mirror of
https://git.FreeBSD.org/src.git
synced 2024-12-22 11:17:19 +00:00
2217 lines
54 KiB
C
2217 lines
54 KiB
C
/* $Header: /src/pub/tcsh/sh.proc.c,v 3.81 2003/11/09 03:02:46 christos Exp $ */
|
|
/*
|
|
* sh.proc.c: Job manipulations
|
|
*/
|
|
/*-
|
|
* Copyright (c) 1980, 1991 The Regents of the University of California.
|
|
* All rights reserved.
|
|
*
|
|
* Redistribution and use in source and binary forms, with or without
|
|
* modification, are permitted provided that the following conditions
|
|
* are met:
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
* notice, this list of conditions and the following disclaimer.
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
* documentation and/or other materials provided with the distribution.
|
|
* 3. Neither the name of the University nor the names of its contributors
|
|
* may be used to endorse or promote products derived from this software
|
|
* without specific prior written permission.
|
|
*
|
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
* SUCH DAMAGE.
|
|
*/
|
|
#include "sh.h"
|
|
|
|
RCSID("$Id: sh.proc.c,v 3.81 2003/11/09 03:02:46 christos Exp $")
|
|
|
|
#include "ed.h"
|
|
#include "tc.h"
|
|
#include "tc.wait.h"
|
|
|
|
#ifdef WINNT_NATIVE
|
|
#undef POSIX
|
|
#define POSIX
|
|
#endif /* WINNT_NATIVE */
|
|
#ifdef aiws
|
|
# undef HZ
|
|
# define HZ 16
|
|
#endif /* aiws */
|
|
|
|
#if defined(_BSD) || (defined(IRIS4D) && __STDC__) || defined(__lucid) || defined(linux)
|
|
# define BSDWAIT
|
|
#endif /* _BSD || (IRIS4D && __STDC__) || __lucid || linux */
|
|
#ifndef WTERMSIG
|
|
# define WTERMSIG(w) (((union wait *) &(w))->w_termsig)
|
|
# ifndef BSDWAIT
|
|
# define BSDWAIT
|
|
# endif /* !BSDWAIT */
|
|
#endif /* !WTERMSIG */
|
|
#ifndef WEXITSTATUS
|
|
# define WEXITSTATUS(w) (((union wait *) &(w))->w_retcode)
|
|
#endif /* !WEXITSTATUS */
|
|
#ifndef WSTOPSIG
|
|
# define WSTOPSIG(w) (((union wait *) &(w))->w_stopsig)
|
|
#endif /* !WSTOPSIG */
|
|
|
|
#ifdef __osf__
|
|
# ifndef WCOREDUMP
|
|
# define WCOREDUMP(x) (_W_INT(x) & WCOREFLAG)
|
|
# endif
|
|
#endif
|
|
|
|
#ifndef WCOREDUMP
|
|
# ifdef BSDWAIT
|
|
# define WCOREDUMP(w) (((union wait *) &(w))->w_coredump)
|
|
# else /* !BSDWAIT */
|
|
# define WCOREDUMP(w) ((w) & 0200)
|
|
# endif /* !BSDWAIT */
|
|
#endif /* !WCOREDUMP */
|
|
|
|
/*
|
|
* C Shell - functions that manage processes, handling hanging, termination
|
|
*/
|
|
|
|
#define BIGINDEX 9 /* largest desirable job index */
|
|
|
|
#ifdef BSDTIMES
|
|
# ifdef convex
|
|
/* use 'cvxrusage' to get parallel statistics */
|
|
static struct cvxrusage zru = {{0L, 0L}, {0L, 0L}, 0L, 0L, 0L, 0L,
|
|
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
|
|
{0L, 0L}, 0LL, 0LL, 0LL, 0LL, 0L, 0L, 0L,
|
|
0LL, 0LL, {0L, 0L, 0L, 0L, 0L}};
|
|
# else
|
|
# if defined(SUNOS4) || defined(hp9000) || (defined(__alpha) && defined(__osf__))
|
|
static struct rusage zru = {{0L, 0L}, {0L, 0L}, 0L, 0L, 0L, 0L,
|
|
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L};
|
|
|
|
# else /* !SUNOS4 && !hp9000 && !(__alpha && __osf__) */
|
|
# ifdef masscomp
|
|
/*
|
|
* Initialization of this structure under RTU 4.1A & RTU 5.0 is problematic
|
|
* because the first two elements are unions of a time_t and a struct timeval.
|
|
* So we'll just have to trust the loader to do the "right thing", DAS DEC-90.
|
|
*/
|
|
static struct rusage zru;
|
|
# else /* masscomp */
|
|
static struct rusage zru = {{0L, 0L}, {0L, 0L}, 0, 0, 0, 0, 0, 0, 0,
|
|
0, 0, 0, 0, 0, 0};
|
|
# endif /* masscomp */
|
|
# endif /* SUNOS4 || hp9000 || (__alpha && __osf__) */
|
|
# endif /* convex */
|
|
#else /* !BSDTIMES */
|
|
# ifdef _SEQUENT_
|
|
static struct process_stats zru = {{0L, 0L}, {0L, 0L}, 0, 0, 0, 0, 0, 0, 0,
|
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
|
|
# else /* !_SEQUENT_ */
|
|
# ifdef _SX
|
|
static struct tms zru = {0, 0, 0, 0}, lru = {0, 0, 0, 0};
|
|
# else /* !_SX */
|
|
static struct tms zru = {0L, 0L, 0L, 0L}, lru = {0L, 0L, 0L, 0L};
|
|
# endif /* !_SX */
|
|
# endif /* !_SEQUENT_ */
|
|
#endif /* !BSDTIMES */
|
|
|
|
#ifndef RUSAGE_CHILDREN
|
|
# define RUSAGE_CHILDREN -1
|
|
#endif /* RUSAGE_CHILDREN */
|
|
|
|
static void pflushall __P((void));
|
|
static void pflush __P((struct process *));
|
|
static void pfree __P((struct process *));
|
|
static void pclrcurr __P((struct process *));
|
|
static void padd __P((struct command *));
|
|
static int pprint __P((struct process *, int));
|
|
static void ptprint __P((struct process *));
|
|
static void pads __P((Char *));
|
|
static void pkill __P((Char **, int));
|
|
static struct process *pgetcurr __P((struct process *));
|
|
static void okpcntl __P((void));
|
|
static void setttypgrp __P((int));
|
|
|
|
/*
|
|
* pchild - called at interrupt level by the SIGCHLD signal
|
|
* indicating that at least one child has terminated or stopped
|
|
* thus at least one wait system call will definitely return a
|
|
* childs status. Top level routines (like pwait) must be sure
|
|
* to mask interrupts when playing with the proclist data structures!
|
|
*/
|
|
sigret_t
|
|
/*ARGSUSED*/
|
|
pchild(snum)
|
|
int snum;
|
|
{
|
|
register struct process *pp;
|
|
register struct process *fp;
|
|
register int pid;
|
|
#if defined(BSDJOBS) || (!defined(BSDTIMES) && (defined(ODT) || defined(aiws) || defined(uts)))
|
|
extern int insource;
|
|
#endif /* BSDJOBS || (!BSDTIMES && (ODT || aiws || uts)) */
|
|
#ifdef BSDWAIT
|
|
union wait w;
|
|
#else /* !BSDWAIT */
|
|
int w;
|
|
#endif /* !BSDWAIT */
|
|
int jobflags;
|
|
#ifdef BSDTIMES
|
|
struct sysrusage ru;
|
|
#else /* !BSDTIMES */
|
|
# ifdef _SEQUENT_
|
|
struct process_stats ru;
|
|
struct process_stats cpst1, cpst2;
|
|
timeval_t tv;
|
|
# else /* !_SEQUENT_ */
|
|
struct tms proctimes;
|
|
|
|
USE(snum);
|
|
if (!timesdone) {
|
|
timesdone++;
|
|
(void) times(&shtimes);
|
|
}
|
|
# endif /* !_SEQUENT_ */
|
|
#endif /* !BSDTIMES */
|
|
|
|
#ifdef JOBDEBUG
|
|
xprintf("pchild()\n");
|
|
#endif /* JOBDEBUG */
|
|
|
|
/* Christos on where the signal(SIGCHLD, pchild) shoud be:
|
|
*
|
|
* I think that it should go *after* the wait, unlike most signal handlers.
|
|
*
|
|
* In release two (for which I have manuals), it says that wait will remove
|
|
* the first child from the queue of dead children.
|
|
* All the rest of the children that die while in the signal handler of the
|
|
* SIGC(H)LD, will be placed in the queue. If signal is called to re-establish
|
|
* the signal handler, and there are items in the queue, the process will
|
|
* receive another SIGC(H)LD before signal returns. BTW this is from the
|
|
* manual page on comp-sim... Maybe it is not applicable to the hp's, but
|
|
* I read on the news in comp.unix.wizards or comp.unix.questions yesterday
|
|
* that another person was claiming the the signal() call should be after
|
|
* the wait().
