linux/kernel/exit.c

1791 lines
46 KiB
C
Raw Normal View History

/*
* linux/kernel/exit.c
*
* Copyright (C) 1991, 1992 Linus Torvalds
*/
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/interrupt.h>
#include <linux/module.h>
#include <linux/capability.h>
#include <linux/completion.h>
#include <linux/personality.h>
#include <linux/tty.h>
#include <linux/iocontext.h>
#include <linux/key.h>
#include <linux/security.h>
#include <linux/cpu.h>
#include <linux/acct.h>
#include <linux/tsacct_kern.h>
#include <linux/file.h>
#include <linux/fdtable.h>
#include <linux/binfmts.h>
#include <linux/nsproxy.h>
#include <linux/pid_namespace.h>
#include <linux/ptrace.h>
#include <linux/profile.h>
#include <linux/mount.h>
#include <linux/proc_fs.h>
#include <linux/kthread.h>
#include <linux/mempolicy.h>
#include <linux/taskstats_kern.h>
#include <linux/delayacct.h>
#include <linux/freezer.h>
#include <linux/cgroup.h>
#include <linux/syscalls.h>
#include <linux/signal.h>
#include <linux/posix-timers.h>
#include <linux/cn_proc.h>
#include <linux/mutex.h>
#include <linux/futex.h>
#include <linux/pipe_fs_i.h>
#include <linux/audit.h> /* for audit_free() */
#include <linux/resource.h>
#include <linux/blkdev.h>
#include <linux/task_io_accounting_ops.h>
#include <linux/tracehook.h>
#include <linux/fs_struct.h>
CRED: Inaugurate COW credentials Inaugurate copy-on-write credentials management. This uses RCU to manage the credentials pointer in the task_struct with respect to accesses by other tasks. A process may only modify its own credentials, and so does not need locking to access or modify its own credentials. A mutex (cred_replace_mutex) is added to the task_struct to control the effect of PTRACE_ATTACHED on credential calculations, particularly with respect to execve(). With this patch, the contents of an active credentials struct may not be changed directly; rather a new set of credentials must be prepared, modified and committed using something like the following sequence of events: struct cred *new = prepare_creds(); int ret = blah(new); if (ret < 0) { abort_creds(new); return ret; } return commit_creds(new); There are some exceptions to this rule: the keyrings pointed to by the active credentials may be instantiated - keyrings violate the COW rule as managing COW keyrings is tricky, given that it is possible for a task to directly alter the keys in a keyring in use by another task. To help enforce this, various pointers to sets of credentials, such as those in the task_struct, are declared const. The purpose of this is compile-time discouragement of altering credentials through those pointers. Once a set of credentials has been made public through one of these pointers, it may not be modified, except under special circumstances: (1) Its reference count may incremented and decremented. (2) The keyrings to which it points may be modified, but not replaced. The only safe way to modify anything else is to create a replacement and commit using the functions described in Documentation/credentials.txt (which will be added by a later patch). This patch and the preceding patches have been tested with the LTP SELinux testsuite. This patch makes several logical sets of alteration: (1) execve(). This now prepares and commits credentials in various places in the security code rather than altering the current creds directly. (2) Temporary credential overrides. do_coredump() and sys_faccessat() now prepare their own credentials and temporarily override the ones currently on the acting thread, whilst preventing interference from other threads by holding cred_replace_mutex on the thread being dumped. This will be replaced in a future patch by something that hands down the credentials directly to the functions being called, rather than altering the task's objective credentials. (3) LSM interface. A number of functions have been changed, added or removed: (*) security_capset_check(), ->capset_check() (*) security_capset_set(), ->capset_set() Removed in favour of security_capset(). (*) security_capset(), ->capset() New. This is passed a pointer to the new creds, a pointer to the old creds and the proposed capability sets. It should fill in the new creds or return an error. All pointers, barring the pointer to the new creds, are now const. (*) security_bprm_apply_creds(), ->bprm_apply_creds() Changed; now returns a value, which will cause the process to be killed if it's an error. (*) security_task_alloc(), ->task_alloc_security() Removed in favour of security_prepare_creds(). (*) security_cred_free(), ->cred_free() New. Free security data attached to cred->security. (*) security_prepare_creds(), ->cred_prepare() New. Duplicate any security data attached to cred->security. (*) security_commit_creds(), ->cred_commit() New. Apply any security effects for the upcoming installation of new security by commit_creds(). (*) security_task_post_setuid(), ->task_post_setuid() Removed in favour of security_task_fix_setuid(). (*) security_task_fix_setuid(), ->task_fix_setuid() Fix up the proposed new credentials for setuid(). This is used by cap_set_fix_setuid() to implicitly adjust capabilities in line with setuid() changes. Changes are made to the new credentials, rather than the task itself as in security_task_post_setuid(). (*) security_task_reparent_to_init(), ->task_reparent_to_init() Removed. Instead the task being reparented to init is referred directly to init's credentials. NOTE! This results in the loss of some state: SELinux's osid no longer records the sid of the thread that forked it. (*) security_key_alloc(), ->key_alloc() (*) security_key_permission(), ->key_permission() Changed. These now take cred pointers rather than task pointers to refer to the security context. (4) sys_capset(). This has been simplified and uses less locking. The LSM functions it calls have been merged. (5) reparent_to_kthreadd(). This gives the current thread the same credentials as init by simply using commit_thread() to point that way. (6) __sigqueue_alloc() and switch_uid() __sigqueue_alloc() can't stop the target task from changing its creds beneath it, so this function gets a reference to the currently applicable user_struct which it then passes into the sigqueue struct it returns if successful. switch_uid() is now called from commit_creds(), and possibly should be folded into that. commit_creds() should take care of protecting __sigqueue_alloc(). (7) [sg]et[ug]id() and co and [sg]et_current_groups. The set functions now all use prepare_creds(), commit_creds() and abort_creds() to build and check a new set of credentials before applying it. security_task_set[ug]id() is called inside the prepared section. This guarantees that nothing else will affect the creds until we've finished. The calling of set_dumpable() has been moved into commit_creds(). Much of the functionality of set_user() has been moved into commit_creds(). The get functions all simply access the data directly. (8) security_task_prctl() and cap_task_prctl(). security_task_prctl() has been modified to return -ENOSYS if it doesn't want to handle a function, or otherwise return the return value directly rather than through an argument. Additionally, cap_task_prctl() now prepares a new set of credentials, even if it doesn't end up using it. (9) Keyrings. A number of changes have been made to the keyrings code: (a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have all been dropped and built in to the credentials functions directly. They may want separating out again later. (b) key_alloc() and search_process_keyrings() now take a cred pointer rather than a task pointer to specify the security context. (c) copy_creds() gives a new thread within the same thread group a new thread keyring if its parent had one, otherwise it discards the thread keyring. (d) The authorisation key now points directly to the credentials to extend the search into rather pointing to the task that carries them. (e) Installing thread, process or session keyrings causes a new set of credentials to be created, even though it's not strictly necessary for process or session keyrings (they're shared). (10) Usermode helper. The usermode helper code now carries a cred struct pointer in its subprocess_info struct instead of a new session keyring pointer. This set of credentials is derived from init_cred and installed on the new process after it has been cloned. call_usermodehelper_setup() allocates the new credentials and call_usermodehelper_freeinfo() discards them if they haven't been used. A special cred function (prepare_usermodeinfo_creds()) is provided specifically for call_usermodehelper_setup() to call. call_usermodehelper_setkeys() adjusts the credentials to sport the supplied keyring as the new session keyring. (11) SELinux. SELinux has a number of changes, in addition to those to support the LSM interface changes mentioned above: (a) selinux_setprocattr() no longer does its check for whether the current ptracer can access processes with the new SID inside the lock that covers getting the ptracer's SID. Whilst this lock ensures that the check is done with the ptracer pinned, the result is only valid until the lock is released, so there's no point doing it inside the lock. (12) is_single_threaded(). This function has been extracted from selinux_setprocattr() and put into a file of its own in the lib/ directory as join_session_keyring() now wants to use it too. The code in SELinux just checked to see whether a task shared mm_structs with other tasks (CLONE_VM), but that isn't good enough. We really want to know if they're part of the same thread group (CLONE_THREAD). (13) nfsd. The NFS server daemon now has to use the COW credentials to set the credentials it is going to use. It really needs to pass the credentials down to the functions it calls, but it can't do that until other patches in this series have been applied. