freebsd-dev/sys/kern/kern_exit.c

639 lines
16 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 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
1999-08-28 01:08:13 +00:00
* $FreeBSD$
1994-05-24 10:09:53 +00:00
*/
#include "opt_compat.h"
#include "opt_ktrace.h"
1994-05-24 10:09:53 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/sysproto.h>
#include <sys/kernel.h>
#include <sys/malloc.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/vnode.h>
#include <sys/resourcevar.h>
#include <sys/signalvar.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/shm.h>
#include <sys/sem.h>
#include <sys/aio.h>
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
#include <sys/jail.h>
#include <machine/limits.h> /* for UCHAR_MAX = typeof(p_priority)_MAX */
1994-05-24 10:09:53 +00:00
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <sys/lock.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
VM level code cleanups. 1) Start using TSM. Struct procs continue to point to upages structure, after being freed. Struct vmspace continues to point to pte object and kva space for kstack. u_map is now superfluous. 2) vm_map's don't need to be reference counted. They always exist either in the kernel or in a vmspace. The vmspaces are managed by reference counts. 3) Remove the "wired" vm_map nonsense. 4) No need to keep a cache of kernel stack kva's. 5) Get rid of strange looking ++var, and change to var++. 6) Change more data structures to use our "zone" allocator. Added struct proc, struct vmspace and struct vnode. This saves a significant amount of kva space and physical memory. Additionally, this enables TSM for the zone managed memory. 7) Keep ioopt disabled for now. 8) Remove the now bogus "single use" map concept. 9) Use generation counts or id's for data structures residing in TSM, where it allows us to avoid unneeded restart overhead during traversals, where blocking might occur. 10) Account better for memory deficits, so the pageout daemon will be able to make enough memory available (experimental.) 11) Fix some vnode locking problems. (From Tor, I think.) 12) Add a check in ufs_lookup, to avoid lots of unneeded calls to bcmp. (experimental.) 13) Significantly shrink, cleanup, and make slightly faster the vm_fault.c code. Use generation counts, get rid of unneded collpase operations, and clean up the cluster code. 14) Make vm_zone more suitable for TSM. This commit is partially as a result of discussions and contributions from other people, including DG, Tor Egge, PHK, and probably others that I have forgotten to attribute (so let me know, if I forgot.) This is not the infamous, final cleanup of the vnode stuff, but a necessary step. Vnode mgmt should be correct, but things might still change, and there is still some missing stuff (like ioopt, and physical backing of non-merged cache files, debugging of layering concepts.)
1998-01-22 17:30:44 +00:00
#include <vm/vm_zone.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");
static int wait1 __P((struct proc *, struct wait_args *, int));
/*
* callout list for things to do at exit time
*/
typedef struct exit_list_element {
struct exit_list_element *next;
exitlist_fn function;
} *ele_p;
static ele_p exit_list;
1994-05-24 10:09:53 +00:00
/*
* exit --
* Death of process.
*/
void
exit(p, uap)
1994-05-24 10:09:53 +00:00
struct proc *p;
struct rexit_args /* {
int rval;
} */ *uap;
1994-05-24 10:09:53 +00:00
{
exit1(p, W_EXITCODE(uap->rval, 0));
/* 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
1994-05-24 10:09:53 +00:00
exit1(p, rv)
register struct proc *p;
int rv;
{
register struct proc *q, *nq;
register struct vmspace *vm;
ele_p ep = exit_list;
1994-05-24 10:09:53 +00:00
if (p->p_pid == 1) {
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!");
}
aio_proc_rundown(p);
/* are we a task leader? */
if(p == p->p_leader) {
struct kill_args killArgs;
killArgs.signum = SIGKILL;
q = p->p_peers;
while(q) {
killArgs.pid = q->p_pid;
/*
* The interface for kill is better
* than the internal signal
*/
kill(p, &killArgs);
nq = q;
q = q->p_peers;
}
while (p->p_peers)
tsleep((caddr_t)p, PWAIT, "exit1", 0);
}
1994-05-24 10:09:53 +00:00
#ifdef PGINPROF
vmsizmon();
#endif
STOPEVENT(p, S_EXIT, rv);
/*
* 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?
*/
while (ep) {
(*ep->function)(p);
ep = ep->next;
}
1994-05-24 10:09:53 +00:00
if (p->p_flag & P_PROFIL)
stopprofclock(p);
MALLOC(p->p_ru, struct rusage *, sizeof(struct rusage),
M_ZOMBIE, M_WAITOK);
/*
* If parent is waiting for us to exit or exec,
* P_PPWAIT is set; we will wakeup the parent below.