|
|
*/
|
|
|
|
loop:
|
|
errno = 0; /* reset, just in case */
|
|
#ifdef JOBDEBUG
|
|
xprintf("Waiting...\n");
|
|
flush();
|
|
#endif /* JOBDEBUG */
|
|
#ifndef WINNT_NATIVE
|
|
# ifdef BSDJOBS
|
|
# ifdef BSDTIMES
|
|
# ifdef convex
|
|
/* use 'cvxwait' to get parallel statistics */
|
|
pid = cvxwait(&w,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG), &ru);
|
|
# else
|
|
/* both a wait3 and rusage */
|
|
# if !defined(BSDWAIT) || defined(NeXT) || defined(MACH) || defined(linux) || (defined(IRIS4D) && (__STDC__ || defined(FUNCPROTO)) && SYSVREL <= 3) || defined(__lucid) || defined(__osf__)
|
|
pid = wait3(&w,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG), &ru);
|
|
# else /* BSDWAIT */
|
|
pid = wait3(&w.w_status,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG), &ru);
|
|
# endif /* BSDWAIT */
|
|
# endif /* convex */
|
|
# else /* !BSDTIMES */
|
|
# ifdef _SEQUENT_
|
|
(void) get_process_stats(&tv, PS_SELF, 0, &cpst1);
|
|
pid = waitpid(-1, &w,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG));
|
|
(void) get_process_stats(&tv, PS_SELF, 0, &cpst2);
|
|
pr_stat_sub(&cpst2, &cpst1, &ru);
|
|
# else /* !_SEQUENT_ */
|
|
# ifndef POSIX
|
|
/* we have a wait3, but no rusage stuff */
|
|
pid = wait3(&w.w_status,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG), 0);
|
|
# else /* POSIX */
|
|
pid = waitpid(-1, &w,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG));
|
|
# endif /* POSIX */
|
|
# endif /* !_SEQUENT_ */
|
|
# endif /* !BSDTIMES */
|
|
# else /* !BSDJOBS */
|
|
# ifdef BSDTIMES
|
|
# define HAVEwait3
|
|
/* both a wait3 and rusage */
|
|
# ifdef hpux
|
|
pid = wait3(&w.w_status, WNOHANG, 0);
|
|
# else /* !hpux */
|
|
pid = wait3(&w.w_status, WNOHANG, &ru);
|
|
# endif /* !hpux */
|
|
# else /* !BSDTIMES */
|
|
# ifdef ODT /* For Sco Unix 3.2.0 or ODT 1.0 */
|
|
# define HAVEwait3
|
|
pid = waitpid(-1, &w,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG));
|
|
# endif /* ODT */
|
|
# if defined(aiws) || defined(uts)
|
|
# define HAVEwait3
|
|
pid = wait3(&w.w_status,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG), 0);
|
|
# endif /* aiws || uts */
|
|
# ifndef HAVEwait3
|
|
# ifdef UNRELSIGS
|
|
/* no wait3, therefore no rusage */
|
|
/* on Sys V, this may hang. I hope it's not going to be a problem */
|
|
# ifdef _MINIX
|
|
pid = wait(&w);
|
|
# else /* !_MINIX */
|
|
pid = ourwait(&w.w_status);
|
|
# endif /* _MINIX */
|
|
# else /* !UNRELSIGS */
|
|
/*
|
|
* XXX: for greater than 3 we should use waitpid().
|
|
* but then again, SVR4 falls into the POSIX/BSDJOBS category.
|
|
*/
|
|
pid = wait(&w.w_status);
|
|
# endif /* !UNRELSIGS */
|
|
# endif /* !HAVEwait3 */
|
|
# endif /* !BSDTIMES */
|
|
# ifndef BSDSIGS
|
|
(void) sigset(SIGCHLD, pchild);
|
|
# endif /* !BSDSIGS */
|
|
# endif /* !BSDJOBS */
|
|
#else /* WINNT_NATIVE */
|
|
{
|
|
extern int insource;
|
|
pid = waitpid(-1, &w,
|
|
(setintr && (intty || insource) ? WNOHANG | WUNTRACED : WNOHANG));
|
|
}
|
|
#endif /* WINNT_NATIVE */
|
|
|
|
#ifdef JOBDEBUG
|
|
xprintf("parent %d pid %d, retval %x termsig %x retcode %x\n",
|
|
getpid(), pid, w, WTERMSIG(w), WEXITSTATUS(w));
|
|
flush();
|
|
#endif /* JOBDEBUG */
|
|
|
|
if ((pid == 0) || (pid == -1)) {
|
|
#ifdef JOBDEBUG
|
|
xprintf("errno == %d\n", errno);
|
|
#endif /* JOBDEBUG */
|
|
if (errno == EINTR) {
|
|
errno = 0;
|
|
goto loop;
|
|
}
|
|
pnoprocesses = pid == -1;
|
|
#ifndef SIGVOID
|
|
return (0);
|
|
#else /* !SIGVOID */
|
|
return;
|
|
#endif /* !SIGVOID */
|
|
}
|
|
for (pp = proclist.p_next; pp != NULL; pp = pp->p_next)
|
|
if (pid == pp->p_procid)
|
|
goto found;
|
|
#if !defined(BSDJOBS) && !defined(WINNT_NATIVE)
|
|
/* this should never have happened */
|
|
stderror(ERR_SYNC, pid);
|
|
xexit(0);
|
|
#else /* BSDJOBS || WINNT_NATIVE */
|
|
goto loop;
|
|
#endif /* !BSDJOBS && !WINNT_NATIVE */
|
|
found:
|
|
pp->p_flags &= ~(PRUNNING | PSTOPPED | PREPORTED);
|
|
if (WIFSTOPPED(w)) {
|
|
pp->p_flags |= PSTOPPED;
|
|
pp->p_reason = WSTOPSIG(w);
|
|
}
|
|
else {
|
|
if (pp->p_flags & (PTIME | PPTIME) || adrof(STRtime))
|
|
#ifndef BSDTIMES
|
|
# ifdef _SEQUENT_
|
|
(void) get_process_stats(&pp->p_etime, PS_SELF, NULL, NULL);
|
|
# else /* !_SEQUENT_ */
|
|
# ifndef COHERENT
|
|
pp->p_etime = times(&proctimes);
|
|
# else /* COHERENT */
|
|
pp->p_etime = HZ * time(NULL);
|
|
times(&proctimes);
|
|
# endif /* COHERENT */
|
|
# endif /* !_SEQUENT_ */
|
|
#else /* BSDTIMES */
|
|
(void) gettimeofday(&pp->p_etime, NULL);
|
|
#endif /* BSDTIMES */
|
|
|
|
|
|
#if defined(BSDTIMES) || defined(_SEQUENT_)
|
|
pp->p_rusage = ru;
|
|
#else /* !BSDTIMES && !_SEQUENT_ */
|
|
(void) times(&proctimes);
|
|
pp->p_utime = proctimes.tms_cutime - shtimes.tms_cutime;
|
|
pp->p_stime = proctimes.tms_cstime - shtimes.tms_cstime;
|
|
shtimes = proctimes;
|
|
#endif /* !BSDTIMES && !_SEQUENT_ */
|
|
if (WIFSIGNALED(w)) {
|
|
if (WTERMSIG(w) == SIGINT)
|
|
pp->p_flags |= PINTERRUPTED;
|
|
else
|
|
pp->p_flags |= PSIGNALED;
|
|
if (WCOREDUMP(w))
|
|
pp->p_flags |= PDUMPED;
|
|
pp->p_reason = WTERMSIG(w);
|
|
}
|
|
else {
|
|
pp->p_reason = WEXITSTATUS(w);
|
|
if (pp->p_reason != 0)
|
|
pp->p_flags |= PAEXITED;
|
|
else
|
|
pp->p_flags |= PNEXITED;
|
|
}
|
|
}
|
|
jobflags = 0;
|
|
fp = pp;
|
|
do {
|
|
if ((fp->p_flags & (PPTIME | PRUNNING | PSTOPPED)) == 0 &&
|
|
!child && adrof(STRtime) &&
|
|
#ifdef BSDTIMES
|
|
fp->p_rusage.ru_utime.tv_sec + fp->p_rusage.ru_stime.tv_sec
|
|
#else /* !BSDTIMES */
|
|
# ifdef _SEQUENT_
|
|
fp->p_rusage.ps_utime.tv_sec + fp->p_rusage.ps_stime.tv_sec
|
|
# else /* !_SEQUENT_ */
|
|
# ifndef POSIX
|
|
(fp->p_utime + fp->p_stime) / HZ
|
|
# else /* POSIX */
|
|
(fp->p_utime + fp->p_stime) / clk_tck
|
|
# endif /* POSIX */
|
|
# endif /* !_SEQUENT_ */
|
|
#endif /* !BSDTIMES */
|
|
>= atoi(short2str(varval(STRtime))))
|
|
fp->p_flags |= PTIME;
|
|
jobflags |= fp->p_flags;
|
|
} while ((fp = fp->p_friends) != pp);
|
|
pp->p_flags &= ~PFOREGND;
|
|
if (pp == pp->p_friends && (pp->p_flags & PPTIME)) {
|
|
pp->p_flags &= ~PPTIME;
|
|
pp->p_flags |= PTIME;
|
|
}
|
|
if ((jobflags & (PRUNNING | PREPORTED)) == 0) {
|
|
fp = pp;
|
|
do {
|
|
if (fp->p_flags & PSTOPPED)
|
|
fp->p_flags |= PREPORTED;
|
|
} while ((fp = fp->p_friends) != pp);
|
|
while (fp->p_procid != fp->p_jobid)
|
|
fp = fp->p_friends;
|
|
if (jobflags & PSTOPPED) {
|
|
if (pcurrent && pcurrent != fp)
|
|
pprevious = pcurrent;
|
|
pcurrent = fp;
|
|
}
|
|
else
|
|
pclrcurr(fp);
|
|
if (jobflags & PFOREGND) {
|
|
if (!(jobflags & (PSIGNALED | PSTOPPED | PPTIME) ||
|
|
#ifdef notdef
|
|
jobflags & PAEXITED ||
|
|
#endif /* notdef */
|
|
!eq(dcwd->di_name, fp->p_cwd->di_name))) {
|
|
/* PWP: print a newline after ^C */
|
|
if (jobflags & PINTERRUPTED) {
|
|
#ifdef SHORT_STRINGS
|
|
xputchar('\r' | QUOTE), xputchar('\n');
|
|
#else /* !SHORT_STRINGS */
|
|
xprintf("\215\n"); /* \215 is a quoted ^M */
|
|
#endif /* !SHORT_STRINGS */
|
|
}
|
|
#ifdef notdef
|
|
else if ((jobflags & (PTIME|PSTOPPED)) == PTIME)
|
|
ptprint(fp);
|
|
#endif /* notdef */
|
|
}
|
|
}
|
|
else {
|
|
if (jobflags & PNOTIFY || adrof(STRnotify)) {
|
|
#ifdef SHORT_STRINGS
|
|
xputchar('\r' | QUOTE), xputchar('\n');
|
|
#else /* !SHORT_STRINGS */
|
|
xprintf("\215\n"); /* \215 is a quoted ^M */
|
|
#endif /* !SHORT_STRINGS */
|
|
(void) pprint(pp, NUMBER | NAME | REASON);
|
|
if ((jobflags & PSTOPPED) == 0)
|
|
pflush(pp);
|
|
{
|
|
extern Char GettingInput;
|
|
|
|
if (GettingInput) {
|
|
errno = 0;
|
|
(void) Rawmode();
|
|
#ifdef notdef
|
|
/*
|
|
* don't really want to do that, because it
|
|
* will erase our message in case of multi-line
|
|
* input
|
|
*/
|
|
ClearLines();
|
|
#endif /* notdef */
|
|
ClearDisp();
|
|
Refresh();
|
|
}
|
|
}
|
|
}
|
|
else {
|
|
fp->p_flags |= PNEEDNOTE;
|
|
neednote++;
|
|
}
|
|
}
|
|
}
|
|
#if defined(BSDJOBS) || defined(HAVEwait3)
|
|
goto loop;
|
|
#endif /* BSDJOBS || HAVEwait3 */
|
|
}
|
|
|
|
void
|
|
pnote()
|
|
{
|
|
register struct process *pp;
|
|
int flags;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask;
|
|
#endif /* BSDSIGS */
|
|
|
|
neednote = 0;
|
|
for (pp = proclist.p_next; pp != NULL; pp = pp->p_next) {
|
|
if (pp->p_flags & PNEEDNOTE) {
|
|
#ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGCHLD));
|
|
#else /* !BSDSIGS */
|
|
(void) sighold(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
pp->p_flags &= ~PNEEDNOTE;
|
|
flags = pprint(pp, NUMBER | NAME | REASON);
|
|
if ((flags & (PRUNNING | PSTOPPED)) == 0)
|
|
pflush(pp);
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
static void
|
|
pfree(pp)
|
|
struct process *pp;
|
|
{
|
|
xfree((ptr_t) pp->p_command);
|
|
if (pp->p_cwd && --pp->p_cwd->di_count == 0)
|
|
if (pp->p_cwd->di_next == 0)
|
|
dfree(pp->p_cwd);
|
|
xfree((ptr_t) pp);
|
|
}
|
|
|
|
|
|
/*
|
|
* pwait - wait for current job to terminate, maintaining integrity
|
|
* of current and previous job indicators.
