freebsd-dev/sys/kern/kern_exit.c

758 lines
19 KiB
C
Raw Normal View History

/*
1994-05-24 10:09:53 +00:00
* Copyright (c) 1982, 1986, 1989, 1991, 1993
* The Regents of the University of California. All rights reserved.
* (c) UNIX System Laboratories, Inc.
* All or some portions of this file are derived from material licensed
* to the University of California by American Telephone and Telegraph
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
* the permission of UNIX System Laboratories, Inc.
*
* 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.
* 4. 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.
*
* @(#)kern_exit.c 8.7 (Berkeley) 2/12/94
*/
2003-06-11 00:56:59 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_compat.h"
#include "opt_ktrace.h"
#include "opt_mac.h"
1994-05-24 10:09:53 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/sysproto.h>
#include <sys/eventhandler.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/lock.h>
#include <sys/mutex.h>
1994-05-24 10:09:53 +00:00
#include <sys/proc.h>
#include <sys/pioctl.h>
1994-05-24 10:09:53 +00:00
#include <sys/tty.h>
#include <sys/wait.h>
#include <sys/vmmeter.h>
#include <sys/vnode.h>
1994-05-24 10:09:53 +00:00
#include <sys/resourcevar.h>
#include <sys/signalvar.h>
#include <sys/sched.h>
#include <sys/sx.h>
1994-05-24 10:09:53 +00:00
#include <sys/ptrace.h>
#include <sys/acct.h> /* for acct_process() function prototype */
#include <sys/filedesc.h>
#include <sys/mac.h>
#include <sys/shm.h>
#include <sys/sem.h>
#ifdef KTRACE
#include <sys/ktrace.h>
#endif
1994-05-24 10:09:53 +00:00
#include <vm/vm.h>
#include <vm/vm_extern.h>
#include <vm/vm_param.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
#include <vm/vm_page.h>
#include <vm/uma.h>
#include <sys/user.h>
1994-05-24 10:09:53 +00:00
/* Required to be non-static for SysVR4 emulator */
MALLOC_DEFINE(M_ZOMBIE, "zombie", "zombie proc status");
1994-05-24 10:09:53 +00:00
/*
* exit --
* Death of process.
*
* MPSAFE
1994-05-24 10:09:53 +00:00
*/
void
2003-03-19 00:49:40 +00:00
sys_exit(struct thread *td, struct sys_exit_args *uap)
1994-05-24 10:09:53 +00:00
{
exit1(td, W_EXITCODE(uap->rval, 0));
1994-05-24 10:09:53 +00:00
/* NOTREACHED */
}
/*
* Exit: deallocate address space and other resources, change proc state
* to zombie, and unlink proc from allproc and parent's lists. Save exit
* status and rusage for wait(). Check for child processes and orphan them.
*/
void
2003-03-19 00:49:40 +00:00
exit1(struct thread *td, int rv)
1994-05-24 10:09:53 +00:00
{
2004-09-22 15:24:33 +00:00
struct bintime new_switchtime;
struct proc *p, *nq, *q;
struct tty *tp;
struct vnode *ttyvp;
2003-03-19 00:49:40 +00:00
struct vmspace *vm;
struct vnode *vtmp;
#ifdef KTRACE
struct vnode *tracevp;
struct ucred *tracecred;
#endif
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
struct plimit *plim;
int refcnt;
1994-05-24 10:09:53 +00:00
/*
* Drop Giant if caller has it. Eventually we should warn about
* being called with Giant held.
*/
while (mtx_owned(&Giant))
mtx_unlock(&Giant);
p = td->td_proc;
if (p == initproc) {
printf("init died (signal %d, exit %d)\n",
1994-05-24 10:09:53 +00:00
WTERMSIG(rv), WEXITSTATUS(rv));
panic("Going nowhere without my init!");
}
/*
* MUST abort all other threads before proceeding past here.