*/
p->p_flag &= ~(P_TRACED | P_PPWAIT);
p->p_flag |= P_WEXIT;
SIGEMPTYSET(p->p_siglist);
if (timevalisset(&p->p_realtimer.it_value))
init_main.c subr_autoconf.c: Add support for "interrupt driven configuration hooks". A component of the kernel can register a hook, most likely during auto-configuration, and receive a callback once interrupt services are available. This callback will occur before the root and dump devices are configured, so the configuration task can affect the selection of those two devices or complete any tasks that need to be performed prior to launching init. System boot is posponed so long as a hook is registered. The hook owner is responsible for removing the hook once their task is complete or the system boot can continue. kern_acct.c kern_clock.c kern_exit.c kern_synch.c kern_time.c: Change the interface and implementation for the kernel callout service. The new implemntaion is based on the work of Adam M. Costello and George Varghese, published in a technical report entitled "Redesigning the BSD Callout and Timer Facilities". The interface used in FreeBSD is a little different than the one outlined in the paper. The new function prototypes are: struct callout_handle timeout(void (*func)(void *), void *arg, int ticks); void untimeout(void (*func)(void *), void *arg, struct callout_handle handle); If a client wishes to remove a timeout, it must store the callout_handle returned by timeout and pass it to untimeout. The new implementation gives 0(1) insert and removal of callouts making this interface scale well even for applications that keep 100s of callouts outstanding. See the updated timeout.9 man page for more details.
1997-09-21 22:00:25 +00:00
untimeout(realitexpire, (caddr_t)p, p->p_ithandle);
1994-05-24 10:09:53 +00:00
/*
* Reset any sigio structures pointing to us as a result of
* F_SETOWN with our pid.
*/
funsetownlst(&p->p_sigiolst);
1994-05-24 10:09:53 +00:00
/*
* Close open files and release open-file table.
* This may block!
*/
fdfree(p);
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((caddr_t)p->p_leader);
}
/*
* XXX Shutdown SYSV semaphores
*/
semexit(p);
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.
*/
if (vm->vm_refcnt == 1) {
The biggie: Get rid of the UPAGES from the top of the per-process address space. (!) Have each process use the kernel stack and pcb in the kvm space. Since the stacks are at a different address, we cannot copy the stack at fork() and allow the child to return up through the function call tree to return to user mode - create a new execution context and have the new process begin executing from cpu_switch() and go to user mode directly. In theory this should speed up fork a bit. Context switch the tss_esp0 pointer in the common tss. This is a lot simpler since than swithching the gdt[GPROC0_SEL].sd.sd_base pointer to each process's tss since the esp0 pointer is a 32 bit pointer, and the sd_base setting is split into three different bit sections at non-aligned boundaries and requires a lot of twiddling to reset. The 8K of memory at the top of the process space is now empty, and unmapped (and unmappable, it's higher than VM_MAXUSER_ADDRESS). Simplity the pmap code to manage process contexts, we no longer have to double map the UPAGES, this simplifies and should measuably speed up fork(). The following parts came from John Dyson: Set PG_G on the UPAGES that are now in kernel context, and invalidate them when swapping them out. Move the upages object (upobj) from the vmspace to the proc structure. Now that the UPAGES (pcb and kernel stack) are out of user space, make rfork(..RFMEM..) do what was intended by sharing the vmspace entirely via reference counting rather than simply inheriting the mappings.
1997-04-07 07:16:06 +00:00
if (vm->vm_shm)
shmexit(p);
pmap_remove_pages(vmspace_pmap(vm), VM_MIN_ADDRESS,
VM_MAXUSER_ADDRESS);
(void) vm_map_remove(&vm->vm_map, VM_MIN_ADDRESS,
VM_MAXUSER_ADDRESS);
}
1994-05-24 10:09:53 +00:00
if (SESS_LEADER(p)) {
register struct session *sp = p->p_session;
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)) {
1994-05-24 10:09:53 +00:00
if (sp->s_ttyp->t_pgrp)
pgsignal(sp->s_ttyp->t_pgrp, SIGHUP, 1);
(void) ttywait(sp->s_ttyp);
/*
* The tty could have been revoked
* if we blocked.
*/
if (sp->s_ttyvp)
VOP_REVOKE(sp->s_ttyvp, REVOKEALL);
1994-05-24 10:09:53 +00:00
}
if (sp->s_ttyvp)
vrele(sp->s_ttyvp);
sp->s_ttyvp = NULL;
/*
* s_ttyp is not zero'd; we use this to indicate
* that the session once had a controlling terminal.