|
|
*/
|
|
void
|
|
pwait()
|
|
{
|
|
register struct process *fp, *pp;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask;
|
|
#endif /* BSDSIGS */
|
|
|
|
/*
|
|
* Here's where dead procs get flushed.
|
|
*/
|
|
#ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGCHLD));
|
|
#else /* !BSDSIGS */
|
|
(void) sighold(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
for (pp = (fp = &proclist)->p_next; pp != NULL; pp = (fp = pp)->p_next)
|
|
if (pp->p_procid == 0) {
|
|
fp->p_next = pp->p_next;
|
|
pfree(pp);
|
|
pp = fp;
|
|
}
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
# ifdef notdef
|
|
if (setintr)
|
|
sigignore(SIGINT);
|
|
# endif /* notdef */
|
|
#endif /* !BSDSIGS */
|
|
pjwait(pcurrjob);
|
|
}
|
|
|
|
|
|
/*
|
|
* pjwait - wait for a job to finish or become stopped
|
|
* It is assumed to be in the foreground state (PFOREGND)
|
|
*/
|
|
void
|
|
pjwait(pp)
|
|
register struct process *pp;
|
|
{
|
|
register struct process *fp;
|
|
int jobflags, reason;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask;
|
|
#endif /* BSDSIGS */
|
|
#ifdef UNRELSIGS
|
|
signalfun_t inthandler;
|
|
#endif /* UNRELSIGS */
|
|
while (pp->p_procid != pp->p_jobid)
|
|
pp = pp->p_friends;
|
|
fp = pp;
|
|
|
|
do {
|
|
if ((fp->p_flags & (PFOREGND | PRUNNING)) == PRUNNING)
|
|
xprintf(CGETS(17, 1, "BUG: waiting for background job!\n"));
|
|
} while ((fp = fp->p_friends) != pp);
|
|
/*
|
|
* Now keep pausing as long as we are not interrupted (SIGINT), and the
|
|
* target process, or any of its friends, are running
|
|
*/
|
|
fp = pp;
|
|
#ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGCHLD));
|
|
#endif /* BSDSIGS */
|
|
#ifdef UNRELSIGS
|
|
if (setintr)
|
|
inthandler = signal(SIGINT, SIG_IGN);
|
|
#endif /* UNRELSIGS */
|
|
for (;;) {
|
|
#ifndef BSDSIGS
|
|
(void) sighold(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
jobflags = 0;
|
|
do
|
|
jobflags |= fp->p_flags;
|
|
while ((fp = (fp->p_friends)) != pp);
|
|
if ((jobflags & PRUNNING) == 0)
|
|
break;
|
|
#ifdef JOBDEBUG
|
|
xprintf("%d starting to sigpause for SIGCHLD on %d\n",
|
|
getpid(), fp->p_procid);
|
|
#endif /* JOBDEBUG */
|
|
#ifdef BSDSIGS
|
|
/* (void) sigpause(sigblock((sigmask_t) 0) &~ sigmask(SIGCHLD)); */
|
|
(void) sigpause(omask & ~sigmask(SIGCHLD));
|
|
#else /* !BSDSIGS */
|
|
(void) sigpause(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
}
|
|
#ifdef JOBDEBUG
|
|
xprintf("%d returned from sigpause loop\n", getpid());
|
|
#endif /* JOBDEBUG */
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
#ifdef UNRELSIGS
|
|
if (setintr)
|
|
(void) signal(SIGINT, inthandler);
|
|
#endif /* UNRELSIGS */
|
|
#ifdef BSDJOBS
|
|
if (tpgrp > 0) /* get tty back */
|
|
(void) tcsetpgrp(FSHTTY, tpgrp);
|
|
#endif /* BSDJOBS */
|
|
if ((jobflags & (PSIGNALED | PSTOPPED | PTIME)) ||
|
|
!eq(dcwd->di_name, fp->p_cwd->di_name)) {
|
|
if (jobflags & PSTOPPED) {
|
|
xputchar('\n');
|
|
if (adrof(STRlistjobs)) {
|
|
Char *jobcommand[3];
|
|
|
|
jobcommand[0] = STRjobs;
|
|
if (eq(varval(STRlistjobs), STRlong))
|
|
jobcommand[1] = STRml;
|
|
else
|
|
jobcommand[1] = NULL;
|
|
jobcommand[2] = NULL;
|
|
|
|
dojobs(jobcommand, NULL);
|
|
(void) pprint(pp, SHELLDIR);
|
|
}
|
|
else
|
|
(void) pprint(pp, AREASON | SHELLDIR);
|
|
}
|
|
else
|
|
(void) pprint(pp, AREASON | SHELLDIR);
|
|
}
|
|
if ((jobflags & (PINTERRUPTED | PSTOPPED)) && setintr &&
|
|
(!gointr || !eq(gointr, STRminus))) {
|
|
if ((jobflags & PSTOPPED) == 0)
|
|
pflush(pp);
|
|
pintr1(0);
|
|
/* NOTREACHED */
|
|
}
|
|
reason = 0;
|
|
fp = pp;
|
|
do {
|
|
if (fp->p_reason)
|
|
reason = fp->p_flags & (PSIGNALED | PINTERRUPTED) ?
|
|
fp->p_reason | META : fp->p_reason;
|
|
} while ((fp = fp->p_friends) != pp);
|
|
/*
|
|
* Don't report on backquoted jobs, cause it will mess up
|
|
* their output.
|
|
*/
|
|
if ((reason != 0) && (adrof(STRprintexitvalue)) &&
|
|
(pp->p_flags & PBACKQ) == 0)
|
|
xprintf(CGETS(17, 2, "Exit %d\n"), reason);
|
|
set(STRstatus, putn(reason), VAR_READWRITE);
|
|
if (reason && exiterr)
|
|
exitstat();
|
|
pflush(pp);
|
|
}
|
|
|
|
/*
|
|
* dowait - wait for all processes to finish
|
|
*/
|
|
|
|
/*ARGSUSED*/
|
|
void
|
|
dowait(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask;
|
|
#endif /* BSDSIGS */
|
|
|
|
USE(c);
|
|
USE(v);
|
|
pjobs++;
|
|
#ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGCHLD));
|
|
loop:
|
|
#else /* !BSDSIGS */
|
|
if (setintr)
|
|
(void) sigrelse(SIGINT);
|
|
loop:
|
|
(void) sighold(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
for (pp = proclist.p_next; pp; pp = pp->p_next)
|
|
if (pp->p_procid && /* pp->p_procid == pp->p_jobid && */
|
|
pp->p_flags & PRUNNING) {
|
|
#ifdef BSDSIGS
|
|
(void) sigpause((sigmask_t) 0);
|
|
#else /* !BSDSIGS */
|
|
(void) sigpause(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
goto loop;
|
|
}
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
pjobs = 0;
|
|
}
|
|
|
|
/*
|
|
* pflushall - flush all jobs from list (e.g. at fork())
|
|
*/
|
|
static void
|
|
pflushall()
|
|
{
|
|
register struct process *pp;
|
|
|
|
for (pp = proclist.p_next; pp != NULL; pp = pp->p_next)
|
|
if (pp->p_procid)
|
|
pflush(pp);
|
|
}
|
|
|
|
/*
|
|
* pflush - flag all process structures in the same job as the
|
|
* the argument process for deletion. The actual free of the
|
|
* space is not done here since pflush is called at interrupt level.
|
|
*/
|
|
static void
|
|
pflush(pp)
|
|
register struct process *pp;
|
|
{
|
|
register struct process *np;
|
|
register int idx;
|
|
|
|
if (pp->p_procid == 0) {
|
|
xprintf(CGETS(17, 3, "BUG: process flushed twice"));
|
|
return;
|
|
}
|
|
while (pp->p_procid != pp->p_jobid)
|
|
pp = pp->p_friends;
|
|
pclrcurr(pp);
|
|
if (pp == pcurrjob)
|
|
pcurrjob = 0;
|
|
idx = pp->p_index;
|
|
np = pp;
|
|
do {
|
|
np->p_index = np->p_procid = 0;
|
|
np->p_flags &= ~PNEEDNOTE;
|
|
} while ((np = np->p_friends) != pp);
|
|
if (idx == pmaxindex) {
|
|
for (np = proclist.p_next, idx = 0; np; np = np->p_next)
|
|
if (np->p_index > idx)
|
|
idx = np->p_index;
|
|
pmaxindex = idx;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* pclrcurr - make sure the given job is not the current or previous job;
|
|
* pp MUST be the job leader
|
|
*/
|
|
static void
|
|
pclrcurr(pp)
|
|
register struct process *pp;
|
|
{
|
|
if (pp == pcurrent) {
|
|
if (pprevious != NULL) {
|
|
pcurrent = pprevious;
|
|
pprevious = pgetcurr(pp);
|
|
}
|
|
else {
|
|
pcurrent = pgetcurr(pp);
|
|
pprevious = pgetcurr(pp);
|
|
}
|
|
}
|
|
else if (pp == pprevious)
|
|
pprevious = pgetcurr(pp);
|
|
}
|
|
|
|
/* +4 here is 1 for '\0', 1 ea for << >& >> */
|
|
static Char command[PMAXLEN + 4];
|
|
static int cmdlen;
|
|
static Char *cmdp;
|
|
|
|
/* GrP
|
|
* unparse - Export padd() functionality
|
|
*/
|
|
Char *
|
|
unparse(t)
|
|
register struct command *t;
|
|
{
|
|
cmdp = command;
|
|
cmdlen = 0;
|
|
padd(t);
|
|
*cmdp++ = '\0';
|
|
return Strsave(command);
|
|
}
|
|
|
|
|
|
/*
|
|
* palloc - allocate a process structure and fill it up.