*/
PROC_LOCK(p);
Refactor a bunch of scheduler code to give basically the same behaviour but with slightly cleaned up interfaces. The KSE structure has become the same as the "per thread scheduler private data" structure. In order to not make the diffs too great one is #defined as the other at this time. The KSE (or td_sched) structure is now allocated per thread and has no allocation code of its own. Concurrency for a KSEGRP is now kept track of via a simple pair of counters rather than using KSE structures as tokens. Since the KSE structure is different in each scheduler, kern_switch.c is now included at the end of each scheduler. Nothing outside the scheduler knows the contents of the KSE (aka td_sched) structure. The fields in the ksegrp structure that are to do with the scheduler's queueing mechanisms are now moved to the kg_sched structure. (per ksegrp scheduler private data structure). In other words how the scheduler queues and keeps track of threads is no-one's business except the scheduler's. This should allow people to write experimental schedulers with completely different internal structuring. A scheduler call sched_set_concurrency(kg, N) has been added that notifies teh scheduler that no more than N threads from that ksegrp should be allowed to be on concurrently scheduled. This is also used to enforce 'fainess' at this time so that a ksegrp with 10000 threads can not swamp a the run queue and force out a process with 1 thread, since the current code will not set the concurrency above NCPU, and both schedulers will not allow more than that many onto the system run queue at a time. Each scheduler should eventualy develop their own methods to do this now that they are effectively separated. Rejig libthr's kernel interface to follow the same code paths as linkse for scope system threads. This has slightly hurt libthr's performance but I will work to recover as much of it as I can. Thread exit code has been cleaned up greatly. exit and exec code now transitions a process back to 'standard non-threaded mode' before taking the next step. Reviewed by: scottl, peter MFC after: 1 week
2004-09-05 02:09:54 +00:00
if (p->p_flag & P_HADTHREADS) {
retry:
/*
* First check if some other thread got here before us..
* if so, act apropriatly, (exit or suspend);
*/
thread_suspend_check(0);
/*
* Kill off the other threads. This requires
* Some co-operation from other parts of the kernel
* so it may not be instant.
* With this state set:
* Any thread entering the kernel from userspace will
* thread_exit() in trap(). Any thread attempting to
* sleep will return immediatly with EINTR or EWOULDBLOCK,
* which will hopefully force them to back out to userland,
* freeing resources as they go, and anything attempting
* to return to userland will thread_exit() from userret().
* thread_exit() will unsuspend us when the last other
* thread exits.
* If there is already a thread singler after resumption,
* calling thread_single will fail, in the case, we just
* re-check all suspension request, the thread should
* either be suspended there or exit.
*/
if (thread_single(SINGLE_EXIT))
goto retry;
/*
* All other activity in this process is now stopped.
Refactor a bunch of scheduler code to give basically the same behaviour but with slightly cleaned up interfaces. The KSE structure has become the same as the "per thread scheduler private data" structure. In order to not make the diffs too great one is #defined as the other at this time. The KSE (or td_sched) structure is now allocated per thread and has no allocation code of its own. Concurrency for a KSEGRP is now kept track of via a simple pair of counters rather than using KSE structures as tokens. Since the KSE structure is different in each scheduler, kern_switch.c is now included at the end of each scheduler. Nothing outside the scheduler knows the contents of the KSE (aka td_sched) structure. The fields in the ksegrp structure that are to do with the scheduler's queueing mechanisms are now moved to the kg_sched structure. (per ksegrp scheduler private data structure). In other words how the scheduler queues and keeps track of threads is no-one's business except the scheduler's. This should allow people to write experimental schedulers with completely different internal structuring. A scheduler call sched_set_concurrency(kg, N) has been added that notifies teh scheduler that no more than N threads from that ksegrp should be allowed to be on concurrently scheduled. This is also used to enforce 'fainess' at this time so that a ksegrp with 10000 threads can not swamp a the run queue and force out a process with 1 thread, since the current code will not set the concurrency above NCPU, and both schedulers will not allow more than that many onto the system run queue at a time. Each scheduler should eventualy develop their own methods to do this now that they are effectively separated. Rejig libthr's kernel interface to follow the same code paths as linkse for scope system threads. This has slightly hurt libthr's performance but I will work to recover as much of it as I can. Thread exit code has been cleaned up greatly. exit and exec code now transitions a process back to 'standard non-threaded mode' before taking the next step. Reviewed by: scottl, peter MFC after: 1 week
2004-09-05 02:09:54 +00:00
* Threading support has been turned off.
