2005-01-06 23:35:40 +00:00
|
|
|
/*-
|
2017-11-20 19:43:44 +00:00
|
|
|
* SPDX-License-Identifier: BSD-3-Clause
|
|
|
|
*
|
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.
|
2016-09-15 13:16:20 +00:00
|
|
|
* 3. Neither the name of the University nor the names of its contributors
|
1994-05-24 10:09:53 +00:00
|
|
|
* 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_fork.c 8.6 (Berkeley) 4/8/94
|
|
|
|
*/
|
|
|
|
|
2003-06-11 00:56:59 +00:00
|
|
|
#include <sys/cdefs.h>
|
|
|
|
__FBSDID("$FreeBSD$");
|
|
|
|
|
1996-01-03 21:42:35 +00:00
|
|
|
#include "opt_ktrace.h"
|
2009-09-01 11:41:51 +00:00
|
|
|
#include "opt_kstack_pages.h"
|
1996-01-03 21:42:35 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/param.h>
|
|
|
|
#include <sys/systm.h>
|
2018-12-07 12:22:32 +00:00
|
|
|
#include <sys/bitstring.h>
|
1995-11-12 06:43:28 +00:00
|
|
|
#include <sys/sysproto.h>
|
2003-03-24 21:15:35 +00:00
|
|
|
#include <sys/eventhandler.h>
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
#include <sys/fcntl.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/filedesc.h>
|
2009-05-27 14:11:23 +00:00
|
|
|
#include <sys/jail.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/kernel.h>
|
2003-08-15 21:29:06 +00:00
|
|
|
#include <sys/kthread.h>
|
1997-04-26 15:59:50 +00:00
|
|
|
#include <sys/sysctl.h>
|
Rework the witness code to work with sx locks as well as mutexes.
- Introduce lock classes and lock objects. Each lock class specifies a
name and set of flags (or properties) shared by all locks of a given
type. Currently there are three lock classes: spin mutexes, sleep
mutexes, and sx locks. A lock object specifies properties of an
additional lock along with a lock name and all of the extra stuff needed
to make witness work with a given lock. This abstract lock stuff is
defined in sys/lock.h. The lockmgr constants, types, and prototypes have
been moved to sys/lockmgr.h. For temporary backwards compatability,
sys/lock.h includes sys/lockmgr.h.
- Replace proc->p_spinlocks with a per-CPU list, PCPU(spinlocks), of spin
locks held. By making this per-cpu, we do not have to jump through
magic hoops to deal with sched_lock changing ownership during context
switches.
- Replace proc->p_heldmtx, formerly a list of held sleep mutexes, with
proc->p_sleeplocks, which is a list of held sleep locks including sleep
mutexes and sx locks.
- Add helper macros for logging lock events via the KTR_LOCK KTR logging
level so that the log messages are consistent.
- Add some new flags that can be passed to mtx_init():
- MTX_NOWITNESS - specifies that this lock should be ignored by witness.
This is used for the mutex that blocks a sx lock for example.
- MTX_QUIET - this is not new, but you can pass this to mtx_init() now
and no events will be logged for this lock, so that one doesn't have
to change all the individual mtx_lock/unlock() operations.
- All lock objects maintain an initialized flag. Use this flag to export
a mtx_initialized() macro that can be safely called from drivers. Also,
we on longer walk the all_mtx list if MUTEX_DEBUG is defined as witness
performs the corresponding checks using the initialized flag.
- The lock order reversal messages have been improved to output slightly
more accurate file and line numbers.
2001-03-28 09:03:24 +00:00
|
|
|
#include <sys/lock.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/malloc.h>
|
2000-10-20 07:58:15 +00:00
|
|
|
#include <sys/mutex.h>
|
2006-11-06 13:42:10 +00:00
|
|
|
#include <sys/priv.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/proc.h>
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
#include <sys/procdesc.h>
|
2015-10-06 19:29:05 +00:00
|
|
|
#include <sys/ptrace.h>
|
2011-03-29 17:47:25 +00:00
|
|
|
#include <sys/racct.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/resourcevar.h>
|
2002-10-12 05:32:24 +00:00
|
|
|
#include <sys/sched.h>
|
2001-01-24 10:47:14 +00:00
|
|
|
#include <sys/syscall.h>
|
2003-08-15 21:29:06 +00:00
|
|
|
#include <sys/vmmeter.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/vnode.h>
|
|
|
|
#include <sys/acct.h>
|
2000-09-07 01:33:02 +00:00
|
|
|
#include <sys/ktr.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/ktrace.h>
|
2018-01-14 15:01:25 +00:00
|
|
|
#include <sys/unistd.h>
|
2008-05-24 06:22:16 +00:00
|
|
|
#include <sys/sdt.h>
|
2001-03-07 02:30:39 +00:00
|
|
|
#include <sys/sx.h>
|
2011-03-08 19:01:45 +00:00
|
|
|
#include <sys/sysent.h>
|
2004-11-27 06:51:39 +00:00
|
|
|
#include <sys/signalvar.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2006-02-02 01:32:58 +00:00
|
|
|
#include <security/audit/audit.h>
|
2006-10-22 11:52:19 +00:00
|
|
|
#include <security/mac/mac_framework.h>
|
2006-02-02 01:32:58 +00:00
|
|
|
|
1994-10-10 01:00:49 +00:00
|
|
|
#include <vm/vm.h>
|
1996-02-23 18:49:25 +00:00
|
|
|
#include <vm/pmap.h>
|
|
|
|
#include <vm/vm_map.h>
|
1995-12-07 12:48:31 +00:00
|
|
|
#include <vm/vm_extern.h>
|
2002-03-20 04:09:59 +00:00
|
|
|
#include <vm/uma.h>
|
1994-10-10 01:00:49 +00:00
|
|
|
|
2008-05-24 06:22:16 +00:00
|
|
|
#ifdef KDTRACE_HOOKS
|
|
|
|
#include <sys/dtrace_bsd.h>
|
|
|
|
dtrace_fork_func_t dtrace_fasttrap_fork;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
SDT_PROVIDER_DECLARE(proc);
|
2015-12-16 23:39:27 +00:00
|
|
|
SDT_PROBE_DEFINE3(proc, , , create, "struct proc *", "struct proc *", "int");
|
1999-01-26 02:38:12 +00:00
|
|
|
|
1995-11-12 06:43:28 +00:00
|
|
|
#ifndef _SYS_SYSPROTO_H_
|
1995-10-08 00:06:22 +00:00
|
|
|
struct fork_args {
|
1996-08-22 03:50:33 +00:00
|
|
|
int dummy;
|
1995-10-08 00:06:22 +00:00
|
|
|
};
|
1995-11-12 06:43:28 +00:00
|
|
|
#endif
|
1995-10-08 00:06:22 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/* ARGSUSED */
|
1994-05-25 09:21:21 +00:00
|
|
|
int
|
2011-09-16 13:58:51 +00:00
|
|
|
sys_fork(struct thread *td, struct fork_args *uap)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
2016-02-04 04:22:18 +00:00
|
|
|
struct fork_req fr;
|
2016-02-04 04:25:30 +00:00
|
|
|
int error, pid;
|
1997-11-20 16:36:17 +00:00
|
|
|
|
2016-02-04 04:22:18 +00:00
|
|
|
bzero(&fr, sizeof(fr));
|
|
|
|
fr.fr_flags = RFFDG | RFPROC;
|
2016-02-04 04:25:30 +00:00
|
|
|
fr.fr_pidp = &pid;
|
2016-02-04 04:22:18 +00:00
|
|
|
error = fork1(td, &fr);
|
1999-06-30 15:33:41 +00:00
|
|
|
if (error == 0) {
|
2016-02-04 04:25:30 +00:00
|
|
|
td->td_retval[0] = pid;
|
2001-09-12 08:38:13 +00:00
|
|
|
td->td_retval[1] = 0;
|
1999-06-30 15:33:41 +00:00
|
|
|
}
|
2003-08-15 21:29:06 +00:00
|
|
|
return (error);
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
|
|
|
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
/* ARGUSED */
|
|
|
|
int
|
2016-02-06 09:01:03 +00:00
|
|
|
sys_pdfork(struct thread *td, struct pdfork_args *uap)
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
{
|
2016-02-04 04:22:18 +00:00
|
|
|
struct fork_req fr;
|
2016-02-04 04:25:30 +00:00
|
|
|
int error, fd, pid;
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
|
2016-02-04 04:22:18 +00:00
|
|
|
bzero(&fr, sizeof(fr));
|
|
|
|
fr.fr_flags = RFFDG | RFPROC | RFPROCDESC;
|
2016-02-04 04:25:30 +00:00
|
|
|
fr.fr_pidp = &pid;
|
2016-02-04 04:22:18 +00:00
|
|
|
fr.fr_pd_fd = &fd;
|
|
|
|
fr.fr_pd_flags = uap->flags;
|
Add BSM record conversion for a number of syscalls:
- thr_kill(2) and thr_exit(2) generally (no argument auditing here.
- A set of syscalls for the process descriptor family, specifically:
pdfork(2), pdgetpid(2) and pdkill(2)
For these syscalls, audit the file descriptor. In the case of pdfork(2)
a pointer to an integer (file descriptor) is passed in as an argument.
We audit the post initialized file descriptor (not the random garbage
that would have been passed in). We will also audit the child process
which was created from the fork operation (similar to what is done for
the fork(2) syscall).
pdkill(2) we audit the signal value and fd, and finally pdgetpid(2)
just the file descriptor:
- Following is a sample of the produced audit trails:
header,111,11,pdfork(2),0,Sat May 16 03:07:50 2020, + 394 msec
argument,0,0x39d,child PID
argument,2,0x2,flags
argument,1,0x8,fd
subject,root,root,0,root,0,924,0,0,0.0.0.0
return,success,925
header,79,11,pdgetpid(2),0,Sat May 16 03:07:50 2020, + 394 msec
argument,1,0x8,fd
subject,root,root,0,root,0,924,0,0,0.0.0.0
return,success,0
trailer,79
header,135,11,pdkill(2),0,Sat May 16 03:07:50 2020, + 395 msec
argument,1,0x8,fd
argument,2,0xf,signal
process_ex,root,root,0,root,0,925,0,0,0.0.0.0
subject,root,root,0,root,0,924,0,0,0.0.0.0
return,success,0
trailer,135
MFC after: 1 week
2020-05-16 03:45:15 +00:00
|
|
|
AUDIT_ARG_FFLAGS(uap->flags);
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
/*
|
|
|
|
* It is necessary to return fd by reference because 0 is a valid file
|
|
|
|
* descriptor number, and the child needs to be able to distinguish
|
|
|
|
* itself from the parent using the return value.
|
|
|
|
*/
|
2016-02-04 04:22:18 +00:00
|
|
|
error = fork1(td, &fr);
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
if (error == 0) {
|
2016-02-04 04:25:30 +00:00
|
|
|
td->td_retval[0] = pid;
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
td->td_retval[1] = 0;
|
|
|
|
error = copyout(&fd, uap->fdp, sizeof(fd));
|
|
|
|
}
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/* ARGSUSED */
|
1994-05-25 09:21:21 +00:00
|
|
|
int
|
2011-09-16 13:58:51 +00:00
|
|
|
sys_vfork(struct thread *td, struct vfork_args *uap)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
2016-02-04 04:22:18 +00:00
|
|
|
struct fork_req fr;
|
2016-02-04 04:25:30 +00:00
|
|
|
int error, pid;
|
1997-11-20 16:36:17 +00:00
|
|
|
|
2016-02-04 04:22:18 +00:00
|
|
|
bzero(&fr, sizeof(fr));
|
|
|
|
fr.fr_flags = RFFDG | RFPROC | RFPPWAIT | RFMEM;
|
2016-02-04 04:25:30 +00:00
|
|
|
fr.fr_pidp = &pid;
|
2016-02-04 04:22:18 +00:00
|
|
|
error = fork1(td, &fr);
|
1999-06-30 15:33:41 +00:00
|
|
|
if (error == 0) {
|
2016-02-04 04:25:30 +00:00
|
|
|
td->td_retval[0] = pid;
|
2001-09-12 08:38:13 +00:00
|
|
|
td->td_retval[1] = 0;
|
1999-06-30 15:33:41 +00:00
|
|
|
}
|
2003-08-15 21:29:06 +00:00
|
|
|
return (error);
|
1996-02-23 18:49:25 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
|
1996-02-23 18:49:25 +00:00
|
|
|
int
|
2011-09-16 13:58:51 +00:00
|
|
|
sys_rfork(struct thread *td, struct rfork_args *uap)
|
1996-02-23 18:49:25 +00:00
|
|
|
{
|
2016-02-04 04:22:18 +00:00
|
|
|
struct fork_req fr;
|
2016-02-04 04:25:30 +00:00
|
|
|
int error, pid;
|
1997-11-20 16:36:17 +00:00
|
|
|
|
2004-03-04 09:56:29 +00:00
|
|
|
/* Don't allow kernel-only flags. */
|
2001-12-19 00:53:23 +00:00
|
|
|
if ((uap->flags & RFKERNELONLY) != 0)
|
|
|
|
return (EINVAL);
|
2019-09-25 19:20:41 +00:00
|
|
|
/* RFSPAWN must not appear with others */
|
|
|
|
if ((uap->flags & RFSPAWN) != 0 && uap->flags != RFSPAWN)
|
|
|
|
return (EINVAL);
|
2004-03-04 09:56:29 +00:00
|
|
|
|
2009-06-27 13:58:44 +00:00
|
|
|
AUDIT_ARG_FFLAGS(uap->flags);
|
2016-02-04 04:22:18 +00:00
|
|
|
bzero(&fr, sizeof(fr));
|
2019-09-25 19:20:41 +00:00
|
|
|
if ((uap->flags & RFSPAWN) != 0) {
|
|
|
|
fr.fr_flags = RFFDG | RFPROC | RFPPWAIT | RFMEM;
|
|
|
|
fr.fr_flags2 = FR2_DROPSIG_CAUGHT;
|
|
|
|
} else {
|
|
|
|
fr.fr_flags = uap->flags;
|
|
|
|
}
|
2016-02-04 04:25:30 +00:00
|
|
|
fr.fr_pidp = &pid;
|
2016-02-04 04:22:18 +00:00
|
|
|
error = fork1(td, &fr);
|
1999-06-30 15:33:41 +00:00
|
|
|
if (error == 0) {
|
2016-02-04 04:25:30 +00:00
|
|
|
td->td_retval[0] = pid;
|
2001-09-12 08:38:13 +00:00
|
|
|
td->td_retval[1] = 0;
|
1999-06-30 15:33:41 +00:00
|
|
|
}
|
2003-08-15 21:29:06 +00:00
|
|
|
return (error);
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
|
|
|
|
2019-05-04 19:04:17 +00:00
|
|
|
int __exclusive_cache_line nprocs = 1; /* process 0 */
|
2001-06-11 21:54:19 +00:00
|
|
|
int lastpid = 0;
|
2018-01-14 15:01:25 +00:00
|
|
|
SYSCTL_INT(_kern, OID_AUTO, lastpid, CTLFLAG_RD, &lastpid, 0,
|
2001-02-12 17:59:01 +00:00
|
|
|
"Last used PID");
|
1994-05-24 10:09:53 +00:00
|
|
|
|
1999-12-06 11:13:50 +00:00
|
|
|
/*
|
2001-06-11 21:54:19 +00:00
|
|
|
* Random component to lastpid generation. We mix in a random factor to make
|
1999-12-06 11:13:50 +00:00
|
|
|
* it a little harder to predict. We sanity check the modulus value to avoid
|
|
|
|
* doing it in critical paths. Don't let it be too small or we pointlessly
|
|
|
|
* waste randomness entropy, and don't let it be impossibly large. Using a
|
|
|
|
* modulus that is too big causes a LOT more process table scans and slows
|
|
|
|
* down fork processing as the pidchecked caching is defeated.
