        sigemptyset(&rmp->mp_ignore);   
 18214	                }
 18215	                else {                                  /* system process */
 18216	                        rmp->mp_pid = get_free_pid();
 18217	                        rmp->mp_flags |= IN_USE | DONT_SWAP | PRIV_PROC; 
 18218	                        sigfillset(&rmp->mp_ignore);    
 18219	                }
 18220	                sigemptyset(&rmp->mp_sigmask);
 18221	                sigemptyset(&rmp->mp_catch);
 18222	                sigemptyset(&rmp->mp_sig2mess);
 18223	
 18224	                /* Get memory map for this process from the kernel. */
 18225	                if ((s=get_mem_map(ip->proc_nr, rmp->mp_seg)) != OK)
 18226	                        panic(__FILE__,"couldn't get process entry",s);
 18227	                if (rmp->mp_seg[T].mem_len != 0) rmp->mp_flags |= SEPARATE;
 18228	                minix_clicks += rmp->mp_seg[S].mem_phys + 
 18229	                        rmp->mp_seg[S].mem_len - rmp->mp_seg[T].mem_phys;
 18230	                patch_mem_chunks(mem_chunks, rmp->mp_seg);
 18231	
 18232	                /* Tell FS about this system process. */

_________________________ Page 879    File: servers/pm/main.c _________________________
 18233	                mess.PR_PROC_NR = ip->proc_nr;
 18234	                mess.PR_PID = rmp->mp_pid;
 18235	                if (OK != (s=send(FS_PROC_NR, &mess)))
 18236	                        panic(__FILE__,"can't sync up with FS", s);
 18237	                printf(" %s", ip->proc_name);   /* display process name */
 18238	        }
 18239	  }
 18240	  printf(".\n");                                /* last process done */
 18241	
 18242	  /* Override some details. PM is somewhat special. */
 18243	  mproc[PM_PROC_NR].mp_pid = PM_PID;            /* magically override pid */
 18244	  mproc[PM_PROC_NR].mp_parent = PM_PROC_NR;     /* PM doesn't have parent */
 18245	
 18246	  /* Tell FS that no more system processes follow and synchronize. */
 18247	  mess.PR_PROC_NR = NONE;
 18248	  if (sendrec(FS_PROC_NR, &mess) != OK || mess.m_type != OK)
 18249	        panic(__FILE__,"can't sync up with FS", NO_NUM);
 18250	
 18251	  /* Initialize tables to all physical memory and print memory information. */
 18252	  printf("Physical memory:");
 18253	  mem_init(mem_chunks, &free_clicks);
 18254	  total_clicks = minix_clicks + free_clicks;
 18255	  printf(" total %u KB,", click_to_round_k(total_clicks));
 18256	  printf(" system %u KB,", click_to_round_k(minix_clicks));
 18257	  printf(" free %u KB.\n", click_to_round_k(free_clicks));
 18258	}
	
 18260	/*===========================================================================*
 18261	 *                              get_nice_value                               *
 18262	 *===========================================================================*/
 18263	PRIVATE int get_nice_value(queue)
 18264	int queue;                              /* store mem chunks here */
 18265	{
 18266	/* Processes in the boot image have a priority assigned. The PM doesn't know
 18267	 * about priorities, but uses 'nice' values instead. The priority is between 
 18268	 * MIN_USER_Q and MAX_USER_Q. We have to scale between PRIO_MIN and PRIO_MAX.
 18269	 */ 
 18270	  int nice_val = (queue - USER_Q) * (PRIO_MAX-PRIO_MIN+1) / 
 18271	      (MIN_USER_Q-MAX_USER_Q+1);
 18272	  if (nice_val > PRIO_MAX) nice_val = PRIO_MAX; /* shouldn't happen */
 18273	  if (nice_val < PRIO_MIN) nice_val = PRIO_MIN; /* shouldn't happen */
 18274	  return nice_val;
 18275	}
	
 18277	/*===========================================================================*
 18278	 *                              get_mem_chunks                               *
 18279	 *===========================================================================*/
 18280	PRIVATE void get_mem_chunks(mem_chunks)
 18281	struct memory *mem_chunks;                      /* store mem chunks here */
 18282	{
 18283	/* Initialize the free memory list from the 'memory' boot variable.  Translate
 18284	 * the byte offsets and sizes in this list to clicks, properly truncated. Also
 18285	 * make sure that we don't exceed the maximum address space of the 286 or the
 18286	 * 8086, i.e. when running in 16-bit protected mode or real mode.
 18287	 */
 18288	  long base, size, limit;
 18289	  char *s, *end;                        /* use to parse boot variable */ 
 18290	  int i, done = 0;
 18291	  struct memory *memp;
 18292	

_________________________ Page 880    File: servers/pm/main.c _________________________
 18293	  /* Initialize everything to zero. */
 18294	  for (i = 0; i < NR_MEMS; i++) {
 18295	        memp = &mem_chunks[i];          /* next mem chunk is stored here */
 18296	        memp->base = memp->size = 0;
 18297	  }
 18298	  
 18299	  /* The available memory is determined by MINIX' boot loader as a list of 
 18300	   * (base:size)-pairs in boothead.s. The 'memory' boot variable is set in
 18301	   * in boot.s.  The format is "b0:s0,b1:s1,b2:s2", where b0:s0 is low mem,
 18302	   * b1:s1 is mem between 1M and 16M, b2:s2 is mem above 16M. Pairs b1:s1 
 18303	   * and b2:s2 are combined if the memory is adjacent. 
 18304	   */
 18305	  s = find_param("memory");             /* get memory boot variable */
 18306	  for (i = 0; i < NR_MEMS && !done; i++) {
 18307	        memp = &mem_chunks[i];          /* next mem chunk is stored here */
 18308	        base = size = 0;                /* initialize next base:size pair */
 18309	        if (*s != 0) {                  /* get fresh data, unless at end */     
 18310	
 18311	            /* Read fresh base and expect colon as next char. */ 
 18312	            base = strtoul(s, &end, 0x10);              /* get number */
 18313	            if (end != s && *end == ':') s = ++end;     /* skip ':' */ 
 18314	            else *s=0;                  /* terminate, should not happen */
 18315	
 18316	            /* Read fresh size and expect comma or assume end. */ 
 18317	            size = strtoul(s, &end, 0x10);              /* get number */
 18318	            if (end != s && *end == ',') s = ++end;     /* skip ',' */
 18319	            else done = 1;
 18320	        }
 18321	        limit = base + size;    
 18322	        base = (base + CLICK_SIZE-1) & ~(long)(CLICK_SIZE-1);
 18323	        limit &= ~(long)(CLICK_SIZE-1);
 18324	        if (limit <= base) continue;
 18325	        memp->base = base >> CLICK_SHIFT;
 18326	        memp->size = (limit - base) >> CLICK_SHIFT;
 18327	  }
 18328	}
	
 18330	/*===========================================================================*
 18331	 *                              patch_mem_chunks                             *
 18332	 *===========================================================================*/
 18333	PRIVATE void patch_mem_chunks(mem_chunks, map_ptr)
 18334	struct memory *mem_chunks;                      /* store mem chunks here */
 18335	struct mem_map *map_ptr;                        /* memory to remove */
 18336	{
 18337	/* Remove server memory from the free memory list. The boot monitor
 18338	 * promises to put processes at the start of memory chunks. The 
 18339	 * tasks all use same base address, so only the first task changes
 18340	 * the memory lists. The servers and init have their own memory
 18341	 * spaces and their memory will be removed from the list. 
 18342	 */
 18343	  struct memory *memp;
 18344	  for (memp = mem_chunks; memp < &mem_chunks[NR_MEMS]; memp++) {
 18345	        if (memp->base == map_ptr[T].mem_phys) {
 18346	                memp->base += map_ptr[T].mem_len + map_ptr[D].mem_len;
 18347	                memp->size -= map_ptr[T].mem_len + map_ptr[D].mem_len;
 18348	        }
 18349	  }
 18350	}

_________________________ Page 881    File: servers/pm/main.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/forkexit.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 18400	/* This file deals with creating processes (via FORK) and deleting them (via
 18401	 * EXIT/WAIT).  When a process forks, a new slot in the 'mproc' table is
 18402	 * allocated for it, and a copy of the parent's core image is made for the
 18403	 * child.  Then the kernel and file system are informed.  A process is removed
 18404	 * from the 'mproc' table when two events have occurred: (1) it has exited or
 18405	 * been killed by a signal, and (2) the parent has done a WAIT.  If the process
 18406	 * exits first, it continues to occupy a slot until the parent does a WAIT.
 18407	 *
 18408	 * The entry points into this file are:
 18409	 *   do_fork:    perform the FORK system call
 18410	 *   do_pm_exit: perform the EXIT system call (by calling pm_exit())
 18411	 *   pm_exit:    actually do the exiting
 18412	 *   do_wait:    perform the WAITPID or WAIT system call
 18413	 */
 18414	
 18415	#include "pm.h"
 18416	#include <sys/wait.h>
 18417	#include <minix/callnr.h>
 18418	#include <minix/com.h>
 18419	#include <signal.h>
 18420	#include "mproc.h"
 18421	#include "param.h"
 18422	
 18423	#define LAST_FEW            2   /* last few slots reserved for superuser */
 18424	
 18425	FORWARD _PROTOTYPE (void cleanup, (register struct mproc *child) );
 18426	
 18427	/*===========================================================================*
 18428	 *                              do_fork                                      *
 18429	 *===========================================================================*/
 18430	PUBLIC int do_fork()
 18431	{
 18432	/* The process pointed to by 'mp' has forked.  Create a child process. */
 18433	  register struct mproc *rmp;   /* pointer to parent */
 18434	  register struct mproc *rmc;   /* pointer to child */
 18435	  int child_nr, s;
 18436	  phys_clicks prog_clicks, child_base;
 18437	  phys_bytes prog_bytes, parent_abs, child_abs; /* Intel only */
 18438	  pid_t new_pid;
 18439	
 18440	 /* If tables might fill up during FORK, don't even start since recovery half
 18441	  * way through is such a nuisance.
 18442	  */
 18443	  rmp = mp;
 18444	  if ((procs_in_use == NR_PROCS) || 
 18445	                (procs_in_use >= NR_PROCS-LAST_FEW && rmp->mp_effuid != 0))
 18446	  {
 18447	        printf("PM: warning, process table is full!\n");
 18448	        return(EAGAIN);
 18449	  }
 18450	
 18451	  /* Determine how much memory to allocate.  Only the data and stack need to
 18452	   * be copied, because the text segment is either shared or of zero length.
 18453	   */
 18454	  prog_clicks = (phys_clicks) rmp->mp_seg[S].mem_len;

_________________________ Page 882    File: servers/pm/forkexit.c ________________________
 18455	  prog_clicks += (rmp->mp_seg[S].mem_vir - rmp->mp_seg[D].mem_vir);
 18456	  prog_bytes = (phys_bytes) prog_clicks << CLICK_SHIFT;
 18457	  if ( (child_base = alloc_mem(prog_clicks)) == NO_MEM) return(ENOMEM);
 18458	
 18459	  /* Create a copy of the parent's core image for the child. */
 18460	  child_abs = (phys_bytes) child_base << CLICK_SHIFT;
 18461	  parent_abs = (phys_bytes) rmp->mp_seg[D].mem_phys << CLICK_SHIFT;
 18462	  s = sys_abscopy(parent_abs, child_abs, prog_bytes);
 18463	  if (s < 0) panic(__FILE__,"do_fork can't copy", s);
 18464	
 18465	  /* Find a slot in 'mproc' for the child process.  A slot must exist. */
 18466	  for (rmc = &mproc[0]; rmc < &mproc[NR_PROCS]; rmc++)
 18467	        if ( (rmc->mp_flags & IN_USE) == 0) break;
 18468	
 18469	  /* Set up the child and its memory map; copy its 'mproc' slot from parent. */
 18470	  child_nr = (int)(rmc - mproc);        /* slot number of the child */
 18471	  procs_in_use++;
 18472	  *rmc = *rmp;                  /* copy parent's process slot to child's */
 18473	  rmc->mp_parent = who;                 /* record child's parent */
 18474	  /* inherit only these flags */
 18475	  rmc->mp_flags &= (IN_USE|SEPARATE|PRIV_PROC|DONT_SWAP);
 18476	  rmc->mp_child_utime = 0;              /* reset administration */
 18477	  rmc->mp_child_stime = 0;              /* reset administration */
 18478	
 18479	  /* A separate I&D child keeps the parents text segment.  The data and stack
 18480	   * segments must refer to the new copy.
 18481	   */
 18482	  if (!(rmc->mp_flags & SEPARATE)) rmc->mp_seg[T].mem_phys = child_base;
 18483	  rmc->mp_seg[D].mem_phys = child_base;
 18484	  rmc->mp_seg[S].mem_phys = rmc->mp_seg[D].mem_phys + 
 18485	                        (rmp->mp_seg[S].mem_vir - rmp->mp_seg[D].mem_vir);
 18486	  rmc->mp_exitstatus = 0;
 18487	  rmc->mp_sigstatus = 0;
 18488	
 18489	  /* Find a free pid for the child and put it in the table. */
 18490	  new_pid = get_free_pid();
 18491	  rmc->mp_pid = new_pid;        /* assign pid to child */
 18492	
 18493	  /* Tell kernel and file system about the (now successful) FORK. */
 18494	  sys_fork(who, child_nr);
 18495	  tell_fs(FORK, who, child_nr, rmc->mp_pid);
 18496	
 18497	  /* Report child's memory map to kernel. */
 18498	  sys_newmap(child_nr, rmc->mp_seg);
 18499	
 18500	  /* Reply to child to wake it up. */
 18501	  setreply(child_nr, 0);                /* only parent gets details */
 18502	  rmp->mp_reply.procnr = child_nr;      /* child's process number */
 18503	  return(new_pid);                      /* child's pid */
 18504	}
	
 18506	/*===========================================================================*
 18507	 *                              do_pm_exit                                   *
 18508	 *===========================================================================*/
 18509	PUBLIC int do_pm_exit()
 18510	{
 18511	/* Perform the exit(status) system call. The real work is done by pm_exit(),
 18512	 * which is also called when a process is killed by a signal.
 18513	 */
 18514	  pm_exit(mp, m_in.status);

_________________________ Page 883    File: servers/pm/forkexit.c ________________________
 18515	  return(SUSPEND);              /* can't communicate from beyond the grave */
 18516	}
	
 18518	/*===========================================================================*
 18519	 *                              pm_exit                                      *
 18520	 *===========================================================================*/
 18521	PUBLIC void pm_exit(rmp, exit_status)
 18522	register struct mproc *rmp;     /* pointer to the process to be terminated */
 18523	int exit_status;                /* the process' exit status (for parent) */
 18524	{
 18525	/* A process is done.  Release most of the process' possessions.  If its
 18526	 * parent is waiting, release the rest, else keep the process slot and
 18527	 * become a zombie.
 18528	 */
 18529	  register int proc_nr;
 18530	  int parent_waiting, right_child;
 18531	  pid_t pidarg, procgrp;
 18532	  struct mproc *p_mp;
 18533	  clock_t t[5];
 18534	
 18535	  proc_nr = (int) (rmp - mproc);        /* get process slot number */
 18536	
 18537	  /* Remember a session leader's process group. */
 18538	  procgrp = (rmp->mp_pid == mp->mp_procgrp) ? mp->mp_procgrp : 0;
 18539	
 18540	  /* If the exited process has a timer pending, kill it. */
 18541	  if (rmp->mp_flags & ALARM_ON) set_alarm(proc_nr, (unsigned) 0);
 18542	
 18543	  /* Do accounting: fetch usage times and accumulate at parent. */
 18544	  sys_times(proc_nr, t);
 18545	  p_mp = &mproc[rmp->mp_parent];                        /* process' parent */
 18546	  p_mp->mp_child_utime += t[0] + rmp->mp_child_utime;   /* add user time */
 18547	  p_mp->mp_child_stime += t[1] + rmp->mp_child_stime;   /* add system time */
 18548	
 18549	  /* Tell the kernel and FS that the process is no longer runnable. */
 18550	  tell_fs(EXIT, proc_nr, 0, 0);  /* file system can free the proc slot */
 18551	  sys_exit(proc_nr);
 18552	
 18553	  /* Pending reply messages for the dead process cannot be delivered. */
 18554	  rmp->mp_flags &= ~REPLY;
 18555	  
 18556	  /* Release the memory occupied by the child. */
 18557	  if (find_share(rmp, rmp->mp_ino, rmp->mp_dev, rmp->mp_ctime) == NULL) {
 18558	        /* No other process shares the text segment, so free it. */
 18559	        free_mem(rmp->mp_seg[T].mem_phys, rmp->mp_seg[T].mem_len);
 18560	  }
 18561	  /* Free the data and stack segments. */
 18562	  free_mem(rmp->mp_seg[D].mem_phys,
 18563	      rmp->mp_seg[S].mem_vir 
 18564	        + rmp->mp_seg[S].mem_len - rmp->mp_seg[D].mem_vir);
 18565	
 18566	  /* The process slot can only be freed if the parent has done a WAIT. */
 18567	  rmp->mp_exitstatus = (char) exit_status;
 18568	
 18569	  pidarg = p_mp->mp_wpid;               /* who's being waited for? */
 18570	  parent_waiting = p_mp->mp_flags & WAITING;
 18571	  right_child =                         /* child meets one of the 3 tests? */
 18572	        (pidarg == -1 || pidarg == rmp->mp_pid || -pidarg == rmp->mp_procgrp);
 18573	
 18574	  if (parent_waiting && right_child) {

_________________________ Page 884    File: servers/pm/forkexit.c ________________________
 18575	        cleanup(rmp);                   /* tell parent and release child slot */
 18576	  } else {
 18577	        rmp->mp_flags = IN_USE|ZOMBIE;  /* parent not waiting, zombify child */
 18578	        sig_proc(p_mp, SIGCHLD);        /* send parent a "child died" signal */
 18579	  }
 18580	
 18581	  /* If the process has children, disinherit them.  INIT is the new parent. */
 18582	  for (rmp = &mproc[0]; rmp < &mproc[NR_PROCS]; rmp++) {
 18583	        if (rmp->mp_flags & IN_USE && rmp->mp_parent == proc_nr) {
 18584	                /* 'rmp' now points to a child to be disinherited. */
 18585	                rmp->mp_parent = INIT_PROC_NR;
 18586	                parent_waiting = mproc[INIT_PROC_NR].mp_flags & WAITING;
 18587	                if (parent_waiting && (rmp->mp_flags & ZOMBIE)) cleanup
 18588	        }
 18589	  }
 18590	
 18591	  /* Send a hangup to the process' process group if it was a session leader. */
 18592	  if (procgrp != 0) check_sig(-procgrp, SIGHUP);
 18593	}
	
 18595	/*===========================================================================*
 18596	 *                              do_waitpid                                   *
 18597	 *===========================================================================*/
 18598	PUBLIC int do_waitpid()
 18599	{
 18600	/* A process wants to wait for a child to terminate. If a child is already 
 18601	 * waiting, go clean it up and let this WAIT call terminate.  Otherwise, 
 18602	 * really wait. 
 18603	 * A process calling WAIT never gets a reply in the usual way at the end
 18604	 * of the main loop (unless WNOHANG is set or no qualifying child exists).
 18605	 * If a child has already exited, the routine cleanup() sends the reply
 18606	 * to awaken the caller.
 18607	 * Both WAIT and WAITPID are handled by this code.
 18608	 */
 18609	  register struct mproc *rp;
 18610	  int pidarg, options, children;
 18611	
 18612	  /* Set internal variables, depending on whether this is WAIT or WAITPID. */
 18613	  pidarg  = (call_nr == WAIT ? -1 : m_in.pid);     /* 1st param of waitpid */
 18614	  options = (call_nr == WAIT ?  0 : m_in.sig_nr);  /* 3rd param of waitpid */
 18615	  if (pidarg == 0) pidarg = -mp->mp_procgrp;    /* pidarg < 0 ==> proc grp */
 18616	
 18617	  /* Is there a child waiting to be collected? At this point, pidarg != 0:
 18618	   *    pidarg  >  0 means pidarg is pid of a specific process to wait for
 18619	   *    pidarg == -1 means wait for any child
 18620	   *    pidarg  < -1 means wait for any child whose process group = -pidarg
 18621	   */
 18622	  children = 0;
 18623	  for (rp = &mproc[0]; rp < &mproc[NR_PROCS]; rp++) {
 18624	        if ( (rp->mp_flags & IN_USE) && rp->mp_parent == who) {
 18625	                /* The value of pidarg determines which children qualify. */
 18626	                if (pidarg  > 0 && pidarg != rp->mp_pid) continue;
 18627	                if (pidarg < -1 && -pidarg != rp->mp_procgrp) continue;
 18628	
 18629	                children++;             /* this child is acceptable */
 18630	                if (rp->mp_flags & ZOMBIE) {
 18631	                        /* This child meets the pid test and has exited. */
 18632	                        cleanup(rp);    /* this child has already exited */
 18633	                        return(SUSPEND);
 18634	                }

_________________________ Page 885    File: servers/pm/forkexit.c ________________________
 18635	                if ((rp->mp_flags & STOPPED) && rp->mp_sigstatus) {
 18636	                        /* This child meets the pid test and is being traced.*/
 18637	                        mp->mp_reply.reply_res2 = 0177|(rp->mp_sigstatus << 8
 18638	                        rp->mp_sigstatus = 0;
 18639	                        return(rp->mp_pid);
 18640	                }
 18641	        }
 18642	  }
 18643	
 18644	  /* No qualifying child has exited.  Wait for one, unless none exists. */
 18645	  if (children > 0) {
 18646	        /* At least 1 child meets the pid test exists, but has not exited. */
 18647	        if (options & WNOHANG) return(0);    /* parent does not want to wait */
 18648	        mp->mp_flags |= WAITING;             /* parent wants to wait */
 18649	        mp->mp_wpid = (pid_t) pidarg;        /* save pid for later */
 18650	        return(SUSPEND);                     /* do not reply, let it wait */
 18651	  } else {
 18652	        /* No child even meets the pid test.  Return error immediately. */
 18653	        return(ECHILD);                      /* no - parent has no children */
 18654	  }
 18655	}
	
 18657	/*===========================================================================*
 18658	 *                              cleanup                                      *
 18659	 *===========================================================================*/
 18660	PRIVATE void cleanup(child)
 18661	register struct mproc *child;   /* tells which process is exiting */
 18662	{
 18663	/* Finish off the exit of a process.  The process has exited or been killed
 18664	 * by a signal, and its parent is waiting.
 18665	 */
 18666	  struct mproc *parent = &mproc[child->mp_parent];
 18667	  int exitstatus;
 18668	
 18669	  /* Wake up the parent by sending the reply message. */
 18670	  exitstatus = (child->mp_exitstatus << 8) | (child->mp_sigstatus & 0377)
 18671	  parent->mp_reply.reply_res2 = exitstatus;
 18672	  setreply(child->mp_parent, child->mp_pid);
 18673	  parent->mp_flags &= ~WAITING;         /* parent no longer waiting */
 18674	
 18675	  /* Release the process table entry and reinitialize some field. */
 18676	  child->mp_pid = 0;
 18677	  child->mp_flags = 0;
 18678	  child->mp_child_utime = 0;
 18679	  child->mp_child_stime = 0;
 18680	  procs_in_use--;
 18681	}
	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/exec.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 18700	/* This file handles the EXEC system call.  It performs the work as follows:
 18701	 *    - see if the permissions allow the file to be executed
 18702	 *    - read the header and extract the sizes
 18703	 *    - fetch the initial args and environment from the user space
 18704	 *    - allocate the memory for the new process

_________________________ Page 886    File: servers/pm/exec.c _________________________
 18705	 *    - copy the initial stack from PM to the process
 18706	 *    - read in the text and data segments and copy to the process
 18707	 *    - take care of setuid and setgid bits
 18708	 *    - fix up 'mproc' table
 18709	 *    - tell kernel about EXEC
 18710	 *    - save offset to initial argc (for ps)
 18711	 *
 18712	 * The entry points into this file are:
 18713	 *   do_exec:    perform the EXEC system call
 18714	 *   rw_seg:     read or write a segment from or to a file
 18715	 *   find_share: find a process whose text segment can be shared
 18716	 */
 18717	
 18718	#include "pm.h"
 18719	#include <sys/stat.h>
 18720	#include <minix/callnr.h>
 18721	#include <minix/com.h>
 18722	#include <a.out.h>
 18723	#include <signal.h>
 18724	#include <string.h>
 18725	#include "mproc.h"
 18726	#include "param.h"
 18727	
 18728	FORWARD _PROTOTYPE( int new_mem, (struct mproc *sh_mp, vir_bytes text_bytes,
 18729	                vir_bytes data_bytes, vir_bytes bss_bytes,
 18730	                vir_bytes stk_bytes, phys_bytes tot_bytes)              );
 18731	FORWARD _PROTOTYPE( void patch_ptr, (char stack[ARG_MAX], vir_bytes base) );
 18732	FORWARD _PROTOTYPE( int insert_arg, (char stack[ARG_MAX],
 18733	                vir_bytes *stk_bytes, char *arg, int replace)           );
 18734	FORWARD _PROTOTYPE( char *patch_stack, (int fd, char stack[ARG_MAX],
 18735	                vir_bytes *stk_bytes, char *script)                     );
 18736	FORWARD _PROTOTYPE( int read_header, (int fd, int *ft, vir_bytes *text_bytes,
 18737	                vir_bytes *data_bytes, vir_bytes *bss_bytes,
 18738	                phys_bytes *tot_bytes, long *sym_bytes, vir_clicks sc,
 18739	                vir_bytes *pc)                                          );
 18740	
 18741	#define ESCRIPT (-2000) /* Returned by read_header for a #! script. */
 18742	#define PTRSIZE sizeof(char *) /* Size of pointers in argv[] and envp[]. */
 18743	
 18744	/*===========================================================================*
 18745	 *                              do_exec                                      *
 18746	 *===========================================================================*/
 18747	PUBLIC int do_exec()
 18748	{
 18749	/* Perform the execve(name, argv, envp) call.  The user library builds a
 18750	 * complete stack image, including pointers, args, environ, etc.  The stack
 18751	 * is copied to a buffer inside PM, and then to the new core image.
 18752	 */
 18753	  register struct mproc *rmp;
 18754	  struct mproc *sh_mp;
 18755	  int m, r, fd, ft, sn;
 18756	  static char mbuf[ARG_MAX];    /* buffer for stack and zeroes */
 18757	  static char name_buf[PATH_MAX]; /* the name of the file to exec */
 18758	  char *new_sp, *name, *basename;
 18759	  vir_bytes src, dst, text_bytes, data_bytes, bss_bytes, stk_bytes, vsp;
 18760	  phys_bytes tot_bytes;         /* total space for program, including gap */
 18761	  long sym_bytes;
 18762	  vir_clicks sc;
 18763	  struct stat s_buf[2], *s_p;
 18764	  vir_bytes pc;

_________________________ Page 887    File: servers/pm/exec.c _________________________
 18765	
 18766	  /* Do some validity checks. */
 18767	  rmp = mp;
 18768	  stk_bytes = (vir_bytes) m_in.stack_bytes;
 18769	  if (stk_bytes > ARG_MAX) return(ENOMEM);      /* stack too big */
 18770	  if (m_in.exec_len <= 0 || m_in.exec_len > PATH_MAX) return(EINVAL);
 18771	
 18772	  /* Get the exec file name and see if the file is executable. */
 18773	  src = (vir_bytes) m_in.exec_name;
 18774	  dst = (vir_bytes) name_buf;
 18775	  r = sys_datacopy(who, (vir_bytes) src,
 18776	                PM_PROC_NR, (vir_bytes) dst, (phys_bytes) m_in.exec_len);
 18777	  if (r != OK) return(r);       /* file name not in user data segment */
 18778	
 18779	  /* Fetch the stack from the user before destroying the old core image. */
 18780	  src = (vir_bytes) m_in.stack_ptr;
 18781	  dst = (vir_bytes) mbuf;
 18782	  r = sys_datacopy(who, (vir_bytes) src,
 18783	                        PM_PROC_NR, (vir_bytes) dst, (phys_bytes)stk_bytes);
 18784	  /* can't fetch stack (e.g. bad virtual addr) */
 18785	  if (r != OK) return(EACCES);  
 18786	
 18787	  r = 0;        /* r = 0 (first attempt), or 1 (interpreted script) */
 18788	  name = name_buf;      /* name of file to exec. */
 18789	  do {
 18790	        s_p = &s_buf[r];
 18791	        tell_fs(CHDIR, who, FALSE, 0);  /* switch to the user's FS environ */
 18792	        fd = allowed(name, s_p, X_BIT); /* is file executable? */
 18793	        if (fd < 0)  return(fd);                /* file was not executable */
 18794	
 18795	        /* Read the file header and extract the segment sizes. */
 18796	        sc = (stk_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 18797	
 18798	        m = read_header(fd, &ft, &text_bytes, &data_bytes, &bss_byt
 18799	                                        &tot_bytes, &sym_bytes, sc, &pc
 18800	        if (m != ESCRIPT || ++r > 1) break;
 18801	  } while ((name = patch_stack(fd, mbuf, &stk_bytes, name_buf)) != NULL);
 18802	
 18803	  if (m < 0) {
 18804	        close(fd);              /* something wrong with header */
 18805	        return(stk_bytes > ARG_MAX ? ENOMEM : ENOEXEC);
 18806	  }
 18807	
 18808	  /* Can the process' text be shared with that of one already running? */
 18809	  sh_mp = find_share(rmp, s_p->st_ino, s_p->st_dev, s_p->st_ctime);
 18810	
 18811	  /* Allocate new memory and release old memory.  Fix map and tell kernel. */
 18812	  r = new_mem(sh_mp, text_bytes, data_bytes, bss_bytes, stk_bytes, tot_bytes);
 18813	  if (r != OK) {
 18814	        close(fd);              /* insufficient core or program too big */
 18815	        return(r);
 18816	  }
 18817	
 18818	  /* Save file identification to allow it to be shared. */
 18819	  rmp->mp_ino = s_p->st_ino;
 18820	  rmp->mp_dev = s_p->st_dev;
 18821	  rmp->mp_ctime = s_p->st_ctime;
 18822	
 18823	  /* Patch up stack and copy it from PM to new core image. */
 18824	  vsp = (vir_bytes) rmp->mp_seg[S].mem_vir << CLICK_SHIFT;

_________________________ Page 888    File: servers/pm/exec.c _________________________
 18825	  vsp += (vir_bytes) rmp->mp_seg[S].mem_len << CLICK_SHIFT;
 18826	  vsp -= stk_bytes;
 18827	  patch_ptr(mbuf, vsp);
 18828	  src = (vir_bytes) mbuf;
 18829	  r = sys_datacopy(PM_PROC_NR, (vir_bytes) src,
 18830	                        who, (vir_bytes) vsp, (phys_bytes)stk_bytes);
 18831	  if (r != OK) panic(__FILE__,"do_exec stack copy err on", who);
 18832	
 18833	  /* Read in text and data segments. */
 18834	  if (sh_mp != NULL) {
 18835	        lseek(fd, (off_t) text_bytes, SEEK_CUR);  /* shared: skip text */
 18836	  } else {
 18837	        rw_seg(0, fd, who, T, text_bytes);
 18838	  }
 18839	  rw_seg(0, fd, who, D, data_bytes);
 18840	
 18841	  close(fd);                    /* don't need exec file any more */
 18842	
 18843	  /* Take care of setuid/setgid bits. */
 18844	  if ((rmp->mp_flags & TRACED) == 0) { /* suppress if tracing */
 18845	        if (s_buf[0].st_mode & I_SET_UID_BIT) {
 18846	                rmp->mp_effuid = s_buf[0].st_uid;
 18847	                tell_fs(SETUID,who, (int)rmp->mp_realuid, (int)rmp->mp_effuid);
 18848	        }
 18849	        if (s_buf[0].st_mode & I_SET_GID_BIT) {
 18850	                rmp->mp_effgid = s_buf[0].st_gid;
 18851	                tell_fs(SETGID,who, (int)rmp->mp_realgid, (int)rmp->mp_effgid);
 18852	        }
 18853	  }
 18854	
 18855	  /* Save offset to initial argc (for ps) */
 18856	  rmp->mp_procargs = vsp;
 18857	
 18858	  /* Fix 'mproc' fields, tell kernel that exec is done,  reset caught sigs. */
 18859	  for (sn = 1; sn <= _NSIG; sn++) {
 18860	        if (sigismember(&rmp->mp_catch, sn)) {
 18861	                sigdelset(&rmp->mp_catch, sn);
 18862	                rmp->mp_sigact[sn].sa_handler = SIG_DFL;
 18863	                sigemptyset(&rmp->mp_sigact[sn].sa_mask);
 18864	        }
 18865	  }
 18866	
 18867	  rmp->mp_flags &= ~SEPARATE;   /* turn off SEPARATE bit */
 18868	  rmp->mp_flags |= ft;          /* turn it on for separate I & D files */
 18869	  new_sp = (char *) vsp;
 18870	
 18871	  tell_fs(EXEC, who, 0, 0);     /* allow FS to handle FD_CLOEXEC files */
 18872	
 18873	  /* System will save command line for debugging, ps(1) output, etc. */
 18874	  basename = strrchr(name, '/');
 18875	  if (basename == NULL) basename = name; else basename++;
 18876	  strncpy(rmp->mp_name, basename, PROC_NAME_LEN-1);
 18877	  rmp->mp_name[PROC_NAME_LEN] = '\0';
 18878	  sys_exec(who, new_sp, basename, pc);
 18879	
 18880	  /* Cause a signal if this process is traced. */
 18881	  if (rmp->mp_flags & TRACED) check_sig(rmp->mp_pid, SIGTRAP);
 18882	
 18883	  return(SUSPEND);              /* no reply, new program just runs */
 18884	}

_________________________ Page 889    File: servers/pm/exec.c _________________________
	
 18886	/*===========================================================================*
 18887	 *                              read_header                                  *
 18888	 *===========================================================================*/
 18889	PRIVATE int read_header(fd, ft, text_bytes, data_bytes, bss_bytes, 
 18890	                                                tot_bytes, sym_bytes, sc, pc)
 18891	int fd;                         /* file descriptor for reading exec file */
 18892	int *ft;                        /* place to return ft number */
 18893	vir_bytes *text_bytes;          /* place to return text size */
 18894	vir_bytes *data_bytes;          /* place to return initialized data size */
 18895	vir_bytes *bss_bytes;           /* place to return bss size */
 18896	phys_bytes *tot_bytes;          /* place to return total size */
 18897	long *sym_bytes;                /* place to return symbol table size */
 18898	vir_clicks sc;                  /* stack size in clicks */
 18899	vir_bytes *pc;                  /* program entry point (initial PC) */
 18900	{
 18901	/* Read the header and extract the text, data, bss and total sizes from it. */
 18902	
 18903	  int m, ct;
 18904	  vir_clicks tc, dc, s_vir, dvir;
 18905	  phys_clicks totc;
 18906	  struct exec hdr;              /* a.out header is read in here */
 18907	
 18908	  /* Read the header and check the magic number.  The standard MINIX header 
 18909	   * is defined in <a.out.h>.  It consists of 8 chars followed by 6 longs.
 18910	   * Then come 4 more longs that are not used here.
 18911	   *    Byte 0: magic number 0x01
 18912	   *    Byte 1: magic number 0x03
 18913	   *    Byte 2: normal = 0x10 (not checked, 0 is OK), separate I/D = 0x20
 18914	   *    Byte 3: CPU type, Intel 16 bit = 0x04, Intel 32 bit = 0x10, 
 18915	   *            Motorola = 0x0B, Sun SPARC = 0x17
 18916	   *    Byte 4: Header length = 0x20
 18917	   *    Bytes 5-7 are not used.
 18918	   *
 18919	   *    Now come the 6 longs
 18920	   *    Bytes  8-11: size of text segments in bytes
 18921	   *    Bytes 12-15: size of initialized data segment in bytes
 18922	   *    Bytes 16-19: size of bss in bytes
 18923	   *    Bytes 20-23: program entry point
 18924	   *    Bytes 24-27: total memory allocated to program (text, data + stack)
 18925	   *    Bytes 28-31: size of symbol table in bytes
 18926	   * The longs are represented in a machine dependent order,
 18927	   * little-endian on the 8088, big-endian on the 68000.
 18928	   * The header is followed directly by the text and data segments, and the 
 18929	   * symbol table (if any). The sizes are given in the header. Only the 
 18930	   * text and data segments are copied into memory by exec. The header is 
 18931	   * used here only. The symbol table is for the benefit of a debugger and 
 18932	   * is ignored here.
 18933	   */
 18934	
 18935	  if ((m= read(fd, &hdr, A_MINHDR)) < 2) return(ENOEXEC);
 18936	
 18937	  /* Interpreted script? */
 18938	  if (((char *) &hdr)[0] == '#' && ((char *) &hdr)[1] == '!') retur
 18939	
 18940	  if (m != A_MINHDR) return(ENOEXEC);
 18941	
 18942	  /* Check magic number, cpu type, and flags. */
 18943	  if (BADMAG(hdr)) return(ENOEXEC);
 18944	  if (hdr.a_cpu != A_I80386) return(ENOEXEC);

_________________________ Page 890    File: servers/pm/exec.c _________________________
 18945	  if ((hdr.a_flags & ~(A_NSYM | A_EXEC | A_SEP)) != 0) return(ENOEXEC);
 18946	
 18947	  *ft = ( (hdr.a_flags & A_SEP) ? SEPARATE : 0);    /* separate I & D or no
 18948	
 18949	  /* Get text and data sizes. */
 18950	  *text_bytes = (vir_bytes) hdr.a_text; /* text size in bytes */
 18951	  *data_bytes = (vir_bytes) hdr.a_data; /* data size in bytes */
 18952	  *bss_bytes  = (vir_bytes) hdr.a_bss;  /* bss size in bytes */
 18953	  *tot_bytes  = hdr.a_total;            /* total bytes to allocate for prog */
 18954	  *sym_bytes  = hdr.a_syms;             /* symbol table size in bytes */
 18955	  if (*tot_bytes == 0) return(ENOEXEC);
 18956	
 18957	  if (*ft != SEPARATE) {
 18958	        /* If I & D space is not separated, it is all considered data. Text=0*/
 18959	        *data_bytes += *text_bytes;
 18960	        *text_bytes = 0;
 18961	  }
 18962	  *pc = hdr.a_entry;    /* initial address to start execution */
 18963	
 18964	  /* Check to see if segment sizes are feasible. */
 18965	  tc = ((unsigned long) *text_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 18966	  dc = (*data_bytes + *bss_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 18967	  totc = (*tot_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 18968	  if (dc >= totc) return(ENOEXEC);      /* stack must be at least 1 click */
 18969	  dvir = (*ft == SEPARATE ? 0 : tc);
 18970	  s_vir = dvir + (totc - sc);
 18971	  m = (dvir + dc > s_vir) ? ENOMEM : OK;
 18972	  ct = hdr.a_hdrlen & BYTE;             /* header length */
 18973	  if (ct > A_MINHDR) lseek(fd, (off_t) ct, SEEK_SET); /* skip unused hdr */
 18974	  return(m);
 18975	}
	
 18977	/*===========================================================================*
 18978	 *                              new_mem                                      *
 18979	 *===========================================================================*/
 18980	PRIVATE int new_mem(sh_mp, text_bytes, data_bytes,
 18981	        bss_bytes,stk_bytes,tot_bytes)
 18982	struct mproc *sh_mp;            /* text can be shared with this process */
 18983	vir_bytes text_bytes;           /* text segment size in bytes */
 18984	vir_bytes data_bytes;           /* size of initialized data in bytes */
 18985	vir_bytes bss_bytes;            /* size of bss in bytes */
 18986	vir_bytes stk_bytes;            /* size of initial stack segment in bytes */
 18987	phys_bytes tot_bytes;           /* total memory to allocate, including gap */
 18988	{
 18989	/* Allocate new memory and release the old memory.  Change the map and report
 18990	 * the new map to the kernel.  Zero the new core image's bss, gap and stack.
 18991	 */
 18992	
 18993	  register struct mproc *rmp = mp;
 18994	  vir_clicks text_clicks, data_clicks, gap_clicks, stack_clicks, tot_clicks;
 18995	  phys_clicks new_base;
 18996	  phys_bytes bytes, base, bss_offset;
 18997	  int s;
 18998	
 18999	  /* No need to allocate text if it can be shared. */
 19000	  if (sh_mp != NULL) text_bytes = 0;
 19001	
 19002	  /* Allow the old data to be swapped out to make room.  (Which is really a
 19003	   * waste of time, because we are going to throw it away anyway.)
 19004	   */

_________________________ Page 891    File: servers/pm/exec.c _________________________
 19005	  rmp->mp_flags |= WAITING;
 19006	
 19007	  /* Acquire the new memory.  Each of the 4 parts: text, (data+bss), gap,
 19008	   * and stack occupies an integral number of clicks, starting at click
 19009	   * boundary.  The data and bss parts are run together with no space.
 19010	   */
 19011	  text_clicks = ((unsigned long) text_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 19012	  data_clicks = (data_bytes + bss_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 19013	  stack_clicks = (stk_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 19014	  tot_clicks = (tot_bytes + CLICK_SIZE - 1) >> CLICK_SHIFT;
 19015	  gap_clicks = tot_clicks - data_clicks - stack_clicks;
 19016	  if ( (int) gap_clicks < 0) return(ENOMEM);
 19017	
 19018	  /* Try to allocate memory for the new process. */
 19019	  new_base = alloc_mem(text_clicks + tot_clicks);
 19020	  if (new_base == NO_MEM) return(ENOMEM);
 19021	
 19022	  /* We've got memory for the new core image.  Release the old one. */
 19023	  rmp = mp;
 19024	
 19025	  if (find_share(rmp, rmp->mp_ino, rmp->mp_dev, rmp->mp_ctime) == NULL) {
 19026	        /* No other process shares the text segment, so free it. */
 19027	        free_mem(rmp->mp_seg[T].mem_phys, rmp->mp_seg[T].mem_len);
 19028	  }
 19029	  /* Free the data and stack segments. */
 19030	  free_mem(rmp->mp_seg[D].mem_phys,
 19031	   rmp->mp_seg[S].mem_vir + rmp->mp_seg[S].mem_len - rmp->mp_seg[D].mem_vir);
 19032	
 19033	  /* We have now passed the point of no return.  The old core image has been
 19034	   * forever lost, memory for a new core image has been allocated.  Set up
 19035	   * and report new map.
 19036	   */
 19037	  if (sh_mp != NULL) {
 19038	        /* Share the text segment. */
 19039	        rmp->mp_seg[T] = sh_mp->mp_seg[T];
 19040	  } else {
 19041	        rmp->mp_seg[T].mem_phys = new_base;
 19042	        rmp->mp_seg[T].mem_vir = 0;
 19043	        rmp->mp_seg[T].mem_len = text_clicks;
 19044	  }
 19045	  rmp->mp_seg[D].mem_phys = new_base + text_clicks;
 19046	  rmp->mp_seg[D].mem_vir = 0;
 19047	  rmp->mp_seg[D].mem_len = data_clicks;
 19048	  rmp->mp_seg[S].mem_phys = rmp->mp_seg[D].mem_phys + data_clicks + gap_clicks;
 19049	  rmp->mp_seg[S].mem_vir = rmp->mp_seg[D].mem_vir + data_clicks + gap_clicks;
 19050	  rmp->mp_seg[S].mem_len = stack_clicks;
 19051	
 19052	  sys_newmap(who, rmp->mp_seg);   /* report new map to the kernel */
 19053	
 19054	  /* The old memory may have been swapped out, but the new memory is real. */
 19055	  rmp->mp_flags &= ~(WAITING|ONSWAP|SWAPIN);
 19056	
 19057	  /* Zero the bss, gap, and stack segment. */
 19058	  bytes = (phys_bytes)(data_clicks + gap_clicks + stack_clicks) << CLICK_SHIF
 19059	  base = (phys_bytes) rmp->mp_seg[D].mem_phys << CLICK_SHIFT;
 19060	  bss_offset = (data_bytes >> CLICK_SHIFT) << CLICK_SHIFT;
 19061	  base += bss_offset;
 19062	  bytes -= bss_offset;
 19063	
 19064	  if ((s=sys_memset(0, base, bytes)) != OK) {

_________________________ Page 892    File: servers/pm/exec.c _________________________
 19065	        panic(__FILE__,"new_mem can't zero", s);
 19066	  }
 19067	
 19068	  return(OK);
 19069	}
	
 19071	/*===========================================================================*
 19072	 *                              patch_ptr                                    *
 19073	 *===========================================================================*/
 19074	PRIVATE void patch_ptr(stack, base)
 19075	char stack[ARG_MAX];            /* pointer to stack image within PM */
 19076	vir_bytes base;                 /* virtual address of stack base inside user */
 19077	{
 19078	/* When doing an exec(name, argv, envp) call, the user builds up a stack
 19079	 * image with arg and env pointers relative to the start of the stack.  Now
 19080	 * these pointers must be relocated, since the stack is not positioned at
 19081	 * address 0 in the user's address space.
 19082	 */
 19083	
 19084	  char **ap, flag;
 19085	  vir_bytes v;
 19086	
 19087	  flag = 0;                     /* counts number of 0-pointers seen */
 19088	  ap = (char **) stack;         /* points initially to 'nargs' */
 19089	  ap++;                         /* now points to argv[0] */
 19090	  while (flag < 2) {
 19091	        if (ap >= (char **) &stack[ARG_MAX]) return;    /* too bad */
 19092	        if (*ap != NULL) {
 19093	                v = (vir_bytes) *ap;    /* v is relative pointer */
 19094	                v += base;              /* relocate it */
 19095	                *ap = (char *) v;       /* put it back */
 19096	        } else {
 19097	                flag++;
 19098	        }
 19099	        ap++;
 19100	  }
 19101	}
	
 19103	/*===========================================================================*
 19104	 *                              insert_arg                                   *
 19105	 *===========================================================================*/
 19106	PRIVATE int insert_arg(stack, stk_bytes, arg, replace)
 19107	char stack[ARG_MAX];            /* pointer to stack image within PM */
 19108	vir_bytes *stk_bytes;           /* size of initial stack */
 19109	char *arg;                      /* argument to prepend/replace as new argv[0] */
 19110	int replace;
 19111	{
 19112	/* Patch the stack so that arg will become argv[0].  Be careful, the stack may
 19113	 * be filled with garbage, although it normally looks like this:
 19114	 *      nargs argv[0] ... argv[nargs-1] NULL envp[0] ... NULL
 19115	 * followed by the strings "pointed" to by the argv[i] and the envp[i].  The
 19116	 * pointers are really offsets from the start of stack.
 19117	 * Return true iff the operation succeeded.
 19118	 */
 19119	  int offset, a0, a1, old_bytes = *stk_bytes;
 19120	
 19121	  /* Prepending arg adds at least one string and a zero byte. */
 19122	  offset = strlen(arg) + 1;
 19123	
 19124	  a0 = (int) ((char **) stack)[1];      /* argv[0] */

_________________________ Page 893    File: servers/pm/exec.c _________________________
 19125	  if (a0 < 4 * PTRSIZE || a0 >= old_bytes) return(FALSE);
 19126	
 19127	  a1 = a0;                      /* a1 will point to the strings to be moved */
 19128	  if (replace) {
 19129	        /* Move a1 to the end of argv[0][] (argv[1] if nargs > 1). */
 19130	        do {
 19131	                if (a1 == old_bytes) return(FALSE);
 19132	                --offset;
 19133	        } while (stack[a1++] != 0);
 19134	  } else {
 19135	        offset += PTRSIZE;      /* new argv[0] needs new pointer in argv[] */
 19136	        a0 += PTRSIZE;          /* location of new argv[0][]. */
 19137	  }
 19138	
 19139	  /* stack will grow by offset bytes (or shrink by -offset bytes) */
 19140	  if ((*stk_bytes += offset) > ARG_MAX) return(FALSE);
 19141	
 19142	  /* Reposition the strings by offset bytes */
 19143	  memmove(stack + a1 + offset, stack + a1, old_bytes - a1);
 19144	
 19145	  strcpy(stack + a0, arg);      /* Put arg in the new space. */
 19146	
 19147	  if (!replace) {
 19148	        /* Make space for a new argv[0]. */
 19149	        memmove(stack + 2 * PTRSIZE, stack + 1 * PTRSIZE, a0 - 2 * PTRSIZE);
 19150	
 19151	        ((char **) stack)[0]++; /* nargs++; */
 19152	  }
 19153	  /* Now patch up argv[] and envp[] by offset. */
 19154	  patch_ptr(stack, (vir_bytes) offset);
 19155	  ((char **) stack)[1] = (char *) a0;   /* set argv[0] correctly */
 19156	  return(TRUE);
 19157	}
	
 19159	/*===========================================================================*
 19160	 *                              patch_stack                                  *
 19161	 *===========================================================================*/
 19162	PRIVATE char *patch_stack(fd, stack, stk_bytes, script)
 19163	int fd;                         /* file descriptor to open script file */
 19164	char stack[ARG_MAX];            /* pointer to stack image within PM */
 19165	vir_bytes *stk_bytes;           /* size of initial stack */
 19166	char *script;                   /* name of script to interpret */
 19167	{
 19168	/* Patch the argument vector to include the path name of the script to be
 19169	 * interpreted, and all strings on the #! line.  Returns the path name of
 19170	 * the interpreter.
 19171	 */
 19172	  char *sp, *interp = NULL;
 19173	  int n;
 19174	  enum { INSERT=FALSE, REPLACE=TRUE };
 19175	
 19176	  /* Make script[] the new argv[0]. */
 19177	  if (!insert_arg(stack, stk_bytes, script, REPLACE)) return(NULL);
 19178	
 19179	  if (lseek(fd, 2L, 0) == -1                    /* just behind the #! */
 19180	    || (n= read(fd, script, PATH_MAX)) < 0      /* read line one */
 19181	    || (sp= memchr(script, '\n', n)) == NULL)   /* must be a proper line */
 19182	        return(NULL);
 19183	
 19184	  /* Move sp backwards through script[], prepending each string to stack. */

_________________________ Page 894    File: servers/pm/exec.c _________________________
 19185	  for (;;) {
 19186	        /* skip spaces behind argument. */
 19187	        while (sp > script && (*--sp == ' ' || *sp == '\t')) {}
 19188	        if (sp == script) break;
 19189	
 19190	        sp[1] = 0;
 19191	        /* Move to the start of the argument. */
 19192	        while (sp > script && sp[-1] != ' ' && sp[-1] != '\t') --sp
 19193	
 19194	        interp = sp;
 19195	        if (!insert_arg(stack, stk_bytes, sp, INSERT)) return(NULL);
 19196	  }
 19197	
 19198	  /* Round *stk_bytes up to the size of a pointer for alignment contraints. */
 19199	  *stk_bytes= ((*stk_bytes + PTRSIZE - 1) / PTRSIZE) * PTRSIZE;
 19200	
 19201	  close(fd);
 19202	  return(interp);
 19203	}
	
 19205	/*===========================================================================*
 19206	 *                              rw_seg                                       *
 19207	 *===========================================================================*/
 19208	PUBLIC void rw_seg(rw, fd, proc, seg, seg_bytes0)
 19209	int rw;                         /* 0 = read, 1 = write */
 19210	int fd;                         /* file descriptor to read from / write to */
 19211	int proc;                       /* process number */
 19212	int seg;                        /* T, D, or S */
 19213	phys_bytes seg_bytes0;          /* how much is to be transferred? */
 19214	{
 19215	/* Transfer text or data from/to a file and copy to/from a process segment.
 19216	 * This procedure is a little bit tricky.  The logical way to transfer a
 19217	 * segment would be block by block and copying each block to/from the user
 19218	 * space one at a time.  This is too slow, so we do something dirty here,
 19219	 * namely send the user space and virtual address to the file system in the
 19220	 * upper 10 bits of the file descriptor, and pass it the user virtual address
 19221	 * instead of a PM address.  The file system extracts these parameters when 
 19222	 * gets a read or write call from the process manager, which is the only 
 19223	 * process that is permitted to use this trick.  The file system then copies 
 19224	 * the whole segment directly to/from user space, bypassing PM completely.
 19225	 *
 19226	 * The byte count on read is usually smaller than the segment count, because
 19227	 * a segment is padded out to a click multiple, and the data segment is only
 19228	 * partially initialized.
 19229	 */
 19230	
 19231	  int new_fd, bytes, r;
 19232	  char *ubuf_ptr;
 19233	  struct mem_map *sp = &mproc[proc].mp_seg[seg];
 19234	  phys_bytes seg_bytes = seg_bytes0;
 19235	
 19236	  new_fd = (proc << 7) | (seg << 5) | fd;
 19237	  ubuf_ptr = (char *) ((vir_bytes) sp->mem_vir << CLICK_SHIFT);
 19238	
 19239	  while (seg_bytes != 0) {
 19240	#define PM_CHUNK_SIZE 8192
 19241	        bytes = MIN((INT_MAX / PM_CHUNK_SIZE) * PM_CHUNK_SIZE, seg_bytes);
 19242	        if (rw == 0) {
 19243	                r = read(new_fd, ubuf_ptr, bytes);
 19244	        } else {

_________________________ Page 895    File: servers/pm/exec.c _________________________
 19245	                r = write(new_fd, ubuf_ptr, bytes);
 19246	        }
 19247	        if (r != bytes) break;
 19248	        ubuf_ptr += bytes;
 19249	        seg_bytes -= bytes;
 19250	  }
 19251	}
	