*/
}
p->p_flag |= P_WEXIT;
PROC_UNLOCK(p);
/* Are we a task leader? */
if (p == p->p_leader) {
mtx_lock(&ppeers_lock);
q = p->p_peers;
while (q != NULL) {
PROC_LOCK(q);
psignal(q, SIGKILL);
PROC_UNLOCK(q);
q = q->p_peers;
}
2003-03-19 00:33:38 +00:00
while (p->p_peers != NULL)
msleep(p, &ppeers_lock, PWAIT, "exit1", 0);
mtx_unlock(&ppeers_lock);
}
PROC_LOCK(p);
_STOPEVENT(p, S_EXIT, rv);
wakeup(&p->p_stype); /* Wakeup anyone in procfs' PIOCWAIT */
PROC_UNLOCK(p);
2003-03-19 00:33:38 +00:00
/*
* Check if any loadable modules need anything done at process exit.
* e.g. SYSV IPC stuff
* XXX what if one of these generates an error?
*/
EVENTHANDLER_INVOKE(process_exit, p);
1994-05-24 10:09:53 +00:00
MALLOC(p->p_ru, struct rusage *, sizeof(struct rusage),
M_ZOMBIE, M_WAITOK);
1994-05-24 10:09:53 +00:00
/*
* If parent is waiting for us to exit or exec,
* P_PPWAIT is set; we will wakeup the parent below.
*/
PROC_LOCK(p);
stopprofclock(p);
1994-05-24 10:09:53 +00:00
p->p_flag &= ~(P_TRACED | P_PPWAIT);
SIGEMPTYSET(p->p_siglist);
SIGEMPTYSET(td->td_siglist);
/*
* Stop the real interval timer. If the handler is currently
* executing, prevent it from rearming itself and let it finish.
*/
if (timevalisset(&p->p_realtimer.it_value) &&
callout_stop(&p->p_itcallout) == 0) {
timevalclear(&p->p_realtimer.it_interval);
msleep(&p->p_itcallout, &p->p_mtx, PWAIT, "ritwait", 0);
KASSERT(!timevalisset(&p->p_realtimer.it_value),
("realtime timer is still armed"));
}
PROC_UNLOCK(p);
1994-05-24 10:09:53 +00:00
/*
* Reset any sigio structures pointing to us as a result of
* F_SETOWN with our pid.
*/
mtx_lock(&Giant); /* XXX: not sure if needed */
funsetownlst(&p->p_sigiolst);
1994-05-24 10:09:53 +00:00
/*
* Close open files and release open-file table.
* This may block!
*/
fdfree(td);
mtx_unlock(&Giant);
1994-05-24 10:09:53 +00:00
/*
* Remove ourself from our leader's peer list and wake our leader.
*/
mtx_lock(&ppeers_lock);
if (p->p_leader->p_peers) {
q = p->p_leader;
while (q->p_peers != p)
q = q->p_peers;
q->p_peers = p->p_peers;
wakeup(p->p_leader);
}
mtx_unlock(&ppeers_lock);
1994-05-24 10:09:53 +00:00
/* The next two chunks should probably be moved to vmspace_exit. */
vm = p->p_vmspace;
/*
* Release user portion of address space.
* This releases references to vnodes,
* which could cause I/O if the file has been unlinked.
* Need to do this early enough that we can still sleep.
* Can't free the entire vmspace as the kernel stack
* may be mapped within that space also.
*
* Processes sharing the same vmspace may exit in one order, and
* get cleaned up by vmspace_exit() in a different order. The
* last exiting process to reach this point releases as much of
* the environment as it can, and the last process cleaned up
* by vmspace_exit() (which decrements exitingcnt) cleans up the
* remainder.
1994-05-24 10:09:53 +00:00
*/
atomic_add_int(&vm->vm_exitingcnt, 1);
do
refcnt = vm->vm_refcnt;
while (!atomic_cmpset_int(&vm->vm_refcnt, refcnt, refcnt - 1));
if (refcnt == 1) {
shmexit(vm);
pmap_remove_pages(vmspace_pmap(vm), vm_map_min(&vm->vm_map),
vm_map_max(&vm->vm_map));
(void) vm_map_remove(&vm->vm_map, vm_map_min(&vm->vm_map),
vm_map_max(&vm->vm_map));
}
1994-05-24 10:09:53 +00:00
mtx_lock(&Giant);
sx_xlock(&proctree_lock);
1994-05-24 10:09:53 +00:00
if (SESS_LEADER(p)) {
2003-03-19 00:49:40 +00:00
struct session *sp;
1994-05-24 10:09:53 +00:00
sp = p->p_session;
1994-05-24 10:09:53 +00:00
if (sp->s_ttyvp) {
/*
* Controlling process.