|
|
* an important assumption is made that the process is running.
|
|
*/
|
|
void
|
|
palloc(pid, t)
|
|
int pid;
|
|
register struct command *t;
|
|
{
|
|
register struct process *pp;
|
|
int i;
|
|
|
|
pp = (struct process *) xcalloc(1, (size_t) sizeof(struct process));
|
|
pp->p_procid = pid;
|
|
pp->p_flags = ((t->t_dflg & F_AMPERSAND) ? 0 : PFOREGND) | PRUNNING;
|
|
if (t->t_dflg & F_TIME)
|
|
pp->p_flags |= PPTIME;
|
|
if (t->t_dflg & F_BACKQ)
|
|
pp->p_flags |= PBACKQ;
|
|
if (t->t_dflg & F_HUP)
|
|
pp->p_flags |= PHUP;
|
|
cmdp = command;
|
|
cmdlen = 0;
|
|
padd(t);
|
|
*cmdp++ = 0;
|
|
if (t->t_dflg & F_PIPEOUT) {
|
|
pp->p_flags |= PPOU;
|
|
if (t->t_dflg & F_STDERR)
|
|
pp->p_flags |= PDIAG;
|
|
}
|
|
pp->p_command = Strsave(command);
|
|
if (pcurrjob) {
|
|
struct process *fp;
|
|
|
|
/* careful here with interrupt level */
|
|
pp->p_cwd = 0;
|
|
pp->p_index = pcurrjob->p_index;
|
|
pp->p_friends = pcurrjob;
|
|
pp->p_jobid = pcurrjob->p_procid;
|
|
for (fp = pcurrjob; fp->p_friends != pcurrjob; fp = fp->p_friends)
|
|
continue;
|
|
fp->p_friends = pp;
|
|
}
|
|
else {
|
|
pcurrjob = pp;
|
|
pp->p_jobid = pid;
|
|
pp->p_friends = pp;
|
|
pp->p_cwd = dcwd;
|
|
dcwd->di_count++;
|
|
if (pmaxindex < BIGINDEX)
|
|
pp->p_index = ++pmaxindex;
|
|
else {
|
|
struct process *np;
|
|
|
|
for (i = 1;; i++) {
|
|
for (np = proclist.p_next; np; np = np->p_next)
|
|
if (np->p_index == i)
|
|
goto tryagain;
|
|
pp->p_index = i;
|
|
if (i > pmaxindex)
|
|
pmaxindex = i;
|
|
break;
|
|
tryagain:;
|
|
}
|
|
}
|
|
if (pcurrent == NULL)
|
|
pcurrent = pp;
|
|
else if (pprevious == NULL)
|
|
pprevious = pp;
|
|
}
|
|
pp->p_next = proclist.p_next;
|
|
proclist.p_next = pp;
|
|
#ifdef BSDTIMES
|
|
(void) gettimeofday(&pp->p_btime, NULL);
|
|
#else /* !BSDTIMES */
|
|
# ifdef _SEQUENT_
|
|
(void) get_process_stats(&pp->p_btime, PS_SELF, NULL, NULL);
|
|
# else /* !_SEQUENT_ */
|
|
{
|
|
struct tms tmptimes;
|
|
|
|
# ifndef COHERENT
|
|
pp->p_btime = times(&tmptimes);
|
|
# else /* !COHERENT */
|
|
pp->p_btime = HZ * time(NULL);
|
|
times(&tmptimes);
|
|
# endif /* !COHERENT */
|
|
}
|
|
# endif /* !_SEQUENT_ */
|
|
#endif /* !BSDTIMES */
|
|
}
|
|
|
|
static void
|
|
padd(t)
|
|
register struct command *t;
|
|
{
|
|
Char **argp;
|
|
|
|
if (t == 0)
|
|
return;
|
|
switch (t->t_dtyp) {
|
|
|
|
case NODE_PAREN:
|
|
pads(STRLparensp);
|
|
padd(t->t_dspr);
|
|
pads(STRspRparen);
|
|
break;
|
|
|
|
case NODE_COMMAND:
|
|
for (argp = t->t_dcom; *argp; argp++) {
|
|
pads(*argp);
|
|
if (argp[1])
|
|
pads(STRspace);
|
|
}
|
|
break;
|
|
|
|
case NODE_OR:
|
|
case NODE_AND:
|
|
case NODE_PIPE:
|
|
case NODE_LIST:
|
|
padd(t->t_dcar);
|
|
switch (t->t_dtyp) {
|
|
case NODE_OR:
|
|
pads(STRspor2sp);
|
|
break;
|
|
case NODE_AND:
|
|
pads(STRspand2sp);
|
|
break;
|
|
case NODE_PIPE:
|
|
pads(STRsporsp);
|
|
break;
|
|
case NODE_LIST:
|
|
pads(STRsemisp);
|
|
break;
|
|
default:
|
|
break;
|
|
}
|
|
padd(t->t_dcdr);
|
|
return;
|
|
|
|
default:
|
|
break;
|
|
}
|
|
if ((t->t_dflg & F_PIPEIN) == 0 && t->t_dlef) {
|
|
pads((t->t_dflg & F_READ) ? STRspLarrow2sp : STRspLarrowsp);
|
|
pads(t->t_dlef);
|
|
}
|
|
if ((t->t_dflg & F_PIPEOUT) == 0 && t->t_drit) {
|
|
pads((t->t_dflg & F_APPEND) ? STRspRarrow2 : STRspRarrow);
|
|
if (t->t_dflg & F_STDERR)
|
|
pads(STRand);
|
|
pads(STRspace);
|
|
pads(t->t_drit);
|
|
}
|
|
}
|
|
|
|
static void
|
|
pads(cp)
|
|
Char *cp;
|
|
{
|
|
register int i;
|
|
|
|
/*
|
|
* Avoid the Quoted Space alias hack! Reported by:
|
|
* sam@john-bigboote.ICS.UCI.EDU (Sam Horrocks)
|
|
*/
|
|
if (cp[0] == STRQNULL[0])
|
|
cp++;
|
|
|
|
i = (int) Strlen(cp);
|
|
|
|
if (cmdlen >= PMAXLEN)
|
|
return;
|
|
if (cmdlen + i >= PMAXLEN) {
|
|
(void) Strcpy(cmdp, STRsp3dots);
|
|
cmdlen = PMAXLEN;
|
|
cmdp += 4;
|
|
return;
|
|
}
|
|
(void) Strcpy(cmdp, cp);
|
|
cmdp += i;
|
|
cmdlen += i;
|
|
}
|
|
|
|
/*
|
|
* psavejob - temporarily save the current job on a one level stack
|
|
* so another job can be created. Used for { } in exp6
|
|
* and `` in globbing.
|
|
*/
|
|
void
|
|
psavejob()
|
|
{
|
|
pholdjob = pcurrjob;
|
|
pcurrjob = NULL;
|
|
}
|
|
|
|
/*
|
|
* prestjob - opposite of psavejob. This may be missed if we are interrupted
|
|
* somewhere, but pendjob cleans up anyway.
|
|
*/
|
|
void
|
|
prestjob()
|
|
{
|
|
pcurrjob = pholdjob;
|
|
pholdjob = NULL;
|
|
}
|
|
|
|
/*
|
|
* pendjob - indicate that a job (set of commands) has been completed
|
|
* or is about to begin.
|
|
*/
|
|
void
|
|
pendjob()
|
|
{
|
|
register struct process *pp, *tp;
|
|
|
|
if (pcurrjob && (pcurrjob->p_flags & (PFOREGND | PSTOPPED)) == 0) {
|
|
pp = pcurrjob;
|
|
while (pp->p_procid != pp->p_jobid)
|
|
pp = pp->p_friends;
|
|
xprintf("[%d]", pp->p_index);
|
|
tp = pp;
|
|
do {
|
|
xprintf(" %d", pp->p_procid);
|
|
pp = pp->p_friends;
|
|
} while (pp != tp);
|
|
xputchar('\n');
|
|
}
|
|
pholdjob = pcurrjob = 0;
|
|
}
|
|
|
|
/*
|
|
* pprint - print a job
|
|
*/
|
|
|
|
/*
|
|
* Hacks have been added for SVR4 to deal with pipe's being spawned in
|
|
* reverse order
|
|
*
|
|
* David Dawes (dawes@physics.su.oz.au) Oct 1991
|
|
*/
|
|
|
|
static int
|
|
pprint(pp, flag)
|
|
register struct process *pp;
|
|
bool flag;
|
|
{
|
|
int status, reason;
|
|
struct process *tp;
|
|
extern char *linp, linbuf[];
|
|
int jobflags, pstatus, pcond;
|
|
char *format;
|
|
|
|
#ifdef BACKPIPE
|
|
struct process *pipehead = NULL, *pipetail = NULL, *pmarker = NULL;
|
|
int inpipe = 0;
|
|
#endif /* BACKPIPE */
|
|
|
|
while (pp->p_procid != pp->p_jobid)
|
|
pp = pp->p_friends;
|
|
if (pp == pp->p_friends && (pp->p_flags & PPTIME)) {
|
|
pp->p_flags &= ~PPTIME;
|
|
pp->p_flags |= PTIME;
|
|
}
|
|
tp = pp;
|
|
status = reason = -1;
|
|
jobflags = 0;
|
|
do {
|
|
#ifdef BACKPIPE
|
|
/*
|
|
* The pipeline is reversed, so locate the real head of the pipeline
|
|
* if pp is at the tail of a pipe (and not already in a pipeline)
|
|
*/
|
|
if ((pp->p_friends->p_flags & PPOU) && !inpipe && (flag & NAME)) {
|
|
inpipe = 1;
|
|
pipetail = pp;
|
|
do
|
|
pp = pp->p_friends;
|
|
while (pp->p_friends->p_flags & PPOU);
|
|
pipehead = pp;
|
|
pmarker = pp;
|
|
/*
|
|
* pmarker is used to hold the place of the proc being processed, so
|
|
* we can search for the next one downstream later.