 19253	/*===========================================================================*
 19254	 *                              find_share                                   *
 19255	 *===========================================================================*/
 19256	PUBLIC struct mproc *find_share(mp_ign, ino, dev, ctime)
 19257	struct mproc *mp_ign;           /* process that should not be looked at */
 19258	ino_t ino;                      /* parameters that uniquely identify a file */
 19259	dev_t dev;
 19260	time_t ctime;
 19261	{
 19262	/* Look for a process that is the file <ino, dev, ctime> in execution.  Don't
 19263	 * accidentally "find" mp_ign, because it is the process on whose behalf this
 19264	 * call is made.
 19265	 */
 19266	  struct mproc *sh_mp;
 19267	  for (sh_mp = &mproc[0]; sh_mp < &mproc[NR_PROCS]; sh_mp++) {
 19268	
 19269	        if (!(sh_mp->mp_flags & SEPARATE)) continue;
 19270	        if (sh_mp == mp_ign) continue;
 19271	        if (sh_mp->mp_ino != ino) continue;
 19272	        if (sh_mp->mp_dev != dev) continue;
 19273	        if (sh_mp->mp_ctime != ctime) continue;
 19274	        return sh_mp;
 19275	  }
 19276	  return(NULL);
 19277	}



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/break.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 19300	/* The MINIX model of memory allocation reserves a fixed amount of memory for
 19301	 * the combined text, data, and stack segments.  The amount used for a child
 19302	 * process created by FORK is the same as the parent had.  If the child does
 19303	 * an EXEC later, the new size is taken from the header of the file EXEC'ed.
 19304	 *
 19305	 * The layout in memory consists of the text segment, followed by the data
 19306	 * segment, followed by a gap (unused memory), followed by the stack segment.
 19307	 * The data segment grows upward and the stack grows downward, so each can
 19308	 * take memory from the gap.  If they meet, the process must be killed.  The
 19309	 * procedures in this file deal with the growth of the data and stack segments.
 19310	 *
 19311	 * The entry points into this file are:
 19312	 *   do_brk:      BRK/SBRK system calls to grow or shrink the data segment
 19313	 *   adjust:      see if a proposed segment adjustment is allowed
 19314	 *   size_ok:     see if the segment sizes are feasible
 19315	 */
 19316	
 19317	#include "pm.h"
 19318	#include <signal.h>
 19319	#include "mproc.h"

_________________________ Page 896    File: servers/pm/break.c _________________________
 19320	#include "param.h"
 19321	
 19322	#define DATA_CHANGED       1    /* flag value when data segment size changed */
 19323	#define STACK_CHANGED      2    /* flag value when stack size changed */
 19324	
 19325	/*===========================================================================*
 19326	 *                              do_brk                                       *
 19327	 *===========================================================================*/
 19328	PUBLIC int do_brk()
 19329	{
 19330	/* Perform the brk(addr) system call.
 19331	 *
 19332	 * The call is complicated by the fact that on some machines (e.g., 8088),
 19333	 * the stack pointer can grow beyond the base of the stack segment without
 19334	 * anybody noticing it.
 19335	 * The parameter, 'addr' is the new virtual address in D space.
 19336	 */
 19337	
 19338	  register struct mproc *rmp;
 19339	  int r;
 19340	  vir_bytes v, new_sp;
 19341	  vir_clicks new_clicks;
 19342	
 19343	  rmp = mp;
 19344	  v = (vir_bytes) m_in.addr;
 19345	  new_clicks = (vir_clicks) ( ((long) v + CLICK_SIZE - 1) >> CLICK_SHIFT);
 19346	  if (new_clicks < rmp->mp_seg[D].mem_vir) {
 19347	        rmp->mp_reply.reply_ptr = (char *) -1;
 19348	        return(ENOMEM);
 19349	  }
 19350	  new_clicks -= rmp->mp_seg[D].mem_vir;
 19351	  if ((r=get_stack_ptr(who, &new_sp)) != OK)    /* ask kernel for sp value */
 19352	        panic(__FILE__,"couldn't get stack pointer", r);
 19353	  r = adjust(rmp, new_clicks, new_sp);
 19354	  rmp->mp_reply.reply_ptr = (r == OK ? m_in.addr : (char *) -1);
 19355	  return(r);                    /* return new address or -1 */
 19356	}
	
 19358	/*===========================================================================*
 19359	 *                              adjust                                       *
 19360	 *===========================================================================*/
 19361	PUBLIC int adjust(rmp, data_clicks, sp)
 19362	register struct mproc *rmp;     /* whose memory is being adjusted? */
 19363	vir_clicks data_clicks;         /* how big is data segment to become? */
 19364	vir_bytes sp;                   /* new value of sp */
 19365	{
 19366	/* See if data and stack segments can coexist, adjusting them if need be.
 19367	 * Memory is never allocated or freed.  Instead it is added or removed from the
 19368	 * gap between data segment and stack segment.  If the gap size becomes
 19369	 * negative, the adjustment of data or stack fails and ENOMEM is returned.
 19370	 */
 19371	
 19372	  register struct mem_map *mem_sp, *mem_dp;
 19373	  vir_clicks sp_click, gap_base, lower, old_clicks;
 19374	  int changed, r, ft;
 19375	  long base_of_stack, delta;    /* longs avoid certain problems */
 19376	
 19377	  mem_dp = &rmp->mp_seg[D];     /* pointer to data segment map */
 19378	  mem_sp = &rmp->mp_seg[S];     /* pointer to stack segment map */
 19379	  changed = 0;                  /* set when either segment changed */

_________________________ Page 897    File: servers/pm/break.c _________________________
 19380	
 19381	  if (mem_sp->mem_len == 0) return(OK); /* don't bother init */
 19382	
 19383	  /* See if stack size has gone negative (i.e., sp too close to 0xFFFF...) */
 19384	  base_of_stack = (long) mem_sp->mem_vir + (long) mem_sp->mem_len;
 19385	  sp_click = sp >> CLICK_SHIFT; /* click containing sp */
 19386	  if (sp_click >= base_of_stack) return(ENOMEM);        /* sp too high */
 19387	
 19388	  /* Compute size of gap between stack and data segments. */
 19389	  delta = (long) mem_sp->mem_vir - (long) sp_click;
 19390	  lower = (delta > 0 ? sp_click : mem_sp->mem_vir);
 19391	
 19392	  /* Add a safety margin for future stack growth. Impossible to do right. */
 19393	#define SAFETY_BYTES  (384 * sizeof(char *))
 19394	#define SAFETY_CLICKS ((SAFETY_BYTES + CLICK_SIZE - 1) / CLICK_SIZE)
 19395	  gap_base = mem_dp->mem_vir + data_clicks + SAFETY_CLICKS;
 19396	  if (lower < gap_base) return(ENOMEM); /* data and stack collided */
 19397	
 19398	  /* Update data length (but not data orgin) on behalf of brk() system call. */
 19399	  old_clicks = mem_dp->mem_len;
 19400	  if (data_clicks != mem_dp->mem_len) {
 19401	        mem_dp->mem_len = data_clicks;
 19402	        changed |= DATA_CHANGED;
 19403	  }
 19404	
 19405	  /* Update stack length and origin due to change in stack pointer. */
 19406	  if (delta > 0) {
 19407	        mem_sp->mem_vir -= delta;
 19408	        mem_sp->mem_phys -= delta;
 19409	        mem_sp->mem_len += delta;
 19410	        changed |= STACK_CHANGED;
 19411	  }
 19412	
 19413	  /* Do the new data and stack segment sizes fit in the address space? */
 19414	  ft = (rmp->mp_flags & SEPARATE);
 19415	  r = (rmp->mp_seg[D].mem_vir + rmp->mp_seg[D].mem_len > 
 19416	          rmp->mp_seg[S].mem_vir) ? ENOMEM : OK;
 19417	  if (r == OK) {
 19418	        if (changed) sys_newmap((int)(rmp - mproc), rmp->mp_seg);
 19419	        return(OK);
 19420	  }
 19421	
 19422	  /* New sizes don't fit or require too many page/segment registers. Restore.*/
 19423	  if (changed & DATA_CHANGED) mem_dp->mem_len = old_clicks;
 19424	  if (changed & STACK_CHANGED) {
 19425	        mem_sp->mem_vir += delta;
 19426	        mem_sp->mem_phys += delta;
 19427	        mem_sp->mem_len -= delta;
 19428	  }
 19429	  return(ENOMEM);
 19430	}

_________________________ Page 898    File: servers/pm/break.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/signal.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 19500	/* This file handles signals, which are asynchronous events and are generally
 19501	 * a messy and unpleasant business.  Signals can be generated by the KILL
 19502	 * system call, or from the keyboard (SIGINT) or from the clock (SIGALRM).
 19503	 * In all cases control eventually passes to check_sig() to see which processes
 19504	 * can be signaled.  The actual signaling is done by sig_proc().
 19505	 *
 19506	 * The entry points into this file are:
 19507	 *   do_sigaction:   perform the SIGACTION system call
 19508	 *   do_sigpending:  perform the SIGPENDING system call
 19509	 *   do_sigprocmask: perform the SIGPROCMASK system call
 19510	 *   do_sigreturn:   perform the SIGRETURN system call
 19511	 *   do_sigsuspend:  perform the SIGSUSPEND system call
 19512	 *   do_kill:   perform the KILL system call
 19513	 *   do_alarm:  perform the ALARM system call by calling set_alarm()
 19514	 *   set_alarm: tell the clock task to start or stop a timer
 19515	 *   do_pause:  perform the PAUSE system call
 19516	 *   ksig_pending: the kernel notified about pending signals
 19517	 *   sig_proc:  interrupt or terminate a signaled process
 19518	 *   check_sig: check which processes to signal with sig_proc()
 19519	 *   check_pending:  check if a pending signal can now be delivered
 19520	 */
 19521	
 19522	#include "pm.h"
 19523	#include <sys/stat.h>
 19524	#include <sys/ptrace.h>
 19525	#include <minix/callnr.h>
 19526	#include <minix/com.h>
 19527	#include <signal.h>
 19528	#include <sys/sigcontext.h>
 19529	#include <string.h>
 19530	#include "mproc.h"
 19531	#include "param.h"
 19532	
 19533	#define CORE_MODE       0777    /* mode to use on core image files */
 19534	#define DUMPED          0200    /* bit set in status when core dumped */
 19535	
 19536	FORWARD _PROTOTYPE( void dump_core, (struct mproc *rmp)                 );
 19537	FORWARD _PROTOTYPE( void unpause, (int pro)                             );
 19538	FORWARD _PROTOTYPE( void handle_sig, (int proc_nr, sigset_t sig_map)    );
 19539	FORWARD _PROTOTYPE( void cause_sigalrm, (struct timer *tp)              );
 19540	
 19541	/*===========================================================================*
 19542	 *                              do_sigaction                                 *
 19543	 *===========================================================================*/
 19544	PUBLIC int do_sigaction()
 19545	{
 19546	  int r;
 19547	  struct sigaction svec;
 19548	  struct sigaction *svp;
 19549	
 19550	  if (m_in.sig_nr == SIGKILL) return(OK);
 19551	  if (m_in.sig_nr < 1 || m_in.sig_nr > _NSIG) return (EINVAL);
 19552	  svp = &mp->mp_sigact[m_in.sig_nr];
 19553	  if ((struct sigaction *) m_in.sig_osa != (struct sigaction *) NULL) {
 19554	        r = sys_datacopy(PM_PROC_NR,(vir_bytes) svp,

_________________________ Page 899    File: servers/pm/signal.c _________________________
 19555	                who, (vir_bytes) m_in.sig_osa, (phys_bytes) sizeof(svec));
 19556	        if (r != OK) return(r);
 19557	  }
 19558	
 19559	  if ((struct sigaction *) m_in.sig_nsa == (struct sigaction *) NULL) 
 19560	        return(OK);
 19561	
 19562	  /* Read in the sigaction structure. */
 19563	  r = sys_datacopy(who, (vir_bytes) m_in.sig_nsa,
 19564	                PM_PROC_NR, (vir_bytes) &svec, (phys_bytes) sizeof(svec));
 19565	  if (r != OK) return(r);
 19566	
 19567	  if (svec.sa_handler == SIG_IGN) {
 19568	        sigaddset(&mp->mp_ignore, m_in.sig_nr);
 19569	        sigdelset(&mp->mp_sigpending, m_in.sig_nr);
 19570	        sigdelset(&mp->mp_catch, m_in.sig_nr);
 19571	        sigdelset(&mp->mp_sig2mess, m_in.sig_nr);
 19572	  } else if (svec.sa_handler == SIG_DFL) {
 19573	        sigdelset(&mp->mp_ignore, m_in.sig_nr);
 19574	        sigdelset(&mp->mp_catch, m_in.sig_nr);
 19575	        sigdelset(&mp->mp_sig2mess, m_in.sig_nr);
 19576	  } else if (svec.sa_handler == SIG_MESS) {
 19577	        if (! (mp->mp_flags & PRIV_PROC)) return(EPERM);
 19578	        sigdelset(&mp->mp_ignore, m_in.sig_nr);
 19579	        sigaddset(&mp->mp_sig2mess, m_in.sig_nr);
 19580	        sigdelset(&mp->mp_catch, m_in.sig_nr);
 19581	  } else {
 19582	        sigdelset(&mp->mp_ignore, m_in.sig_nr);
 19583	        sigaddset(&mp->mp_catch, m_in.sig_nr);
 19584	        sigdelset(&mp->mp_sig2mess, m_in.sig_nr);
 19585	  }
 19586	  mp->mp_sigact[m_in.sig_nr].sa_handler = svec.sa_handler;
 19587	  sigdelset(&svec.sa_mask, SIGKILL);
 19588	  mp->mp_sigact[m_in.sig_nr].sa_mask = svec.sa_mask;
 19589	  mp->mp_sigact[m_in.sig_nr].sa_flags = svec.sa_flags;
 19590	  mp->mp_sigreturn = (vir_bytes) m_in.sig_ret;
 19591	  return(OK);
 19592	}
	
 19594	/*===========================================================================*
 19595	 *                              do_sigpending                                *
 19596	 *===========================================================================*/
 19597	PUBLIC int do_sigpending()
 19598	{
 19599	  mp->mp_reply.reply_mask = (long) mp->mp_sigpending;
 19600	  return OK;
 19601	}
	
 19603	/*===========================================================================*
 19604	 *                              do_sigprocmask                               *
 19605	 *===========================================================================*/
 19606	PUBLIC int do_sigprocmask()
 19607	{
 19608	/* Note that the library interface passes the actual mask in sigmask_set,
 19609	 * not a pointer to the mask, in order to save a copy.  Similarly,
 19610	 * the old mask is placed in the return message which the library
 19611	 * interface copies (if requested) to the user specified address.
 19612	 *
 19613	 * The library interface must set SIG_INQUIRE if the 'act' argument
 19614	 * is NULL.

_________________________ Page 900    File: servers/pm/signal.c _________________________
 19615	 */
 19616	
 19617	  int i;
 19618	
 19619	  mp->mp_reply.reply_mask = (long) mp->mp_sigmask;
 19620	
 19621	  switch (m_in.sig_how) {
 19622	      case SIG_BLOCK:
 19623	        sigdelset((sigset_t *)&m_in.sig_set, SIGKILL);
 19624	        for (i = 1; i <= _NSIG; i++) {
 19625	                if (sigismember((sigset_t *)&m_in.sig_set, i))
 19626	                        sigaddset(&mp->mp_sigmask, i);
 19627	        }
 19628	        break;
 19629	
 19630	      case SIG_UNBLOCK:
 19631	        for (i = 1; i <= _NSIG; i++) {
 19632	                if (sigismember((sigset_t *)&m_in.sig_set, i))
 19633	                        sigdelset(&mp->mp_sigmask, i);
 19634	        }
 19635	        check_pending(mp);
 19636	        break;
 19637	
 19638	      case SIG_SETMASK:
 19639	        sigdelset((sigset_t *) &m_in.sig_set, SIGKILL);
 19640	        mp->mp_sigmask = (sigset_t) m_in.sig_set;
 19641	        check_pending(mp);
 19642	        break;
 19643	
 19644	      case SIG_INQUIRE:
 19645	        break;
 19646	
 19647	      default:
 19648	        return(EINVAL);
 19649	        break;
 19650	  }
 19651	  return OK;
 19652	}
	
 19654	/*===========================================================================*
 19655	 *                              do_sigsuspend                                *
 19656	 *===========================================================================*/
 19657	PUBLIC int do_sigsuspend()
 19658	{
 19659	  mp->mp_sigmask2 = mp->mp_sigmask;     /* save the old mask */
 19660	  mp->mp_sigmask = (sigset_t) m_in.sig_set;
 19661	  sigdelset(&mp->mp_sigmask, SIGKILL);
 19662	  mp->mp_flags |= SIGSUSPENDED;
 19663	  check_pending(mp);
 19664	  return(SUSPEND);
 19665	}
	
 19667	/*===========================================================================*
 19668	 *                              do_sigreturn                                 *
 19669	 *===========================================================================*/
 19670	PUBLIC int do_sigreturn()
 19671	{
 19672	/* A user signal handler is done.  Restore context and check for
 19673	 * pending unblocked signals.
 19674	 */

_________________________ Page 901    File: servers/pm/signal.c _________________________
 19675	
 19676	  int r;
 19677	
 19678	  mp->mp_sigmask = (sigset_t) m_in.sig_set;
 19679	  sigdelset(&mp->mp_sigmask, SIGKILL);
 19680	
 19681	  r = sys_sigreturn(who, (struct sigmsg *) m_in.sig_context);
 19682	  check_pending(mp);
 19683	  return(r);
 19684	}
	
 19686	/*===========================================================================*
 19687	 *                              do_kill                                      *
 19688	 *===========================================================================*/
 19689	PUBLIC int do_kill()
 19690	{
 19691	/* Perform the kill(pid, signo) system call. */
 19692	
 19693	  return check_sig(m_in.pid, m_in.sig_nr);
 19694	}
	
 19696	/*===========================================================================*
 19697	 *                              ksig_pending                                 *
 19698	 *===========================================================================*/
 19699	PUBLIC int ksig_pending()
 19700	{
 19701	/* Certain signals, such as segmentation violations originate in the kernel.
 19702	 * When the kernel detects such signals, it notifies the PM to take further 
 19703	 * action. The PM requests the kernel to send messages with the process
 19704	 * slot and bit map for all signaled processes. The File System, for example,
 19705	 * uses this mechanism to signal writing on broken pipes (SIGPIPE). 
 19706	 *
 19707	 * The kernel has notified the PM about pending signals. Request pending
 19708	 * signals until all signals are handled. If there are no more signals,
 19709	 * NONE is returned in the process number field.
 19710	 */ 
 19711	 int proc_nr;
 19712	 sigset_t sig_map;
 19713	
 19714	 while (TRUE) {
 19715	   sys_getksig(&proc_nr, &sig_map);     /* get an arbitrary pending signal 
 19716	   if (NONE == proc_nr) {               /* stop if no more pending signals */
 19717	        break;
 19718	   } else {
 19719	        handle_sig(proc_nr, sig_map);   /* handle the received signal */
 19720	        sys_endksig(proc_nr);           /* tell kernel it's done */
 19721	   }
 19722	 } 
 19723	 return(SUSPEND);                       /* prevents sending reply */
 19724	}
	
 19726	/*===========================================================================*
 19727	 *                              handle_sig                                   *
 19728	 *===========================================================================*/
 19729	PRIVATE void handle_sig(proc_nr, sig_map)
 19730	int proc_nr;
 19731	sigset_t sig_map;
 19732	{
 19733	  register struct mproc *rmp;
 19734	  int i;

_________________________ Page 902    File: servers/pm/signal.c _________________________
 19735	  pid_t proc_id, id;
 19736	
 19737	  rmp = &mproc[proc_nr];
 19738	  if ((rmp->mp_flags & (IN_USE | ZOMBIE)) != IN_USE) return;
 19739	  proc_id = rmp->mp_pid;
 19740	  mp = &mproc[0];                       /* pretend signals are from PM */
 19741	  mp->mp_procgrp = rmp->mp_procgrp;     /* get process group right */
 19742	
 19743	  /* Check each bit in turn to see if a signal is to be sent.  Unlike
 19744	   * kill(), the kernel may collect several unrelated signals for a
 19745	   * process and pass them to PM in one blow.  Thus loop on the bit
 19746	   * map. For SIGINT and SIGQUIT, use proc_id 0 to indicate a broadcast
 19747	   * to the recipient's process group.  For SIGKILL, use proc_id -1 to
 19748	   * indicate a systemwide broadcast.
 19749	   */
 19750	  for (i = 1; i <= _NSIG; i++) {
 19751	        if (!sigismember(&sig_map, i)) continue;
 19752	        switch (i) {
 19753	            case SIGINT:
 19754	            case SIGQUIT:
 19755	                id = 0; break;  /* broadcast to process group */
 19756	            case SIGKILL:
 19757	                id = -1; break; /* broadcast to all except INIT */
 19758	            default:
 19759	                id = proc_id;
 19760	                break;
 19761	        }
 19762	        check_sig(id, i);
 19763	  }
 19764	}
	
 19766	/*===========================================================================*
 19767	 *                              do_alarm                                     *
 19768	 *===========================================================================*/
 19769	PUBLIC int do_alarm()
 19770	{
 19771	/* Perform the alarm(seconds) system call. */
 19772	  return(set_alarm(who, m_in.seconds));
 19773	}
	
 19775	/*===========================================================================*
 19776	 *                              set_alarm                                    *
 19777	 *===========================================================================*/
 19778	PUBLIC int set_alarm(proc_nr, sec)
 19779	int proc_nr;                    /* process that wants the alarm */
 19780	int sec;                        /* how many seconds delay before the signal */
 19781	{
 19782	/* This routine is used by do_alarm() to set the alarm timer.  It is also used
 19783	 * to turn the timer off when a process exits with the timer still on.
 19784	 */
 19785	  clock_t ticks;        /* number of ticks for alarm */
 19786	  clock_t exptime;      /* needed for remaining time on previous alarm */
 19787	  clock_t uptime;       /* current system time */
 19788	  int remaining;        /* previous time left in seconds */
 19789	  int s;
 19790	
 19791	  /* First determine remaining time of previous alarm, if set. */
 19792	  if (mproc[proc_nr].mp_flags & ALARM_ON) {
 19793	        if ( (s=getuptime(&uptime)) != OK) 
 19794	                panic(__FILE__,"set_alarm couldn't get uptime", s);

_________________________ Page 903    File: servers/pm/signal.c _________________________
 19795	        exptime = *tmr_exp_time(&mproc[proc_nr].mp_timer);
 19796	        remaining = (int) ((exptime - uptime + (HZ-1))/HZ);
 19797	        if (remaining < 0) remaining = 0;       
 19798	  } else {
 19799	        remaining = 0; 
 19800	  }
 19801	
 19802	  /* Tell the clock task to provide a signal message when the time comes.
 19803	   *
 19804	   * Large delays cause a lot of problems.  First, the alarm system call
 19805	   * takes an unsigned seconds count and the library has cast it to an int.
 19806	   * That probably works, but on return the library will convert "negative"
 19807	   * unsigneds to errors.  Presumably no one checks for these errors, so
 19808	   * force this call through.  Second, If unsigned and long have the same
 19809	   * size, converting from seconds to ticks can easily overflow.  Finally,
 19810	   * the kernel has similar overflow bugs adding ticks.
 19811	   *
 19812	   * Fixing this requires a lot of ugly casts to fit the wrong interface
 19813	   * types and to avoid overflow traps.  ALRM_EXP_TIME has the right type
 19814	   * (clock_t) although it is declared as long.  How can variables like
 19815	   * this be declared properly without combinatorial explosion of message
 19816	   * types?
 19817	   */
 19818	  ticks = (clock_t) (HZ * (unsigned long) (unsigned) sec);
 19819	  if ( (unsigned long) ticks / HZ != (unsigned) sec)
 19820	        ticks = LONG_MAX;       /* eternity (really TMR_NEVER) */
 19821	
 19822	  if (ticks != 0) {
 19823	        pm_set_timer(&mproc[proc_nr].mp_timer, ticks, cause_sigalrm, proc_nr);
 19824	        mproc[proc_nr].mp_flags |=  ALARM_ON;
 19825	  } else if (mproc[proc_nr].mp_flags & ALARM_ON) {
 19826	        pm_cancel_timer(&mproc[proc_nr].mp_timer);
 19827	        mproc[proc_nr].mp_flags &= ~ALARM_ON;
 19828	  }
 19829	  return(remaining);
 19830	}
	
 19832	/*===========================================================================*
 19833	 *                              cause_sigalrm                                *
 19834	 *===========================================================================*/
 19835	PRIVATE void cause_sigalrm(tp)
 19836	struct timer *tp;
 19837	{
 19838	  int proc_nr;
 19839	  register struct mproc *rmp;
 19840	
 19841	  proc_nr = tmr_arg(tp)->ta_int;        /* get process from timer */
 19842	  rmp = &mproc[proc_nr];
 19843	
 19844	  if ((rmp->mp_flags & (IN_USE | ZOMBIE)) != IN_USE) return;
 19845	  if ((rmp->mp_flags & ALARM_ON) == 0) return;
 19846	  rmp->mp_flags &= ~ALARM_ON;
 19847	  check_sig(rmp->mp_pid, SIGALRM);
 19848	}
	
 19850	/*===========================================================================*
 19851	 *                              do_pause                                     *
 19852	 *===========================================================================*/
 19853	PUBLIC int do_pause()
 19854	{

_________________________ Page 904    File: servers/pm/signal.c _________________________
 19855	/* Perform the pause() system call. */
 19856	
 19857	  mp->mp_flags |= PAUSED;
 19858	  return(SUSPEND);
 19859	}
	
 19861	/*===========================================================================*
 19862	 *                              sig_proc                                     *
 19863	 *===========================================================================*/
 19864	PUBLIC void sig_proc(rmp, signo)
 19865	register struct mproc *rmp;     /* pointer to the process to be signaled */
 19866	int signo;                      /* signal to send to process (1 to _NSIG) */
 19867	{
 19868	/* Send a signal to a process.  Check to see if the signal is to be caught,
 19869	 * ignored, tranformed into a message (for system processes) or blocked.  
 19870	 *  - If the signal is to be transformed into a message, request the KERNEL to
 19871	 * send the target process a system notification with the pending signal as an 
 19872	 * argument. 
 19873	 *  - If the signal is to be caught, request the KERNEL to push a sigcontext 
 19874	 * structure and a sigframe structure onto the catcher's stack.  Also, KERNEL 
 19875	 * will reset the program counter and stack pointer, so that when the process 
 19876	 * next runs, it will be executing the signal handler. When the signal handler 
 19877	 * returns,  sigreturn(2) will be called.  Then KERNEL will restore the signal 
 19878	 * context from the sigcontext structure.
 19879	 * If there is insufficient stack space, kill the process.
 19880	 */
 19881	
 19882	  vir_bytes new_sp;
 19883	  int s;
 19884	  int slot;
 19885	  int sigflags;
 19886	  struct sigmsg sm;
 19887	
 19888	  slot = (int) (rmp - mproc);
 19889	  if ((rmp->mp_flags & (IN_USE | ZOMBIE)) != IN_USE) {
 19890	        printf("PM: signal %d sent to %s process %d\n",
 19891	                signo, (rmp->mp_flags & ZOMBIE) ? "zombie" : "dead", slot);
 19892	        panic(__FILE__,"", NO_NUM);
 19893	  }
 19894	  if ((rmp->mp_flags & TRACED) && signo != SIGKILL) {
 19895	        /* A traced process has special handling. */
 19896	        unpause(slot);
 19897	        stop_proc(rmp, signo);  /* a signal causes it to stop */
 19898	        return;
 19899	  }
 19900	  /* Some signals are ignored by default. */
 19901	  if (sigismember(&rmp->mp_ignore, signo)) { 
 19902	        return;
 19903	  }
 19904	  if (sigismember(&rmp->mp_sigmask, signo)) {
 19905	        /* Signal should be blocked. */
 19906	        sigaddset(&rmp->mp_sigpending, signo);
 19907	        return;
 19908	  }
 19909	  sigflags = rmp->mp_sigact[signo].sa_flags;
 19910	  if (sigismember(&rmp->mp_catch, signo)) {
 19911	        if (rmp->mp_flags & SIGSUSPENDED)
 19912	                sm.sm_mask = rmp->mp_sigmask2;
 19913	        else
 19914	                sm.sm_mask = rmp->mp_sigmask;

_________________________ Page 905    File: servers/pm/signal.c _________________________
 19915	        sm.sm_signo = signo;
 19916	        sm.sm_sighandler = (vir_bytes) rmp->mp_sigact[signo].sa_handler;
 19917	        sm.sm_sigreturn = rmp->mp_sigreturn;
 19918	        if ((s=get_stack_ptr(slot, &new_sp)) != OK)
 19919	                panic(__FILE__,"couldn't get new stack pointer",s);
 19920	        sm.sm_stkptr = new_sp;
 19921	
 19922	        /* Make room for the sigcontext and sigframe struct. */
 19923	        new_sp -= sizeof(struct sigcontext)
 19924	                                 + 3 * sizeof(char *) + 2 * sizeof(int);
 19925	
 19926	        if (adjust(rmp, rmp->mp_seg[D].mem_len, new_sp) != OK)
 19927	                goto doterminate;
 19928	
 19929	        rmp->mp_sigmask |= rmp->mp_sigact[signo].sa_mask;
 19930	        if (sigflags & SA_NODEFER)
 19931	                sigdelset(&rmp->mp_sigmask, signo);
 19932	        else
 19933	                sigaddset(&rmp->mp_sigmask, signo);
 19934	
 19935	        if (sigflags & SA_RESETHAND) {
 19936	                sigdelset(&rmp->mp_catch, signo);
 19937	                rmp->mp_sigact[signo].sa_handler = SIG_DFL;
 19938	        }
 19939	
 19940	        if (OK == (s=sys_sigsend(slot, &sm))) {
 19941	
 19942	                sigdelset(&rmp->mp_sigpending, signo);
 19943	                /* If process is hanging on PAUSE, WAIT, SIGSUSPEND, tty, 
 19944	                 * pipe, etc., release it.
 19945	                 */
 19946	                unpause(slot);
 19947	                return;
 19948	        }
 19949	        panic(__FILE__, "warning, sys_sigsend failed", s);
 19950	  }
 19951	  else if (sigismember(&rmp->mp_sig2mess, signo)) {
 19952	        if (OK != (s=sys_kill(slot,signo)))
 19953	                panic(__FILE__, "warning, sys_kill failed", s);
 19954	        return;
 19955	  }
 19956	
 19957	doterminate:
 19958	  /* Signal should not or cannot be caught.  Take default action. */
 19959	  if (sigismember(&ign_sset, signo)) return;
 19960	
 19961	  rmp->mp_sigstatus = (char) signo;
 19962	  if (sigismember(&core_sset, signo)) {
 19963	        /* Switch to the user's FS environment and dump core. */
 19964	        tell_fs(CHDIR, slot, FALSE, 0);
 19965	        dump_core(rmp);
 19966	  }
 19967	  pm_exit(rmp, 0);              /* terminate process */
 19968	}
	
 19970	/*===========================================================================*
 19971	 *                              check_sig                                    *
 19972	 *===========================================================================*/
 19973	PUBLIC int check_sig(proc_id, signo)
 19974	pid_t proc_id;                  /* pid of proc to sig, or 0 or -1, or -pgrp */

_________________________ Page 906    File: servers/pm/signal.c _________________________
 19975	int signo;                      /* signal to send to process (0 to _NSIG) */
 19976	{
 19977	/* Check to see if it is possible to send a signal.  The signal may have to be
 19978	 * sent to a group of processes.  This routine is invoked by the KILL system
 19979	 * call, and also when the kernel catches a DEL or other signal.
 19980	 */
 19981	
 19982	  register struct mproc *rmp;
 19983	  int count;                    /* count # of signals sent */
 19984	  int error_code;
 19985	
 19986	  if (signo < 0 || signo > _NSIG) return(EINVAL);
 19987	
 19988	  /* Return EINVAL for attempts to send SIGKILL to INIT alone. */
 19989	  if (proc_id == INIT_PID && signo == SIGKILL) return(EINVAL);
 19990	
 19991	  /* Search the proc table for processes to signal.  (See forkexit.c about
 19992	   * pid magic.)
 19993	   */
 19994	  count = 0;
 19995	  error_code = ESRCH;
 19996	  for (rmp = &mproc[0]; rmp < &mproc[NR_PROCS]; rmp++) {
 19997	        if (!(rmp->mp_flags & IN_USE)) continue;
 19998	        if ((rmp->mp_flags & ZOMBIE) && signo != 0) continue;
 19999	
 20000	        /* Check for selection. */
 20001	        if (proc_id > 0 && proc_id != rmp->mp_pid) continue;
 20002	        if (proc_id == 0 && mp->mp_procgrp != rmp->mp_procgrp) continue;
 20003	        if (proc_id == -1 && rmp->mp_pid <= INIT_PID) continue;
 20004	        if (proc_id < -1 && rmp->mp_procgrp != -proc_id) continue;
 20005	
 20006	        /* Check for permission. */
 20007	        if (mp->mp_effuid != SUPER_USER
 20008	            && mp->mp_realuid != rmp->mp_realuid
 20009	            && mp->mp_effuid != rmp->mp_realuid
 20010	            && mp->mp_realuid != rmp->mp_effuid
 20011	            && mp->mp_effuid != rmp->mp_effuid) {
 20012	                error_code = EPERM;
 20013	                continue;
 20014	        }
 20015	
 20016	        count++;
 20017	        if (signo == 0) continue;
 20018	
 20019	        /* 'sig_proc' will handle the disposition of the signal.  The
 20020	         * signal may be caught, blocked, ignored, or cause process
 20021	         * termination, possibly with core dump.
 20022	         */
 20023	        sig_proc(rmp, signo);
 20024	
 20025	        if (proc_id > 0) break; /* only one process being signaled */
 20026	  }
 20027	
 20028	  /* If the calling process has killed itself, don't reply. */
 20029	  if ((mp->mp_flags & (IN_USE | ZOMBIE)) != IN_USE) return(SUSPEND);
 20030	  return(count > 0 ? OK : error_code);
 20031	}
	

_________________________ Page 907    File: servers/pm/signal.c _________________________
 20033	/*===========================================================================*
 20034	 *                              check_pending                                *
 20035	 *===========================================================================*/
 20036	PUBLIC void check_pending(rmp)
 20037	register struct mproc *rmp;
 20038	{
 20039	  /* Check to see if any pending signals have been unblocked.  The
 20040	   * first such signal found is delivered.
 20041	   *
 20042	   * If multiple pending unmasked signals are found, they will be
 20043	   * delivered sequentially.
 20044	   *
 20045	   * There are several places in this file where the signal mask is
 20046	   * changed.  At each such place, check_pending() should be called to
 20047	   * check for newly unblocked signals.
 20048	   */
 20049	
 20050	  int i;
 20051	
 20052	  for (i = 1; i <= _NSIG; i++) {
 20053	        if (sigismember(&rmp->mp_sigpending, i) &&
 20054	                !sigismember(&rmp->mp_sigmask, i)) {
 20055	                sigdelset(&rmp->mp_sigpending, i);
 20056	                sig_proc(rmp, i);
 20057	                break;
 20058	        }
 20059	  }
 20060	}
	
 20062	/*===========================================================================*
 20063	 *                              unpause                                      *
 20064	 *===========================================================================*/
 20065	PRIVATE void unpause(pro)
 20066	int pro;                        /* which process number */
 20067	{
 20068	/* A signal is to be sent to a process.  If that process is hanging on a
 20069	 * system call, the system call must be terminated with EINTR.  Possible
 20070	 * calls are PAUSE, WAIT, READ and WRITE, the latter two for pipes and ttys.
 20071	 * First check if the process is hanging on an PM call.  If not, tell FS,
 20072	 * so it can check for READs and WRITEs from pipes, ttys and the like.
 20073	 */
 20074	
 20075	  register struct mproc *rmp;
 20076	
 20077	  rmp = &mproc[pro];
 20078	
 20079	  /* Check to see if process is hanging on a PAUSE, WAIT or SIGSUSPEND call. */
 20080	  if (rmp->mp_flags & (PAUSED | WAITING | SIGSUSPENDED)) {
 20081	        rmp->mp_flags &= ~(PAUSED | WAITING | SIGSUSPENDED);
 20082	        setreply(pro, EINTR);
 20083	        return;
 20084	  }
 20085	
 20086	  /* Process is not hanging on an PM call.  Ask FS to take a look. */
 20087	  tell_fs(UNPAUSE, pro, 0, 0);
 20088	}
	

_________________________ Page 908    File: servers/pm/signal.c _________________________
 20090	/*===========================================================================*
 20091	 *                              dump_core                                    *
 20092	 *===========================================================================*/
 20093	PRIVATE void dump_core(rmp)
 20094	register struct mproc *rmp;     /* whose core is to be dumped */
 20095	{
 20096	/* Make a core dump on the file "core", if possible. */
 20097	
 20098	  int s, fd, seg, slot;
 20099	  vir_bytes current_sp;
 20100	  long trace_data, trace_off;
 20101	
 20102	  slot = (int) (rmp - mproc);
 20103	
 20104	  /* Can core file be written?  We are operating in the user's FS environment,
 20105	   * so no special permission checks are needed.
 20106	   */
 20107	  if (rmp->mp_realuid != rmp->mp_effuid) return;
 20108	  if ( (fd = open(core_name, O_WRONLY | O_CREAT | O_TRUNC | O_NONBLOCK,
 20109	                                                CORE_MODE)) < 0) return;
 20110	  rmp->mp_sigstatus |= DUMPED;
 20111	
 20112	  /* Make sure the stack segment is up to date.
 20113	   * We don't want adjust() to fail unless current_sp is preposterous,
 20114	   * but it might fail due to safety checking.  Also, we don't really want 
 20115	   * the adjust() for sending a signal to fail due to safety checking.  
 20116	   * Maybe make SAFETY_BYTES a parameter.
 20117	   */
 20118	  if ((s=get_stack_ptr(slot, &current_sp)) != OK)
 20119	        panic(__FILE__,"couldn't get new stack pointer",s);
 20120	  adjust(rmp, rmp->mp_seg[D].mem_len, current_sp);
 20121	
 20122	  /* Write the memory map of all segments to begin the core file. */
 20123	  if (write(fd, (char *) rmp->mp_seg, (unsigned) sizeof rmp->mp_seg)
 20124	      != (unsigned) sizeof rmp->mp_seg) {
 20125	        close(fd);
 20126	        return;
 20127	  }
 20128	
 20129	  /* Write out the whole kernel process table entry to get the regs. */
 20130	  trace_off = 0;
 20131	  while (sys_trace(T_GETUSER, slot, trace_off, &trace_data) == OK) {
 20132	        if (write(fd, (char *) &trace_data, (unsigned) sizeof (long))
 20133	            != (unsigned) sizeof (long)) {
 20134	                close(fd);
 20135	                return;
 20136	        }
 20137	        trace_off += sizeof (long);
 20138	  }
 20139	
 20140	  /* Loop through segments and write the segments themselves out. */
 20141	  for (seg = 0; seg < NR_LOCAL_SEGS; seg++) {
 20142	        rw_seg(1, fd, slot, seg,
 20143	                (phys_bytes) rmp->mp_seg[seg].mem_len << CLICK_SHIFT);
 20144	  }
 20145	  close(fd);
 20146	}
	

_________________________ Page 909    File: servers/pm/signal.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/timers.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 20200	/* PM watchdog timer management. These functions in this file provide
 20201	 * a convenient interface to the timers library that manages a list of
 20202	 * watchdog timers. All details of scheduling an alarm at the CLOCK task 
 20203	 * are hidden behind this interface.
 20204	 * Only system processes are allowed to set an alarm timer at the kernel. 
 20205	 * Therefore, the PM maintains a local list of timers for user processes
 20206	 * that requested an alarm signal. 
 20207	 * 
 20208	 * The entry points into this file are:
 20209	 *   pm_set_timer:      reset and existing or set a new watchdog timer
 20210	 *   pm_expire_timers:  check for expired timers and run watchdog functions
 20211	 *   pm_cancel_timer:   remove a time from the list of timers
 20212	 *
 20213	 */
 20214	
 20215	#include "pm.h"
 20216	
 20217	#include <timers.h>
 20218	#include <minix/syslib.h>
 20219	#include <minix/com.h>
 20220	
 20221	PRIVATE timer_t *pm_timers = NULL;
 20222	
 20223	/*===========================================================================*
 20224	 *                              pm_set_timer                                 *
 20225	 *===========================================================================*/
 20226	PUBLIC void pm_set_timer(timer_t *tp, int ticks, tmr_func_t watchdog, int arg)
 20227	{
 20228	        int r;
 20229	        clock_t now, prev_time = 0, next_time;
 20230	
 20231	        if ((r = getuptime(&now)) != OK)
 20232	                panic(__FILE__, "PM couldn't get uptime", NO_NUM);
 20233	
 20234	        /* Set timer argument and add timer to the list. */
 20235	        tmr_arg(tp)->ta_int = arg;
 20236	        prev_time = tmrs_settimer(&pm_timers,tp,now+ticks,watchdog,&next_ti
 20237	
 20238	        /* Reschedule our synchronous alarm if necessary. */
 20239	        if (! prev_time || prev_time > next_time) {
 20240	                if (sys_setalarm(next_time, 1) != OK)
 20241	                        panic(__FILE__, "PM set timer couldn't set alarm.", NO_NUM)
 20242	        }
 20243	
 20244	        return;
 20245	}
	
 20247	/*===========================================================================*
 20248	 *                              pm_expire_timers                             *
 20249	 *===========================================================================*/
 20250	PUBLIC void pm_expire_timers(clock_t now)
 20251	{
 20252	        clock_t next_time;
 20253	
 20254	        /* Check for expired timers and possibly reschedule an alarm. */

_________________________ Page 910    File: servers/pm/timers.c _________________________
 20255	        tmrs_exptimers(&pm_timers, now, &next_time);
 20256	        if (next_time > 0) {
 20257	                if (sys_setalarm(next_time, 1) != OK)
 20258	                        panic(__FILE__, "PM expire timer couldn't set alarm.", NO_N
 20259	        }
 20260	}
	
 20262	/*===========================================================================*
 20263	 *                              pm_cancel_timer                              *
 20264	 *===========================================================================*/
 20265	PUBLIC void pm_cancel_timer(timer_t *tp)
 20266	{
 20267	        clock_t next_time, prev_time;
 20268	        prev_time = tmrs_clrtimer(&pm_timers, tp, &next_time);
 20269	
 20270	        /* If the earliest timer has been removed, we have to set the alarm to  
 20271	     * the next timer, or cancel the alarm altogether if the last timer has 
 20272	     * been cancelled (next_time will be 0 then).
 20273	         */
 20274	        if (prev_time < next_time || ! next_time) {
 20275	                if (sys_setalarm(next_time, 1) != OK)
 20276	                        panic(__FILE__, "PM expire timer couldn't set alarm.", NO_N
 20277	        }
 20278	}


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/time.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 20300	/* This file takes care of those system calls that deal with time.
 20301	 *
 20302	 * The entry points into this file are
 20303	 *   do_time:           perform the TIME system call
 20304	 *   do_stime:          perform the STIME system call
 20305	 *   do_times:          perform the TIMES system call
 20306	 */
 20307	
 20308	#include "pm.h"
 20309	#include <minix/callnr.h>
 20310	#include <minix/com.h>
 20311	#include <signal.h>
 20312	#include "mproc.h"
 20313	#include "param.h"
 20314	
 20315	PRIVATE time_t boottime;
 20316	
 20317	/*===========================================================================*
 20318	 *                              do_time                                      *
 20319	 *===========================================================================*/
 20320	PUBLIC int do_time()
 20321	{
 20322	/* Perform the time(tp) system call. This returns the time in seconds since 
 20323	 * 1.1.1970.  MINIX is an astrophysically naive system that assumes the earth 
 20324	 * rotates at a constant rate and that such things as leap seconds do not 
 20325	 * exist.
 20326	 */
 20327	  clock_t uptime;
 20328	  int s;
 20329	

_________________________ Page 911    File: servers/pm/time.c _________________________
 20330	  if ( (s=getuptime(&uptime)) != OK) 
 20331	        panic(__FILE__,"do_time couldn't get uptime", s);
 20332	
 20333	  mp->mp_reply.reply_time = (time_t) (boottime + (uptime/HZ));
 20334	  mp->mp_reply.reply_utime = (uptime%HZ)*1000000/HZ;
 20335	  return(OK);
 20336	}
	
 20338	/*===========================================================================*
 20339	 *                              do_stime                                     *
 20340	 *===========================================================================*/
 20341	PUBLIC int do_stime()
 20342	{
 20343	/* Perform the stime(tp) system call. Retrieve the system's uptime (ticks 
 20344	 * since boot) and store the time in seconds at system boot in the global
 20345	 * variable 'boottime'.
 20346	 */
 20347	  clock_t uptime;
 20348	  int s;
 20349	
 20350	  if (mp->mp_effuid != SUPER_USER) { 
 20351	      return(EPERM);
 20352	  }
 20353	  if ( (s=getuptime(&uptime)) != OK) 
 20354	      panic(__FILE__,"do_stime couldn't get uptime", s);
 20355	  boottime = (long) m_in.stime - (uptime/HZ);
 20356	
 20357	  /* Also inform FS about the new system time. */
 20358	  tell_fs(STIME, boottime, 0, 0);
 20359	
 20360	  return(OK);
 20361	}
	
 20363	/*===========================================================================*
 20364	 *                              do_times                                     *
 20365	 *===========================================================================*/
 20366	PUBLIC int do_times()
 20367	{
 20368	/* Perform the times(buffer) system call. */
 20369	  register struct mproc *rmp = mp;
 20370	  clock_t t[5];
 20371	  int s;
 20372	
 20373	  if (OK != (s=sys_times(who, t)))
 20374	      panic(__FILE__,"do_times couldn't get times", s);
 20375	  rmp->mp_reply.reply_t1 = t[0];                /* user time */
 20376	  rmp->mp_reply.reply_t2 = t[1];                /* system time */
 20377	  rmp->mp_reply.reply_t3 = rmp->mp_child_utime; /* child user time */
 20378	  rmp->mp_reply.reply_t4 = rmp->mp_child_stime; /* child system time */
 20379	  rmp->mp_reply.reply_t5 = t[4];                /* uptime since boot */
 20380	
 20381	  return(OK);
 20382	}
	

_________________________ Page 912    File: servers/pm/time.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/getset.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 20400	/* This file handles the 4 system calls that get and set uids and gids.
 20401	 * It also handles getpid(), setsid(), and getpgrp().  The code for each
 20402	 * one is so tiny that it hardly seemed worthwhile to make each a separate
 20403	 * function.
 20404	 */
 20405	
 20406	#include "pm.h"
 20407	#include <minix/callnr.h>
 20408	#include <signal.h>
 20409	#include "mproc.h"
 20410	#include "param.h"
 20411	
 20412	/*===========================================================================*
 20413	 *                              do_getset                                    *
 20414	 *===========================================================================*/
 20415	PUBLIC int do_getset()
 20416	{
 20417	/* Handle GETUID, GETGID, GETPID, GETPGRP, SETUID, SETGID, SETSID.  The four
 20418	 * GETs and SETSID return their primary results in 'r'.  GETUID, GETGID, and
 20419	 * GETPID also return secondary results (the effective IDs, or the parent
 20420	 * process ID) in 'reply_res2', which is returned to the user.
 20421	 */
 20422	
 20423	  register struct mproc *rmp = mp;
 20424	  register int r;
 20425	
 20426	  switch(call_nr) {
 20427	        case GETUID:
 20428	                r = rmp->mp_realuid;
 20429	                rmp->mp_reply.reply_res2 = rmp->mp_effuid;
 20430	                break;
 20431	
 20432	        case GETGID:
 20433	                r = rmp->mp_realgid;
 20434	                rmp->mp_reply.reply_res2 = rmp->mp_effgid;
 20435	                break;
 20436	
 20437	        case GETPID:
 20438	                r = mproc[who].mp_pid;
 20439	                rmp->mp_reply.reply_res2 = mproc[rmp->mp_parent].mp_pid;
 20440	                break;
 20441	
 20442	        case SETUID:
 20443	                if (rmp->mp_realuid != (uid_t) m_in.usr_id && 
 20444	                                rmp->mp_effuid != SUPER_USER)
 20445	                        return(EPERM);
 20446	                rmp->mp_realuid = (uid_t) m_in.usr_id;
 20447	                rmp->mp_effuid = (uid_t) m_in.usr_id;
 20448	                tell_fs(SETUID, who, rmp->mp_realuid, rmp->mp_effuid);
 20449	                r = OK;
 20450	                break;
 20451	
 20452	        case SETGID:
 20453	                if (rmp->mp_realgid != (gid_t) m_in.grp_id && 
 20454	                                rmp->mp_effuid != SUPER_USER)

_________________________ Page 913    File: servers/pm/getset.c _________________________
 20455	                        return(EPERM);
 20456	                rmp->mp_realgid = (gid_t) m_in.grp_id;
 20457	                rmp->mp_effgid = (gid_t) m_in.grp_id;
 20458	                tell_fs(SETGID, who, rmp->mp_realgid, rmp->mp_effgid);
 20459	                r = OK;
 20460	                break;
 20461	
 20462	        case SETSID:
 20463	                if (rmp->mp_procgrp == rmp->mp_pid) return(EPERM);
 20464	                rmp->mp_procgrp = rmp->mp_pid;
 20465	                tell_fs(SETSID, who, 0, 0);
 20466	                /* fall through */
 20467	
 20468	        case GETPGRP:
 20469	                r = rmp->mp_procgrp;
 20470	                break;
 20471	
 20472	        default:
 20473	                r = EINVAL;
 20474	                break;  
 20475	  }
 20476	  return(r);
 20477	}



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/pm/misc.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 20500	/* Miscellaneous system calls.                          Author: Kees J. Bot
 20501	 *                                                              31 Mar 2000
 20502	 * The entry points into this file are:
 20503	 *   do_reboot: kill all processes, then reboot system
 20504	 *   do_svrctl: process manager control
 20505	 *   do_getsysinfo: request copy of PM data structure  (Jorrit N. Herder)
 20506	 *   do_getprocnr: lookup process slot number  (Jorrit N. Herder)
 20507	 *   do_memalloc: allocate a chunk of memory  (Jorrit N. Herder)
 20508	 *   do_memfree: deallocate a chunk of memory  (Jorrit N. Herder)
 20509	 *   do_getsetpriority: get/set process priority
 20510	 */
 20511	
 20512	#include "pm.h"
 20513	#include <minix/callnr.h>
 20514	#include <signal.h>
 20515	#include <sys/svrctl.h>
 20516	#include <sys/resource.h>
 20517	#include <minix/com.h>
 20518	#include <string.h>
 20519	#include "mproc.h"
 20520	#include "param.h"
 20521	
 20522	/*===========================================================================*
 20523	 *                              do_allocmem                                  *
 20524	 *===========================================================================*/
 20525	PUBLIC int do_allocmem()
 20526	{
 20527	  vir_clicks mem_clicks;
 20528	  phys_clicks mem_base;
 20529	

_________________________ Page 914    File: servers/pm/misc.c _________________________
 20530	  mem_clicks = (m_in.memsize + CLICK_SIZE -1 ) >> CLICK_SHIFT;
 20531	  mem_base = alloc_mem(mem_clicks);
 20532	  if (mem_base == NO_MEM) return(ENOMEM);
 20533	  mp->mp_reply.membase =  (phys_bytes) (mem_base << CLICK_SHIFT);
 20534	  return(OK);
 20535	}
	
 20537	/*===========================================================================*
 20538	 *                              do_freemem                                   *
 20539	 *===========================================================================*/
 20540	PUBLIC int do_freemem()
 20541	{
 20542	  vir_clicks mem_clicks;
 20543	  phys_clicks mem_base;
 20544	
 20545	  mem_clicks = (m_in.memsize + CLICK_SIZE -1 ) >> CLICK_SHIFT;
 20546	  mem_base = (m_in.membase + CLICK_SIZE -1 ) >> CLICK_SHIFT;
 20547	  free_mem(mem_base, mem_clicks);
 20548	  return(OK);
 20549	}
	
 20551	/*===========================================================================*
 20552	 *                              do_getsysinfo                                *
 20553	 *===========================================================================*/
 20554	PUBLIC int do_getsysinfo()
 20555	{
 20556	  struct mproc *proc_addr;
 20557	  vir_bytes src_addr, dst_addr;
 20558	  struct kinfo kinfo;
 20559	  size_t len;
 20560	  int s;
 20561	
 20562	  switch(m_in.info_what) {
 20563	  case SI_KINFO:                        /* kernel info is obtained via PM */
 20564	        sys_getkinfo(&kinfo);
 20565	        src_addr = (vir_bytes) &kinfo;
 20566	        len = sizeof(struct kinfo);
 20567	        break;
 20568	  case SI_PROC_ADDR:                    /* get address of PM process table */
 20569	        proc_addr = &mproc[0];
 20570	        src_addr = (vir_bytes) &proc_addr;
 20571	        len = sizeof(struct mproc *);
 20572	        break; 
 20573	  case SI_PROC_TAB:                     /* copy entire process table */
 20574	        src_addr = (vir_bytes) mproc;
 20575	        len = sizeof(struct mproc) * NR_PROCS;
 20576	        break;
 20577	  default:
 20578	        return(EINVAL);
 20579	  }
 20580	
 20581	  dst_addr = (vir_bytes) m_in.info_where;
 20582	  if (OK != (s=sys_datacopy(SELF, src_addr, who, dst_addr, len)))
 20583	        return(s);
 20584	  return(OK);
 20585	}
	