* Signal foreground pgrp,
* drain controlling terminal
* and revoke access to controlling terminal.
*/
if (sp->s_ttyp && (sp->s_ttyp->t_session == sp)) {
tp = sp->s_ttyp;
if (sp->s_ttyp->t_pgrp) {
PGRP_LOCK(sp->s_ttyp->t_pgrp);
1994-05-24 10:09:53 +00:00
pgsignal(sp->s_ttyp->t_pgrp, SIGHUP, 1);
PGRP_UNLOCK(sp->s_ttyp->t_pgrp);
}
/* XXX tp should be locked. */
sx_xunlock(&proctree_lock);
(void) ttywait(tp);
sx_xlock(&proctree_lock);
1994-05-24 10:09:53 +00:00
/*
* The tty could have been revoked
* if we blocked.
*/
if (sp->s_ttyvp) {
ttyvp = sp->s_ttyvp;
SESS_LOCK(p->p_session);
sp->s_ttyvp = NULL;
SESS_UNLOCK(p->p_session);
sx_xunlock(&proctree_lock);
VOP_REVOKE(ttyvp, REVOKEALL);
vrele(ttyvp);
sx_xlock(&proctree_lock);
}
}
if (sp->s_ttyvp) {
ttyvp = sp->s_ttyvp;
SESS_LOCK(p->p_session);
sp->s_ttyvp = NULL;
SESS_UNLOCK(p->p_session);
vrele(ttyvp);
1994-05-24 10:09:53 +00:00
}
/*
* s_ttyp is not zero'd; we use this to indicate
* that the session once had a controlling terminal.
* (for logging and informational purposes)
*/
}
SESS_LOCK(p->p_session);
1994-05-24 10:09:53 +00:00
sp->s_leader = NULL;
SESS_UNLOCK(p->p_session);
}
1994-05-24 10:09:53 +00:00
fixjobc(p, p->p_pgrp, 0);
sx_xunlock(&proctree_lock);
(void)acct_process(td);
mtx_unlock(&Giant);
1994-05-24 10:09:53 +00:00
#ifdef KTRACE
1995-05-30 08:16:23 +00:00
/*
1994-05-24 10:09:53 +00:00
* release trace file
*/
PROC_LOCK(p);
mtx_lock(&ktrace_mtx);
1994-05-24 10:09:53 +00:00
p->p_traceflag = 0; /* don't trace the vrele() */
tracevp = p->p_tracevp;
p->p_tracevp = NULL;
tracecred = p->p_tracecred;
p->p_tracecred = NULL;
mtx_unlock(&ktrace_mtx);
PROC_UNLOCK(p);
if (tracevp != NULL) {
mtx_lock(&Giant);
vrele(tracevp);
mtx_unlock(&Giant);
}
if (tracecred != NULL)
crfree(tracecred);
1994-05-24 10:09:53 +00:00
#endif
/*
* Release reference to text vnode
*/
if ((vtmp = p->p_textvp) != NULL) {
p->p_textvp = NULL;
mtx_lock(&Giant);
vrele(vtmp);
mtx_unlock(&Giant);
}
/*
* Release our limits structure.
*/
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
PROC_LOCK(p);
plim = p->p_limit;
p->p_limit = NULL;
PROC_UNLOCK(p);
lim_free(plim);
1994-05-24 10:09:53 +00:00
/*
* Remove proc from allproc queue and pidhash chain.
* Place onto zombproc. Unlink from parent's child list.
*/
sx_xlock(&allproc_lock);
LIST_REMOVE(p, p_list);
LIST_INSERT_HEAD(&zombproc, p, p_list);
LIST_REMOVE(p, p_hash);
sx_xunlock(&allproc_lock);
1994-05-24 10:09:53 +00:00
sx_xlock(&proctree_lock);
q = LIST_FIRST(&p->p_children);
if (q != NULL) /* only need this if any child is S_ZOMB */
wakeup(initproc);
for (; q != NULL; q = nq) {
nq = LIST_NEXT(q, p_sibling);
PROC_LOCK(q);
proc_reparent(q, initproc);
q->p_sigparent = SIGCHLD;
1994-05-24 10:09:53 +00:00
/*
* Traced processes are killed
* since their existence means someone is screwing up.