* (for logging and informational purposes)
*/
}
sp->s_leader = NULL;
}
fixjobc(p, p->p_pgrp, 0);
(void)acct_process(p);
#ifdef KTRACE
1995-05-30 08:16:23 +00:00
/*
1994-05-24 10:09:53 +00:00
* release trace file
*/
p->p_traceflag = 0; /* don't trace the vrele() */
if (p->p_tracep)
vrele(p->p_tracep);
#endif
/*
* Remove proc from allproc queue and pidhash chain.
* Place onto zombproc. Unlink from parent's child list.
*/
LIST_REMOVE(p, p_list);
LIST_INSERT_HEAD(&zombproc, p, p_list);
1994-05-24 10:09:53 +00:00
p->p_stat = SZOMB;
LIST_REMOVE(p, p_hash);
1994-05-24 10:09:53 +00:00
q = p->p_children.lh_first;
if (q) /* only need this if any child is S_ZOMB */
1994-05-24 10:09:53 +00:00
wakeup((caddr_t) initproc);
for (; q != 0; q = nq) {
nq = q->p_sibling.le_next;
LIST_REMOVE(q, p_sibling);
LIST_INSERT_HEAD(&initproc->p_children, q, p_sibling);
1994-05-24 10:09:53 +00:00
q->p_pptr = 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);
}
}
/*
* Save exit status and final rusage info, adding in child rusage
* info and self times.
*/
p->p_xstat = rv;
*p->p_ru = p->p_stats->p_ru;
calcru(p, &p->p_ru->ru_utime, &p->p_ru->ru_stime, NULL);
ruadd(p->p_ru, &p->p_stats->p_cru);
/*
* Pretend that an mi_switch() to the next process occurs now. We
* must set `switchtime' directly since we will call cpu_switch()
* directly. Set it now so that the rest of the exit time gets
* counted somewhere if possible.
*/
microuptime(&switchtime);
switchticks = ticks;
1994-05-24 10:09:53 +00:00
/*
* Notify parent that we're gone. If parent has the PS_NOCLDWAIT
* flag set, notify process 1 instead (and hope it will handle
* this situation).
1994-05-24 10:09:53 +00:00
*/
if (p->p_pptr->p_procsig->ps_flag & PS_NOCLDWAIT) {
struct proc *pp = p->p_pptr;
proc_reparent(p, initproc);
/*
* 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((caddr_t)pp);
}
if (p->p_sigparent && p->p_pptr != initproc) {
psignal(p->p_pptr, p->p_sigparent);
} else {
psignal(p->p_pptr, SIGCHLD);
}
1994-05-24 10:09:53 +00:00
wakeup((caddr_t)p->p_pptr);
#if defined(tahoe)
/* move this to cpu_exit */
p->p_addr->u_pcb.pcb_savacc.faddr = (float *)NULL;
#endif
/*
* Clear curproc after we've done all operations
* that could block, and before tearing down the rest
* of the process state that might be used from clock, etc.
* Also, can't clear curproc while we're still runnable,
* as we're not on a run queue (we are current, just not
* a proper proc any longer!).
*
* Other substructures are freed from wait().
*/
SET_CURPROC(NULL);
if (--p->p_limit->p_refcnt == 0) {
1994-05-24 10:09:53 +00:00
FREE(p->p_limit, M_SUBPROC);
p->p_limit = NULL;
}
1994-05-24 10:09:53 +00:00
/*
* Finally, call machine-dependent code to release the remaining
* resources including address space, the kernel stack and pcb.
* The address space is released by "vmspace_free(p->p_vmspace)";
* This is machine-dependent, as we may have to change stacks
* or ensure that the current one isn't reallocated before we
* finish. cpu_exit will end with a call to cpu_switch(), finishing
1994-05-24 10:09:53 +00:00
* our execution (pun intended).