_________________________ Page 915    File: servers/pm/misc.c _________________________
 20587	/*===========================================================================*
 20588	 *                              do_getprocnr                                 *
 20589	 *===========================================================================*/
 20590	PUBLIC int do_getprocnr()
 20591	{
 20592	  register struct mproc *rmp;
 20593	  static char search_key[PROC_NAME_LEN+1];
 20594	  int key_len;
 20595	  int s;
 20596	
 20597	  if (m_in.pid >= 0) {                          /* lookup process by pid */
 20598	        for (rmp = &mproc[0]; rmp < &mproc[NR_PROCS]; rmp++) {
 20599	                if ((rmp->mp_flags & IN_USE) && (rmp->mp_pid==m_in.pid)
 20600	                        mp->mp_reply.procnr = (int) (rmp - mproc);
 20601	                        return(OK);
 20602	                } 
 20603	        }
 20604	        return(ESRCH);                  
 20605	  } else if (m_in.namelen > 0) {                /* lookup process by name */
 20606	        key_len = MIN(m_in.namelen, PROC_NAME_LEN);
 20607	        if (OK != (s=sys_datacopy(who, (vir_bytes) m_in.addr, 
 20608	                        SELF, (vir_bytes) search_key, key_len))) 
 20609	                return(s);
 20610	        search_key[key_len] = '\0';     /* terminate for safety */
 20611	        for (rmp = &mproc[0]; rmp < &mproc[NR_PROCS]; rmp++) {
 20612	                if ((rmp->mp_flags & IN_USE) && 
 20613	                        strncmp(rmp->mp_name, search_key, key_len)==0) {
 20614	                        mp->mp_reply.procnr = (int) (rmp - mproc);
 20615	                        return(OK);
 20616	                } 
 20617	        }
 20618	        return(ESRCH);                  
 20619	  } else {                              /* return own process number */
 20620	        mp->mp_reply.procnr = who;
 20621	  }
 20622	  return(OK);
 20623	}
	
 20625	/*===========================================================================*
 20626	 *                              do_reboot                                    *
 20627	 *===========================================================================*/
 20628	#define REBOOT_CODE     "delay; boot"
 20629	PUBLIC int do_reboot()
 20630	{
 20631	  char monitor_code[32*sizeof(char *)];         
 20632	  int code_len;
 20633	  int abort_flag;
 20634	
 20635	  if (mp->mp_effuid != SUPER_USER) return(EPERM);
 20636	
 20637	  switch (m_in.reboot_flag) {
 20638	  case RBT_HALT:
 20639	  case RBT_PANIC:
 20640	  case RBT_RESET:
 20641	        abort_flag = m_in.reboot_flag;
 20642	        break;
 20643	  case RBT_REBOOT:
 20644	        code_len = strlen(REBOOT_CODE) + 1;
 20645	        strncpy(monitor_code, REBOOT_CODE, code_len);        
 20646	        abort_flag = RBT_MONITOR;

_________________________ Page 916    File: servers/pm/misc.c _________________________
 20647	        break;
 20648	  case RBT_MONITOR:
 20649	        code_len = m_in.reboot_strlen + 1;
 20650	        if (code_len > sizeof(monitor_code)) return(EINVAL);
 20651	        if (sys_datacopy(who, (vir_bytes) m_in.reboot_code,
 20652	                PM_PROC_NR, (vir_bytes) monitor_code,
 20653	                (phys_bytes) (code_len)) != OK) return(EFAULT);
 20654	        if (monitor_code[code_len-1] != 0) return(EINVAL);
 20655	        abort_flag = RBT_MONITOR;
 20656	        break;
 20657	  default:
 20658	        return(EINVAL);
 20659	  }
 20660	
 20661	  check_sig(-1, SIGKILL);               /* kill all processes except init */
 20662	  tell_fs(REBOOT,0,0,0);                /* tell FS to prepare for shutdown */
 20663	
 20664	  /* Ask the kernel to abort. All system services, including the PM, will 
 20665	   * get a HARD_STOP notification. Await the notification in the main loop.
 20666	   */
 20667	  sys_abort(abort_flag, PM_PROC_NR, monitor_code, code_len);
 20668	  return(SUSPEND);                      /* don't reply to killed process */
 20669	}
	
 20671	/*===========================================================================*
 20672	 *                              do_getsetpriority                            *
 20673	 *===========================================================================*/
 20674	PUBLIC int do_getsetpriority()
 20675	{
 20676	        int arg_which, arg_who, arg_pri;
 20677	        int rmp_nr;
 20678	        struct mproc *rmp;
 20679	
 20680	        arg_which = m_in.m1_i1;
 20681	        arg_who = m_in.m1_i2;
 20682	        arg_pri = m_in.m1_i3;   /* for SETPRIORITY */
 20683	
 20684	        /* Code common to GETPRIORITY and SETPRIORITY. */
 20685	
 20686	        /* Only support PRIO_PROCESS for now. */
 20687	        if (arg_which != PRIO_PROCESS)
 20688	                return(EINVAL);
 20689	
 20690	        if (arg_who == 0)
 20691	                rmp_nr = who;
 20692	        else
 20693	                if ((rmp_nr = proc_from_pid(arg_who)) < 0)
 20694	                        return(ESRCH);
 20695	
 20696	        rmp = &mproc[rmp_nr];
 20697	
 20698	        if (mp->mp_effuid != SUPER_USER &&
 20699	           mp->mp_effuid != rmp->mp_effuid && mp->mp_effuid != rmp->mp_real
 20700	                return EPERM;
 20701	
 20702	        /* If GET, that's it. */
 20703	        if (call_nr == GETPRIORITY) {
 20704	                return(rmp->mp_nice - PRIO_MIN);
 20705	        }
 20706	

_________________________ Page 917    File: servers/pm/misc.c _________________________
 20707	        /* Only root is allowed to reduce the nice level. */
 20708	        if (rmp->mp_nice > arg_pri && mp->mp_effuid != SUPER_USER)
 20709	                return(EACCES);
 20710	        
 20711	        /* We're SET, and it's allowed. Do it and tell kernel. */
 20712	        rmp->mp_nice = arg_pri;
 20713	        return sys_nice(rmp_nr, arg_pri);
 20714	}
	
 20716	/*===========================================================================*
 20717	 *                              do_svrctl                                    *
 20718	 *===========================================================================*/
 20719	PUBLIC int do_svrctl()
 20720	{
 20721	  int s, req;
 20722	  vir_bytes ptr;
 20723	#define MAX_LOCAL_PARAMS 2
 20724	  static struct {
 20725	        char name[30];
 20726	        char value[30];
 20727	  } local_param_overrides[MAX_LOCAL_PARAMS];
 20728	  static int local_params = 0;
 20729	
 20730	  req = m_in.svrctl_req;
 20731	  ptr = (vir_bytes) m_in.svrctl_argp;
 20732	
 20733	  /* Is the request indeed for the MM? */
 20734	  if (((req >> 8) & 0xFF) != 'M') return(EINVAL);
 20735	
 20736	  /* Control operations local to the PM. */
 20737	  switch(req) {
 20738	  case MMSETPARAM:
 20739	  case MMGETPARAM: {
 20740	      struct sysgetenv sysgetenv;
 20741	      char search_key[64];
 20742	      char *val_start;
 20743	      size_t val_len;
 20744	      size_t copy_len;
 20745	
 20746	      /* Copy sysgetenv structure to PM. */
 20747	      if (sys_datacopy(who, ptr, SELF, (vir_bytes) &sysgetenv, 
 20748	              sizeof(sysgetenv)) != OK) return(EFAULT);  
 20749	
 20750	      /* Set a param override? */
 20751	      if (req == MMSETPARAM) {
 20752	        if (local_params >= MAX_LOCAL_PARAMS) return ENOSPC;
 20753	        if (sysgetenv.keylen <= 0
 20754	         || sysgetenv.keylen >=
 20755	                 sizeof(local_param_overrides[local_params].name)
 20756	         || sysgetenv.vallen <= 0
 20757	         || sysgetenv.vallen >=
 20758	                 sizeof(local_param_overrides[local_params].value))
 20759	                return EINVAL;
 20760	                
 20761	          if ((s = sys_datacopy(who, (vir_bytes) sysgetenv.key,
 20762	            SELF, (vir_bytes) local_param_overrides[local_params].name,
 20763	               sysgetenv.keylen)) != OK)
 20764	                return s;
 20765	          if ((s = sys_datacopy(who, (vir_bytes) sysgetenv.val,
 20766	            SELF, (vir_bytes) local_param_overrides[local_params].value,

_________________________ Page 918    File: servers/pm/misc.c _________________________
 20767	              sysgetenv.keylen)) != OK)
 20768	                return s;
 20769	            local_param_overrides[local_params].name[sysgetenv.keylen] = '\0';
 20770	            local_param_overrides[local_params].value[sysgetenv.vallen] = '\0';
 20771	
 20772	        local_params++;
 20773	
 20774	        return OK;
 20775	      }
 20776	
 20777	      if (sysgetenv.keylen == 0) {      /* copy all parameters */
 20778	          val_start = monitor_params;
 20779	          val_len = sizeof(monitor_params);
 20780	      } 
 20781	      else {                            /* lookup value for key */
 20782	          int p;
 20783	          /* Try to get a copy of the requested key. */
 20784	          if (sysgetenv.keylen > sizeof(search_key)) return(EINVAL);
 20785	          if ((s = sys_datacopy(who, (vir_bytes) sysgetenv.key,
 20786	                  SELF, (vir_bytes) search_key, sysgetenv.keylen)) != OK)
 20787	              return(s);
 20788	
 20789	          /* Make sure key is null-terminated and lookup value.
 20790	           * First check local overrides.
 20791	           */
 20792	          search_key[sysgetenv.keylen-1]= '\0';
 20793	          for(p = 0; p < local_params; p++) {
 20794	                if (!strcmp(search_key, local_param_overrides[p].name)) {
 20795	                        val_start = local_param_overrides[p].value;
 20796	                        break;
 20797	                }
 20798	          }
 20799	          if (p >= local_params && (val_start = find_param(search_key)) == 
 20800	               return(ESRCH);
 20801	          val_len = strlen(val_start) + 1;
 20802	      }
 20803	
 20804	      /* See if it fits in the client's buffer. */
 20805	      if (val_len > sysgetenv.vallen)
 20806	        return E2BIG;
 20807	
 20808	      /* Value found, make the actual copy (as far as possible). */
 20809	      copy_len = MIN(val_len, sysgetenv.vallen); 
 20810	      if ((s=sys_datacopy(SELF, (vir_bytes) val_start, 
 20811	              who, (vir_bytes) sysgetenv.val, copy_len)) != OK)
 20812	          return(s);
 20813	
 20814	      return OK;
 20815	  }
 20816	  default:
 20817	        return(EINVAL);
 20818	  }
 20819	}

_________________________ Page 919    File: servers/pm/misc.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/fs.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 20900	/* This is the master header for fs.  It includes some other files
 20901	 * and defines the principal constants.
 20902	 */
 20903	#define _POSIX_SOURCE      1    /* tell headers to include POSIX stuff */
 20904	#define _MINIX             1    /* tell headers to include MINIX stuff */
 20905	#define _SYSTEM            1    /* tell headers that this is the kernel */
 20906	
 20907	#define VERBOSE            0    /* show messages during initialization? */
 20908	
 20909	/* The following are so basic, all the *.c files get them automatically. */
 20910	#include <minix/config.h>       /* MUST be first */
 20911	#include <ansi.h>               /* MUST be second */
 20912	#include <sys/types.h>
 20913	#include <minix/const.h>
 20914	#include <minix/type.h>
 20915	#include <minix/dmap.h>
 20916	
 20917	#include <limits.h>
 20918	#include <errno.h>
 20919	
 20920	#include <minix/syslib.h>
 20921	#include <minix/sysutil.h>
 20922	
 20923	#include "const.h"
 20924	#include "type.h"
 20925	#include "proto.h"
 20926	#include "glo.h"




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/const.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21000	/* Tables sizes */
 21001	#define V1_NR_DZONES       7    /* # direct zone numbers in a V1 inode */
 21002	#define V1_NR_TZONES       9    /* total # zone numbers in a V1 inode */
 21003	#define V2_NR_DZONES       7    /* # direct zone numbers in a V2 inode */
 21004	#define V2_NR_TZONES      10    /* total # zone numbers in a V2 inode */
 21005	
 21006	#define NR_FILPS         128    /* # slots in filp table */
 21007	#define NR_INODES         64    /* # slots in "in core" inode table */
 21008	#define NR_SUPERS          8    /* # slots in super block table */
 21009	#define NR_LOCKS           8    /* # slots in the file locking table */
 21010	
 21011	/* The type of sizeof may be (unsigned) long.  Use the following macro for
 21012	 * taking the sizes of small objects so that there are no surprises like
 21013	 * (small) long constants being passed to routines expecting an int.
 21014	 */
 21015	#define usizeof(t) ((unsigned) sizeof(t))
 21016	
 21017	/* File system types. */
 21018	#define SUPER_MAGIC   0x137F    /* magic number contained in super-block */
 21019	#define SUPER_REV     0x7F13    /* magic # when 68000 disk read on PC or vv */

_________________________ Page 920    File: servers/fs/const.h _________________________
 21020	#define SUPER_V2      0x2468    /* magic # for V2 file systems */
 21021	#define SUPER_V2_REV  0x6824    /* V2 magic written on PC, read on 68K or vv */
 21022	#define SUPER_V3      0x4d5a    /* magic # for V3 file systems */
 21023	
 21024	#define V1                 1    /* version number of V1 file systems */ 
 21025	#define V2                 2    /* version number of V2 file systems */ 
 21026	#define V3                 3    /* version number of V3 file systems */ 
 21027	
 21028	/* Miscellaneous constants */
 21029	#define SU_UID   ((uid_t) 0)    /* super_user's uid_t */
 21030	#define SYS_UID  ((uid_t) 0)    /* uid_t for processes MM and INIT */
 21031	#define SYS_GID  ((gid_t) 0)    /* gid_t for processes MM and INIT */
 21032	#define NORMAL             0    /* forces get_block to do disk read */
 21033	#define NO_READ            1    /* prevents get_block from doing disk read */
 21034	#define PREFETCH           2    /* tells get_block not to read or mark dev */
 21035	
 21036	#define XPIPE   (-NR_TASKS-1)   /* used in fp_task when susp'd on pipe */
 21037	#define XLOCK   (-NR_TASKS-2)   /* used in fp_task when susp'd on lock */
 21038	#define XPOPEN  (-NR_TASKS-3)   /* used in fp_task when susp'd on pipe open */
 21039	#define XSELECT (-NR_TASKS-4)   /* used in fp_task when susp'd on select */
 21040	
 21041	#define NO_BIT   ((bit_t) 0)    /* returned by alloc_bit() to signal failure */
 21042	
 21043	#define DUP_MASK        0100    /* mask to distinguish dup2 from dup */
 21044	
 21045	#define LOOK_UP            0 /* tells search_dir to lookup string */
 21046	#define ENTER              1 /* tells search_dir to make dir entry */
 21047	#define DELETE             2 /* tells search_dir to delete entry */
 21048	#define IS_EMPTY           3 /* tells search_dir to ret. OK or ENOTEMPTY */  
 21049	
 21050	#define CLEAN              0    /* disk and memory copies identical */
 21051	#define DIRTY              1    /* disk and memory copies differ */
 21052	#define ATIME            002    /* set if atime field needs updating */
 21053	#define CTIME            004    /* set if ctime field needs updating */
 21054	#define MTIME            010    /* set if mtime field needs updating */
 21055	
 21056	#define BYTE_SWAP          0    /* tells conv2/conv4 to swap bytes */
 21057	
 21058	#define END_OF_FILE   (-104)    /* eof detected */
 21059	
 21060	#define ROOT_INODE         1            /* inode number for root directory */
 21061	#define BOOT_BLOCK  ((block_t) 0)       /* block number of boot block */
 21062	#define SUPER_BLOCK_BYTES (1024)        /* bytes offset */
 21063	#define START_BLOCK     2               /* first block of FS (not counting SB) */
 21064	
 21065	#define DIR_ENTRY_SIZE       usizeof (struct direct)  /* # bytes/dir entry   */
 21066	#define NR_DIR_ENTRIES(b)   ((b)/DIR_ENTRY_SIZE)  /* # dir entries/blk   */
 21067	#define SUPER_SIZE      usizeof (struct super_block)  /* super_block size    */
 21068	#define PIPE_SIZE(b)          (V1_NR_DZONES*(b))  /* pipe size in bytes  */
 21069	
 21070	#define FS_BITMAP_CHUNKS(b) ((b)/usizeof (bitchunk_t))/* # map chunks/blk   */
 21071	#define FS_BITCHUNK_BITS                (usizeof(bitchunk_t) * CHAR_BIT)
 21072	#define FS_BITS_PER_BLOCK(b)    (FS_BITMAP_CHUNKS(b) * FS_BITCHUNK_BITS)
 21073	
 21074	/* Derived sizes pertaining to the V1 file system. */
 21075	#define V1_ZONE_NUM_SIZE           usizeof (zone1_t)  /* # bytes in V1 zone  */
 21076	#define V1_INODE_SIZE             usizeof (d1_inode)  /* bytes in V1 dsk ino */
 21077	
 21078	/* # zones/indir block */
 21079	#define V1_INDIRECTS (STATIC_BLOCK_SIZE/V1_ZONE_NUM_SIZE)  

_________________________ Page 921    File: servers/fs/const.h _________________________
 21080	
 21081	/* # V1 dsk inodes/blk */
 21082	#define V1_INODES_PER_BLOCK (STATIC_BLOCK_SIZE/V1_INODE_SIZE)
 21083	
 21084	/* Derived sizes pertaining to the V2 file system. */
 21085	#define V2_ZONE_NUM_SIZE            usizeof (zone_t)  /* # bytes in V2 zone  */
 21086	#define V2_INODE_SIZE             usizeof (d2_inode)  /* bytes in V2 dsk ino */
 21087	#define V2_INDIRECTS(b)   ((b)/V2_ZONE_NUM_SIZE)  /* # zones/indir block */
 21088	#define V2_INODES_PER_BLOCK(b) ((b)/V2_INODE_SIZE)/* # V2 dsk inodes/blk */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/type.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21100	/* Declaration of the V1 inode as it is on the disk (not in core). */
 21101	typedef struct {                /* V1.x disk inode */
 21102	  mode_t d1_mode;               /* file type, protection, etc. */
 21103	  uid_t d1_uid;                 /* user id of the file's owner */
 21104	  off_t d1_size;                /* current file size in bytes */
 21105	  time_t d1_mtime;              /* when was file data last changed */
 21106	  u8_t d1_gid;                  /* group number */
 21107	  u8_t d1_nlinks;               /* how many links to this file */
 21108	  u16_t d1_zone[V1_NR_TZONES];  /* block nums for direct, ind, and dbl ind */
 21109	} d1_inode;
 21110	
 21111	/* Declaration of the V2 inode as it is on the disk (not in core). */
 21112	typedef struct {                /* V2.x disk inode */
 21113	  mode_t d2_mode;               /* file type, protection, etc. */
 21114	  u16_t d2_nlinks;              /* how many links to this file. HACK! */
 21115	  uid_t d2_uid;                 /* user id of the file's owner. */
 21116	  u16_t d2_gid;                 /* group number HACK! */
 21117	  off_t d2_size;                /* current file size in bytes */
 21118	  time_t d2_atime;              /* when was file data last accessed */
 21119	  time_t d2_mtime;              /* when was file data last changed */
 21120	  time_t d2_ctime;              /* when was inode data last changed */
 21121	  zone_t d2_zone[V2_NR_TZONES]; /* block nums for direct, ind, and dbl ind */
 21122	} d2_inode;



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/proto.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21200	/* Function prototypes. */
 21201	
 21202	#include "timers.h"
 21203	
 21204	/* Structs used in prototypes must be declared as such first. */
 21205	struct buf;
 21206	struct filp;            
 21207	struct inode;
 21208	struct super_block;
 21209	
 21210	/* cache.c */
 21211	_PROTOTYPE( zone_t alloc_zone, (Dev_t dev, zone_t z)                    );
 21212	_PROTOTYPE( void flushall, (Dev_t dev)                                  );
 21213	_PROTOTYPE( void free_zone, (Dev_t dev, zone_t numb)                    );
 21214	_PROTOTYPE( struct buf *get_block, (Dev_t dev, block_t block,int only_search));

_________________________ Page 922    File: servers/fs/proto.h _________________________
 21215	_PROTOTYPE( void invalidate, (Dev_t device)                             );
 21216	_PROTOTYPE( void put_block, (struct buf *bp, int block_type)            );
 21217	_PROTOTYPE( void rw_block, (struct buf *bp, int rw_flag)                );
 21218	_PROTOTYPE( void rw_scattered, (Dev_t dev,
 21219	                        struct buf **bufq, int bufqsize, int rw_flag)   );
 21220	
 21221	/* device.c */
 21222	_PROTOTYPE( int dev_open, (Dev_t dev, int proc, int flags)              );
 21223	_PROTOTYPE( void dev_close, (Dev_t dev)                                 );
 21224	_PROTOTYPE( int dev_io, (int op, Dev_t dev, int proc, void *buf,
 21225	                        off_t pos, int bytes, int flags)                );
 21226	_PROTOTYPE( int gen_opcl, (int op, Dev_t dev, int proc, int flags)      );
 21227	_PROTOTYPE( void gen_io, (int task_nr, message *mess_ptr)               );
 21228	_PROTOTYPE( int no_dev, (int op, Dev_t dev, int proc, int flags)        );
 21229	_PROTOTYPE( int tty_opcl, (int op, Dev_t dev, int proc, int flags)      );
 21230	_PROTOTYPE( int ctty_opcl, (int op, Dev_t dev, int proc, int flags)     );
 21231	_PROTOTYPE( int clone_opcl, (int op, Dev_t dev, int proc, int flags)    );
 21232	_PROTOTYPE( void ctty_io, (int task_nr, message *mess_ptr)              );
 21233	_PROTOTYPE( int do_ioctl, (void)                                        );
 21234	_PROTOTYPE( int do_setsid, (void)                                       );
 21235	_PROTOTYPE( void dev_status, (message *)                                );
 21236	
 21237	/* dmp.c */
 21238	_PROTOTYPE( int do_fkey_pressed, (void)                                 );
 21239	
 21240	/* dmap.c */
 21241	_PROTOTYPE( int do_devctl, (void)                                       );
 21242	_PROTOTYPE( void build_dmap, (void)                                     );
 21243	_PROTOTYPE( int map_driver, (int major, int proc_nr, int dev_style)     );
 21244	
 21245	/* filedes.c */
 21246	_PROTOTYPE( struct filp *find_filp, (struct inode *rip, mode_t bits)    );
 21247	_PROTOTYPE( int get_fd, (int start, mode_t bits, int *k, struct filp **fpt) );
 21248	_PROTOTYPE( struct filp *get_filp, (int fild)                           );
 21249	
 21250	/* inode.c */
 21251	_PROTOTYPE( struct inode *alloc_inode, (dev_t dev, mode_t bits)         );
 21252	_PROTOTYPE( void dup_inode, (struct inode *ip)                          );
 21253	_PROTOTYPE( void free_inode, (Dev_t dev, Ino_t numb)                    );
 21254	_PROTOTYPE( struct inode *get_inode, (Dev_t dev, int numb)              );
 21255	_PROTOTYPE( void put_inode, (struct inode *rip)                         );
 21256	_PROTOTYPE( void update_times, (struct inode *rip)                      );
 21257	_PROTOTYPE( void rw_inode, (struct inode *rip, int rw_flag)             );
 21258	_PROTOTYPE( void wipe_inode, (struct inode *rip)                        );
 21259	
 21260	/* link.c */
 21261	_PROTOTYPE( int do_link, (void)                                         );
 21262	_PROTOTYPE( int do_unlink, (void)                                       );
 21263	_PROTOTYPE( int do_rename, (void)                                       );
 21264	_PROTOTYPE( void truncate, (struct inode *rip)                          );
 21265	
 21266	/* lock.c */
 21267	_PROTOTYPE( int lock_op, (struct filp *f, int req)                      );
 21268	_PROTOTYPE( void lock_revive, (void)                                    );
 21269	
 21270	/* main.c */
 21271	_PROTOTYPE( int main, (void)                                            );
 21272	_PROTOTYPE( void reply, (int whom, int result)                          );
 21273	
 21274	/* misc.c */

_________________________ Page 923    File: servers/fs/proto.h _________________________
 21275	_PROTOTYPE( int do_dup, (void)                                          );
 21276	_PROTOTYPE( int do_exit, (void)                                         );
 21277	_PROTOTYPE( int do_fcntl, (void)                                        );
 21278	_PROTOTYPE( int do_fork, (void)                                         );
 21279	_PROTOTYPE( int do_exec, (void)                                         );
 21280	_PROTOTYPE( int do_revive, (void)                                       );
 21281	_PROTOTYPE( int do_set, (void)                                          );
 21282	_PROTOTYPE( int do_sync, (void)                                         );
 21283	_PROTOTYPE( int do_fsync, (void)                                        );
 21284	_PROTOTYPE( int do_reboot, (void)                                       );
 21285	_PROTOTYPE( int do_svrctl, (void)                                       );
 21286	_PROTOTYPE( int do_getsysinfo, (void)                                   );
 21287	
 21288	/* mount.c */
 21289	_PROTOTYPE( int do_mount, (void)                                        );
 21290	_PROTOTYPE( int do_umount, (void)                                       );
 21291	_PROTOTYPE( int unmount, (Dev_t dev)                                    );
 21292	
 21293	/* open.c */
 21294	_PROTOTYPE( int do_close, (void)                                        );
 21295	_PROTOTYPE( int do_creat, (void)                                        );
 21296	_PROTOTYPE( int do_lseek, (void)                                        );
 21297	_PROTOTYPE( int do_mknod, (void)                                        );
 21298	_PROTOTYPE( int do_mkdir, (void)                                        );
 21299	_PROTOTYPE( int do_open, (void)                                         );
 21300	
 21301	/* path.c */
 21302	_PROTOTYPE( struct inode *advance,(struct inode *dirp, char string[NAME_MAX]));
 21303	_PROTOTYPE( int search_dir, (struct inode *ldir_ptr,
 21304	                        char string [NAME_MAX], ino_t *numb, int flag)  );
 21305	_PROTOTYPE( struct inode *eat_path, (char *path)                        );
 21306	_PROTOTYPE( struct inode *last_dir, (char *path, char string [NAME_MAX]));
 21307	
 21308	/* pipe.c */
 21309	_PROTOTYPE( int do_pipe, (void)                                         );
 21310	_PROTOTYPE( int do_unpause, (void)                                      );
 21311	_PROTOTYPE( int pipe_check, (struct inode *rip, int rw_flag,
 21312	                        int oflags, int bytes, off_t position, int *canwrite, int n
 21313	_PROTOTYPE( void release, (struct inode *ip, int call_nr, int count)    );
 21314	_PROTOTYPE( void revive, (int proc_nr, int bytes)                       );
 21315	_PROTOTYPE( void suspend, (int task)                                    );
 21316	_PROTOTYPE( int select_request_pipe, (struct filp *f, int *ops, int bl) );
 21317	_PROTOTYPE( int select_cancel_pipe, (struct filp *f)                    );
 21318	_PROTOTYPE( int select_match_pipe, (struct filp *f)                     );
 21319	
 21320	/* protect.c */
 21321	_PROTOTYPE( int do_access, (void)                                       );
 21322	_PROTOTYPE( int do_chmod, (void)                                        );
 21323	_PROTOTYPE( int do_chown, (void)                                        );
 21324	_PROTOTYPE( int do_umask, (void)                                        );
 21325	_PROTOTYPE( int forbidden, (struct inode *rip, mode_t access_desired)   );
 21326	_PROTOTYPE( int read_only, (struct inode *ip)                           );
 21327	
 21328	/* read.c */
 21329	_PROTOTYPE( int do_read, (void)                                         );
 21330	_PROTOTYPE( struct buf *rahead, (struct inode *rip, block_t baseblock,
 21331	                        off_t position, unsigned bytes_ahead)           );
 21332	_PROTOTYPE( void read_ahead, (void)                                     );
 21333	_PROTOTYPE( block_t read_map, (struct inode *rip, off_t position)       );
 21334	_PROTOTYPE( int read_write, (int rw_flag)                               );

_________________________ Page 924    File: servers/fs/proto.h _________________________
 21335	_PROTOTYPE( zone_t rd_indir, (struct buf *bp, int index)                );
 21336	
 21337	/* stadir.c */
 21338	_PROTOTYPE( int do_chdir, (void)                                        );
 21339	_PROTOTYPE( int do_fchdir, (void)                                       );
 21340	_PROTOTYPE( int do_chroot, (void)                                       );
 21341	_PROTOTYPE( int do_fstat, (void)                                        );
 21342	_PROTOTYPE( int do_stat, (void)                                         );
 21343	_PROTOTYPE( int do_fstatfs, (void)                                      );
 21344	
 21345	/* super.c */
 21346	_PROTOTYPE( bit_t alloc_bit, (struct super_block *sp, int map, bit_t origin));
 21347	_PROTOTYPE( void free_bit, (struct super_block *sp, int map,
 21348	                                                bit_t bit_returned)     );
 21349	_PROTOTYPE( struct super_block *get_super, (Dev_t dev)                  );
 21350	_PROTOTYPE( int mounted, (struct inode *rip)                            );
 21351	_PROTOTYPE( int read_super, (struct super_block *sp)                    );
 21352	_PROTOTYPE( int get_block_size, (dev_t dev)                             );
 21353	
 21354	/* time.c */
 21355	_PROTOTYPE( int do_stime, (void)                                        );
 21356	_PROTOTYPE( int do_utime, (void)                                        );
 21357	
 21358	/* utility.c */
 21359	_PROTOTYPE( time_t clock_time, (void)                                   );
 21360	_PROTOTYPE( unsigned conv2, (int norm, int w)                           );
 21361	_PROTOTYPE( long conv4, (int norm, long x)                              );
 21362	_PROTOTYPE( int fetch_name, (char *path, int len, int flag)             );
 21363	_PROTOTYPE( int no_sys, (void)                                          );
 21364	_PROTOTYPE( void panic, (char *who, char *mess, int num)                );
 21365	
 21366	/* write.c */
 21367	_PROTOTYPE( void clear_zone, (struct inode *rip, off_t pos, int flag)   );
 21368	_PROTOTYPE( int do_write, (void)                                        );
 21369	_PROTOTYPE( struct buf *new_block, (struct inode *rip, off_t position)  );
 21370	_PROTOTYPE( void zero_block, (struct buf *bp)                           );
 21371	
 21372	/* select.c */
 21373	_PROTOTYPE( int do_select, (void)                                       );
 21374	_PROTOTYPE( int select_callback, (struct filp *, int ops)               );
 21375	_PROTOTYPE( void select_forget, (int fproc)                             );
 21376	_PROTOTYPE( void select_timeout_check, (timer_t *)                      );
 21377	_PROTOTYPE( void init_select, (void)                                    );
 21378	_PROTOTYPE( int select_notified, (int major, int minor, int ops)        );
 21379	
 21380	/* timers.c */
 21381	_PROTOTYPE( void fs_set_timer, (timer_t *tp, int delta, tmr_func_t watchdog, int ar
 21382	_PROTOTYPE( void fs_expire_timers, (clock_t now)                        );
 21383	_PROTOTYPE( void fs_cancel_timer, (timer_t *tp)                         );
 21384	_PROTOTYPE( void fs_init_timer, (timer_t *tp)                           );
 21385	
 21386	/* cdprobe.c */
 21387	_PROTOTYPE( int cdprobe, (void)                                         );

_________________________ Page 925    File: servers/fs/proto.h _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/glo.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21400	/* EXTERN should be extern except for the table file */
 21401	#ifdef _TABLE
 21402	#undef EXTERN
 21403	#define EXTERN
 21404	#endif
 21405	
 21406	/* File System global variables */
 21407	EXTERN struct fproc *fp;        /* pointer to caller's fproc struct */
 21408	EXTERN int super_user;          /* 1 if caller is super_user, else 0 */
 21409	EXTERN int susp_count;          /* number of procs suspended on pipe */
 21410	EXTERN int nr_locks;            /* number of locks currently in place */
 21411	EXTERN int reviving;            /* number of pipe processes to be revived */
 21412	EXTERN off_t rdahedpos;         /* position to read ahead */
 21413	EXTERN struct inode *rdahed_inode;      /* pointer to inode to read ahead */
 21414	EXTERN Dev_t root_dev;          /* device number of the root device */
 21415	EXTERN time_t boottime;         /* time in seconds at system boot */
 21416	
 21417	/* The parameters of the call are kept here. */
 21418	EXTERN message m_in;            /* the input message itself */
 21419	EXTERN message m_out;           /* the output message used for reply */
 21420	EXTERN int who;                 /* caller's proc number */
 21421	EXTERN int call_nr;             /* system call number */
 21422	EXTERN char user_path[PATH_MAX];/* storage for user path name */
 21423	
 21424	/* The following variables are used for returning results to the caller. */
 21425	EXTERN int err_code;            /* temporary storage for error number */
 21426	EXTERN int rdwt_err;            /* status of last disk i/o request */
 21427	
 21428	/* Data initialized elsewhere. */
 21429	extern _PROTOTYPE (int (*call_vec[]), (void) ); /* sys call table */
 21430	extern char dot1[2];   /* dot1 (&dot1[0]) and dot2 (&dot2[0]) have a specia
 21431	extern char dot2[3];   /* meaning to search_dir: no access permission check. */




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/fproc.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21500	/* This is the per-process information.  A slot is reserved for each potential
 21501	 * process. Thus NR_PROCS must be the same as in the kernel. It is not 
 21502	 * possible or even necessary to tell when a slot is free here.
 21503	 */
 21504	EXTERN struct fproc {
 21505	  mode_t fp_umask;              /* mask set by umask system call */
 21506	  struct inode *fp_workdir;     /* pointer to working directory's inode */
 21507	  struct inode *fp_rootdir;     /* pointer to current root dir (see chroot) */
 21508	  struct filp *fp_filp[OPEN_MAX];/* the file descriptor table */
 21509	  uid_t fp_realuid;             /* real user id */
 21510	  uid_t fp_effuid;              /* effective user id */
 21511	  gid_t fp_realgid;             /* real group id */
 21512	  gid_t fp_effgid;              /* effective group id */
 21513	  dev_t fp_tty;                 /* major/minor of controlling tty */
 21514	  int fp_fd;                    /* place to save fd if rd/wr can't finish */

_________________________ Page 926    File: servers/fs/fproc.h _________________________
 21515	  char *fp_buffer;              /* place to save buffer if rd/wr can't finish*/
 21516	  int  fp_nbytes;               /* place to save bytes if rd/wr can't finish */
 21517	  int  fp_cum_io_partial;       /* partial byte count if rd/wr can't finish */
 21518	  char fp_suspended;            /* set to indicate process hanging */
 21519	  char fp_revived;              /* set to indicate process being revived */
 21520	  char fp_task;                 /* which task is proc suspended on */
 21521	  char fp_sesldr;               /* true if proc is a session leader */
 21522	  pid_t fp_pid;                 /* process id */
 21523	  long fp_cloexec;              /* bit map for POSIX Table 6-2 FD_CLOEXEC */
 21524	} fproc[NR_PROCS];
 21525	
 21526	/* Field values. */
 21527	#define NOT_SUSPENDED      0    /* process is not suspended on pipe or task */
 21528	#define SUSPENDED          1    /* process is suspended on pipe or task */
 21529	#define NOT_REVIVING       0    /* process is not being revived */
 21530	#define REVIVING           1    /* process is being revived from suspension */
 21531	#define PID_FREE           0    /* process slot free */
 21532	
 21533	/* Check is process number is acceptable - includes system processes. */
 21534	#define isokprocnr(n)   ((unsigned)((n)+NR_TASKS) < NR_PROCS + NR_TASKS)
 21535	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/buf.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21600	/* Buffer (block) cache.  To acquire a block, a routine calls get_block(),
 21601	 * telling which block it wants.  The block is then regarded as "in use"
 21602	 * and has its 'b_count' field incremented.  All the blocks that are not
 21603	 * in use are chained together in an LRU list, with 'front' pointing
 21604	 * to the least recently used block, and 'rear' to the most recently used
 21605	 * block.  A reverse chain, using the field b_prev is also maintained.
 21606	 * Usage for LRU is measured by the time the put_block() is done.  The second
 21607	 * parameter to put_block() can violate the LRU order and put a block on the
 21608	 * front of the list, if it will probably not be needed soon.  If a block
 21609	 * is modified, the modifying routine must set b_dirt to DIRTY, so the block
 21610	 * will eventually be rewritten to the disk.
 21611	 */
 21612	
 21613	#include <sys/dir.h>                    /* need struct direct */
 21614	#include <dirent.h>
 21615	
 21616	EXTERN struct buf {
 21617	  /* Data portion of the buffer. */
 21618	  union {
 21619	    char b__data[MAX_BLOCK_SIZE];                    /* ordinary user data */
 21620	/* directory block */
 21621	    struct direct b__dir[NR_DIR_ENTRIES(MAX_BLOCK_SIZE)];    
 21622	/* V1 indirect block */
 21623	    zone1_t b__v1_ind[V1_INDIRECTS];         
 21624	/* V2 indirect block */
 21625	    zone_t  b__v2_ind[V2_INDIRECTS(MAX_BLOCK_SIZE)];         
 21626	/* V1 inode block */
 21627	    d1_inode b__v1_ino[V1_INODES_PER_BLOCK]; 
 21628	/* V2 inode block */
 21629	    d2_inode b__v2_ino[V2_INODES_PER_BLOCK(MAX_BLOCK_SIZE)]; 

_________________________ Page 927    File: servers/fs/buf.h _________________________
 21630	/* bit map block */
 21631	    bitchunk_t b__bitmap[FS_BITMAP_CHUNKS(MAX_BLOCK_SIZE)];  
 21632	  } b;
 21633	
 21634	  /* Header portion of the buffer. */
 21635	  struct buf *b_next;           /* used to link all free bufs in a chain */
 21636	  struct buf *b_prev;           /* used to link all free bufs the other way */
 21637	  struct buf *b_hash;           /* used to link bufs on hash chains */
 21638	  block_t b_blocknr;            /* block number of its (minor) device */
 21639	  dev_t b_dev;                  /* major | minor device where block resides */
 21640	  char b_dirt;                  /* CLEAN or DIRTY */
 21641	  char b_count;                 /* number of users of this buffer */
 21642	} buf[NR_BUFS];
 21643	
 21644	/* A block is free if b_dev == NO_DEV. */
 21645	
 21646	#define NIL_BUF ((struct buf *) 0)      /* indicates absence of a buffer */
 21647	
 21648	/* These defs make it possible to use to bp->b_data instead of bp->b.b__data */
 21649	#define b_data   b.b__data
 21650	#define b_dir    b.b__dir
 21651	#define b_v1_ind b.b__v1_ind
 21652	#define b_v2_ind b.b__v2_ind
 21653	#define b_v1_ino b.b__v1_ino
 21654	#define b_v2_ino b.b__v2_ino
 21655	#define b_bitmap b.b__bitmap
 21656	
 21657	EXTERN struct buf *buf_hash[NR_BUF_HASH];       /* the buffer hash table */
 21658	
 21659	EXTERN struct buf *front;       /* points to least recently used free block */
 21660	EXTERN struct buf *rear;        /* points to most recently used free block */
 21661	EXTERN int bufs_in_use;         /* # bufs currently in use (not on free list)*/
 21662	
 21663	/* When a block is released, the type of usage is passed to put_block(). */
 21664	#define WRITE_IMMED   0100 /* block should be written to disk now */
 21665	#define ONE_SHOT      0200 /* set if block not likely to be needed soon */
 21666	
 21667	#define INODE_BLOCK        0                             /* inode block */
 21668	#define DIRECTORY_BLOCK    1                             /* directory block */
 21669	#define INDIRECT_BLOCK     2                             /* pointer block */
 21670	#define MAP_BLOCK          3                             /* bit map */
 21671	#define FULL_DATA_BLOCK    5                             /* data, fully used */
 21672	#define PARTIAL_DATA_BLOCK 6                             /* data, partly used*/
 21673	
 21674	#define HASH_MASK (NR_BUF_HASH - 1)     /* mask for hashing block numbers */

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/file.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21700	/* This is the filp table.  It is an intermediary between file descriptors and
 21701	 * inodes.  A slot is free if filp_count == 0.
 21702	 */
 21703	
 21704	EXTERN struct filp {
 21705	  mode_t filp_mode;             /* RW bits, telling how file is opened */
 21706	  int filp_flags;               /* flags from open and fcntl */
 21707	  int filp_count;               /* how many file descriptors share this slot?*/
 21708	  struct inode *filp_ino;       /* pointer to the inode */
 21709	  off_t filp_pos;               /* file position */

_________________________ Page 928    File: servers/fs/file.h _________________________
 21710	
 21711	  /* the following fields are for select() and are owned by the generic
 21712	   * select() code (i.e., fd-type-specific select() code can't touch these).
 21713	   */
 21714	  int filp_selectors;           /* select()ing processes blocking on this fd */
 21715	  int filp_select_ops;          /* interested in these SEL_* operations */
 21716	
 21717	  /* following are for fd-type-specific select() */
 21718	  int filp_pipe_select_ops;
 21719	} filp[NR_FILPS];
 21720	
 21721	#define FILP_CLOSED     0       /* filp_mode: associated device closed */
 21722	
 21723	#define NIL_FILP (struct filp *) 0      /* indicates absence of a filp slot */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/lock.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21800	/* This is the file locking table.  Like the filp table, it points to the
 21801	 * inode table, however, in this case to achieve advisory locking.
 21802	 */
 21803	EXTERN struct file_lock {
 21804	  short lock_type;              /* F_RDLOCK or F_WRLOCK; 0 means unused slot */
 21805	  pid_t lock_pid;               /* pid of the process holding the lock */
 21806	  struct inode *lock_inode;     /* pointer to the inode locked */
 21807	  off_t lock_first;             /* offset of first byte locked */
 21808	  off_t lock_last;              /* offset of last byte locked */
 21809	} file_lock[NR_LOCKS];

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/inode.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 21900	/* Inode table.  This table holds inodes that are currently in use.  In some
 21901	 * cases they have been opened by an open() or creat() system call, in other
 21902	 * cases the file system itself needs the inode for one reason or another,
 21903	 * such as to search a directory for a path name.
 21904	 * The first part of the struct holds fields that are present on the
 21905	 * disk; the second part holds fields not present on the disk.
 21906	 * The disk inode part is also declared in "type.h" as 'd1_inode' for V1
 21907	 * file systems and 'd2_inode' for V2 file systems.
 21908	 */
 21909	
 21910	EXTERN struct inode {
 21911	  mode_t i_mode;                /* file type, protection, etc. */
 21912	  nlink_t i_nlinks;             /* how many links to this file */
 21913	  uid_t i_uid;                  /* user id of the file's owner */
 21914	  gid_t i_gid;                  /* group number */
 21915	  off_t i_size;                 /* current file size in bytes */
 21916	  time_t i_atime;               /* time of last access (V2 only) */
 21917	  time_t i_mtime;               /* when was file data last changed */
 21918	  time_t i_ctime;               /* when was inode itself changed (V2 only)*/
 21919	  zone_t i_zone[V2_NR_TZONES]; /* zone numbers for direct, ind, and dbl ind */
 21920	  
 21921	  /* The following items are not present on the disk. */
 21922	  dev_t i_dev;                  /* which device is the inode on */
 21923	  ino_t i_num;                  /* inode number on its (minor) device */
 21924	  int i_count;                  /* # times inode used; 0 means slot is free */

_________________________ Page 929    File: servers/fs/inode.h _________________________
 21925	  int i_ndzones;                /* # direct zones (Vx_NR_DZONES) */
 21926	  int i_nindirs;                /* # indirect zones per indirect block */
 21927	  struct super_block *i_sp;     /* pointer to super block for inode's device */
 21928	  char i_dirt;                  /* CLEAN or DIRTY */
 21929	  char i_pipe;                  /* set to I_PIPE if pipe */
 21930	  char i_mount;                 /* this bit is set if file mounted on */
 21931	  char i_seek;                  /* set on LSEEK, cleared on READ/WRITE */
 21932	  char i_update;                /* the ATIME, CTIME, and MTIME bits are here */
 21933	} inode[NR_INODES];
 21934	
 21935	#define NIL_INODE (struct inode *) 0    /* indicates absence of inode slot */
 21936	
 21937	/* Field values.  Note that CLEAN and DIRTY are defined in "const.h" */
 21938	#define NO_PIPE            0    /* i_pipe is NO_PIPE if inode is not a pipe */
 21939	#define I_PIPE             1    /* i_pipe is I_PIPE if inode is a pipe */
 21940	#define NO_MOUNT           0    /* i_mount is NO_MOUNT if file not mounted on*/
 21941	#define I_MOUNT            1    /* i_mount is I_MOUNT if file mounted on */
 21942	#define NO_SEEK            0    /* i_seek = NO_SEEK if last op was not SEEK */
 21943	#define ISEEK              1    /* i_seek = ISEEK if last op was SEEK */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/param.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 22000	/* The following names are synonyms for the variables in the input message. */
 22001	#define acc_time      m2_l1
 22002	#define addr          m1_i3
 22003	#define buffer        m1_p1
 22004	#define child         m1_i2
 22005	#define co_mode       m1_i1
 22006	#define eff_grp_id    m1_i3
 22007	#define eff_user_id   m1_i3
 22008	#define erki          m1_p1
 22009	#define fd            m1_i1
 22010	#define fd2           m1_i2
 22011	#define ioflags       m1_i3
 22012	#define group         m1_i3
 22013	#define real_grp_id   m1_i2
 22014	#define ls_fd         m2_i1
 22015	#define mk_mode       m1_i2
 22016	#define mk_z0         m1_i3
 22017	#define mode          m3_i2
 22018	#define c_mode        m1_i3
 22019	#define c_name        m1_p1
 22020	#define name          m3_p1
 22021	#define name1         m1_p1
 22022	#define name2         m1_p2
 22023	#define name_length   m3_i1
 22024	#define name1_length  m1_i1
 22025	#define name2_length  m1_i2
 22026	#define nbytes        m1_i2
 22027	#define owner         m1_i2
 22028	#define parent        m1_i1
 22029	#define pathname      m3_ca1
 22030	#define pid           m1_i3
 22031	#define pro           m1_i1
 22032	#define ctl_req       m4_l1
 22033	#define driver_nr     m4_l2
 22034	#define dev_nr        m4_l3

_________________________ Page 930    File: servers/fs/param.h _________________________
 22035	#define dev_style     m4_l4
 22036	#define rd_only       m1_i3
 22037	#define real_user_id  m1_i2
 22038	#define request       m1_i2
 22039	#define sig           m1_i2
 22040	#define slot1         m1_i1
 22041	#define tp            m2_l1
 22042	#define utime_actime  m2_l1
 22043	#define utime_modtime m2_l2
 22044	#define utime_file    m2_p1
 22045	#define utime_length  m2_i1
 22046	#define utime_strlen  m2_i2
 22047	#define whence        m2_i2
 22048	#define svrctl_req    m2_i1
 22049	#define svrctl_argp   m2_p1
 22050	#define pm_stime      m1_i1
 22051	#define info_what     m1_i1
 22052	#define info_where    m1_p1
 22053	
 22054	/* The following names are synonyms for the variables in the output message. */
 22055	#define reply_type    m_type
 22056	#define reply_l1      m2_l1
 22057	#define reply_i1      m1_i1
 22058	#define reply_i2      m1_i2
 22059	#define reply_t1      m4_l1
 22060	#define reply_t2      m4_l2
 22061	#define reply_t3      m4_l3
 22062	#define reply_t4      m4_l4
 22063	#define reply_t5      m4_l5


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/super.h
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 22100	/* Super block table.  The root file system and every mounted file system
 22101	 * has an entry here.  The entry holds information about the sizes of the bit
 22102	 * maps and inodes.  The s_ninodes field gives the number of inodes available
 22103	 * for files and directories, including the root directory.  Inode 0 is 
 22104	 * on the disk, but not used.  Thus s_ninodes = 4 means that 5 bits will be
 22105	 * used in the bit map, bit 0, which is always 1 and not used, and bits 1-4
 22106	 * for files and directories.  The disk layout is:
 22107	 *
 22108	 *    Item        # blocks
 22109	 *    boot block      1
 22110	 *    super block     1    (offset 1kB)
 22111	 *    inode map     s_imap_blocks
 22112	 *    zone map      s_zmap_blocks
 22113	 *    inodes        (s_ninodes + 'inodes per block' - 1)/'inodes per block'
 22114	 *    unused        whatever is needed to fill out the current zone
 22115	 *    data zones    (s_zones - s_firstdatazone) << s_log_zone_size
 22116	 *
 22117	 * A super_block slot is free if s_dev == NO_DEV. 
 22118	 */
 22119	
 22120	EXTERN struct super_block {
 22121	  ino_t s_ninodes;              /* # usable inodes on the minor device */
 22122	  zone1_t  s_nzones;            /* total device size, including bit maps etc */
 22123	  short s_imap_blocks;          /* # of blocks used by inode bit map */
 22124	  short s_zmap_blocks;          /* # of blocks used by zone bit map */

_________________________ Page 931    File: servers/fs/super.h _________________________
 22125	  zone1_t s_firstdatazone;      /* number of first data zone */
 22126	  short s_log_zone_size;        /* log2 of blocks/zone */
 22127	  short s_pad;                  /* try to avoid compiler-dependent padding */
 22128	  off_t s_max_size;             /* maximum file size on this device */
 22129	  zone_t s_zones;               /* number of zones (replaces s_nzones in V2) */
 22130	  short s_magic;                /* magic number to recognize super-blocks */
 22131	
 22132	  /* The following items are valid on disk only for V3 and above */
 22133	
 22134	  /* The block size in bytes. Minimum MIN_BLOCK SIZE. SECTOR_SIZE
 22135	   * multiple. If V1 or V2 filesystem, this should be
 22136	   * initialised to STATIC_BLOCK_SIZE. Maximum MAX_BLOCK_SIZE.
 22137	   */
 22138	  short s_pad2;                 /* try to avoid compiler-dependent padding */
 22139	  unsigned short s_block_size;  /* block size in bytes. */
 22140	  char s_disk_version;          /* filesystem format sub-version */
 22141	
 22142	  /* The following items are only used when the super_block is in memory. */
 22143	  struct inode *s_isup;         /* inode for root dir of mounted file sys */
 22144	  struct inode *s_imount;       /* inode mounted on */
 22145	  unsigned s_inodes_per_block;  /* precalculated from magic number */
 22146	  dev_t s_dev;                  /* whose super block is this? */
 22147	  int s_rd_only;                /* set to 1 iff file sys mounted read only */
 22148	  int s_native;                 /* set to 1 iff not byte swapped file system */
 22149	  int s_version;                /* file system version, zero means bad magic */
 22150	  int s_ndzones;                /* # direct zones in an inode */
 22151	  int s_nindirs;                /* # indirect zones per indirect block */
 22152	  bit_t s_isearch;              /* inodes below this bit number are in use */
 22153	  bit_t s_zsearch;              /* all zones below this bit number are in use*/
 22154	} super_block[NR_SUPERS];
 22155	
 22156	#define NIL_SUPER (struct super_block *) 0
 22157	#define IMAP            0       /* operating on the inode bit map */
 22158	#define ZMAP            1       /* operating on the zone bit map */


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/table.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 22200	/* This file contains the table used to map system call numbers onto the
 22201	 * routines that perform them.
 22202	 */
 22203	
 22204	#define _TABLE
 22205	
 22206	#include "fs.h"
 22207	#include <minix/callnr.h>
 22208	#include <minix/com.h>
 22209	#include "buf.h"
 22210	#include "file.h"
 22211	#include "fproc.h"
 22212	#include "inode.h"
 22213	#include "lock.h"
 22214	#include "super.h"
 22215	
 22216	PUBLIC _PROTOTYPE (int (*call_vec[]), (void) ) = {
 22217	        no_sys,         /*  0 = unused  */
 22218	        do_exit,        /*  1 = exit    */
 22219	        do_fork,        /*  2 = fork    */

_________________________ Page 932    File: servers/fs/table.c _________________________
 22220	        do_read,        /*  3 = read    */
 22221	        do_write,       /*  4 = write   */
 22222	        do_open,        /*  5 = open    */
 22223	        do_close,       /*  6 = close   */
 22224	        no_sys,         /*  7 = wait    */
 22225	        do_creat,       /*  8 = creat   */
 22226	        do_link,        /*  9 = link    */
 22227	        do_unlink,      /* 10 = unlink  */
 22228	        no_sys,         /* 11 = waitpid */
 22229	        do_chdir,       /* 12 = chdir   */
 22230	        no_sys,         /* 13 = time    */
 22231	        do_mknod,       /* 14 = mknod   */
 22232	        do_chmod,       /* 15 = chmod   */
 22233	        do_chown,       /* 16 = chown   */
 22234	        no_sys,         /* 17 = break   */
 22235	        do_stat,        /* 18 = stat    */
 22236	        do_lseek,       /* 19 = lseek   */
 22237	        no_sys,         /* 20 = getpid  */
 22238	        do_mount,       /* 21 = mount   */
 22239	        do_umount,      /* 22 = umount  */
 22240	        do_set,         /* 23 = setuid  */
 22241	        no_sys,         /* 24 = getuid  */
 22242	        do_stime,       /* 25 = stime   */
 22243	        no_sys,         /* 26 = ptrace  */
 22244	        no_sys,         /* 27 = alarm   */
 22245	        do_fstat,       /* 28 = fstat   */
 22246	        no_sys,         /* 29 = pause   */
 22247	        do_utime,       /* 30 = utime   */
 22248	        no_sys,         /* 31 = (stty)  */
 22249	        no_sys,         /* 32 = (gtty)  */
 22250	        do_access,      /* 33 = access  */
 22251	        no_sys,         /* 34 = (nice)  */
 22252	        no_sys,         /* 35 = (ftime) */
 22253	        do_sync,        /* 36 = sync    */
 22254	        no_sys,         /* 37 = kill    */
 22255	        do_rename,      /* 38 = rename  */
 22256	        do_mkdir,       /* 39 = mkdir   */
 22257	        do_unlink,      /* 40 = rmdir   */
 22258	        do_dup,         /* 41 = dup     */
 22259	        do_pipe,        /* 42 = pipe    */
 22260	        no_sys,         /* 43 = times   */
 22261	        no_sys,         /* 44 = (prof)  */
 22262	        no_sys,         /* 45 = unused  */
 22263	        do_set,         /* 46 = setgid  */
 22264	        no_sys,         /* 47 = getgid  */
 22265	        no_sys,         /* 48 = (signal)*/
 22266	        no_sys,         /* 49 = unused  */
 22267	        no_sys,         /* 50 = unused  */
 22268	        no_sys,         /* 51 = (acct)  */
 22269	        no_sys,         /* 52 = (phys)  */
 22270	        no_sys,         /* 53 = (lock)  */
 22271	        do_ioctl,       /* 54 = ioctl   */
 22272	        do_fcntl,       /* 55 = fcntl   */
 22273	        no_sys,         /* 56 = (mpx)   */
 22274	        no_sys,         /* 57 = unused  */
 22275	        no_sys,         /* 58 = unused  */
 22276	        do_exec,        /* 59 = execve  */
 22277	        do_umask,       /* 60 = umask   */
 22278	        do_chroot,      /* 61 = chroot  */
 22279	        do_setsid,      /* 62 = setsid  */