|
|
*/
|
|
}
|
|
pcond = (int) (tp != pp || (inpipe && tp == pp));
|
|
#else /* !BACKPIPE */
|
|
pcond = (int) (tp != pp);
|
|
#endif /* BACKPIPE */
|
|
|
|
jobflags |= pp->p_flags;
|
|
pstatus = (int) (pp->p_flags & PALLSTATES);
|
|
if (pcond && linp != linbuf && !(flag & FANCY) &&
|
|
((pstatus == status && pp->p_reason == reason) ||
|
|
!(flag & REASON)))
|
|
xputchar(' ');
|
|
else {
|
|
if (pcond && linp != linbuf)
|
|
xputchar('\n');
|
|
if (flag & NUMBER) {
|
|
#ifdef BACKPIPE
|
|
pcond = ((pp == tp && !inpipe) ||
|
|
(inpipe && pipetail == tp && pp == pipehead));
|
|
#else /* BACKPIPE */
|
|
pcond = (pp == tp);
|
|
#endif /* BACKPIPE */
|
|
if (pcond)
|
|
xprintf("[%d]%s %c ", pp->p_index,
|
|
pp->p_index < 10 ? " " : "",
|
|
pp == pcurrent ? '+' :
|
|
(pp == pprevious ? '-' : ' '));
|
|
else
|
|
xprintf(" ");
|
|
}
|
|
if (flag & FANCY) {
|
|
#ifdef TCF
|
|
extern char *sitename();
|
|
|
|
#endif /* TCF */
|
|
xprintf("%5d ", pp->p_procid);
|
|
#ifdef TCF
|
|
xprintf("%11s ", sitename(pp->p_procid));
|
|
#endif /* TCF */
|
|
}
|
|
if (flag & (REASON | AREASON)) {
|
|
if (flag & NAME)
|
|
format = "%-30s";
|
|
else
|
|
format = "%s";
|
|
if (pstatus == status) {
|
|
if (pp->p_reason == reason) {
|
|
xprintf(format, "");
|
|
goto prcomd;
|
|
}
|
|
else
|
|
reason = (int) pp->p_reason;
|
|
}
|
|
else {
|
|
status = pstatus;
|
|
reason = (int) pp->p_reason;
|
|
}
|
|
switch (status) {
|
|
|
|
case PRUNNING:
|
|
xprintf(format, CGETS(17, 4, "Running "));
|
|
break;
|
|
|
|
case PINTERRUPTED:
|
|
case PSTOPPED:
|
|
case PSIGNALED:
|
|
/*
|
|
* tell what happened to the background job
|
|
* From: Michael Schroeder
|
|
* <mlschroe@immd4.informatik.uni-erlangen.de>
|
|
*/
|
|
if ((flag & REASON)
|
|
|| ((flag & AREASON)
|
|
&& reason != SIGINT
|
|
&& (reason != SIGPIPE
|
|
|| (pp->p_flags & PPOU) == 0))) {
|
|
char *ptr;
|
|
char buf[1024];
|
|
|
|
if ((ptr = mesg[pp->p_reason & ASCII].pname) == NULL)
|
|
xsnprintf(ptr = buf, sizeof(buf), "%s %d",
|
|
CGETS(17, 5, "Signal"), pp->p_reason & ASCII);
|
|
xprintf(format, ptr);
|
|
}
|
|
else
|
|
reason = -1;
|
|
break;
|
|
|
|
case PNEXITED:
|
|
case PAEXITED:
|
|
if (flag & REASON) {
|
|
if (pp->p_reason)
|
|
xprintf(CGETS(17, 6, "Exit %-25d"), pp->p_reason);
|
|
else
|
|
xprintf(format, CGETS(17, 7, "Done"));
|
|
}
|
|
break;
|
|
|
|
default:
|
|
xprintf(CGETS(17, 8, "BUG: status=%-9o"),
|
|
status);
|
|
}
|
|
}
|
|
}
|
|
prcomd:
|
|
if (flag & NAME) {
|
|
xprintf("%S", pp->p_command);
|
|
if (pp->p_flags & PPOU)
|
|
xprintf(" |");
|
|
if (pp->p_flags & PDIAG)
|
|
xprintf("&");
|
|
}
|
|
if (flag & (REASON | AREASON) && pp->p_flags & PDUMPED)
|
|
xprintf(CGETS(17, 9, " (core dumped)"));
|
|
if (tp == pp->p_friends) {
|
|
if (flag & AMPERSAND)
|
|
xprintf(" &");
|
|
if (flag & JOBDIR &&
|
|
!eq(tp->p_cwd->di_name, dcwd->di_name)) {
|
|
xprintf(CGETS(17, 10, " (wd: "));
|
|
dtildepr(tp->p_cwd->di_name);
|
|
xprintf(")");
|
|
}
|
|
}
|
|
if (pp->p_flags & PPTIME && !(status & (PSTOPPED | PRUNNING))) {
|
|
if (linp != linbuf)
|
|
xprintf("\n\t");
|
|
#if defined(BSDTIMES) || defined(_SEQUENT_)
|
|
prusage(&zru, &pp->p_rusage, &pp->p_etime,
|
|
&pp->p_btime);
|
|
#else /* !BSDTIMES && !SEQUENT */
|
|
lru.tms_utime = pp->p_utime;
|
|
lru.tms_stime = pp->p_stime;
|
|
lru.tms_cutime = 0;
|
|
lru.tms_cstime = 0;
|
|
prusage(&zru, &lru, pp->p_etime,
|
|
pp->p_btime);
|
|
#endif /* !BSDTIMES && !SEQUENT */
|
|
|
|
}
|
|
#ifdef BACKPIPE
|
|
pcond = ((tp == pp->p_friends && !inpipe) ||
|
|
(inpipe && pipehead->p_friends == tp && pp == pipetail));
|
|
#else /* !BACKPIPE */
|
|
pcond = (tp == pp->p_friends);
|
|
#endif /* BACKPIPE */
|
|
if (pcond) {
|
|
if (linp != linbuf)
|
|
xputchar('\n');
|
|
if (flag & SHELLDIR && !eq(tp->p_cwd->di_name, dcwd->di_name)) {
|
|
xprintf(CGETS(17, 11, "(wd now: "));
|
|
dtildepr(dcwd->di_name);
|
|
xprintf(")\n");
|
|
}
|
|
}
|
|
#ifdef BACKPIPE
|
|
if (inpipe) {
|
|
/*
|
|
* if pmaker == pipetail, we are finished that pipeline, and
|
|
* can now skip to past the head
|
|
*/
|
|
if (pmarker == pipetail) {
|
|
inpipe = 0;
|
|
pp = pipehead;
|
|
}
|
|
else {
|
|
/*
|
|
* set pp to one before the one we want next, so the while below
|
|
* increments to the correct spot.
|
|
*/
|
|
do
|
|
pp = pp->p_friends;
|
|
while (pp->p_friends->p_friends != pmarker);
|
|
pmarker = pp->p_friends;
|
|
}
|
|
}
|
|
pcond = ((pp = pp->p_friends) != tp || inpipe);
|
|
#else /* !BACKPIPE */
|
|
pcond = ((pp = pp->p_friends) != tp);
|
|
#endif /* BACKPIPE */
|
|
} while (pcond);
|
|
|
|
if (jobflags & PTIME && (jobflags & (PSTOPPED | PRUNNING)) == 0) {
|
|
if (jobflags & NUMBER)
|
|
xprintf(" ");
|
|
ptprint(tp);
|
|
}
|
|
return (jobflags);
|
|
}
|
|
|
|
/*
|
|
* All 4.3 BSD derived implementations are buggy and I've had enough.