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: James Morris <jmorris@namei.org> Signed-off-by: James Morris <jmorris@namei.org>
2008-11-13 23:39:23 +00:00
#include <linux/init_task.h>
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 10:02:48 +00:00
#include <linux/perf_event.h>
#include <trace/events/sched.h>
hw-breakpoints: Rewrite the hw-breakpoints layer on top of perf events This patch rebase the implementation of the breakpoints API on top of perf events instances. Each breakpoints are now perf events that handle the register scheduling, thread/cpu attachment, etc.. The new layering is now made as follows: ptrace kgdb ftrace perf syscall \ | / / \ | / / / Core breakpoint API / / | / | / Breakpoints perf events | | Breakpoints PMU ---- Debug Register constraints handling (Part of core breakpoint API) | | Hardware debug registers Reasons of this rewrite: - Use the centralized/optimized pmu registers scheduling, implying an easier arch integration - More powerful register handling: perf attributes (pinned/flexible events, exclusive/non-exclusive, tunable period, etc...) Impact: - New perf ABI: the hardware breakpoints counters - Ptrace breakpoints setting remains tricky and still needs some per thread breakpoints references. Todo (in the order): - Support breakpoints perf counter events for perf tools (ie: implement perf_bpcounter_event()) - Support from perf tools Changes in v2: - Follow the perf "event " rename - The ptrace regression have been fixed (ptrace breakpoint perf events weren't released when a task ended) - Drop the struct hw_breakpoint and store generic fields in perf_event_attr. - Separate core and arch specific headers, drop asm-generic/hw_breakpoint.h and create linux/hw_breakpoint.h - Use new generic len/type for breakpoint - Handle off case: when breakpoints api is not supported by an arch Changes in v3: - Fix broken CONFIG_KVM, we need to propagate the breakpoint api changes to kvm when we exit the guest and restore the bp registers to the host. Changes in v4: - Drop the hw_breakpoint_restore() stub as it is only used by KVM - EXPORT_SYMBOL_GPL hw_breakpoint_restore() as KVM can be built as a module - Restore the breakpoints unconditionally on kvm guest exit: TIF_DEBUG_THREAD doesn't anymore cover every cases of running breakpoints and vcpu->arch.switch_db_regs might not always be set when the guest used debug registers. (Waiting for a reliable optimization) Changes in v5: - Split-up the asm-generic/hw-breakpoint.h moving to linux/hw_breakpoint.h into a separate patch - Optimize the breakpoints restoring while switching from kvm guest to host. We only want to restore the state if we have active breakpoints to the host, otherwise we don't care about messed-up address registers. - Add asm/hw_breakpoint.h to Kbuild - Fix bad breakpoint type in trace_selftest.c Changes in v6: - Fix wrong header inclusion in trace.h (triggered a build error with CONFIG_FTRACE_SELFTEST Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Prasad <prasad@linux.vnet.ibm.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jan Kiszka <jan.kiszka@web.de> Cc: Jiri Slaby <jirislaby@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Avi Kivity <avi@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Paul Mundt <lethal@linux-sh.org>
2009-09-09 17:22:48 +00:00
#include <linux/hw_breakpoint.h>
#include <linux/oom.h>
#include <linux/writeback.h>
#include <linux/shm.h>
#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm/pgtable.h>
#include <asm/mmu_context.h>
static void exit_mm(struct task_struct * tsk);
static void __unhash_process(struct task_struct *p, bool group_dead)
{
nr_threads--;
detach_pid(p, PIDTYPE_PID);
if (group_dead) {
detach_pid(p, PIDTYPE_PGID);
detach_pid(p, PIDTYPE_SID);
list_del_rcu(&p->tasks);
list_del_init(&p->sibling);
__this_cpu_dec(process_counts);
pidns: guarantee that the pidns init will be the last pidns process reaped Today we have a twofold bug. Sometimes release_task on pid == 1 in a pid namespace can run before other processes in a pid namespace have had release task called. With the result that pid_ns_release_proc can be called before the last proc_flus_task() is done using upid->ns->proc_mnt, resulting in the use of a stale pointer. This same set of circumstances can lead to waitpid(...) returning for a processes started with clone(CLONE_NEWPID) before the every process in the pid namespace has actually exited. To fix this modify zap_pid_ns_processess wait until all other processes in the pid namespace have exited, even EXIT_DEAD zombies. The delay_group_leader and related tests ensure that the thread gruop leader will be the last thread of a process group to be reaped, or to become EXIT_DEAD and self reap. With the change to zap_pid_ns_processes we get the guarantee that pid == 1 in a pid namespace will be the last task that release_task is called on. With pid == 1 being the last task to pass through release_task pid_ns_release_proc can no longer be called too early nor can wait return before all of the EXIT_DEAD tasks in a pid namespace have exited. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Louis Rilling <louis.rilling@kerlabs.com> Cc: Mike Galbraith <efault@gmx.de> Acked-by: Pavel Emelyanov <xemul@parallels.com> Tested-by: Andrew Wagin <avagin@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-06-20 19:53:03 +00:00
/*
* If we are the last child process in a pid namespace to be
* reaped, notify the reaper sleeping zap_pid_ns_processes().
*/
if (IS_ENABLED(CONFIG_PID_NS)) {
struct task_struct *parent = p->real_parent;
if ((task_active_pid_ns(parent)->child_reaper == parent) &&
pidns: guarantee that the pidns init will be the last pidns process reaped Today we have a twofold bug. Sometimes release_task on pid == 1 in a pid namespace can run before other processes in a pid namespace have had release task called. With the result that pid_ns_release_proc can be called before the last proc_flus_task() is done using upid->ns->proc_mnt, resulting in the use of a stale pointer. This same set of circumstances can lead to waitpid(...) returning for a processes started with clone(CLONE_NEWPID) before the every process in the pid namespace has actually exited. To fix this modify zap_pid_ns_processess wait until all other processes in the pid namespace have exited, even EXIT_DEAD zombies. The delay_group_leader and related tests ensure that the thread gruop leader will be the last thread of a process group to be reaped, or to become EXIT_DEAD and self reap. With the change to zap_pid_ns_processes we get the guarantee that pid == 1 in a pid namespace will be the last task that release_task is called on. With pid == 1 being the last task to pass through release_task pid_ns_release_proc can no longer be called too early nor can wait return before all of the EXIT_DEAD tasks in a pid namespace have exited. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Louis Rilling <louis.rilling@kerlabs.com> Cc: Mike Galbraith <efault@gmx.de> Acked-by: Pavel Emelyanov <xemul@parallels.com> Tested-by: Andrew Wagin <avagin@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-06-20 19:53:03 +00:00
list_empty(&parent->children) &&
(parent->flags & PF_EXITING))
wake_up_process(parent);
}
}
list_del_rcu(&p->thread_group);
}
/*
* This function expects the tasklist_lock write-locked.