|
|
|
|
*/
|
1999-11-28 17:51:09 +00:00
|
|
|
static int randompid = 0;
|
1999-12-06 11:13:50 +00:00
|
|
|
|
|
|
|
static int
|
2000-07-04 11:25:35 +00:00
|
|
|
sysctl_kern_randompid(SYSCTL_HANDLER_ARGS)
|
1999-12-06 11:13:50 +00:00
|
|
|
{
|
2000-12-04 10:23:29 +00:00
|
|
|
int error, pid;
|
|
|
|
|
2004-02-26 00:27:04 +00:00
|
|
|
error = sysctl_wire_old_buffer(req, sizeof(int));
|
|
|
|
if (error != 0)
|
|
|
|
return(error);
|
2002-05-02 15:13:45 +00:00
|
|
|
sx_xlock(&allproc_lock);
|
2000-12-04 10:23:29 +00:00
|
|
|
pid = randompid;
|
|
|
|
error = sysctl_handle_int(oidp, &pid, 0, req);
|
2002-05-02 15:13:45 +00:00
|
|
|
if (error == 0 && req->newptr != NULL) {
|
2017-09-10 15:01:29 +00:00
|
|
|
if (pid == 0)
|
|
|
|
randompid = 0;
|
|
|
|
else if (pid == 1)
|
|
|
|
/* generate a random PID modulus between 100 and 1123 */
|
|
|
|
randompid = 100 + arc4random() % 1024;
|
|
|
|
else if (pid < 0 || pid > pid_max - 100)
|
|
|
|
/* out of range */
|
|
|
|
randompid = pid_max - 100;
|
2018-01-14 15:01:25 +00:00
|
|
|
else if (pid < 100)
|
2017-09-10 15:01:29 +00:00
|
|
|
/* Make it reasonable */
|
|
|
|
randompid = 100;
|
|
|
|
else
|
|
|
|
randompid = pid;
|
2002-05-02 15:13:45 +00:00
|
|
|
}
|
|
|
|
sx_xunlock(&allproc_lock);
|
2000-12-04 10:23:29 +00:00
|
|
|
return (error);
|
1999-12-06 11:13:50 +00:00
|
|
|
}
|
|
|
|
|
2020-02-26 14:26:36 +00:00
|
|
|
SYSCTL_PROC(_kern, OID_AUTO, randompid,
|
|
|
|
CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, 0, 0,
|
|
|
|
sysctl_kern_randompid, "I",
|
|
|
|
"Random PID modulus. Special values: 0: disable, 1: choose random value");
|
1999-11-28 17:51:09 +00:00
|
|
|
|
2018-12-07 12:22:32 +00:00
|
|
|
extern bitstr_t proc_id_pidmap;
|
|
|
|
extern bitstr_t proc_id_grpidmap;
|
|
|
|
extern bitstr_t proc_id_sessidmap;
|
|
|
|
extern bitstr_t proc_id_reapmap;
|
|
|
|
|
2018-12-19 20:29:52 +00:00
|
|
|
/*
|
|
|
|
* Find an unused process ID
|
|
|
|
*
|
|
|
|
* If RFHIGHPID is set (used during system boot), do not allocate
|
|
|
|
* low-numbered pids.
|
|
|
|
*/
|
2010-12-10 08:33:56 +00:00
|
|
|
static int
|
|
|
|
fork_findpid(int flags)
|
|
|
|
{
|
2018-12-07 12:22:32 +00:00
|
|
|
pid_t result;
|
2019-08-17 18:19:49 +00:00
|
|
|
int trypid, random;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Avoid calling arc4random with procid_lock held.
|
|
|
|
*/
|
|
|
|
random = 0;
|
|
|
|
if (__predict_false(randompid))
|
|
|
|
random = arc4random() % randompid;
|
|
|
|
|
|
|
|
mtx_lock(&procid_lock);
|
2010-12-10 08:33:56 +00:00
|
|
|
|
|
|
|
trypid = lastpid + 1;
|
|
|
|
if (flags & RFHIGHPID) {
|
|
|
|
if (trypid < 10)
|
|
|
|
trypid = 10;
|
|
|
|
} else {
|
2019-08-17 18:19:49 +00:00
|
|
|
trypid += random;
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
retry:
|
2019-08-17 17:42:01 +00:00
|
|
|
if (trypid >= pid_max)
|
|
|
|
trypid = 2;
|
2010-12-10 08:33:56 +00:00
|
|
|
|
2018-12-07 12:22:32 +00:00
|
|
|
bit_ffc_at(&proc_id_pidmap, trypid, pid_max, &result);
|
2018-12-08 10:22:12 +00:00
|
|
|
if (result == -1) {
|
2019-08-17 17:42:01 +00:00
|
|
|
KASSERT(trypid != 2, ("unexpectedly ran out of IDs"));
|
|
|
|
trypid = 2;
|
2018-12-07 12:22:32 +00:00
|
|
|
goto retry;
|
2018-12-08 10:22:12 +00:00
|
|
|
}
|
2018-12-07 12:22:32 +00:00
|
|
|
if (bit_test(&proc_id_grpidmap, result) ||
|
|
|
|
bit_test(&proc_id_sessidmap, result) ||
|
|
|
|
bit_test(&proc_id_reapmap, result)) {
|
2018-12-19 20:29:52 +00:00
|
|
|
trypid = result + 1;
|
2018-12-07 12:22:32 +00:00
|
|
|
goto retry;
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* RFHIGHPID does not mess with the lastpid counter during boot.
|
|
|
|
*/
|
2018-12-07 12:22:32 +00:00
|
|
|
if ((flags & RFHIGHPID) == 0)
|
|
|
|
lastpid = result;
|
2010-12-10 08:33:56 +00:00
|
|
|
|
2018-12-07 12:22:32 +00:00
|
|
|
bit_set(&proc_id_pidmap, result);
|
|
|
|
mtx_unlock(&procid_lock);
|
2018-11-29 02:52:08 +00:00
|
|
|
|
2018-12-07 12:22:32 +00:00
|
|
|
return (result);
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
|
2010-12-06 16:39:18 +00:00
|
|
|
static int
|
2011-01-02 12:16:57 +00:00
|
|
|
fork_norfproc(struct thread *td, int flags)
|
2010-12-06 16:39:18 +00:00
|
|
|
{
|
|
|
|
int error;
|
|
|
|
struct proc *p1;
|
|
|
|
|
2010-12-06 19:15:38 +00:00
|
|
|
KASSERT((flags & RFPROC) == 0,
|
|
|
|
("fork_norfproc called with RFPROC set"));
|
2010-12-06 16:39:18 +00:00
|
|
|
p1 = td->td_proc;
|
|
|
|
|
|
|
|
if (((p1->p_flag & (P_HADTHREADS|P_SYSTEM)) == P_HADTHREADS) &&
|
|
|
|
(flags & (RFCFDG | RFFDG))) {
|
|
|
|
PROC_LOCK(p1);
|
2014-12-13 16:18:29 +00:00
|
|
|
if (thread_single(p1, SINGLE_BOUNDARY)) {
|
2010-12-06 16:39:18 +00:00
|
|
|
PROC_UNLOCK(p1);
|
|
|
|
return (ERESTART);
|
|
|
|
}
|
|
|
|
PROC_UNLOCK(p1);
|
|
|
|
}
|
|
|
|
|
|
|
|
error = vm_forkproc(td, NULL, NULL, NULL, flags);
|
|
|
|
if (error)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Close all file descriptors.
|
|
|
|
*/
|
|
|
|
if (flags & RFCFDG) {
|
|
|
|
struct filedesc *fdtmp;
|
2020-07-15 10:24:04 +00:00
|
|
|
fdtmp = fdinit(td->td_proc->p_fd, false, NULL);
|
Merge Capsicum overhaul:
- Capability is no longer separate descriptor type. Now every descriptor
has set of its own capability rights.
- The cap_new(2) system call is left, but it is no longer documented and
should not be used in new code.
- The new syscall cap_rights_limit(2) should be used instead of
cap_new(2), which limits capability rights of the given descriptor
without creating a new one.
- The cap_getrights(2) syscall is renamed to cap_rights_get(2).
- If CAP_IOCTL capability right is present we can further reduce allowed
ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed
ioctls can be retrived with cap_ioctls_get(2) syscall.
- If CAP_FCNTL capability right is present we can further reduce fcntls
that can be used with the new cap_fcntls_limit(2) syscall and retrive
them with cap_fcntls_get(2).
- To support ioctl and fcntl white-listing the filedesc structure was
heavly modified.
- The audit subsystem, kdump and procstat tools were updated to
recognize new syscalls.
- Capability rights were revised and eventhough I tried hard to provide
backward API and ABI compatibility there are some incompatible changes
that are described in detail below:
CAP_CREATE old behaviour:
- Allow for openat(2)+O_CREAT.
- Allow for linkat(2).
- Allow for symlinkat(2).
CAP_CREATE new behaviour:
- Allow for openat(2)+O_CREAT.
Added CAP_LINKAT:
- Allow for linkat(2). ABI: Reuses CAP_RMDIR bit.
- Allow to be target for renameat(2).
Added CAP_SYMLINKAT:
- Allow for symlinkat(2).
Removed CAP_DELETE. Old behaviour:
- Allow for unlinkat(2) when removing non-directory object.
- Allow to be source for renameat(2).
Removed CAP_RMDIR. Old behaviour:
- Allow for unlinkat(2) when removing directory.
Added CAP_RENAMEAT:
- Required for source directory for the renameat(2) syscall.
Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR):
- Allow for unlinkat(2) on any object.
- Required if target of renameat(2) exists and will be removed by this
call.
Removed CAP_MAPEXEC.
CAP_MMAP old behaviour:
- Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and
PROT_WRITE.
CAP_MMAP new behaviour:
- Allow for mmap(2)+PROT_NONE.
Added CAP_MMAP_R:
- Allow for mmap(PROT_READ).
Added CAP_MMAP_W:
- Allow for mmap(PROT_WRITE).
Added CAP_MMAP_X:
- Allow for mmap(PROT_EXEC).
Added CAP_MMAP_RW:
- Allow for mmap(PROT_READ | PROT_WRITE).
Added CAP_MMAP_RX:
- Allow for mmap(PROT_READ | PROT_EXEC).
Added CAP_MMAP_WX:
- Allow for mmap(PROT_WRITE | PROT_EXEC).
Added CAP_MMAP_RWX:
- Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC).
Renamed CAP_MKDIR to CAP_MKDIRAT.
Renamed CAP_MKFIFO to CAP_MKFIFOAT.
Renamed CAP_MKNODE to CAP_MKNODEAT.
CAP_READ old behaviour:
- Allow pread(2).
- Disallow read(2), readv(2) (if there is no CAP_SEEK).
CAP_READ new behaviour:
- Allow read(2), readv(2).
- Disallow pread(2) (CAP_SEEK was also required).
CAP_WRITE old behaviour:
- Allow pwrite(2).
- Disallow write(2), writev(2) (if there is no CAP_SEEK).
CAP_WRITE new behaviour:
- Allow write(2), writev(2).
- Disallow pwrite(2) (CAP_SEEK was also required).