_________________________ Page 933    File: servers/fs/table.c _________________________
 22280	        no_sys,         /* 63 = getpgrp */
 22281	
 22282	        no_sys,         /* 64 = KSIG: signals originating in the kernel */
 22283	        do_unpause,     /* 65 = UNPAUSE */
 22284	        no_sys,         /* 66 = unused  */
 22285	        do_revive,      /* 67 = REVIVE  */
 22286	        no_sys,         /* 68 = TASK_REPLY      */
 22287	        no_sys,         /* 69 = unused */
 22288	        no_sys,         /* 70 = unused */
 22289	        no_sys,         /* 71 = si */
 22290	        no_sys,         /* 72 = sigsuspend */
 22291	        no_sys,         /* 73 = sigpending */
 22292	        no_sys,         /* 74 = sigprocmask */
 22293	        no_sys,         /* 75 = sigreturn */
 22294	        do_reboot,      /* 76 = reboot */
 22295	        do_svrctl,      /* 77 = svrctl */
 22296	
 22297	        no_sys,         /* 78 = unused */
 22298	        do_getsysinfo,  /* 79 = getsysinfo */
 22299	        no_sys,         /* 80 = unused */
 22300	        do_devctl,      /* 81 = devctl */
 22301	        do_fstatfs,     /* 82 = fstatfs */
 22302	        no_sys,         /* 83 = memalloc */
 22303	        no_sys,         /* 84 = memfree */
 22304	        do_select,      /* 85 = select */
 22305	        do_fchdir,      /* 86 = fchdir */
 22306	        do_fsync,       /* 87 = fsync */
 22307	        no_sys,         /* 88 = getpriority */
 22308	        no_sys,         /* 89 = setpriority */
 22309	        no_sys,         /* 90 = gettimeofday */
 22310	};
 22311	/* This should not fail with "array size is negative": */
 22312	extern int dummy[sizeof(call_vec) == NCALLS * sizeof(call_vec[0]) ? 1 : -1];
 22313	


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/cache.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 22400	/* The file system maintains a buffer cache to reduce the number of disk
 22401	 * accesses needed.  Whenever a read or write to the disk is done, a check is
 22402	 * first made to see if the block is in the cache.  This file manages the
 22403	 * cache.
 22404	 *
 22405	 * The entry points into this file are:
 22406	 *   get_block:   request to fetch a block for reading or writing from cache
 22407	 *   put_block:   return a block previously requested with get_block
 22408	 *   alloc_zone:  allocate a new zone (to increase the length of a file)
 22409	 *   free_zone:   release a zone (when a file is removed)
 22410	 *   rw_block:    read or write a block from the disk itself
 22411	 *   invalidate:  remove all the cache blocks on some device
 22412	 */
 22413	
 22414	#include "fs.h"
 22415	#include <minix/com.h>
 22416	#include "buf.h"
 22417	#include "file.h"
 22418	#include "fproc.h"
 22419	#include "super.h"

_________________________ Page 934    File: servers/fs/cache.c _________________________
 22420	
 22421	FORWARD _PROTOTYPE( void rm_lru, (struct buf *bp) );
 22422	
 22423	/*===========================================================================*
 22424	 *                              get_block                                    *
 22425	 *===========================================================================*/
 22426	PUBLIC struct buf *get_block(dev, block, only_search)
 22427	register dev_t dev;             /* on which device is the block? */
 22428	register block_t block;         /* which block is wanted? */
 22429	int only_search;                /* if NO_READ, don't read, else act normal */
 22430	{
 22431	/* Check to see if the requested block is in the block cache.  If so, return
 22432	 * a pointer to it.  If not, evict some other block and fetch it (unless
 22433	 * 'only_search' is 1).  All the blocks in the cache that are not in use
 22434	 * are linked together in a chain, with 'front' pointing to the least recently
 22435	 * used block and 'rear' to the most recently used block.  If 'only_search' is
 22436	 * 1, the block being requested will be overwritten in its entirety, so it is
 22437	 * only necessary to see if it is in the cache; if it is not, any free buffer
 22438	 * will do.  It is not necessary to actually read the block in from disk.
 22439	 * If 'only_search' is PREFETCH, the block need not be read from the disk,
 22440	 * and the device is not to be marked on the block, so callers can tell if
 22441	 * the block returned is valid.
 22442	 * In addition to the LRU chain, there is also a hash chain to link together
 22443	 * blocks whose block numbers end with the same bit strings, for fast lookup.
 22444	 */
 22445	
 22446	  int b;
 22447	  register struct buf *bp, *prev_ptr;
 22448	
 22449	  /* Search the hash chain for (dev, block). Do_read() can use 
 22450	   * get_block(NO_DEV ...) to get an unnamed block to fill with zeros when
 22451	   * someone wants to read from a hole in a file, in which case this search
 22452	   * is skipped
 22453	   */
 22454	  if (dev != NO_DEV) {
 22455	        b = (int) block & HASH_MASK;
 22456	        bp = buf_hash[b];
 22457	        while (bp != NIL_BUF) {
 22458	                if (bp->b_blocknr == block && bp->b_dev == dev) {
 22459	                        /* Block needed has been found. */
 22460	                        if (bp->b_count == 0) rm_lru(bp);
 22461	                        bp->b_count++;  /* record that block is in use */
 22462	
 22463	                        return(bp);
 22464	                } else {
 22465	                        /* This block is not the one sought. */
 22466	                        bp = bp->b_hash; /* move to next block on hash chain */
 22467	                }
 22468	        }
 22469	  }
 22470	
 22471	  /* Desired block is not on available chain.  Take oldest block ('front'). */
 22472	  if ((bp = front) == NIL_BUF) panic(__FILE__,"all buffers in use", NR_BUFS);
 22473	  rm_lru(bp);
 22474	
 22475	  /* Remove the block that was just taken from its hash chain. */
 22476	  b = (int) bp->b_blocknr & HASH_MASK;
 22477	  prev_ptr = buf_hash[b];
 22478	  if (prev_ptr == bp) {
 22479	        buf_hash[b] = bp->b_hash;

_________________________ Page 935    File: servers/fs/cache.c _________________________
 22480	  } else {
 22481	        /* The block just taken is not on the front of its hash chain. */
 22482	        while (prev_ptr->b_hash != NIL_BUF)
 22483	                if (prev_ptr->b_hash == bp) {
 22484	                        prev_ptr->b_hash = bp->b_hash;  /* found it */
 22485	                        break;
 22486	                } else {
 22487	                        prev_ptr = prev_ptr->b_hash;    /* keep looking */
 22488	                }
 22489	  }
 22490	
 22491	  /* If the block taken is dirty, make it clean by writing it to the disk.
 22492	   * Avoid hysteresis by flushing all other dirty blocks for the same device.
 22493	   */
 22494	  if (bp->b_dev != NO_DEV) {
 22495	        if (bp->b_dirt == DIRTY) flushall(bp->b_dev);
 22496	  }
 22497	
 22498	  /* Fill in block's parameters and add it to the hash chain where it goes. */
 22499	  bp->b_dev = dev;              /* fill in device number */
 22500	  bp->b_blocknr = block;        /* fill in block number */
 22501	  bp->b_count++;                /* record that block is being used */
 22502	  b = (int) bp->b_blocknr & HASH_MASK;
 22503	  bp->b_hash = buf_hash[b];
 22504	  buf_hash[b] = bp;             /* add to hash list */
 22505	
 22506	  /* Go get the requested block unless searching or prefetching. */
 22507	  if (dev != NO_DEV) {
 22508	        if (only_search == PREFETCH) bp->b_dev = NO_DEV;
 22509	        else
 22510	        if (only_search == NORMAL) {
 22511	                rw_block(bp, READING);
 22512	        }
 22513	  }
 22514	  return(bp);                   /* return the newly acquired block */
 22515	}
	
 22517	/*===========================================================================*
 22518	 *                              put_block                                    *
 22519	 *===========================================================================*/
 22520	PUBLIC void put_block(bp, block_type)
 22521	register struct buf *bp;        /* pointer to the buffer to be released */
 22522	int block_type;                 /* INODE_BLOCK, DIRECTORY_BLOCK, or whatever */
 22523	{
 22524	/* Return a block to the list of available blocks.   Depending on 'block_type'
 22525	 * it may be put on the front or rear of the LRU chain.  Blocks that are
 22526	 * expected to be needed again shortly (e.g., partially full data blocks)
 22527	 * go on the rear; blocks that are unlikely to be needed again shortly
 22528	 * (e.g., full data blocks) go on the front.  Blocks whose loss can hurt
 22529	 * the integrity of the file system (e.g., inode blocks) are written to
 22530	 * disk immediately if they are dirty.
 22531	 */
 22532	  if (bp == NIL_BUF) return;    /* it is easier to check here than in caller */
 22533	
 22534	  bp->b_count--;                /* there is one use fewer now */
 22535	  if (bp->b_count != 0) return; /* block is still in use */
 22536	
 22537	  bufs_in_use--;                /* one fewer block buffers in use */
 22538	
 22539	  /* Put this block back on the LRU chain.  If the ONE_SHOT bit is set in

_________________________ Page 936    File: servers/fs/cache.c _________________________
 22540	   * 'block_type', the block is not likely to be needed again shortly, so put
 22541	   * it on the front of the LRU chain where it will be the first one to be
 22542	   * taken when a free buffer is needed later.
 22543	   */
 22544	  if (bp->b_dev == DEV_RAM || block_type & ONE_SHOT) {
 22545	        /* Block probably won't be needed quickly. Put it on front of chain.
 22546	         * It will be the next block to be evicted from the cache.
 22547	         */
 22548	        bp->b_prev = NIL_BUF;
 22549	        bp->b_next = front;
 22550	        if (front == NIL_BUF)
 22551	                rear = bp;      /* LRU chain was empty */
 22552	        else
 22553	                front->b_prev = bp;
 22554	        front = bp;
 22555	  } else {
 22556	        /* Block probably will be needed quickly.  Put it on rear of chain.
 22557	         * It will not be evicted from the cache for a long time.
 22558	         */
 22559	        bp->b_prev = rear;
 22560	        bp->b_next = NIL_BUF;
 22561	        if (rear == NIL_BUF)
 22562	                front = bp;
 22563	        else
 22564	                rear->b_next = bp;
 22565	        rear = bp;
 22566	  }
 22567	
 22568	  /* Some blocks are so important (e.g., inodes, indirect blocks) that they
 22569	   * should be written to the disk immediately to avoid messing up the file
 22570	   * system in the event of a crash.
 22571	   */
 22572	  if ((block_type & WRITE_IMMED) && bp->b_dirt==DIRTY && bp->b_
 22573	                rw_block(bp, WRITING);
 22574	  } 
 22575	}
	
 22577	/*===========================================================================*
 22578	 *                              alloc_zone                                   *
 22579	 *===========================================================================*/
 22580	PUBLIC zone_t alloc_zone(dev, z)
 22581	dev_t dev;                      /* device where zone wanted */
 22582	zone_t z;                       /* try to allocate new zone near this one */
 22583	{
 22584	/* Allocate a new zone on the indicated device and return its number. */
 22585	
 22586	  int major, minor;
 22587	  bit_t b, bit;
 22588	  struct super_block *sp;
 22589	
 22590	  /* Note that the routine alloc_bit() returns 1 for the lowest possible
 22591	   * zone, which corresponds to sp->s_firstdatazone.  To convert a value
 22592	   * between the bit number, 'b', used by alloc_bit() and the zone number, 'z',
 22593	   * stored in the inode, use the formula:
 22594	   *     z = b + sp->s_firstdatazone - 1
 22595	   * Alloc_bit() never returns 0, since this is used for NO_BIT (failure).
 22596	   */
 22597	  sp = get_super(dev);
 22598	
 22599	  /* If z is 0, skip initial part of the map known to be fully in use. */

_________________________ Page 937    File: servers/fs/cache.c _________________________
 22600	  if (z == sp->s_firstdatazone) {
 22601	        bit = sp->s_zsearch;
 22602	  } else {
 22603	        bit = (bit_t) z - (sp->s_firstdatazone - 1);
 22604	  }
 22605	  b = alloc_bit(sp, ZMAP, bit);
 22606	  if (b == NO_BIT) {
 22607	        err_code = ENOSPC;
 22608	        major = (int) (sp->s_dev >> MAJOR) & BYTE;
 22609	        minor = (int) (sp->s_dev >> MINOR) & BYTE;
 22610	        printf("No space on %sdevice %d/%d\n",
 22611	                sp->s_dev == root_dev ? "root " : "", major, minor);
 22612	        return(NO_ZONE);
 22613	  }
 22614	  if (z == sp->s_firstdatazone) sp->s_zsearch = b;      /* for next time */
 22615	  return(sp->s_firstdatazone - 1 + (zone_t) b);
 22616	}
	
 22618	/*===========================================================================*
 22619	 *                              free_zone                                    *
 22620	 *===========================================================================*/
 22621	PUBLIC void free_zone(dev, numb)
 22622	dev_t dev;                              /* device where zone located */
 22623	zone_t numb;                            /* zone to be returned */
 22624	{
 22625	/* Return a zone. */
 22626	
 22627	  register struct super_block *sp;
 22628	  bit_t bit;
 22629	
 22630	  /* Locate the appropriate super_block and return bit. */
 22631	  sp = get_super(dev);
 22632	  if (numb < sp->s_firstdatazone || numb >= sp->s_zones) return;
 22633	  bit = (bit_t) (numb - (sp->s_firstdatazone - 1));
 22634	  free_bit(sp, ZMAP, bit);
 22635	  if (bit < sp->s_zsearch) sp->s_zsearch = bit;
 22636	}
	
 22638	/*===========================================================================*
 22639	 *                              rw_block                                     *
 22640	 *===========================================================================*/
 22641	PUBLIC void rw_block(bp, rw_flag)
 22642	register struct buf *bp;        /* buffer pointer */
 22643	int rw_flag;                    /* READING or WRITING */
 22644	{
 22645	/* Read or write a disk block. This is the only routine in which actual disk
 22646	 * I/O is invoked. If an error occurs, a message is printed here, but the error
 22647	 * is not reported to the caller.  If the error occurred while purging a block
 22648	 * from the cache, it is not clear what the caller could do about it anyway.
 22649	 */
 22650	
 22651	  int r, op;
 22652	  off_t pos;
 22653	  dev_t dev;
 22654	  int block_size;
 22655	
 22656	  block_size = get_block_size(bp->b_dev);
 22657	
 22658	  if ( (dev = bp->b_dev) != NO_DEV) {
 22659	        pos = (off_t) bp->b_blocknr * block_size;

_________________________ Page 938    File: servers/fs/cache.c _________________________
 22660	        op = (rw_flag == READING ? DEV_READ : DEV_WRITE);
 22661	        r = dev_io(op, dev, FS_PROC_NR, bp->b_data, pos, block_size, 0);
 22662	        if (r != block_size) {
 22663	            if (r >= 0) r = END_OF_FILE;
 22664	            if (r != END_OF_FILE)
 22665	              printf("Unrecoverable disk error on device %d/%d, block %ld\n",
 22666	                        (dev>>MAJOR)&BYTE, (dev>>MINOR)&BYTE, bp->b_blocknr
 22667	                bp->b_dev = NO_DEV;     /* invalidate block */
 22668	
 22669	                /* Report read errors to interested parties. */
 22670	                if (rw_flag == READING) rdwt_err = r;
 22671	        }
 22672	  }
 22673	
 22674	  bp->b_dirt = CLEAN;
 22675	}
	
 22677	/*===========================================================================*
 22678	 *                              invalidate                                   *
 22679	 *===========================================================================*/
 22680	PUBLIC void invalidate(device)
 22681	dev_t device;                   /* device whose blocks are to be purged */
 22682	{
 22683	/* Remove all the blocks belonging to some device from the cache. */
 22684	
 22685	  register struct buf *bp;
 22686	
 22687	  for (bp = &buf[0]; bp < &buf[NR_BUFS]; bp++)
 22688	        if (bp->b_dev == device) bp->b_dev = NO_DEV;
 22689	}
	
 22691	/*===========================================================================*
 22692	 *                              flushall                                     *
 22693	 *===========================================================================*/
 22694	PUBLIC void flushall(dev)
 22695	dev_t dev;                      /* device to flush */
 22696	{
 22697	/* Flush all dirty blocks for one device. */
 22698	
 22699	  register struct buf *bp;
 22700	  static struct buf *dirty[NR_BUFS];    /* static so it isn't on stack */
 22701	  int ndirty;
 22702	
 22703	  for (bp = &buf[0], ndirty = 0; bp < &buf[NR_BUFS]; bp++)
 22704	        if (bp->b_dirt == DIRTY && bp->b_dev == dev) dirty[ndirty++] = bp;
 22705	  rw_scattered(dev, dirty, ndirty, WRITING);
 22706	}
	
 22708	/*===========================================================================*
 22709	 *                              rw_scattered                                 *
 22710	 *===========================================================================*/
 22711	PUBLIC void rw_scattered(dev, bufq, bufqsize, rw_flag)
 22712	dev_t dev;                      /* major-minor device number */
 22713	struct buf **bufq;              /* pointer to array of buffers */
 22714	int bufqsize;                   /* number of buffers */
 22715	int rw_flag;                    /* READING or WRITING */
 22716	{
 22717	/* Read or write scattered data from a device. */
 22718	
 22719	  register struct buf *bp;

_________________________ Page 939    File: servers/fs/cache.c _________________________
 22720	  int gap;
 22721	  register int i;
 22722	  register iovec_t *iop;
 22723	  static iovec_t iovec[NR_IOREQS];  /* static so it isn't on stack */
 22724	  int j, r;
 22725	  int block_size;
 22726	
 22727	  block_size = get_block_size(dev);
 22728	
 22729	  /* (Shell) sort buffers on b_blocknr. */
 22730	  gap = 1;
 22731	  do
 22732	        gap = 3 * gap + 1;
 22733	  while (gap <= bufqsize);
 22734	  while (gap != 1) {
 22735	        gap /= 3;
 22736	        for (j = gap; j < bufqsize; j++) {
 22737	                for (i = j - gap;
 22738	                     i >= 0 && bufq[i]->b_blocknr > bufq[i + gap]->b_blockn
 22739	                     i -= gap) {
 22740	                        bp = bufq[i];
 22741	                        bufq[i] = bufq[i + gap];
 22742	                        bufq[i + gap] = bp;
 22743	                }
 22744	        }
 22745	  }
 22746	
 22747	  /* Set up I/O vector and do I/O.  The result of dev_io is OK if everything
 22748	   * went fine, otherwise the error code for the first failed transfer.
 22749	   */  
 22750	  while (bufqsize > 0) {
 22751	        for (j = 0, iop = iovec; j < NR_IOREQS && j < bufqsize; j++, 
 22752	                bp = bufq[j];
 22753	                if (bp->b_blocknr != bufq[0]->b_blocknr + j) break;
 22754	                iop->iov_addr = (vir_bytes) bp->b_data;
 22755	                iop->iov_size = block_size;
 22756	        }
 22757	        r = dev_io(rw_flag == WRITING ? DEV_SCATTER : DEV_GATHER,
 22758	                dev, FS_PROC_NR, iovec,
 22759	                (off_t) bufq[0]->b_blocknr * block_size, j, 0);
 22760	
 22761	        /* Harvest the results.  Dev_io reports the first error it may have
 22762	         * encountered, but we only care if it's the first block that failed.
 22763	         */
 22764	        for (i = 0, iop = iovec; i < j; i++, iop++) {
 22765	                bp = bufq[i];
 22766	                if (iop->iov_size != 0) {
 22767	                        /* Transfer failed. An error? Do we care? */
 22768	                        if (r != OK && i == 0) {
 22769	                                printf(
 22770	                                "fs: I/O error on device %d/%d, block %lu\n",
 22771	                                        (dev>>MAJOR)&BYTE, (dev>>MINOR)&BYT
 22772	                                        bp->b_blocknr);
 22773	                                bp->b_dev = NO_DEV;     /* invalidate block */
 22774	                        }
 22775	                        break;
 22776	                }
 22777	                if (rw_flag == READING) {
 22778	                        bp->b_dev = dev;        /* validate block */
 22779	                        put_block(bp, PARTIAL_DATA_BLOCK);

_________________________ Page 940    File: servers/fs/cache.c _________________________
 22780	                } else {
 22781	                        bp->b_dirt = CLEAN;
 22782	                }
 22783	        }
 22784	        bufq += i;
 22785	        bufqsize -= i;
 22786	        if (rw_flag == READING) {
 22787	                /* Don't bother reading more than the device is willing to
 22788	                 * give at this time.  Don't forget to release those extras.
 22789	                 */
 22790	                while (bufqsize > 0) {
 22791	                        put_block(*bufq++, PARTIAL_DATA_BLOCK);
 22792	                        bufqsize--;
 22793	                }
 22794	        }
 22795	        if (rw_flag == WRITING && i == 0) {
 22796	                /* We're not making progress, this means we might keep
 22797	                 * looping. Buffers remain dirty if un-written. Buffers are
 22798	                 * lost if invalidate()d or LRU-removed while dirty. This
 22799	                 * is better than keeping unwritable blocks around forever..
 22800	                 */
 22801	                break;
 22802	        }
 22803	  }
 22804	}
	
 22806	/*===========================================================================*
 22807	 *                              rm_lru                                       *
 22808	 *===========================================================================*/
 22809	PRIVATE void rm_lru(bp)
 22810	struct buf *bp;
 22811	{
 22812	/* Remove a block from its LRU chain. */
 22813	  struct buf *next_ptr, *prev_ptr;
 22814	
 22815	  bufs_in_use++;
 22816	  next_ptr = bp->b_next;        /* successor on LRU chain */
 22817	  prev_ptr = bp->b_prev;        /* predecessor on LRU chain */
 22818	  if (prev_ptr != NIL_BUF)
 22819	        prev_ptr->b_next = next_ptr;
 22820	  else
 22821	        front = next_ptr;       /* this block was at front of chain */
 22822	
 22823	  if (next_ptr != NIL_BUF)
 22824	        next_ptr->b_prev = prev_ptr;
 22825	  else
 22826	        rear = prev_ptr;        /* this block was at rear of chain */
 22827	}



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/inode.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 22900	/* This file manages the inode table.  There are procedures to allocate and
 22901	 * deallocate inodes, acquire, erase, and release them, and read and write
 22902	 * them from the disk.
 22903	 *
 22904	 * The entry points into this file are

_________________________ Page 941    File: servers/fs/inode.c _________________________
 22905	 *   get_inode:    search inode table for a given inode; if not there,
 22906	 *                 read it
 22907	 *   put_inode:    indicate that an inode is no longer needed in memory
 22908	 *   alloc_inode:  allocate a new, unused inode
 22909	 *   wipe_inode:   erase some fields of a newly allocated inode
 22910	 *   free_inode:   mark an inode as available for a new file
 22911	 *   update_times: update atime, ctime, and mtime
 22912	 *   rw_inode:     read a disk block and extract an inode, or corresp. write
 22913	 *   old_icopy:    copy to/from in-core inode struct and disk inode (V1.x)
 22914	 *   new_icopy:    copy to/from in-core inode struct and disk inode (V2.x)
 22915	 *   dup_inode:    indicate that someone else is using an inode table entry
 22916	 */
 22917	
 22918	#include "fs.h"
 22919	#include "buf.h"
 22920	#include "file.h"
 22921	#include "fproc.h"
 22922	#include "inode.h"
 22923	#include "super.h"
 22924	
 22925	FORWARD _PROTOTYPE( void old_icopy, (struct inode *rip, d1_inode *dip,
 22926	                                                int direction, int norm));
 22927	FORWARD _PROTOTYPE( void new_icopy, (struct inode *rip, d2_inode *dip,
 22928	                                                int direction, int norm));
 22929	
 22930	/*===========================================================================*
 22931	 *                              get_inode                                    *
 22932	 *===========================================================================*/
 22933	PUBLIC struct inode *get_inode(dev, numb)
 22934	dev_t dev;                      /* device on which inode resides */
 22935	int numb;                       /* inode number (ANSI: may not be unshort) */
 22936	{
 22937	/* Find a slot in the inode table, load the specified inode into it, and
 22938	 * return a pointer to the slot.  If 'dev' == NO_DEV, just return a free slot.
 22939	 */
 22940	
 22941	  register struct inode *rip, *xp;
 22942	
 22943	  /* Search the inode table both for (dev, numb) and a free slot. */
 22944	  xp = NIL_INODE;
 22945	  for (rip = &inode[0]; rip < &inode[NR_INODES]; rip++) {
 22946	        if (rip->i_count > 0) { /* only check used slots for (dev, numb) */
 22947	                if (rip->i_dev == dev && rip->i_num == numb) {
 22948	                        /* This is the inode that we are looking for. */
 22949	                        rip->i_count++;
 22950	                        return(rip);    /* (dev, numb) found */
 22951	                }
 22952	        } else {
 22953	                xp = rip;       /* remember this free slot for later */
 22954	        }
 22955	  }
 22956	
 22957	  /* Inode we want is not currently in use.  Did we find a free slot? */
 22958	  if (xp == NIL_INODE) {        /* inode table completely full */
 22959	        err_code = ENFILE;
 22960	        return(NIL_INODE);
 22961	  }
 22962	
 22963	  /* A free inode slot has been located.  Load the inode into it. */
 22964	  xp->i_dev = dev;

_________________________ Page 942    File: servers/fs/inode.c _________________________
 22965	  xp->i_num = numb;
 22966	  xp->i_count = 1;
 22967	  if (dev != NO_DEV) rw_inode(xp, READING);     /* get inode from disk */
 22968	  xp->i_update = 0;             /* all the times are initially up-to-date */
 22969	
 22970	  return(xp);
 22971	}
	
 22973	/*===========================================================================*
 22974	 *                              put_inode                                    *
 22975	 *===========================================================================*/
 22976	PUBLIC void put_inode(rip)
 22977	register struct inode *rip;     /* pointer to inode to be released */
 22978	{
 22979	/* The caller is no longer using this inode.  If no one else is using it either
 22980	 * write it back to the disk immediately.  If it has no links, truncate it and
 22981	 * return it to the pool of available inodes.
 22982	 */
 22983	
 22984	  if (rip == NIL_INODE) return; /* checking here is easier than in caller */
 22985	  if (--rip->i_count == 0) {    /* i_count == 0 means no one is using it now */
 22986	        if (rip->i_nlinks == 0) {
 22987	                /* i_nlinks == 0 means free the inode. */
 22988	                truncate(rip);  /* return all the disk blocks */
 22989	                rip->i_mode = I_NOT_ALLOC;      /* clear I_TYPE field */
 22990	                rip->i_dirt = DIRTY;
 22991	                free_inode(rip->i_dev, rip->i_num);
 22992	        } else {
 22993	                if (rip->i_pipe == I_PIPE) truncate(rip);
 22994	        }
 22995	        rip->i_pipe = NO_PIPE;  /* should always be cleared */
 22996	        if (rip->i_dirt == DIRTY) rw_inode(rip, WRITING);
 22997	  }
 22998	}
	
 23000	/*===========================================================================*
 23001	 *                              alloc_inode                                  *
 23002	 *===========================================================================*/
 23003	PUBLIC struct inode *alloc_inode(dev_t dev, mode_t bits)
 23004	{
 23005	/* Allocate a free inode on 'dev', and return a pointer to it. */
 23006	
 23007	  register struct inode *rip;
 23008	  register struct super_block *sp;
 23009	  int major, minor, inumb;
 23010	  bit_t b;
 23011	
 23012	  sp = get_super(dev);  /* get pointer to super_block */
 23013	  if (sp->s_rd_only) {  /* can't allocate an inode on a read only device. */
 23014	        err_code = EROFS;
 23015	        return(NIL_INODE);
 23016	  }
 23017	
 23018	  /* Acquire an inode from the bit map. */
 23019	  b = alloc_bit(sp, IMAP, sp->s_isearch);
 23020	  if (b == NO_BIT) {
 23021	        err_code = ENFILE;
 23022	        major = (int) (sp->s_dev >> MAJOR) & BYTE;
 23023	        minor = (int) (sp->s_dev >> MINOR) & BYTE;
 23024	        printf("Out of i-nodes on %sdevice %d/%d\n",

_________________________ Page 943    File: servers/fs/inode.c _________________________
 23025	                sp->s_dev == root_dev ? "root " : "", major, minor);
 23026	        return(NIL_INODE);
 23027	  }
 23028	  sp->s_isearch = b;            /* next time start here */
 23029	  inumb = (int) b;              /* be careful not to pass unshort as param */
 23030	
 23031	  /* Try to acquire a slot in the inode table. */
 23032	  if ((rip = get_inode(NO_DEV, inumb)) == NIL_INODE) {
 23033	        /* No inode table slots available.  Free the inode just allocated. */
 23034	        free_bit(sp, IMAP, b);
 23035	  } else {
 23036	        /* An inode slot is available. Put the inode just allocated into it. */
 23037	        rip->i_mode = bits;             /* set up RWX bits */
 23038	        rip->i_nlinks = 0;              /* initial no links */
 23039	        rip->i_uid = fp->fp_effuid;     /* file's uid is owner's */
 23040	        rip->i_gid = fp->fp_effgid;     /* ditto group id */
 23041	        rip->i_dev = dev;               /* mark which device it is on */
 23042	        rip->i_ndzones = sp->s_ndzones; /* number of direct zones */
 23043	        rip->i_nindirs = sp->s_nindirs; /* number of indirect zones per blk*/
 23044	        rip->i_sp = sp;                 /* pointer to super block */
 23045	
 23046	        /* Fields not cleared already are cleared in wipe_inode().  They have
 23047	         * been put there because truncate() needs to clear the same fields if
 23048	         * the file happens to be open while being truncated.  It saves space
 23049	         * not to repeat the code twice.
 23050	         */
 23051	        wipe_inode(rip);
 23052	  }
 23053	
 23054	  return(rip);
 23055	}
	
 23057	/*===========================================================================*
 23058	 *                              wipe_inode                                   *
 23059	 *===========================================================================*/
 23060	PUBLIC void wipe_inode(rip)
 23061	register struct inode *rip;     /* the inode to be erased */
 23062	{
 23063	/* Erase some fields in the inode.  This function is called from alloc_inode()
 23064	 * when a new inode is to be allocated, and from truncate(), when an existing
 23065	 * inode is to be truncated.
 23066	 */
 23067	
 23068	  register int i;
 23069	
 23070	  rip->i_size = 0;
 23071	  rip->i_update = ATIME | CTIME | MTIME;        /* update all times later */
 23072	  rip->i_dirt = DIRTY;
 23073	  for (i = 0; i < V2_NR_TZONES; i++) rip->i_zone[i] = NO_ZONE;
 23074	}
	
 23076	/*===========================================================================*
 23077	 *                              free_inode                                   *
 23078	 *===========================================================================*/
 23079	PUBLIC void free_inode(dev, inumb)
 23080	dev_t dev;                      /* on which device is the inode */
 23081	ino_t inumb;                    /* number of inode to be freed */
 23082	{
 23083	/* Return an inode to the pool of unallocated inodes. */
 23084	

_________________________ Page 944    File: servers/fs/inode.c _________________________
 23085	  register struct super_block *sp;
 23086	  bit_t b;
 23087	
 23088	  /* Locate the appropriate super_block. */
 23089	  sp = get_super(dev);
 23090	  if (inumb <= 0 || inumb > sp->s_ninodes) return;
 23091	  b = inumb;
 23092	  free_bit(sp, IMAP, b);
 23093	  if (b < sp->s_isearch) sp->s_isearch = b;
 23094	}
	
 23096	/*===========================================================================*
 23097	 *                              update_times                                 *
 23098	 *===========================================================================*/
 23099	PUBLIC void update_times(rip)
 23100	register struct inode *rip;     /* pointer to inode to be read/written */
 23101	{
 23102	/* Various system calls are required by the standard to update atime, ctime,
 23103	 * or mtime.  Since updating a time requires sending a message to the clock
 23104	 * task--an expensive business--the times are marked for update by setting
 23105	 * bits in i_update.  When a stat, fstat, or sync is done, or an inode is 
 23106	 * released, update_times() may be called to actually fill in the times.
 23107	 */
 23108	
 23109	  time_t cur_time;
 23110	  struct super_block *sp;
 23111	
 23112	  sp = rip->i_sp;               /* get pointer to super block. */
 23113	  if (sp->s_rd_only) return;    /* no updates for read-only file systems */
 23114	
 23115	  cur_time = clock_time();
 23116	  if (rip->i_update & ATIME) rip->i_atime = cur_time;
 23117	  if (rip->i_update & CTIME) rip->i_ctime = cur_time;
 23118	  if (rip->i_update & MTIME) rip->i_mtime = cur_time;
 23119	  rip->i_update = 0;            /* they are all up-to-date now */
 23120	}
	
 23122	/*===========================================================================*
 23123	 *                              rw_inode                                     *
 23124	 *===========================================================================*/
 23125	PUBLIC void rw_inode(rip, rw_flag)
 23126	register struct inode *rip;     /* pointer to inode to be read/written */
 23127	int rw_flag;                    /* READING or WRITING */
 23128	{
 23129	/* An entry in the inode table is to be copied to or from the disk. */
 23130	
 23131	  register struct buf *bp;
 23132	  register struct super_block *sp;
 23133	  d1_inode *dip;
 23134	  d2_inode *dip2;
 23135	  block_t b, offset;
 23136	
 23137	  /* Get the block where the inode resides. */
 23138	  sp = get_super(rip->i_dev);   /* get pointer to super block */
 23139	  rip->i_sp = sp;               /* inode must contain super block pointer */
 23140	  offset = sp->s_imap_blocks + sp->s_zmap_blocks + 2;
 23141	  b = (block_t) (rip->i_num - 1)/sp->s_inodes_per_block + offset;
 23142	  bp = get_block(rip->i_dev, b, NORMAL);
 23143	  dip  = bp->b_v1_ino + (rip->i_num - 1) % V1_INODES_PER_BLOCK;
 23144	  dip2 = bp->b_v2_ino + (rip->i_num - 1) %

_________________________ Page 945    File: servers/fs/inode.c _________________________
 23145	         V2_INODES_PER_BLOCK(sp->s_block_size);
 23146	
 23147	  /* Do the read or write. */
 23148	  if (rw_flag == WRITING) {
 23149	        if (rip->i_update) update_times(rip);   /* times need updating */
 23150	        if (sp->s_rd_only == FALSE) bp->b_dirt = DIRTY;
 23151	  }
 23152	
 23153	  /* Copy the inode from the disk block to the in-core table or vice versa.
 23154	   * If the fourth parameter below is FALSE, the bytes are swapped.
 23155	   */
 23156	  if (sp->s_version == V1)
 23157	        old_icopy(rip, dip,  rw_flag, sp->s_native);
 23158	  else
 23159	        new_icopy(rip, dip2, rw_flag, sp->s_native);
 23160	  
 23161	  put_block(bp, INODE_BLOCK);
 23162	  rip->i_dirt = CLEAN;
 23163	}
	
 23165	/*===========================================================================*
 23166	 *                              old_icopy                                    *
 23167	 *===========================================================================*/
 23168	PRIVATE void old_icopy(rip, dip, direction, norm)
 23169	register struct inode *rip;     /* pointer to the in-core inode struct */
 23170	register d1_inode *dip;         /* pointer to the d1_inode inode struct */
 23171	int direction;                  /* READING (from disk) or WRITING (to disk) */
 23172	int norm;                       /* TRUE = do not swap bytes; FALSE = swap */
 23173	
 23174	{
 23175	/* The V1.x IBM disk, the V1.x 68000 disk, and the V2 disk (same for IBM and
 23176	 * 68000) all have different inode layouts.  When an inode is read or written
 23177	 * this routine handles the conversions so that the information in the inode
 23178	 * table is independent of the disk structure from which the inode came.
 23179	 * The old_icopy routine copies to and from V1 disks.
 23180	 */
 23181	
 23182	  int i;
 23183	
 23184	  if (direction == READING) {
 23185	        /* Copy V1.x inode to the in-core table, swapping bytes if need be. */
 23186	        rip->i_mode    = conv2(norm, (int) dip->d1_mode);
 23187	        rip->i_uid     = conv2(norm, (int) dip->d1_uid );
 23188	        rip->i_size    = conv4(norm,       dip->d1_size);
 23189	        rip->i_mtime   = conv4(norm,       dip->d1_mtime);
 23190	        rip->i_atime   = rip->i_mtime;
 23191	        rip->i_ctime   = rip->i_mtime;
 23192	        rip->i_nlinks  = dip->d1_nlinks;                /* 1 char */
 23193	        rip->i_gid     = dip->d1_gid;                   /* 1 char */
 23194	        rip->i_ndzones = V1_NR_DZONES;
 23195	        rip->i_nindirs = V1_INDIRECTS;
 23196	        for (i = 0; i < V1_NR_TZONES; i++)
 23197	                rip->i_zone[i] = conv2(norm, (int) dip->d1_zone[i]);
 23198	  } else {
 23199	        /* Copying V1.x inode to disk from the in-core table. */
 23200	        dip->d1_mode   = conv2(norm, (int) rip->i_mode);
 23201	        dip->d1_uid    = conv2(norm, (int) rip->i_uid );
 23202	        dip->d1_size   = conv4(norm,       rip->i_size);
 23203	        dip->d1_mtime  = conv4(norm,       rip->i_mtime);
 23204	        dip->d1_nlinks = rip->i_nlinks;                 /* 1 char */

_________________________ Page 946    File: servers/fs/inode.c _________________________
 23205	        dip->d1_gid    = rip->i_gid;                    /* 1 char */
 23206	        for (i = 0; i < V1_NR_TZONES; i++)
 23207	                dip->d1_zone[i] = conv2(norm, (int) rip->i_zone[i]);
 23208	  }
 23209	}
	
 23211	/*===========================================================================*
 23212	 *                              new_icopy                                    *
 23213	 *===========================================================================*/
 23214	PRIVATE void new_icopy(rip, dip, direction, norm)
 23215	register struct inode *rip;     /* pointer to the in-core inode struct */
 23216	register d2_inode *dip; /* pointer to the d2_inode struct */
 23217	int direction;                  /* READING (from disk) or WRITING (to disk) */
 23218	int norm;                       /* TRUE = do not swap bytes; FALSE = swap */
 23219	
 23220	{
 23221	/* Same as old_icopy, but to/from V2 disk layout. */
 23222	
 23223	  int i;
 23224	
 23225	  if (direction == READING) {
 23226	        /* Copy V2.x inode to the in-core table, swapping bytes if need be. */
 23227	        rip->i_mode    = conv2(norm,dip->d2_mode);
 23228	        rip->i_uid     = conv2(norm,dip->d2_uid);
 23229	        rip->i_nlinks  = conv2(norm,dip->d2_nlinks);
 23230	        rip->i_gid     = conv2(norm,dip->d2_gid);
 23231	        rip->i_size    = conv4(norm,dip->d2_size);
 23232	        rip->i_atime   = conv4(norm,dip->d2_atime);
 23233	        rip->i_ctime   = conv4(norm,dip->d2_ctime);
 23234	        rip->i_mtime   = conv4(norm,dip->d2_mtime);
 23235	        rip->i_ndzones = V2_NR_DZONES;
 23236	        rip->i_nindirs = V2_INDIRECTS(rip->i_sp->s_block_size);
 23237	        for (i = 0; i < V2_NR_TZONES; i++)
 23238	                rip->i_zone[i] = conv4(norm, (long) dip->d2_zone[i]);
 23239	  } else {
 23240	        /* Copying V2.x inode to disk from the in-core table. */
 23241	        dip->d2_mode   = conv2(norm,rip->i_mode);
 23242	        dip->d2_uid    = conv2(norm,rip->i_uid);
 23243	        dip->d2_nlinks = conv2(norm,rip->i_nlinks);
 23244	        dip->d2_gid    = conv2(norm,rip->i_gid);
 23245	        dip->d2_size   = conv4(norm,rip->i_size);
 23246	        dip->d2_atime  = conv4(norm,rip->i_atime);
 23247	        dip->d2_ctime  = conv4(norm,rip->i_ctime);
 23248	        dip->d2_mtime  = conv4(norm,rip->i_mtime);
 23249	        for (i = 0; i < V2_NR_TZONES; i++)
 23250	                dip->d2_zone[i] = conv4(norm, (long) rip->i_zone[i]);
 23251	  }
 23252	}
	
 23254	/*===========================================================================*
 23255	 *                              dup_inode                                    *
 23256	 *===========================================================================*/
 23257	PUBLIC void dup_inode(ip)
 23258	struct inode *ip;               /* The inode to be duplicated. */
 23259	{
 23260	/* This routine is a simplified form of get_inode() for the case where
 23261	 * the inode pointer is already known.
 23262	 */
 23263	
 23264	  ip->i_count++;

_________________________ Page 947    File: servers/fs/inode.c _________________________
 23265	}





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/super.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 23300	/* This file manages the super block table and the related data structures,
 23301	 * namely, the bit maps that keep track of which zones and which inodes are
 23302	 * allocated and which are free.  When a new inode or zone is needed, the
 23303	 * appropriate bit map is searched for a free entry.
 23304	 *
 23305	 * The entry points into this file are
 23306	 *   alloc_bit:       somebody wants to allocate a zone or inode; find one
 23307	 *   free_bit:        indicate that a zone or inode is available for allocation
 23308	 *   get_super:       search the 'superblock' table for a device
 23309	 *   mounted:         tells if file inode is on mounted (or ROOT) file system
 23310	 *   read_super:      read a superblock
 23311	 */
 23312	
 23313	#include "fs.h"
 23314	#include <string.h>
 23315	#include <minix/com.h>
 23316	#include "buf.h"
 23317	#include "inode.h"
 23318	#include "super.h"
 23319	#include "const.h"
 23320	
 23321	/*===========================================================================*
 23322	 *                              alloc_bit                                    *
 23323	 *===========================================================================*/
 23324	PUBLIC bit_t alloc_bit(sp, map, origin)
 23325	struct super_block *sp;         /* the filesystem to allocate from */
 23326	int map;                        /* IMAP (inode map) or ZMAP (zone map) */
 23327	bit_t origin;                   /* number of bit to start searching at */
 23328	{
 23329	/* Allocate a bit from a bit map and return its bit number. */
 23330	
 23331	  block_t start_block;          /* first bit block */
 23332	  bit_t map_bits;               /* how many bits are there in the bit map? */
 23333	  unsigned bit_blocks;          /* how many blocks are there in the bit map? */
 23334	  unsigned block, word, bcount;
 23335	  struct buf *bp;
 23336	  bitchunk_t *wptr, *wlim, k;
 23337	  bit_t i, b;
 23338	
 23339	  if (sp->s_rd_only)
 23340	        panic(__FILE__,"can't allocate bit on read-only filesys.", NO_NUM);
 23341	
 23342	  if (map == IMAP) {
 23343	        start_block = START_BLOCK;
 23344	        map_bits = sp->s_ninodes + 1;
 23345	        bit_blocks = sp->s_imap_blocks;
 23346	  } else {
 23347	        start_block = START_BLOCK + sp->s_imap_blocks;
 23348	        map_bits = sp->s_zones - (sp->s_firstdatazone - 1);
 23349	        bit_blocks = sp->s_zmap_blocks;

_________________________ Page 948    File: servers/fs/super.c _________________________
 23350	  }
 23351	
 23352	  /* Figure out where to start the bit search (depends on 'origin'). */
 23353	  if (origin >= map_bits) origin = 0;   /* for robustness */
 23354	
 23355	  /* Locate the starting place. */
 23356	  block = origin / FS_BITS_PER_BLOCK(sp->s_block_size);
 23357	  word = (origin % FS_BITS_PER_BLOCK(sp->s_block_size)) / FS_BITCHUNK_BITS;
 23358	
 23359	  /* Iterate over all blocks plus one, because we start in the middle. */
 23360	  bcount = bit_blocks + 1;
 23361	  do {
 23362	        bp = get_block(sp->s_dev, start_block + block, NORMAL);
 23363	        wlim = &bp->b_bitmap[FS_BITMAP_CHUNKS(sp->s_block_size)];
 23364	
 23365	        /* Iterate over the words in block. */
 23366	        for (wptr = &bp->b_bitmap[word]; wptr < wlim; wptr++) {
 23367	
 23368	                /* Does this word contain a free bit? */
 23369	                if (*wptr == (bitchunk_t) ~0) continue;
 23370	
 23371	                /* Find and allocate the free bit. */
 23372	                k = conv2(sp->s_native, (int) *wptr);
 23373	                for (i = 0; (k & (1 << i)) != 0; ++i) {}
 23374	
 23375	                /* Bit number from the start of the bit map. */
 23376	                b = ((bit_t) block * FS_BITS_PER_BLOCK(sp->s_block_size))
 23377	                    + (wptr - &bp->b_bitmap[0]) * FS_BITCHUNK_BITS
 23378	                    + i;
 23379	
 23380	                /* Don't allocate bits beyond the end of the map. */
 23381	                if (b >= map_bits) break;
 23382	
 23383	                /* Allocate and return bit number. */
 23384	                k |= 1 << i;
 23385	                *wptr = conv2(sp->s_native, (int) k);
 23386	                bp->b_dirt = DIRTY;
 23387	                put_block(bp, MAP_BLOCK);
 23388	                return(b);
 23389	        }
 23390	        put_block(bp, MAP_BLOCK);
 23391	        if (++block >= bit_blocks) block = 0;   /* last block, wrap around */
 23392	        word = 0;
 23393	  } while (--bcount > 0);
 23394	  return(NO_BIT);               /* no bit could be allocated */
 23395	}
	
 23397	/*===========================================================================*
 23398	 *                              free_bit                                     *
 23399	 *===========================================================================*/
 23400	PUBLIC void free_bit(sp, map, bit_returned)
 23401	struct super_block *sp;         /* the filesystem to operate on */
 23402	int map;                        /* IMAP (inode map) or ZMAP (zone map) */
 23403	bit_t bit_returned;             /* number of bit to insert into the map */
 23404	{
 23405	/* Return a zone or inode by turning off its bitmap bit. */
 23406	
 23407	  unsigned block, word, bit;
 23408	  struct buf *bp;
 23409	  bitchunk_t k, mask;

_________________________ Page 949    File: servers/fs/super.c _________________________
 23410	  block_t start_block;
 23411	
 23412	  if (sp->s_rd_only)
 23413	        panic(__FILE__,"can't free bit on read-only filesys.", NO_NUM);
 23414	
 23415	  if (map == IMAP) {
 23416	        start_block = START_BLOCK;
 23417	  } else {
 23418	        start_block = START_BLOCK + sp->s_imap_blocks;
 23419	  }
 23420	  block = bit_returned / FS_BITS_PER_BLOCK(sp->s_block_size);
 23421	  word = (bit_returned % FS_BITS_PER_BLOCK(sp->s_block_size))
 23422	         / FS_BITCHUNK_BITS;
 23423	
 23424	  bit = bit_returned % FS_BITCHUNK_BITS;
 23425	  mask = 1 << bit;
 23426	
 23427	  bp = get_block(sp->s_dev, start_block + block, NORMAL);
 23428	
 23429	  k = conv2(sp->s_native, (int) bp->b_bitmap[word]);
 23430	  if (!(k & mask)) {
 23431	        panic(__FILE__,map == IMAP ? "tried to free unused inode" :
 23432	              "tried to free unused block", NO_NUM);
 23433	  }
 23434	
 23435	  k &= ~mask;
 23436	  bp->b_bitmap[word] = conv2(sp->s_native, (int) k);
 23437	  bp->b_dirt = DIRTY;
 23438	
 23439	  put_block(bp, MAP_BLOCK);
 23440	}
	
 23442	/*===========================================================================*
 23443	 *                              get_super                                    *
 23444	 *===========================================================================*/
 23445	PUBLIC struct super_block *get_super(dev)
 23446	dev_t dev;                      /* device number whose super_block is sought */
 23447	{
 23448	/* Search the superblock table for this device.  It is supposed to be there. */
 23449	
 23450	  register struct super_block *sp;
 23451	
 23452	  if (dev == NO_DEV)
 23453	        panic(__FILE__,"request for super_block of NO_DEV", NO_NUM);
 23454	
 23455	  for (sp = &super_block[0]; sp < &super_block[NR_SUPERS]; sp++)
 23456	        if (sp->s_dev == dev) return(sp);
 23457	
 23458	  /* Search failed.  Something wrong. */
 23459	  panic(__FILE__,"can't find superblock for device (in decimal)", (int) dev);
 23460	
 23461	  return(NIL_SUPER);            /* to keep the compiler and lint quiet */
 23462	}
	
 23464	/*===========================================================================*
 23465	 *                              get_block_size                               *
 23466	 *===========================================================================*/
 23467	PUBLIC int get_block_size(dev_t dev)
 23468	{
 23469	/* Search the superblock table for this device. */

_________________________ Page 950    File: servers/fs/super.c _________________________
 23470	
 23471	  register struct super_block *sp;
 23472	
 23473	  if (dev == NO_DEV)
 23474	        panic(__FILE__,"request for block size of NO_DEV", NO_NUM);
 23475	
 23476	  for (sp = &super_block[0]; sp < &super_block[NR_SUPERS]; sp++) {
 23477	        if (sp->s_dev == dev) {
 23478	                return(sp->s_block_size);
 23479	        }
 23480	  }
 23481	
 23482	  /* no mounted filesystem? use this block size then. */
 23483	  return MIN_BLOCK_SIZE;
 23484	}
	
 23486	/*===========================================================================*
 23487	 *                              mounted                                      *
 23488	 *===========================================================================*/
 23489	PUBLIC int mounted(rip)
 23490	register struct inode *rip;     /* pointer to inode */
 23491	{
 23492	/* Report on whether the given inode is on a mounted (or ROOT) file system. */
 23493	
 23494	  register struct super_block *sp;
 23495	  register dev_t dev;
 23496	
 23497	  dev = (dev_t) rip->i_zone[0];
 23498	  if (dev == root_dev) return(TRUE);    /* inode is on root file system */
 23499	
 23500	  for (sp = &super_block[0]; sp < &super_block[NR_SUPERS]; sp++)
 23501	        if (sp->s_dev == dev) return(TRUE);
 23502	
 23503	  return(FALSE);
 23504	}
	
 23506	/*===========================================================================*
 23507	 *                              read_super                                   *
 23508	 *===========================================================================*/
 23509	PUBLIC int read_super(sp)
 23510	register struct super_block *sp; /* pointer to a superblock */
 23511	{
 23512	/* Read a superblock. */
 23513	  dev_t dev;
 23514	  int magic;
 23515	  int version, native, r;
 23516	  static char sbbuf[MIN_BLOCK_SIZE];
 23517	
 23518	  dev = sp->s_dev;              /* save device (will be overwritten by copy) */
 23519	  if (dev == NO_DEV)
 23520	        panic(__FILE__,"request for super_block of NO_DEV", NO_NUM);
 23521	  r = dev_io(DEV_READ, dev, FS_PROC_NR,
 23522	        sbbuf, SUPER_BLOCK_BYTES, MIN_BLOCK_SIZE, 0);
 23523	  if (r != MIN_BLOCK_SIZE) {
 23524	        return EINVAL;
 23525	  }
 23526	  memcpy(sp, sbbuf, sizeof(*sp));
 23527	  sp->s_dev = NO_DEV;           /* restore later */
 23528	  magic = sp->s_magic;          /* determines file system type */
 23529	

_________________________ Page 951    File: servers/fs/super.c _________________________
 23530	  /* Get file system version and type. */
 23531	  if (magic == SUPER_MAGIC || magic == conv2(BYTE_SWAP, SUPER_MAGIC)) {
 23532	        version = V1;
 23533	        native  = (magic == SUPER_MAGIC);
 23534	  } else if (magic == SUPER_V2 || magic == conv2(BYTE_SWAP, SUPER_V2)) {
 23535	        version = V2;
 23536	        native  = (magic == SUPER_V2);
 23537	  } else if (magic == SUPER_V3) {
 23538	        version = V3;
 23539	        native = 1;
 23540	  } else {
 23541	        return(EINVAL);
 23542	  }
 23543	
 23544	  /* If the super block has the wrong byte order, swap the fields; the magic
 23545	   * number doesn't need conversion. */
 23546	  sp->s_ninodes =       conv4(native, sp->s_ninodes);
 23547	  sp->s_nzones =        conv2(native, (int) sp->s_nzones);
 23548	  sp->s_imap_blocks =   conv2(native, (int) sp->s_imap_blocks);
 23549	  sp->s_zmap_blocks =   conv2(native, (int) sp->s_zmap_blocks);
 23550	  sp->s_firstdatazone = conv2(native, (int) sp->s_firstdatazone);
 23551	  sp->s_log_zone_size = conv2(native, (int) sp->s_log_zone_size);
 23552	  sp->s_max_size =      conv4(native, sp->s_max_size);
 23553	  sp->s_zones =         conv4(native, sp->s_zones);
 23554	
 23555	  /* In V1, the device size was kept in a short, s_nzones, which limited
 23556	   * devices to 32K zones.  For V2, it was decided to keep the size as a
 23557	   * long.  However, just changing s_nzones to a long would not work, since
 23558	   * then the position of s_magic in the super block would not be the same
 23559	   * in V1 and V2 file systems, and there would be no way to tell whether
 23560	   * a newly mounted file system was V1 or V2.  The solution was to introduce
 23561	   * a new variable, s_zones, and copy the size there.
 23562	   *
 23563	   * Calculate some other numbers that depend on the version here too, to
 23564	   * hide some of the differences.
 23565	   */
 23566	  if (version == V1) {
 23567	        sp->s_block_size = STATIC_BLOCK_SIZE;
 23568	        sp->s_zones = sp->s_nzones;     /* only V1 needs this copy */
 23569	        sp->s_inodes_per_block = V1_INODES_PER_BLOCK;
 23570	        sp->s_ndzones = V1_NR_DZONES;
 23571	        sp->s_nindirs = V1_INDIRECTS;
 23572	  } else {
 23573	        if (version == V2)
 23574	                sp->s_block_size = STATIC_BLOCK_SIZE;
 23575	        if (sp->s_block_size < MIN_BLOCK_SIZE)
 23576	                return EINVAL;
 23577	        sp->s_inodes_per_block = V2_INODES_PER_BLOCK(sp->s_block_size);
 23578	        sp->s_ndzones = V2_NR_DZONES;
 23579	        sp->s_nindirs = V2_INDIRECTS(sp->s_block_size);
 23580	  }
 23581	
 23582	  if (sp->s_block_size < MIN_BLOCK_SIZE) {
 23583	        return EINVAL;
 23584	  }
 23585	  if (sp->s_block_size > MAX_BLOCK_SIZE) {
 23586	        printf("Filesystem block size is %d kB; maximum filesystem\n"
 23587	        "block size is %d kB. This limit can be increased by recompiling.\n",
 23588	        sp->s_block_size/1024, MAX_BLOCK_SIZE/1024);
 23589	        return EINVAL;