*/
static void __exit_signal(struct task_struct *tsk)
{
struct signal_struct *sig = tsk->signal;
bool group_dead = thread_group_leader(tsk);
struct sighand_struct *sighand;
struct tty_struct *uninitialized_var(tty);
sighand = rcu_dereference_check(tsk->sighand,
lockdep_tasklist_lock_is_held());
spin_lock(&sighand->siglock);
posix_cpu_timers_exit(tsk);
if (group_dead) {
posix_cpu_timers_exit_group(tsk);
tty = sig->tty;
sig->tty = NULL;
} else {
/*
* This can only happen if the caller is de_thread().
* FIXME: this is the temporary hack, we should teach
* posix-cpu-timers to handle this case correctly.
*/
if (unlikely(has_group_leader_pid(tsk)))
posix_cpu_timers_exit_group(tsk);
/*
* If there is any task waiting for the group exit
* then notify it:
*/
if (sig->notify_count > 0 && !--sig->notify_count)
wake_up_process(sig->group_exit_task);
if (tsk == sig->curr_target)
sig->curr_target = next_thread(tsk);
/*
* Accumulate here the counters for all threads but the
* group leader as they die, so they can be added into
* the process-wide totals when those are taken.
* The group leader stays around as a zombie as long
* as there are other threads. When it gets reaped,
* the exit.c code will add its counts into these totals.
* We won't ever get here for the group leader, since it
* will have been the last reference on the signal_struct.
*/
sig->utime += tsk->utime;
sig->stime += tsk->stime;
sig->gtime += tsk->gtime;
sig->min_flt += tsk->min_flt;
sig->maj_flt += tsk->maj_flt;
sig->nvcsw += tsk->nvcsw;
sig->nivcsw += tsk->nivcsw;
sig->inblock += task_io_get_inblock(tsk);
sig->oublock += task_io_get_oublock(tsk);
task_io_accounting_add(&sig->ioac, &tsk->ioac);
sig->sum_sched_runtime += tsk->se.sum_exec_runtime;
}
sig->nr_threads--;
__unhash_process(tsk, group_dead);
/*
* Do this under ->siglock, we can race with another thread
* doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals.
*/
flush_sigqueue(&tsk->pending);
tsk->sighand = NULL;
spin_unlock(&sighand->siglock);
__cleanup_sighand(sighand);
clear_tsk_thread_flag(tsk,TIF_SIGPENDING);
if (group_dead) {
flush_sigqueue(&sig->shared_pending);
tty_kref_put(tty);
}
}
[PATCH] task: RCU protect task->usage A big problem with rcu protected data structures that are also reference counted is that you must jump through several hoops to increase the reference count. I think someone finally implemented atomic_inc_not_zero(&count) to automate the common case. Unfortunately this means you must special case the rcu access case. When data structures are only visible via rcu in a manner that is not determined by the reference count on the object (i.e. tasks are visible until their zombies are reaped) there is a much simpler technique we can employ. Simply delaying the decrement of the reference count until the rcu interval is over. What that means is that the proc code that looks up a task and later wants to sleep can now do: rcu_read_lock(); task = find_task_by_pid(some_pid); if (task) { get_task_struct(task); } rcu_read_unlock(); The effect on the rest of the kernel is that put_task_struct becomes cheaper and immediate, and in the case where the task has been reaped it frees the task immediate instead of unnecessarily waiting an until the rcu interval is over. Cleanup of task_struct does not happen when its reference count drops to zero, instead cleanup happens when release_task is called. Tasks can only be looked up via rcu before release_task is called. All rcu protected members of task_struct are freed by release_task. Therefore we can move call_rcu from put_task_struct into release_task. And we can modify release_task to not immediately release the reference count but instead have it call put_task_struct from the function it gives to call_rcu. The end result: - get_task_struct is safe in an rcu context where we have just looked up the task. - put_task_struct() simplifies into its old pre rcu self. This reorganization also makes put_task_struct uncallable from modules as it is not exported but it does not appear to be called from any modules so this should not be an issue, and is trivially fixed. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-31 10:31:37 +00:00
static void delayed_put_task_struct(struct rcu_head *rhp)
{
struct task_struct *tsk = container_of(rhp, struct task_struct, rcu);
perf_event_delayed_put(tsk);
trace_sched_process_free(tsk);
put_task_struct(tsk);
[PATCH] task: RCU protect task->usage A big problem with rcu protected data structures that are also reference counted is that you must jump through several hoops to increase the reference count. I think someone finally implemented atomic_inc_not_zero(&count) to automate the common case. Unfortunately this means you must special case the rcu access case. When data structures are only visible via rcu in a manner that is not determined by the reference count on the object (i.e. tasks are visible until their zombies are reaped) there is a much simpler technique we can employ. Simply delaying the decrement of the reference count until the rcu interval is over. What that means is that the proc code that looks up a task and later wants to sleep can now do: rcu_read_lock(); task = find_task_by_pid(some_pid); if (task) { get_task_struct(task); } rcu_read_unlock(); The effect on the rest of the kernel is that put_task_struct becomes cheaper and immediate, and in the case where the task has been reaped it frees the task immediate instead of unnecessarily waiting an until the rcu interval is over. Cleanup of task_struct does not happen when its reference count drops to zero, instead cleanup happens when release_task is called. Tasks can only be looked up via rcu before release_task is called. All rcu protected members of task_struct are freed by release_task. Therefore we can move call_rcu from put_task_struct into release_task. And we can modify release_task to not immediately release the reference count but instead have it call put_task_struct from the function it gives to call_rcu. The end result: - get_task_struct is safe in an rcu context where we have just looked up the task. - put_task_struct() simplifies into its old pre rcu self. This reorganization also makes put_task_struct uncallable from modules as it is not exported but it does not appear to be called from any modules so this should not be an issue, and is trivially fixed. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-31 10:31:37 +00:00
}
void release_task(struct task_struct * p)
{
struct task_struct *leader;
int zap_leader;
repeat:
/* don't need to get the RCU readlock here - the process is dead and
* can't be modifying its own credentials. But shut RCU-lockdep up */
rcu_read_lock();
atomic_dec(&__task_cred(p)->user->processes);
rcu_read_unlock();
proc_flush_task(p);
write_lock_irq(&tasklist_lock);
ptrace_release_task(p);
__exit_signal(p);
/*
* If we are the last non-leader member of the thread
* group, and the leader is zombie, then notify the
* group leader's parent process. (if it wants notification.)
*/
zap_leader = 0;
leader = p->group_leader;
if (leader != p && thread_group_empty(leader) && leader->exit_state == EXIT_ZOMBIE) {
/*
* If we were the last child thread and the leader has
* exited already, and the leader's parent ignores SIGCHLD,
* then we are the one who should release the leader.