Added convinient defines:
#define CAP_PREAD (CAP_SEEK | CAP_READ)
#define CAP_PWRITE (CAP_SEEK | CAP_WRITE)
#define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ)
#define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE)
#define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL)
#define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W)
#define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X)
#define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X)
#define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X)
#define CAP_RECV CAP_READ
#define CAP_SEND CAP_WRITE
#define CAP_SOCK_CLIENT \
(CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \
CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN)
#define CAP_SOCK_SERVER \
(CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \
CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \
CAP_SETSOCKOPT | CAP_SHUTDOWN)
Added defines for backward API compatibility:
#define CAP_MAPEXEC CAP_MMAP_X
#define CAP_DELETE CAP_UNLINKAT
#define CAP_MKDIR CAP_MKDIRAT
#define CAP_RMDIR CAP_UNLINKAT
#define CAP_MKFIFO CAP_MKFIFOAT
#define CAP_MKNOD CAP_MKNODAT
#define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER)
Sponsored by: The FreeBSD Foundation
Reviewed by: Christoph Mallon <christoph.mallon@gmx.de>
Many aspects discussed with: rwatson, benl, jonathan
ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
|
|
|
fdescfree(td);
|
2010-12-06 16:39:18 +00:00
|
|
|
p1->p_fd = fdtmp;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Unshare file descriptors (from parent).
|
|
|
|
*/
|
2014-06-28 05:41:53 +00:00
|
|
|
if (flags & RFFDG)
|
|
|
|
fdunshare(td);
|
2010-12-06 16:39:18 +00:00
|
|
|
|
|
|
|
fail:
|
|
|
|
if (((p1->p_flag & (P_HADTHREADS|P_SYSTEM)) == P_HADTHREADS) &&
|
2010-12-06 16:45:36 +00:00
|
|
|
(flags & (RFCFDG | RFFDG))) {
|
2010-12-06 16:39:18 +00:00
|
|
|
PROC_LOCK(p1);
|
2014-12-13 16:18:29 +00:00
|
|
|
thread_single_end(p1, SINGLE_BOUNDARY);
|
2010-12-06 16:39:18 +00:00
|
|
|
PROC_UNLOCK(p1);
|
|
|
|
}
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
|
2010-12-10 08:33:56 +00:00
|
|
|
static void
|
2016-02-04 04:25:30 +00:00
|
|
|
do_fork(struct thread *td, struct fork_req *fr, struct proc *p2, struct thread *td2,
|
|
|
|
struct vmspace *vm2, struct file *fp_procdesc)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
2010-12-10 08:33:56 +00:00
|
|
|
struct proc *p1, *pptr;
|
2001-03-07 05:21:47 +00:00
|
|
|
struct filedesc *fd;
|
2003-06-02 16:05:32 +00:00
|
|
|
struct filedesc_to_leader *fdtol;
|
2002-05-02 15:13:45 +00:00
|
|
|
struct sigacts *newsigacts;
|
1997-04-13 01:48:35 +00:00
|
|
|
|
2003-08-15 21:29:06 +00:00
|
|
|
p1 = td->td_proc;
|
|
|
|
|
Close race conditions between fork() and [sg]etpriority()'s
PRIO_USER case, possibly also other places that deferences
p_ucred.
In the past, we insert a new process into the allproc list right
after PID allocation, and release the allproc_lock sx. Because
most content in new proc's structure is not yet initialized,
this could lead to undefined result if we do not handle PRS_NEW
with care.
The problem with PRS_NEW state is that it does not provide fine
grained information about how much initialization is done for a
new process. By defination, after PRIO_USER setpriority(), all
processes that belongs to given user should have their nice value
set to the specified value. Therefore, if p_{start,end}copy
section was done for a PRS_NEW process, we can not safely ignore
it because p_nice is in this area. On the other hand, we should
be careful on PRS_NEW processes because we do not allow non-root
users to lower their nice values, and without a successful copy
of the copy section, we can get stale values that is inherted
from the uninitialized area of the process structure.
This commit tries to close the race condition by grabbing proc
mutex *before* we release allproc_lock xlock, and do copy as
well as zero immediately after the allproc_lock xunlock. This
guarantees that the new process would have its p_copy and p_zero
sections, as well as user credential informaion initialized. In
getpriority() case, instead of grabbing PROC_LOCK for a PRS_NEW
process, we just skip the process in question, because it does
not affect the final result of the call, as the p_nice value
would be copied from its parent, and we will see it during
allproc traverse.
Other potential solutions are still under evaluation.
Discussed with: davidxu, jhb, rwatson
PR: kern/108071
MFC after: 2 weeks
2007-02-26 03:38:09 +00:00
|
|
|
PROC_LOCK(p1);
|
|
|
|
bcopy(&p1->p_startcopy, &p2->p_startcopy,
|
|
|
|
__rangeof(struct proc, p_startcopy, p_endcopy));
|
2008-07-23 08:45:25 +00:00
|
|
|
pargs_hold(p2->p_args);
|
Close race conditions between fork() and [sg]etpriority()'s
PRIO_USER case, possibly also other places that deferences
p_ucred.
In the past, we insert a new process into the allproc list right
after PID allocation, and release the allproc_lock sx. Because
most content in new proc's structure is not yet initialized,
this could lead to undefined result if we do not handle PRS_NEW
with care.
The problem with PRS_NEW state is that it does not provide fine
grained information about how much initialization is done for a
new process. By defination, after PRIO_USER setpriority(), all
processes that belongs to given user should have their nice value
set to the specified value. Therefore, if p_{start,end}copy
section was done for a PRS_NEW process, we can not safely ignore
it because p_nice is in this area. On the other hand, we should
be careful on PRS_NEW processes because we do not allow non-root
users to lower their nice values, and without a successful copy
of the copy section, we can get stale values that is inherted
from the uninitialized area of the process structure.
This commit tries to close the race condition by grabbing proc
mutex *before* we release allproc_lock xlock, and do copy as
well as zero immediately after the allproc_lock xunlock. This
guarantees that the new process would have its p_copy and p_zero
sections, as well as user credential informaion initialized. In
getpriority() case, instead of grabbing PROC_LOCK for a PRS_NEW
process, we just skip the process in question, because it does
not affect the final result of the call, as the p_nice value
would be copied from its parent, and we will see it during
allproc traverse.
Other potential solutions are still under evaluation.
Discussed with: davidxu, jhb, rwatson
PR: kern/108071
MFC after: 2 weeks
2007-02-26 03:38:09 +00:00
|
|
|
PROC_UNLOCK(p1);
|
|
|
|
|
|
|
|
bzero(&p2->p_startzero,
|
|
|
|
__rangeof(struct proc, p_startzero, p_endzero));
|
|
|
|
|
2009-05-27 14:11:23 +00:00
|
|
|
/* Tell the prison that we exist. */
|
|
|
|
prison_proc_hold(p2->p_ucred->cr_prison);
|
MFp4:
Bring in updated jail support from bz_jail branch.
This enhances the current jail implementation to permit multiple
addresses per jail. In addtion to IPv4, IPv6 is supported as well.
Due to updated checks it is even possible to have jails without
an IP address at all, which basically gives one a chroot with
restricted process view, no networking,..
SCTP support was updated and supports IPv6 in jails as well.
Cpuset support permits jails to be bound to specific processor
sets after creation.
Jails can have an unrestricted (no duplicate protection, etc.) name
in addition to the hostname. The jail name cannot be changed from
within a jail and is considered to be used for management purposes
or as audit-token in the future.
DDB 'show jails' command was added to aid debugging.
Proper compat support permits 32bit jail binaries to be used on 64bit
systems to manage jails. Also backward compatibility was preserved where
possible: for jail v1 syscalls, as well as with user space management
utilities.
Both jail as well as prison version were updated for the new features.
A gap was intentionally left as the intermediate versions had been
used by various patches floating around the last years.
Bump __FreeBSD_version for the afore mentioned and in kernel changes.
Special thanks to:
- Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches
and Olivier Houchard (cognet) for initial single-IPv6 patches.
- Jeff Roberson (jeff) and Randall Stewart (rrs) for their
help, ideas and review on cpuset and SCTP support.
- Robert Watson (rwatson) for lots and lots of help, discussions,
suggestions and review of most of the patch at various stages.
- John Baldwin (jhb) for his help.
- Simon L. Nielsen (simon) as early adopter testing changes
on cluster machines as well as all the testers and people
who provided feedback the last months on freebsd-jail and
other channels.
- My employer, CK Software GmbH, for the support so I could work on this.
Reviewed by: (see above)
MFC after: 3 months (this is just so that I get the mail)
X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
|
|
|
|
2019-08-17 18:19:49 +00:00
|
|
|
p2->p_state = PRS_NEW; /* protect against others */
|
|
|
|
p2->p_pid = fork_findpid(fr->fr_flags);
|
|
|
|
AUDIT_ARG_PID(p2->p_pid);
|
|
|
|
|
|
|
|
sx_xlock(&allproc_lock);
|
|
|
|
LIST_INSERT_HEAD(&allproc, p2, p_list);
|
|
|
|
allproc_gen++;
|
|
|
|
sx_xunlock(&allproc_lock);
|
|
|
|
|
|
|
|
sx_xlock(PIDHASHLOCK(p2->p_pid));
|
|
|
|
LIST_INSERT_HEAD(PIDHASH(p2->p_pid), p2, p_hash);
|
|
|
|
sx_xunlock(PIDHASHLOCK(p2->p_pid));
|
Close race conditions between fork() and [sg]etpriority()'s
PRIO_USER case, possibly also other places that deferences
p_ucred.
In the past, we insert a new process into the allproc list right
after PID allocation, and release the allproc_lock sx. Because
most content in new proc's structure is not yet initialized,
this could lead to undefined result if we do not handle PRS_NEW
with care.
The problem with PRS_NEW state is that it does not provide fine
grained information about how much initialization is done for a
new process. By defination, after PRIO_USER setpriority(), all
processes that belongs to given user should have their nice value
set to the specified value. Therefore, if p_{start,end}copy
section was done for a PRS_NEW process, we can not safely ignore
it because p_nice is in this area. On the other hand, we should
be careful on PRS_NEW processes because we do not allow non-root
users to lower their nice values, and without a successful copy
of the copy section, we can get stale values that is inherted
from the uninitialized area of the process structure.
This commit tries to close the race condition by grabbing proc
mutex *before* we release allproc_lock xlock, and do copy as
well as zero immediately after the allproc_lock xunlock. This
guarantees that the new process would have its p_copy and p_zero
sections, as well as user credential informaion initialized. In
getpriority() case, instead of grabbing PROC_LOCK for a PRS_NEW
process, we just skip the process in question, because it does
not affect the final result of the call, as the p_nice value
would be copied from its parent, and we will see it during
allproc traverse.
Other potential solutions are still under evaluation.
Discussed with: davidxu, jhb, rwatson
PR: kern/108071
MFC after: 2 weeks
2007-02-26 03:38:09 +00:00
|
|
|
|
2018-02-20 02:18:30 +00:00
|
|
|
tidhash_add(td2);
|
|
|
|
|
2002-05-02 15:13:45 +00:00
|
|
|
/*
|
|
|
|
* Malloc things while we don't hold any locks.
|
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFSIGSHARE)
|
2002-05-02 15:13:45 +00:00
|
|
|
newsigacts = NULL;
|
- Merge struct procsig with struct sigacts.
- Move struct sigacts out of the u-area and malloc() it using the
M_SUBPROC malloc bucket.
- Add a small sigacts_*() API for managing sigacts structures: sigacts_alloc(),
sigacts_free(), sigacts_copy(), sigacts_share(), and sigacts_shared().
- Remove the p_sigignore, p_sigacts, and p_sigcatch macros.
- Add a mutex to struct sigacts that protects all the members of the struct.
- Add sigacts locking.
- Remove Giant from nosys(), kill(), killpg(), and kern_sigaction() now
that sigacts is locked.
- Several in-kernel functions such as psignal(), tdsignal(), trapsignal(),
and thread_stopped() are now MP safe.
Reviewed by: arch@
Approved by: re (rwatson)
2003-05-13 20:36:02 +00:00
|
|
|
else
|
|
|
|
newsigacts = sigacts_alloc();
|
2002-05-02 15:13:45 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Copy filedesc.
|
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFCFDG) {
|
2020-07-15 10:24:04 +00:00
|
|
|
fd = fdinit(p1->p_fd, false, NULL);
|
2003-06-02 16:05:32 +00:00
|
|
|
fdtol = NULL;
|
2016-02-04 04:25:30 +00:00
|
|
|
} else if (fr->fr_flags & RFFDG) {
|
2004-11-08 12:43:23 +00:00
|
|
|
fd = fdcopy(p1->p_fd);
|
2003-06-02 16:05:32 +00:00
|
|
|
fdtol = NULL;
|
|
|
|
} else {
|
2003-01-01 01:01:14 +00:00
|
|
|
fd = fdshare(p1->p_fd);
|
2003-06-02 16:05:32 +00:00
|
|
|
if (p1->p_fdtol == NULL)
|
2011-01-02 12:16:57 +00:00
|
|
|
p1->p_fdtol = filedesc_to_leader_alloc(NULL, NULL,
|
|
|
|
p1->p_leader);
|
2016-02-04 04:25:30 +00:00
|
|
|
if ((fr->fr_flags & RFTHREAD) != 0) {
|
2003-06-02 16:05:32 +00:00
|
|
|
/*
|
2011-01-02 12:16:57 +00:00
|
|
|
* Shared file descriptor table, and shared
|
|
|
|
* process leaders.
|
2003-06-02 16:05:32 +00:00
|
|
|
*/
|
|
|
|
fdtol = p1->p_fdtol;
|
Replace custom file descriptor array sleep lock constructed using a mutex
and flags with an sxlock. This leads to a significant and measurable
performance improvement as a result of access to shared locking for
frequent lookup operations, reduced general overhead, and reduced overhead
in the event of contention. All of these are imported for threaded
applications where simultaneous access to a shared file descriptor array
occurs frequently. Kris has reported 2x-4x transaction rate improvements
on 8-core MySQL benchmarks; smaller improvements can be expected for many
workloads as a result of reduced overhead.
- Generally eliminate the distinction between "fast" and regular
acquisisition of the filedesc lock; the plan is that they will now all
be fast. Change all locking instances to either shared or exclusive
locks.