_________________________ Page 952    File: servers/fs/super.c _________________________
 23590	  }
 23591	  if ((sp->s_block_size % 512) != 0) {
 23592	        return EINVAL;
 23593	  }
 23594	  if (SUPER_SIZE > sp->s_block_size) {
 23595	        return EINVAL;
 23596	  }
 23597	  if ((sp->s_block_size % V2_INODE_SIZE) != 0 ||
 23598	     (sp->s_block_size % V1_INODE_SIZE) != 0) {
 23599	        return EINVAL;
 23600	  }
 23601	
 23602	  sp->s_isearch = 0;            /* inode searches initially start at 0 */
 23603	  sp->s_zsearch = 0;            /* zone searches initially start at 0 */
 23604	  sp->s_version = version;
 23605	  sp->s_native  = native;
 23606	
 23607	  /* Make a few basic checks to see if super block looks reasonable. */
 23608	  if (sp->s_imap_blocks < 1 || sp->s_zmap_blocks < 1
 23609	                                || sp->s_ninodes < 1 || sp->s_zones < 1
 23610	                                || (unsigned) sp->s_log_zone_size > 4) {
 23611	        printf("not enough imap or zone map blocks, \n");
 23612	        printf("or not enough inodes, or not enough zones, "
 23613	                "or zone size too large\n");
 23614	        return(EINVAL);
 23615	  }
 23616	  sp->s_dev = dev;              /* restore device number */
 23617	  return(OK);
 23618	}


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/filedes.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 23700	/* This file contains the procedures that manipulate file descriptors.
 23701	 *
 23702	 * The entry points into this file are
 23703	 *   get_fd:    look for free file descriptor and free filp slots
 23704	 *   get_filp:  look up the filp entry for a given file descriptor
 23705	 *   find_filp: find a filp slot that points to a given inode
 23706	 */
 23707	
 23708	#include "fs.h"
 23709	#include "file.h"
 23710	#include "fproc.h"
 23711	#include "inode.h"
 23712	
 23713	/*===========================================================================*
 23714	 *                              get_fd                                       *
 23715	 *===========================================================================*/
 23716	PUBLIC int get_fd(int start, mode_t bits, int *k, struct filp **fpt)
 23717	{
 23718	/* Look for a free file descriptor and a free filp slot.  Fill in the mode word
 23719	 * in the latter, but don't claim either one yet, since the open() or creat()
 23720	 * may yet fail.
 23721	 */
 23722	
 23723	  register struct filp *f;
 23724	  register int i;

_________________________ Page 953    File: servers/fs/filedes.c _________________________
 23725	
 23726	  *k = -1;                      /* we need a way to tell if file desc found */
 23727	
 23728	  /* Search the fproc fp_filp table for a free file descriptor. */
 23729	  for (i = start; i < OPEN_MAX; i++) {
 23730	        if (fp->fp_filp[i] == NIL_FILP) {
 23731	                /* A file descriptor has been located. */
 23732	                *k = i;
 23733	                break;
 23734	        }
 23735	  }
 23736	
 23737	  /* Check to see if a file descriptor has been found. */
 23738	  if (*k < 0) return(EMFILE);   /* this is why we initialized k to -1 */
 23739	
 23740	  /* Now that a file descriptor has been found, look for a free filp slot. */
 23741	  for (f = &filp[0]; f < &filp[NR_FILPS]; f++) {
 23742	        if (f->filp_count == 0) {
 23743	                f->filp_mode = bits;
 23744	                f->filp_pos = 0L;
 23745	                f->filp_selectors = 0;
 23746	                f->filp_select_ops = 0;
 23747	                f->filp_pipe_select_ops = 0;
 23748	                f->filp_flags = 0;
 23749	                *fpt = f;
 23750	                return(OK);
 23751	        }
 23752	  }
 23753	
 23754	  /* If control passes here, the filp table must be full.  Report that back. */
 23755	  return(ENFILE);
 23756	}
	
 23758	/*===========================================================================*
 23759	 *                              get_filp                                     *
 23760	 *===========================================================================*/
 23761	PUBLIC struct filp *get_filp(fild)
 23762	int fild;                       /* file descriptor */
 23763	{
 23764	/* See if 'fild' refers to a valid file descr.  If so, return its filp ptr. */
 23765	
 23766	  err_code = EBADF;
 23767	  if (fild < 0 || fild >= OPEN_MAX ) return(NIL_FILP);
 23768	  return(fp->fp_filp[fild]);    /* may also be NIL_FILP */
 23769	}
	
 23771	/*===========================================================================*
 23772	 *                              find_filp                                    *
 23773	 *===========================================================================*/
 23774	PUBLIC struct filp *find_filp(register struct inode *rip, mode_t bits)
 23775	{
 23776	/* Find a filp slot that refers to the inode 'rip' in a way as described
 23777	 * by the mode bit 'bits'. Used for determining whether somebody is still
 23778	 * interested in either end of a pipe.  Also used when opening a FIFO to
 23779	 * find partners to share a filp field with (to shared the file position).
 23780	 * Like 'get_fd' it performs its job by linear search through the filp table.
 23781	 */
 23782	
 23783	  register struct filp *f;
 23784	

_________________________ Page 954    File: servers/fs/filedes.c _________________________
 23785	  for (f = &filp[0]; f < &filp[NR_FILPS]; f++) {
 23786	        if (f->filp_count != 0 && f->filp_ino == rip && (f->filp_mo
 23787	                return(f);
 23788	        }
 23789	  }
 23790	
 23791	  /* If control passes here, the filp wasn't there.  Report that back. */
 23792	  return(NIL_FILP);
 23793	}


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/lock.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 23800	/* This file handles advisory file locking as required by POSIX.
 23801	 *
 23802	 * The entry points into this file are
 23803	 *   lock_op:   perform locking operations for FCNTL system call
 23804	 *   lock_revive: revive processes when a lock is released
 23805	 */
 23806	
 23807	#include "fs.h"
 23808	#include <minix/com.h>
 23809	#include <fcntl.h>
 23810	#include <unistd.h>
 23811	#include "file.h"
 23812	#include "fproc.h"
 23813	#include "inode.h"
 23814	#include "lock.h"
 23815	#include "param.h"
 23816	
 23817	/*===========================================================================*
 23818	 *                              lock_op                                      *
 23819	 *===========================================================================*/
 23820	PUBLIC int lock_op(f, req)
 23821	struct filp *f;
 23822	int req;                        /* either F_SETLK or F_SETLKW */
 23823	{
 23824	/* Perform the advisory locking required by POSIX. */
 23825	
 23826	  int r, ltype, i, conflict = 0, unlocking = 0;
 23827	  mode_t mo;
 23828	  off_t first, last;
 23829	  struct flock flock;
 23830	  vir_bytes user_flock;
 23831	  struct file_lock *flp, *flp2, *empty;
 23832	
 23833	  /* Fetch the flock structure from user space. */
 23834	  user_flock = (vir_bytes) m_in.name1;
 23835	  r = sys_datacopy(who, (vir_bytes) user_flock,
 23836	        FS_PROC_NR, (vir_bytes) &flock, (phys_bytes) sizeof(flock));
 23837	  if (r != OK) return(EINVAL);
 23838	
 23839	  /* Make some error checks. */
 23840	  ltype = flock.l_type;
 23841	  mo = f->filp_mode;
 23842	  if (ltype != F_UNLCK && ltype != F_RDLCK && ltype != F_WRLCK) ret
 23843	  if (req == F_GETLK && ltype == F_UNLCK) return(EINVAL);
 23844	  if ( (f->filp_ino->i_mode & I_TYPE) != I_REGULAR) return(EINVAL);

_________________________ Page 955    File: servers/fs/lock.c _________________________
 23845	  if (req != F_GETLK && ltype == F_RDLCK && (mo & R_BIT) == 0) 
 23846	  if (req != F_GETLK && ltype == F_WRLCK && (mo & W_BIT) == 0) 
 23847	
 23848	  /* Compute the first and last bytes in the lock region. */
 23849	  switch (flock.l_whence) {
 23850	        case SEEK_SET:  first = 0; break;
 23851	        case SEEK_CUR:  first = f->filp_pos; break;
 23852	        case SEEK_END:  first = f->filp_ino->i_size; break;
 23853	        default:        return(EINVAL);
 23854	  }
 23855	  /* Check for overflow. */
 23856	  if (((long)flock.l_start > 0) && ((first + flock.l_start) < first))
 23857	        return(EINVAL);
 23858	  if (((long)flock.l_start < 0) && ((first + flock.l_start) > first))
 23859	        return(EINVAL);
 23860	  first = first + flock.l_start;
 23861	  last = first + flock.l_len - 1;
 23862	  if (flock.l_len == 0) last = MAX_FILE_POS;
 23863	  if (last < first) return(EINVAL);
 23864	
 23865	  /* Check if this region conflicts with any existing lock. */
 23866	  empty = (struct file_lock *) 0;
 23867	  for (flp = &file_lock[0]; flp < & file_lock[NR_LOCKS]; flp++) {
 23868	        if (flp->lock_type == 0) {
 23869	                if (empty == (struct file_lock *) 0) empty = flp;
 23870	                continue;       /* 0 means unused slot */
 23871	        }
 23872	        if (flp->lock_inode != f->filp_ino) continue;   /* different file */
 23873	        if (last < flp->lock_first) continue;   /* new one is in front */
 23874	        if (first > flp->lock_last) continue;   /* new one is afterwards */
 23875	        if (ltype == F_RDLCK && flp->lock_type == F_RDLCK) continue;
 23876	        if (ltype != F_UNLCK && flp->lock_pid == fp->fp_pid) continue;
 23877	  
 23878	        /* There might be a conflict.  Process it. */
 23879	        conflict = 1;
 23880	        if (req == F_GETLK) break;
 23881	
 23882	        /* If we are trying to set a lock, it just failed. */
 23883	        if (ltype == F_RDLCK || ltype == F_WRLCK) {
 23884	                if (req == F_SETLK) {
 23885	                        /* For F_SETLK, just report back failure. */
 23886	                        return(EAGAIN);
 23887	                } else {
 23888	                        /* For F_SETLKW, suspend the process. */
 23889	                        suspend(XLOCK);
 23890	                        return(SUSPEND);
 23891	                }
 23892	        }
 23893	
 23894	        /* We are clearing a lock and we found something that overlaps. */
 23895	        unlocking = 1;
 23896	        if (first <= flp->lock_first && last >= flp->lock_last) {
 23897	                flp->lock_type = 0;     /* mark slot as unused */
 23898	                nr_locks--;             /* number of locks is now 1 less */
 23899	                continue;
 23900	        }
 23901	
 23902	        /* Part of a locked region has been unlocked. */
 23903	        if (first <= flp->lock_first) {
 23904	                flp->lock_first = last + 1;

_________________________ Page 956    File: servers/fs/lock.c _________________________
 23905	                continue;
 23906	        }
 23907	
 23908	        if (last >= flp->lock_last) {
 23909	                flp->lock_last = first - 1;
 23910	                continue;
 23911	        }
 23912	        
 23913	        /* Bad luck. A lock has been split in two by unlocking the middle. */
 23914	        if (nr_locks == NR_LOCKS) return(ENOLCK);
 23915	        for (i = 0; i < NR_LOCKS; i++)
 23916	                if (file_lock[i].lock_type == 0) break;
 23917	        flp2 = &file_lock[i];
 23918	        flp2->lock_type = flp->lock_type;
 23919	        flp2->lock_pid = flp->lock_pid;
 23920	        flp2->lock_inode = flp->lock_inode;
 23921	        flp2->lock_first = last + 1;
 23922	        flp2->lock_last = flp->lock_last;
 23923	        flp->lock_last = first - 1;
 23924	        nr_locks++;
 23925	  }
 23926	  if (unlocking) lock_revive();
 23927	
 23928	  if (req == F_GETLK) {
 23929	        if (conflict) {
 23930	                /* GETLK and conflict. Report on the conflicting lock. */
 23931	                flock.l_type = flp->lock_type;
 23932	                flock.l_whence = SEEK_SET;
 23933	                flock.l_start = flp->lock_first;
 23934	                flock.l_len = flp->lock_last - flp->lock_first + 1;
 23935	                flock.l_pid = flp->lock_pid;
 23936	
 23937	        } else {
 23938	                /* It is GETLK and there is no conflict. */
 23939	                flock.l_type = F_UNLCK;
 23940	        }
 23941	
 23942	        /* Copy the flock structure back to the caller. */
 23943	        r = sys_datacopy(FS_PROC_NR, (vir_bytes) &flock,
 23944	                who, (vir_bytes) user_flock, (phys_bytes) sizeof(flock));
 23945	        return(r);
 23946	  }
 23947	
 23948	  if (ltype == F_UNLCK) return(OK);     /* unlocked a region with no locks */
 23949	
 23950	  /* There is no conflict.  If space exists, store new lock in the table. */
 23951	  if (empty == (struct file_lock *) 0) return(ENOLCK);  /* table full */
 23952	  empty->lock_type = ltype;
 23953	  empty->lock_pid = fp->fp_pid;
 23954	  empty->lock_inode = f->filp_ino;
 23955	  empty->lock_first = first;
 23956	  empty->lock_last = last;
 23957	  nr_locks++;
 23958	  return(OK);
 23959	}
	

_________________________ Page 957    File: servers/fs/lock.c _________________________
 23961	/*===========================================================================*
 23962	 *                              lock_revive                                  *
 23963	 *===========================================================================*/
 23964	PUBLIC void lock_revive()
 23965	{
 23966	/* Go find all the processes that are waiting for any kind of lock and 
 23967	 * revive them all.  The ones that are still blocked will block again when 
 23968	 * they run.  The others will complete.  This strategy is a space-time 
 23969	 * tradeoff.  Figuring out exactly which ones to unblock now would take 
 23970	 * extra code, and the only thing it would win would be some performance in 
 23971	 * extremely rare circumstances (namely, that somebody actually used 
 23972	 * locking).
 23973	 */
 23974	
 23975	  int task;
 23976	  struct fproc *fptr;
 23977	
 23978	  for (fptr = &fproc[INIT_PROC_NR + 1]; fptr < &fproc[NR_PROCS]; fptr++)
 23979	        task = -fptr->fp_task;
 23980	        if (fptr->fp_suspended == SUSPENDED && task == XLOCK) {
 23981	                revive( (int) (fptr - fproc), 0);
 23982	        }
 23983	  }
 23984	}


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/main.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 24000	/* This file contains the main program of the File System.  It consists of
 24001	 * a loop that gets messages requesting work, carries out the work, and sends
 24002	 * replies.
 24003	 *
 24004	 * The entry points into this file are:
 24005	 *   main:      main program of the File System
 24006	 *   reply:     send a reply to a process after the requested work is done
 24007	 *
 24008	 */
 24009	
 24010	struct super_block;             /* proto.h needs to know this */
 24011	
 24012	#include "fs.h"
 24013	#include <fcntl.h>
 24014	#include <string.h>
 24015	#include <stdio.h>
 24016	#include <signal.h>
 24017	#include <stdlib.h>
 24018	#include <sys/ioc_memory.h>
 24019	#include <sys/svrctl.h>
 24020	#include <minix/callnr.h>
 24021	#include <minix/com.h>
 24022	#include <minix/keymap.h>
 24023	#include <minix/const.h>
 24024	#include "buf.h"
 24025	#include "file.h"
 24026	#include "fproc.h"
 24027	#include "inode.h"
 24028	#include "param.h"
 24029	#include "super.h"

_________________________ Page 958    File: servers/fs/main.c _________________________
 24030	
 24031	FORWARD _PROTOTYPE( void fs_init, (void)                                );
 24032	FORWARD _PROTOTYPE( int igetenv, (char *var, int optional)              );
 24033	FORWARD _PROTOTYPE( void get_work, (void)                               );
 24034	FORWARD _PROTOTYPE( void load_ram, (void)                               );
 24035	FORWARD _PROTOTYPE( void load_super, (Dev_t super_dev)                  );
 24036	
 24037	/*===========================================================================*
 24038	 *                              main                                         *
 24039	 *===========================================================================*/
 24040	PUBLIC int main()
 24041	{
 24042	/* This is the main program of the file system.  The main loop consists of
 24043	 * three major activities: getting new work, processing the work, and sending
 24044	 * the reply.  This loop never terminates as long as the file system runs.
 24045	 */
 24046	  sigset_t sigset;
 24047	  int error;
 24048	
 24049	  fs_init();
 24050	
 24051	  /* This is the main loop that gets work, processes it, and sends replies. */
 24052	  while (TRUE) {
 24053	        get_work();             /* sets who and call_nr */
 24054	
 24055	        fp = &fproc[who];       /* pointer to proc table struct */
 24056	        super_user = (fp->fp_effuid == SU_UID ? TRUE : FALSE);   /* su? */
 24057	
 24058	        /* Check for special control messages first. */
 24059	        if (call_nr == SYS_SIG) { 
 24060	                sigset = m_in.NOTIFY_ARG;
 24061	                if (sigismember(&sigset, SIGKSTOP)) {
 24062	                        do_sync();
 24063	                        sys_exit(0);            /* never returns */
 24064	                }
 24065	        } else if (call_nr == SYN_ALARM) {
 24066	                /* Not a user request; system has expired one of our timers,
 24067	                 * currently only in use for select(). Check it.
 24068	                 */
 24069	                fs_expire_timers(m_in.NOTIFY_TIMESTAMP);
 24070	        } else if ((call_nr & NOTIFY_MESSAGE)) {
 24071	                /* Device notifies us of an event. */
 24072	                dev_status(&m_in);
 24073	        } else {
 24074	                /* Call the internal function that does the work. */
 24075	                if (call_nr < 0 || call_nr >= NCALLS) { 
 24076	                        error = ENOSYS;
 24077	                        printf("FS, warning illegal %d system call by %d\n", call_n
 24078	                } else if (fp->fp_pid == PID_FREE) {
 24079	                        error = ENOSYS;
 24080	                        printf("FS, bad process, who = %d, call_nr = %d, slot1 = %d
 24081	                                 who, call_nr, m_in.slot1);
 24082	                } else {
 24083	                        error = (*call_vec[call_nr])();
 24084	                }
 24085	
 24086	                /* Copy the results back to the user and send reply. */
 24087	                if (error != SUSPEND) { reply(who, error); }
 24088	                if (rdahed_inode != NIL_INODE) {
 24089	                        read_ahead(); /* do block read ahead */

_________________________ Page 959    File: servers/fs/main.c _________________________
 24090	                }
 24091	        }
 24092	  }
 24093	  return(OK);                           /* shouldn't come here */
 24094	}
	
 24096	/*===========================================================================*
 24097	 *                              get_work                                     *
 24098	 *===========================================================================*/
 24099	PRIVATE void get_work()
 24100	{  
 24101	  /* Normally wait for new input.  However, if 'reviving' is
 24102	   * nonzero, a suspended process must be awakened.
 24103	   */
 24104	  register struct fproc *rp;
 24105	
 24106	  if (reviving != 0) {
 24107	        /* Revive a suspended process. */
 24108	        for (rp = &fproc[0]; rp < &fproc[NR_PROCS]; rp++) 
 24109	                if (rp->fp_revived == REVIVING) {
 24110	                        who = (int)(rp - fproc);
 24111	                        call_nr = rp->fp_fd & BYTE;
 24112	                        m_in.fd = (rp->fp_fd >>8) & BYTE;
 24113	                        m_in.buffer = rp->fp_buffer;
 24114	                        m_in.nbytes = rp->fp_nbytes;
 24115	                        rp->fp_suspended = NOT_SUSPENDED; /*no longer hanging*/
 24116	                        rp->fp_revived = NOT_REVIVING;
 24117	                        reviving--;
 24118	                        return;
 24119	                }
 24120	        panic(__FILE__,"get_work couldn't revive anyone", NO_NUM);
 24121	  }
 24122	
 24123	  /* Normal case.  No one to revive. */
 24124	  if (receive(ANY, &m_in) != OK) panic(__FILE__,"fs receive error", NO_NUM);
 24125	  who = m_in.m_source;
 24126	  call_nr = m_in.m_type;
 24127	}
	
 24129	/*===========================================================================*
 24130	 *                              buf_pool                                     *
 24131	 *===========================================================================*/
 24132	PRIVATE void buf_pool(void)
 24133	{
 24134	/* Initialize the buffer pool. */
 24135	
 24136	  register struct buf *bp;
 24137	
 24138	  bufs_in_use = 0;
 24139	  front = &buf[0];
 24140	  rear = &buf[NR_BUFS - 1];
 24141	
 24142	  for (bp = &buf[0]; bp < &buf[NR_BUFS]; bp++) {
 24143	        bp->b_blocknr = NO_BLOCK;
 24144	        bp->b_dev = NO_DEV;
 24145	        bp->b_next = bp + 1;
 24146	        bp->b_prev = bp - 1;
 24147	  }
 24148	  buf[0].b_prev = NIL_BUF;
 24149	  buf[NR_BUFS - 1].b_next = NIL_BUF;

_________________________ Page 960    File: servers/fs/main.c _________________________
 24150	
 24151	  for (bp = &buf[0]; bp < &buf[NR_BUFS]; bp++) bp->b_hash = bp->b_next;
 24152	  buf_hash[0] = front;
 24153	
 24154	}
	
 24156	/*===========================================================================*
 24157	 *                              reply                                        *
 24158	 *===========================================================================*/
 24159	PUBLIC void reply(whom, result)
 24160	int whom;                       /* process to reply to */
 24161	int result;                     /* result of the call (usually OK or error #) */
 24162	{
 24163	/* Send a reply to a user process. It may fail (if the process has just
 24164	 * been killed by a signal), so don't check the return code.  If the send
 24165	 * fails, just ignore it.
 24166	 */
 24167	  int s;
 24168	  m_out.reply_type = result;
 24169	  s = send(whom, &m_out);
 24170	  if (s != OK) printf("FS: couldn't send reply %d: %d\n", result, s);
 24171	}
	
 24173	/*===========================================================================*
 24174	 *                              fs_init                                      *
 24175	 *===========================================================================*/
 24176	PRIVATE void fs_init()
 24177	{
 24178	/* Initialize global variables, tables, etc. */
 24179	  register struct inode *rip;
 24180	  register struct fproc *rfp;
 24181	  message mess;
 24182	  int s;
 24183	
 24184	  /* Initialize the process table with help of the process manager messages. 
 24185	   * Expect one message for each system process with its slot number and pid. 
 24186	   * When no more processes follow, the magic process number NONE is sent. 
 24187	   * Then, stop and synchronize with the PM.
 24188	   */
 24189	  do {
 24190	        if (OK != (s=receive(PM_PROC_NR, &mess)))
 24191	                panic(__FILE__,"FS couldn't receive from PM", s);
 24192	        if (NONE == mess.PR_PROC_NR) break; 
 24193	
 24194	        rfp = &fproc[mess.PR_PROC_NR];
 24195	        rfp->fp_pid = mess.PR_PID;
 24196	        rfp->fp_realuid = (uid_t) SYS_UID;
 24197	        rfp->fp_effuid = (uid_t) SYS_UID;
 24198	        rfp->fp_realgid = (gid_t) SYS_GID;
 24199	        rfp->fp_effgid = (gid_t) SYS_GID;
 24200	        rfp->fp_umask = ~0;
 24201	   
 24202	  } while (TRUE);                       /* continue until process NONE */
 24203	  mess.m_type = OK;                     /* tell PM that we succeeded */
 24204	  s=send(PM_PROC_NR, &mess);            /* send synchronization message */
 24205	
 24206	  /* All process table entries have been set. Continue with FS initialization.
 24207	   * Certain relations must hold for the file system to work at all. Some 
 24208	   * extra block_size requirements are checked at super-block-read-in time.
 24209	   */

_________________________ Page 961    File: servers/fs/main.c _________________________
 24210	  if (OPEN_MAX > 127) panic(__FILE__,"OPEN_MAX > 127", NO_NUM);
 24211	  if (NR_BUFS < 6) panic(__FILE__,"NR_BUFS < 6", NO_NUM);
 24212	  if (V1_INODE_SIZE != 32) panic(__FILE__,"V1 inode size != 32", NO_NUM);
 24213	  if (V2_INODE_SIZE != 64) panic(__FILE__,"V2 inode size != 64", NO_NUM);
 24214	  if (OPEN_MAX > 8 * sizeof(long))
 24215	         panic(__FILE__,"Too few bits in fp_cloexec", NO_NUM);
 24216	
 24217	  /* The following initializations are needed to let dev_opcl succeed .*/
 24218	  fp = (struct fproc *) NULL;
 24219	  who = FS_PROC_NR;
 24220	
 24221	  buf_pool();                   /* initialize buffer pool */
 24222	  build_dmap();                 /* build device table and map boot driver */
 24223	  load_ram();                   /* init RAM disk, load if it is root */
 24224	  load_super(root_dev);         /* load super block for root device */
 24225	  init_select();                /* init select() structures */
 24226	
 24227	  /* The root device can now be accessed; set process directories. */
 24228	  for (rfp=&fproc[0]; rfp < &fproc[NR_PROCS]; rfp++) {
 24229	        if (rfp->fp_pid != PID_FREE) {
 24230	                rip = get_inode(root_dev, ROOT_INODE);
 24231	                dup_inode(rip);
 24232	                rfp->fp_rootdir = rip;
 24233	                rfp->fp_workdir = rip;
 24234	        }
 24235	  }
 24236	}
	
 24238	/*===========================================================================*
 24239	 *                              igetenv                                      *
 24240	 *===========================================================================*/
 24241	PRIVATE int igetenv(key, optional)
 24242	char *key;
 24243	int optional;
 24244	{
 24245	/* Ask kernel for an integer valued boot environment variable. */
 24246	  char value[64];
 24247	  int i;
 24248	
 24249	  if ((i = env_get_param(key, value, sizeof(value))) != OK) {
 24250	      if (!optional)
 24251	        printf("FS: Warning, couldn't get monitor param: %d\n", i);
 24252	      return 0;
 24253	  }
 24254	  return(atoi(value));
 24255	}
	
 24257	/*===========================================================================*
 24258	 *                              load_ram                                     *
 24259	 *===========================================================================*/
 24260	PRIVATE void load_ram(void)
 24261	{
 24262	/* Allocate a RAM disk with size given in the boot parameters. If a RAM disk 
 24263	 * image is given, the copy the entire image device block-by-block to a RAM 
 24264	 * disk with the same size as the image.
 24265	 * If the root device is not set, the RAM disk will be used as root instead. 
 24266	 */
 24267	  register struct buf *bp, *bp1;
 24268	  u32_t lcount, ram_size_kb;
 24269	  zone_t zones;

_________________________ Page 962    File: servers/fs/main.c _________________________
 24270	  struct super_block *sp, *dsp;
 24271	  block_t b;
 24272	  Dev_t image_dev;
 24273	  static char sbbuf[MIN_BLOCK_SIZE];
 24274	  int block_size_image, block_size_ram, ramfs_block_size;
 24275	  int s;
 24276	
 24277	  /* Get some boot environment variables. */
 24278	  root_dev = igetenv("rootdev", 0);
 24279	  image_dev = igetenv("ramimagedev", 0);
 24280	  ram_size_kb = igetenv("ramsize", 0);
 24281	
 24282	  /* Open the root device. */
 24283	  if (dev_open(root_dev, FS_PROC_NR, R_BIT|W_BIT) != OK)
 24284	        panic(__FILE__,"Cannot open root device",NO_NUM);
 24285	
 24286	  /* If we must initialize a ram disk, get details from the image device. */
 24287	  if (root_dev == DEV_RAM) {
 24288	        u32_t fsmax, probedev;
 24289	
 24290	        /* If we are running from CD, see if we can find it. */
 24291	        if (igetenv("cdproberoot", 1) && (probedev=cdprobe()) != NO_DEV) {
 24292	                char devnum[10];
 24293	                struct sysgetenv env;
 24294	
 24295	                /* If so, this is our new RAM image device. */
 24296	                image_dev = probedev;
 24297	
 24298	                /* Tell PM about it, so userland can find out about it
 24299	                 * with sysenv interface.
 24300	                 */
 24301	                env.key = "cdproberoot";
 24302	                env.keylen = strlen(env.key);
 24303	                sprintf(devnum, "%d", (int) probedev);
 24304	                env.val = devnum;
 24305	                env.vallen = strlen(devnum);
 24306	                svrctl(MMSETPARAM, &env);
 24307	        }
 24308	
 24309	        /* Open image device for RAM root. */
 24310	        if (dev_open(image_dev, FS_PROC_NR, R_BIT) != OK)
 24311	                panic(__FILE__,"Cannot open RAM image device", NO_NUM);
 24312	
 24313	        /* Get size of RAM disk image from the super block. */
 24314	        sp = &super_block[0];
 24315	        sp->s_dev = image_dev;
 24316	        if (read_super(sp) != OK) 
 24317	                panic(__FILE__,"Bad RAM disk image FS", NO_NUM);
 24318	
 24319	        lcount = sp->s_zones << sp->s_log_zone_size;    /* # blks on root dev
 24320	
 24321	        /* Stretch the RAM disk file system to the boot parameters size, but
 24322	         * no further than the last zone bit map block allows.
 24323	         */
 24324	        if (ram_size_kb*1024 < lcount*sp->s_block_size)
 24325	                ram_size_kb = lcount*sp->s_block_size/1024;
 24326	        fsmax = (u32_t) sp->s_zmap_blocks * CHAR_BIT * sp->s_block_size;
 24327	        fsmax = (fsmax + (sp->s_firstdatazone-1)) << sp->s_log_zone_size;
 24328	        if (ram_size_kb*1024 > fsmax*sp->s_block_size)
 24329	                ram_size_kb = fsmax*sp->s_block_size/1024;

_________________________ Page 963    File: servers/fs/main.c _________________________
 24330	  }
 24331	
 24332	  /* Tell RAM driver how big the RAM disk must be. */
 24333	  m_out.m_type = DEV_IOCTL;
 24334	  m_out.PROC_NR = FS_PROC_NR;
 24335	  m_out.DEVICE = RAM_DEV;
 24336	  m_out.REQUEST = MIOCRAMSIZE;                  /* I/O control to use */
 24337	  m_out.POSITION = (ram_size_kb * 1024);        /* request in bytes */
 24338	  if ((s=sendrec(MEM_PROC_NR, &m_out)) != OK)
 24339	        panic("FS","sendrec from MEM failed", s);
 24340	  else if (m_out.REP_STATUS != OK) {
 24341	        /* Report and continue, unless RAM disk is required as root FS. */
 24342	        if (root_dev != DEV_RAM) {
 24343	                report("FS","can't set RAM disk size", m_out.REP_STATUS);
 24344	                return;
 24345	        } else {
 24346	                panic(__FILE__,"can't set RAM disk size", m_out.REP_STATUS);
 24347	        }
 24348	  }
 24349	
 24350	  /* See if we must load the RAM disk image, otherwise return. */
 24351	  if (root_dev != DEV_RAM)
 24352	        return;
 24353	
 24354	  /* Copy the blocks one at a time from the image to the RAM disk. */
 24355	  printf("Loading RAM disk onto /dev/ram:\33[23CLoaded:    0 KB");
 24356	
 24357	  inode[0].i_mode = I_BLOCK_SPECIAL;    /* temp inode for rahead() */
 24358	  inode[0].i_size = LONG_MAX;
 24359	  inode[0].i_dev = image_dev;
 24360	  inode[0].i_zone[0] = image_dev;
 24361	
 24362	  block_size_ram = get_block_size(DEV_RAM);
 24363	  block_size_image = get_block_size(image_dev);
 24364	
 24365	  /* RAM block size has to be a multiple of the root image block
 24366	   * size to make copying easier.
 24367	   */
 24368	  if (block_size_image % block_size_ram) {
 24369	        printf("\nram block size: %d image block size: %d\n", 
 24370	                block_size_ram, block_size_image);
 24371	        panic(__FILE__, "ram disk block size must be a multiple of "
 24372	                "the image disk block size", NO_NUM);
 24373	  }
 24374	
 24375	  /* Loading blocks from image device. */
 24376	  for (b = 0; b < (block_t) lcount; b++) {
 24377	        int rb, factor;
 24378	        bp = rahead(&inode[0], b, (off_t)block_size_image * b, block_size_image
 24379	        factor = block_size_image/block_size_ram;
 24380	        for(rb = 0; rb < factor; rb++) {
 24381	                bp1 = get_block(root_dev, b * factor + rb, NO_READ);
 24382	                memcpy(bp1->b_data, bp->b_data + rb * block_size_ram,
 24383	                        (size_t) block_size_ram);
 24384	                bp1->b_dirt = DIRTY;
 24385	                put_block(bp1, FULL_DATA_BLOCK);
 24386	        }
 24387	        put_block(bp, FULL_DATA_BLOCK);
 24388	        if (b % 11 == 0)
 24389	        printf("\b\b\b\b\b\b\b\b\b%6ld KB", ((long) b * block_size_image)/1024L);

_________________________ Page 964    File: servers/fs/main.c _________________________
 24390	  }
 24391	
 24392	  /* Commit changes to RAM so dev_io will see it. */
 24393	  do_sync();
 24394	
 24395	  printf("\rRAM disk of %u KB loaded onto /dev/ram.", (unsigned) ram_size_kb);
 24396	  if (root_dev == DEV_RAM) printf(" Using RAM disk as root FS.");
 24397	  printf("  \n");
 24398	
 24399	  /* Invalidate and close the image device. */
 24400	  invalidate(image_dev);
 24401	  dev_close(image_dev);
 24402	
 24403	  /* Resize the RAM disk root file system. */
 24404	  if (dev_io(DEV_READ, root_dev, FS_PROC_NR,
 24405	        sbbuf, SUPER_BLOCK_BYTES, MIN_BLOCK_SIZE, 0) != MIN_BLOCK_SIZE) {
 24406	        printf("WARNING: ramdisk read for resizing failed\n");
 24407	  }
 24408	  dsp = (struct super_block *) sbbuf;
 24409	  if (dsp->s_magic == SUPER_V3)
 24410	        ramfs_block_size = dsp->s_block_size;
 24411	  else
 24412	        ramfs_block_size = STATIC_BLOCK_SIZE;
 24413	  zones = (ram_size_kb * 1024 / ramfs_block_size) >> sp->s_log_zone_size;
 24414	
 24415	  dsp->s_nzones = conv2(sp->s_native, (u16_t) zones);
 24416	  dsp->s_zones = conv4(sp->s_native, zones);
 24417	  if (dev_io(DEV_WRITE, root_dev, FS_PROC_NR,
 24418	        sbbuf, SUPER_BLOCK_BYTES, MIN_BLOCK_SIZE, 0) != MIN_BLOCK_SIZE) {
 24419	        printf("WARNING: ramdisk write for resizing failed\n");
 24420	  }
 24421	}
	
 24423	/*===========================================================================*
 24424	 *                              load_super                                   *
 24425	 *===========================================================================*/
 24426	PRIVATE void load_super(super_dev)
 24427	dev_t super_dev;                        /* place to get superblock from */
 24428	{
 24429	  int bad;
 24430	  register struct super_block *sp;
 24431	  register struct inode *rip;
 24432	
 24433	  /* Initialize the super_block table. */
 24434	  for (sp = &super_block[0]; sp < &super_block[NR_SUPERS]; sp++)
 24435	        sp->s_dev = NO_DEV;
 24436	
 24437	  /* Read in super_block for the root file system. */
 24438	  sp = &super_block[0];
 24439	  sp->s_dev = super_dev;
 24440	
 24441	  /* Check super_block for consistency. */
 24442	  bad = (read_super(sp) != OK);
 24443	  if (!bad) {
 24444	        rip = get_inode(super_dev, ROOT_INODE); /* inode for root dir */
 24445	        if ( (rip->i_mode & I_TYPE) != I_DIRECTORY || rip->i_nlinks < 3) bad
 24446	  }
 24447	  if (bad) panic(__FILE__,"Invalid root file system", NO_NUM);
 24448	
 24449	  sp->s_imount = rip;

_________________________ Page 965    File: servers/fs/main.c _________________________
 24450	  dup_inode(rip);
 24451	  sp->s_isup = rip;
 24452	  sp->s_rd_only = 0;
 24453	  return;
 24454	}

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/open.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 24500	/* This file contains the procedures for creating, opening, closing, and
 24501	 * seeking on files.
 24502	 *
 24503	 * The entry points into this file are
 24504	 *   do_creat:  perform the CREAT system call
 24505	 *   do_open:   perform the OPEN system call
 24506	 *   do_mknod:  perform the MKNOD system call
 24507	 *   do_mkdir:  perform the MKDIR system call
 24508	 *   do_close:  perform the CLOSE system call
 24509	 *   do_lseek:  perform the LSEEK system call
 24510	 */
 24511	
 24512	#include "fs.h"
 24513	#include <sys/stat.h>
 24514	#include <fcntl.h>
 24515	#include <minix/callnr.h>
 24516	#include <minix/com.h>
 24517	#include "buf.h"
 24518	#include "file.h"
 24519	#include "fproc.h"
 24520	#include "inode.h"
 24521	#include "lock.h"
 24522	#include "param.h"
 24523	#include "super.h"
 24524	
 24525	#define offset m2_l1
 24526	
 24527	PRIVATE char mode_map[] = {R_BIT, W_BIT, R_BIT|W_BIT, 0};
 24528	
 24529	FORWARD _PROTOTYPE( int common_open, (int oflags, mode_t omode)         );
 24530	FORWARD _PROTOTYPE( int pipe_open, (struct inode *rip,mode_t bits,int oflags));
 24531	FORWARD _PROTOTYPE( struct inode *new_node, (char *path, mode_t bits,
 24532	                                                        zone_t z0)      );
 24533	
 24534	/*===========================================================================*
 24535	 *                              do_creat                                     *
 24536	 *===========================================================================*/
 24537	PUBLIC int do_creat()
 24538	{
 24539	/* Perform the creat(name, mode) system call. */
 24540	  int r;
 24541	
 24542	  if (fetch_name(m_in.name, m_in.name_length, M3) != OK) return(err_code);
 24543	  r = common_open(O_WRONLY | O_CREAT | O_TRUNC, (mode_t) m_in.mode);
 24544	  return(r);
 24545	}
	

_________________________ Page 966    File: servers/fs/open.c _________________________
 24547	/*===========================================================================*
 24548	 *                              do_open                                      *
 24549	 *===========================================================================*/
 24550	PUBLIC int do_open()
 24551	{
 24552	/* Perform the open(name, flags,...) system call. */
 24553	
 24554	  int create_mode = 0;          /* is really mode_t but this gives problems */
 24555	  int r;
 24556	
 24557	  /* If O_CREAT is set, open has three parameters, otherwise two. */
 24558	  if (m_in.mode & O_CREAT) {
 24559	        create_mode = m_in.c_mode;      
 24560	        r = fetch_name(m_in.c_name, m_in.name1_length, M1);
 24561	  } else {
 24562	        r = fetch_name(m_in.name, m_in.name_length, M3);
 24563	  }
 24564	
 24565	  if (r != OK) return(err_code); /* name was bad */
 24566	  r = common_open(m_in.mode, create_mode);
 24567	  return(r);
 24568	}
	
 24570	/*===========================================================================*
 24571	 *                              common_open                                  *
 24572	 *===========================================================================*/
 24573	PRIVATE int common_open(register int oflags, mode_t omode)
 24574	{
 24575	/* Common code from do_creat and do_open. */
 24576	
 24577	  register struct inode *rip;
 24578	  int r, b, exist = TRUE;
 24579	  dev_t dev;
 24580	  mode_t bits;
 24581	  off_t pos;
 24582	  struct filp *fil_ptr, *filp2;
 24583	
 24584	  /* Remap the bottom two bits of oflags. */
 24585	  bits = (mode_t) mode_map[oflags & O_ACCMODE];
 24586	
 24587	  /* See if file descriptor and filp slots are available. */
 24588	  if ( (r = get_fd(0, bits, &m_in.fd, &fil_ptr)) != OK) return(r);
 24589	
 24590	  /* If O_CREATE is set, try to make the file. */ 
 24591	  if (oflags & O_CREAT) {
 24592	        /* Create a new inode by calling new_node(). */
 24593	        omode = I_REGULAR | (omode & ALL_MODES & fp->fp_umask);
 24594	        rip = new_node(user_path, omode, NO_ZONE);
 24595	        r = err_code;
 24596	        if (r == OK) exist = FALSE;      /* we just created the file */
 24597	        else if (r != EEXIST) return(r); /* other error */
 24598	        else exist = !(oflags & O_EXCL); /* file exists, if the O_EXCL 
 24599	                                            flag is set this is an error */
 24600	  } else {
 24601	         /* Scan path name. */
 24602	        if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 24603	  }
 24604	
 24605	  /* Claim the file descriptor and filp slot and fill them in. */
 24606	  fp->fp_filp[m_in.fd] = fil_ptr;

_________________________ Page 967    File: servers/fs/open.c _________________________
 24607	  fil_ptr->filp_count = 1;
 24608	  fil_ptr->filp_ino = rip;
 24609	  fil_ptr->filp_flags = oflags;
 24610	
 24611	  /* Only do the normal open code if we didn't just create the file. */
 24612	  if (exist) {
 24613	        /* Check protections. */
 24614	        if ((r = forbidden(rip, bits)) == OK) {
 24615	                /* Opening reg. files directories and special files differ. */
 24616	                switch (rip->i_mode & I_TYPE) {
 24617	                   case I_REGULAR: 
 24618	                        /* Truncate regular file if O_TRUNC. */
 24619	                        if (oflags & O_TRUNC) {
 24620	                                if ((r = forbidden(rip, W_BIT)) !=OK) break;
 24621	                                truncate(rip);
 24622	                                wipe_inode(rip);
 24623	                                /* Send the inode from the inode cache to the
 24624	                                 * block cache, so it gets written on the next
 24625	                                 * cache flush.
 24626	                                 */
 24627	                                rw_inode(rip, WRITING);
 24628	                        }
 24629	                        break;
 24630	 
 24631	                   case I_DIRECTORY: 
 24632	                        /* Directories may be read but not written. */
 24633	                        r = (bits & W_BIT ? EISDIR : OK);
 24634	                        break;
 24635	
 24636	                   case I_CHAR_SPECIAL:
 24637	                   case I_BLOCK_SPECIAL:
 24638	                        /* Invoke the driver for special processing. */
 24639	                        dev = (dev_t) rip->i_zone[0];
 24640	                        r = dev_open(dev, who, bits | (oflags & ~O_ACCMODE));
 24641	                        break;
 24642	
 24643	                   case I_NAMED_PIPE:
 24644	                        oflags |= O_APPEND;     /* force append mode */
 24645	                        fil_ptr->filp_flags = oflags;
 24646	                        r = pipe_open(rip, bits, oflags);
 24647	                        if (r != ENXIO) {
 24648	                                /* See if someone else is doing a rd or wt on
 24649	                                 * the FIFO.  If so, use its filp entry so the
 24650	                                 * file position will be automatically shared.
 24651	                                 */
 24652	                                b = (bits & R_BIT ? R_BIT : W_BIT);
 24653	                                fil_ptr->filp_count = 0; /* don't find self */
 24654	                                if ((filp2 = find_filp(rip, b)) != NIL_FILP) {
 24655	                                        /* Co-reader or writer found. Use it.*/
 24656	                                        fp->fp_filp[m_in.fd] = filp2;
 24657	                                        filp2->filp_count++;
 24658	                                        filp2->filp_ino = rip;
 24659	                                        filp2->filp_flags = oflags;
 24660	
 24661	                                        /* i_count was incremented incorrectly
 24662	                                         * by eatpath above, not knowing that
 24663	                                         * we were going to use an existing
 24664	                                         * filp entry.  Correct this error.
 24665	                                         */
 24666	                                        rip->i_count--;

_________________________ Page 968    File: servers/fs/open.c _________________________
 24667	                                } else {
 24668	                                        /* Nobody else found.  Restore filp. */
 24669	                                        fil_ptr->filp_count = 1;
 24670	                                        if (b == R_BIT)
 24671	                                             pos = rip->i_zone[V2_NR_DZONES+0];
 24672	                                        else
 24673	                                             pos = rip->i_zone[V2_NR_DZONES+1];
 24674	                                        fil_ptr->filp_pos = pos;
 24675	                                }
 24676	                        }
 24677	                        break;
 24678	                }
 24679	        }
 24680	  }
 24681	
 24682	  /* If error, release inode. */
 24683	  if (r != OK) {
 24684	        if (r == SUSPEND) return(r);            /* Oops, just suspended */
 24685	        fp->fp_filp[m_in.fd] = NIL_FILP;
 24686	        fil_ptr->filp_count= 0;
 24687	        put_inode(rip);
 24688	        return(r);
 24689	  }
 24690	  
 24691	  return(m_in.fd);
 24692	}
	
 24694	/*===========================================================================*
 24695	 *                              new_node                                     *
 24696	 *===========================================================================*/
 24697	PRIVATE struct inode *new_node(char *path, mode_t bits, zone_t z0)
 24698	{
 24699	/* New_node() is called by common_open(), do_mknod(), and do_mkdir().  
 24700	 * In all cases it allocates a new inode, makes a directory entry for it on 
 24701	 * the path 'path', and initializes it.  It returns a pointer to the inode if 
 24702	 * it can do this; otherwise it returns NIL_INODE.  It always sets 'err_code'
 24703	 * to an appropriate value (OK or an error code).
 24704	 */
 24705	
 24706	  register struct inode *rlast_dir_ptr, *rip;
 24707	  register int r;
 24708	  char string[NAME_MAX];
 24709	
 24710	  /* See if the path can be opened down to the last directory. */
 24711	  if ((rlast_dir_ptr = last_dir(path, string)) == NIL_INODE) return(NIL_INODE);
 24712	
 24713	  /* The final directory is accessible. Get final component of the path. */
 24714	  rip = advance(rlast_dir_ptr, string);
 24715	  if ( rip == NIL_INODE && err_code == ENOENT) {
 24716	        /* Last path component does not exist.  Make new directory entry. */
 24717	        if ( (rip = alloc_inode(rlast_dir_ptr->i_dev, bits)) == NIL_INODE) {
 24718	                /* Can't creat new inode: out of inodes. */
 24719	                put_inode(rlast_dir_ptr);
 24720	                return(NIL_INODE);
 24721	        }
 24722	
 24723	        /* Force inode to the disk before making directory entry to make
 24724	         * the system more robust in the face of a crash: an inode with
 24725	         * no directory entry is much better than the opposite.
 24726	         */

_________________________ Page 969    File: servers/fs/open.c _________________________
 24727	        rip->i_nlinks++;
 24728	        rip->i_zone[0] = z0;            /* major/minor device numbers */
 24729	        rw_inode(rip, WRITING);         /* force inode to disk now */
 24730	
 24731	        /* New inode acquired.  Try to make directory entry. */
 24732	        if ((r = search_dir(rlast_dir_ptr, string, &rip->i_num,ENTER)) != OK) {
 24733	                put_inode(rlast_dir_ptr);
 24734	                rip->i_nlinks--;        /* pity, have to free disk inode */
 24735	                rip->i_dirt = DIRTY;    /* dirty inodes are written out */
 24736	                put_inode(rip); /* this call frees the inode */
 24737	                err_code = r;
 24738	                return(NIL_INODE);
 24739	        }
 24740	
 24741	  } else {
 24742	        /* Either last component exists, or there is some problem. */
 24743	        if (rip != NIL_INODE)
 24744	                r = EEXIST;
 24745	        else
 24746	                r = err_code;
 24747	  }
 24748	
 24749	  /* Return the directory inode and exit. */
 24750	  put_inode(rlast_dir_ptr);
 24751	  err_code = r;
 24752	  return(rip);
 24753	}
	
 24755	/*===========================================================================*
 24756	 *                              pipe_open                                    *
 24757	 *===========================================================================*/
 24758	PRIVATE int pipe_open(register struct inode *rip, register mode_t bits,
 24759	        register int oflags)
 24760	{
 24761	/*  This function is called from common_open. It checks if
 24762	 *  there is at least one reader/writer pair for the pipe, if not
 24763	 *  it suspends the caller, otherwise it revives all other blocked
 24764	 *  processes hanging on the pipe.
 24765	 */
 24766	
 24767	  rip->i_pipe = I_PIPE; 
 24768	  if (find_filp(rip, bits & W_BIT ? R_BIT : W_BIT) == NIL_FILP) { 
 24769	        if (oflags & O_NONBLOCK) {
 24770	                if (bits & W_BIT) return(ENXIO);
 24771	        } else {
 24772	                suspend(XPOPEN);        /* suspend caller */
 24773	                return(SUSPEND);
 24774	        }
 24775	  } else if (susp_count > 0) {/* revive blocked processes */
 24776	        release(rip, OPEN, susp_count);
 24777	        release(rip, CREAT, susp_count);
 24778	  }
 24779	  return(OK);
 24780	}
	
 24782	/*===========================================================================*
 24783	 *                              do_mknod                                     *
 24784	 *===========================================================================*/
 24785	PUBLIC int do_mknod()
 24786	{

_________________________ Page 970    File: servers/fs/open.c _________________________
 24787	/* Perform the mknod(name, mode, addr) system call. */
 24788	
 24789	  register mode_t bits, mode_bits;
 24790	  struct inode *ip;
 24791	
 24792	  /* Only the super_user may make nodes other than fifos. */
 24793	  mode_bits = (mode_t) m_in.mk_mode;            /* mode of the inode */
 24794	  if (!super_user && ((mode_bits & I_TYPE) != I_NAMED_PIPE)) return(EPE
 24795	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 24796	  bits = (mode_bits & I_TYPE) | (mode_bits & ALL_MODES & fp->fp_umask);
 24797	  ip = new_node(user_path, bits, (zone_t) m_in.mk_z0);
 24798	  put_inode(ip);
 24799	  return(err_code);
 24800	}
	
 24802	/*===========================================================================*
 24803	 *                              do_mkdir                                     *
 24804	 *===========================================================================*/
 24805	PUBLIC int do_mkdir()
 24806	{
 24807	/* Perform the mkdir(name, mode) system call. */
 24808	
 24809	  int r1, r2;                   /* status codes */
 24810	  ino_t dot, dotdot;            /* inode numbers for . and .. */
 24811	  mode_t bits;                  /* mode bits for the new inode */
 24812	  char string[NAME_MAX];        /* last component of the new dir's path name */
 24813	  register struct inode *rip, *ldirp;
 24814	
 24815	  /* Check to see if it is possible to make another link in the parent dir. */
 24816	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 24817	  ldirp = last_dir(user_path, string);  /* pointer to new dir's parent */
 24818	  if (ldirp == NIL_INODE) return(err_code);
 24819	  if (ldirp->i_nlinks >= (ldirp->i_sp->s_version == V1 ?
 24820	         CHAR_MAX : SHRT_MAX)) {
 24821	        put_inode(ldirp);       /* return parent */
 24822	        return(EMLINK);
 24823	  }
 24824	
 24825	  /* Next make the inode. If that fails, return error code. */
 24826	  bits = I_DIRECTORY | (m_in.mode & RWX_MODES & fp->fp_umask);
 24827	  rip = new_node(user_path, bits, (zone_t) 0);
 24828	  if (rip == NIL_INODE || err_code == EEXIST) {
 24829	        put_inode(rip);         /* can't make dir: it already exists */
 24830	        put_inode(ldirp);       /* return parent too */
 24831	        return(err_code);
 24832	  }
 24833	
 24834	  /* Get the inode numbers for . and .. to enter in the directory. */
 24835	  dotdot = ldirp->i_num;        /* parent's inode number */
 24836	  dot = rip->i_num;             /* inode number of the new dir itself */
 24837	
 24838	  /* Now make dir entries for . and .. unless the disk is completely full. */
 24839	  /* Use dot1 and dot2, so the mode of the directory isn't important. */
 24840	  rip->i_mode = bits;   /* set mode */
 24841	  r1 = search_dir(rip, dot1, &dot, ENTER);      /* enter . in the new dir */
 24842	  r2 = search_dir(rip, dot2, &dotdot, ENTER);   /* enter .. in the new dir */
 24843	
 24844	  /* If both . and .. were successfully entered, increment the link counts. */
 24845	  if (r1 == OK && r2 == OK) {
 24846	        /* Normal case.  It was possible to enter . and .. in the new dir. */