*/
zap_leader = do_notify_parent(leader, leader->exit_signal);
if (zap_leader)
leader->exit_state = EXIT_DEAD;
}
write_unlock_irq(&tasklist_lock);
release_thread(p);
[PATCH] task: RCU protect task->usage A big problem with rcu protected data structures that are also reference counted is that you must jump through several hoops to increase the reference count. I think someone finally implemented atomic_inc_not_zero(&count) to automate the common case. Unfortunately this means you must special case the rcu access case. When data structures are only visible via rcu in a manner that is not determined by the reference count on the object (i.e. tasks are visible until their zombies are reaped) there is a much simpler technique we can employ. Simply delaying the decrement of the reference count until the rcu interval is over. What that means is that the proc code that looks up a task and later wants to sleep can now do: rcu_read_lock(); task = find_task_by_pid(some_pid); if (task) { get_task_struct(task); } rcu_read_unlock(); The effect on the rest of the kernel is that put_task_struct becomes cheaper and immediate, and in the case where the task has been reaped it frees the task immediate instead of unnecessarily waiting an until the rcu interval is over. Cleanup of task_struct does not happen when its reference count drops to zero, instead cleanup happens when release_task is called. Tasks can only be looked up via rcu before release_task is called. All rcu protected members of task_struct are freed by release_task. Therefore we can move call_rcu from put_task_struct into release_task. And we can modify release_task to not immediately release the reference count but instead have it call put_task_struct from the function it gives to call_rcu. The end result: - get_task_struct is safe in an rcu context where we have just looked up the task. - put_task_struct() simplifies into its old pre rcu self. This reorganization also makes put_task_struct uncallable from modules as it is not exported but it does not appear to be called from any modules so this should not be an issue, and is trivially fixed. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-31 10:31:37 +00:00
call_rcu(&p->rcu, delayed_put_task_struct);
p = leader;
if (unlikely(zap_leader))
goto repeat;
}
/*
* This checks not only the pgrp, but falls back on the pid if no
* satisfactory pgrp is found. I dunno - gdb doesn't work correctly
* without this...
*
* The caller must hold rcu lock or the tasklist lock.
*/
struct pid *session_of_pgrp(struct pid *pgrp)
{
struct task_struct *p;
struct pid *sid = NULL;
p = pid_task(pgrp, PIDTYPE_PGID);
if (p == NULL)
p = pid_task(pgrp, PIDTYPE_PID);
if (p != NULL)
sid = task_session(p);
return sid;
}
/*
* Determine if a process group is "orphaned", according to the POSIX
* definition in 2.2.2.52. Orphaned process groups are not to be affected
* by terminal-generated stop signals. Newly orphaned process groups are
* to receive a SIGHUP and a SIGCONT.
*
* "I ask you, have you ever known what it is to be an orphan?"
*/
static int will_become_orphaned_pgrp(struct pid *pgrp, struct task_struct *ignored_task)
{
struct task_struct *p;
do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
if ((p == ignored_task) ||
(p->exit_state && thread_group_empty(p)) ||
is_global_init(p->real_parent))
continue;
if (task_pgrp(p->real_parent) != pgrp &&
task_session(p->real_parent) == task_session(p))
return 0;
} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
return 1;
}
int is_current_pgrp_orphaned(void)
{
int retval;
read_lock(&tasklist_lock);
retval = will_become_orphaned_pgrp(task_pgrp(current), NULL);
read_unlock(&tasklist_lock);
return retval;
}
static bool has_stopped_jobs(struct pid *pgrp)
{
struct task_struct *p;
do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
if (p->signal->flags & SIGNAL_STOP_STOPPED)
return true;
} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
return false;
}
/*
* Check to see if any process groups have become orphaned as
* a result of our exiting, and if they have any stopped jobs,
* send them a SIGHUP and then a SIGCONT. (POSIX 3.2.2.2)
*/
static void
kill_orphaned_pgrp(struct task_struct *tsk, struct task_struct *parent)
{
struct pid *pgrp = task_pgrp(tsk);
struct task_struct *ignored_task = tsk;
if (!parent)
/* exit: our father is in a different pgrp than
* we are and we were the only connection outside.
*/
parent = tsk->real_parent;
else
/* reparent: our child is in a different pgrp than
* we are, and it was the only connection outside.
*/
ignored_task = NULL;
if (task_pgrp(parent) != pgrp &&
task_session(parent) == task_session(tsk) &&
will_become_orphaned_pgrp(pgrp, ignored_task) &&
has_stopped_jobs(pgrp)) {
__kill_pgrp_info(SIGHUP, SEND_SIG_PRIV, pgrp);
__kill_pgrp_info(SIGCONT, SEND_SIG_PRIV, pgrp);
}
}
/**
* reparent_to_kthreadd - Reparent the calling kernel thread to kthreadd
*
* If a kernel thread is launched as a result of a system call, or if
* it ever exits, it should generally reparent itself to kthreadd so it
* isn't in the way of other processes and is correctly cleaned up on exit.
*
* The various task state such as scheduling policy and priority may have
* been inherited from a user process, so we reset them to sane values here.
*
* NOTE that reparent_to_kthreadd() gives the caller full capabilities.
*/
static void reparent_to_kthreadd(void)
{
write_lock_irq(&tasklist_lock);
ptrace_unlink(current);
/* Reparent to init */
current->real_parent = current->parent = kthreadd_task;
list_move_tail(&current->sibling, &current->real_parent->children);
/* Set the exit signal to SIGCHLD so we signal init on exit */
current->exit_signal = SIGCHLD;
if (task_nice(current) < 0)
set_user_nice(current, 0);
/* cpus_allowed? */
/* rt_priority? */
/* signals? */
memcpy(current->signal->rlim, init_task.signal->rlim,
sizeof(current->signal->rlim));
CRED: Inaugurate COW credentials Inaugurate copy-on-write credentials management. This uses RCU to manage the credentials pointer in the task_struct with respect to accesses by other tasks. A process may only modify its own credentials, and so does not need locking to access or modify its own credentials. A mutex (cred_replace_mutex) is added to the task_struct to control the effect of PTRACE_ATTACHED on credential calculations, particularly with respect to execve(). With this patch, the contents of an active credentials struct may not be changed directly; rather a new set of credentials must be prepared, modified and committed using something like the following sequence of events: struct cred *new = prepare_creds(); int ret = blah(new); if (ret < 0) { abort_creds(new); return ret; } return commit_creds(new); There are some exceptions to this rule: the keyrings pointed to by the active credentials may be instantiated - keyrings violate the COW rule as managing COW keyrings is tricky, given that it is possible for a task to directly alter the keys in a keyring in use by another task. To help enforce this, various pointers to sets of credentials, such as those in the task_struct, are declared const. The purpose of this is compile-time discouragement of altering credentials through those pointers. Once a set of credentials has been made public through one of these pointers, it may not be modified, except under special circumstances: (1) Its reference count may incremented and decremented. (2) The keyrings to which it points may be modified, but not replaced. The only safe way to modify anything else is to create a replacement and commit using the functions described in Documentation/credentials.txt (which will be added by a later patch). This patch and the preceding patches have been tested with the LTP SELinux testsuite. This patch makes several logical sets of alteration: (1) execve(). This now prepares and commits credentials in various places in the security code rather than altering the current creds directly. (2) Temporary credential overrides. do_coredump() and sys_faccessat() now prepare their own credentials and temporarily override the ones currently on the acting thread, whilst preventing interference from other threads by holding cred_replace_mutex on the thread being dumped. This will be replaced in a future patch by something that hands down the credentials directly to the functions being called, rather than altering the task's objective credentials. (3) LSM interface. A number of functions have been changed, added or removed: (*) security_capset_check(), ->capset_check() (*) security_capset_set(), ->capset_set() Removed in favour of