*/
if (q->p_flag & P_TRACED) {
q->p_flag &= ~P_TRACED;
psignal(q, SIGKILL);
}
PROC_UNLOCK(q);
1994-05-24 10:09:53 +00:00
}
/*
* Save exit status and final rusage info, adding in child rusage
* info and self times.
*/
mtx_lock(&Giant);
PROC_LOCK(p);
1994-05-24 10:09:53 +00:00
p->p_xstat = rv;
Add code to support debugging threaded process. 1. Add tm_lwpid into kse_thr_mailbox to indicate which kernel thread current user thread is running on. Add tm_dflags into kse_thr_mailbox, the flags is written by debugger, it tells UTS and kernel what should be done when the process is being debugged, current, there two flags TMDF_SSTEP and TMDF_DONOTRUNUSER. TMDF_SSTEP is used to tell kernel to turn on single stepping, or turn off if it is not set. TMDF_DONOTRUNUSER is used to tell kernel to schedule upcall whenever possible, to UTS, it means do not run the user thread until debugger clears it, this behaviour is necessary because gdb wants to resume only one thread when the thread's pc is at a breakpoint, and thread needs to go forward, in order to avoid other threads sneak pass the breakpoints, it needs to remove breakpoint, only wants one thread to go. Also, add km_lwp to kse_mailbox, the lwp id is copied to kse_thr_mailbox at context switch time when process is not being debugged, so when process is attached, debugger can map kernel thread to user thread. 2. Add p_xthread to proc strcuture and td_xsig to thread structure. p_xthread is used by a thread when it wants to report event to debugger, every thread can set the pointer, especially, when it is used in ptracestop, it is the last thread reporting event will win the race. Every thread has a td_xsig to exchange signal with debugger, thread uses TDF_XSIG flag to indicate it is reporting signal to debugger, if the flag is not cleared, thread will keep retrying until it is cleared by debugger, p_xthread may be used by debugger to indicate CURRENT thread. The p_xstat is still in proc structure to keep wait() to work, in future, we may just use td_xsig. 3. Add TDF_DBSUSPEND flag, the flag is used by debugger to suspend a thread. When process stops, debugger can set the flag for thread, thread will check the flag in thread_suspend_check, enters a loop, unless it is cleared by debugger, process is detached or process is existing. The flag is also checked in ptracestop, so debugger can temporarily suspend a thread even if the thread wants to exchange signal. 4. Current, in ptrace, we always resume all threads, but if a thread has already a TDF_DBSUSPEND flag set by debugger, it won't run. Encouraged by: marcel, julian, deischen
2004-07-13 07:20:10 +00:00
p->p_xthread = td;
1994-05-24 10:09:53 +00:00
*p->p_ru = p->p_stats->p_ru;
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock_spin(&sched_lock);
1994-05-24 10:09:53 +00:00
calcru(p, &p->p_ru->ru_utime, &p->p_ru->ru_stime, NULL);
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock_spin(&sched_lock);
1994-05-24 10:09:53 +00:00
ruadd(p->p_ru, &p->p_stats->p_cru);
mtx_unlock(&Giant);
/*
* Notify interested parties of our demise.
*/
KNOTE_LOCKED(&p->p_klist, NOTE_EXIT);
2004-09-22 15:24:33 +00:00
/*
* Just delete all entries in the p_klist. At this point we won't
* report any more events, and there are nasty race conditions that
* can beat us if we don't.
*/
knlist_clear(&p->p_klist, 1);
1994-05-24 10:09:53 +00:00
/*
* Notify parent that we're gone. If parent has the PS_NOCLDWAIT
* flag set, or if the handler is set to SIG_IGN, notify process
* 1 instead (and hope it will handle this situation).