*/
cpu_exit(p);
}
#ifdef COMPAT_43
int
owait(p, uap)
1994-05-24 10:09:53 +00:00
struct proc *p;
register struct owait_args /* {
int dummy;
} */ *uap;
1994-05-24 10:09:53 +00:00
{
struct wait_args w;
1994-05-24 10:09:53 +00:00
w.options = 0;
w.rusage = NULL;
w.pid = WAIT_ANY;
w.status = NULL;
return (wait1(p, &w, 1));
1994-05-24 10:09:53 +00:00
}
#endif /* COMPAT_43 */
1994-05-24 10:09:53 +00:00
int
wait4(p, uap)
1994-05-24 10:09:53 +00:00
struct proc *p;
struct wait_args *uap;
{
return (wait1(p, uap, 0));
1994-05-24 10:09:53 +00:00
}
static int
wait1(q, uap, compat)
register struct proc *q;
register struct wait_args /* {
int pid;
int *status;
int options;
struct rusage *rusage;
} */ *uap;
int compat;
1994-05-24 10:09:53 +00:00
{
register int nfound;
register struct proc *p, *t;
int status, error;
1994-05-24 10:09:53 +00:00
if (uap->pid == 0)
uap->pid = -q->p_pgid;
if (uap->options &~ (WUNTRACED|WNOHANG|WLINUXCLONE))
1994-05-24 10:09:53 +00:00
return (EINVAL);
loop:
nfound = 0;
for (p = q->p_children.lh_first; p != 0; p = p->p_sibling.le_next) {
1994-05-24 10:09:53 +00:00
if (uap->pid != WAIT_ANY &&
p->p_pid != uap->pid && p->p_pgid != -uap->pid)
continue;
/* 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) ^ ((uap->options & WLINUXCLONE) != 0))
continue;
1994-05-24 10:09:53 +00:00
nfound++;
if (p->p_stat == SZOMB) {
/* charge childs scheduling cpu usage to parent */
if (curproc->p_pid != 1) {
curproc->p_estcpu = min(curproc->p_estcpu +
p->p_estcpu, UCHAR_MAX);
}
q->p_retval[0] = p->p_pid;
#ifdef COMPAT_43
if (compat)
q->p_retval[1] = p->p_xstat;
1994-05-24 10:09:53 +00:00
else
#endif
if (uap->status) {
status = p->p_xstat; /* convert to int */
if ((error = copyout((caddr_t)&status,
(caddr_t)uap->status, sizeof(status))))
1994-05-24 10:09:53 +00:00
return (error);
}
if (uap->rusage && (error = copyout((caddr_t)p->p_ru,
(caddr_t)uap->rusage, sizeof (struct rusage))))
return (error);
/*
* If we got the child via a ptrace 'attach',
* we need to give it back to the old parent.
*/
if (p->p_oppid && (t = pfind(p->p_oppid))) {
p->p_oppid = 0;
proc_reparent(p, t);
psignal(t, SIGCHLD);
wakeup((caddr_t)t);
return (0);
}
p->p_xstat = 0;
ruadd(&q->p_stats->p_cru, p->p_ru);
FREE(p->p_ru, M_ZOMBIE);
p->p_ru = NULL;
1994-05-24 10:09:53 +00:00
/*
* Decrement the count of procs running with this uid.
*/
(void)chgproccnt(p->p_cred->p_ruid, -1);
/*
* Release reference to text vnode
*/
if (p->p_textvp)
vrele(p->p_textvp);
1994-05-24 10:09:53 +00:00
/*
* Free up credentials.
*/
if (--p->p_cred->p_refcnt == 0) {
crfree(p->p_cred->pc_ucred);
FREE(p->p_cred, M_SUBPROC);
p->p_cred = NULL;
1994-05-24 10:09:53 +00:00
}
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
/*
* Destroy empty prisons
*/
1999-08-27 19:47:41 +00:00
if (p->p_prison && !--p->p_prison->pr_ref) {
if (p->p_prison->pr_linux != NULL)
FREE(p->p_prison->pr_linux, M_PRISON);
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
FREE(p->p_prison, M_PRISON);
1999-08-27 19:47:41 +00:00
}
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
1994-05-24 10:09:53 +00:00
/*
* Finally finished with old proc entry.
* Unlink it from its process group and free it.
*/
leavepgrp(p);
LIST_REMOVE(p, p_list); /* off zombproc */
LIST_REMOVE(p, p_sibling);
1994-05-24 10:09:53 +00:00
if (--p->p_procsig->ps_refcnt == 0) {
if (p->p_sigacts != &p->p_addr->u_sigacts)
FREE(p->p_sigacts, M_SUBPROC);
FREE(p->p_procsig, M_SUBPROC);
p->p_procsig = NULL;
}
1994-05-24 10:09:53 +00:00
/*
* Give machine-dependent layer a chance
* to free anything that cpu_exit couldn't
* release while still running in process context.