security_capset(). (*) security_capset(), ->capset() New. This is passed a pointer to the new creds, a pointer to the old creds and the proposed capability sets. It should fill in the new creds or return an error. All pointers, barring the pointer to the new creds, are now const. (*) security_bprm_apply_creds(), ->bprm_apply_creds() Changed; now returns a value, which will cause the process to be killed if it's an error. (*) security_task_alloc(), ->task_alloc_security() Removed in favour of security_prepare_creds(). (*) security_cred_free(), ->cred_free() New. Free security data attached to cred->security. (*) security_prepare_creds(), ->cred_prepare() New. Duplicate any security data attached to cred->security. (*) security_commit_creds(), ->cred_commit() New. Apply any security effects for the upcoming installation of new security by commit_creds(). (*) security_task_post_setuid(), ->task_post_setuid() Removed in favour of security_task_fix_setuid(). (*) security_task_fix_setuid(), ->task_fix_setuid() Fix up the proposed new credentials for setuid(). This is used by cap_set_fix_setuid() to implicitly adjust capabilities in line with setuid() changes. Changes are made to the new credentials, rather than the task itself as in security_task_post_setuid(). (*) security_task_reparent_to_init(), ->task_reparent_to_init() Removed. Instead the task being reparented to init is referred directly to init's credentials. NOTE! This results in the loss of some state: SELinux's osid no longer records the sid of the thread that forked it. (*) security_key_alloc(), ->key_alloc() (*) security_key_permission(), ->key_permission() Changed. These now take cred pointers rather than task pointers to refer to the security context. (4) sys_capset(). This has been simplified and uses less locking. The LSM functions it calls have been merged. (5) reparent_to_kthreadd(). This gives the current thread the same credentials as init by simply using commit_thread() to point that way. (6) __sigqueue_alloc() and switch_uid() __sigqueue_alloc() can't stop the target task from changing its creds beneath it, so this function gets a reference to the currently applicable user_struct which it then passes into the sigqueue struct it returns if successful. switch_uid() is now called from commit_creds(), and possibly should be folded into that. commit_creds() should take care of protecting __sigqueue_alloc(). (7) [sg]et[ug]id() and co and [sg]et_current_groups. The set functions now all use prepare_creds(), commit_creds() and abort_creds() to build and check a new set of credentials before applying it. security_task_set[ug]id() is called inside the prepared section. This guarantees that nothing else will affect the creds until we've finished. The calling of set_dumpable() has been moved into commit_creds(). Much of the functionality of set_user() has been moved into commit_creds(). The get functions all simply access the data directly. (8) security_task_prctl() and cap_task_prctl(). security_task_prctl() has been modified to return -ENOSYS if it doesn't want to handle a function, or otherwise return the return value directly rather than through an argument. Additionally, cap_task_prctl() now prepares a new set of credentials, even if it doesn't end up using it. (9) Keyrings. A number of changes have been made to the keyrings code: (a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have all been dropped and built in to the credentials functions directly. They may want separating out again later. (b) key_alloc() and search_process_keyrings() now take a cred pointer rather than a task pointer to specify the security context. (c) copy_creds() gives a new thread within the same thread group a new thread keyring if its parent had one, otherwise it discards the thread keyring. (d) The authorisation key now points directly to the credentials to extend the search into rather pointing to the task that carries them. (e) Installing thread, process or session keyrings causes a new set of credentials to be created, even though it's not strictly necessary for process or session keyrings (they're shared). (10) Usermode helper. The usermode helper code now carries a cred struct pointer in its subprocess_info struct instead of a new session keyring pointer. This set of credentials is derived from init_cred and installed on the new process after it has been cloned. call_usermodehelper_setup() allocates the new credentials and call_usermodehelper_freeinfo() discards them if they haven't been used. A special cred function (prepare_usermodeinfo_creds()) is provided specifically for call_usermodehelper_setup() to call. call_usermodehelper_setkeys() adjusts the credentials to sport the supplied keyring as the new session keyring. (11) SELinux. SELinux has a number of changes, in addition to those to support the LSM interface changes mentioned above: (a) selinux_setprocattr() no longer does its check for whether the current ptracer can access processes with the new SID inside the lock that covers getting the ptracer's SID. Whilst this lock ensures that the check is done with the ptracer pinned, the result is only valid until the lock is released, so there's no point doing it inside the lock. (12) is_single_threaded(). This function has been extracted from selinux_setprocattr() and put into a file of its own in the lib/ directory as join_session_keyring() now wants to use it too. The code in SELinux just checked to see whether a task shared mm_structs with other tasks (CLONE_VM), but that isn't good enough. We really want to know if they're part of the same thread group (CLONE_THREAD). (13) nfsd. The NFS server daemon now has to use the COW credentials to set the credentials it is going to use. It really needs to pass the credentials down to the functions it calls, but it can't do that until other patches in this series have been applied. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: James Morris <jmorris@namei.org> Signed-off-by: James Morris <jmorris@namei.org>
2008-11-13 23:39:23 +00:00
atomic_inc(&init_cred.usage);
commit_creds(&init_cred);
write_unlock_irq(&tasklist_lock);
}
void __set_special_pids(struct pid *pid)
{
struct task_struct *curr = current->group_leader;
connector: fix regression introduced by sid connector Since commit 02b51df1b07b4e9ca823c89284e704cadb323cd1 (proc connector: add event for process becoming session leader) we have the following warning: Badness at kernel/softirq.c:143 [...] Krnl PSW : 0404c00180000000 00000000001481d4 (local_bh_enable+0xb0/0xe0) [...] Call Trace: ([<000000013fe04100>] 0x13fe04100) [<000000000048a946>] sk_filter+0x9a/0xd0 [<000000000049d938>] netlink_broadcast+0x2c0/0x53c [<00000000003ba9ae>] cn_netlink_send+0x272/0x2b0 [<00000000003baef0>] proc_sid_connector+0xc4/0xd4 [<0000000000142604>] __set_special_pids+0x58/0x90 [<0000000000159938>] sys_setsid+0xb4/0xd8 [<00000000001187fe>] sysc_noemu+0x10/0x16 [<00000041616cb266>] 0x41616cb266 The warning is ---> WARN_ON_ONCE(in_irq() || irqs_disabled()); The network code must not be called with disabled interrupts but sys_setsid holds the tasklist_lock with spinlock_irq while calling the connector. After a discussion we agreed that we can move proc_sid_connector from __set_special_pids to sys_setsid. We also agreed that it is sufficient to change the check from task_session(curr) != pid into err > 0, since if we don't change the session, this means we were already the leader and return -EPERM. One last thing: There is also daemonize(), and some people might want to get a notification in that case. Since daemonize() is only needed if a user space does kernel_thread this does not look important (and there seems to be no consensus if this connector should be called in daemonize). If we really want this, we can add proc_sid_connector to daemonize() in an additional patch (Scott?) Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Scott James Remnant <scott@ubuntu.com> Cc: Matt Helsley <matthltc@us.ibm.com> Cc: David S. Miller <davem@davemloft.net> Acked-by: Oleg Nesterov <oleg@redhat.com> Acked-by: Evgeniy Polyakov <zbr@ioremap.net> Acked-by: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-10-26 23:49:34 +00:00
if (task_session(curr) != pid)
change_pid(curr, PIDTYPE_SID, pid);
if (task_pgrp(curr) != pid)
change_pid(curr, PIDTYPE_PGID, pid);
}
static void set_special_pids(struct pid *pid)
{
write_lock_irq(&tasklist_lock);
__set_special_pids(pid);
write_unlock_irq(&tasklist_lock);
}
/*
* Let kernel threads use this to say that they allow a certain signal.
* Must not be used if kthread was cloned with CLONE_SIGHAND.