- Correct a bug (pointed out by kib) in fdfree() where previously msleep()
was called without the mutex held; sx_sleep() is now always called with
the sxlock held exclusively.
- Universally hold the struct file lock over changes to struct file,
rather than the filedesc lock or no lock. Always update the f_ops
field last. A further memory barrier is required here in the future
(discussed with jhb).
- Improve locking and reference management in linux_at(), which fails to
properly acquire vnode references before using vnode pointers. Annotate
improper use of vn_fullpath(), which will be replaced at a future date.
In fcntl(), we conservatively acquire an exclusive lock, even though in
some cases a shared lock may be sufficient, which should be revisited.
The dropping of the filedesc lock in fdgrowtable() is no longer required
as the sxlock can be held over the sleep operation; we should consider
removing that (pointed out by attilio).
Tested by: kris
Discussed with: jhb, kris, attilio, jeff
2007-04-04 09:11:34 +00:00
|
|
|
FILEDESC_XLOCK(p1->p_fd);
|
2003-06-02 16:05:32 +00:00
|
|
|
fdtol->fdl_refcount++;
|
Replace custom file descriptor array sleep lock constructed using a mutex
and flags with an sxlock. This leads to a significant and measurable
performance improvement as a result of access to shared locking for
frequent lookup operations, reduced general overhead, and reduced overhead
in the event of contention. All of these are imported for threaded
applications where simultaneous access to a shared file descriptor array
occurs frequently. Kris has reported 2x-4x transaction rate improvements
on 8-core MySQL benchmarks; smaller improvements can be expected for many
workloads as a result of reduced overhead.
- Generally eliminate the distinction between "fast" and regular
acquisisition of the filedesc lock; the plan is that they will now all
be fast. Change all locking instances to either shared or exclusive
locks.
- Correct a bug (pointed out by kib) in fdfree() where previously msleep()
was called without the mutex held; sx_sleep() is now always called with
the sxlock held exclusively.
- Universally hold the struct file lock over changes to struct file,
rather than the filedesc lock or no lock. Always update the f_ops
field last. A further memory barrier is required here in the future
(discussed with jhb).
- Improve locking and reference management in linux_at(), which fails to
properly acquire vnode references before using vnode pointers. Annotate
improper use of vn_fullpath(), which will be replaced at a future date.
In fcntl(), we conservatively acquire an exclusive lock, even though in
some cases a shared lock may be sufficient, which should be revisited.
The dropping of the filedesc lock in fdgrowtable() is no longer required
as the sxlock can be held over the sleep operation; we should consider
removing that (pointed out by attilio).
Tested by: kris
Discussed with: jhb, kris, attilio, jeff
2007-04-04 09:11:34 +00:00
|
|
|
FILEDESC_XUNLOCK(p1->p_fd);
|
2003-06-02 16:05:32 +00:00
|
|
|
} else {
|
2018-01-14 15:01:25 +00:00
|
|
|
/*
|
2011-01-02 12:16:57 +00:00
|
|
|
* Shared file descriptor table, and different
|
|
|
|
* process leaders.
|
2003-06-02 16:05:32 +00:00
|
|
|
*/
|
|
|
|
fdtol = filedesc_to_leader_alloc(p1->p_fdtol,
|
2011-01-02 12:16:57 +00:00
|
|
|
p1->p_fd, p2);
|
2003-06-02 16:05:32 +00:00
|
|
|
}
|
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Make a proc table entry for the new process.
|
|
|
|
* Start by zeroing the section of proc that is zero-initialized,
|
|
|
|
* then copy the section that is copied directly from the parent.
|
|
|
|
*/
|
2002-10-02 07:44:29 +00:00
|
|
|
|
2003-05-01 21:16:38 +00:00
|
|
|
PROC_LOCK(p2);
|
|
|
|
PROC_LOCK(p1);
|
|
|
|
|
2002-02-07 20:58:47 +00:00
|
|
|
bzero(&td2->td_startzero,
|
2004-11-20 23:00:59 +00:00
|
|
|
__rangeof(struct thread, td_startzero, td_endzero));
|
2001-09-12 08:38:13 +00:00
|
|
|
|
2002-02-07 20:58:47 +00:00
|
|
|
bcopy(&td->td_startcopy, &td2->td_startcopy,
|
2004-11-20 23:00:59 +00:00
|
|
|
__rangeof(struct thread, td_startcopy, td_endcopy));
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2007-11-15 02:13:44 +00:00
|
|
|
bcopy(&p2->p_comm, &td2->td_name, sizeof(td2->td_name));
|
2004-01-03 02:02:26 +00:00
|
|
|
td2->td_sigstk = td->td_sigstk;
|
2007-09-17 05:31:39 +00:00
|
|
|
td2->td_flags = TDF_INMEM;
|
2010-12-09 02:42:02 +00:00
|
|
|
td2->td_lend_user_pri = PRI_MAX;
|
2004-01-03 02:02:26 +00:00
|
|
|
|
Change the curvnet variable from a global const struct vnet *,
previously always pointing to the default vnet context, to a
dynamically changing thread-local one. The currvnet context
should be set on entry to networking code via CURVNET_SET() macros,
and reverted to previous state via CURVNET_RESTORE(). Recursions
on curvnet are permitted, though strongly discuouraged.
This change should have no functional impact on nooptions VIMAGE
kernel builds, where CURVNET_* macros expand to whitespace.
The curthread->td_vnet (aka curvnet) variable's purpose is to be an
indicator of the vnet context in which the current network-related
operation takes place, in case we cannot deduce the current vnet
context from any other source, such as by looking at mbuf's
m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so
far curvnet has turned out to be an invaluable consistency checking
aid: it helps to catch cases when sockets, ifnets or any other
vnet-aware structures may have leaked from one vnet to another.
The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros
was a result of an empirical iterative process, whith an aim to
reduce recursions on CURVNET_SET() to a minimum, while still reducing
the scope of CURVNET_SET() to networking only operations - the
alternative would be calling CURVNET_SET() on each system call entry.
In general, curvnet has to be set in three typicall cases: when
processing socket-related requests from userspace or from within the
kernel; when processing inbound traffic flowing from device drivers
to upper layers of the networking stack, and when executing
timer-driven networking functions.
This change also introduces a DDB subcommand to show the list of all
vnet instances.
Approved by: julian (mentor)
2009-05-05 10:56:12 +00:00
|
|
|
#ifdef VIMAGE
|
|
|
|
td2->td_vnet = NULL;
|
|
|
|
td2->td_vnet_lpush = NULL;
|
|
|
|
#endif
|
|
|
|
|
2011-01-06 22:24:00 +00:00
|
|
|
/*
|
|
|
|
* Allow the scheduler to initialize the child.
|
|
|
|
*/
|
|
|
|
thread_lock(td);
|
|
|
|
sched_fork(td, td2);
|
|
|
|
thread_unlock(td);
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Duplicate sub-structures as needed.
|
|
|
|
* Increase reference counts on shared objects.
|
|
|
|
*/
|
2007-09-17 05:31:39 +00:00
|
|
|
p2->p_flag = P_INMEM;
|
Implement Address Space Layout Randomization (ASLR)
With this change, randomization can be enabled for all non-fixed
mappings. It means that the base address for the mapping is selected
with a guaranteed amount of entropy (bits). If the mapping was
requested to be superpage aligned, the randomization honours the
superpage attributes.
Although the value of ASLR is diminshing over time as exploit authors
work out simple ASLR bypass techniques, it elimintates the trivial
exploitation of certain vulnerabilities, at least in theory. This
implementation is relatively small and happens at the correct
architectural level. Also, it is not expected to introduce
regressions in existing cases when turned off (default for now), or
cause any significant maintaince burden.
The randomization is done on a best-effort basis - that is, the
allocator falls back to a first fit strategy if fragmentation prevents
entropy injection. It is trivial to implement a strong mode where
failure to guarantee the requested amount of entropy results in
mapping request failure, but I do not consider that to be usable.
I have not fine-tuned the amount of entropy injected right now. It is
only a quantitive change that will not change the implementation. The
current amount is controlled by aslr_pages_rnd.
To not spoil coalescing optimizations, to reduce the page table
fragmentation inherent to ASLR, and to keep the transient superpage
promotion for the malloced memory, locality clustering is implemented
for anonymous private mappings, which are automatically grouped until
fragmentation kicks in. The initial location for the anon group range
is, of course, randomized. This is controlled by vm.cluster_anon,
enabled by default.
The default mode keeps the sbrk area unpopulated by other mappings,
but this can be turned off, which gives much more breathing bits on
architectures with small address space, such as i386. This is tied
with the question of following an application's hint about the mmap(2)
base address. Testing shows that ignoring the hint does not affect the
function of common applications, but I would expect more demanding
code could break. By default sbrk is preserved and mmap hints are
satisfied, which can be changed by using the
kern.elf{32,64}.aslr.honor_sbrk sysctl.
ASLR is enabled on per-ABI basis, and currently it is only allowed on
FreeBSD native i386 and amd64 (including compat 32bit) ABIs. Support
for additional architectures will be added after further testing.
Both per-process and per-image controls are implemented:
- procctl(2) adds PROC_ASLR_CTL/PROC_ASLR_STATUS;
- NT_FREEBSD_FCTL_ASLR_DISABLE feature control note bit makes it possible
to force ASLR off for the given binary. (A tool to edit the feature
control note is in development.)
Global controls are:
- kern.elf{32,64}.aslr.enable - for non-fixed mappings done by mmap(2);
- kern.elf{32,64}.aslr.pie_enable - for PIE image activation mappings;
- kern.elf{32,64}.aslr.honor_sbrk - allow to use sbrk area for mmap(2);
- vm.cluster_anon - enables anon mapping clustering.
PR: 208580 (exp runs)
Exp-runs done by: antoine
Reviewed by: markj (previous version)
Discussed with: emaste
Tested by: pho
MFC after: 1 month
Sponsored by: The FreeBSD Foundation
Differential revision: https://reviews.freebsd.org/D5603
2019-02-10 17:19:45 +00:00
|
|
|
p2->p_flag2 = p1->p_flag2 & (P2_ASLR_DISABLE | P2_ASLR_ENABLE |
|
2019-07-10 19:57:48 +00:00
|
|
|
P2_ASLR_IGNSTART | P2_NOTRACE | P2_NOTRACE_EXEC |
|
2019-09-03 18:56:25 +00:00
|
|
|
P2_PROTMAX_ENABLE | P2_PROTMAX_DISABLE | P2_TRAPCAP |
|
|
|
|
P2_STKGAP_DISABLE | P2_STKGAP_DISABLE_EXEC);
|
2007-09-21 04:10:23 +00:00
|
|
|
p2->p_swtick = ticks;
|
2003-04-22 20:54:04 +00:00
|
|
|
if (p1->p_flag & P_PROFIL)
|
|
|
|
startprofclock(p2);
|
1999-11-16 20:31:58 +00:00
|
|
|
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFSIGSHARE) {
|
- Merge struct procsig with struct sigacts.
- Move struct sigacts out of the u-area and malloc() it using the
M_SUBPROC malloc bucket.
- Add a small sigacts_*() API for managing sigacts structures: sigacts_alloc(),
sigacts_free(), sigacts_copy(), sigacts_share(), and sigacts_shared().
- Remove the p_sigignore, p_sigacts, and p_sigcatch macros.
- Add a mutex to struct sigacts that protects all the members of the struct.
- Add sigacts locking.
- Remove Giant from nosys(), kill(), killpg(), and kern_sigaction() now
that sigacts is locked.
- Several in-kernel functions such as psignal(), tdsignal(), trapsignal(),
and thread_stopped() are now MP safe.
Reviewed by: arch@
Approved by: re (rwatson)
2003-05-13 20:36:02 +00:00
|
|
|
p2->p_sigacts = sigacts_hold(p1->p_sigacts);
|
1998-12-19 02:55:34 +00:00
|
|
|
} else {
|
- Merge struct procsig with struct sigacts.
- Move struct sigacts out of the u-area and malloc() it using the
M_SUBPROC malloc bucket.
- Add a small sigacts_*() API for managing sigacts structures: sigacts_alloc(),
sigacts_free(), sigacts_copy(), sigacts_share(), and sigacts_shared().
- Remove the p_sigignore, p_sigacts, and p_sigcatch macros.
- Add a mutex to struct sigacts that protects all the members of the struct.
- Add sigacts locking.
- Remove Giant from nosys(), kill(), killpg(), and kern_sigaction() now
that sigacts is locked.
- Several in-kernel functions such as psignal(), tdsignal(), trapsignal(),
and thread_stopped() are now MP safe.