_________________________ Page 971    File: servers/fs/open.c _________________________
 24847	        rip->i_nlinks++;        /* this accounts for . */
 24848	        ldirp->i_nlinks++;      /* this accounts for .. */
 24849	        ldirp->i_dirt = DIRTY;  /* mark parent's inode as dirty */
 24850	  } else {
 24851	        /* It was not possible to enter . or .. probably disk was full. */
 24852	        (void) search_dir(ldirp, string, (ino_t *) 0, DELETE);
 24853	        rip->i_nlinks--;        /* undo the increment done in new_node() */
 24854	  }
 24855	  rip->i_dirt = DIRTY;          /* either way, i_nlinks has changed */
 24856	
 24857	  put_inode(ldirp);             /* return the inode of the parent dir */
 24858	  put_inode(rip);               /* return the inode of the newly made dir */
 24859	  return(err_code);             /* new_node() always sets 'err_code' */
 24860	}
	
 24862	/*===========================================================================*
 24863	 *                              do_close                                     *
 24864	 *===========================================================================*/
 24865	PUBLIC int do_close()
 24866	{
 24867	/* Perform the close(fd) system call. */
 24868	
 24869	  register struct filp *rfilp;
 24870	  register struct inode *rip;
 24871	  struct file_lock *flp;
 24872	  int rw, mode_word, lock_count;
 24873	  dev_t dev;
 24874	
 24875	  /* First locate the inode that belongs to the file descriptor. */
 24876	  if ( (rfilp = get_filp(m_in.fd)) == NIL_FILP) return(err_code);
 24877	  rip = rfilp->filp_ino;        /* 'rip' points to the inode */
 24878	
 24879	  if (rfilp->filp_count - 1 == 0 && rfilp->filp_mode != FILP_CLOSED) {
 24880	        /* Check to see if the file is special. */
 24881	        mode_word = rip->i_mode & I_TYPE;
 24882	        if (mode_word == I_CHAR_SPECIAL || mode_word == I_BLOCK_SPECIAL) {
 24883	                dev = (dev_t) rip->i_zone[0];
 24884	                if (mode_word == I_BLOCK_SPECIAL)  {
 24885	                        /* Invalidate cache entries unless special is mounted
 24886	                         * or ROOT
 24887	                         */
 24888	                        if (!mounted(rip)) {
 24889	                                (void) do_sync();       /* purge cache */
 24890	                                invalidate(dev);
 24891	                        }    
 24892	                }
 24893	                /* Do any special processing on device close. */
 24894	                dev_close(dev);
 24895	        }
 24896	  }
 24897	
 24898	  /* If the inode being closed is a pipe, release everyone hanging on it. */
 24899	  if (rip->i_pipe == I_PIPE) {
 24900	        rw = (rfilp->filp_mode & R_BIT ? WRITE : READ);
 24901	        release(rip, rw, NR_PROCS);
 24902	  }
 24903	
 24904	  /* If a write has been done, the inode is already marked as DIRTY. */
 24905	  if (--rfilp->filp_count == 0) {
 24906	        if (rip->i_pipe == I_PIPE && rip->i_count > 1) {

_________________________ Page 972    File: servers/fs/open.c _________________________
 24907	                /* Save the file position in the i-node in case needed later.
 24908	                 * The read and write positions are saved separately.  The
 24909	                 * last 3 zones in the i-node are not used for (named) pipes.
 24910	                 */
 24911	                if (rfilp->filp_mode == R_BIT)
 24912	                        rip->i_zone[V2_NR_DZONES+0] = (zone_t) rfilp->filp_pos;
 24913	                else
 24914	                        rip->i_zone[V2_NR_DZONES+1] = (zone_t) rfilp->filp_pos;
 24915	        }
 24916	        put_inode(rip);
 24917	  }
 24918	
 24919	  fp->fp_cloexec &= ~(1L << m_in.fd);   /* turn off close-on-exec bit */
 24920	  fp->fp_filp[m_in.fd] = NIL_FILP;
 24921	
 24922	  /* Check to see if the file is locked.  If so, release all locks. */
 24923	  if (nr_locks == 0) return(OK);
 24924	  lock_count = nr_locks;        /* save count of locks */
 24925	  for (flp = &file_lock[0]; flp < &file_lock[NR_LOCKS]; flp++) {
 24926	        if (flp->lock_type == 0) continue;      /* slot not in use */
 24927	        if (flp->lock_inode == rip && flp->lock_pid == fp->fp_pid) {
 24928	                flp->lock_type = 0;
 24929	                nr_locks--;
 24930	        }
 24931	  }
 24932	  if (nr_locks < lock_count) lock_revive();     /* lock released */
 24933	  return(OK);
 24934	}
	
 24936	/*===========================================================================*
 24937	 *                              do_lseek                                     *
 24938	 *===========================================================================*/
 24939	PUBLIC int do_lseek()
 24940	{
 24941	/* Perform the lseek(ls_fd, offset, whence) system call. */
 24942	
 24943	  register struct filp *rfilp;
 24944	  register off_t pos;
 24945	
 24946	  /* Check to see if the file descriptor is valid. */
 24947	  if ( (rfilp = get_filp(m_in.ls_fd)) == NIL_FILP) return(err_code);
 24948	
 24949	  /* No lseek on pipes. */
 24950	  if (rfilp->filp_ino->i_pipe == I_PIPE) return(ESPIPE);
 24951	
 24952	  /* The value of 'whence' determines the start position to use. */
 24953	  switch(m_in.whence) {
 24954	        case 0: pos = 0;        break;
 24955	        case 1: pos = rfilp->filp_pos;  break;
 24956	        case 2: pos = rfilp->filp_ino->i_size;  break;
 24957	        default: return(EINVAL);
 24958	  }
 24959	
 24960	  /* Check for overflow. */
 24961	  if (((long)m_in.offset > 0) && ((long)(pos + m_in.offset) < (long)pos)
 24962	        return(EINVAL);
 24963	  if (((long)m_in.offset < 0) && ((long)(pos + m_in.offset) > (long)pos)
 24964	        return(EINVAL);
 24965	  pos = pos + m_in.offset;
 24966	

_________________________ Page 973    File: servers/fs/open.c _________________________
 24967	  if (pos != rfilp->filp_pos)
 24968	        rfilp->filp_ino->i_seek = ISEEK;        /* inhibit read ahead */
 24969	  rfilp->filp_pos = pos;
 24970	  m_out.reply_l1 = pos;         /* insert the long into the output message */
 24971	  return(OK);
 24972	}





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/read.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 25000	/* This file contains the heart of the mechanism used to read (and write)
 25001	 * files.  Read and write requests are split up into chunks that do not cross
 25002	 * block boundaries.  Each chunk is then processed in turn.  Reads on special
 25003	 * files are also detected and handled.
 25004	 *
 25005	 * The entry points into this file are
 25006	 *   do_read:    perform the READ system call by calling read_write
 25007	 *   read_write: actually do the work of READ and WRITE
 25008	 *   read_map:   given an inode and file position, look up its zone number
 25009	 *   rd_indir:   read an entry in an indirect block 
 25010	 *   read_ahead: manage the block read ahead business
 25011	 */
 25012	
 25013	#include "fs.h"
 25014	#include <fcntl.h>
 25015	#include <minix/com.h>
 25016	#include "buf.h"
 25017	#include "file.h"
 25018	#include "fproc.h"
 25019	#include "inode.h"
 25020	#include "param.h"
 25021	#include "super.h"
 25022	
 25023	FORWARD _PROTOTYPE( int rw_chunk, (struct inode *rip, off_t position,
 25024	        unsigned off, int chunk, unsigned left, int rw_flag,
 25025	        char *buff, int seg, int usr, int block_size, int *completed));
 25026	
 25027	/*===========================================================================*
 25028	 *                              do_read                                      *
 25029	 *===========================================================================*/
 25030	PUBLIC int do_read()
 25031	{
 25032	  return(read_write(READING));
 25033	}
	
 25035	/*===========================================================================*
 25036	 *                              read_write                                   *
 25037	 *===========================================================================*/
 25038	PUBLIC int read_write(rw_flag)
 25039	int rw_flag;                    /* READING or WRITING */
 25040	{
 25041	/* Perform read(fd, buffer, nbytes) or write(fd, buffer, nbytes) call. */
 25042	
 25043	  register struct inode *rip;
 25044	  register struct filp *f;

_________________________ Page 974    File: servers/fs/read.c _________________________
 25045	  off_t bytes_left, f_size, position;
 25046	  unsigned int off, cum_io;
 25047	  int op, oflags, r, chunk, usr, seg, block_spec, char_spec;
 25048	  int regular, partial_pipe = 0, partial_cnt = 0;
 25049	  mode_t mode_word;
 25050	  struct filp *wf;
 25051	  int block_size;
 25052	  int completed, r2 = OK;
 25053	  phys_bytes p;
 25054	
 25055	  /* left unfinished rw_chunk()s from previous call! this can't happen.
 25056	   * it means something has gone wrong we can't repair now.
 25057	   */
 25058	  if (bufs_in_use < 0) {
 25059	        panic(__FILE__,"start - bufs_in_use negative", bufs_in_use);
 25060	  }
 25061	
 25062	  /* MM loads segments by putting funny things in upper 10 bits of 'fd'. */
 25063	  if (who == PM_PROC_NR && (m_in.fd & (~BYTE)) ) {
 25064	        usr = m_in.fd >> 7;
 25065	        seg = (m_in.fd >> 5) & 03;
 25066	        m_in.fd &= 037;         /* get rid of user and segment bits */
 25067	  } else {
 25068	        usr = who;              /* normal case */
 25069	        seg = D;
 25070	  }
 25071	
 25072	  /* If the file descriptor is valid, get the inode, size and mode. */
 25073	  if (m_in.nbytes < 0) return(EINVAL);
 25074	  if ((f = get_filp(m_in.fd)) == NIL_FILP) return(err_code);
 25075	  if (((f->filp_mode) & (rw_flag == READING ? R_BIT : W_BIT)) == 0) {
 25076	        return(f->filp_mode == FILP_CLOSED ? EIO : EBADF);
 25077	  }
 25078	  if (m_in.nbytes == 0)
 25079	         return(0);     /* so char special files need not check for 0*/
 25080	
 25081	  /* check if user process has the memory it needs.
 25082	   * if not, copying will fail later.
 25083	   * do this after 0-check above because umap doesn't want to map 0 bytes.
 25084	   */
 25085	  if ((r = sys_umap(usr, seg, (vir_bytes) m_in.buffer, m_in.nbytes, &p)) != OK)
 25086	        return r;
 25087	  position = f->filp_pos;
 25088	  oflags = f->filp_flags;
 25089	  rip = f->filp_ino;
 25090	  f_size = rip->i_size;
 25091	  r = OK;
 25092	  if (rip->i_pipe == I_PIPE) {
 25093	        /* fp->fp_cum_io_partial is only nonzero when doing partial writes */
 25094	        cum_io = fp->fp_cum_io_partial; 
 25095	  } else {
 25096	        cum_io = 0;
 25097	  }
 25098	  op = (rw_flag == READING ? DEV_READ : DEV_WRITE);
 25099	  mode_word = rip->i_mode & I_TYPE;
 25100	  regular = mode_word == I_REGULAR || mode_word == I_NAMED_PIPE;
 25101	
 25102	  if ((char_spec = (mode_word == I_CHAR_SPECIAL ? 1 : 0))) {
 25103	        if (rip->i_zone[0] == NO_DEV)
 25104	                panic(__FILE__,"read_write tries to read from "

_________________________ Page 975    File: servers/fs/read.c _________________________
 25105	                        "character device NO_DEV", NO_NUM);
 25106	        block_size = get_block_size(rip->i_zone[0]);
 25107	  }
 25108	  if ((block_spec = (mode_word == I_BLOCK_SPECIAL ? 1 : 0))) {
 25109	        f_size = ULONG_MAX;
 25110	        if (rip->i_zone[0] == NO_DEV)
 25111	                panic(__FILE__,"read_write tries to read from "
 25112	                " block device NO_DEV", NO_NUM);
 25113	        block_size = get_block_size(rip->i_zone[0]);
 25114	  }
 25115	
 25116	  if (!char_spec && !block_spec)
 25117	        block_size = rip->i_sp->s_block_size;
 25118	
 25119	  rdwt_err = OK;                /* set to EIO if disk error occurs */
 25120	
 25121	  /* Check for character special files. */
 25122	  if (char_spec) {
 25123	        dev_t dev;
 25124	        dev = (dev_t) rip->i_zone[0];
 25125	        r = dev_io(op, dev, usr, m_in.buffer, position, m_in.nbytes, oflags);
 25126	        if (r >= 0) {
 25127	                cum_io = r;
 25128	                position += r;
 25129	                r = OK;
 25130	        }
 25131	  } else {
 25132	        if (rw_flag == WRITING && block_spec == 0) {
 25133	                /* Check in advance to see if file will grow too big. */
 25134	                if (position > rip->i_sp->s_max_size - m_in.nbytes) 
 25135	                        return(EFBIG);
 25136	
 25137	                /* Check for O_APPEND flag. */
 25138	                if (oflags & O_APPEND) position = f_size;
 25139	
 25140	                /* Clear the zone containing present EOF if hole about
 25141	                 * to be created.  This is necessary because all unwritten
 25142	                 * blocks prior to the EOF must read as zeros.
 25143	                 */
 25144	                if (position > f_size) clear_zone(rip, f_size, 0);
 25145	        }
 25146	
 25147	        /* Pipes are a little different.  Check. */
 25148	        if (rip->i_pipe == I_PIPE) {
 25149	               r = pipe_check(rip, rw_flag, oflags,
 25150	                        m_in.nbytes, position, &partial_cnt, 0);
 25151	               if (r <= 0) return(r);
 25152	        }
 25153	
 25154	        if (partial_cnt > 0) partial_pipe = 1;
 25155	
 25156	        /* Split the transfer into chunks that don't span two blocks. */
 25157	        while (m_in.nbytes != 0) {
 25158	
 25159	                off = (unsigned int) (position % block_size);/* offset in blk*/
 25160	                if (partial_pipe) {  /* pipes only */
 25161	                        chunk = MIN(partial_cnt, block_size - off);
 25162	                } else
 25163	                        chunk = MIN(m_in.nbytes, block_size - off);
 25164	                if (chunk < 0) chunk = block_size - off;

_________________________ Page 976    File: servers/fs/read.c _________________________
 25165	
 25166	                if (rw_flag == READING) {
 25167	                        bytes_left = f_size - position;
 25168	                        if (position >= f_size) break;  /* we are beyond EOF */
 25169	                        if (chunk > bytes_left) chunk = (int) bytes_left;
 25170	                }
 25171	
 25172	                /* Read or write 'chunk' bytes. */
 25173	                r = rw_chunk(rip, position, off, chunk, (unsigned) m_in.nbytes,
 25174	                             rw_flag, m_in.buffer, seg, usr, block_size, &compl
 25175	
 25176	                if (r != OK) break;     /* EOF reached */
 25177	                if (rdwt_err < 0) break;
 25178	
 25179	                /* Update counters and pointers. */
 25180	                m_in.buffer += chunk;   /* user buffer address */
 25181	                m_in.nbytes -= chunk;   /* bytes yet to be read */
 25182	                cum_io += chunk;        /* bytes read so far */
 25183	                position += chunk;      /* position within the file */
 25184	
 25185	                if (partial_pipe) {
 25186	                        partial_cnt -= chunk;
 25187	                        if (partial_cnt <= 0)  break;
 25188	                }
 25189	        }
 25190	  }
 25191	
 25192	  /* On write, update file size and access time. */
 25193	  if (rw_flag == WRITING) {
 25194	        if (regular || mode_word == I_DIRECTORY) {
 25195	                if (position > f_size) rip->i_size = position;
 25196	        }
 25197	  } else {
 25198	        if (rip->i_pipe == I_PIPE) {
 25199	                if ( position >= rip->i_size) {
 25200	                        /* Reset pipe pointers. */
 25201	                        rip->i_size = 0;        /* no data left */
 25202	                        position = 0;           /* reset reader(s) */
 25203	                        wf = find_filp(rip, W_BIT);
 25204	                        if (wf != NIL_FILP) wf->filp_pos = 0;
 25205	                }
 25206	        }
 25207	  }
 25208	  f->filp_pos = position;
 25209	
 25210	  /* Check to see if read-ahead is called for, and if so, set it up. */
 25211	  if (rw_flag == READING && rip->i_seek == NO_SEEK && position % bl
 25212	                && (regular || mode_word == I_DIRECTORY)) {
 25213	        rdahed_inode = rip;
 25214	        rdahedpos = position;
 25215	  }
 25216	  rip->i_seek = NO_SEEK;
 25217	
 25218	  if (rdwt_err != OK) r = rdwt_err;     /* check for disk error */
 25219	  if (rdwt_err == END_OF_FILE) r = OK;
 25220	
 25221	  /* if user-space copying failed, read/write failed. */
 25222	  if (r == OK && r2 != OK) {
 25223	        r = r2;
 25224	  }

_________________________ Page 977    File: servers/fs/read.c _________________________
 25225	  if (r == OK) {
 25226	        if (rw_flag == READING) rip->i_update |= ATIME;
 25227	        if (rw_flag == WRITING) rip->i_update |= CTIME | MTIME;
 25228	        rip->i_dirt = DIRTY;            /* inode is thus now dirty */
 25229	        if (partial_pipe) {
 25230	                partial_pipe = 0;
 25231	                        /* partial write on pipe with */
 25232	                /* O_NONBLOCK, return write count */
 25233	                if (!(oflags & O_NONBLOCK)) {
 25234	                        fp->fp_cum_io_partial = cum_io;
 25235	                        suspend(XPIPE);   /* partial write on pipe with */
 25236	                        return(SUSPEND);  /* nbyte > PIPE_SIZE - non-atomic */
 25237	                }
 25238	        }
 25239	        fp->fp_cum_io_partial = 0;
 25240	        return(cum_io);
 25241	  }
 25242	  if (bufs_in_use < 0) {
 25243	        panic(__FILE__,"end - bufs_in_use negative", bufs_in_use);
 25244	  }
 25245	  return(r);
 25246	}
	
 25248	/*===========================================================================*
 25249	 *                              rw_chunk                                     *
 25250	 *===========================================================================*/
 25251	PRIVATE int rw_chunk(rip, position, off, chunk, left, rw_flag, buff,
 25252	 seg, usr, block_size, completed)
 25253	register struct inode *rip;     /* pointer to inode for file to be rd/wr */
 25254	off_t position;                 /* position within file to read or write */
 25255	unsigned off;                   /* off within the current block */
 25256	int chunk;                      /* number of bytes to read or write */
 25257	unsigned left;                  /* max number of bytes wanted after position */
 25258	int rw_flag;                    /* READING or WRITING */
 25259	char *buff;                     /* virtual address of the user buffer */
 25260	int seg;                        /* T or D segment in user space */
 25261	int usr;                        /* which user process */
 25262	int block_size;                 /* block size of FS operating on */
 25263	int *completed;                 /* number of bytes copied */
 25264	{
 25265	/* Read or write (part of) a block. */
 25266	
 25267	  register struct buf *bp;
 25268	  register int r = OK;
 25269	  int n, block_spec;
 25270	  block_t b;
 25271	  dev_t dev;
 25272	
 25273	  *completed = 0;
 25274	
 25275	  block_spec = (rip->i_mode & I_TYPE) == I_BLOCK_SPECIAL;
 25276	  if (block_spec) {
 25277	        b = position/block_size;
 25278	        dev = (dev_t) rip->i_zone[0];
 25279	  } else {
 25280	        b = read_map(rip, position);
 25281	        dev = rip->i_dev;
 25282	  }
 25283	
 25284	  if (!block_spec && b == NO_BLOCK) {

_________________________ Page 978    File: servers/fs/read.c _________________________
 25285	        if (rw_flag == READING) {
 25286	                /* Reading from a nonexistent block.  Must read as all zeros.*/
 25287	                bp = get_block(NO_DEV, NO_BLOCK, NORMAL);    /* get a buffer */
 25288	                zero_block(bp);
 25289	        } else {
 25290	                /* Writing to a nonexistent block. Create and enter in inode.*/
 25291	                if ((bp= new_block(rip, position)) == NIL_BUF)return(err_code);
 25292	        }
 25293	  } else if (rw_flag == READING) {
 25294	        /* Read and read ahead if convenient. */
 25295	        bp = rahead(rip, b, position, left);
 25296	  } else {
 25297	        /* Normally an existing block to be partially overwritten is first read
 25298	         * in.  However, a full block need not be read in.  If it is already in
 25299	         * the cache, acquire it, otherwise just acquire a free buffer.
 25300	         */
 25301	        n = (chunk == block_size ? NO_READ : NORMAL);
 25302	        if (!block_spec && off == 0 && position >= rip->i_size) n =
 25303	        bp = get_block(dev, b, n);
 25304	  }
 25305	
 25306	  /* In all cases, bp now points to a valid buffer. */
 25307	  if (bp == NIL_BUF) {
 25308	        panic(__FILE__,"bp not valid in rw_chunk, this can't happen", NO_NUM);
 25309	  }
 25310	  if (rw_flag == WRITING && chunk != block_size && !block_spec & ...
 25311	                                        position >= rip->i_size && off == 0
 25312	        zero_block(bp);
 25313	  }
 25314	
 25315	  if (rw_flag == READING) {
 25316	        /* Copy a chunk from the block buffer to user space. */
 25317	        r = sys_vircopy(FS_PROC_NR, D, (phys_bytes) (bp->b_data+off),
 25318	                        usr, seg, (phys_bytes) buff,
 25319	                        (phys_bytes) chunk);
 25320	  } else {
 25321	        /* Copy a chunk from user space to the block buffer. */
 25322	        r = sys_vircopy(usr, seg, (phys_bytes) buff,
 25323	                        FS_PROC_NR, D, (phys_bytes) (bp->b_data+off),
 25324	                        (phys_bytes) chunk);
 25325	        bp->b_dirt = DIRTY;
 25326	  }
 25327	  n = (off + chunk == block_size ? FULL_DATA_BLOCK : PARTIAL_DATA_BLOCK);
 25328	  put_block(bp, n);
 25329	
 25330	  return(r);
 25331	}
	
	
 25334	/*===========================================================================*
 25335	 *                              read_map                                     *
 25336	 *===========================================================================*/
 25337	PUBLIC block_t read_map(rip, position)
 25338	register struct inode *rip;     /* ptr to inode to map from */
 25339	off_t position;                 /* position in file whose blk wanted */
 25340	{
 25341	/* Given an inode and a position within the corresponding file, locate the
 25342	 * block (not zone) number in which that position is to be found and return it.
 25343	 */
 25344	

_________________________ Page 979    File: servers/fs/read.c _________________________
 25345	  register struct buf *bp;
 25346	  register zone_t z;
 25347	  int scale, boff, dzones, nr_indirects, index, zind, ex;
 25348	  block_t b;
 25349	  long excess, zone, block_pos;
 25350	  
 25351	  scale = rip->i_sp->s_log_zone_size;   /* for block-zone conversion */
 25352	  block_pos = position/rip->i_sp->s_block_size; /* relative blk # in file */
 25353	  zone = block_pos >> scale;    /* position's zone */
 25354	  boff = (int) (block_pos - (zone << scale) ); /* relative blk # within zone 
 25355	  dzones = rip->i_ndzones;
 25356	  nr_indirects = rip->i_nindirs;
 25357	
 25358	  /* Is 'position' to be found in the inode itself? */
 25359	  if (zone < dzones) {
 25360	        zind = (int) zone;      /* index should be an int */
 25361	        z = rip->i_zone[zind];
 25362	        if (z == NO_ZONE) return(NO_BLOCK);
 25363	        b = ((block_t) z << scale) + boff;
 25364	        return(b);
 25365	  }
 25366	
 25367	  /* It is not in the inode, so it must be single or double indirect. */
 25368	  excess = zone - dzones;       /* first Vx_NR_DZONES don't count */
 25369	
 25370	  if (excess < nr_indirects) {
 25371	        /* 'position' can be located via the single indirect block. */
 25372	        z = rip->i_zone[dzones];
 25373	  } else {
 25374	        /* 'position' can be located via the double indirect block. */
 25375	        if ( (z = rip->i_zone[dzones+1]) == NO_ZONE) return(NO_BLOCK);
 25376	        excess -= nr_indirects;                 /* single indir doesn't count*/
 25377	        b = (block_t) z << scale;
 25378	        bp = get_block(rip->i_dev, b, NORMAL);  /* get double indirect block */
 25379	        index = (int) (excess/nr_indirects);
 25380	        z = rd_indir(bp, index);                /* z= zone for single*/
 25381	        put_block(bp, INDIRECT_BLOCK);          /* release double ind block */
 25382	        excess = excess % nr_indirects;         /* index into single ind blk */
 25383	  }
 25384	
 25385	  /* 'z' is zone num for single indirect block; 'excess' is index into it. */
 25386	  if (z == NO_ZONE) return(NO_BLOCK);
 25387	  b = (block_t) z << scale;                     /* b is blk # for single ind 
 25388	  bp = get_block(rip->i_dev, b, NORMAL);        /* get single indirect block */
 25389	  ex = (int) excess;                            /* need an integer */
 25390	  z = rd_indir(bp, ex);                         /* get block pointed to */
 25391	  put_block(bp, INDIRECT_BLOCK);                /* release single indir blk */
 25392	  if (z == NO_ZONE) return(NO_BLOCK);
 25393	  b = ((block_t) z << scale) + boff;
 25394	  return(b);
 25395	}
	
 25397	/*===========================================================================*
 25398	 *                              rd_indir                                     *
 25399	 *===========================================================================*/
 25400	PUBLIC zone_t rd_indir(bp, index)
 25401	struct buf *bp;                 /* pointer to indirect block */
 25402	int index;                      /* index into *bp */
 25403	{
 25404	/* Given a pointer to an indirect block, read one entry.  The reason for

_________________________ Page 980    File: servers/fs/read.c _________________________
 25405	 * making a separate routine out of this is that there are four cases:
 25406	 * V1 (IBM and 68000), and V2 (IBM and 68000).
 25407	 */
 25408	
 25409	  struct super_block *sp;
 25410	  zone_t zone;                  /* V2 zones are longs (shorts in V1) */
 25411	
 25412	  sp = get_super(bp->b_dev);    /* need super block to find file sys type */
 25413	
 25414	  /* read a zone from an indirect block */
 25415	  if (sp->s_version == V1)
 25416	        zone = (zone_t) conv2(sp->s_native, (int)  bp->b_v1_ind[index]);
 25417	  else
 25418	        zone = (zone_t) conv4(sp->s_native, (long) bp->b_v2_ind[index]);
 25419	
 25420	  if (zone != NO_ZONE &&
 25421	                (zone < (zone_t) sp->s_firstdatazone || zone >= sp->s_zones)) {
 25422	        printf("Illegal zone number %ld in indirect block, index %d\n",
 25423	               (long) zone, index);
 25424	        panic(__FILE__,"check file system", NO_NUM);
 25425	  }
 25426	  return(zone);
 25427	}
	
 25429	/*===========================================================================*
 25430	 *                              read_ahead                                   *
 25431	 *===========================================================================*/
 25432	PUBLIC void read_ahead()
 25433	{
 25434	/* Read a block into the cache before it is needed. */
 25435	  int block_size;
 25436	  register struct inode *rip;
 25437	  struct buf *bp;
 25438	  block_t b;
 25439	
 25440	  rip = rdahed_inode;           /* pointer to inode to read ahead from */
 25441	  block_size = get_block_size(rip->i_dev);
 25442	  rdahed_inode = NIL_INODE;     /* turn off read ahead */
 25443	  if ( (b = read_map(rip, rdahedpos)) == NO_BLOCK) return;      /* at EOF */
 25444	  bp = rahead(rip, b, rdahedpos, block_size);
 25445	  put_block(bp, PARTIAL_DATA_BLOCK);
 25446	}
	
 25448	/*===========================================================================*
 25449	 *                              rahead                                       *
 25450	 *===========================================================================*/
 25451	PUBLIC struct buf *rahead(rip, baseblock, position, bytes_ahead)
 25452	register struct inode *rip;     /* pointer to inode for file to be read */
 25453	block_t baseblock;              /* block at current position */
 25454	off_t position;                 /* position within file */
 25455	unsigned bytes_ahead;           /* bytes beyond position for immediate use */
 25456	{
 25457	/* Fetch a block from the cache or the device.  If a physical read is
 25458	 * required, prefetch as many more blocks as convenient into the cache.
 25459	 * This usually covers bytes_ahead and is at least BLOCKS_MINIMUM.
 25460	 * The device driver may decide it knows better and stop reading at a
 25461	 * cylinder boundary (or after an error).  Rw_scattered() puts an optional
 25462	 * flag on all reads to allow this.
 25463	 */
 25464	  int block_size;

_________________________ Page 981    File: servers/fs/read.c _________________________
 25465	/* Minimum number of blocks to prefetch. */
 25466	# define BLOCKS_MINIMUM         (NR_BUFS < 50 ? 18 : 32)
 25467	  int block_spec, scale, read_q_size;
 25468	  unsigned int blocks_ahead, fragment;
 25469	  block_t block, blocks_left;
 25470	  off_t ind1_pos;
 25471	  dev_t dev;
 25472	  struct buf *bp;
 25473	  static struct buf *read_q[NR_BUFS];
 25474	
 25475	  block_spec = (rip->i_mode & I_TYPE) == I_BLOCK_SPECIAL;
 25476	  if (block_spec) {
 25477	        dev = (dev_t) rip->i_zone[0];
 25478	  } else {
 25479	        dev = rip->i_dev;
 25480	  }
 25481	  block_size = get_block_size(dev);
 25482	
 25483	  block = baseblock;
 25484	  bp = get_block(dev, block, PREFETCH);
 25485	  if (bp->b_dev != NO_DEV) return(bp);
 25486	
 25487	  /* The best guess for the number of blocks to prefetch:  A lot.
 25488	   * It is impossible to tell what the device looks like, so we don't even
 25489	   * try to guess the geometry, but leave it to the driver.
 25490	   *
 25491	   * The floppy driver can read a full track with no rotational delay, and it
 25492	   * avoids reading partial tracks if it can, so handing it enough buffers to
 25493	   * read two tracks is perfect.  (Two, because some diskette types have
 25494	   * an odd number of sectors per track, so a block may span tracks.)
 25495	   *
 25496	   * The disk drivers don't try to be smart.  With todays disks it is
 25497	   * impossible to tell what the real geometry looks like, so it is best to
 25498	   * read as much as you can.  With luck the caching on the drive allows
 25499	   * for a little time to start the next read.
 25500	   *
 25501	   * The current solution below is a bit of a hack, it just reads blocks from
 25502	   * the current file position hoping that more of the file can be found.  A
 25503	   * better solution must look at the already available zone pointers and
 25504	   * indirect blocks (but don't call read_map!).
 25505	   */
 25506	
 25507	  fragment = position % block_size;
 25508	  position -= fragment;
 25509	  bytes_ahead += fragment;
 25510	
 25511	  blocks_ahead = (bytes_ahead + block_size - 1) / block_size;
 25512	
 25513	  if (block_spec && rip->i_size == 0) {
 25514	        blocks_left = NR_IOREQS;
 25515	  } else {
 25516	        blocks_left = (rip->i_size - position + block_size - 1) / block_size;
 25517	
 25518	        /* Go for the first indirect block if we are in its neighborhood. */
 25519	        if (!block_spec) {
 25520	                scale = rip->i_sp->s_log_zone_size;
 25521	                ind1_pos = (off_t) rip->i_ndzones * (block_size << scale);
 25522	                if (position <= ind1_pos && rip->i_size > ind1_pos) {
 25523	                        blocks_ahead++;
 25524	                        blocks_left++;

_________________________ Page 982    File: servers/fs/read.c _________________________
 25525	                }
 25526	        }
 25527	  }
 25528	
 25529	  /* No more than the maximum request. */
 25530	  if (blocks_ahead > NR_IOREQS) blocks_ahead = NR_IOREQS;
 25531	
 25532	  /* Read at least the minimum number of blocks, but not after a seek. */
 25533	  if (blocks_ahead < BLOCKS_MINIMUM && rip->i_seek == NO_SEEK)
 25534	        blocks_ahead = BLOCKS_MINIMUM;
 25535	
 25536	  /* Can't go past end of file. */
 25537	  if (blocks_ahead > blocks_left) blocks_ahead = blocks_left;
 25538	
 25539	  read_q_size = 0;
 25540	
 25541	  /* Acquire block buffers. */
 25542	  for (;;) {
 25543	        read_q[read_q_size++] = bp;
 25544	
 25545	        if (--blocks_ahead == 0) break;
 25546	
 25547	        /* Don't trash the cache, leave 4 free. */
 25548	        if (bufs_in_use >= NR_BUFS - 4) break;
 25549	
 25550	        block++;
 25551	
 25552	        bp = get_block(dev, block, PREFETCH);
 25553	        if (bp->b_dev != NO_DEV) {
 25554	                /* Oops, block already in the cache, get out. */
 25555	                put_block(bp, FULL_DATA_BLOCK);
 25556	                break;
 25557	        }
 25558	  }
 25559	  rw_scattered(dev, read_q, read_q_size, READING);
 25560	  return(get_block(dev, baseblock, NORMAL));
 25561	}




++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/write.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 25600	/* This file is the counterpart of "read.c".  It contains the code for writing
 25601	 * insofar as this is not contained in read_write().
 25602	 *
 25603	 * The entry points into this file are
 25604	 *   do_write:     call read_write to perform the WRITE system call
 25605	 *   clear_zone:   erase a zone in the middle of a file
 25606	 *   new_block:    acquire a new block
 25607	 */
 25608	
 25609	#include "fs.h"
 25610	#include <string.h>
 25611	#include "buf.h"
 25612	#include "file.h"
 25613	#include "fproc.h"
 25614	#include "inode.h"

_________________________ Page 983    File: servers/fs/write.c _________________________
 25615	#include "super.h"
 25616	
 25617	FORWARD _PROTOTYPE( int write_map, (struct inode *rip, off_t position,
 25618	                        zone_t new_zone)                                );
 25619	
 25620	FORWARD _PROTOTYPE( void wr_indir, (struct buf *bp, int index, zone_t zone) );
 25621	
 25622	/*===========================================================================*
 25623	 *                              do_write                                     *
 25624	 *===========================================================================*/
 25625	PUBLIC int do_write()
 25626	{
 25627	/* Perform the write(fd, buffer, nbytes) system call. */
 25628	
 25629	  return(read_write(WRITING));
 25630	}
	
 25632	/*===========================================================================*
 25633	 *                              write_map                                    *
 25634	 *===========================================================================*/
 25635	PRIVATE int write_map(rip, position, new_zone)
 25636	register struct inode *rip;     /* pointer to inode to be changed */
 25637	off_t position;                 /* file address to be mapped */
 25638	zone_t new_zone;                /* zone # to be inserted */
 25639	{
 25640	/* Write a new zone into an inode. */
 25641	  int scale, ind_ex, new_ind, new_dbl, zones, nr_indirects, single, zindex, ex;
 25642	  zone_t z, z1;
 25643	  register block_t b;
 25644	  long excess, zone;
 25645	  struct buf *bp;
 25646	
 25647	  rip->i_dirt = DIRTY;          /* inode will be changed */
 25648	  bp = NIL_BUF;
 25649	  scale = rip->i_sp->s_log_zone_size;           /* for zone-block conversion */
 25650	        /* relative zone # to insert */
 25651	  zone = (position/rip->i_sp->s_block_size) >> scale;
 25652	  zones = rip->i_ndzones;       /* # direct zones in the inode */
 25653	  nr_indirects = rip->i_nindirs;/* # indirect zones per indirect block */
 25654	
 25655	  /* Is 'position' to be found in the inode itself? */
 25656	  if (zone < zones) {
 25657	        zindex = (int) zone;    /* we need an integer here */
 25658	        rip->i_zone[zindex] = new_zone;
 25659	        return(OK);
 25660	  }
 25661	
 25662	  /* It is not in the inode, so it must be single or double indirect. */
 25663	  excess = zone - zones;        /* first Vx_NR_DZONES don't count */
 25664	  new_ind = FALSE;
 25665	  new_dbl = FALSE;
 25666	
 25667	  if (excess < nr_indirects) {
 25668	        /* 'position' can be located via the single indirect block. */
 25669	        z1 = rip->i_zone[zones];        /* single indirect zone */
 25670	        single = TRUE;
 25671	  } else {
 25672	        /* 'position' can be located via the double indirect block. */
 25673	        if ( (z = rip->i_zone[zones+1]) == NO_ZONE) {
 25674	                /* Create the double indirect block. */

_________________________ Page 984    File: servers/fs/write.c _________________________
 25675	                if ( (z = alloc_zone(rip->i_dev, rip->i_zone[0])) == NO_ZONE)
 25676	                        return(err_code);
 25677	                rip->i_zone[zones+1] = z;
 25678	                new_dbl = TRUE; /* set flag for later */
 25679	        }
 25680	
 25681	        /* Either way, 'z' is zone number for double indirect block. */
 25682	        excess -= nr_indirects; /* single indirect doesn't count */
 25683	        ind_ex = (int) (excess / nr_indirects);
 25684	        excess = excess % nr_indirects;
 25685	        if (ind_ex >= nr_indirects) return(EFBIG);
 25686	        b = (block_t) z << scale;
 25687	        bp = get_block(rip->i_dev, b, (new_dbl ? NO_READ : NORMAL));
 25688	        if (new_dbl) zero_block(bp);
 25689	        z1 = rd_indir(bp, ind_ex);
 25690	        single = FALSE;
 25691	  }
 25692	
 25693	  /* z1 is now single indirect zone; 'excess' is index. */
 25694	  if (z1 == NO_ZONE) {
 25695	        /* Create indirect block and store zone # in inode or dbl indir blk. */
 25696	        z1 = alloc_zone(rip->i_dev, rip->i_zone[0]);
 25697	        if (single)
 25698	                rip->i_zone[zones] = z1;        /* update inode */
 25699	        else
 25700	                wr_indir(bp, ind_ex, z1);       /* update dbl indir */
 25701	
 25702	        new_ind = TRUE;
 25703	        if (bp != NIL_BUF) bp->b_dirt = DIRTY;  /* if double ind, it is dirty*/
 25704	        if (z1 == NO_ZONE) {
 25705	                put_block(bp, INDIRECT_BLOCK);  /* release dbl indirect blk */
 25706	                return(err_code);       /* couldn't create single ind */
 25707	        }
 25708	  }
 25709	  put_block(bp, INDIRECT_BLOCK);        /* release double indirect blk */
 25710	
 25711	  /* z1 is indirect block's zone number. */
 25712	  b = (block_t) z1 << scale;
 25713	  bp = get_block(rip->i_dev, b, (new_ind ? NO_READ : NORMAL) );
 25714	  if (new_ind) zero_block(bp);
 25715	  ex = (int) excess;                    /* we need an int here */
 25716	  wr_indir(bp, ex, new_zone);
 25717	  bp->b_dirt = DIRTY;
 25718	  put_block(bp, INDIRECT_BLOCK);
 25719	
 25720	  return(OK);
 25721	}
	
 25723	/*===========================================================================*
 25724	 *                              wr_indir                                     *
 25725	 *===========================================================================*/
 25726	PRIVATE void wr_indir(bp, index, zone)
 25727	struct buf *bp;                 /* pointer to indirect block */
 25728	int index;                      /* index into *bp */
 25729	zone_t zone;                    /* zone to write */
 25730	{
 25731	/* Given a pointer to an indirect block, write one entry. */
 25732	
 25733	  struct super_block *sp;
 25734	

_________________________ Page 985    File: servers/fs/write.c _________________________
 25735	  sp = get_super(bp->b_dev);    /* need super block to find file sys type */
 25736	
 25737	  /* write a zone into an indirect block */
 25738	  if (sp->s_version == V1)
 25739	        bp->b_v1_ind[index] = (zone1_t) conv2(sp->s_native, (int)  zone);
 25740	  else
 25741	        bp->b_v2_ind[index] = (zone_t)  conv4(sp->s_native, (long) zone);
 25742	}
	
 25744	/*===========================================================================*
 25745	 *                              clear_zone                                   *
 25746	 *===========================================================================*/
 25747	PUBLIC void clear_zone(rip, pos, flag)
 25748	register struct inode *rip;     /* inode to clear */
 25749	off_t pos;                      /* points to block to clear */
 25750	int flag;                       /* 0 if called by read_write, 1 by new_block */
 25751	{
 25752	/* Zero a zone, possibly starting in the middle.  The parameter 'pos' gives
 25753	 * a byte in the first block to be zeroed.  Clearzone() is called from 
 25754	 * read_write and new_block().
 25755	 */
 25756	
 25757	  register struct buf *bp;
 25758	  register block_t b, blo, bhi;
 25759	  register off_t next;
 25760	  register int scale;
 25761	  register zone_t zone_size;
 25762	
 25763	  /* If the block size and zone size are the same, clear_zone() not needed. */
 25764	  scale = rip->i_sp->s_log_zone_size;
 25765	  if (scale == 0) return;
 25766	
 25767	  zone_size = (zone_t) rip->i_sp->s_block_size << scale;
 25768	  if (flag == 1) pos = (pos/zone_size) * zone_size;
 25769	  next = pos + rip->i_sp->s_block_size - 1;
 25770	
 25771	  /* If 'pos' is in the last block of a zone, do not clear the zone. */
 25772	  if (next/zone_size != pos/zone_size) return;
 25773	  if ( (blo = read_map(rip, next)) == NO_BLOCK) return;
 25774	  bhi = (  ((blo>>scale)+1) << scale)   - 1;
 25775	
 25776	  /* Clear all the blocks between 'blo' and 'bhi'. */
 25777	  for (b = blo; b <= bhi; b++) {
 25778	        bp = get_block(rip->i_dev, b, NO_READ);
 25779	        zero_block(bp);
 25780	        put_block(bp, FULL_DATA_BLOCK);
 25781	  }
 25782	}
	
 25784	/*===========================================================================*
 25785	 *                              new_block                                    *
 25786	 *===========================================================================*/
 25787	PUBLIC struct buf *new_block(rip, position)
 25788	register struct inode *rip;     /* pointer to inode */
 25789	off_t position;                 /* file pointer */
 25790	{
 25791	/* Acquire a new block and return a pointer to it.  Doing so may require
 25792	 * allocating a complete zone, and then returning the initial block.
 25793	 * On the other hand, the current zone may still have some unused blocks.
 25794	 */

_________________________ Page 986    File: servers/fs/write.c _________________________
 25795	
 25796	  register struct buf *bp;
 25797	  block_t b, base_block;
 25798	  zone_t z;
 25799	  zone_t zone_size;
 25800	  int scale, r;
 25801	  struct super_block *sp;
 25802	
 25803	  /* Is another block available in the current zone? */
 25804	  if ( (b = read_map(rip, position)) == NO_BLOCK) {
 25805	        /* Choose first zone if possible. */
 25806	        /* Lose if the file is nonempty but the first zone number is NO_ZONE
 25807	         * corresponding to a zone full of zeros.  It would be better to
 25808	         * search near the last real zone.
 25809	         */
 25810	        if (rip->i_zone[0] == NO_ZONE) {
 25811	                sp = rip->i_sp;
 25812	                z = sp->s_firstdatazone;
 25813	        } else {
 25814	                z = rip->i_zone[0];     /* hunt near first zone */
 25815	        }
 25816	        if ( (z = alloc_zone(rip->i_dev, z)) == NO_ZONE) return(NIL_BUF);
 25817	        if ( (r = write_map(rip, position, z)) != OK) {
 25818	                free_zone(rip->i_dev, z);
 25819	                err_code = r;
 25820	                return(NIL_BUF);
 25821	        }
 25822	
 25823	        /* If we are not writing at EOF, clear the zone, just to be safe. */
 25824	        if ( position != rip->i_size) clear_zone(rip, position, 1);
 25825	        scale = rip->i_sp->s_log_zone_size;
 25826	        base_block = (block_t) z << scale;
 25827	        zone_size = (zone_t) rip->i_sp->s_block_size << scale;
 25828	        b = base_block + (block_t)((position % zone_size)/rip->i_sp->s_block_size);
 25829	  }
 25830	
 25831	  bp = get_block(rip->i_dev, b, NO_READ);
 25832	  zero_block(bp);
 25833	  return(bp);
 25834	}
	
 25836	/*===========================================================================*
 25837	 *                              zero_block                                   *
 25838	 *===========================================================================*/
 25839	PUBLIC void zero_block(bp)
 25840	register struct buf *bp;        /* pointer to buffer to zero */
 25841	{
 25842	/* Zero a block. */
 25843	  memset(bp->b_data, 0, MAX_BLOCK_SIZE);
 25844	  bp->b_dirt = DIRTY;
 25845	}

_________________________ Page 987    File: servers/fs/write.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/pipe.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 25900	/* This file deals with the suspension and revival of processes.  A process can
 25901	 * be suspended because it wants to read or write from a pipe and can't, or
 25902	 * because it wants to read or write from a special file and can't.  When a
 25903	 * process can't continue it is suspended, and revived later when it is able
 25904	 * to continue.
 25905	 *
 25906	 * The entry points into this file are
 25907	 *   do_pipe:     perform the PIPE system call
 25908	 *   pipe_check:  check to see that a read or write on a pipe is feasible now
 25909	 *   suspend:     suspend a process that cannot do a requested read or write
 25910	 *   release:     check to see if a suspended process can be released and do
 25911	 *                it
 25912	 *   revive:      mark a suspended process as able to run again
 25913	 *   do_unpause:  a signal has been sent to a process; see if it suspended
 25914	 */
 25915	
 25916	#include "fs.h"
 25917	#include <fcntl.h>
 25918	#include <signal.h>
 25919	#include <minix/callnr.h>
 25920	#include <minix/com.h>
 25921	#include <sys/select.h>
 25922	#include <sys/time.h>
 25923	#include "file.h"
 25924	#include "fproc.h"
 25925	#include "inode.h"
 25926	#include "param.h"
 25927	#include "super.h"
 25928	#include "select.h"
 25929	
 25930	/*===========================================================================*
 25931	 *                              do_pipe                                      *
 25932	 *===========================================================================*/
 25933	PUBLIC int do_pipe()
 25934	{
 25935	/* Perform the pipe(fil_des) system call. */
 25936	
 25937	  register struct fproc *rfp;
 25938	  register struct inode *rip;
 25939	  int r;
 25940	  struct filp *fil_ptr0, *fil_ptr1;
 25941	  int fil_des[2];               /* reply goes here */
 25942	
 25943	  /* Acquire two file descriptors. */
 25944	  rfp = fp;
 25945	  if ( (r = get_fd(0, R_BIT, &fil_des[0], &fil_ptr0)) != OK) return(r);
 25946	  rfp->fp_filp[fil_des[0]] = fil_ptr0;
 25947	  fil_ptr0->filp_count = 1;
 25948	  if ( (r = get_fd(0, W_BIT, &fil_des[1], &fil_ptr1)) != OK) {
 25949	        rfp->fp_filp[fil_des[0]] = NIL_FILP;
 25950	        fil_ptr0->filp_count = 0;
 25951	        return(r);
 25952	  }
 25953	  rfp->fp_filp[fil_des[1]] = fil_ptr1;
 25954	  fil_ptr1->filp_count = 1;

_________________________ Page 988    File: servers/fs/pipe.c _________________________
 25955	
 25956	  /* Make the inode on the pipe device. */
 25957	  if ( (rip = alloc_inode(root_dev, I_REGULAR) ) == NIL_INODE) {
 25958	        rfp->fp_filp[fil_des[0]] = NIL_FILP;
 25959	        fil_ptr0->filp_count = 0;
 25960	        rfp->fp_filp[fil_des[1]] = NIL_FILP;
 25961	        fil_ptr1->filp_count = 0;
 25962	        return(err_code);
 25963	  }
 25964	
 25965	  if (read_only(rip) != OK) 
 25966	        panic(__FILE__,"pipe device is read only", NO_NUM);
 25967	 
 25968	  rip->i_pipe = I_PIPE;
 25969	  rip->i_mode &= ~I_REGULAR;
 25970	  rip->i_mode |= I_NAMED_PIPE;  /* pipes and FIFOs have this bit set */
 25971	  fil_ptr0->filp_ino = rip;
 25972	  fil_ptr0->filp_flags = O_RDONLY;
 25973	  dup_inode(rip);               /* for double usage */
 25974	  fil_ptr1->filp_ino = rip;
 25975	  fil_ptr1->filp_flags = O_WRONLY;
 25976	  rw_inode(rip, WRITING);       /* mark inode as allocated */
 25977	  m_out.reply_i1 = fil_des[0];
 25978	  m_out.reply_i2 = fil_des[1];
 25979	  rip->i_update = ATIME | CTIME | MTIME;
 25980	  return(OK);
 25981	}
	
 25983	/*===========================================================================*
 25984	 *                              pipe_check                                   *
 25985	 *===========================================================================*/
 25986	PUBLIC int pipe_check(rip, rw_flag, oflags, bytes, position, canwrite, notouch)
 25987	register struct inode *rip;     /* the inode of the pipe */
 25988	int rw_flag;                    /* READING or WRITING */
 25989	int oflags;                     /* flags set by open or fcntl */
 25990	register int bytes;             /* bytes to be read or written (all chunks) */
 25991	register off_t position;        /* current file position */
 25992	int *canwrite;                  /* return: number of bytes we can write */
 25993	int notouch;                    /* check only */
 25994	{
 25995	/* Pipes are a little different.  If a process reads from an empty pipe for
 25996	 * which a writer still exists, suspend the reader.  If the pipe is empty
 25997	 * and there is no writer, return 0 bytes.  If a process is writing to a
 25998	 * pipe and no one is reading from it, give a broken pipe error.
 25999	 */
 26000	
 26001	  /* If reading, check for empty pipe. */
 26002	  if (rw_flag == READING) {
 26003	        if (position >= rip->i_size) {
 26004	                /* Process is reading from an empty pipe. */
 26005	                int r = 0;
 26006	                if (find_filp(rip, W_BIT) != NIL_FILP) {
 26007	                        /* Writer exists */
 26008	                        if (oflags & O_NONBLOCK) {
 26009	                                r = EAGAIN;
 26010	                        } else {
 26011	                                if (!notouch)
 26012	                                        suspend(XPIPE); /* block reader */
 26013	                                r = SUSPEND;
 26014	                        }

_________________________ Page 989    File: servers/fs/pipe.c _________________________
 26015	                        /* If need be, activate sleeping writers. */
 26016	                        if (susp_count > 0 && !notouch)
 26017	                                release(rip, WRITE, susp_count);
 26018	                }
 26019	                return(r);
 26020	        }
 26021	  } else {
 26022	        /* Process is writing to a pipe. */
 26023	        if (find_filp(rip, R_BIT) == NIL_FILP) {
 26024	                /* Tell kernel to generate a SIGPIPE signal. */
 26025	                if (!notouch)
 26026	                        sys_kill((int)(fp - fproc), SIGPIPE);
 26027	                return(EPIPE);
 26028	        }
 26029	
 26030	        if (position + bytes > PIPE_SIZE(rip->i_sp->s_block_size)) {
 26031	                if ((oflags & O_NONBLOCK)
 26032	                 && bytes < PIPE_SIZE(rip->i_sp->s_block_size))
 26033	                        return(EAGAIN);
 26034	                else if ((oflags & O_NONBLOCK)
 26035	                && bytes > PIPE_SIZE(rip->i_sp->s_block_size)) {
 26036	                if ( (*canwrite = (PIPE_SIZE(rip->i_sp->s_block_size) 
 26037	                        - position)) > 0)  {
 26038	                                /* Do a partial write. Need to wakeup reader */
 26039	                                if (!notouch)
 26040	                                        release(rip, READ, susp_count);
 26041	                                return(1);
 26042	                        } else {
 26043	                                return(EAGAIN);
 26044	                        }
 26045	                     }
 26046	                if (bytes > PIPE_SIZE(rip->i_sp->s_block_size)) {
 26047	                        if ((*canwrite = PIPE_SIZE(rip->i_sp->s_block_size) 
 26048	                                - position) > 0) {
 26049	                                /* Do a partial write. Need to wakeup reader
 26050	                                 * since we'll suspend ourself in read_write()
 26051	                                 */
 26052	                                release(rip, READ, susp_count);
 26053	                                return(1);
 26054	                        }
 26055	                }
 26056	                if (!notouch)
 26057	                        suspend(XPIPE); /* stop writer -- pipe full */
 26058	                return(SUSPEND);
 26059	        }
 26060	
 26061	        /* Writing to an empty pipe.  Search for suspended reader. */
 26062	        if (position == 0 && !notouch)
 26063	                release(rip, READ, susp_count);
 26064	  }
 26065	
 26066	  *canwrite = 0;
 26067	  return(1);
 26068	}
	
 26070	/*===========================================================================*
 26071	 *                              suspend                                      *
 26072	 *===========================================================================*/
 26073	PUBLIC void suspend(task)
 26074	int task;                       /* who is proc waiting for? (PIPE = pipe) */

_________________________ Page 990    File: servers/fs/pipe.c _________________________
 26075	{
 26076	/* Take measures to suspend the processing of the present system call.
 26077	 * Store the parameters to be used upon resuming in the process table.
 26078	 * (Actually they are not used when a process is waiting for an I/O device,
 26079	 * but they are needed for pipes, and it is not worth making the distinction.)
 26080	 * The SUSPEND pseudo error should be returned after calling suspend().
 26081	 */
 26082	
 26083	  if (task == XPIPE || task == XPOPEN) susp_count++;/* #procs susp'ed on pipe*/
 26084	  fp->fp_suspended = SUSPENDED;
 26085	  fp->fp_fd = m_in.fd << 8 | call_nr;
 26086	  fp->fp_task = -task;
 26087	  if (task == XLOCK) {
 26088	        fp->fp_buffer = (char *) m_in.name1;    /* third arg to fcntl() */
 26089	        fp->fp_nbytes = m_in.request;           /* second arg to fcntl() */
 26090	  } else {
 26091	        fp->fp_buffer = m_in.buffer;            /* for reads and writes */
 26092	        fp->fp_nbytes = m_in.nbytes;
 26093	  }
 26094	}
	