*/
int allow_signal(int sig)
{
if (!valid_signal(sig) || sig < 1)
return -EINVAL;
spin_lock_irq(&current->sighand->siglock);
/* This is only needed for daemonize()'ed kthreads */
sigdelset(&current->blocked, sig);
/*
* Kernel threads handle their own signals. Let the signal code
* know it'll be handled, so that they don't get converted to
* SIGKILL or just silently dropped.
*/
current->sighand->action[(sig)-1].sa.sa_handler = (void __user *)2;
recalc_sigpending();
spin_unlock_irq(&current->sighand->siglock);
return 0;
}
EXPORT_SYMBOL(allow_signal);
int disallow_signal(int sig)
{
if (!valid_signal(sig) || sig < 1)
return -EINVAL;
spin_lock_irq(&current->sighand->siglock);
current->sighand->action[(sig)-1].sa.sa_handler = SIG_IGN;
recalc_sigpending();
spin_unlock_irq(&current->sighand->siglock);
return 0;
}
EXPORT_SYMBOL(disallow_signal);
/*
* Put all the gunge required to become a kernel thread without
* attached user resources in one place where it belongs.
*/
void daemonize(const char *name, ...)
{
va_list args;
sigset_t blocked;
va_start(args, name);
vsnprintf(current->comm, sizeof(current->comm), name, args);
va_end(args);
/*
* If we were started as result of loading a module, close all of the
* user space pages. We don't need them, and if we didn't close them
* they would be locked into memory.
*/
exit_mm(current);
/*
* We don't want to get frozen, in case system-wide hibernation
* or suspend transition begins right now.
*/
current->flags |= (PF_NOFREEZE | PF_KTHREAD);
if (current->nsproxy != &init_nsproxy) {
get_nsproxy(&init_nsproxy);
switch_task_namespaces(current, &init_nsproxy);
}
set_special_pids(&init_struct_pid);
[PATCH] tty: ->signal->tty locking Fix the locking of signal->tty. Use ->sighand->siglock to protect ->signal->tty; this lock is already used by most other members of ->signal/->sighand. And unless we are 'current' or the tasklist_lock is held we need ->siglock to access ->signal anyway. (NOTE: sys_unshare() is broken wrt ->sighand locking rules) Note that tty_mutex is held over tty destruction, so while holding tty_mutex any tty pointer remains valid. Otherwise the lifetime of ttys are governed by their open file handles. This leaves some holes for tty access from signal->tty (or any other non file related tty access). It solves the tty SLAB scribbles we were seeing. (NOTE: the change from group_send_sig_info to __group_send_sig_info needs to be examined by someone familiar with the security framework, I think it is safe given the SEND_SIG_PRIV from other __group_send_sig_info invocations) [schwidefsky@de.ibm.com: 3270 fix] [akpm@osdl.org: various post-viro fixes] Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Alan Cox <alan@redhat.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Cc: Prarit Bhargava <prarit@redhat.com> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Roland McGrath <roland@redhat.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Jan Kara <jack@ucw.cz> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-08 10:36:04 +00:00
proc_clear_tty(current);
/* Block and flush all signals */
sigfillset(&blocked);
sigprocmask(SIG_BLOCK, &blocked, NULL);
flush_signals(current);
/* Become as one with the init task */
daemonize_fs_struct();
daemonize_descriptors();
reparent_to_kthreadd();
}
EXPORT_SYMBOL(daemonize);
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
#ifdef CONFIG_MM_OWNER
/*
memcg: clear mm->owner when last possible owner leaves The following crash was reported: > Call Trace: > [<ffffffff81139792>] mem_cgroup_from_task+0x15/0x17 > [<ffffffff8113a75a>] __mem_cgroup_try_charge+0x148/0x4b4 > [<ffffffff810493f3>] ? need_resched+0x23/0x2d > [<ffffffff814cbf43>] ? preempt_schedule+0x46/0x4f > [<ffffffff8113afe8>] mem_cgroup_charge_common+0x9a/0xce > [<ffffffff8113b6d1>] mem_cgroup_newpage_charge+0x5d/0x5f > [<ffffffff81134024>] khugepaged+0x5da/0xfaf > [<ffffffff81078ea0>] ? __init_waitqueue_head+0x4b/0x4b > [<ffffffff81133a4a>] ? add_mm_counter.constprop.5+0x13/0x13 > [<ffffffff81078625>] kthread+0xa8/0xb0 > [<ffffffff814d13e8>] ? sub_preempt_count+0xa1/0xb4 > [<ffffffff814d5664>] kernel_thread_helper+0x4/0x10 > [<ffffffff814ce858>] ? retint_restore_args+0x13/0x13 > [<ffffffff8107857d>] ? __init_kthread_worker+0x5a/0x5a What happens is that khugepaged tries to charge a huge page against an mm whose last possible owner has already exited, and the memory controller crashes when the stale mm->owner is used to look up the cgroup to charge. mm->owner has never been set to NULL with the last owner going away, but nobody cared until khugepaged came along. Even then it wasn't a problem because the final mmput() on an mm was forced to acquire and release mmap_sem in write-mode, preventing an exiting owner to go away while the mmap_sem was held, and until "692e0b3 mm: thp: optimize memcg charge in khugepaged", the memory cgroup charge was protected by mmap_sem in read-mode. Instead of going back to relying on the mmap_sem to enforce lifetime of a task, this patch ensures that mm->owner is properly set to NULL when the last possible owner is exiting, which the memory controller can handle just fine. [akpm@linux-foundation.org: tweak comments] Signed-off-by: Hugh Dickins <hughd@google.com> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reported-by: Hugh Dickins <hughd@google.com> Reported-by: Dave Jones <davej@redhat.com> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-15 22:08:43 +00:00
* A task is exiting. If it owned this mm, find a new owner for the mm.
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
*/
void mm_update_next_owner(struct mm_struct *mm)
{
struct task_struct *c, *g, *p = current;
retry:
memcg: clear mm->owner when last possible owner leaves The following crash was reported: > Call Trace: > [<ffffffff81139792>] mem_cgroup_from_task+0x15/0x17 > [<ffffffff8113a75a>] __mem_cgroup_try_charge+0x148/0x4b4 > [<ffffffff810493f3>] ? need_resched+0x23/0x2d > [<ffffffff814cbf43>] ? preempt_schedule+0x46/0x4f > [<ffffffff8113afe8>] mem_cgroup_charge_common+0x9a/0xce > [<ffffffff8113b6d1>] mem_cgroup_newpage_charge+0x5d/0x5f > [<ffffffff81134024>] khugepaged+0x5da/0xfaf > [<ffffffff81078ea0>] ? __init_waitqueue_head+0x4b/0x4b > [<ffffffff81133a4a>] ? add_mm_counter.constprop.5+0x13/0x13 > [<ffffffff81078625>] kthread+0xa8/0xb0 > [<ffffffff814d13e8>] ? sub_preempt_count+0xa1/0xb4 > [<ffffffff814d5664>] kernel_thread_helper+0x4/0x10 > [<ffffffff814ce858>] ? retint_restore_args+0x13/0x13 > [<ffffffff8107857d>] ? __init_kthread_worker+0x5a/0x5a What happens is that khugepaged tries to charge a huge page against an mm whose last possible owner has already exited, and the memory controller crashes when the stale mm->owner is used to look up the cgroup to charge. mm->owner has never been set to NULL with the last owner going away, but nobody cared until khugepaged came along. Even then it wasn't a problem because the final mmput() on an mm was forced to acquire and release mmap_sem in write-mode, preventing an exiting owner to go away while the mmap_sem was held, and until "692e0b3 mm: thp: optimize memcg charge in khugepaged", the memory cgroup charge was protected by mmap_sem in read-mode. Instead of going back to relying on the mmap_sem to enforce lifetime of a task, this patch ensures that mm->owner is properly set to NULL when the last possible owner is exiting, which the memory controller can handle just fine. [akpm@linux-foundation.org: tweak comments] Signed-off-by: Hugh Dickins <hughd@google.com> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reported-by: Hugh Dickins <hughd@google.com> Reported-by: Dave Jones <davej@redhat.com> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-15 22:08:43 +00:00
/*
* If the exiting or execing task is not the owner, it's
* someone else's problem.