*/
cpu_wait(p);
VM level code cleanups. 1) Start using TSM. Struct procs continue to point to upages structure, after being freed. Struct vmspace continues to point to pte object and kva space for kstack. u_map is now superfluous. 2) vm_map's don't need to be reference counted. They always exist either in the kernel or in a vmspace. The vmspaces are managed by reference counts. 3) Remove the "wired" vm_map nonsense. 4) No need to keep a cache of kernel stack kva's. 5) Get rid of strange looking ++var, and change to var++. 6) Change more data structures to use our "zone" allocator. Added struct proc, struct vmspace and struct vnode. This saves a significant amount of kva space and physical memory. Additionally, this enables TSM for the zone managed memory. 7) Keep ioopt disabled for now. 8) Remove the now bogus "single use" map concept. 9) Use generation counts or id's for data structures residing in TSM, where it allows us to avoid unneeded restart overhead during traversals, where blocking might occur. 10) Account better for memory deficits, so the pageout daemon will be able to make enough memory available (experimental.) 11) Fix some vnode locking problems. (From Tor, I think.) 12) Add a check in ufs_lookup, to avoid lots of unneeded calls to bcmp. (experimental.) 13) Significantly shrink, cleanup, and make slightly faster the vm_fault.c code. Use generation counts, get rid of unneded collpase operations, and clean up the cluster code. 14) Make vm_zone more suitable for TSM. This commit is partially as a result of discussions and contributions from other people, including DG, Tor Egge, PHK, and probably others that I have forgotten to attribute (so let me know, if I forgot.) This is not the infamous, final cleanup of the vnode stuff, but a necessary step. Vnode mgmt should be correct, but things might still change, and there is still some missing stuff (like ioopt, and physical backing of non-merged cache files, debugging of layering concepts.)
1998-01-22 17:30:44 +00:00
zfree(proc_zone, p);
1994-05-24 10:09:53 +00:00
nprocs--;
return (0);
}
if (p->p_stat == SSTOP && (p->p_flag & P_WAITED) == 0 &&
(p->p_flag & P_TRACED || uap->options & WUNTRACED)) {
p->p_flag |= P_WAITED;
q->p_retval[0] = p->p_pid;
#ifdef COMPAT_43
if (compat) {
q->p_retval[1] = W_STOPCODE(p->p_xstat);
1994-05-24 10:09:53 +00:00
error = 0;
} else
#endif
if (uap->status) {
status = W_STOPCODE(p->p_xstat);
error = copyout((caddr_t)&status,
(caddr_t)uap->status, sizeof(status));
} else
error = 0;
return (error);
}
}
if (nfound == 0)
return (ECHILD);
if (uap->options & WNOHANG) {
q->p_retval[0] = 0;
1994-05-24 10:09:53 +00:00
return (0);
}
if ((error = tsleep((caddr_t)q, PWAIT | PCATCH, "wait", 0)))
1994-05-24 10:09:53 +00:00
return (error);
goto loop;
}
/*
* make process 'parent' the new parent of process 'child'.
*/
void
proc_reparent(child, parent)
register struct proc *child;
register struct proc *parent;
{
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;
}
/*
* The next two functions are to handle adding/deleting items on the
* exit callout list
*
* at_exit():
* Take the arguments given and put them onto the exit callout list,
* However first make sure that it's not already there.
* returns 0 on success.
*/
int
1997-08-26 00:15:04 +00:00
at_exit(function)
exitlist_fn function;
{
ele_p ep;
/* Be noisy if the programmer has lost track of things */
if (rm_at_exit(function))
printf("exit callout entry already present\n");
ep = malloc(sizeof(*ep), M_TEMP, M_NOWAIT);
if (ep == NULL)
return (ENOMEM);
ep->next = exit_list;
ep->function = function;
exit_list = ep;
return (0);
}
/*
* Scan the exit callout list for the given items and remove them.
* Returns the number of items removed.
* Logically this can only be 0 or 1.
*/
int
1997-08-26 00:15:04 +00:00
rm_at_exit(function)
exitlist_fn function;
{
ele_p *epp, ep;
int count;
count = 0;
epp = &exit_list;
ep = *epp;
while (ep) {
if (ep->function == function) {
*epp = ep->next;
free(ep, M_TEMP);
count++;
} else {
epp = &ep->next;
}
ep = *epp;
}
return (count);
}
void check_sigacts (void)
{
struct proc *p = curproc;
struct sigacts *pss;
int s;
if (p->p_procsig->ps_refcnt == 1 &&
p->p_sigacts != &p->p_addr->u_sigacts) {
pss = p->p_sigacts;
s = splhigh();
p->p_addr->u_sigacts = *pss;
p->p_sigacts = &p->p_addr->u_sigacts;
splx(s);
FREE(pss, M_SUBPROC);
}
}