|
|
* The following implementation produces similar code and works in all
|
|
* cases. The 4.3BSD one works only for <, >, !=
|
|
*/
|
|
# undef timercmp
|
|
# define timercmp(tvp, uvp, cmp) \
|
|
(((tvp)->tv_sec == (uvp)->tv_sec) ? \
|
|
((tvp)->tv_usec cmp (uvp)->tv_usec) : \
|
|
((tvp)->tv_sec cmp (uvp)->tv_sec))
|
|
|
|
static void
|
|
ptprint(tp)
|
|
register struct process *tp;
|
|
{
|
|
#ifdef BSDTIMES
|
|
struct timeval tetime, diff;
|
|
static struct timeval ztime;
|
|
struct sysrusage ru;
|
|
register struct process *pp = tp;
|
|
|
|
ru = zru;
|
|
tetime = ztime;
|
|
do {
|
|
ruadd(&ru, &pp->p_rusage);
|
|
tvsub(&diff, &pp->p_etime, &pp->p_btime);
|
|
if (timercmp(&diff, &tetime, >))
|
|
tetime = diff;
|
|
} while ((pp = pp->p_friends) != tp);
|
|
prusage(&zru, &ru, &tetime, &ztime);
|
|
#else /* !BSDTIMES */
|
|
# ifdef _SEQUENT_
|
|
timeval_t tetime, diff;
|
|
static timeval_t ztime;
|
|
struct process_stats ru;
|
|
register struct process *pp = tp;
|
|
|
|
ru = zru;
|
|
tetime = ztime;
|
|
do {
|
|
ruadd(&ru, &pp->p_rusage);
|
|
tvsub(&diff, &pp->p_etime, &pp->p_btime);
|
|
if (timercmp(&diff, &tetime, >))
|
|
tetime = diff;
|
|
} while ((pp = pp->p_friends) != tp);
|
|
prusage(&zru, &ru, &tetime, &ztime);
|
|
# else /* !_SEQUENT_ */
|
|
# ifndef POSIX
|
|
static time_t ztime = 0;
|
|
static time_t zu_time = 0;
|
|
static time_t zs_time = 0;
|
|
time_t tetime, diff;
|
|
time_t u_time, s_time;
|
|
|
|
# else /* POSIX */
|
|
static clock_t ztime = 0;
|
|
static clock_t zu_time = 0;
|
|
static clock_t zs_time = 0;
|
|
clock_t tetime, diff;
|
|
clock_t u_time, s_time;
|
|
|
|
# endif /* POSIX */
|
|
struct tms zts, rts;
|
|
register struct process *pp = tp;
|
|
|
|
u_time = zu_time;
|
|
s_time = zs_time;
|
|
tetime = ztime;
|
|
do {
|
|
u_time += pp->p_utime;
|
|
s_time += pp->p_stime;
|
|
diff = pp->p_etime - pp->p_btime;
|
|
if (diff > tetime)
|
|
tetime = diff;
|
|
} while ((pp = pp->p_friends) != tp);
|
|
zts.tms_utime = zu_time;
|
|
zts.tms_stime = zs_time;
|
|
zts.tms_cutime = 0;
|
|
zts.tms_cstime = 0;
|
|
rts.tms_utime = u_time;
|
|
rts.tms_stime = s_time;
|
|
rts.tms_cutime = 0;
|
|
rts.tms_cstime = 0;
|
|
prusage(&zts, &rts, tetime, ztime);
|
|
# endif /* !_SEQUENT_ */
|
|
#endif /* !BSDTIMES */
|
|
}
|
|
|
|
/*
|
|
* dojobs - print all jobs
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dojobs(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
register int flag = NUMBER | NAME | REASON;
|
|
int i;
|
|
|
|
USE(c);
|
|
if (chkstop)
|
|
chkstop = 2;
|
|
if (*++v) {
|
|
if (v[1] || !eq(*v, STRml))
|
|
stderror(ERR_JOBS);
|
|
flag |= FANCY | JOBDIR;
|
|
}
|
|
for (i = 1; i <= pmaxindex; i++)
|
|
for (pp = proclist.p_next; pp; pp = pp->p_next)
|
|
if (pp->p_index == i && pp->p_procid == pp->p_jobid) {
|
|
pp->p_flags &= ~PNEEDNOTE;
|
|
if (!(pprint(pp, flag) & (PRUNNING | PSTOPPED)))
|
|
pflush(pp);
|
|
break;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* dofg - builtin - put the job into the foreground
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dofg(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
|
|
USE(c);
|
|
okpcntl();
|
|
++v;
|
|
do {
|
|
pp = pfind(*v);
|
|
if (!pstart(pp, 1)) {
|
|
pp->p_procid = 0;
|
|
stderror(ERR_NAME|ERR_BADJOB, pp->p_command, strerror(errno));
|
|
continue;
|
|
}
|
|
#ifndef BSDSIGS
|
|
# ifdef notdef
|
|
if (setintr)
|
|
sigignore(SIGINT);
|
|
# endif
|
|
#endif /* !BSDSIGS */
|
|
pjwait(pp);
|
|
} while (*v && *++v);
|
|
}
|
|
|
|
/*
|
|
* %... - builtin - put the job into the foreground
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dofg1(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
|
|
USE(c);
|
|
okpcntl();
|
|
pp = pfind(v[0]);
|
|
if (!pstart(pp, 1)) {
|
|
pp->p_procid = 0;
|
|
stderror(ERR_NAME|ERR_BADJOB, pp->p_command, strerror(errno));
|
|
return;
|
|
}
|
|
#ifndef BSDSIGS
|
|
# ifdef notdef
|
|
if (setintr)
|
|
sigignore(SIGINT);
|
|
# endif
|
|
#endif /* !BSDSIGS */
|
|
pjwait(pp);
|
|
}
|
|
|
|
/*
|
|
* dobg - builtin - put the job into the background
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dobg(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
|
|
USE(c);
|
|
okpcntl();
|
|
++v;
|
|
do {
|
|
pp = pfind(*v);
|
|
if (!pstart(pp, 0)) {
|
|
pp->p_procid = 0;
|
|
stderror(ERR_NAME|ERR_BADJOB, pp->p_command, strerror(errno));
|
|
}
|
|
} while (*v && *++v);
|
|
}
|
|
|
|
/*
|
|
* %... & - builtin - put the job into the background
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dobg1(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
|
|
USE(c);
|
|
pp = pfind(v[0]);
|
|
if (!pstart(pp, 0)) {
|
|
pp->p_procid = 0;
|
|
stderror(ERR_NAME|ERR_BADJOB, pp->p_command, strerror(errno));
|
|
}
|
|
}
|
|
|
|
/*
|
|
* dostop - builtin - stop the job
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dostop(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
USE(c);
|
|
#ifdef BSDJOBS
|
|
pkill(++v, SIGSTOP);
|
|
#endif /* BSDJOBS */
|
|
}
|
|
|
|
/*
|
|
* dokill - builtin - superset of kill (1)
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
dokill(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register int signum, len = 0;
|
|
register char *name;
|
|
Char *sigptr;
|
|
extern int T_Cols;
|
|
extern int nsig;
|
|
|
|
USE(c);
|
|
v++;
|
|
if (v[0] && v[0][0] == '-') {
|
|
if (v[0][1] == 'l') {
|
|
for (signum = 0; signum <= nsig; signum++) {
|
|
if ((name = mesg[signum].iname) != NULL) {
|
|
len += strlen(name) + 1;
|
|
if (len >= T_Cols - 1) {
|
|
xputchar('\n');
|
|
len = strlen(name) + 1;
|
|
}
|
|
xprintf("%s ", name);
|
|
}
|
|
}
|
|
xputchar('\n');
|
|
return;
|
|
}
|
|
sigptr = &v[0][1];
|
|
if (v[0][1] == 's') {
|
|
if (v[1]) {
|
|
v++;
|
|
sigptr = &v[0][0];
|
|
} else {
|
|
stderror(ERR_NAME | ERR_TOOFEW);
|
|
}
|
|
}
|
|
if (Isdigit(*sigptr)) {
|
|
char *ep;
|
|
signum = strtoul(short2str(sigptr), &ep, 0);
|
|
if (*ep || signum < 0 || signum > (MAXSIG-1))
|
|
stderror(ERR_NAME | ERR_BADSIG);
|
|
}
|
|
else {
|
|
for (signum = 0; signum <= nsig; signum++)
|
|
if (mesg[signum].iname &&
|
|
eq(sigptr, str2short(mesg[signum].iname)))
|
|
goto gotsig;
|
|
setname(short2str(sigptr));
|
|
stderror(ERR_NAME | ERR_UNKSIG);
|
|
}
|
|
gotsig:
|
|
v++;
|
|
}
|
|
else
|
|
signum = SIGTERM;
|
|
pkill(v, signum);
|
|
}
|
|
|
|
static void
|
|
pkill(v, signum)
|
|
Char **v;
|
|
int signum;
|
|
{
|
|
register struct process *pp, *np;
|
|
int jobflags = 0, err1 = 0;
|
|
pid_t pid;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask;
|
|
#endif /* BSDSIGS */
|
|
Char *cp, **vp;
|
|
|
|
#ifdef BSDSIGS
|
|
omask = sigmask(SIGCHLD);
|
|
if (setintr)
|
|
omask |= sigmask(SIGINT);
|
|
omask = sigblock(omask) & ~omask;
|
|
#else /* !BSDSIGS */
|
|
if (setintr)
|
|
(void) sighold(SIGINT);
|
|
(void) sighold(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
|
|
/* Avoid globbing %?x patterns */
|
|
for (vp = v; vp && *vp; vp++)
|
|
if (**vp == '%')
|
|
(void) quote(*vp);
|
|
|
|
gflag = 0, tglob(v);
|
|
if (gflag) {
|
|
v = globall(v);
|
|
if (v == 0)
|
|
stderror(ERR_NAME | ERR_NOMATCH);
|
|
}
|
|
else {
|
|
v = gargv = saveblk(v);
|
|
trim(v);
|
|
}
|
|
|
|
|
|
while (v && (cp = *v)) {
|
|
if (*cp == '%') {
|
|
np = pp = pfind(cp);
|
|
do
|
|
jobflags |= np->p_flags;
|
|
while ((np = np->p_friends) != pp);
|
|
#ifdef BSDJOBS
|
|
switch (signum) {
|
|
|
|
case SIGSTOP:
|
|
case SIGTSTP:
|
|
case SIGTTIN:
|
|
case SIGTTOU:
|
|
if ((jobflags & PRUNNING) == 0) {
|
|
# ifdef SUSPENDED
|
|
xprintf(CGETS(17, 12, "%S: Already suspended\n"), cp);
|
|
# else /* !SUSPENDED */
|
|
xprintf(CGETS(17, 13, "%S: Already stopped\n"), cp);
|
|
# endif /* !SUSPENDED */
|
|
err1++;
|
|
goto cont;
|
|
}
|
|
break;
|
|
/*
|
|
* suspend a process, kill -CONT %, then type jobs; the shell
|
|
* says it is suspended, but it is running; thanks jaap..