*/
if (mm->owner != p)
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
return;
memcg: clear mm->owner when last possible owner leaves The following crash was reported: > Call Trace: > [<ffffffff81139792>] mem_cgroup_from_task+0x15/0x17 > [<ffffffff8113a75a>] __mem_cgroup_try_charge+0x148/0x4b4 > [<ffffffff810493f3>] ? need_resched+0x23/0x2d > [<ffffffff814cbf43>] ? preempt_schedule+0x46/0x4f > [<ffffffff8113afe8>] mem_cgroup_charge_common+0x9a/0xce > [<ffffffff8113b6d1>] mem_cgroup_newpage_charge+0x5d/0x5f > [<ffffffff81134024>] khugepaged+0x5da/0xfaf > [<ffffffff81078ea0>] ? __init_waitqueue_head+0x4b/0x4b > [<ffffffff81133a4a>] ? add_mm_counter.constprop.5+0x13/0x13 > [<ffffffff81078625>] kthread+0xa8/0xb0 > [<ffffffff814d13e8>] ? sub_preempt_count+0xa1/0xb4 > [<ffffffff814d5664>] kernel_thread_helper+0x4/0x10 > [<ffffffff814ce858>] ? retint_restore_args+0x13/0x13 > [<ffffffff8107857d>] ? __init_kthread_worker+0x5a/0x5a What happens is that khugepaged tries to charge a huge page against an mm whose last possible owner has already exited, and the memory controller crashes when the stale mm->owner is used to look up the cgroup to charge. mm->owner has never been set to NULL with the last owner going away, but nobody cared until khugepaged came along. Even then it wasn't a problem because the final mmput() on an mm was forced to acquire and release mmap_sem in write-mode, preventing an exiting owner to go away while the mmap_sem was held, and until "692e0b3 mm: thp: optimize memcg charge in khugepaged", the memory cgroup charge was protected by mmap_sem in read-mode. Instead of going back to relying on the mmap_sem to enforce lifetime of a task, this patch ensures that mm->owner is properly set to NULL when the last possible owner is exiting, which the memory controller can handle just fine. [akpm@linux-foundation.org: tweak comments] Signed-off-by: Hugh Dickins <hughd@google.com> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reported-by: Hugh Dickins <hughd@google.com> Reported-by: Dave Jones <davej@redhat.com> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-15 22:08:43 +00:00
/*
* The current owner is exiting/execing and there are no other
* candidates. Do not leave the mm pointing to a possibly
* freed task structure.
*/
if (atomic_read(&mm->mm_users) <= 1) {
mm->owner = NULL;
return;
}
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
read_lock(&tasklist_lock);
/*
* Search in the children
*/
list_for_each_entry(c, &p->children, sibling) {
if (c->mm == mm)
goto assign_new_owner;
}
/*
* Search in the siblings
*/
list_for_each_entry(c, &p->real_parent->children, sibling) {
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
if (c->mm == mm)
goto assign_new_owner;
}
/*
* Search through everything else. We should not get
* here often
*/
do_each_thread(g, c) {
if (c->mm == mm)
goto assign_new_owner;
} while_each_thread(g, c);
read_unlock(&tasklist_lock);
mm owner: fix race between swapoff and exit There's a race between mm->owner assignment and swapoff, more easily seen when task slab poisoning is turned on. The condition occurs when try_to_unuse() runs in parallel with an exiting task. A similar race can occur with callers of get_task_mm(), such as /proc/<pid>/<mmstats> or ptrace or page migration. CPU0 CPU1 try_to_unuse looks at mm = task0->mm increments mm->mm_users task 0 exits mm->owner needs to be updated, but no new owner is found (mm_users > 1, but no other task has task->mm = task0->mm) mm_update_next_owner() leaves mmput(mm) decrements mm->mm_users task0 freed dereferencing mm->owner fails The fix is to notify the subsystem via mm_owner_changed callback(), if no new owner is found, by specifying the new task as NULL. Jiri Slaby: mm->owner was set to NULL prior to calling cgroup_mm_owner_callbacks(), but must be set after that, so as not to pass NULL as old owner causing oops. Daisuke Nishimura: mm_update_next_owner() may set mm->owner to NULL, but mem_cgroup_from_task() and its callers need to take account of this situation to avoid oops. Hugh Dickins: Lockdep warning and hang below exec_mmap() when testing these patches. exit_mm() up_reads mmap_sem before calling mm_update_next_owner(), so exec_mmap() now needs to do the same. And with that repositioning, there's now no point in mm_need_new_owner() allowing for NULL mm. Reported-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Signed-off-by: Jiri Slaby <jirislaby@gmail.com> Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Signed-off-by: Hugh Dickins <hugh@veritas.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Paul Menage <menage@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-28 22:09:31 +00:00
/*
* We found no owner yet mm_users > 1: this implies that we are
* most likely racing with swapoff (try_to_unuse()) or /proc or
* ptrace or page migration (get_task_mm()). Mark owner as NULL.
mm owner: fix race between swapoff and exit There's a race between mm->owner assignment and swapoff, more easily seen when task slab poisoning is turned on. The condition occurs when try_to_unuse() runs in parallel with an exiting task. A similar race can occur with callers of get_task_mm(), such as /proc/<pid>/<mmstats> or ptrace or page migration. CPU0 CPU1 try_to_unuse looks at mm = task0->mm increments mm->mm_users task 0 exits mm->owner needs to be updated, but no new owner is found (mm_users > 1, but no other task has task->mm = task0->mm) mm_update_next_owner() leaves mmput(mm) decrements mm->mm_users task0 freed dereferencing mm->owner fails The fix is to notify the subsystem via mm_owner_changed callback(), if no new owner is found, by specifying the new task as NULL. Jiri Slaby: mm->owner was set to NULL prior to calling cgroup_mm_owner_callbacks(), but must be set after that, so as not to pass NULL as old owner causing oops. Daisuke Nishimura: mm_update_next_owner() may set mm->owner to NULL, but mem_cgroup_from_task() and its callers need to take account of this situation to avoid oops. Hugh Dickins: Lockdep warning and hang below exec_mmap() when testing these patches. exit_mm() up_reads mmap_sem before calling mm_update_next_owner(), so exec_mmap() now needs to do the same. And with that repositioning, there's now no point in mm_need_new_owner() allowing for NULL mm. Reported-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Signed-off-by: Jiri Slaby <jirislaby@gmail.com> Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Signed-off-by: Hugh Dickins <hugh@veritas.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Paul Menage <menage@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-28 22:09:31 +00:00
*/
mm->owner = NULL;
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
return;
assign_new_owner:
BUG_ON(c == p);
get_task_struct(c);
/*
* The task_lock protects c->mm from changing.