1994-05-24 10:09:53 +00:00
*/
PROC_LOCK(p->p_pptr);
mtx_lock(&p->p_pptr->p_sigacts->ps_mtx);
if (p->p_pptr->p_sigacts->ps_flag & (PS_NOCLDWAIT | PS_CLDSIGIGN)) {
struct proc *pp;
mtx_unlock(&p->p_pptr->p_sigacts->ps_mtx);
pp = p->p_pptr;
PROC_UNLOCK(pp);
proc_reparent(p, initproc);
p->p_sigparent = SIGCHLD;
PROC_LOCK(p->p_pptr);
/*
* If this was the last child of our parent, notify
* parent, so in case he was wait(2)ing, he will
* continue.
*/
if (LIST_EMPTY(&pp->p_children))
wakeup(pp);
} else
mtx_unlock(&p->p_pptr->p_sigacts->ps_mtx);
if (p->p_pptr == initproc)
2002-10-02 23:12:01 +00:00
psignal(p->p_pptr, SIGCHLD);
else if (p->p_sigparent != 0)
psignal(p->p_pptr, p->p_sigparent);
PROC_UNLOCK(p->p_pptr);
/*
* If this is a kthread, then wakeup anyone waiting for it to exit.
*/
if (p->p_flag & P_KTHREAD)
wakeup(p);
PROC_UNLOCK(p);
2003-03-19 00:33:38 +00:00
1994-05-24 10:09:53 +00:00
/*
* Finally, call machine-dependent code to release the remaining
* resources including address space.
* The address space is released by "vmspace_exitfree(p)" in
* vm_waitproc().
1994-05-24 10:09:53 +00:00
*/
cpu_exit(td);
PROC_LOCK(p);
PROC_LOCK(p->p_pptr);
sx_xunlock(&proctree_lock);
while (mtx_owned(&Giant))
Change the preemption code for software interrupt thread schedules and mutex releases to not require flags for the cases when preemption is not allowed: The purpose of the MTX_NOSWITCH and SWI_NOSWITCH flags is to prevent switching to a higher priority thread on mutex releease and swi schedule, respectively when that switch is not safe. Now that the critical section API maintains a per-thread nesting count, the kernel can easily check whether or not it should switch without relying on flags from the programmer. This fixes a few bugs in that all current callers of swi_sched() used SWI_NOSWITCH, when in fact, only the ones called from fast interrupt handlers and the swi_sched of softclock needed this flag. Note that to ensure that swi_sched()'s in clock and fast interrupt handlers do not switch, these handlers have to be explicitly wrapped in critical_enter/exit pairs. Presently, just wrapping the handlers is sufficient, but in the future with the fully preemptive kernel, the interrupt must be EOI'd before critical_exit() is called. (critical_exit() can switch due to a deferred preemption in a fully preemptive kernel.) I've tested the changes to the interrupt code on i386 and alpha. I have not tested ia64, but the interrupt code is almost identical to the alpha code, so I expect it will work fine. PowerPC and ARM do not yet have interrupt code in the tree so they shouldn't be broken. Sparc64 is broken, but that's been ok'd by jake and tmm who will be fixing the interrupt code for sparc64 shortly. Reviewed by: peter Tested on: i386, alpha
2002-01-05 08:47:13 +00:00
mtx_unlock(&Giant);
/*
2003-04-17 22:22:47 +00:00
* We have to wait until after acquiring all locks before
* changing p_state. We need to avoid any possibly context
* switches while marked as a zombie including blocking on
* a mutex.
*/
mtx_lock_spin(&sched_lock);
p->p_state = PRS_ZOMBIE;
critical_enter();
mtx_unlock_spin(&sched_lock);
wakeup(p->p_pptr);
PROC_UNLOCK(p->p_pptr);
mtx_lock_spin(&sched_lock);
critical_exit();
/* Do the same timestamp bookkeeping that mi_switch() would do. */
binuptime(&new_switchtime);
bintime_add(&p->p_runtime, &new_switchtime);
bintime_sub(&p->p_runtime, PCPU_PTR(switchtime));
PCPU_SET(switchtime, new_switchtime);
PCPU_SET(switchticks, ticks);
cnt.v_swtch++;
sched_exit(p->p_pptr, td);
/*
2004-09-22 15:24:33 +00:00
* Hopefully no one will try to deliver a signal to the process this
* late in the game.
*/
knlist_destroy(&p->p_klist);
/*
* Make sure the scheduler takes this thread out of its tables etc.