|
|
*/
|
|
case SIGCONT:
|
|
if (!pstart(pp, 0)) {
|
|
pp->p_procid = 0;
|
|
stderror(ERR_NAME|ERR_BADJOB, pp->p_command,
|
|
strerror(errno));
|
|
}
|
|
goto cont;
|
|
default:
|
|
break;
|
|
}
|
|
#endif /* BSDJOBS */
|
|
if (killpg(pp->p_jobid, signum) < 0) {
|
|
xprintf("%S: %s\n", cp, strerror(errno));
|
|
err1++;
|
|
}
|
|
#ifdef BSDJOBS
|
|
if (signum == SIGTERM || signum == SIGHUP)
|
|
(void) killpg(pp->p_jobid, SIGCONT);
|
|
#endif /* BSDJOBS */
|
|
}
|
|
else if (!(Isdigit(*cp) || *cp == '-'))
|
|
stderror(ERR_NAME | ERR_JOBARGS);
|
|
else {
|
|
char *ep;
|
|
#ifndef WINNT_NATIVE
|
|
pid = strtol(short2str(cp), &ep, 10);
|
|
#else
|
|
pid = strtoul(short2str(cp), &ep, 0);
|
|
#endif /* WINNT_NATIVE */
|
|
if (*ep)
|
|
stderror(ERR_NAME | ERR_JOBARGS);
|
|
else if (kill(pid, signum) < 0) {
|
|
xprintf("%d: %s\n", pid, strerror(errno));
|
|
err1++;
|
|
goto cont;
|
|
}
|
|
#ifdef BSDJOBS
|
|
if (signum == SIGTERM || signum == SIGHUP)
|
|
(void) kill(pid, SIGCONT);
|
|
#endif /* BSDJOBS */
|
|
}
|
|
cont:
|
|
v++;
|
|
}
|
|
if (gargv)
|
|
blkfree(gargv), gargv = 0;
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
if (setintr)
|
|
(void) sigrelse(SIGINT);
|
|
#endif /* !BSDSIGS */
|
|
if (err1)
|
|
stderror(ERR_SILENT);
|
|
}
|
|
|
|
/*
|
|
* pstart - start the job in foreground/background
|
|
*/
|
|
int
|
|
pstart(pp, foregnd)
|
|
register struct process *pp;
|
|
int foregnd;
|
|
{
|
|
int rv = 0;
|
|
register struct process *np;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask;
|
|
#endif /* BSDSIGS */
|
|
/* We don't use jobflags in this function right now (see below) */
|
|
/* long jobflags = 0; */
|
|
|
|
#ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGCHLD));
|
|
#else /* !BSDSIGS */
|
|
(void) sighold(SIGCHLD);
|
|
#endif
|
|
np = pp;
|
|
do {
|
|
/* We don't use jobflags in this function right now (see below) */
|
|
/* jobflags |= np->p_flags; */
|
|
if (np->p_flags & (PRUNNING | PSTOPPED)) {
|
|
np->p_flags |= PRUNNING;
|
|
np->p_flags &= ~PSTOPPED;
|
|
if (foregnd)
|
|
np->p_flags |= PFOREGND;
|
|
else
|
|
np->p_flags &= ~PFOREGND;
|
|
}
|
|
} while ((np = np->p_friends) != pp);
|
|
if (!foregnd)
|
|
pclrcurr(pp);
|
|
(void) pprint(pp, foregnd ? NAME | JOBDIR : NUMBER | NAME | AMPERSAND);
|
|
|
|
/* GrP run jobcmd hook if foregrounding */
|
|
if (foregnd) {
|
|
job_cmd(pp->p_command);
|
|
}
|
|
|
|
#ifdef BSDJOBS
|
|
if (foregnd) {
|
|
rv = tcsetpgrp(FSHTTY, pp->p_jobid);
|
|
}
|
|
/*
|
|
* 1. child process of csh (shell script) receives SIGTTIN/SIGTTOU
|
|
* 2. parent process (csh) receives SIGCHLD
|
|
* 3. The "csh" signal handling function pchild() is invoked
|
|
* with a SIGCHLD signal.
|
|
* 4. pchild() calls wait3(WNOHANG) which returns 0.
|
|
* The child process is NOT ready to be waited for at this time.
|
|
* pchild() returns without picking-up the correct status
|
|
* for the child process which generated the SIGCHILD.
|
|
* 5. CONSEQUENCE : csh is UNaware that the process is stopped
|
|
* 6. THIS LINE HAS BEEN COMMENTED OUT : if (jobflags&PSTOPPED)
|
|
* (beto@aixwiz.austin.ibm.com - aug/03/91)
|
|
* 7. I removed the line completely and added extra checks for
|
|
* pstart, so that if a job gets attached to and dies inside
|
|
* a debugger it does not confuse the shell. [christos]
|
|
* 8. on the nec sx-4 there seems to be a problem, which requires
|
|
* a syscall(151, getpid(), getpid()) in osinit. Don't ask me
|
|
* what this is doing. [schott@rzg.mpg.de]
|
|
*/
|
|
|
|
if (rv != -1)
|
|
rv = killpg(pp->p_jobid, SIGCONT);
|
|
#endif /* BSDJOBS */
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
return rv != -1;
|
|
}
|
|
|
|
void
|
|
panystop(neednl)
|
|
bool neednl;
|
|
{
|
|
register struct process *pp;
|
|
|
|
chkstop = 2;
|
|
for (pp = proclist.p_next; pp; pp = pp->p_next)
|
|
if (pp->p_flags & PSTOPPED)
|
|
stderror(ERR_STOPPED, neednl ? "\n" : "");
|
|
}
|
|
|
|
struct process *
|
|
pfind(cp)
|
|
Char *cp;
|
|
{
|
|
register struct process *pp, *np;
|
|
|
|
if (cp == 0 || cp[1] == 0 || eq(cp, STRcent2) || eq(cp, STRcentplus)) {
|
|
if (pcurrent == NULL)
|
|
stderror(ERR_NAME | ERR_JOBCUR);
|
|
return (pcurrent);
|
|
}
|
|
if (eq(cp, STRcentminus) || eq(cp, STRcenthash)) {
|
|
if (pprevious == NULL)
|
|
stderror(ERR_NAME | ERR_JOBPREV);
|
|
return (pprevious);
|
|
}
|
|
if (Isdigit(cp[1])) {
|
|
int idx = atoi(short2str(cp + 1));
|
|
|
|
for (pp = proclist.p_next; pp; pp = pp->p_next)
|
|
if (pp->p_index == idx && pp->p_procid == pp->p_jobid)
|
|
return (pp);
|
|
stderror(ERR_NAME | ERR_NOSUCHJOB);
|
|
}
|
|
np = NULL;
|
|
for (pp = proclist.p_next; pp; pp = pp->p_next)
|
|
if (pp->p_procid == pp->p_jobid) {
|
|
if (cp[1] == '?') {
|
|
register Char *dp;
|
|
|
|
for (dp = pp->p_command; *dp; dp++) {
|
|
if (*dp != cp[2])
|
|
continue;
|
|
if (prefix(cp + 2, dp))
|
|
goto match;
|
|
}
|
|
}
|
|
else if (prefix(cp + 1, pp->p_command)) {
|
|
match:
|
|
if (np)
|
|
stderror(ERR_NAME | ERR_AMBIG);
|
|
np = pp;
|
|
}
|
|
}
|
|
if (np)
|
|
return (np);
|
|
stderror(ERR_NAME | (cp[1] == '?' ? ERR_JOBPAT : ERR_NOSUCHJOB));
|
|
/* NOTREACHED */
|
|
return (0);
|
|
}
|
|
|
|
|
|
/*
|
|
* pgetcurr - find most recent job that is not pp, preferably stopped
|
|
*/
|
|
static struct process *
|
|
pgetcurr(pp)
|
|
register struct process *pp;
|
|
{
|
|
register struct process *np;
|
|
register struct process *xp = NULL;
|
|
|
|
for (np = proclist.p_next; np; np = np->p_next)
|
|
if (np != pcurrent && np != pp && np->p_procid &&
|
|
np->p_procid == np->p_jobid) {
|
|
if (np->p_flags & PSTOPPED)
|
|
return (np);
|
|
if (xp == NULL)
|
|
xp = np;
|
|
}
|
|
return (xp);
|
|
}
|
|
|
|
/*
|
|
* donotify - flag the job so as to report termination asynchronously
|
|
*/
|
|
/*ARGSUSED*/
|
|
void
|
|
donotify(v, c)
|
|
Char **v;
|
|
struct command *c;
|
|
{
|
|
register struct process *pp;
|
|
|
|
USE(c);
|
|
pp = pfind(*++v);
|
|
pp->p_flags |= PNOTIFY;
|
|
}
|
|
|
|
/*
|
|
* Do the fork and whatever should be done in the child side that
|
|
* should not be done if we are not forking at all (like for simple builtin's)
|
|
* Also do everything that needs any signals fiddled with in the parent side
|
|
*
|
|
* Wanttty tells whether process and/or tty pgrps are to be manipulated:
|
|
* -1: leave tty alone; inherit pgrp from parent
|
|
* 0: already have tty; manipulate process pgrps only
|
|
* 1: want to claim tty; manipulate process and tty pgrps
|
|
* It is usually just the value of tpgrp.
|
|
*/
|
|
|
|
int
|
|
pfork(t, wanttty)
|
|
struct command *t; /* command we are forking for */
|
|
int wanttty;
|
|
{
|
|
register int pid;
|
|
bool ignint = 0;
|
|
int pgrp;
|
|
#ifdef BSDSIGS
|
|
sigmask_t omask = 0;
|
|
#endif /* BSDSIGS */
|
|
#ifdef SIGSYNCH
|
|
sigvec_t osv;
|
|
static sigvec_t nsv = {synch_handler, (sigset_t) ~0, 0};
|
|
#endif /* SIGSYNCH */
|
|
|
|
/*
|
|
* A child will be uninterruptible only under very special conditions.
|
|
* Remember that the semantics of '&' is implemented by disconnecting the
|
|
* process from the tty so signals do not need to ignored just for '&'.
|
|
* Thus signals are set to default action for children unless: we have had
|
|
* an "onintr -" (then specifically ignored) we are not playing with
|
|
* signals (inherit action)
|
|
*/
|
|
if (setintr)
|
|
ignint = (tpgrp == -1 && (t->t_dflg & F_NOINTERRUPT))
|
|
|| (gointr && eq(gointr, STRminus));
|
|
|
|
#ifdef COHERENT
|
|
ignint |= gointr && eq(gointr, STRminus);
|
|
#endif /* COHERENT */
|
|
|
|
/*
|
|
* Check for maximum nesting of 16 processes to avoid Forking loops
|
|
*/
|
|
if (child == 16)
|
|
stderror(ERR_NESTING, 16);
|
|
#ifdef SIGSYNCH
|
|
if (mysigvec(SIGSYNCH, &nsv, &osv))
|
|
stderror(ERR_SYSTEM, "pfork: sigvec set", strerror(errno));
|
|
#endif /* SIGSYNCH */
|
|
/*
|
|
* Hold SIGCHLD until we have the process installed in our table.
|
|
*/
|
|
if (wanttty < 0) {
|
|
#ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGCHLD));
|
|
#else /* !BSDSIGS */
|
|
(void) sighold(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
}
|
|
while ((pid = fork()) == -1)
|
|
if (setintr == 0)
|
|
(void) sleep(FORKSLEEP);
|
|
else {
|
|
if (wanttty < 0)
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
(void) sigrelse(SIGINT);
|
|
#endif /* !BSDSIGS */
|
|
stderror(ERR_NOPROC);
|
|
}
|
|
if (pid == 0) {
|
|
settimes();
|
|
pgrp = pcurrjob ? pcurrjob->p_jobid : getpid();
|
|
pflushall();
|
|
pcurrjob = NULL;
|
|
#if !defined(BSDTIMES) && !defined(_SEQUENT_)
|
|
timesdone = 0;
|
|
#endif /* !defined(BSDTIMES) && !defined(_SEQUENT_) */
|
|
child++;
|
|
if (setintr) {
|
|
setintr = 0; /* until I think otherwise */
|
|
#ifndef BSDSIGS
|
|
if (wanttty < 0)
|
|
(void) sigrelse(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
/*
|
|
* Children just get blown away on SIGINT, SIGQUIT unless "onintr
|
|
* -" seen.