 26096	/*===========================================================================*
 26097	 *                              release                                      *
 26098	 *===========================================================================*/
 26099	PUBLIC void release(ip, call_nr, count)
 26100	register struct inode *ip;      /* inode of pipe */
 26101	int call_nr;                    /* READ, WRITE, OPEN or CREAT */
 26102	int count;                      /* max number of processes to release */
 26103	{
 26104	/* Check to see if any process is hanging on the pipe whose inode is in 'ip'.
 26105	 * If one is, and it was trying to perform the call indicated by 'call_nr',
 26106	 * release it.
 26107	 */
 26108	
 26109	  register struct fproc *rp;
 26110	  struct filp *f;
 26111	
 26112	  /* Trying to perform the call also includes SELECTing on it with that
 26113	   * operation.
 26114	   */
 26115	  if (call_nr == READ || call_nr == WRITE) {
 26116	          int op;
 26117	          if (call_nr == READ)
 26118	                op = SEL_RD;
 26119	          else
 26120	                op = SEL_WR;
 26121	          for(f = &filp[0]; f < &filp[NR_FILPS]; f++) {
 26122	                if (f->filp_count < 1 || !(f->filp_pipe_select_ops & op) ||
 26123	                   f->filp_ino != ip)
 26124	                        continue;
 26125	                 select_callback(f, op);
 26126	                f->filp_pipe_select_ops &= ~op;
 26127	        }
 26128	  }
 26129	
 26130	  /* Search the proc table. */
 26131	  for (rp = &fproc[0]; rp < &fproc[NR_PROCS]; rp++) {
 26132	        if (rp->fp_suspended == SUSPENDED &&
 26133	                        rp->fp_revived == NOT_REVIVING &&
 26134	                        (rp->fp_fd & BYTE) == call_nr &&

_________________________ Page 991    File: servers/fs/pipe.c _________________________
 26135	                        rp->fp_filp[rp->fp_fd>>8]->filp_ino == ip) {
 26136	                revive((int)(rp - fproc), 0);
 26137	                susp_count--;   /* keep track of who is suspended */
 26138	                if (--count == 0) return;
 26139	        }
 26140	  }
 26141	}
	
 26143	/*===========================================================================*
 26144	 *                              revive                                       *
 26145	 *===========================================================================*/
 26146	PUBLIC void revive(proc_nr, returned)
 26147	int proc_nr;                    /* process to revive */
 26148	int returned;                   /* if hanging on task, how many bytes read */
 26149	{
 26150	/* Revive a previously blocked process. When a process hangs on tty, this
 26151	 * is the way it is eventually released.
 26152	 */
 26153	
 26154	  register struct fproc *rfp;
 26155	  register int task;
 26156	
 26157	  if (proc_nr < 0 || proc_nr >= NR_PROCS)
 26158	        panic(__FILE__,"revive err", proc_nr);
 26159	  rfp = &fproc[proc_nr];
 26160	  if (rfp->fp_suspended == NOT_SUSPENDED || rfp->fp_revived == REVIVING)return;
 26161	
 26162	  /* The 'reviving' flag only applies to pipes.  Processes waiting for TTY get
 26163	   * a message right away.  The revival process is different for TTY and pipes.
 26164	   * For select and TTY revival, the work is already done, for pipes it is not:
 26165	   *  the proc must be restarted so it can try again.
 26166	   */
 26167	  task = -rfp->fp_task;
 26168	  if (task == XPIPE || task == XLOCK) {
 26169	        /* Revive a process suspended on a pipe or lock. */
 26170	        rfp->fp_revived = REVIVING;
 26171	        reviving++;             /* process was waiting on pipe or lock */
 26172	  } else {
 26173	        rfp->fp_suspended = NOT_SUSPENDED;
 26174	        if (task == XPOPEN) /* process blocked in open or create */
 26175	                reply(proc_nr, rfp->fp_fd>>8);
 26176	        else if (task == XSELECT) {
 26177	                reply(proc_nr, returned);
 26178	        } else {
 26179	                /* Revive a process suspended on TTY or other device. */
 26180	                rfp->fp_nbytes = returned;      /*pretend it wants only what there 
 26181	                reply(proc_nr, returned);       /* unblock the process */
 26182	        }
 26183	  }
 26184	}
	
 26186	/*===========================================================================*
 26187	 *                              do_unpause                                   *
 26188	 *===========================================================================*/
 26189	PUBLIC int do_unpause()
 26190	{
 26191	/* A signal has been sent to a user who is paused on the file system.
 26192	 * Abort the system call with the EINTR error message.
 26193	 */
 26194	

_________________________ Page 992    File: servers/fs/pipe.c _________________________
 26195	  register struct fproc *rfp;
 26196	  int proc_nr, task, fild;
 26197	  struct filp *f;
 26198	  dev_t dev;
 26199	  message mess;
 26200	
 26201	  if (who > PM_PROC_NR) return(EPERM);
 26202	  proc_nr = m_in.pro;
 26203	  if (proc_nr < 0 || proc_nr >= NR_PROCS)
 26204	        panic(__FILE__,"unpause err 1", proc_nr);
 26205	  rfp = &fproc[proc_nr];
 26206	  if (rfp->fp_suspended == NOT_SUSPENDED) return(OK);
 26207	  task = -rfp->fp_task;
 26208	
 26209	  switch (task) {
 26210	        case XPIPE:             /* process trying to read or write a pipe */
 26211	                break;
 26212	
 26213	        case XLOCK:             /* process trying to set a lock with FCNTL */
 26214	                break;
 26215	
 26216	        case XSELECT:           /* process blocking on select() */
 26217	                select_forget(proc_nr);
 26218	                break;
 26219	
 26220	        case XPOPEN:            /* process trying to open a fifo */
 26221	                break;
 26222	
 26223	        default:                /* process trying to do device I/O (e.g. tty)*/
 26224	                fild = (rfp->fp_fd >> 8) & BYTE;/* extract file descriptor */
 26225	                if (fild < 0 || fild >= OPEN_MAX)
 26226	                        panic(__FILE__,"unpause err 2",NO_NUM);
 26227	                f = rfp->fp_filp[fild];
 26228	                dev = (dev_t) f->filp_ino->i_zone[0];   /* device hung on */
 26229	                mess.TTY_LINE = (dev >> MINOR) & BYTE;
 26230	                mess.PROC_NR = proc_nr;
 26231	
 26232	                /* Tell kernel R or W. Mode is from current call, not open. */
 26233	                mess.COUNT = (rfp->fp_fd & BYTE) == READ ? R_BIT : W_BIT;
 26234	                mess.m_type = CANCEL;
 26235	                fp = rfp;       /* hack - ctty_io uses fp */
 26236	                (*dmap[(dev >> MAJOR) & BYTE].dmap_io)(task, &mess);
 26237	  }
 26238	
 26239	  rfp->fp_suspended = NOT_SUSPENDED;
 26240	  reply(proc_nr, EINTR);        /* signal interrupted call */
 26241	  return(OK);
 26242	}
	
 26244	/*===========================================================================*
 26245	 *                              select_request_pipe                          *
 26246	 *===========================================================================*/
 26247	PUBLIC int select_request_pipe(struct filp *f, int *ops, int block)
 26248	{
 26249	        int orig_ops, r = 0, err, canwrite;
 26250	        orig_ops = *ops;
 26251	        if ((*ops & SEL_RD)) {
 26252	                if ((err = pipe_check(f->filp_ino, READING, 0,
 26253	                        1, f->filp_pos, &canwrite, 1)) != SUSPEND)
 26254	                        r |= SEL_RD;

_________________________ Page 993    File: servers/fs/pipe.c _________________________
 26255	                if (err < 0 && err != SUSPEND && (*ops & SEL
 26256	                        r |= SEL_ERR;
 26257	        }
 26258	        if ((*ops & SEL_WR)) {
 26259	                if ((err = pipe_check(f->filp_ino, WRITING, 0,
 26260	                        1, f->filp_pos, &canwrite, 1)) != SUSPEND)
 26261	                        r |= SEL_WR;
 26262	                if (err < 0 && err != SUSPEND && (*ops & SEL
 26263	                        r |= SEL_ERR;
 26264	        }
 26265	
 26266	        *ops = r;
 26267	
 26268	        if (!r && block) {
 26269	                f->filp_pipe_select_ops |= orig_ops;
 26270	        }
 26271	
 26272	        return SEL_OK;
 26273	}
	
 26275	/*===========================================================================*
 26276	 *                              select_match_pipe                            *
 26277	 *===========================================================================*/
 26278	PUBLIC int select_match_pipe(struct filp *f)
 26279	{
 26280	        /* recognize either pipe or named pipe (FIFO) */
 26281	        if (f && f->filp_ino && (f->filp_ino->i_mode & I_NAMED_
 26282	                return 1;
 26283	        return 0;
 26284	}
	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/path.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 26300	/* This file contains the procedures that look up path names in the directory
 26301	 * system and determine the inode number that goes with a given path name.
 26302	 *
 26303	 *  The entry points into this file are
 26304	 *   eat_path:   the 'main' routine of the path-to-inode conversion mechanism
 26305	 *   last_dir:   find the final directory on a given path
 26306	 *   advance:    parse one component of a path name
 26307	 *   search_dir: search a directory for a string and return its inode number
 26308	 */
 26309	
 26310	#include "fs.h"
 26311	#include <string.h>
 26312	#include <minix/callnr.h>
 26313	#include "buf.h"
 26314	#include "file.h"
 26315	#include "fproc.h"
 26316	#include "inode.h"
 26317	#include "super.h"
 26318	
 26319	PUBLIC char dot1[2] = ".";      /* used for search_dir to bypass the access */

_________________________ Page 994    File: servers/fs/path.c _________________________
 26320	PUBLIC char dot2[3] = "..";     /* permissions for . and ..                 */
 26321	
 26322	FORWARD _PROTOTYPE( char *get_name, (char *old_name, char string [NAME_MAX]) );
 26323	
 26324	/*===========================================================================*
 26325	 *                              eat_path                                     *
 26326	 *===========================================================================*/
 26327	PUBLIC struct inode *eat_path(path)
 26328	char *path;                     /* the path name to be parsed */
 26329	{
 26330	/* Parse the path 'path' and put its inode in the inode table. If not possible,
 26331	 * return NIL_INODE as function value and an error code in 'err_code'.
 26332	 */
 26333	
 26334	  register struct inode *ldip, *rip;
 26335	  char string[NAME_MAX];        /* hold 1 path component name here */
 26336	
 26337	  /* First open the path down to the final directory. */
 26338	  if ( (ldip = last_dir(path, string)) == NIL_INODE) {
 26339	        return(NIL_INODE);      /* we couldn't open final directory */
 26340	        }
 26341	
 26342	  /* The path consisting only of "/" is a special case, check for it. */
 26343	  if (string[0] == '\0') return(ldip);
 26344	
 26345	  /* Get final component of the path. */
 26346	  rip = advance(ldip, string);
 26347	  put_inode(ldip);
 26348	  return(rip);
 26349	}
	
 26351	/*===========================================================================*
 26352	 *                              last_dir                                     *
 26353	 *===========================================================================*/
 26354	PUBLIC struct inode *last_dir(path, string)
 26355	char *path;                     /* the path name to be parsed */
 26356	char string[NAME_MAX];          /* the final component is returned here */
 26357	{
 26358	/* Given a path, 'path', located in the fs address space, parse it as
 26359	 * far as the last directory, fetch the inode for the last directory into
 26360	 * the inode table, and return a pointer to the inode.  In
 26361	 * addition, return the final component of the path in 'string'.
 26362	 * If the last directory can't be opened, return NIL_INODE and
 26363	 * the reason for failure in 'err_code'.
 26364	 */
 26365	
 26366	  register struct inode *rip;
 26367	  register char *new_name;
 26368	  register struct inode *new_ip;
 26369	
 26370	  /* Is the path absolute or relative?  Initialize 'rip' accordingly. */
 26371	  rip = (*path == '/' ? fp->fp_rootdir : fp->fp_workdir);
 26372	
 26373	  /* If dir has been removed or path is empty, return ENOENT. */
 26374	  if (rip->i_nlinks == 0 || *path == '\0') {
 26375	        err_code = ENOENT;
 26376	        return(NIL_INODE);
 26377	  }
 26378	
 26379	  dup_inode(rip);               /* inode will be returned with put_inode */

_________________________ Page 995    File: servers/fs/path.c _________________________
 26380	
 26381	  /* Scan the path component by component. */
 26382	  while (TRUE) {
 26383	        /* Extract one component. */
 26384	        if ( (new_name = get_name(path, string)) == (char*) 0) {
 26385	                put_inode(rip); /* bad path in user space */
 26386	                return(NIL_INODE);
 26387	        }
 26388	        if (*new_name == '\0') {
 26389	                if ( (rip->i_mode & I_TYPE) == I_DIRECTORY) {
 26390	                        return(rip);    /* normal exit */
 26391	                } else {
 26392	                        /* last file of path prefix is not a directory */
 26393	                        put_inode(rip);
 26394	                        err_code = ENOTDIR;                     
 26395	                        return(NIL_INODE);
 26396	                }
 26397	        }
 26398	
 26399	        /* There is more path.  Keep parsing. */
 26400	        new_ip = advance(rip, string);
 26401	        put_inode(rip);         /* rip either obsolete or irrelevant */
 26402	        if (new_ip == NIL_INODE) return(NIL_INODE);
 26403	
 26404	        /* The call to advance() succeeded.  Fetch next component. */
 26405	        path = new_name;
 26406	        rip = new_ip;
 26407	  }
 26408	}
	
 26410	/*===========================================================================*
 26411	 *                              get_name                                     *
 26412	 *===========================================================================*/
 26413	PRIVATE char *get_name(old_name, string)
 26414	char *old_name;                 /* path name to parse */
 26415	char string[NAME_MAX];          /* component extracted from 'old_name' */
 26416	{
 26417	/* Given a pointer to a path name in fs space, 'old_name', copy the next
 26418	 * component to 'string' and pad with zeros.  A pointer to that part of
 26419	 * the name as yet unparsed is returned.  Roughly speaking,
 26420	 * 'get_name' = 'old_name' - 'string'.
 26421	 *
 26422	 * This routine follows the standard convention that /usr/ast, /usr//ast,
 26423	 * //usr///ast and /usr/ast/ are all equivalent.
 26424	 */
 26425	
 26426	  register int c;
 26427	  register char *np, *rnp;
 26428	
 26429	  np = string;                  /* 'np' points to current position */
 26430	  rnp = old_name;               /* 'rnp' points to unparsed string */
 26431	  while ( (c = *rnp) == '/') rnp++;     /* skip leading slashes */
 26432	
 26433	  /* Copy the unparsed path, 'old_name', to the array, 'string'. */
 26434	  while ( rnp < &old_name[PATH_MAX]  &&  c != '/'   &&  c !=
 26435	        if (np < &string[NAME_MAX]) *np++ = c;
 26436	        c = *++rnp;             /* advance to next character */
 26437	  }
 26438	
 26439	  /* To make /usr/ast/ equivalent to /usr/ast, skip trailing slashes. */

_________________________ Page 996    File: servers/fs/path.c _________________________
 26440	  while (c == '/' && rnp < &old_name[PATH_MAX]) c = *++rnp;
 26441	
 26442	  if (np < &string[NAME_MAX]) *np = '\0';       /* Terminate string */
 26443	
 26444	  if (rnp >= &old_name[PATH_MAX]) {
 26445	        err_code = ENAMETOOLONG;
 26446	        return((char *) 0);
 26447	  }
 26448	  return(rnp);
 26449	}
	
 26451	/*===========================================================================*
 26452	 *                              advance                                      *
 26453	 *===========================================================================*/
 26454	PUBLIC struct inode *advance(dirp, string)
 26455	struct inode *dirp;             /* inode for directory to be searched */
 26456	char string[NAME_MAX];          /* component name to look for */
 26457	{
 26458	/* Given a directory and a component of a path, look up the component in
 26459	 * the directory, find the inode, open it, and return a pointer to its inode
 26460	 * slot.  If it can't be done, return NIL_INODE.
 26461	 */
 26462	
 26463	  register struct inode *rip;
 26464	  struct inode *rip2;
 26465	  register struct super_block *sp;
 26466	  int r, inumb;
 26467	  dev_t mnt_dev;
 26468	  ino_t numb;
 26469	
 26470	  /* If 'string' is empty, yield same inode straight away. */
 26471	  if (string[0] == '\0') { return(get_inode(dirp->i_dev, (int) dirp->i_num)); }
 26472	
 26473	  /* Check for NIL_INODE. */
 26474	  if (dirp == NIL_INODE) { return(NIL_INODE); }
 26475	
 26476	  /* If 'string' is not present in the directory, signal error. */
 26477	  if ( (r = search_dir(dirp, string, &numb, LOOK_UP)) != OK) {
 26478	        err_code = r;
 26479	        return(NIL_INODE);
 26480	  }
 26481	
 26482	  /* Don't go beyond the current root directory, unless the string is dot2. */
 26483	  if (dirp == fp->fp_rootdir && strcmp(string, "..") == 0 && string
 26484	                return(get_inode(dirp->i_dev, (int) dirp->i_num));
 26485	
 26486	  /* The component has been found in the directory.  Get inode. */
 26487	  if ( (rip = get_inode(dirp->i_dev, (int) numb)) == NIL_INODE)  {
 26488	        return(NIL_INODE);
 26489	        }
 26490	
 26491	  if (rip->i_num == ROOT_INODE)
 26492	        if (dirp->i_num == ROOT_INODE) {
 26493	            if (string[1] == '.') {
 26494	                for (sp = &super_block[1]; sp < &super_block[NR_SUPERS];
 26495	                        if (sp->s_dev == rip->i_dev) {
 26496	                                /* Release the root inode.  Replace by the
 26497	                                 * inode mounted on.
 26498	                                 */
 26499	                                put_inode(rip);

_________________________ Page 997    File: servers/fs/path.c _________________________
 26500	                                mnt_dev = sp->s_imount->i_dev;
 26501	                                inumb = (int) sp->s_imount->i_num;
 26502	                                rip2 = get_inode(mnt_dev, inumb);
 26503	                                rip = advance(rip2, string);
 26504	                                put_inode(rip2);
 26505	                                break;
 26506	                        }
 26507	                }
 26508	            }
 26509	        }
 26510	  if (rip == NIL_INODE) return(NIL_INODE);
 26511	
 26512	  /* See if the inode is mounted on.  If so, switch to root directory of the
 26513	   * mounted file system.  The super_block provides the linkage between the
 26514	   * inode mounted on and the root directory of the mounted file system.
 26515	   */
 26516	  while (rip != NIL_INODE && rip->i_mount == I_MOUNT) {
 26517	        /* The inode is indeed mounted on. */
 26518	        for (sp = &super_block[0]; sp < &super_block[NR_SUPERS]; sp++) {
 26519	                if (sp->s_imount == rip) {
 26520	                        /* Release the inode mounted on.  Replace by the
 26521	                         * inode of the root inode of the mounted device.
 26522	                         */
 26523	                        put_inode(rip);
 26524	                        rip = get_inode(sp->s_dev, ROOT_INODE);
 26525	                        break;
 26526	                }
 26527	        }
 26528	  }
 26529	  return(rip);          /* return pointer to inode's component */
 26530	}
	
 26532	/*===========================================================================*
 26533	 *                              search_dir                                   *
 26534	 *===========================================================================*/
 26535	PUBLIC int search_dir(ldir_ptr, string, numb, flag)
 26536	register struct inode *ldir_ptr; /* ptr to inode for dir to search */
 26537	char string[NAME_MAX];           /* component to search for */
 26538	ino_t *numb;                     /* pointer to inode number */
 26539	int flag;                        /* LOOK_UP, ENTER, DELETE or IS_EMPTY */
 26540	{
 26541	/* This function searches the directory whose inode is pointed to by 'ldip':
 26542	 * if (flag == ENTER)  enter 'string' in the directory with inode # '*numb';
 26543	 * if (flag == DELETE) delete 'string' from the directory;
 26544	 * if (flag == LOOK_UP) search for 'string' and return inode # in 'numb';
 26545	 * if (flag == IS_EMPTY) return OK if only . and .. in dir else ENOTEMPTY;
 26546	 *
 26547	 *    if 'string' is dot1 or dot2, no access permissions are checked.
 26548	 */
 26549	
 26550	  register struct direct *dp = NULL;
 26551	  register struct buf *bp = NULL;
 26552	  int i, r, e_hit, t, match;
 26553	  mode_t bits;
 26554	  off_t pos;
 26555	  unsigned new_slots, old_slots;
 26556	  block_t b;
 26557	  struct super_block *sp;
 26558	  int extended = 0;
 26559	

_________________________ Page 998    File: servers/fs/path.c _________________________
 26560	  /* If 'ldir_ptr' is not a pointer to a dir inode, error. */
 26561	  if ( (ldir_ptr->i_mode & I_TYPE) != I_DIRECTORY) return(ENOTDIR);
 26562	
 26563	  r = OK;
 26564	
 26565	  if (flag != IS_EMPTY) {
 26566	        bits = (flag == LOOK_UP ? X_BIT : W_BIT | X_BIT);
 26567	
 26568	        if (string == dot1 || string == dot2) {
 26569	                if (flag != LOOK_UP) r = read_only(ldir_ptr);
 26570	                                     /* only a writable device is required. */
 26571	        }
 26572	        else r = forbidden(ldir_ptr, bits); /* check access permissions */
 26573	  }
 26574	  if (r != OK) return(r);
 26575	  
 26576	  /* Step through the directory one block at a time. */
 26577	  old_slots = (unsigned) (ldir_ptr->i_size/DIR_ENTRY_SIZE);
 26578	  new_slots = 0;
 26579	  e_hit = FALSE;
 26580	  match = 0;                    /* set when a string match occurs */
 26581	
 26582	  for (pos = 0; pos < ldir_ptr->i_size; pos += ldir_ptr->i_sp->s_block_size) {
 26583	        b = read_map(ldir_ptr, pos);    /* get block number */
 26584	
 26585	        /* Since directories don't have holes, 'b' cannot be NO_BLOCK. */
 26586	        bp = get_block(ldir_ptr->i_dev, b, NORMAL);     /* get a dir block */
 26587	
 26588	        if (bp == NO_BLOCK)
 26589	                panic(__FILE__,"get_block returned NO_BLOCK", NO_NUM);
 26590	
 26591	        /* Search a directory block. */
 26592	        for (dp = &bp->b_dir[0];
 26593	                dp < &bp->b_dir[NR_DIR_ENTRIES(ldir_ptr->i_sp->s_block_size)
 26594	                dp++) {
 26595	                if (++new_slots > old_slots) { /* not found, but room left */
 26596	                        if (flag == ENTER) e_hit = TRUE;
 26597	                        break;
 26598	                }
 26599	
 26600	                /* Match occurs if string found. */
 26601	                if (flag != ENTER && dp->d_ino != 0) {
 26602	                        if (flag == IS_EMPTY) {
 26603	                                /* If this test succeeds, dir is not empty. */
 26604	                                if (strcmp(dp->d_name, "." ) != 0 &&
 26605	                                    strcmp(dp->d_name, "..") != 0) match = 1;
 26606	                        } else {
 26607	                                if (strncmp(dp->d_name, string, NAME_MAX) == 0) {
 26608	                                        match = 1;
 26609	                                }
 26610	                        }
 26611	                }
 26612	
 26613	                if (match) {
 26614	                        /* LOOK_UP or DELETE found what it wanted. */
 26615	                        r = OK;
 26616	                        if (flag == IS_EMPTY) r = ENOTEMPTY;
 26617	                        else if (flag == DELETE) {
 26618	                                /* Save d_ino for recovery. */
 26619	                                t = NAME_MAX - sizeof(ino_t);

_________________________ Page 999    File: servers/fs/path.c _________________________
 26620	                                *((ino_t *) &dp->d_name[t]) = dp->d_ino;
 26621	                                dp->d_ino = 0;  /* erase entry */
 26622	                                bp->b_dirt = DIRTY;
 26623	                                ldir_ptr->i_update |= CTIME | MTIME;
 26624	                                ldir_ptr->i_dirt = DIRTY;
 26625	                        } else {
 26626	                                sp = ldir_ptr->i_sp;    /* 'flag' is LOOK_UP */
 26627	                                *numb = conv4(sp->s_native, (int) dp->d_ino);
 26628	                        }
 26629	                        put_block(bp, DIRECTORY_BLOCK);
 26630	                        return(r);
 26631	                }
 26632	
 26633	                /* Check for free slot for the benefit of ENTER. */
 26634	                if (flag == ENTER && dp->d_ino == 0) {
 26635	                        e_hit = TRUE;   /* we found a free slot */
 26636	                        break;
 26637	                }
 26638	        }
 26639	
 26640	        /* The whole block has been searched or ENTER has a free slot. */
 26641	        if (e_hit) break;       /* e_hit set if ENTER can be performed now */
 26642	        put_block(bp, DIRECTORY_BLOCK); /* otherwise, continue searching dir */
 26643	  }
 26644	
 26645	  /* The whole directory has now been searched. */
 26646	  if (flag != ENTER) {
 26647	        return(flag == IS_EMPTY ? OK : ENOENT);
 26648	  }
 26649	
 26650	  /* This call is for ENTER.  If no free slot has been found so far, try to
 26651	   * extend directory.
 26652	   */
 26653	  if (e_hit == FALSE) { /* directory is full and no room left in last block */
 26654	        new_slots++;            /* increase directory size by 1 entry */
 26655	        if (new_slots == 0) return(EFBIG); /* dir size limited by slot count */
 26656	        if ( (bp = new_block(ldir_ptr, ldir_ptr->i_size)) == NIL_BUF)
 26657	                return(err_code);
 26658	        dp = &bp->b_dir[0];
 26659	        extended = 1;
 26660	  }
 26661	
 26662	  /* 'bp' now points to a directory block with space. 'dp' points to slot. */
 26663	  (void) memset(dp->d_name, 0, (size_t) NAME_MAX); /* clear entry */
 26664	  for (i = 0; string[i] && i < NAME_MAX; i++) dp->d_name[i] = string[i];
 26665	  sp = ldir_ptr->i_sp; 
 26666	  dp->d_ino = conv4(sp->s_native, (int) *numb);
 26667	  bp->b_dirt = DIRTY;
 26668	  put_block(bp, DIRECTORY_BLOCK);
 26669	  ldir_ptr->i_update |= CTIME | MTIME;  /* mark mtime for update later */
 26670	  ldir_ptr->i_dirt = DIRTY;
 26671	  if (new_slots > old_slots) {
 26672	        ldir_ptr->i_size = (off_t) new_slots * DIR_ENTRY_SIZE;
 26673	        /* Send the change to disk if the directory is extended. */
 26674	        if (extended) rw_inode(ldir_ptr, WRITING);
 26675	  }
 26676	  return(OK);
 26677	}

_________________________ Page 1000    File: servers/fs/path.c _________________________

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/mount.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 26700	/* This file performs the MOUNT and UMOUNT system calls.
 26701	 *
 26702	 * The entry points into this file are
 26703	 *   do_mount:  perform the MOUNT system call
 26704	 *   do_umount: perform the UMOUNT system call
 26705	 */
 26706	
 26707	#include "fs.h"
 26708	#include <fcntl.h>
 26709	#include <minix/com.h>
 26710	#include <sys/stat.h>
 26711	#include "buf.h"
 26712	#include "file.h"
 26713	#include "fproc.h"
 26714	#include "inode.h"
 26715	#include "param.h"
 26716	#include "super.h"
 26717	
 26718	FORWARD _PROTOTYPE( dev_t name_to_dev, (char *path)                     );
 26719	
 26720	/*===========================================================================*
 26721	 *                              do_mount                                     *
 26722	 *===========================================================================*/
 26723	PUBLIC int do_mount()
 26724	{
 26725	/* Perform the mount(name, mfile, rd_only) system call. */
 26726	
 26727	  register struct inode *rip, *root_ip;
 26728	  struct super_block *xp, *sp;
 26729	  dev_t dev;
 26730	  mode_t bits;
 26731	  int rdir, mdir;               /* TRUE iff {root|mount} file is dir */
 26732	  int r, found;
 26733	
 26734	  /* Only the super-user may do MOUNT. */
 26735	  if (!super_user) return(EPERM);
 26736	
 26737	  /* If 'name' is not for a block special file, return error. */
 26738	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 26739	  if ( (dev = name_to_dev(user_path)) == NO_DEV) return(err_code);
 26740	
 26741	  /* Scan super block table to see if dev already mounted & find a free slot.*/
 26742	  sp = NIL_SUPER;
 26743	  found = FALSE;
 26744	  for (xp = &super_block[0]; xp < &super_block[NR_SUPERS]; xp++) {
 26745	        if (xp->s_dev == dev) found = TRUE;     /* is it mounted already? */
 26746	        if (xp->s_dev == NO_DEV) sp = xp;       /* record free slot */
 26747	  }
 26748	  if (found) return(EBUSY);     /* already mounted */
 26749	  if (sp == NIL_SUPER) return(ENFILE);  /* no super block available */
 26750	
 26751	  /* Open the device the file system lives on. */
 26752	  if (dev_open(dev, who, m_in.rd_only ? R_BIT : (R_BIT|W_BIT)) != OK) 
 26753	        return(EINVAL);
 26754	

_________________________ Page 1001    File: servers/fs/mount.c _________________________
 26755	  /* Make the cache forget about blocks it has open on the filesystem */
 26756	  (void) do_sync();
 26757	  invalidate(dev);
 26758	
 26759	  /* Fill in the super block. */
 26760	  sp->s_dev = dev;              /* read_super() needs to know which dev */
 26761	  r = read_super(sp);
 26762	
 26763	  /* Is it recognized as a Minix filesystem? */
 26764	  if (r != OK) {
 26765	        dev_close(dev);
 26766	        sp->s_dev = NO_DEV;
 26767	        return(r);
 26768	  }
 26769	
 26770	  /* Now get the inode of the file to be mounted on. */
 26771	  if (fetch_name(m_in.name2, m_in.name2_length, M1) != OK) {
 26772	        dev_close(dev);
 26773	        sp->s_dev = NO_DEV;
 26774	        return(err_code);
 26775	  }
 26776	  if ( (rip = eat_path(user_path)) == NIL_INODE) {
 26777	        dev_close(dev);
 26778	        sp->s_dev = NO_DEV;
 26779	        return(err_code);
 26780	  }
 26781	
 26782	  /* It may not be busy. */
 26783	  r = OK;
 26784	  if (rip->i_count > 1) r = EBUSY;
 26785	
 26786	  /* It may not be special. */
 26787	  bits = rip->i_mode & I_TYPE;
 26788	  if (bits == I_BLOCK_SPECIAL || bits == I_CHAR_SPECIAL) r = ENOTDIR;
 26789	
 26790	  /* Get the root inode of the mounted file system. */
 26791	  root_ip = NIL_INODE;          /* if 'r' not OK, make sure this is defined */
 26792	  if (r == OK) {
 26793	        if ( (root_ip = get_inode(dev, ROOT_INODE)) == NIL_INODE) r = err_code;
 26794	  }
 26795	  if (root_ip != NIL_INODE && root_ip->i_mode == 0) {
 26796	        r = EINVAL;
 26797	  }
 26798	
 26799	  /* File types of 'rip' and 'root_ip' may not conflict. */
 26800	  if (r == OK) {
 26801	        mdir = ((rip->i_mode & I_TYPE) == I_DIRECTORY);  /* TRUE iff dir */
 26802	        rdir = ((root_ip->i_mode & I_TYPE) == I_DIRECTORY);
 26803	        if (!mdir && rdir) r = EISDIR;
 26804	  }
 26805	
 26806	  /* If error, return the super block and both inodes; release the maps. */
 26807	  if (r != OK) {
 26808	        put_inode(rip);
 26809	        put_inode(root_ip);
 26810	        (void) do_sync();
 26811	        invalidate(dev);
 26812	        dev_close(dev);
 26813	        sp->s_dev = NO_DEV;
 26814	        return(r);

_________________________ Page 1002    File: servers/fs/mount.c _________________________
 26815	  }
 26816	
 26817	  /* Nothing else can go wrong.  Perform the mount. */
 26818	  rip->i_mount = I_MOUNT;       /* this bit says the inode is mounted on */
 26819	  sp->s_imount = rip;
 26820	  sp->s_isup = root_ip;
 26821	  sp->s_rd_only = m_in.rd_only;
 26822	  return(OK);
 26823	}
	
 26825	/*===========================================================================*
 26826	 *                              do_umount                                    *
 26827	 *===========================================================================*/
 26828	PUBLIC int do_umount()
 26829	{
 26830	/* Perform the umount(name) system call. */
 26831	  dev_t dev;
 26832	
 26833	  /* Only the super-user may do UMOUNT. */
 26834	  if (!super_user) return(EPERM);
 26835	
 26836	  /* If 'name' is not for a block special file, return error. */
 26837	  if (fetch_name(m_in.name, m_in.name_length, M3) != OK) return(err_code);
 26838	  if ( (dev = name_to_dev(user_path)) == NO_DEV) return(err_code);
 26839	
 26840	  return(unmount(dev));
 26841	}
	
 26843	/*===========================================================================*
 26844	 *                              unmount                                      *
 26845	 *===========================================================================*/
 26846	PUBLIC int unmount(dev)
 26847	Dev_t dev;
 26848	{
 26849	/* Unmount a file system by device number. */
 26850	  register struct inode *rip;
 26851	  struct super_block *sp, *sp1;
 26852	  int count;
 26853	
 26854	  /* See if the mounted device is busy.  Only 1 inode using it should be
 26855	   * open -- the root inode -- and that inode only 1 time.
 26856	   */
 26857	  count = 0;
 26858	  for (rip = &inode[0]; rip< &inode[NR_INODES]; rip++)
 26859	        if (rip->i_count > 0 && rip->i_dev == dev) count += rip->i_count;
 26860	  if (count > 1) return(EBUSY); /* can't umount a busy file system */
 26861	
 26862	  /* Find the super block. */
 26863	  sp = NIL_SUPER;
 26864	  for (sp1 = &super_block[0]; sp1 < &super_block[NR_SUPERS]; sp1++) {
 26865	        if (sp1->s_dev == dev) {
 26866	                sp = sp1;
 26867	                break;
 26868	        }
 26869	  }
 26870	
 26871	  /* Sync the disk, and invalidate cache. */
 26872	  (void) do_sync();             /* force any cached blocks out of memory */
 26873	  invalidate(dev);              /* invalidate cache entries for this dev */
 26874	  if (sp == NIL_SUPER) {

_________________________ Page 1003    File: servers/fs/mount.c _________________________
 26875	        return(EINVAL);
 26876	  }
 26877	
 26878	  /* Close the device the file system lives on. */
 26879	  dev_close(dev);
 26880	
 26881	  /* Finish off the unmount. */
 26882	  sp->s_imount->i_mount = NO_MOUNT;     /* inode returns to normal */
 26883	  put_inode(sp->s_imount);      /* release the inode mounted on */
 26884	  put_inode(sp->s_isup);        /* release the root inode of the mounted fs */
 26885	  sp->s_imount = NIL_INODE;
 26886	  sp->s_dev = NO_DEV;
 26887	  return(OK);
 26888	}
	
 26890	/*===========================================================================*
 26891	 *                              name_to_dev                                  *
 26892	 *===========================================================================*/
 26893	PRIVATE dev_t name_to_dev(path)
 26894	char *path;                     /* pointer to path name */
 26895	{
 26896	/* Convert the block special file 'path' to a device number.  If 'path'
 26897	 * is not a block special file, return error code in 'err_code'.
 26898	 */
 26899	
 26900	  register struct inode *rip;
 26901	  register dev_t dev;
 26902	
 26903	  /* If 'path' can't be opened, give up immediately. */
 26904	  if ( (rip = eat_path(path)) == NIL_INODE) return(NO_DEV);
 26905	
 26906	  /* If 'path' is not a block special file, return error. */
 26907	  if ( (rip->i_mode & I_TYPE) != I_BLOCK_SPECIAL) {
 26908	        err_code = ENOTBLK;
 26909	        put_inode(rip);
 26910	        return(NO_DEV);
 26911	  }
 26912	
 26913	  /* Extract the device number. */
 26914	  dev = (dev_t) rip->i_zone[0];
 26915	  put_inode(rip);
 26916	  return(dev);
 26917	}



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/link.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 27000	/* This file handles the LINK and UNLINK system calls.  It also deals with
 27001	 * deallocating the storage used by a file when the last UNLINK is done to a
 27002	 * file and the blocks must be returned to the free block pool.
 27003	 *
 27004	 * The entry points into this file are
 27005	 *   do_link:   perform the LINK system call
 27006	 *   do_unlink: perform the UNLINK and RMDIR system calls
 27007	 *   do_rename: perform the RENAME system call
 27008	 *   truncate:  release all the blocks associated with an inode
 27009	 */

_________________________ Page 1004    File: servers/fs/link.c _________________________
 27010	
 27011	#include "fs.h"
 27012	#include <sys/stat.h>
 27013	#include <string.h>
 27014	#include <minix/com.h>
 27015	#include <minix/callnr.h>
 27016	#include "buf.h"
 27017	#include "file.h"
 27018	#include "fproc.h"
 27019	#include "inode.h"
 27020	#include "param.h"
 27021	#include "super.h"
 27022	
 27023	#define SAME 1000
 27024	
 27025	FORWARD _PROTOTYPE( int remove_dir, (struct inode *rldirp, struct inode *rip,
 27026	                        char dir_name[NAME_MAX])                        );
 27027	
 27028	FORWARD _PROTOTYPE( int unlink_file, (struct inode *dirp, struct inode *rip,
 27029	                        char file_name[NAME_MAX])                       );
 27030	
 27031	/*===========================================================================*
 27032	 *                              do_link                                      *
 27033	 *===========================================================================*/
 27034	PUBLIC int do_link()
 27035	{
 27036	/* Perform the link(name1, name2) system call. */
 27037	
 27038	  register struct inode *ip, *rip;
 27039	  register int r;
 27040	  char string[NAME_MAX];
 27041	  struct inode *new_ip;
 27042	
 27043	  /* See if 'name' (file to be linked) exists. */
 27044	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 27045	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 27046	
 27047	  /* Check to see if the file has maximum number of links already. */
 27048	  r = OK;
 27049	  if (rip->i_nlinks >= (rip->i_sp->s_version == V1 ? CHAR_MAX : SHRT_MAX))
 27050	        r = EMLINK;
 27051	
 27052	  /* Only super_user may link to directories. */
 27053	  if (r == OK)
 27054	        if ( (rip->i_mode & I_TYPE) == I_DIRECTORY && !super_user) r = 
 27055	
 27056	  /* If error with 'name', return the inode. */
 27057	  if (r != OK) {
 27058	        put_inode(rip);
 27059	        return(r);
 27060	  }
 27061	
 27062	  /* Does the final directory of 'name2' exist? */
 27063	  if (fetch_name(m_in.name2, m_in.name2_length, M1) != OK) {
 27064	        put_inode(rip);
 27065	        return(err_code);
 27066	  }
 27067	  if ( (ip = last_dir(user_path, string)) == NIL_INODE) r = err_code;
 27068	
 27069	  /* If 'name2' exists in full (even if no space) set 'r' to error. */

_________________________ Page 1005    File: servers/fs/link.c _________________________
 27070	  if (r == OK) {
 27071	        if ( (new_ip = advance(ip, string)) == NIL_INODE) {
 27072	                r = err_code;
 27073	                if (r == ENOENT) r = OK;
 27074	        } else {
 27075	                put_inode(new_ip);
 27076	                r = EEXIST;
 27077	        }
 27078	  }
 27079	
 27080	  /* Check for links across devices. */
 27081	  if (r == OK)
 27082	        if (rip->i_dev != ip->i_dev) r = EXDEV;
 27083	
 27084	  /* Try to link. */
 27085	  if (r == OK)
 27086	        r = search_dir(ip, string, &rip->i_num, ENTER);
 27087	
 27088	  /* If success, register the linking. */
 27089	  if (r == OK) {
 27090	        rip->i_nlinks++;
 27091	        rip->i_update |= CTIME;
 27092	        rip->i_dirt = DIRTY;
 27093	  }
 27094	
 27095	  /* Done.  Release both inodes. */
 27096	  put_inode(rip);
 27097	  put_inode(ip);
 27098	  return(r);
 27099	}
	
 27101	/*===========================================================================*
 27102	 *                              do_unlink                                    *
 27103	 *===========================================================================*/
 27104	PUBLIC int do_unlink()
 27105	{
 27106	/* Perform the unlink(name) or rmdir(name) system call. The code for these two
 27107	 * is almost the same.  They differ only in some condition testing.  Unlink()
 27108	 * may be used by the superuser to do dangerous things; rmdir() may not.
 27109	 */
 27110	
 27111	  register struct inode *rip;
 27112	  struct inode *rldirp;
 27113	  int r;
 27114	  char string[NAME_MAX];
 27115	
 27116	  /* Get the last directory in the path. */
 27117	  if (fetch_name(m_in.name, m_in.name_length, M3) != OK) return(err_code);
 27118	  if ( (rldirp = last_dir(user_path, string)) == NIL_INODE)
 27119	        return(err_code);
 27120	
 27121	  /* The last directory exists.  Does the file also exist? */
 27122	  r = OK;
 27123	  if ( (rip = advance(rldirp, string)) == NIL_INODE) r = err_code;
 27124	
 27125	  /* If error, return inode. */
 27126	  if (r != OK) {
 27127	        put_inode(rldirp);
 27128	        return(r);
 27129	  }

_________________________ Page 1006    File: servers/fs/link.c _________________________
 27130	
 27131	  /* Do not remove a mount point. */
 27132	  if (rip->i_num == ROOT_INODE) {
 27133	        put_inode(rldirp);
 27134	        put_inode(rip);
 27135	        return(EBUSY);
 27136	  }
 27137	
 27138	  /* Now test if the call is allowed, separately for unlink() and rmdir(). */
 27139	  if (call_nr == UNLINK) {
 27140	        /* Only the su may unlink directories, but the su can unlink any dir.*/
 27141	        if ( (rip->i_mode & I_TYPE) == I_DIRECTORY && !super_user) r = 
 27142	
 27143	        /* Don't unlink a file if it is the root of a mounted file system. */
 27144	        if (rip->i_num == ROOT_INODE) r = EBUSY;
 27145	
 27146	        /* Actually try to unlink the file; fails if parent is mode 0 etc. */
 27147	        if (r == OK) r = unlink_file(rldirp, rip, string);
 27148	
 27149	  } else {
 27150	        r = remove_dir(rldirp, rip, string); /* call is RMDIR */
 27151	  }
 27152	
 27153	  /* If unlink was possible, it has been done, otherwise it has not. */
 27154	  put_inode(rip);
 27155	  put_inode(rldirp);
 27156	  return(r);
 27157	}
	
 27159	/*===========================================================================*
 27160	 *                              do_rename                                    *
 27161	 *===========================================================================*/
 27162	PUBLIC int do_rename()
 27163	{
 27164	/* Perform the rename(name1, name2) system call. */
 27165	
 27166	  struct inode *old_dirp, *old_ip;      /* ptrs to old dir, file inodes */
 27167	  struct inode *new_dirp, *new_ip;      /* ptrs to new dir, file inodes */
 27168	  struct inode *new_superdirp, *next_new_superdirp;
 27169	  int r = OK;                           /* error flag; initially no error */
 27170	  int odir, ndir;                       /* TRUE iff {old|new} file is dir */
 27171	  int same_pdir;                        /* TRUE iff parent dirs are the same */
 27172	  char old_name[NAME_MAX], new_name[NAME_MAX];
 27173	  ino_t numb;
 27174	  int r1;
 27175	  
 27176	  /* See if 'name1' (existing file) exists.  Get dir and file inodes. */
 27177	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 27178	  if ( (old_dirp = last_dir(user_path, old_name))==NIL_INODE) return(err_code);
 27179	
 27180	  if ( (old_ip = advance(old_dirp, old_name)) == NIL_INODE) r = err_code;
 27181	
 27182	  /* See if 'name2' (new name) exists.  Get dir and file inodes. */
 27183	  if (fetch_name(m_in.name2, m_in.name2_length, M1) != OK) r = err_code;
 27184	  if ( (new_dirp = last_dir(user_path, new_name)) == NIL_INODE) r = err_code;
 27185	  new_ip = advance(new_dirp, new_name); /* not required to exist */
 27186	
 27187	  if (old_ip != NIL_INODE)
 27188	        odir = ((old_ip->i_mode & I_TYPE) == I_DIRECTORY);  /* TRUE iff dir */
 27189	

_________________________ Page 1007    File: servers/fs/link.c _________________________
 27190	  /* If it is ok, check for a variety of possible errors. */
 27191	  if (r == OK) {
 27192	        same_pdir = (old_dirp == new_dirp);
 27193	
 27194	        /* The old inode must not be a superdirectory of the new last dir. */
 27195	        if (odir && !same_pdir) {
 27196	                dup_inode(new_superdirp = new_dirp);
 27197	                while (TRUE) {          /* may hang in a file system loop */
 27198	                        if (new_superdirp == old_ip) {
 27199	                                r = EINVAL;
 27200	                                break;
 27201	                        }
 27202	                        next_new_superdirp = advance(new_superdirp, dot2);
 27203	                        put_inode(new_superdirp);
 27204	                        if (next_new_superdirp == new_superdirp)
 27205	                                break;  /* back at system root directory */
 27206	                        new_superdirp = next_new_superdirp;
 27207	                        if (new_superdirp == NIL_INODE) {
 27208	                                /* Missing ".." entry.  Assume the worst. */
 27209	                                r = EINVAL;
 27210	                                break;
 27211	                        }
 27212	                }       
 27213	                put_inode(new_superdirp);
 27214	        }       
 27215	
 27216	        /* The old or new name must not be . or .. */
 27217	        if (strcmp(old_name, ".")==0 || strcmp(old_name, "..")==0 ||
 27218	            strcmp(new_name, ".")==0 || strcmp(new_name, "..")==0) r = EINVAL;
 27219	
 27220	        /* Both parent directories must be on the same device. */
 27221	        if (old_dirp->i_dev != new_dirp->i_dev) r = EXDEV;
 27222	
 27223	        /* Parent dirs must be writable, searchable and on a writable device */
 27224	        if ((r1 = forbidden(old_dirp, W_BIT | X_BIT)) != OK ||
 27225	            (r1 = forbidden(new_dirp, W_BIT | X_BIT)) != OK) r = r1;
 27226	
 27227	        /* Some tests apply only if the new path exists. */
 27228	        if (new_ip == NIL_INODE) {
 27229	                /* don't rename a file with a file system mounted on it. */
 27230	                if (old_ip->i_dev != old_dirp->i_dev) r = EXDEV;
 27231	                if (odir && new_dirp->i_nlinks >=
 27232	                    (new_dirp->i_sp->s_version == V1 ? CHAR_MAX : SHRT_MAX) && ...
 27233	                    !same_pdir && r == OK) r = EMLINK;
 27234	        } else {
 27235	                if (old_ip == new_ip) r = SAME; /* old=new */
 27236	
 27237	                /* has the old file or new file a file system mounted on it? */
 27238	                if (old_ip->i_dev != new_ip->i_dev) r = EXDEV;
 27239	
 27240	                ndir = ((new_ip->i_mode & I_TYPE) == I_DIRECTORY); /* dir ? */
 27241	                if (odir == TRUE && ndir == FALSE) r = ENOTDIR;
 27242	                if (odir == FALSE && ndir == TRUE) r = EISDIR;
 27243	        }
 27244	  }
 27245	
 27246	  /* If a process has another root directory than the system root, we might
 27247	   * "accidently" be moving it's working directory to a place where it's
 27248	   * root directory isn't a super directory of it anymore. This can make
 27249	   * the function chroot useless. If chroot will be used often we should

_________________________ Page 1008    File: servers/fs/link.c _________________________
 27250	   * probably check for it here.
 27251	   */
 27252	
 27253	  /* The rename will probably work. Only two things can go wrong now:
 27254	   * 1. being unable to remove the new file. (when new file already exists)
 27255	   * 2. being unable to make the new directory entry. (new file doesn't exists)
 27256	   *     [directory has to grow by one block and cannot because the disk
 27257	   *      is completely full].
 27258	   */
 27259	  if (r == OK) {
 27260	        if (new_ip != NIL_INODE) {
 27261	                  /* There is already an entry for 'new'. Try to remove it. */
 27262	                if (odir) 
 27263	                        r = remove_dir(new_dirp, new_ip, new_name);
 27264	                else 
 27265	                        r = unlink_file(new_dirp, new_ip, new_name);
 27266	        }
 27267	        /* if r is OK, the rename will succeed, while there is now an
 27268	         * unused entry in the new parent directory.
 27269	         */
 27270	  }
 27271	
 27272	  if (r == OK) {
 27273	        /* If the new name will be in the same parent directory as the old one,
 27274	         * first remove the old name to free an entry for the new name,
 27275	         * otherwise first try to create the new name entry to make sure
 27276	         * the rename will succeed.
 27277	         */
 27278	        numb = old_ip->i_num;           /* inode number of old file */
 27279	
 27280	        if (same_pdir) {
 27281	                r = search_dir(old_dirp, old_name, (ino_t *) 0, DELETE);
 27282	                                                /* shouldn't go wrong. */
 27283	                if (r==OK) (void) search_dir(old_dirp, new_name, &numb, ENTER);
 27284	        } else {
 27285	                r = search_dir(new_dirp, new_name, &numb, ENTER);
 27286	                if (r == OK)
 27287	                    (void) search_dir(old_dirp, old_name, (ino_t *) 0, DELETE);
 27288	        }
 27289	  }
 27290	  /* If r is OK, the ctime and mtime of old_dirp and new_dirp have been marked
 27291	   * for update in search_dir.
 27292	   */
 27293	
 27294	  if (r == OK && odir && !same_pdir) {
 27295	        /* Update the .. entry in the directory (still points to old_dirp). */
 27296	        numb = new_dirp->i_num;
 27297	        (void) unlink_file(old_ip, NIL_INODE, dot2);
 27298	        if (search_dir(old_ip, dot2, &numb, ENTER) == OK) {
 27299	                /* New link created. */
 27300	                new_dirp->i_nlinks++;
 27301	                new_dirp->i_dirt = DIRTY;
 27302	        }
 27303	  }
 27304	        
 27305	  /* Release the inodes. */
 27306	  put_inode(old_dirp);
 27307	  put_inode(old_ip);
 27308	  put_inode(new_dirp);
 27309	  put_inode(new_ip);

_________________________ Page 1009    File: servers/fs/link.c _________________________
 27310	  return(r == SAME ? OK : r);
 27311	}
	
 27313	/*===========================================================================*
 27314	 *                              truncate                                     *
 27315	 *===========================================================================*/
 27316	PUBLIC void truncate(rip)
 27317	register struct inode *rip;     /* pointer to inode to be truncated */
 27318	{
 27319	/* Remove all the zones from the inode 'rip' and mark it dirty. */
 27320	
 27321	  register block_t b;
 27322	  zone_t z, zone_size, z1;
 27323	  off_t position;
 27324	  int i, scale, file_type, waspipe, single, nr_indirects;
 27325	  struct buf *bp;
 27326	  dev_t dev;
 27327	
 27328	  file_type = rip->i_mode & I_TYPE;     /* check to see if file is special */
 27329	  if (file_type == I_CHAR_SPECIAL || file_type == I_BLOCK_SPECIAL) return;
 27330	  dev = rip->i_dev;             /* device on which inode resides */
 27331	  scale = rip->i_sp->s_log_zone_size;
 27332	  zone_size = (zone_t) rip->i_sp->s_block_size << scale;
 27333	  nr_indirects = rip->i_nindirs;
 27334	
 27335	  /* Pipes can shrink, so adjust size to make sure all zones are removed. */
 27336	  waspipe = rip->i_pipe == I_PIPE;      /* TRUE is this was a pipe */
 27337	  if (waspipe) rip->i_size = PIPE_SIZE(rip->i_sp->s_block_size);
 27338	
 27339	  /* Step through the file a zone at a time, finding and freeing the zones. */
 27340	  for (position = 0; position < rip->i_size; position += zone_size) {
 27341	        if ( (b = read_map(rip, position)) != NO_BLOCK) {
 27342	                z = (zone_t) b >> scale;
 27343	                free_zone(dev, z);
 27344	        }
 27345	  }
 27346	
 27347	  /* All the data zones have been freed.  Now free the indirect zones. */
 27348	  rip->i_dirt = DIRTY;
 27349	  if (waspipe) {
 27350	        wipe_inode(rip);        /* clear out inode for pipes */
 27351	        return;                 /* indirect slots contain file positions */
 27352	  }
 27353	  single = rip->i_ndzones;
 27354	  free_zone(dev, rip->i_zone[single]);  /* single indirect zone */
 27355	  if ( (z = rip->i_zone[single+1]) != NO_ZONE) {
 27356	        /* Free all the single indirect zones pointed to by the double. */
 27357	        b = (block_t) z << scale;
 27358	        bp = get_block(dev, b, NORMAL); /* get double indirect zone */
 27359	        for (i = 0; i < nr_indirects; i++) {
 27360	                z1 = rd_indir(bp, i);
 27361	                free_zone(dev, z1);
 27362	        }
 27363	
 27364	        /* Now free the double indirect zone itself. */
 27365	        put_block(bp, INDIRECT_BLOCK);
 27366	        free_zone(dev, z);
 27367	  }
 27368	
 27369	  /* Leave zone numbers for de(1) to recover file after an unlink(2).  */