Reviewed by: arch@
Approved by: re (rwatson)
2003-05-13 20:36:02 +00:00
|
|
|
sigacts_copy(newsigacts, p1->p_sigacts);
|
|
|
|
p2->p_sigacts = newsigacts;
|
2019-09-25 19:20:41 +00:00
|
|
|
if ((fr->fr_flags2 & FR2_DROPSIG_CAUGHT) != 0) {
|
|
|
|
mtx_lock(&p2->p_sigacts->ps_mtx);
|
|
|
|
sig_drop_caught(p2);
|
|
|
|
mtx_unlock(&p2->p_sigacts->ps_mtx);
|
|
|
|
}
|
1998-12-19 02:55:34 +00:00
|
|
|
}
|
2011-07-12 20:37:18 +00:00
|
|
|
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFTSIGZMB)
|
|
|
|
p2->p_sigparent = RFTSIGNUM(fr->fr_flags);
|
|
|
|
else if (fr->fr_flags & RFLINUXTHPN)
|
1998-12-19 02:55:34 +00:00
|
|
|
p2->p_sigparent = SIGUSR1;
|
1999-03-02 00:28:09 +00:00
|
|
|
else
|
|
|
|
p2->p_sigparent = SIGCHLD;
|
1999-01-26 02:38:12 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
p2->p_textvp = p1->p_textvp;
|
2001-03-07 05:21:47 +00:00
|
|
|
p2->p_fd = fd;
|
2003-06-02 16:05:32 +00:00
|
|
|
p2->p_fdtol = fdtol;
|
1996-02-23 18:49:25 +00:00
|
|
|
|
2013-09-19 18:53:42 +00:00
|
|
|
if (p1->p_flag2 & P2_INHERIT_PROTECTED) {
|
|
|
|
p2->p_flag |= P_PROTECTED;
|
|
|
|
p2->p_flag2 |= P2_INHERIT_PROTECTED;
|
|
|
|
}
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
2004-03-04 09:56:29 +00:00
|
|
|
* p_limit is copy-on-write. Bump its refcount.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2007-06-01 01:12:45 +00:00
|
|
|
lim_fork(p1, p2);
|
2004-11-20 02:28:48 +00:00
|
|
|
|
2015-06-10 10:43:59 +00:00
|
|
|
thread_cow_get_proc(td2, p2);
|
|
|
|
|
2004-11-20 02:28:48 +00:00
|
|
|
pstats_fork(p1->p_stats, p2->p_stats);
|
|
|
|
|
2004-10-01 05:01:29 +00:00
|
|
|
PROC_UNLOCK(p1);
|
2004-10-06 23:53:49 +00:00
|
|
|
PROC_UNLOCK(p2);
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2011-01-02 12:16:57 +00:00
|
|
|
/* Bump references to the text vnode (for procfs). */
|
2004-03-08 00:32:34 +00:00
|
|
|
if (p2->p_textvp)
|
2016-12-12 15:37:11 +00:00
|
|
|
vrefact(p2->p_textvp);
|
2004-03-08 00:32:34 +00:00
|
|
|
|
2002-10-15 00:14:32 +00:00
|
|
|
/*
|
2004-03-04 09:56:29 +00:00
|
|
|
* Set up linkage for kernel based threading.
|
2002-10-15 00:14:32 +00:00
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
if ((fr->fr_flags & RFTHREAD) != 0) {
|
2002-10-15 00:14:32 +00:00
|
|
|
mtx_lock(&ppeers_lock);
|
|
|
|
p2->p_peers = p1->p_peers;
|
|
|
|
p1->p_peers = p2;
|
|
|
|
p2->p_leader = p1->p_leader;
|
|
|
|
mtx_unlock(&ppeers_lock);
|
|
|
|
PROC_LOCK(p1->p_leader);
|
|
|
|
if ((p1->p_leader->p_flag & P_WEXIT) != 0) {
|
|
|
|
PROC_UNLOCK(p1->p_leader);
|
|
|
|
/*
|
|
|
|
* The task leader is exiting, so process p1 is
|
|
|
|
* going to be killed shortly. Since p1 obviously
|
|
|
|
* isn't dead yet, we know that the leader is either
|
|
|
|
* sending SIGKILL's to all the processes in this
|
|
|
|
* task or is sleeping waiting for all the peers to
|
|
|
|
* exit. We let p1 complete the fork, but we need
|
|
|
|
* to go ahead and kill the new process p2 since
|
|
|
|
* the task leader may not get a chance to send
|
|
|
|
* SIGKILL to it. We leave it on the list so that
|
|
|
|
* the task leader will wait for this new process
|
|
|
|
* to commit suicide.
|
|
|
|
*/
|
|
|
|
PROC_LOCK(p2);
|
2011-09-16 13:58:51 +00:00
|
|
|
kern_psignal(p2, SIGKILL);
|
2002-10-15 00:14:32 +00:00
|
|
|
PROC_UNLOCK(p2);
|
2002-11-18 14:23:21 +00:00
|
|
|
} else
|
|
|
|
PROC_UNLOCK(p1->p_leader);
|
2002-10-15 00:14:32 +00:00
|
|
|
} else {
|
|
|
|
p2->p_peers = NULL;
|
|
|
|
p2->p_leader = p2;
|
|
|
|
}
|
|
|
|
|
2002-05-02 15:13:45 +00:00
|
|
|
sx_xlock(&proctree_lock);
|
|
|
|
PGRP_LOCK(p1->p_pgrp);
|
|
|
|
PROC_LOCK(p2);
|
|
|
|
PROC_LOCK(p1);
|
|
|
|
|
1997-02-17 10:58:46 +00:00
|
|
|
/*
|
2003-04-22 20:54:04 +00:00
|
|
|
* Preserve some more flags in subprocess. P_PROFIL has already
|
1997-11-20 16:36:17 +00:00
|
|
|
* been preserved.
|
1997-02-17 10:58:46 +00:00
|
|
|
*/
|
2004-01-03 02:02:26 +00:00
|
|
|
p2->p_flag |= p1->p_flag & P_SUGID;
|
Add a way to manage thread signal mask using shared word, instead of syscall.
A new syscall sigfastblock(2) is added which registers a uint32_t
variable as containing the count of blocks for signal delivery. Its
content is read by kernel on each syscall entry and on AST processing,
non-zero count of blocks is interpreted same as the signal mask
blocking all signals.
The biggest downside of the feature that I see is that memory
corruption that affects the registered fast sigblock location, would
cause quite strange application misbehavior. For instance, the process
would be immune to ^C (but killable by SIGKILL).
With consumers (rtld and libthr added), benchmarks do not show a
slow-down of the syscalls in micro-measurements, and macro benchmarks
like buildworld do not demonstrate a difference. Part of the reason is
that buildworld time is dominated by compiler, and clang already links
to libthr. On the other hand, small utilities typically used by shell
scripts have the total number of syscalls cut by half.
The syscall is not exported from the stable libc version namespace on
purpose. It is intended to be used only by our C runtime
implementation internals.
Tested by: pho
Disscussed with: cem, emaste, jilles
Sponsored by: The FreeBSD Foundation
Differential revision: https://reviews.freebsd.org/D12773
2020-02-09 11:53:12 +00:00
|
|
|
td2->td_pflags |= (td->td_pflags & (TDP_ALTSTACK |
|
|
|
|
TDP_SIGFASTBLOCK)) | TDP_FORKING;
|
2002-02-23 11:12:57 +00:00
|
|
|
SESS_LOCK(p1->p_session);
|
1994-05-24 10:09:53 +00:00
|
|
|
if (p1->p_session->s_ttyvp != NULL && p1->p_flag & P_CONTROLT)
|
|
|
|
p2->p_flag |= P_CONTROLT;
|
2002-02-23 11:12:57 +00:00
|
|
|
SESS_UNLOCK(p1->p_session);
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFPPWAIT)
|
1994-05-24 10:09:53 +00:00
|
|
|
p2->p_flag |= P_PPWAIT;
|
1997-11-20 16:36:17 +00:00
|
|
|
|
2004-01-09 23:42:36 +00:00
|
|
|
p2->p_pgrp = p1->p_pgrp;
|
1996-03-11 06:05:03 +00:00
|
|
|
LIST_INSERT_AFTER(p1, p2, p_pglist);
|
2002-04-02 17:12:06 +00:00
|
|
|
PGRP_UNLOCK(p1->p_pgrp);
|
1996-03-11 06:05:03 +00:00
|
|
|
LIST_INIT(&p2->p_children);
|
2012-02-23 11:50:23 +00:00
|
|
|
LIST_INIT(&p2->p_orphans);
|
1996-03-11 06:05:03 +00:00
|
|
|
|
Fix a race between kern_setitimer() and realitexpire(), where the
callout is started before kern_setitimer() acquires process mutex, but
looses a race and kern_setitimer() gets the process mutex before the
callout. Then, assuming that new specified struct itimerval has
it_interval zero, but it_value non-zero, the callout, after it starts
executing again, clears p->p_realtimer.it_value, but kern_setitimer()
already rescheduled the callout.
As the result of the race, both p_realtimer is zero, and the callout
is rescheduled. Then, in the exit1(), the exit code sees that it_value
is zero and does not even try to stop the callout. This allows the
struct proc to be reused and eventually the armed callout is
re-initialized. The consequence is the corrupted callwheel tailq.
Use process mutex to interlock the callout start, which fixes the race.
Reported and tested by: pho
Reviewed by: jhb
MFC after: 2 weeks
2012-12-04 20:49:39 +00:00
|
|
|
callout_init_mtx(&p2->p_itcallout, &p2->p_mtx, 0);
|
2000-11-27 22:52:31 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* This begins the section where we must prevent the parent
|
2004-10-06 23:53:49 +00:00
|
|
|
* from being swapped.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2004-10-06 23:53:49 +00:00
|
|
|
_PHOLD(p1);
|
2001-03-07 02:30:39 +00:00
|
|
|
PROC_UNLOCK(p1);
|
2002-05-02 15:13:45 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Attach the new process to its parent.
|
|
|
|
*
|
|
|
|
* If RFNOWAIT is set, the newly created process becomes a child
|
|
|
|
* of init. This effectively disassociates the child from the
|
|
|
|
* parent.
|
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
if ((fr->fr_flags & RFNOWAIT) != 0) {
|
2014-12-15 12:01:42 +00:00
|
|
|
pptr = p1->p_reaper;
|
|
|
|
p2->p_reaper = pptr;
|
|
|
|
} else {
|
|
|
|
p2->p_reaper = (p1->p_treeflag & P_TREE_REAPER) != 0 ?
|
|
|
|
p1 : p1->p_reaper;
|
2002-05-02 15:13:45 +00:00
|
|
|
pptr = p1;
|
2014-12-15 12:01:42 +00:00
|
|
|
}
|
2002-05-02 15:13:45 +00:00
|
|
|
p2->p_pptr = pptr;
|
2018-11-16 17:07:54 +00:00
|
|
|
p2->p_oppid = pptr->p_pid;
|
2002-05-02 15:13:45 +00:00
|
|
|
LIST_INSERT_HEAD(&pptr->p_children, p2, p_sibling);
|
2014-12-15 12:01:42 +00:00
|
|
|
LIST_INIT(&p2->p_reaplist);
|
|
|
|
LIST_INSERT_HEAD(&p2->p_reaper->p_reaplist, p2, p_reapsibling);
|
2018-12-07 12:22:32 +00:00
|
|
|
if (p2->p_reaper == p1 && p1 != initproc) {
|
2014-12-15 12:01:42 +00:00
|
|
|
p2->p_reapsubtree = p2->p_pid;
|
2018-12-07 12:22:32 +00:00
|
|
|
proc_id_set_cond(PROC_ID_REAP, p2->p_pid);
|
|
|
|
}
|
2002-05-02 15:13:45 +00:00
|
|
|
sx_xunlock(&proctree_lock);
|
|
|
|
|
2003-04-17 22:24:59 +00:00
|
|
|
/* Inform accounting that we have forked. */
|
|
|
|
p2->p_acflag = AFORK;
|
|
|
|
PROC_UNLOCK(p2);
|
|
|
|
|
2011-02-25 22:05:33 +00:00
|
|
|
#ifdef KTRACE
|
|
|
|
ktrprocfork(p1, p2);
|
|
|
|
#endif
|
|
|
|
|
1996-02-23 18:49:25 +00:00
|
|
|
/*
|
1997-04-07 07:16:06 +00:00
|
|
|
* Finish creating the child process. It will return via a different
|
|
|
|
* execution path later. (ie: directly into user mode)
|
1996-02-23 18:49:25 +00:00
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
vm_forkproc(td, p2, td2, vm2, fr->fr_flags);
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags == (RFFDG | RFPROC)) {
|
- Remove 'struct vmmeter' from 'struct pcpu', leaving only global vmmeter
in place. To do per-cpu stats, convert all fields that previously were
maintained in the vmmeters that sit in pcpus to counter(9).
- Since some vmmeter stats may be touched at very early stages of boot,
before we have set up UMA and we can do counter_u64_alloc(), provide an
early counter mechanism:
o Leave one spare uint64_t in struct pcpu, named pc_early_dummy_counter.
o Point counter(9) fields of vmmeter to pcpu[0].pc_early_dummy_counter,
so that at early stages of boot, before counters are allocated we already
point to a counter that can be safely written to.
o For sparc64 that required a whole dummy pcpu[MAXCPU] array.
Further related changes:
- Don't include vmmeter.h into pcpu.h.
- vm.stats.vm.v_swappgsout and vm.stats.vm.v_swappgsin changed to 64-bit,
to match kernel representation.
- struct vmmeter hidden under _KERNEL, and only vmstat(1) is an exclusion.
This is based on benno@'s 4-year old patch:
https://lists.freebsd.org/pipermail/freebsd-arch/2013-July/014471.html
Reviewed by: kib, gallatin, marius, lidl
Differential Revision: https://reviews.freebsd.org/D10156
2017-04-17 17:34:47 +00:00
|
|
|
VM_CNT_INC(v_forks);
|
|
|
|
VM_CNT_ADD(v_forkpages, p2->p_vmspace->vm_dsize +
|
2004-09-03 05:11:32 +00:00
|
|
|
p2->p_vmspace->vm_ssize);
|
2016-02-04 04:25:30 +00:00
|
|
|
} else if (fr->fr_flags == (RFFDG | RFPROC | RFPPWAIT | RFMEM)) {
|
- Remove 'struct vmmeter' from 'struct pcpu', leaving only global vmmeter
in place. To do per-cpu stats, convert all fields that previously were
maintained in the vmmeters that sit in pcpus to counter(9).
- Since some vmmeter stats may be touched at very early stages of boot,
before we have set up UMA and we can do counter_u64_alloc(), provide an
early counter mechanism:
o Leave one spare uint64_t in struct pcpu, named pc_early_dummy_counter.
o Point counter(9) fields of vmmeter to pcpu[0].pc_early_dummy_counter,
so that at early stages of boot, before counters are allocated we already
point to a counter that can be safely written to.
o For sparc64 that required a whole dummy pcpu[MAXCPU] array.