* This will also release this thread's reference to the ucred.
2003-03-19 00:33:38 +00:00
* Other thread parts to release include pcb bits and such.
*/
thread_exit();
1994-05-24 10:09:53 +00:00
}
#ifdef COMPAT_43
/*
* MPSAFE. The dirty work is handled by kern_wait().
*/
int
2003-03-19 00:49:40 +00:00
owait(struct thread *td, struct owait_args *uap __unused)
1994-05-24 10:09:53 +00:00
{
int error, status;
1994-05-24 10:09:53 +00:00
error = kern_wait(td, WAIT_ANY, &status, 0, NULL);
if (error == 0)
td->td_retval[1] = status;
return (error);
1994-05-24 10:09:53 +00:00
}
#endif /* COMPAT_43 */
1994-05-24 10:09:53 +00:00
/*
* MPSAFE. The dirty work is handled by kern_wait().
*/
int
2003-03-19 00:49:40 +00:00
wait4(struct thread *td, struct wait_args *uap)
1994-05-24 10:09:53 +00:00
{
struct rusage ru;
int error, status;
error = kern_wait(td, uap->pid, &status, uap->options, &ru);
if (uap->status != NULL && error == 0)
error = copyout(&status, uap->status, sizeof(status));
if (uap->rusage != NULL && error == 0)
error = copyout(&ru, uap->rusage, sizeof(struct rusage));
return (error);
1994-05-24 10:09:53 +00:00
}
int
2004-09-22 15:24:33 +00:00
kern_wait(struct thread *td, pid_t pid, int *status, int options,
struct rusage *rusage)
1994-05-24 10:09:53 +00:00
{
int nfound;
struct proc *p, *q, *t;
int error;
1994-05-24 10:09:53 +00:00
q = td->td_proc;
if (pid == 0) {
PROC_LOCK(q);
pid = -q->p_pgid;
PROC_UNLOCK(q);
}
if (options &~ (WUNTRACED|WNOHANG|WCONTINUED|WLINUXCLONE))
return (EINVAL);
1994-05-24 10:09:53 +00:00
loop:
nfound = 0;
sx_xlock(&proctree_lock);
LIST_FOREACH(p, &q->p_children, p_sibling) {
PROC_LOCK(p);
if (pid != WAIT_ANY &&
p->p_pid != pid && p->p_pgid != -pid) {
PROC_UNLOCK(p);
1994-05-24 10:09:53 +00:00
continue;
}
/*
2003-03-19 00:33:38 +00:00
* This special case handles a kthread spawned by linux_clone
* (see linux_misc.c). The linux_wait4 and linux_waitpid
* functions need to be able to distinguish between waiting
* on a process and waiting on a thread. It is a thread if
* p_sigparent is not SIGCHLD, and the WLINUXCLONE option
* signifies we want to wait for threads and not processes.
*/
if ((p->p_sigparent != SIGCHLD) ^
((options & WLINUXCLONE) != 0)) {
PROC_UNLOCK(p);
continue;
}
1994-05-24 10:09:53 +00:00
nfound++;
if (p->p_state == PRS_ZOMBIE) {
td->td_retval[0] = p->p_pid;
if (status)
*status = p->p_xstat; /* convert to int */
if (rusage)
bcopy(p->p_ru, rusage, sizeof(struct rusage));
1994-05-24 10:09:53 +00:00
/*
* If we got the child via a ptrace 'attach',
* we need to give it back to the old parent.
*/
PROC_UNLOCK(p);
if (p->p_oppid && (t = pfind(p->p_oppid)) != NULL) {
PROC_LOCK(p);
p->p_oppid = 0;
proc_reparent(p, t);
PROC_UNLOCK(p);
psignal(t, SIGCHLD);
wakeup(t);
PROC_UNLOCK(t);
sx_xunlock(&proctree_lock);
return (0);
1994-05-24 10:09:53 +00:00
}
/*
* Remove other references to this process to ensure
* we have an exclusive reference.
*/
sx_xlock(&allproc_lock);
LIST_REMOVE(p, p_list); /* off zombproc */
sx_xunlock(&allproc_lock);
LIST_REMOVE(p, p_sibling);
leavepgrp(p);
sx_xunlock(&proctree_lock);
/*
* As a side effect of this lock, we know that
* all other writes to this proc are visible now, so
* no more locking is needed for p.