|
|
*/
|
|
(void) signal(SIGINT, ignint ? SIG_IGN : SIG_DFL);
|
|
(void) signal(SIGQUIT, ignint ? SIG_IGN : SIG_DFL);
|
|
#ifdef BSDJOBS
|
|
if (wanttty >= 0) {
|
|
/* make stoppable */
|
|
(void) signal(SIGTSTP, SIG_DFL);
|
|
(void) signal(SIGTTIN, SIG_DFL);
|
|
(void) signal(SIGTTOU, SIG_DFL);
|
|
}
|
|
#endif /* BSDJOBS */
|
|
(void) signal(SIGTERM, parterm);
|
|
}
|
|
else if (tpgrp == -1 && (t->t_dflg & F_NOINTERRUPT)) {
|
|
(void) signal(SIGINT, SIG_IGN);
|
|
(void) signal(SIGQUIT, SIG_IGN);
|
|
}
|
|
#ifdef OREO
|
|
sigignore(SIGIO); /* ignore SIGIO in child too */
|
|
#endif /* OREO */
|
|
|
|
pgetty(wanttty, pgrp);
|
|
/*
|
|
* Nohup and nice apply only to NODE_COMMAND's but it would be nice
|
|
* (?!?) if you could say "nohup (foo;bar)" Then the parser would have
|
|
* to know about nice/nohup/time
|
|
*/
|
|
if (t->t_dflg & F_NOHUP)
|
|
(void) signal(SIGHUP, SIG_IGN);
|
|
if (t->t_dflg & F_NICE) {
|
|
int nval = SIGN_EXTEND_CHAR(t->t_nice);
|
|
#ifdef BSDNICE
|
|
if (setpriority(PRIO_PROCESS, 0, nval) == -1 && errno)
|
|
stderror(ERR_SYSTEM, "setpriority", strerror(errno));
|
|
#else /* !BSDNICE */
|
|
(void) nice(nval);
|
|
#endif /* !BSDNICE */
|
|
}
|
|
#ifdef F_VER
|
|
if (t->t_dflg & F_VER) {
|
|
tsetenv(STRSYSTYPE, t->t_systype ? STRbsd43 : STRsys53);
|
|
dohash(NULL, NULL);
|
|
}
|
|
#endif /* F_VER */
|
|
#ifdef SIGSYNCH
|
|
/* rfw 8/89 now parent can continue */
|
|
if (kill(getppid(), SIGSYNCH))
|
|
stderror(ERR_SYSTEM, "pfork child: kill", strerror(errno));
|
|
#endif /* SIGSYNCH */
|
|
|
|
}
|
|
else {
|
|
#ifdef POSIXJOBS
|
|
if (wanttty >= 0) {
|
|
/*
|
|
* `Walking' process group fix from Beto Appleton.
|
|
* (beto@aixwiz.austin.ibm.com)
|
|
* If setpgid fails at this point that means that
|
|
* our process leader has died. We flush the current
|
|
* job and become the process leader ourselves.
|
|
* The parent will figure that out later.
|
|
*/
|
|
pgrp = pcurrjob ? pcurrjob->p_jobid : pid;
|
|
if (setpgid(pid, pgrp) == -1 && errno == EPERM) {
|
|
pcurrjob = NULL;
|
|
/*
|
|
* We don't care if this causes an error here;
|
|
* then we are already in the right process group
|
|
*/
|
|
(void) setpgid(pid, pgrp = pid);
|
|
}
|
|
}
|
|
#endif /* POSIXJOBS */
|
|
palloc(pid, t);
|
|
#ifdef SIGSYNCH
|
|
/*
|
|
* rfw 8/89 Wait for child to own terminal. Solves half of ugly
|
|
* synchronization problem. With this change, we know that the only
|
|
* reason setpgrp to a previous process in a pipeline can fail is that
|
|
* the previous process has already exited. Without this hack, he may
|
|
* either have exited or not yet started to run. Two uglies become
|
|
* one.
|
|
*/
|
|
(void) sigpause(omask & ~SYNCHMASK);
|
|
if (mysigvec(SIGSYNCH, &osv, NULL))
|
|
stderror(ERR_SYSTEM, "pfork parent: sigvec restore",
|
|
strerror(errno));
|
|
#endif /* SIGSYNCH */
|
|
|
|
if (wanttty < 0) {
|
|
#ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
#else /* !BSDSIGS */
|
|
(void) sigrelse(SIGCHLD);
|
|
#endif /* !BSDSIGS */
|
|
}
|
|
}
|
|
return (pid);
|
|
}
|
|
|
|
static void
|
|
okpcntl()
|
|
{
|
|
if (tpgrp == -1)
|
|
stderror(ERR_JOBCONTROL);
|
|
if (tpgrp == 0)
|
|
stderror(ERR_JOBCTRLSUB);
|
|
}
|
|
|
|
|
|
static void
|
|
setttypgrp(pgrp)
|
|
int pgrp;
|
|
{
|
|
/*
|
|
* If we are piping out a builtin, eg. 'echo | more' things can go
|
|
* out of sequence, i.e. the more can run before the echo. This
|
|
* can happen even if we have vfork, since the echo will be forked
|
|
* with the regular fork. In this case, we need to set the tty
|
|
* pgrp ourselves. If that happens, then the process will be still
|
|
* alive. And the tty process group will already be set.
|
|
* This should fix the famous sequent problem as a side effect:
|
|
* The controlling terminal is lost if all processes in the
|
|
* terminal process group are zombies. In this case tcgetpgrp()
|
|
* returns 0. If this happens we must set the terminal process
|
|
* group again.
|
|
*/
|
|
if (tcgetpgrp(FSHTTY) != pgrp) {
|
|
#ifdef POSIXJOBS
|
|
/*
|
|
* tcsetpgrp will set SIGTTOU to all the the processes in
|
|
* the background according to POSIX... We ignore this here.
|
|
*/
|
|
signalfun_t old = sigset(SIGTTOU, SIG_IGN);
|
|
#endif
|
|
(void) tcsetpgrp(FSHTTY, pgrp);
|
|
# ifdef POSIXJOBS
|
|
(void) sigset(SIGTTOU, old);
|
|
# endif
|
|
|
|
}
|
|
}
|
|
|
|
|
|
/*
|
|
* if we don't have vfork(), things can still go in the wrong order
|
|
* resulting in the famous 'Stopped (tty output)'. But some systems
|
|
* don't permit the setpgid() call, (these are more recent secure
|
|
* systems such as ibm's aix), when they do. Then we'd rather print
|
|
* an error message than hang the shell!
|
|
* I am open to suggestions how to fix that.
|
|
*/
|
|
void
|
|
pgetty(wanttty, pgrp)
|
|
int wanttty, pgrp;
|
|
{
|
|
#ifdef BSDJOBS
|
|
# if defined(BSDSIGS) && defined(POSIXJOBS)
|
|
sigmask_t omask = 0;
|
|
# endif /* BSDSIGS && POSIXJOBS */
|
|
|
|
# ifdef JOBDEBUG
|
|
xprintf("wanttty %d pid %d opgrp%d pgrp %d tpgrp %d\n",
|
|
wanttty, getpid(), pgrp, mygetpgrp(), tcgetpgrp(FSHTTY));
|
|
# endif /* JOBDEBUG */
|
|
# ifdef POSIXJOBS
|
|
/*
|
|
* christos: I am blocking the tty signals till I've set things
|
|
* correctly....
|
|
*/
|
|
if (wanttty > 0)
|
|
# ifdef BSDSIGS
|
|
omask = sigblock(sigmask(SIGTSTP)|sigmask(SIGTTIN));
|
|
# else /* !BSDSIGS */
|
|
{
|
|
(void) sighold(SIGTSTP);
|
|
(void) sighold(SIGTTIN);
|
|
}
|
|
# endif /* !BSDSIGS */
|
|
# endif /* POSIXJOBS */
|
|
|
|
# ifndef POSIXJOBS
|
|
if (wanttty > 0)
|
|
setttypgrp(pgrp);
|
|
# endif /* !POSIXJOBS */
|
|
|
|
/*
|
|
* From: Michael Schroeder <mlschroe@immd4.informatik.uni-erlangen.de>
|
|
* Don't check for tpgrp >= 0 so even non-interactive shells give
|
|
* background jobs process groups Same for the comparison in the other part
|
|
* of the #ifdef
|
|
*/
|
|
if (wanttty >= 0) {
|
|
if (setpgid(0, pgrp) == -1) {
|
|
# ifdef POSIXJOBS
|
|
/* Walking process group fix; see above */
|
|
if (setpgid(0, pgrp = getpid()) == -1) {
|
|
# endif /* POSIXJOBS */
|
|
stderror(ERR_SYSTEM, "setpgid child:\n", strerror(errno));
|
|
xexit(0);
|
|
# ifdef POSIXJOBS
|
|
}
|
|
wanttty = pgrp; /* Now we really want the tty, since we became the
|
|
* the process group leader
|
|
*/
|
|
# endif /* POSIXJOBS */
|
|
}
|
|
}
|
|
|
|
# ifdef POSIXJOBS
|
|
if (wanttty > 0)
|
|
setttypgrp(pgrp);
|
|
# ifdef BSDSIGS
|
|
(void) sigsetmask(omask);
|
|
# else /* BSDSIGS */
|
|
(void) sigrelse(SIGTSTP);
|
|
(void) sigrelse(SIGTTIN);
|
|
# endif /* !BSDSIGS */
|
|
# endif /* POSIXJOBS */
|
|
|
|
# ifdef JOBDEBUG
|
|
xprintf("wanttty %d pid %d pgrp %d tpgrp %d\n",
|
|
wanttty, getpid(), mygetpgrp(), tcgetpgrp(FSHTTY));
|
|
# endif /* JOBDEBUG */
|
|
|
|
if (tpgrp > 0)
|
|
tpgrp = 0; /* gave tty away */
|
|
#endif /* BSDJOBS */
|
|
}
|