Further related changes:
- Don't include vmmeter.h into pcpu.h.
- vm.stats.vm.v_swappgsout and vm.stats.vm.v_swappgsin changed to 64-bit,
to match kernel representation.
- struct vmmeter hidden under _KERNEL, and only vmstat(1) is an exclusion.
This is based on benno@'s 4-year old patch:
https://lists.freebsd.org/pipermail/freebsd-arch/2013-July/014471.html
Reviewed by: kib, gallatin, marius, lidl
Differential Revision: https://reviews.freebsd.org/D10156
2017-04-17 17:34:47 +00:00
|
|
|
VM_CNT_INC(v_vforks);
|
|
|
|
VM_CNT_ADD(v_vforkpages, p2->p_vmspace->vm_dsize +
|
2004-09-03 05:11:32 +00:00
|
|
|
p2->p_vmspace->vm_ssize);
|
2001-01-23 14:32:01 +00:00
|
|
|
} else if (p1 == &proc0) {
|
- Remove 'struct vmmeter' from 'struct pcpu', leaving only global vmmeter
in place. To do per-cpu stats, convert all fields that previously were
maintained in the vmmeters that sit in pcpus to counter(9).
- Since some vmmeter stats may be touched at very early stages of boot,
before we have set up UMA and we can do counter_u64_alloc(), provide an
early counter mechanism:
o Leave one spare uint64_t in struct pcpu, named pc_early_dummy_counter.
o Point counter(9) fields of vmmeter to pcpu[0].pc_early_dummy_counter,
so that at early stages of boot, before counters are allocated we already
point to a counter that can be safely written to.
o For sparc64 that required a whole dummy pcpu[MAXCPU] array.
Further related changes:
- Don't include vmmeter.h into pcpu.h.
- vm.stats.vm.v_swappgsout and vm.stats.vm.v_swappgsin changed to 64-bit,
to match kernel representation.
- struct vmmeter hidden under _KERNEL, and only vmstat(1) is an exclusion.
This is based on benno@'s 4-year old patch:
https://lists.freebsd.org/pipermail/freebsd-arch/2013-July/014471.html
Reviewed by: kib, gallatin, marius, lidl
Differential Revision: https://reviews.freebsd.org/D10156
2017-04-17 17:34:47 +00:00
|
|
|
VM_CNT_INC(v_kthreads);
|
|
|
|
VM_CNT_ADD(v_kthreadpages, p2->p_vmspace->vm_dsize +
|
2004-09-03 05:11:32 +00:00
|
|
|
p2->p_vmspace->vm_ssize);
|
2001-01-23 14:32:01 +00:00
|
|
|
} else {
|
- Remove 'struct vmmeter' from 'struct pcpu', leaving only global vmmeter
in place. To do per-cpu stats, convert all fields that previously were
maintained in the vmmeters that sit in pcpus to counter(9).
- Since some vmmeter stats may be touched at very early stages of boot,
before we have set up UMA and we can do counter_u64_alloc(), provide an
early counter mechanism:
o Leave one spare uint64_t in struct pcpu, named pc_early_dummy_counter.
o Point counter(9) fields of vmmeter to pcpu[0].pc_early_dummy_counter,
so that at early stages of boot, before counters are allocated we already
point to a counter that can be safely written to.
o For sparc64 that required a whole dummy pcpu[MAXCPU] array.
Further related changes:
- Don't include vmmeter.h into pcpu.h.
- vm.stats.vm.v_swappgsout and vm.stats.vm.v_swappgsin changed to 64-bit,
to match kernel representation.
- struct vmmeter hidden under _KERNEL, and only vmstat(1) is an exclusion.
This is based on benno@'s 4-year old patch:
https://lists.freebsd.org/pipermail/freebsd-arch/2013-July/014471.html
Reviewed by: kib, gallatin, marius, lidl
Differential Revision: https://reviews.freebsd.org/D10156
2017-04-17 17:34:47 +00:00
|
|
|
VM_CNT_INC(v_rforks);
|
|
|
|
VM_CNT_ADD(v_rforkpages, p2->p_vmspace->vm_dsize +
|
2004-09-03 05:11:32 +00:00
|
|
|
p2->p_vmspace->vm_ssize);
|
2001-01-23 14:32:01 +00:00
|
|
|
}
|
|
|
|
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
/*
|
|
|
|
* Associate the process descriptor with the process before anything
|
|
|
|
* can happen that might cause that process to need the descriptor.
|
|
|
|
* However, don't do this until after fork(2) can no longer fail.
|
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFPROCDESC)
|
|
|
|
procdesc_new(p2, fr->fr_pd_flags);
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
|
1996-08-19 02:28:24 +00:00
|
|
|
/*
|
1999-04-17 08:36:07 +00:00
|
|
|
* Both processes are set up, now check if any loadable modules want
|
1996-08-22 03:50:33 +00:00
|
|
|
* to adjust anything.
|
1996-08-19 02:28:24 +00:00
|
|
|
*/
|
2017-11-09 22:51:48 +00:00
|
|
|
EVENTHANDLER_DIRECT_INVOKE(process_fork, p1, p2, fr->fr_flags);
|
1996-08-19 02:28:24 +00:00
|
|
|
|
2004-02-05 21:01:37 +00:00
|
|
|
/*
|
|
|
|
* Set the child start time and mark the process as being complete.
|
|
|
|
*/
|
2011-03-24 18:40:11 +00:00
|
|
|
PROC_LOCK(p2);
|
|
|
|
PROC_LOCK(p1);
|
2004-02-05 21:01:37 +00:00
|
|
|
microuptime(&p2->p_stats->p_start);
|
Commit 6/14 of sched_lock decomposition.
- Use thread_lock() rather than sched_lock for per-thread scheduling
sychronization.
- Use the per-process spinlock rather than the sched_lock for per-process
scheduling synchronization.
- Replace the tail-end of fork_exit() with a scheduler specific routine
which can do the appropriate lock manipulations.
Tested by: kris, current@
Tested on: i386, amd64, ULE, 4BSD, libthr, libkse, PREEMPTION, etc.
Discussed with: kris, attilio, kmacy, jhb, julian, bde (small parts each)
2007-06-04 23:53:34 +00:00
|
|
|
PROC_SLOCK(p2);
|
2004-02-05 21:01:37 +00:00
|
|
|
p2->p_state = PRS_NORMAL;
|
Commit 6/14 of sched_lock decomposition.
- Use thread_lock() rather than sched_lock for per-thread scheduling
sychronization.
- Use the per-process spinlock rather than the sched_lock for per-process
scheduling synchronization.
- Replace the tail-end of fork_exit() with a scheduler specific routine
which can do the appropriate lock manipulations.
Tested by: kris, current@
Tested on: i386, amd64, ULE, 4BSD, libthr, libkse, PREEMPTION, etc.
Discussed with: kris, attilio, kmacy, jhb, julian, bde (small parts each)
2007-06-04 23:53:34 +00:00
|
|
|
PROC_SUNLOCK(p2);
|
2011-01-25 10:59:21 +00:00
|
|
|
|
2010-09-09 09:58:05 +00:00
|
|
|
#ifdef KDTRACE_HOOKS
|
|
|
|
/*
|
2013-12-18 01:41:52 +00:00
|
|
|
* Tell the DTrace fasttrap provider about the new process so that any
|
|
|
|
* tracepoints inherited from the parent can be removed. We have to do
|
|
|
|
* this only after p_state is PRS_NORMAL since the fasttrap module will
|
|
|
|
* use pfind() later on.
|
2010-09-09 09:58:05 +00:00
|
|
|
*/
|
2016-02-04 04:25:30 +00:00
|
|
|
if ((fr->fr_flags & RFMEM) == 0 && dtrace_fasttrap_fork)
|
2010-09-09 09:58:05 +00:00
|
|
|
dtrace_fasttrap_fork(p1, p2);
|
|
|
|
#endif
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFPPWAIT) {
|
2012-02-27 21:10:10 +00:00
|
|
|
td->td_pflags |= TDP_RFPPWAIT;
|
|
|
|
td->td_rfppwait_p = p2;
|
2016-07-18 14:53:55 +00:00
|
|
|
td->td_dbgflags |= TDB_VFORK;
|
2012-02-27 21:10:10 +00:00
|
|
|
}
|
2011-03-24 18:40:11 +00:00
|
|
|
PROC_UNLOCK(p2);
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
/*
|
2018-12-08 06:34:12 +00:00
|
|
|
* Tell any interested parties about the new process.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2018-12-08 06:34:12 +00:00
|
|
|
knote_fork(p1->p_klist, p2->p_pid);
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2000-04-16 18:53:38 +00:00
|
|
|
/*
|
2018-12-08 06:34:12 +00:00
|
|
|
* Now can be swapped.
|
2000-04-16 18:53:38 +00:00
|
|
|
*/
|
2018-12-08 06:34:12 +00:00
|
|
|
_PRELE(p1);
|
|
|
|
PROC_UNLOCK(p1);
|
2016-02-04 04:25:30 +00:00
|
|
|
SDT_PROBE3(proc, , , create, p2, p1, fr->fr_flags);
|
2008-05-24 06:22:16 +00:00
|
|
|
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_flags & RFPROCDESC) {
|
|
|
|
procdesc_finit(p2->p_procdesc, fp_procdesc);
|
|
|
|
fdrop(fp_procdesc, td);
|
|
|
|
}
|
2020-09-01 22:12:32 +00:00
|
|
|
|
2018-06-21 21:12:49 +00:00
|
|
|
/*
|
|
|
|
* Speculative check for PTRACE_FORK. PTRACE_FORK is not
|
|
|
|
* synced with forks in progress so it is OK if we miss it
|
|
|
|
* if being set atm.
|
|
|
|
*/
|
|
|
|
if ((p1->p_ptevents & PTRACE_FORK) != 0) {
|
|
|
|
sx_xlock(&proctree_lock);
|
|
|
|
PROC_LOCK(p2);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* p1->p_ptevents & p1->p_pptr are protected by both
|
|
|
|
* process and proctree locks for modifications,
|
|
|
|
* so owning proctree_lock allows the race-free read.
|
|
|
|
*/
|
|
|
|
if ((p1->p_ptevents & PTRACE_FORK) != 0) {
|
|
|
|
/*
|
|
|
|
* Arrange for debugger to receive the fork event.
|
|
|
|
*
|
|
|
|
* We can report PL_FLAG_FORKED regardless of
|
|
|
|
* P_FOLLOWFORK settings, but it does not make a sense
|
|
|
|
* for runaway child.
|
|
|
|
*/
|
|
|
|
td->td_dbgflags |= TDB_FORK;
|
|
|
|
td->td_dbg_forked = p2->p_pid;
|
|
|
|
td2->td_dbgflags |= TDB_STOPATFORK;
|
|
|
|
proc_set_traced(p2, true);
|
|
|
|
CTR2(KTR_PTRACE,
|
|
|
|
"do_fork: attaching to new child pid %d: oppid %d",
|
|
|
|
p2->p_pid, p2->p_oppid);
|
2018-11-16 17:07:54 +00:00
|
|
|
proc_reparent(p2, p1->p_pptr, false);
|
2018-06-21 21:12:49 +00:00
|
|
|
}
|
|
|
|
PROC_UNLOCK(p2);
|
|
|
|
sx_xunlock(&proctree_lock);
|
|
|
|
}
|
2018-11-22 21:29:36 +00:00
|
|
|
|
|
|
|
racct_proc_fork_done(p2);
|
|
|
|
|
2016-02-04 04:25:30 +00:00
|
|
|
if ((fr->fr_flags & RFSTOPPED) == 0) {
|
2018-11-22 21:29:36 +00:00
|
|
|
if (fr->fr_pidp != NULL)
|
|
|
|
*fr->fr_pidp = p2->p_pid;
|
2016-02-04 04:25:30 +00:00
|
|
|
/*
|
|
|
|
* If RFSTOPPED not requested, make child runnable and
|
|
|
|
* add to run queue.
|
|
|
|
*/
|
|
|
|
thread_lock(td2);
|
|
|
|
TD_SET_CAN_RUN(td2);
|
|
|
|
sched_add(td2, SRQ_BORING);
|
|
|
|
} else {
|
|
|
|
*fr->fr_procp = p2;
|
|
|
|
}
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
|
2018-12-19 20:27:26 +00:00
|
|
|
void
|
|
|
|
fork_rfppwait(struct thread *td)
|
|
|
|
{
|
|
|
|
struct proc *p, *p2;
|
|
|
|
|
|
|
|
MPASS(td->td_pflags & TDP_RFPPWAIT);
|
|
|
|
|
|
|
|
p = td->td_proc;
|
|
|
|
/*
|
|
|
|
* Preserve synchronization semantics of vfork. If
|
|
|
|
* waiting for child to exec or exit, fork set
|
|
|
|
* P_PPWAIT on child, and there we sleep on our proc
|
|
|
|
* (in case of exit).
|
|
|
|
*
|
|
|
|
* Do it after the ptracestop() above is finished, to
|
|
|
|
* not block our debugger until child execs or exits
|
|
|
|
* to finish vfork wait.