_________________________ Page 1010    File: servers/fs/link.c _________________________
 27370	}
	
 27372	/*===========================================================================*
 27373	 *                              remove_dir                                   *
 27374	 *===========================================================================*/
 27375	PRIVATE int remove_dir(rldirp, rip, dir_name)
 27376	struct inode *rldirp;                   /* parent directory */
 27377	struct inode *rip;                      /* directory to be removed */
 27378	char dir_name[NAME_MAX];                /* name of directory to be removed */
 27379	{
 27380	  /* A directory file has to be removed. Five conditions have to met:
 27381	   *    - The file must be a directory
 27382	   *    - The directory must be empty (except for . and ..)
 27383	   *    - The final component of the path must not be . or ..
 27384	   *    - The directory must not be the root of a mounted file system
 27385	   *    - The directory must not be anybody's root/working directory
 27386	   */
 27387	
 27388	  int r;
 27389	  register struct fproc *rfp;
 27390	
 27391	  /* search_dir checks that rip is a directory too. */
 27392	  if ((r = search_dir(rip, "", (ino_t *) 0, IS_EMPTY)) != OK) return r;
 27393	
 27394	  if (strcmp(dir_name, ".") == 0 || strcmp(dir_name, "..") == 0)return(EINVAL);
 27395	  if (rip->i_num == ROOT_INODE) return(EBUSY); /* can't remove 'root' */
 27396	  
 27397	  for (rfp = &fproc[INIT_PROC_NR + 1]; rfp < &fproc[NR_PROCS]; rfp++)
 27398	        if (rfp->fp_workdir == rip || rfp->fp_rootdir == rip) return(EBUSY);
 27399	                                /* can't remove anybody's working dir */
 27400	
 27401	  /* Actually try to unlink the file; fails if parent is mode 0 etc. */
 27402	  if ((r = unlink_file(rldirp, rip, dir_name)) != OK) return r;
 27403	
 27404	  /* Unlink . and .. from the dir. The super user can link and unlink any dir,
 27405	   * so don't make too many assumptions about them.
 27406	   */
 27407	  (void) unlink_file(rip, NIL_INODE, dot1);
 27408	  (void) unlink_file(rip, NIL_INODE, dot2);
 27409	  return(OK);
 27410	}
	
 27412	/*===========================================================================*
 27413	 *                              unlink_file                                  *
 27414	 *===========================================================================*/
 27415	PRIVATE int unlink_file(dirp, rip, file_name)
 27416	struct inode *dirp;             /* parent directory of file */
 27417	struct inode *rip;              /* inode of file, may be NIL_INODE too. */
 27418	char file_name[NAME_MAX];       /* name of file to be removed */
 27419	{
 27420	/* Unlink 'file_name'; rip must be the inode of 'file_name' or NIL_INODE. */
 27421	
 27422	  ino_t numb;                   /* inode number */
 27423	  int   r;
 27424	
 27425	  /* If rip is not NIL_INODE, it is used to get faster access to the inode. */
 27426	  if (rip == NIL_INODE) {
 27427	        /* Search for file in directory and try to get its inode. */
 27428	        err_code = search_dir(dirp, file_name, &numb, LOOK_UP);
 27429	        if (err_code == OK) rip = get_inode(dirp->i_dev, (int) numb);

_________________________ Page 1011    File: servers/fs/link.c _________________________
 27430	        if (err_code != OK || rip == NIL_INODE) return(err_code);
 27431	  } else {
 27432	        dup_inode(rip);         /* inode will be returned with put_inode */
 27433	  }
 27434	
 27435	  r = search_dir(dirp, file_name, (ino_t *) 0, DELETE);
 27436	
 27437	  if (r == OK) {
 27438	        rip->i_nlinks--;        /* entry deleted from parent's dir */
 27439	        rip->i_update |= CTIME;
 27440	        rip->i_dirt = DIRTY;
 27441	  }
 27442	
 27443	  put_inode(rip);
 27444	  return(r);
 27445	}





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/stadir.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 27500	/* This file contains the code for performing four system calls relating to
 27501	 * status and directories.
 27502	 *
 27503	 * The entry points into this file are
 27504	 *   do_chdir:  perform the CHDIR system call
 27505	 *   do_chroot: perform the CHROOT system call
 27506	 *   do_stat:   perform the STAT system call
 27507	 *   do_fstat:  perform the FSTAT system call
 27508	 *   do_fstatfs: perform the FSTATFS system call
 27509	 */
 27510	
 27511	#include "fs.h"
 27512	#include <sys/stat.h>
 27513	#include <sys/statfs.h>
 27514	#include <minix/com.h>
 27515	#include "file.h"
 27516	#include "fproc.h"
 27517	#include "inode.h"
 27518	#include "param.h"
 27519	#include "super.h"
 27520	
 27521	FORWARD _PROTOTYPE( int change, (struct inode **iip, char *name_ptr, int len));
 27522	FORWARD _PROTOTYPE( int change_into, (struct inode **iip, struct inode *ip));
 27523	FORWARD _PROTOTYPE( int stat_inode, (struct inode *rip, struct filp *fil_ptr,
 27524	                        char *user_addr)                                );
 27525	
 27526	/*===========================================================================*
 27527	 *                              do_fchdir                                    *
 27528	 *===========================================================================*/
 27529	PUBLIC int do_fchdir()
 27530	{
 27531	        /* Change directory on already-opened fd. */
 27532	        struct filp *rfilp;
 27533	
 27534	        /* Is the file descriptor valid? */

_________________________ Page 1012    File: servers/fs/stadir.c _________________________
 27535	        if ( (rfilp = get_filp(m_in.fd)) == NIL_FILP) return(err_code);
 27536	        return change_into(&fp->fp_workdir, rfilp->filp_ino);
 27537	}
	
 27539	/*===========================================================================*
 27540	 *                              do_chdir                                     *
 27541	 *===========================================================================*/
 27542	PUBLIC int do_chdir()
 27543	{
 27544	/* Change directory.  This function is  also called by MM to simulate a chdir
 27545	 * in order to do EXEC, etc.  It also changes the root directory, the uids and
 27546	 * gids, and the umask. 
 27547	 */
 27548	
 27549	  int r;
 27550	  register struct fproc *rfp;
 27551	
 27552	  if (who == PM_PROC_NR) {
 27553	        rfp = &fproc[m_in.slot1];
 27554	        put_inode(fp->fp_rootdir);
 27555	        dup_inode(fp->fp_rootdir = rfp->fp_rootdir);
 27556	        put_inode(fp->fp_workdir);
 27557	        dup_inode(fp->fp_workdir = rfp->fp_workdir);
 27558	
 27559	        /* MM uses access() to check permissions.  To make this work, pretend
 27560	         * that the user's real ids are the same as the user's effective ids.
 27561	         * FS calls other than access() do not use the real ids, so are not
 27562	         * affected.
 27563	         */
 27564	        fp->fp_realuid =
 27565	        fp->fp_effuid = rfp->fp_effuid;
 27566	        fp->fp_realgid =
 27567	        fp->fp_effgid = rfp->fp_effgid;
 27568	        fp->fp_umask = rfp->fp_umask;
 27569	        return(OK);
 27570	  }
 27571	
 27572	  /* Perform the chdir(name) system call. */
 27573	  r = change(&fp->fp_workdir, m_in.name, m_in.name_length);
 27574	  return(r);
 27575	}
	
 27577	/*===========================================================================*
 27578	 *                              do_chroot                                    *
 27579	 *===========================================================================*/
 27580	PUBLIC int do_chroot()
 27581	{
 27582	/* Perform the chroot(name) system call. */
 27583	
 27584	  register int r;
 27585	
 27586	  if (!super_user) return(EPERM);       /* only su may chroot() */
 27587	  r = change(&fp->fp_rootdir, m_in.name, m_in.name_length);
 27588	  return(r);
 27589	}
	

_________________________ Page 1013    File: servers/fs/stadir.c _________________________
 27591	/*===========================================================================*
 27592	 *                              change                                       *
 27593	 *===========================================================================*/
 27594	PRIVATE int change(iip, name_ptr, len)
 27595	struct inode **iip;             /* pointer to the inode pointer for the dir */
 27596	char *name_ptr;                 /* pointer to the directory name to change to */
 27597	int len;                        /* length of the directory name string */
 27598	{
 27599	/* Do the actual work for chdir() and chroot(). */
 27600	  struct inode *rip;
 27601	
 27602	  /* Try to open the new directory. */
 27603	  if (fetch_name(name_ptr, len, M3) != OK) return(err_code);
 27604	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 27605	  return change_into(iip, rip);
 27606	}
	
 27608	/*===========================================================================*
 27609	 *                              change_into                                  *
 27610	 *===========================================================================*/
 27611	PRIVATE int change_into(iip, rip)
 27612	struct inode **iip;             /* pointer to the inode pointer for the dir */
 27613	struct inode *rip;              /* this is what the inode has to become */
 27614	{
 27615	  register int r;
 27616	
 27617	  /* It must be a directory and also be searchable. */
 27618	  if ( (rip->i_mode & I_TYPE) != I_DIRECTORY)
 27619	        r = ENOTDIR;
 27620	  else
 27621	        r = forbidden(rip, X_BIT);      /* check if dir is searchable */
 27622	
 27623	  /* If error, return inode. */
 27624	  if (r != OK) {
 27625	        put_inode(rip);
 27626	        return(r);
 27627	  }
 27628	
 27629	  /* Everything is OK.  Make the change. */
 27630	  put_inode(*iip);              /* release the old directory */
 27631	  *iip = rip;                   /* acquire the new one */
 27632	  return(OK);
 27633	}
	
 27635	/*===========================================================================*
 27636	 *                              do_stat                                      *
 27637	 *===========================================================================*/
 27638	PUBLIC int do_stat()
 27639	{
 27640	/* Perform the stat(name, buf) system call. */
 27641	
 27642	  register struct inode *rip;
 27643	  register int r;
 27644	
 27645	  /* Both stat() and fstat() use the same routine to do the real work.  That
 27646	   * routine expects an inode, so acquire it temporarily.
 27647	   */
 27648	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 27649	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 27650	  r = stat_inode(rip, NIL_FILP, m_in.name2);    /* actually do the work.*/

_________________________ Page 1014    File: servers/fs/stadir.c _________________________
 27651	  put_inode(rip);               /* release the inode */
 27652	  return(r);
 27653	}
	
 27655	/*===========================================================================*
 27656	 *                              do_fstat                                     *
 27657	 *===========================================================================*/
 27658	PUBLIC int do_fstat()
 27659	{
 27660	/* Perform the fstat(fd, buf) system call. */
 27661	
 27662	  register struct filp *rfilp;
 27663	
 27664	  /* Is the file descriptor valid? */
 27665	  if ( (rfilp = get_filp(m_in.fd)) == NIL_FILP) return(err_code);
 27666	
 27667	  return(stat_inode(rfilp->filp_ino, rfilp, m_in.buffer));
 27668	}
	
 27670	/*===========================================================================*
 27671	 *                              stat_inode                                   *
 27672	 *===========================================================================*/
 27673	PRIVATE int stat_inode(rip, fil_ptr, user_addr)
 27674	register struct inode *rip;     /* pointer to inode to stat */
 27675	struct filp *fil_ptr;           /* filp pointer, supplied by 'fstat' */
 27676	char *user_addr;                /* user space address where stat buf goes */
 27677	{
 27678	/* Common code for stat and fstat system calls. */
 27679	
 27680	  struct stat statbuf;
 27681	  mode_t mo;
 27682	  int r, s;
 27683	
 27684	  /* Update the atime, ctime, and mtime fields in the inode, if need be. */
 27685	  if (rip->i_update) update_times(rip);
 27686	
 27687	  /* Fill in the statbuf struct. */
 27688	  mo = rip->i_mode & I_TYPE;
 27689	
 27690	  /* true iff special */
 27691	  s = (mo == I_CHAR_SPECIAL || mo == I_BLOCK_SPECIAL);
 27692	
 27693	  statbuf.st_dev = rip->i_dev;
 27694	  statbuf.st_ino = rip->i_num;
 27695	  statbuf.st_mode = rip->i_mode;
 27696	  statbuf.st_nlink = rip->i_nlinks;
 27697	  statbuf.st_uid = rip->i_uid;
 27698	  statbuf.st_gid = rip->i_gid;
 27699	  statbuf.st_rdev = (dev_t) (s ? rip->i_zone[0] : NO_DEV);
 27700	  statbuf.st_size = rip->i_size;
 27701	
 27702	  if (rip->i_pipe == I_PIPE) {
 27703	        statbuf.st_mode &= ~I_REGULAR;  /* wipe out I_REGULAR bit for pipes */
 27704	        if (fil_ptr != NIL_FILP && fil_ptr->filp_mode & R_BIT) 
 27705	                statbuf.st_size -= fil_ptr->filp_pos;
 27706	  }
 27707	
 27708	  statbuf.st_atime = rip->i_atime;
 27709	  statbuf.st_mtime = rip->i_mtime;
 27710	  statbuf.st_ctime = rip->i_ctime;

_________________________ Page 1015    File: servers/fs/stadir.c _________________________
 27711	
 27712	  /* Copy the struct to user space. */
 27713	  r = sys_datacopy(FS_PROC_NR, (vir_bytes) &statbuf,
 27714	                who, (vir_bytes) user_addr, (phys_bytes) sizeof(statbuf));
 27715	  return(r);
 27716	}
	
 27718	/*===========================================================================*
 27719	 *                              do_fstatfs                                   *
 27720	 *===========================================================================*/
 27721	PUBLIC int do_fstatfs()
 27722	{
 27723	  /* Perform the fstatfs(fd, buf) system call. */
 27724	  struct statfs st;
 27725	  register struct filp *rfilp;
 27726	  int r;
 27727	
 27728	  /* Is the file descriptor valid? */
 27729	  if ( (rfilp = get_filp(m_in.fd)) == NIL_FILP) return(err_code);
 27730	
 27731	  st.f_bsize = rfilp->filp_ino->i_sp->s_block_size;
 27732	
 27733	  r = sys_datacopy(FS_PROC_NR, (vir_bytes) &st,
 27734	                who, (vir_bytes) m_in.buffer, (phys_bytes) sizeof(st));
 27735	
 27736	   return(r);
 27737	}
	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/protect.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 27800	/* This file deals with protection in the file system.  It contains the code
 27801	 * for four system calls that relate to protection.
 27802	 *
 27803	 * The entry points into this file are
 27804	 *   do_chmod:  perform the CHMOD system call
 27805	 *   do_chown:  perform the CHOWN system call
 27806	 *   do_umask:  perform the UMASK system call
 27807	 *   do_access: perform the ACCESS system call
 27808	 *   forbidden: check to see if a given access is allowed on a given inode
 27809	 */
 27810	
 27811	#include "fs.h"
 27812	#include <unistd.h>
 27813	#include <minix/callnr.h>
 27814	#include "buf.h"
 27815	#include "file.h"
 27816	#include "fproc.h"
 27817	#include "inode.h"
 27818	#include "param.h"
 27819	#include "super.h"
 27820	

_________________________ Page 1016    File: servers/fs/protect.c ________________________
 27821	/*===========================================================================*
 27822	 *                              do_chmod                                     *
 27823	 *===========================================================================*/
 27824	PUBLIC int do_chmod()
 27825	{
 27826	/* Perform the chmod(name, mode) system call. */
 27827	
 27828	  register struct inode *rip;
 27829	  register int r;
 27830	
 27831	  /* Temporarily open the file. */
 27832	  if (fetch_name(m_in.name, m_in.name_length, M3) != OK) return(err_code);
 27833	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 27834	
 27835	  /* Only the owner or the super_user may change the mode of a file.
 27836	   * No one may change the mode of a file on a read-only file system.
 27837	   */
 27838	  if (rip->i_uid != fp->fp_effuid && !super_user)
 27839	        r = EPERM;
 27840	  else
 27841	        r = read_only(rip);
 27842	
 27843	  /* If error, return inode. */
 27844	  if (r != OK)  {
 27845	        put_inode(rip);
 27846	        return(r);
 27847	  }
 27848	
 27849	  /* Now make the change. Clear setgid bit if file is not in caller's grp */
 27850	  rip->i_mode = (rip->i_mode & ~ALL_MODES) | (m_in.mode & ALL_MODES);
 27851	  if (!super_user && rip->i_gid != fp->fp_effgid)rip->i_mode &= ~I_SET_
 27852	  rip->i_update |= CTIME;
 27853	  rip->i_dirt = DIRTY;
 27854	
 27855	  put_inode(rip);
 27856	  return(OK);
 27857	}
	
 27859	/*===========================================================================*
 27860	 *                              do_chown                                     *
 27861	 *===========================================================================*/
 27862	PUBLIC int do_chown()
 27863	{
 27864	/* Perform the chown(name, owner, group) system call. */
 27865	
 27866	  register struct inode *rip;
 27867	  register int r;
 27868	
 27869	  /* Temporarily open the file. */
 27870	  if (fetch_name(m_in.name1, m_in.name1_length, M1) != OK) return(err_code);
 27871	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 27872	
 27873	  /* Not permitted to change the owner of a file on a read-only file sys. */
 27874	  r = read_only(rip);
 27875	  if (r == OK) {
 27876	        /* FS is R/W.  Whether call is allowed depends on ownership, etc. */
 27877	        if (super_user) {
 27878	                /* The super user can do anything. */
 27879	                rip->i_uid = m_in.owner;        /* others later */
 27880	        } else {

_________________________ Page 1017    File: servers/fs/protect.c ________________________
 27881	                /* Regular users can only change groups of their own files. */
 27882	                if (rip->i_uid != fp->fp_effuid) r = EPERM;
 27883	                if (rip->i_uid != m_in.owner) r = EPERM;  /* no giving away */
 27884	                if (fp->fp_effgid != m_in.group) r = EPERM;
 27885	        }
 27886	  }
 27887	  if (r == OK) {
 27888	        rip->i_gid = m_in.group;
 27889	        rip->i_mode &= ~(I_SET_UID_BIT | I_SET_GID_BIT);
 27890	        rip->i_update |= CTIME;
 27891	        rip->i_dirt = DIRTY;
 27892	  }
 27893	
 27894	  put_inode(rip);
 27895	  return(r);
 27896	}
	
 27898	/*===========================================================================*
 27899	 *                              do_umask                                     *
 27900	 *===========================================================================*/
 27901	PUBLIC int do_umask()
 27902	{
 27903	/* Perform the umask(co_mode) system call. */
 27904	  register mode_t r;
 27905	
 27906	  r = ~fp->fp_umask;            /* set 'r' to complement of old mask */
 27907	  fp->fp_umask = ~(m_in.co_mode & RWX_MODES);
 27908	  return(r);                    /* return complement of old mask */
 27909	}
	
 27911	/*===========================================================================*
 27912	 *                              do_access                                    *
 27913	 *===========================================================================*/
 27914	PUBLIC int do_access()
 27915	{
 27916	/* Perform the access(name, mode) system call. */
 27917	
 27918	  struct inode *rip;
 27919	  register int r;
 27920	
 27921	  /* First check to see if the mode is correct. */
 27922	  if ( (m_in.mode & ~(R_OK | W_OK | X_OK)) != 0 && m_in.mode != F_OK)
 27923	        return(EINVAL);
 27924	
 27925	  /* Temporarily open the file whose access is to be checked. */
 27926	  if (fetch_name(m_in.name, m_in.name_length, M3) != OK) return(err_code);
 27927	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 27928	
 27929	  /* Now check the permissions. */
 27930	  r = forbidden(rip, (mode_t) m_in.mode);
 27931	  put_inode(rip);
 27932	  return(r);
 27933	}
	
 27935	/*===========================================================================*
 27936	 *                              forbidden                                    *
 27937	 *===========================================================================*/
 27938	PUBLIC int forbidden(register struct inode *rip, mode_t access_desired)
 27939	{
 27940	/* Given a pointer to an inode, 'rip', and the access desired, determine

_________________________ Page 1018    File: servers/fs/protect.c ________________________
 27941	 * if the access is allowed, and if not why not.  The routine looks up the
 27942	 * caller's uid in the 'fproc' table.  If access is allowed, OK is returned
 27943	 * if it is forbidden, EACCES is returned.
 27944	 */
 27945	
 27946	  register struct inode *old_rip = rip;
 27947	  register struct super_block *sp;
 27948	  register mode_t bits, perm_bits;
 27949	  int r, shift, test_uid, test_gid, type;
 27950	
 27951	  if (rip->i_mount == I_MOUNT)  /* The inode is mounted on. */
 27952	        for (sp = &super_block[1]; sp < &super_block[NR_SUPERS]; sp++)
 27953	                if (sp->s_imount == rip) {
 27954	                        rip = get_inode(sp->s_dev, ROOT_INODE);
 27955	                        break;
 27956	                } /* if */
 27957	
 27958	  /* Isolate the relevant rwx bits from the mode. */
 27959	  bits = rip->i_mode;
 27960	  test_uid = (call_nr == ACCESS ? fp->fp_realuid : fp->fp_effuid);
 27961	  test_gid = (call_nr == ACCESS ? fp->fp_realgid : fp->fp_effgid);
 27962	  if (test_uid == SU_UID) {
 27963	        /* Grant read and write permission.  Grant search permission for
 27964	         * directories.  Grant execute permission (for non-directories) if
 27965	         * and only if one of the 'X' bits is set.
 27966	         */
 27967	        if ( (bits & I_TYPE) == I_DIRECTORY ||
 27968	             bits & ((X_BIT << 6) | (X_BIT << 3) | X_BIT))
 27969	                perm_bits = R_BIT | W_BIT | X_BIT;
 27970	        else
 27971	                perm_bits = R_BIT | W_BIT;
 27972	  } else {
 27973	        if (test_uid == rip->i_uid) shift = 6;          /* owner */
 27974	        else if (test_gid == rip->i_gid ) shift = 3;    /* group */
 27975	        else shift = 0;                                 /* other */
 27976	        perm_bits = (bits >> shift) & (R_BIT | W_BIT | X_BIT);
 27977	  }
 27978	
 27979	  /* If access desired is not a subset of what is allowed, it is refused. */
 27980	  r = OK;
 27981	  if ((perm_bits | access_desired) != perm_bits) r = EACCES;
 27982	
 27983	  /* Check to see if someone is trying to write on a file system that is
 27984	   * mounted read-only.
 27985	   */
 27986	  type = rip->i_mode & I_TYPE;
 27987	  if (r == OK)
 27988	        if (access_desired & W_BIT)
 27989	                r = read_only(rip);
 27990	
 27991	  if (rip != old_rip) put_inode(rip);
 27992	
 27993	  return(r);
 27994	}
	
 27996	/*===========================================================================*
 27997	 *                              read_only                                    *
 27998	 *===========================================================================*/
 27999	PUBLIC int read_only(ip)
 28000	struct inode *ip;               /* ptr to inode whose file sys is to be cked */

_________________________ Page 1019    File: servers/fs/protect.c ________________________
 28001	{
 28002	/* Check to see if the file system on which the inode 'ip' resides is mounted
 28003	 * read only.  If so, return EROFS, else return OK.
 28004	 */
 28005	
 28006	  register struct super_block *sp;
 28007	
 28008	  sp = ip->i_sp;
 28009	  return(sp->s_rd_only ? EROFS : OK);
 28010	}

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/dmap.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 28100	/* This file contains the table with device <-> driver mappings. It also
 28101	 * contains some routines to dynamically add and/ or remove device drivers
 28102	 * or change mappings.  
 28103	 */
 28104	
 28105	#include "fs.h"
 28106	#include "fproc.h"
 28107	#include <string.h>
 28108	#include <stdlib.h>
 28109	#include <ctype.h>
 28110	#include <unistd.h>
 28111	#include <minix/com.h>
 28112	#include "param.h"
 28113	
 28114	/* Some devices may or may not be there in the next table. */
 28115	#define DT(enable, opcl, io, driver, flags) \
 28116	  { (enable?(opcl):no_dev), (enable?(io):0), \
 28117	        (enable?(driver):0), (flags) },
 28118	#define NC(x) (NR_CTRLRS >= (x))
 28119	
 28120	/* The order of the entries here determines the mapping between major device
 28121	 * numbers and tasks.  The first entry (major device 0) is not used.  The
 28122	 * next entry is major device 1, etc.  Character and block devices can be
 28123	 * intermixed at random.  The ordering determines the device numbers in /dev/.
 28124	 * Note that FS knows the device number of /dev/ram/ to load the RAM disk.
 28125	 * Also note that the major device numbers used in /dev/ are NOT the same as 
 28126	 * the process numbers of the device drivers. 
 28127	 */
 28128	/*
 28129	  Driver enabled     Open/Cls  I/O     Driver #     Flags Device  File
 28130	  --------------     --------  ------  -----------  ----- ------  ----       
 28131	 */
 28132	struct dmap dmap[NR_DEVICES];                           /* actual map */ 
 28133	PRIVATE struct dmap init_dmap[] = {
 28134	  DT(1, no_dev,   0,       0,           0)              /* 0 = not used   */
 28135	  DT(1, gen_opcl, gen_io,  MEM_PROC_NR, 0)              /* 1 = /dev/mem   */
 28136	  DT(0, no_dev,   0,       0,           DMAP_MUTABLE)   /* 2 = /dev/fd0   */
 28137	  DT(0, no_dev,   0,       0,           DMAP_MUTABLE)   /* 3 = /dev/c0    */
 28138	  DT(1, tty_opcl, gen_io,  TTY_PROC_NR, 0)              /* 4 = /dev/tty00 */
 28139	  DT(1, ctty_opcl,ctty_io, TTY_PROC_NR, 0)              /* 5 = /dev/tty   */
 28140	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /* 6 = /dev/lp    */
 28141	  DT(1, no_dev,   0,       0,           DMAP_MUTABLE)   /* 7 = /dev/ip    */
 28142	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /* 8 = /dev/c1    */
 28143	  DT(0, 0,        0,       0,           DMAP_MUTABLE)   /* 9 = not used   */
 28144	  DT(0, no_dev,   0,       0,           DMAP_MUTABLE)   /*10 = /dev/c2    */

_________________________ Page 1020    File: servers/fs/dmap.c _________________________
 28145	  DT(0, 0,        0,       0,           DMAP_MUTABLE)   /*11 = not used   */
 28146	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /*12 = /dev/c3    */
 28147	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /*13 = /dev/audio */
 28148	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /*14 = /dev/mixer */
 28149	  DT(1, gen_opcl, gen_io,  LOG_PROC_NR, 0)              /*15 = /dev/klog  */
 28150	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /*16 = /dev/random*/
 28151	  DT(0, no_dev,   0,       NONE,        DMAP_MUTABLE)   /*17 = /dev/cmos  */
 28152	};
 28153	
 28154	/*===========================================================================*
 28155	 *                              do_devctl                                    *
 28156	 *===========================================================================*/
 28157	PUBLIC int do_devctl()
 28158	{
 28159	  int result;
 28160	
 28161	  switch(m_in.ctl_req) {
 28162	  case DEV_MAP:
 28163	      /* Try to update device mapping. */
 28164	      result = map_driver(m_in.dev_nr, m_in.driver_nr, m_in.dev_style);
 28165	      break;
 28166	  case DEV_UNMAP:
 28167	      result = ENOSYS;
 28168	      break;
 28169	  default:
 28170	      result = EINVAL;
 28171	  }
 28172	  return(result);
 28173	}
	
 28175	/*===========================================================================*
 28176	 *                              map_driver                                   *
 28177	 *===========================================================================*/
 28178	PUBLIC int map_driver(major, proc_nr, style)
 28179	int major;                      /* major number of the device */
 28180	int proc_nr;                    /* process number of the driver */
 28181	int style;                      /* style of the device */
 28182	{
 28183	/* Set a new device driver mapping in the dmap table. Given that correct 
 28184	 * arguments are given, this only works if the entry is mutable and the 
 28185	 * current driver is not busy. 
 28186	 * Normal error codes are returned so that this function can be used from
 28187	 * a system call that tries to dynamically install a new driver.
 28188	 */
 28189	  struct dmap *dp;
 28190	
 28191	  /* Get pointer to device entry in the dmap table. */
 28192	  if (major >= NR_DEVICES) return(ENODEV);
 28193	  dp = &dmap[major];            
 28194	        
 28195	  /* See if updating the entry is allowed. */
 28196	  if (! (dp->dmap_flags & DMAP_MUTABLE))  return(EPERM);
 28197	  if (dp->dmap_flags & DMAP_BUSY)  return(EBUSY);
 28198	
 28199	  /* Check process number of new driver. */
 28200	  if (! isokprocnr(proc_nr))  return(EINVAL);
 28201	
 28202	  /* Try to update the entry. */
 28203	  switch (style) {
 28204	  case STYLE_DEV:       dp->dmap_opcl = gen_opcl;       break;

_________________________ Page 1021    File: servers/fs/dmap.c _________________________
 28205	  case STYLE_TTY:       dp->dmap_opcl = tty_opcl;       break;
 28206	  case STYLE_CLONE:     dp->dmap_opcl = clone_opcl;     break;
 28207	  default:              return(EINVAL);
 28208	  }
 28209	  dp->dmap_io = gen_io;
 28210	  dp->dmap_driver = proc_nr;
 28211	  return(OK); 
 28212	}
	
 28214	/*===========================================================================*
 28215	 *                              build_dmap                                   *
 28216	 *===========================================================================*/
 28217	PUBLIC void build_dmap()
 28218	{
 28219	/* Initialize the table with all device <-> driver mappings. Then, map  
 28220	 * the boot driver to a controller and update the dmap table to that
 28221	 * selection. The boot driver and the controller it handles are set at 
 28222	 * the boot monitor.  
 28223	 */
 28224	  char driver[16];
 28225	  char *controller = "c##";
 28226	  int nr, major = -1;
 28227	  int i,s;
 28228	  struct dmap *dp;
 28229	
 28230	  /* Build table with device <-> driver mappings. */
 28231	  for (i=0; i<NR_DEVICES; i++) {
 28232	      dp = &dmap[i];            
 28233	      if (i < sizeof(init_dmap)/sizeof(struct dmap) && 
 28234	              init_dmap[i].dmap_opcl != no_dev) {       /* a preset driver */
 28235	          dp->dmap_opcl = init_dmap[i].dmap_opcl;
 28236	          dp->dmap_io = init_dmap[i].dmap_io;
 28237	          dp->dmap_driver = init_dmap[i].dmap_driver;
 28238	          dp->dmap_flags = init_dmap[i].dmap_flags;
 28239	      } else {                                          /* no default */
 28240	          dp->dmap_opcl = no_dev;
 28241	          dp->dmap_io = 0;
 28242	          dp->dmap_driver = 0;
 28243	          dp->dmap_flags = DMAP_MUTABLE;
 28244	      }
 28245	  }
 28246	
 28247	  /* Get settings of 'controller' and 'driver' at the boot monitor. */
 28248	  if ((s = env_get_param("label", driver, sizeof(driver))) != OK) 
 28249	      panic(__FILE__,"couldn't get boot monitor parameter 'driver'", s);
 28250	  if ((s = env_get_param("controller", controller, sizeof(controller))) != OK) 
 28251	      panic(__FILE__,"couldn't get boot monitor parameter 'controller'", s);
 28252	
 28253	  /* Determine major number to map driver onto. */
 28254	  if (controller[0] == 'f' && controller[1] == 'd') {
 28255	      major = FLOPPY_MAJOR;
 28256	  } 
 28257	  else if (controller[0] == 'c' && isdigit(controller[1])) {
 28258	      if ((nr = (unsigned) atoi(&controller[1])) > NR_CTRLRS)
 28259	          panic(__FILE__,"monitor 'controller' maximum 'c#' is", NR_CTRLRS);
 28260	      major = CTRLR(nr);
 28261	  } 
 28262	  else {
 28263	      panic(__FILE__,"monitor 'controller' syntax is 'c#' of 'fd'", NO_NUM); 
 28264	  }

_________________________ Page 1022    File: servers/fs/dmap.c _________________________
 28265	  
 28266	  /* Now try to set the actual mapping and report to the user. */
 28267	  if ((s=map_driver(major, DRVR_PROC_NR, STYLE_DEV)) != OK)
 28268	      panic(__FILE__,"map_driver failed",s);
 28269	  printf("Boot medium driver: %s driver mapped onto controller %s.\n",
 28270	      driver, controller);
 28271	}
	



++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/device.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 28300	/* When a needed block is not in the cache, it must be fetched from the disk.
 28301	 * Special character files also require I/O.  The routines for these are here.
 28302	 *
 28303	 * The entry points in this file are:
 28304	 *   dev_open:   FS opens a device
 28305	 *   dev_close:  FS closes a device
 28306	 *   dev_io:     FS does a read or write on a device
 28307	 *   dev_status: FS processes callback request alert
 28308	 *   gen_opcl:   generic call to a task to perform an open/close
 28309	 *   gen_io:     generic call to a task to perform an I/O operation
 28310	 *   no_dev:     open/close processing for devices that don't exist
 28311	 *   tty_opcl:   perform tty-specific processing for open/close
 28312	 *   ctty_opcl:  perform controlling-tty-specific processing for open/close
 28313	 *   ctty_io:    perform controlling-tty-specific processing for I/O
 28314	 *   do_ioctl:   perform the IOCTL system call
 28315	 *   do_setsid:  perform the SETSID system call (FS side)
 28316	 */
 28317	
 28318	#include "fs.h"
 28319	#include <fcntl.h>
 28320	#include <minix/callnr.h>
 28321	#include <minix/com.h>
 28322	#include "file.h"
 28323	#include "fproc.h"
 28324	#include "inode.h"
 28325	#include "param.h"
 28326	
 28327	#define ELEMENTS(a) (sizeof(a)/sizeof((a)[0]))
 28328	
 28329	extern int dmap_size;
 28330	
 28331	/*===========================================================================*
 28332	 *                              dev_open                                     *
 28333	 *===========================================================================*/
 28334	PUBLIC int dev_open(dev, proc, flags)
 28335	dev_t dev;                      /* device to open */
 28336	int proc;                       /* process to open for */
 28337	int flags;                      /* mode bits and flags */
 28338	{
 28339	  int major, r;
 28340	  struct dmap *dp;
 28341	
 28342	  /* Determine the major device number call the device class specific
 28343	   * open/close routine.  (This is the only routine that must check the
 28344	   * device number for being in range.  All others can trust this check.)

_________________________ Page 1023    File: servers/fs/device.c _________________________
 28345	   */
 28346	  major = (dev >> MAJOR) & BYTE;
 28347	  if (major >= NR_DEVICES) major = 0;
 28348	  dp = &dmap[major];
 28349	  r = (*dp->dmap_opcl)(DEV_OPEN, dev, proc, flags);
 28350	  if (r == SUSPEND) panic(__FILE__,"suspend on open from", dp->dmap_driver);
 28351	  return(r);
 28352	}
	
 28354	/*===========================================================================*
 28355	 *                              dev_close                                    *
 28356	 *===========================================================================*/
 28357	PUBLIC void dev_close(dev)
 28358	dev_t dev;                      /* device to close */
 28359	{
 28360	  (void) (*dmap[(dev >> MAJOR) & BYTE].dmap_opcl)(DEV_CLOSE, dev, 0, 0);
 28361	}
	
 28363	/*===========================================================================*
 28364	 *                              dev_status                                      *
 28365	 *===========================================================================*/
 28366	PUBLIC void dev_status(message *m)
 28367	{
 28368	        message st;
 28369	        int d, get_more = 1;
 28370	
 28371	        for(d = 0; d < NR_DEVICES; d++)
 28372	                if (dmap[d].dmap_driver == m->m_source)
 28373	                        break;
 28374	
 28375	        if (d >= NR_DEVICES)
 28376	                return;
 28377	
 28378	        do {
 28379	                int r;
 28380	                st.m_type = DEV_STATUS;
 28381	                if ((r=sendrec(m->m_source, &st)) != OK)
 28382	                        panic(__FILE__,"couldn't sendrec for DEV_STATUS", r);
 28383	
 28384	                switch(st.m_type) {
 28385	                        case DEV_REVIVE:
 28386	                                revive(st.REP_PROC_NR, st.REP_STATUS);
 28387	                                break;
 28388	                        case DEV_IO_READY:
 28389	                                select_notified(d, st.DEV_MINOR, st.DEV_SEL_OPS);
 28390	                                break;
 28391	                        default:
 28392	                                printf("FS: unrecognized reply %d to DEV_STATUS\n",
 28393	                                /* Fall through. */
 28394	                        case DEV_NO_STATUS:
 28395	                                get_more = 0;
 28396	                                break;
 28397	                }
 28398	        } while(get_more);
 28399	
 28400	        return;
 28401	}
	

_________________________ Page 1024    File: servers/fs/device.c _________________________
 28403	/*===========================================================================*
 28404	 *                              dev_io                                       *
 28405	 *===========================================================================*/
 28406	PUBLIC int dev_io(op, dev, proc, buf, pos, bytes, flags)
 28407	int op;                         /* DEV_READ, DEV_WRITE, DEV_IOCTL, etc. */
 28408	dev_t dev;                      /* major-minor device number */
 28409	int proc;                       /* in whose address space is buf? */
 28410	void *buf;                      /* virtual address of the buffer */
 28411	off_t pos;                      /* byte position */
 28412	int bytes;                      /* how many bytes to transfer */
 28413	int flags;                      /* special flags, like O_NONBLOCK */
 28414	{
 28415	/* Read or write from a device.  The parameter 'dev' tells which one. */
 28416	  struct dmap *dp;
 28417	  message dev_mess;
 28418	
 28419	  /* Determine task dmap. */
 28420	  dp = &dmap[(dev >> MAJOR) & BYTE];
 28421	
 28422	  /* Set up the message passed to task. */
 28423	  dev_mess.m_type   = op;
 28424	  dev_mess.DEVICE   = (dev >> MINOR) & BYTE;
 28425	  dev_mess.POSITION = pos;
 28426	  dev_mess.PROC_NR  = proc;
 28427	  dev_mess.ADDRESS  = buf;
 28428	  dev_mess.COUNT    = bytes;
 28429	  dev_mess.TTY_FLAGS = flags;
 28430	
 28431	  /* Call the task. */
 28432	  (*dp->dmap_io)(dp->dmap_driver, &dev_mess);
 28433	
 28434	  /* Task has completed.  See if call completed. */
 28435	  if (dev_mess.REP_STATUS == SUSPEND) {
 28436	        if (flags & O_NONBLOCK) {
 28437	                /* Not supposed to block. */
 28438	                dev_mess.m_type = CANCEL;
 28439	                dev_mess.PROC_NR = proc;
 28440	                dev_mess.DEVICE = (dev >> MINOR) & BYTE;
 28441	                (*dp->dmap_io)(dp->dmap_driver, &dev_mess);
 28442	                if (dev_mess.REP_STATUS == EINTR) dev_mess.REP_STATUS = EAGAIN;
 28443	        } else {
 28444	                /* Suspend user. */
 28445	                suspend(dp->dmap_driver);
 28446	                return(SUSPEND);
 28447	        }
 28448	  }
 28449	  return(dev_mess.REP_STATUS);
 28450	}
	
 28452	/*===========================================================================*
 28453	 *                              gen_opcl                                     *
 28454	 *===========================================================================*/
 28455	PUBLIC int gen_opcl(op, dev, proc, flags)
 28456	int op;                         /* operation, DEV_OPEN or DEV_CLOSE */
 28457	dev_t dev;                      /* device to open or close */
 28458	int proc;                       /* process to open/close for */
 28459	int flags;                      /* mode bits and flags */
 28460	{
 28461	/* Called from the dmap struct in table.c on opens & closes of special files.*/
 28462	  struct dmap *dp;

_________________________ Page 1025    File: servers/fs/device.c _________________________
 28463	  message dev_mess;
 28464	
 28465	  /* Determine task dmap. */
 28466	  dp = &dmap[(dev >> MAJOR) & BYTE];
 28467	
 28468	  dev_mess.m_type   = op;
 28469	  dev_mess.DEVICE   = (dev >> MINOR) & BYTE;
 28470	  dev_mess.PROC_NR  = proc;
 28471	  dev_mess.COUNT    = flags;
 28472	
 28473	  /* Call the task. */
 28474	  (*dp->dmap_io)(dp->dmap_driver, &dev_mess);
 28475	
 28476	  return(dev_mess.REP_STATUS);
 28477	}
	
 28479	/*===========================================================================*
 28480	 *                              tty_opcl                                     *
 28481	 *===========================================================================*/
 28482	PUBLIC int tty_opcl(op, dev, proc, flags)
 28483	int op;                         /* operation, DEV_OPEN or DEV_CLOSE */
 28484	dev_t dev;                      /* device to open or close */
 28485	int proc;                       /* process to open/close for */
 28486	int flags;                      /* mode bits and flags */
 28487	{
 28488	/* This procedure is called from the dmap struct on tty open/close. */
 28489	 
 28490	  int r;
 28491	  register struct fproc *rfp;
 28492	
 28493	  /* Add O_NOCTTY to the flags if this process is not a session leader, or
 28494	   * if it already has a controlling tty, or if it is someone elses
 28495	   * controlling tty.
 28496	   */
 28497	  if (!fp->fp_sesldr || fp->fp_tty != 0) {
 28498	        flags |= O_NOCTTY;
 28499	  } else {
 28500	        for (rfp = &fproc[0]; rfp < &fproc[NR_PROCS]; rfp++) {
 28501	                if (rfp->fp_tty == dev) flags |= O_NOCTTY;
 28502	        }
 28503	  }
 28504	
 28505	  r = gen_opcl(op, dev, proc, flags);
 28506	
 28507	  /* Did this call make the tty the controlling tty? */
 28508	  if (r == 1) {
 28509	        fp->fp_tty = dev;
 28510	        r = OK;
 28511	  }
 28512	  return(r);
 28513	}
	
 28515	/*===========================================================================*
 28516	 *                              ctty_opcl                                    *
 28517	 *===========================================================================*/
 28518	PUBLIC int ctty_opcl(op, dev, proc, flags)
 28519	int op;                         /* operation, DEV_OPEN or DEV_CLOSE */
 28520	dev_t dev;                      /* device to open or close */
 28521	int proc;                       /* process to open/close for */
 28522	int flags;                      /* mode bits and flags */

_________________________ Page 1026    File: servers/fs/device.c _________________________
 28523	{
 28524	/* This procedure is called from the dmap struct in table.c on opening/closing
 28525	 * /dev/tty, the magic device that translates to the controlling tty.
 28526	 */
 28527	 
 28528	  return(fp->fp_tty == 0 ? ENXIO : OK);
 28529	}
	
 28531	/*===========================================================================*
 28532	 *                              do_setsid                                    *
 28533	 *===========================================================================*/
 28534	PUBLIC int do_setsid()
 28535	{
 28536	/* Perform the FS side of the SETSID call, i.e. get rid of the controlling
 28537	 * terminal of a process, and make the process a session leader.
 28538	 */
 28539	  register struct fproc *rfp;
 28540	
 28541	  /* Only MM may do the SETSID call directly. */
 28542	  if (who != PM_PROC_NR) return(ENOSYS);
 28543	
 28544	  /* Make the process a session leader with no controlling tty. */
 28545	  rfp = &fproc[m_in.slot1];
 28546	  rfp->fp_sesldr = TRUE;
 28547	  rfp->fp_tty = 0;
 28548	  return(OK);
 28549	}
	
 28551	/*===========================================================================*
 28552	 *                              do_ioctl                                     *
 28553	 *===========================================================================*/
 28554	PUBLIC int do_ioctl()
 28555	{
 28556	/* Perform the ioctl(ls_fd, request, argx) system call (uses m2 fmt). */
 28557	
 28558	  struct filp *f;
 28559	  register struct inode *rip;
 28560	  dev_t dev;
 28561	
 28562	  if ( (f = get_filp(m_in.ls_fd)) == NIL_FILP) return(err_code);
 28563	  rip = f->filp_ino;            /* get inode pointer */
 28564	  if ( (rip->i_mode & I_TYPE) != I_CHAR_SPECIAL
 28565	        && (rip->i_mode & I_TYPE) != I_BLOCK_SPECIAL) return(ENOTTY);
 28566	  dev = (dev_t) rip->i_zone[0];
 28567	
 28568	  return(dev_io(DEV_IOCTL, dev, who, m_in.ADDRESS, 0L, 
 28569	        m_in.REQUEST, f->filp_flags));
 28570	}
	
 28572	/*===========================================================================*
 28573	 *                              gen_io                                       *
 28574	 *===========================================================================*/
 28575	PUBLIC void gen_io(task_nr, mess_ptr)
 28576	int task_nr;                    /* which task to call */
 28577	message *mess_ptr;              /* pointer to message for task */
 28578	{
 28579	/* All file system I/O ultimately comes down to I/O on major/minor device
 28580	 * pairs.  These lead to calls on the following routines via the dmap table.
 28581	 */
 28582	

_________________________ Page 1027    File: servers/fs/device.c _________________________
 28583	  int r, proc_nr;
 28584	  message local_m;
 28585	
 28586	  proc_nr = mess_ptr->PROC_NR;
 28587	  if (! isokprocnr(proc_nr)) {
 28588	      printf("FS: warning, got illegal process number (%d) from %d\n",
 28589	          mess_ptr->PROC_NR, mess_ptr->m_source);
 28590	      return;
 28591	  }
 28592	
 28593	  while ((r = sendrec(task_nr, mess_ptr)) == ELOCKED) {
 28594	        /* sendrec() failed to avoid deadlock. The task 'task_nr' is
 28595	         * trying to send a REVIVE message for an earlier request.
 28596	         * Handle it and go try again.
 28597	         */
 28598	        if ((r = receive(task_nr, &local_m)) != OK) {
 28599	                break;
 28600	        }
 28601	
 28602	        /* If we're trying to send a cancel message to a task which has just
 28603	         * sent a completion reply, ignore the reply and abort the cancel
 28604	         * request. The caller will do the revive for the process.
 28605	         */
 28606	        if (mess_ptr->m_type == CANCEL && local_m.REP_PROC_NR == proc_nr) {
 28607	                return;
 28608	        }
 28609	
 28610	        /* Otherwise it should be a REVIVE. */
 28611	        if (local_m.m_type != REVIVE) {
 28612	                printf(
 28613	                "fs: strange device reply from %d, type = %d, proc = %d (1)\n",
 28614	                        local_m.m_source,
 28615	                        local_m.m_type, local_m.REP_PROC_NR);
 28616	                continue;
 28617	        }
 28618	
 28619	        revive(local_m.REP_PROC_NR, local_m.REP_STATUS);
 28620	  }
 28621	
 28622	  /* The message received may be a reply to this call, or a REVIVE for some
 28623	   * other process.
 28624	   */
 28625	  for (;;) {
 28626	        if (r != OK) {
 28627	                if (r == EDEADDST) return;      /* give up */
 28628	                else panic(__FILE__,"call_task: can't send/receive", r);
 28629	        }
 28630	
 28631	        /* Did the process we did the sendrec() for get a result? */
 28632	        if (mess_ptr->REP_PROC_NR == proc_nr) {
 28633	                break;
 28634	        } else if (mess_ptr->m_type == REVIVE) {
 28635	                /* Otherwise it should be a REVIVE. */
 28636	                revive(mess_ptr->REP_PROC_NR, mess_ptr->REP_STATUS);
 28637	        } else {
 28638	                printf(
 28639	                "fs: strange device reply from %d, type = %d, proc = %d (2)\n",
 28640	                        mess_ptr->m_source,
 28641	                        mess_ptr->m_type, mess_ptr->REP_PROC_NR);
 28642	                return;

_________________________ Page 1028    File: servers/fs/device.c _________________________
 28643	        }
 28644	
 28645	        r = receive(task_nr, mess_ptr);
 28646	  }
 28647	}
	
 28649	/*===========================================================================*
 28650	 *                              ctty_io                                      *
 28651	 *===========================================================================*/
 28652	PUBLIC void ctty_io(task_nr, mess_ptr)
 28653	int task_nr;                    /* not used - for compatibility with dmap_t */
 28654	message *mess_ptr;              /* pointer to message for task */
 28655	{
 28656	/* This routine is only called for one device, namely /dev/tty.  Its job
 28657	 * is to change the message to use the controlling terminal, instead of the
 28658	 * major/minor pair for /dev/tty itself.
 28659	 */
 28660	
 28661	  struct dmap *dp;
 28662	
 28663	  if (fp->fp_tty == 0) {
 28664	        /* No controlling tty present anymore, return an I/O error. */
 28665	        mess_ptr->REP_STATUS = EIO;
 28666	  } else {
 28667	        /* Substitute the controlling terminal device. */
 28668	        dp = &dmap[(fp->fp_tty >> MAJOR) & BYTE];
 28669	        mess_ptr->DEVICE = (fp->fp_tty >> MINOR) & BYTE;
 28670	        (*dp->dmap_io)(dp->dmap_driver, mess_ptr);
 28671	  }
 28672	}
	
 28674	/*===========================================================================*
 28675	 *                              no_dev                                       *
 28676	 *===========================================================================*/
 28677	PUBLIC int no_dev(op, dev, proc, flags)
 28678	int op;                         /* operation, DEV_OPEN or DEV_CLOSE */
 28679	dev_t dev;                      /* device to open or close */
 28680	int proc;                       /* process to open/close for */
 28681	int flags;                      /* mode bits and flags */
 28682	{
 28683	/* Called when opening a nonexistent device. */
 28684	
 28685	  return(ENODEV);
 28686	}
	
 28688	/*===========================================================================*
 28689	 *                              clone_opcl                                   *
 28690	 *===========================================================================*/
 28691	PUBLIC int clone_opcl(op, dev, proc, flags)
 28692	int op;                         /* operation, DEV_OPEN or DEV_CLOSE */
 28693	dev_t dev;                      /* device to open or close */
 28694	int proc;                       /* process to open/close for */
 28695	int flags;                      /* mode bits and flags */
 28696	{
 28697	/* Some devices need special processing upon open.  Such a device is "cloned",
 28698	 * i.e. on a succesful open it is replaced by a new device with a new unique
 28699	 * minor device number.  This new device number identifies a new object (such
 28700	 * as a new network connection) that has been allocated within a task.
 28701	 */
 28702	  struct dmap *dp;

_________________________ Page 1029    File: servers/fs/device.c _________________________
 28703	  int minor;
 28704	  message dev_mess;
 28705	
 28706	  /* Determine task dmap. */
 28707	  dp = &dmap[(dev >> MAJOR) & BYTE];
 28708	  minor = (dev >> MINOR) & BYTE;
 28709	
 28710	  dev_mess.m_type   = op;
 28711	  dev_mess.DEVICE   = minor;
 28712	  dev_mess.PROC_NR  = proc;
 28713	  dev_mess.COUNT    = flags;
 28714	
 28715	  /* Call the task. */
 28716	  (*dp->dmap_io)(dp->dmap_driver, &dev_mess);
 28717	
 28718	  if (op == DEV_OPEN && dev_mess.REP_STATUS >= 0) {
 28719	        if (dev_mess.REP_STATUS != minor) {
 28720	                /* A new minor device number has been returned.  Create a
 28721	                 * temporary device file to hold it.
 28722	                 */
 28723	                struct inode *ip;
 28724	
 28725	                /* Device number of the new device. */
 28726	                dev = (dev & ~(BYTE << MINOR)) | (dev_mess.REP_STATUS < ...
 28727	
 28728	                ip = alloc_inode(root_dev, ALL_MODES | I_CHAR_SPECIAL);
 28729	                if (ip == NIL_INODE) {
 28730	                        /* Oops, that didn't work.  Undo open. */
 28731	                        (void) clone_opcl(DEV_CLOSE, dev, proc, 0);
 28732	                        return(err_code);
 28733	                }
 28734	                ip->i_zone[0] = dev;
 28735	
 28736	                put_inode(fp->fp_filp[m_in.fd]->filp_ino);
 28737	                fp->fp_filp[m_in.fd]->filp_ino = ip;
 28738	        }
 28739	        dev_mess.REP_STATUS = OK;
 28740	  }
 28741	  return(dev_mess.REP_STATUS);
 28742	}
	





++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
                                      servers/fs/time.c
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 28800	/* This file takes care of those system calls that deal with time.
 28801	 *
 28802	 * The entry points into this file are
 28803	 *   do_utime:          perform the UTIME system call
 28804	 *   do_stime:          PM informs FS about STIME system call
 28805	 */
 28806	
 28807	#include "fs.h"
 28808	#include <minix/callnr.h>
 28809	#include <minix/com.h>

_________________________ Page 1030    File: servers/fs/time.c _________________________
 28810	#include "file.h"
 28811	#include "fproc.h"
 28812	#include "inode.h"
 28813	#include "param.h"
 28814	
 28815	/*===========================================================================*
 28816	 *                              do_utime                                     *
 28817	 *===========================================================================*/
 28818	PUBLIC int do_utime()
 28819	{
 28820	/* Perform the utime(name, timep) system call. */
 28821	
 28822	  register struct inode *rip;
 28823	  register int len, r;
 28824	
 28825	  /* Adjust for case of 'timep' being NULL;
 28826	   * utime_strlen then holds the actual size: strlen(name)+1.
 28827	   */
 28828	  len = m_in.utime_length;
 28829	  if (len == 0) len = m_in.utime_strlen;
 28830	
 28831	  /* Temporarily open the file. */
 28832	  if (fetch_name(m_in.utime_file, len, M1) != OK) return(err_code);
 28833	  if ( (rip = eat_path(user_path)) == NIL_INODE) return(err_code);
 28834	
 28835	  /* Only the owner of a file or the super_user can change its time. */
 28836	  r = OK;
 28837	  if (rip->i_uid != fp->fp_effuid && !super_user) r = EPERM;
 28838	  if (m_in.utime_length == 0 && r != OK) r = forbidden(rip, W_BIT);
 28839	  if (read_only(rip) != OK) r = EROFS;  /* not even su can touch if R/O */
 28840	  if (r == OK) {
 28841	        if (m_in.utime_length == 0) {
 28842	                rip->i_atime = clock_time();
 28843	                rip->i_mtime = rip->i_atime;
 28844	        } else {
 28845	                rip->i_atime = m_in.utime_actime;
 28846	                rip->i_mtime = m_in.utime_modtime;
 28847	        }
 28848	        rip->i_update = CTIME;  /* discard any stale ATIME and MTIME flags */
 28849	        rip->i_dirt = DIRTY;
 28850	  }
 28851	
 28852	  put_inode(rip);
 28853	  return(r);
 28854	}
	
 28856	/*===========================================================================*
 28857	 *                              do_stime                                     *
 28858	 *===========================================================================*/
 28859	PUBLIC int do_stime()
 28860	{
 28861	/* Perform the stime(tp) system call. */
 28862	  boottime = (long) m_in.pm_stime; 
 28863	  return(OK);