*/
mtx_lock(&Giant);
PROC_LOCK(p);
p->p_xstat = 0; /* XXX: why? */
PROC_UNLOCK(p);
PROC_LOCK(q);
1994-05-24 10:09:53 +00:00
ruadd(&q->p_stats->p_cru, p->p_ru);
PROC_UNLOCK(q);
1994-05-24 10:09:53 +00:00
FREE(p->p_ru, M_ZOMBIE);
p->p_ru = NULL;
mtx_unlock(&Giant);
1994-05-24 10:09:53 +00:00
/*
* Decrement the count of procs running with this uid.
*/
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
(void)chgproccnt(p->p_ucred->cr_ruidinfo, -1, 0);
1994-05-24 10:09:53 +00:00
/*
* Free credentials, arguments, and sigacts
1994-05-24 10:09:53 +00:00
*/
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
crfree(p->p_ucred);
p->p_ucred = NULL;
pargs_drop(p->p_args);
p->p_args = NULL;
sigacts_free(p->p_sigacts);
p->p_sigacts = NULL;
/*
* do any thread-system specific cleanups
*/
thread_wait(p);
1994-05-24 10:09:53 +00:00
/*
* Give vm and machine-dependent layer a chance
1994-05-24 10:09:53 +00:00
* to free anything that cpu_exit couldn't
* release while still running in process context.
*/
vm_waitproc(p);
#ifdef MAC
mac_destroy_proc(p);
#endif
KASSERT(FIRST_THREAD_IN_PROC(p),
("kern_wait: no residual thread!"));
uma_zfree(proc_zone, p);
sx_xlock(&allproc_lock);
1994-05-24 10:09:53 +00:00
nprocs--;
sx_xunlock(&allproc_lock);
return (0);
1994-05-24 10:09:53 +00:00
}
mtx_lock_spin(&sched_lock);
if (P_SHOULDSTOP(p) && (p->p_suspcount == p->p_numthreads) &&
((p->p_flag & P_WAITED) == 0) &&
(p->p_flag & P_TRACED || options & WUNTRACED)) {
mtx_unlock_spin(&sched_lock);
1994-05-24 10:09:53 +00:00
p->p_flag |= P_WAITED;
sx_xunlock(&proctree_lock);
td->td_retval[0] = p->p_pid;
if (status)
*status = W_STOPCODE(p->p_xstat);
PROC_UNLOCK(p);
return (0);
1994-05-24 10:09:53 +00:00
}
mtx_unlock_spin(&sched_lock);
if (options & WCONTINUED && (p->p_flag & P_CONTINUED)) {
sx_xunlock(&proctree_lock);
td->td_retval[0] = p->p_pid;
p->p_flag &= ~P_CONTINUED;
PROC_UNLOCK(p);
if (status)
*status = SIGCONT;
return (0);
}
PROC_UNLOCK(p);
1994-05-24 10:09:53 +00:00
}
if (nfound == 0) {
sx_xunlock(&proctree_lock);
return (ECHILD);
}
if (options & WNOHANG) {
sx_xunlock(&proctree_lock);
td->td_retval[0] = 0;
return (0);
}
PROC_LOCK(q);
sx_xunlock(&proctree_lock);
error = msleep(q, &q->p_mtx, PWAIT | PCATCH, "wait", 0);
PROC_UNLOCK(q);
if (error)
return (error);
1994-05-24 10:09:53 +00:00
goto loop;
}
/*
* Make process 'parent' the new parent of process 'child'.
* Must be called with an exclusive hold of proctree lock.
1994-05-24 10:09:53 +00:00
*/
void
2003-03-19 00:49:40 +00:00
proc_reparent(struct proc *child, struct proc *parent)
1994-05-24 10:09:53 +00:00
{
sx_assert(&proctree_lock, SX_XLOCKED);
PROC_LOCK_ASSERT(child, MA_OWNED);
1994-05-24 10:09:53 +00:00
if (child->p_pptr == parent)
return;
LIST_REMOVE(child, p_sibling);
LIST_INSERT_HEAD(&parent->p_children, child, p_sibling);
1994-05-24 10:09:53 +00:00
child->p_pptr = parent;
}