|
|
|
|
*/
|
|
|
|
td->td_pflags &= ~TDP_RFPPWAIT;
|
|
|
|
p2 = td->td_rfppwait_p;
|
|
|
|
again:
|
|
|
|
PROC_LOCK(p2);
|
|
|
|
while (p2->p_flag & P_PPWAIT) {
|
|
|
|
PROC_LOCK(p);
|
|
|
|
if (thread_suspend_check_needed()) {
|
|
|
|
PROC_UNLOCK(p2);
|
|
|
|
thread_suspend_check(0);
|
|
|
|
PROC_UNLOCK(p);
|
|
|
|
goto again;
|
|
|
|
} else {
|
|
|
|
PROC_UNLOCK(p);
|
|
|
|
}
|
|
|
|
cv_timedwait(&p2->p_pwait, &p2->p_mtx, hz);
|
|
|
|
}
|
|
|
|
PROC_UNLOCK(p2);
|
|
|
|
|
|
|
|
if (td->td_dbgflags & TDB_VFORK) {
|
|
|
|
PROC_LOCK(p);
|
|
|
|
if (p->p_ptevents & PTRACE_VFORK)
|
|
|
|
ptracestop(td, SIGTRAP, NULL);
|
|
|
|
td->td_dbgflags &= ~TDB_VFORK;
|
|
|
|
PROC_UNLOCK(p);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-12-10 08:33:56 +00:00
|
|
|
int
|
2016-02-04 04:22:18 +00:00
|
|
|
fork1(struct thread *td, struct fork_req *fr)
|
2010-12-10 08:33:56 +00:00
|
|
|
{
|
2015-10-08 11:07:09 +00:00
|
|
|
struct proc *p1, *newproc;
|
2010-12-10 08:33:56 +00:00
|
|
|
struct thread *td2;
|
|
|
|
struct vmspace *vm2;
|
2019-08-17 17:56:43 +00:00
|
|
|
struct ucred *cred;
|
2015-10-08 11:07:09 +00:00
|
|
|
struct file *fp_procdesc;
|
2010-12-10 08:33:56 +00:00
|
|
|
vm_ooffset_t mem_charged;
|
2019-08-17 17:56:43 +00:00
|
|
|
int error, nprocs_new;
|
2010-12-10 08:33:56 +00:00
|
|
|
static int curfail;
|
|
|
|
static struct timeval lastfail;
|
2016-02-04 04:22:18 +00:00
|
|
|
int flags, pages;
|
|
|
|
|
|
|
|
flags = fr->fr_flags;
|
|
|
|
pages = fr->fr_pages;
|
2010-12-10 08:33:56 +00:00
|
|
|
|
2016-02-04 04:25:30 +00:00
|
|
|
if ((flags & RFSTOPPED) != 0)
|
|
|
|
MPASS(fr->fr_procp != NULL && fr->fr_pidp == NULL);
|
|
|
|
else
|
|
|
|
MPASS(fr->fr_procp == NULL);
|
|
|
|
|
2011-07-12 20:37:18 +00:00
|
|
|
/* Check for the undefined or unimplemented flags. */
|
|
|
|
if ((flags & ~(RFFLAGS | RFTSIGFLAGS(RFTSIGMASK))) != 0)
|
|
|
|
return (EINVAL);
|
|
|
|
|
|
|
|
/* Signal value requires RFTSIGZMB. */
|
|
|
|
if ((flags & RFTSIGFLAGS(RFTSIGMASK)) != 0 && (flags & RFTSIGZMB) == 0)
|
|
|
|
return (EINVAL);
|
|
|
|
|
2010-12-10 08:33:56 +00:00
|
|
|
/* Can't copy and clear. */
|
|
|
|
if ((flags & (RFFDG|RFCFDG)) == (RFFDG|RFCFDG))
|
|
|
|
return (EINVAL);
|
|
|
|
|
2011-07-12 20:37:18 +00:00
|
|
|
/* Check the validity of the signal number. */
|
|
|
|
if ((flags & RFTSIGZMB) != 0 && (u_int)RFTSIGNUM(flags) > _SIG_MAXSIG)
|
|
|
|
return (EINVAL);
|
|
|
|
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
if ((flags & RFPROCDESC) != 0) {
|
|
|
|
/* Can't not create a process yet get a process descriptor. */
|
|
|
|
if ((flags & RFPROC) == 0)
|
|
|
|
return (EINVAL);
|
|
|
|
|
|
|
|
/* Must provide a place to put a procdesc if creating one. */
|
2016-02-04 04:22:18 +00:00
|
|
|
if (fr->fr_pd_fd == NULL)
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
return (EINVAL);
|
2016-06-08 02:09:14 +00:00
|
|
|
|
|
|
|
/* Check if we are using supported flags. */
|
|
|
|
if ((fr->fr_pd_flags & ~PD_ALLOWED_AT_FORK) != 0)
|
|
|
|
return (EINVAL);
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
}
|
|
|
|
|
2010-12-10 08:33:56 +00:00
|
|
|
p1 = td->td_proc;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Here we don't create a new process, but we divorce
|
|
|
|
* certain parts of a process from itself.
|
|
|
|
*/
|
2011-01-02 12:16:57 +00:00
|
|
|
if ((flags & RFPROC) == 0) {
|
2016-02-04 04:25:30 +00:00
|
|
|
if (fr->fr_procp != NULL)
|
|
|
|
*fr->fr_procp = NULL;
|
|
|
|
else if (fr->fr_pidp != NULL)
|
|
|
|
*fr->fr_pidp = 0;
|
2011-01-02 12:16:57 +00:00
|
|
|
return (fork_norfproc(td, flags));
|
|
|
|
}
|
2010-12-10 08:33:56 +00:00
|
|
|
|
2015-10-08 11:07:09 +00:00
|
|
|
fp_procdesc = NULL;
|
|
|
|
newproc = NULL;
|
|
|
|
vm2 = NULL;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Increment the nprocs resource before allocations occur.
|
|
|
|
* Although process entries are dynamically created, we still
|
|
|
|
* keep a global limit on the maximum number we will
|
|
|
|
* create. There are hard-limits as to the number of processes
|
|
|
|
* that can run, established by the KVA and memory usage for
|
|
|
|
* the process data.
|
|
|
|
*
|
|
|
|
* Don't allow a nonprivileged user to use the last ten
|
|
|
|
* processes; don't let root exceed the limit.
|
|
|
|
*/
|
|
|
|
nprocs_new = atomic_fetchadd_int(&nprocs, 1) + 1;
|
2019-05-07 15:03:26 +00:00
|
|
|
if (nprocs_new >= maxproc - 10) {
|
|
|
|
if (priv_check_cred(td->td_ucred, PRIV_MAXPROC) != 0 ||
|
|
|
|
nprocs_new >= maxproc) {
|
|
|
|
error = EAGAIN;
|
|
|
|
sx_xlock(&allproc_lock);
|
|
|
|
if (ppsratecheck(&lastfail, &curfail, 1)) {
|
|
|
|
printf("maxproc limit exceeded by uid %u "
|
|
|
|
"(pid %d); see tuning(7) and "
|
|
|
|
"login.conf(5)\n",
|
|
|
|
td->td_ucred->cr_ruid, p1->p_pid);
|
|
|
|
}
|
|
|
|
sx_xunlock(&allproc_lock);
|
|
|
|
goto fail2;
|
2015-10-08 11:07:09 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
/*
|
|
|
|
* If required, create a process descriptor in the parent first; we
|
|
|
|
* will abandon it if something goes wrong. We don't finit() until
|
|
|
|
* later.
|
|
|
|
*/
|
|
|
|
if (flags & RFPROCDESC) {
|
2016-06-08 02:09:14 +00:00
|
|
|
error = procdesc_falloc(td, &fp_procdesc, fr->fr_pd_fd,
|
|
|
|
fr->fr_pd_flags, fr->fr_pd_fcaps);
|
2011-09-17 19:55:32 +00:00
|
|
|
if (error != 0)
|
2015-10-16 14:55:39 +00:00
|
|
|
goto fail2;
|
Add BSM record conversion for a number of syscalls:
- thr_kill(2) and thr_exit(2) generally (no argument auditing here.
- A set of syscalls for the process descriptor family, specifically:
pdfork(2), pdgetpid(2) and pdkill(2)
For these syscalls, audit the file descriptor. In the case of pdfork(2)
a pointer to an integer (file descriptor) is passed in as an argument.
We audit the post initialized file descriptor (not the random garbage
that would have been passed in). We will also audit the child process
which was created from the fork operation (similar to what is done for
the fork(2) syscall).
pdkill(2) we audit the signal value and fd, and finally pdgetpid(2)
just the file descriptor:
- Following is a sample of the produced audit trails:
header,111,11,pdfork(2),0,Sat May 16 03:07:50 2020, + 394 msec
argument,0,0x39d,child PID
argument,2,0x2,flags
argument,1,0x8,fd
subject,root,root,0,root,0,924,0,0,0.0.0.0
return,success,925
header,79,11,pdgetpid(2),0,Sat May 16 03:07:50 2020, + 394 msec
argument,1,0x8,fd
subject,root,root,0,root,0,924,0,0,0.0.0.0
return,success,0
trailer,79
header,135,11,pdkill(2),0,Sat May 16 03:07:50 2020, + 395 msec
argument,1,0x8,fd
argument,2,0xf,signal
process_ex,root,root,0,root,0,925,0,0,0.0.0.0
subject,root,root,0,root,0,924,0,0,0.0.0.0
return,success,0
trailer,135
MFC after: 1 week
2020-05-16 03:45:15 +00:00
|
|
|
AUDIT_ARG_FD(*fr->fr_pd_fd);
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
}
|
|
|
|
|
2010-12-10 08:33:56 +00:00
|
|
|
mem_charged = 0;
|
|
|
|
if (pages == 0)
|
2015-08-10 17:18:21 +00:00
|
|
|
pages = kstack_pages;
|
2010-12-10 08:33:56 +00:00
|
|
|
/* Allocate new proc. */
|
|
|
|
newproc = uma_zalloc(proc_zone, M_WAITOK);
|
|
|
|
td2 = FIRST_THREAD_IN_PROC(newproc);
|
|
|
|
if (td2 == NULL) {
|
|
|
|
td2 = thread_alloc(pages);
|
|
|
|
if (td2 == NULL) {
|
|
|
|
error = ENOMEM;
|
2015-03-21 20:24:03 +00:00
|
|
|
goto fail2;
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
proc_linkup(newproc, td2);
|
|
|
|
} else {
|
|
|
|
if (td2->td_kstack == 0 || td2->td_kstack_pages != pages) {
|
|
|
|
if (td2->td_kstack != 0)
|
|
|
|
vm_thread_dispose(td2);
|
|
|
|
if (!thread_alloc_stack(td2, pages)) {
|
|
|
|
error = ENOMEM;
|
2015-03-21 20:24:03 +00:00
|
|
|
goto fail2;
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if ((flags & RFMEM) == 0) {
|
|
|
|
vm2 = vmspace_fork(p1->p_vmspace, &mem_charged);
|
|
|
|
if (vm2 == NULL) {
|
|
|
|
error = ENOMEM;
|
2015-03-21 20:24:03 +00:00
|
|
|
goto fail2;
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
if (!swap_reserve(mem_charged)) {
|
|
|
|
/*
|
|
|
|
* The swap reservation failed. The accounting
|
|
|
|
* from the entries of the copied vm2 will be
|
2016-04-29 22:15:33 +00:00
|
|
|
* subtracted in vmspace_free(), so force the
|
2010-12-10 08:33:56 +00:00
|
|
|
* reservation there.
|
|
|
|
*/
|
|
|
|
swap_reserve_force(mem_charged);
|
|
|
|
error = ENOMEM;
|
2015-03-21 20:24:03 +00:00
|
|
|
goto fail2;
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
} else
|
|
|
|
vm2 = NULL;
|
|
|
|
|
2011-03-29 17:47:25 +00:00
|
|
|
/*
|
|
|
|
* XXX: This is ugly; when we copy resource usage, we need to bump
|
|
|
|
* per-cred resource counters.
|
|
|
|
*/
|
2020-06-09 23:03:48 +00:00
|
|
|
proc_set_cred_init(newproc, td->td_ucred);
|
2011-03-29 17:47:25 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Initialize resource accounting for the child process.
|
|
|
|
*/
|
|
|
|
error = racct_proc_fork(p1, newproc);
|
|
|
|
if (error != 0) {
|
|
|
|
error = EAGAIN;
|
|
|
|
goto fail1;
|
|
|
|
}
|
|
|
|
|
2011-09-17 20:48:49 +00:00
|
|
|
#ifdef MAC
|
|
|
|
mac_proc_init(newproc);
|
|
|
|
#endif
|
When filt_proc() removes event from the knlist due to the process
exiting (NOTE_EXIT->knlist_remove_inevent()), two things happen:
- knote kn_knlist pointer is reset
- INFLUX knote is removed from the process knlist.
And, there are two consequences:
- KN_LIST_UNLOCK() on such knote is nop
- there is nothing which would block exit1() from processing past the
knlist_destroy() (and knlist_destroy() resets knlist lock pointers).
Both consequences result either in leaked process lock, or
dereferencing NULL function pointers for locking.
Handle this by stopping embedding the process knlist into struct proc.
Instead, the knlist is allocated together with struct proc, but marked
as autodestroy on the zombie reap, by knlist_detach() function. The
knlist is freed when last kevent is removed from the list, in
particular, at the zombie reap time if the list is empty. As result,
the knlist_remove_inevent() is no longer needed and removed.
Other changes:
In filt_procattach(), clear NOTE_EXEC and NOTE_FORK desired events
from kn_sfflags for knote registered by kernel to only get NOTE_CHILD
notifications. The flags leak resulted in excessive
NOTE_EXEC/NOTE_FORK reports.
Fix immediate note activation in filt_procattach(). Condition should
be either the immediate CHILD_NOTE activation, or immediate NOTE_EXIT
report for the exiting process.
In knote_fork(), do not perform racy check for KN_INFLUX before kq
lock is taken. Besides being racy, it did not accounted for notes
just added by scan (KN_SCAN).
Some minor and incomplete style fixes.
Analyzed and tested by: Eric Badger <eric@badgerio.us>
Reviewed by: jhb
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Approved by: re (gjb)
Differential revision: https://reviews.freebsd.org/D6859
2016-06-27 21:52:17 +00:00
|
|
|
newproc->p_klist = knlist_alloc(&newproc->p_mtx);
|
2011-09-17 20:48:49 +00:00
|
|
|
STAILQ_INIT(&newproc->p_ktr);
|
|
|
|
|
2010-12-10 08:33:56 +00:00
|
|
|
/*
|
|
|
|
* Increment the count of procs running with this uid. Don't allow
|
|
|
|
* a nonprivileged user to exceed their current limit.
|
|
|
|
*/
|
2019-08-17 17:56:43 +00:00
|
|
|
cred = td->td_ucred;
|
|
|
|
if (!chgproccnt(cred->cr_ruidinfo, 1, lim_cur(td, RLIMIT_NPROC))) {
|
|
|
|
if (priv_check_cred(cred, PRIV_PROC_LIMIT) != 0)
|
|
|
|
goto fail0;
|
|
|
|
chgproccnt(cred->cr_ruidinfo, 1, 0);
|
2010-12-10 08:33:56 +00:00
|
|
|
}
|
|
|
|
|
2019-08-17 17:56:43 +00:00
|
|
|
do_fork(td, fr, newproc, td2, vm2, fp_procdesc);
|
|
|
|
return (0);
|
|
|
|
fail0:
|
2010-12-10 08:33:56 +00:00
|
|
|
error = EAGAIN;
|
2004-01-25 18:42:18 +00:00
|
|
|
#ifdef MAC
|
2007-10-24 19:04:04 +00:00
|
|
|
mac_proc_destroy(newproc);
|
2004-01-25 18:42:18 +00:00
|
|
|
#endif
|
2011-09-17 20:48:49 +00:00
|
|
|
racct_proc_exit(newproc);
|
2012-05-22 15:58:27 +00:00
|
|
|
fail1:
|
2020-06-09 23:03:48 +00:00
|
|
|
proc_unset_cred(newproc);
|
2015-03-21 20:24:03 +00:00
|
|
|
fail2:
|
2008-03-20 15:24:49 +00:00
|
|
|
if (vm2 != NULL)
|
|
|
|
vmspace_free(vm2);
|
2002-10-15 00:14:32 +00:00
|
|
|
uma_zfree(proc_zone, newproc);
|
2013-02-17 11:47:01 +00:00
|
|
|
if ((flags & RFPROCDESC) != 0 && fp_procdesc != NULL) {
|
2016-02-04 04:22:18 +00:00
|
|
|
fdclose(td, fp_procdesc, *fr->fr_pd_fd);
|
Add experimental support for process descriptors
A "process descriptor" file descriptor is used to manage processes
without using the PID namespace. This is required for Capsicum's
Capability Mode, where the PID namespace is unavailable.
New system calls pdfork(2) and pdkill(2) offer the functional equivalents
of fork(2) and kill(2). pdgetpid(2) allows querying the PID of the remote
process for debugging purposes. The currently-unimplemented pdwait(2) will,
in the future, allow querying rusage/exit status. In the interim, poll(2)
may be used to check (and wait for) process termination.
When a process is referenced by a process descriptor, it does not issue
SIGCHLD to the parent, making it suitable for use in libraries---a common
scenario when using library compartmentalisation from within large
applications (such as web browsers). Some observers may note a similarity
to Mach task ports; process descriptors provide a subset of this behaviour,
but in a UNIX style.
This feature is enabled by "options PROCDESC", but as with several other
Capsicum kernel features, is not enabled by default in GENERIC 9.0.
Reviewed by: jhb, kib
Approved by: re (kib), mentor (rwatson)
Sponsored by: Google Inc
2011-08-18 22:51:30 +00:00
|
|
|
fdrop(fp_procdesc, td);
|
2012-06-19 22:21:59 +00:00
|
|
|
}
|
2015-10-08 11:07:09 +00:00
|
|
|
atomic_add_int(&nprocs, -1);
|
2007-02-27 17:22:30 +00:00
|
|
|
pause("fork", hz / 2);
|
2002-10-15 00:14:32 +00:00
|
|
|
return (error);
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
1996-08-19 02:28:24 +00:00
|
|
|
|
2001-01-24 10:47:14 +00:00
|
|
|
/*
|
|
|
|
* Handle the return of a child process from fork1(). This function
|
|
|
|
* is called from the MD fork_trampoline() entry point.
|
|
|
|
*/
|
|
|
|
void
|
2010-12-06 16:39:18 +00:00
|
|
|
fork_exit(void (*callout)(void *, struct trapframe *), void *arg,
|
|
|
|
struct trapframe *frame)
|
2001-01-24 10:47:14 +00:00
|
|
|
{
|
2002-12-10 02:33:45 +00:00
|
|
|
struct proc *p;
|
2003-08-15 21:29:06 +00:00
|
|
|
struct thread *td;
|
2007-06-12 07:47:09 +00:00
|
|
|
struct thread *dtd;
|
2003-08-15 21:29:06 +00:00
|
|
|
|
2002-12-10 02:33:45 +00:00
|
|
|
td = curthread;
|
|
|
|
p = td->td_proc;
|
2004-02-05 21:01:37 +00:00
|
|
|
KASSERT(p->p_state == PRS_NORMAL, ("executing process is still new"));
|
2003-08-15 21:29:06 +00:00
|
|
|
|
2008-03-12 10:12:01 +00:00
|
|
|
CTR4(KTR_PROC, "fork_exit: new thread %p (td_sched %p, pid %d, %s)",
|
2016-06-05 17:04:03 +00:00
|
|
|
td, td_get_sched(td), p->p_pid, td->td_name);
|
2004-07-27 03:46:31 +00:00
|
|
|
|
Commit 6/14 of sched_lock decomposition.
- Use thread_lock() rather than sched_lock for per-thread scheduling
sychronization.
- Use the per-process spinlock rather than the sched_lock for per-process
scheduling synchronization.
- Replace the tail-end of fork_exit() with a scheduler specific routine
which can do the appropriate lock manipulations.
Tested by: kris, current@
Tested on: i386, amd64, ULE, 4BSD, libthr, libkse, PREEMPTION, etc.
Discussed with: kris, attilio, kmacy, jhb, julian, bde (small parts each)
2007-06-04 23:53:34 +00:00
|
|
|
sched_fork_exit(td);
|
2007-06-12 07:47:09 +00:00
|
|
|
/*
|
|
|
|
* Processes normally resume in mi_switch() after being
|
|
|
|
* cpu_switch()'ed to, but when children start up they arrive here
|
|
|
|
* instead, so we must do much the same things as mi_switch() would.
|
|
|
|
*/
|
|
|
|
if ((dtd = PCPU_GET(deadthread))) {
|
|
|
|
PCPU_SET(deadthread, NULL);
|
|
|
|
thread_stash(dtd);
|
|
|
|
}
|
|
|
|
thread_unlock(td);
|
|
|
|
|
2001-01-24 10:47:14 +00:00
|
|
|
/*
|
2016-06-16 12:05:44 +00:00
|
|
|
* cpu_fork_kthread_handler intercepts this function call to
|
2003-08-15 21:29:06 +00:00
|
|
|
* have this call a non-return function to stay in kernel mode.
|
|
|
|
* initproc has its own fork handler, but it does return.
|
|
|
|
*/
|
2001-02-20 05:26:15 +00:00
|
|
|
KASSERT(callout != NULL, ("NULL callout in fork_exit"));
|
2001-01-26 23:51:41 +00:00
|
|
|
callout(arg, frame);
|
2001-01-24 10:47:14 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Check if a kernel thread misbehaved and returned from its main
|
|
|
|
* function.
|
|
|
|
*/
|
2016-02-09 16:30:16 +00:00
|
|
|
if (p->p_flag & P_KPROC) {
|
2001-01-24 10:47:14 +00:00
|
|
|
printf("Kernel thread \"%s\" (pid %d) exited prematurely.\n",
|
2007-11-14 06:51:33 +00:00
|
|
|
td->td_name, p->p_pid);
|
2016-02-08 23:11:23 +00:00
|
|
|
kthread_exit();
|
2001-01-24 10:47:14 +00:00
|
|
|
}
|
|
|
|
mtx_assert(&Giant, MA_NOTOWNED);
|
2006-08-15 12:10:57 +00:00
|
|
|
|
2011-03-08 19:01:45 +00:00
|
|
|
if (p->p_sysent->sv_schedtail != NULL)
|
|
|
|
(p->p_sysent->sv_schedtail)(td);
|
2015-10-22 09:33:34 +00:00
|
|
|
td->td_pflags &= ~TDP_FORKING;
|
2001-01-24 10:47:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Simplified back end of syscall(), used when returning from fork()
|
2016-08-03 07:10:09 +00:00
|
|
|
* directly into user mode. This function is passed in to fork_exit()
|
|
|
|
* as the first parameter and is called when returning to a new
|
|
|
|
* userland process.
|
2001-01-24 10:47:14 +00:00
|
|
|
*/
|
|
|
|
void
|
2010-12-06 16:39:18 +00:00
|
|
|
fork_return(struct thread *td, struct trapframe *frame)
|
2001-01-24 10:47:14 +00:00
|
|
|
{
|
2018-06-21 21:12:49 +00:00
|
|
|
struct proc *p;
|
2011-01-25 10:59:21 +00:00
|
|
|
|
2015-10-06 19:29:05 +00:00
|
|
|
p = td->td_proc;
|
2011-01-25 10:59:21 +00:00
|
|
|
if (td->td_dbgflags & TDB_STOPATFORK) {
|
|
|
|
PROC_LOCK(p);
|
2018-06-21 21:12:49 +00:00
|
|
|
if ((p->p_flag & P_TRACED) != 0) {
|
2011-01-25 10:59:21 +00:00
|
|
|
/*
|
2018-06-21 21:12:49 +00:00
|
|
|
* Inform the debugger if one is still present.
|
2011-01-25 10:59:21 +00:00
|
|
|
*/
|
When a debugger attaches to the process, SIGSTOP is sent to the
target. Due to a way issignal() selects the next signal to deliver
and report, if the simultaneous or already pending another signal
exists, that signal might be reported by the next waitpid(2) call.
This causes minor annoyance for debuggers, which must be prepared to
take any signal as the first event, then filter SIGSTOP later.
More importantly, for tools like gcore(1), which attach and then
detach without processing events, SIGSTOP might leak to be delivered
after PT_DETACH. This results in the process being unintentionally
stopped after detach, which is fatal for automatic tools.
The solution is to force SIGSTOP to be the first signal reported after
the attach. Attach code is modified to set P2_PTRACE_FSTP to indicate
that the attaching ritual was not yet finished, and issignal() prefers
SIGSTOP in that condition. Also, the thread which handles
P2_PTRACE_FSTP is made to guarantee to own p_xthread during the first
waitpid(2). All that ensures that SIGSTOP is consumed first.
Additionally, if P2_PTRACE_FSTP is still set on detach, which means
that waitpid(2) was not called at all, SIGSTOP is removed from the
queue, ensuring that the process is resumed on detach.
In issignal(), when acting on STOPing signals, remove the signal from
queue before suspending. Otherwise parallel attach could result in
ptracestop() acting on that STOP as if it was the STOP signal from the
attach. Then SIGSTOP from attach leaks again.
As a minor refactoring, some bits of the common attach code is moved
to new helper proc_set_traced().
Reported by: markj
Reviewed by: jhb, markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D7256
2016-07-28 08:41:13 +00:00
|
|
|
td->td_dbgflags |= TDB_CHILD | TDB_SCX | TDB_FSTP;
|
2017-02-20 15:53:16 +00:00
|
|
|
ptracestop(td, SIGSTOP, NULL);
|
2015-10-06 19:29:05 +00:00
|
|
|
td->td_dbgflags &= ~(TDB_CHILD | TDB_SCX);
|
2011-01-25 10:59:21 +00:00
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* ... otherwise clear the request.
|
|
|
|
*/
|
|
|
|
td->td_dbgflags &= ~TDB_STOPATFORK;
|
|
|
|
}
|
|
|
|
PROC_UNLOCK(p);
|
2015-12-29 23:25:26 +00:00
|
|
|
} else if (p->p_flag & P_TRACED || td->td_dbgflags & TDB_BORN) {
|
2015-10-06 19:29:05 +00:00
|
|
|
/*
|
|
|
|
* This is the start of a new thread in a traced
|
|
|
|
* process. Report a system call exit event.
|
|
|
|
*/
|
|
|
|
PROC_LOCK(p);
|
|
|
|
td->td_dbgflags |= TDB_SCX;
|
2016-07-15 15:32:09 +00:00
|
|
|
if ((p->p_ptevents & PTRACE_SCX) != 0 ||
|
2015-12-29 23:25:26 +00:00
|
|
|
(td->td_dbgflags & TDB_BORN) != 0)
|
2017-02-20 15:53:16 +00:00
|
|
|
ptracestop(td, SIGTRAP, NULL);
|
2015-12-29 23:25:26 +00:00
|
|
|
td->td_dbgflags &= ~(TDB_SCX | TDB_BORN);
|
2015-10-06 19:29:05 +00:00
|
|
|
PROC_UNLOCK(p);
|
2011-01-25 10:59:21 +00:00
|
|
|
}
|
2001-01-24 10:47:14 +00:00
|
|
|
|
2006-02-08 08:09:17 +00:00
|
|
|
userret(td, frame);
|
2011-01-25 10:59:21 +00:00
|
|
|
|
2001-01-24 10:47:14 +00:00
|
|
|
#ifdef KTRACE
|
2002-06-07 05:42:25 +00:00
|
|
|
if (KTRPOINT(td, KTR_SYSRET))
|
|
|
|
ktrsysret(SYS_fork, 0, 0);
|
2001-01-24 10:47:14 +00:00
|
|
|
#endif
|
|
|
|
}
|