* We always want mm->owner->mm == mm
*/
task_lock(c);
/*
* Delay read_unlock() till we have the task_lock()
* to ensure that c does not slip away underneath us
*/
read_unlock(&tasklist_lock);
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
if (c->mm != mm) {
task_unlock(c);
put_task_struct(c);
goto retry;
}
mm->owner = c;
task_unlock(c);
put_task_struct(c);
}
#endif /* CONFIG_MM_OWNER */
/*
* Turn us into a lazy TLB process if we
* aren't already..
*/
static void exit_mm(struct task_struct * tsk)
{
struct mm_struct *mm = tsk->mm;
struct core_state *core_state;
mm_release(tsk, mm);
if (!mm)
return;
sync_mm_rss(mm);
/*
* Serialize with any possible pending coredump.
* We must hold mmap_sem around checking core_state
* and clearing tsk->mm. The core-inducing thread
* will increment ->nr_threads for each thread in the
* group with ->mm != NULL.
*/
down_read(&mm->mmap_sem);
core_state = mm->core_state;
if (core_state) {
struct core_thread self;
up_read(&mm->mmap_sem);
self.task = tsk;
self.next = xchg(&core_state->dumper.next, &self);
/*
* Implies mb(), the result of xchg() must be visible
* to core_state->dumper.
*/
if (atomic_dec_and_test(&core_state->nr_threads))
complete(&core_state->startup);
for (;;) {
set_task_state(tsk, TASK_UNINTERRUPTIBLE);
if (!self.task) /* see coredump_finish() */
break;
schedule();
}
__set_task_state(tsk, TASK_RUNNING);
down_read(&mm->mmap_sem);
}
atomic_inc(&mm->mm_count);
BUG_ON(mm != tsk->active_mm);
/* more a memory barrier than a real lock */
task_lock(tsk);
tsk->mm = NULL;
up_read(&mm->mmap_sem);
enter_lazy_tlb(mm, current);
task_unlock(tsk);
cgroups: add an owner to the mm_struct Remove the mem_cgroup member from mm_struct and instead adds an owner. This approach was suggested by Paul Menage. The advantage of this approach is that, once the mm->owner is known, using the subsystem id, the cgroup can be determined. It also allows several control groups that are virtually grouped by mm_struct, to exist independent of the memory controller i.e., without adding mem_cgroup's for each controller, to mm_struct. A new config option CONFIG_MM_OWNER is added and the memory resource controller selects this config option. This patch also adds cgroup callbacks to notify subsystems when mm->owner changes. The mm_cgroup_changed callback is called with the task_lock() of the new task held and is called just prior to changing the mm->owner. I am indebted to Paul Menage for the several reviews of this patchset and helping me make it lighter and simpler. This patch was tested on a powerpc box, it was compiled with both the MM_OWNER config turned on and off. After the thread group leader exits, it's moved to init_css_state by cgroup_exit(), thus all future charges from runnings threads would be redirected to the init_css_set's subsystem. Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com> Cc: Pavel Emelianov <xemul@openvz.org> Cc: Hugh Dickins <hugh@veritas.com> Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp> Cc: Hirokazu Takahashi <taka@valinux.co.jp> Cc: David Rientjes <rientjes@google.com>, Cc: Balbir Singh <balbir@linux.vnet.ibm.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Reviewed-by: Paul Menage <menage@google.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 08:00:16 +00:00
mm_update_next_owner(mm);
mmput(mm);
}
/*
prctl: add PR_{SET,GET}_CHILD_SUBREAPER to allow simple process supervision Userspace service managers/supervisors need to track their started services. Many services daemonize by double-forking and get implicitly re-parented to PID 1. The service manager will no longer be able to receive the SIGCHLD signals for them, and is no longer in charge of reaping the children with wait(). All information about the children is lost at the moment PID 1 cleans up the re-parented processes. With this prctl, a service manager process can mark itself as a sort of 'sub-init', able to stay as the parent for all orphaned processes created by the started services. All SIGCHLD signals will be delivered to the service manager. Receiving SIGCHLD and doing wait() is in cases of a service-manager much preferred over any possible asynchronous notification about specific PIDs, because the service manager has full access to the child process data in /proc and the PID can not be re-used until the wait(), the service-manager itself is in charge of, has happened. As a side effect, the relevant parent PID information does not get lost by a double-fork, which results in a more elaborate process tree and 'ps' output: before: # ps afx 253 ? Ss 0:00 /bin/dbus-daemon --system --nofork 294 ? Sl 0:00 /usr/libexec/polkit-1/polkitd 328 ? S 0:00 /usr/sbin/modem-manager 608 ? Sl 0:00 /usr/libexec/colord 658 ? Sl 0:00 /usr/libexec/upowerd 819 ? Sl 0:00 /usr/libexec/imsettings-daemon 916 ? Sl 0:00 /usr/libexec/udisks-daemon 917 ? S 0:00 \_ udisks-daemon: not polling any devices after: # ps afx 294 ? Ss 0:00 /bin/dbus-daemon --system --nofork 426 ? Sl 0:00 \_ /usr/libexec/polkit-1/polkitd 449 ? S 0:00 \_ /usr/sbin/modem-manager 635 ? Sl 0:00 \_ /usr/libexec/colord 705 ? Sl 0:00 \_ /usr/libexec/upowerd 959 ? Sl 0:00 \_ /usr/libexec/udisks-daemon 960 ? S 0:00 | \_ udisks-daemon: not polling any devices 977 ? Sl 0:00 \_ /usr/libexec/packagekitd This prctl is orthogonal to PID namespaces. PID namespaces are isolated from each other, while a service management process usually requires the services to live in the same namespace, to be able to talk to each other. Users of this will be the systemd per-user instance, which provides init-like functionality for the user's login session and D-Bus, which activates bus services on-demand. Both need init-like capabilities to be able to properly keep track of the services they start. Many thanks to Oleg for several rounds of review and insights. [akpm@linux-foundation.org: fix comment layout and spelling] [akpm@linux-foundation.org: add lengthy code comment from Oleg] Reviewed-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Lennart Poettering <lennart@poettering.net> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Acked-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-23 22:01:54 +00:00
* When we die, we re-parent all our children, and try to:
* 1. give them to another thread in our thread group, if such a member exists
* 2. give it to the first ancestor process which prctl'd itself as a
* child_subreaper for its children (like a service manager)
* 3. give it to the init process (PID 1) in our pid namespace
*/
static struct task_struct *find_new_reaper(struct task_struct *father)
__releases(&tasklist_lock)
__acquires(&tasklist_lock)
{
struct pid_namespace *pid_ns = task_active_pid_ns(father);
struct task_struct *thread;
thread = father;
while_each_thread(father, thread) {
if (thread->flags & PF_EXITING)
continue;
if (unlikely(pid_ns->child_reaper == father))
pid_ns->child_reaper = thread;
return thread;
}
if (unlikely(pid_ns->child_reaper == father)) {
write_unlock_irq(&tasklist_lock);
if (unlikely(pid_ns == &init_pid_ns)) {
panic("Attempted to kill init! exitcode=0x%08x\n",
father->signal->group_exit_code ?:
father->exit_code);
}
zap_pid_ns_processes(pid_ns);
write_lock_irq(&tasklist_lock);
prctl: add PR_{SET,GET}_CHILD_SUBREAPER to allow simple process supervision Userspace service managers/supervisors need to track their started services. Many services daemonize by double-forking and get implicitly re-parented to PID 1. The service manager will no longer be able to receive the SIGCHLD signals for them, and is no longer in charge of reaping the children with wait(). All information about the children is lost at the moment PID 1 cleans up the re-parented processes. With this prctl, a service manager process can mark itself as a sort of 'sub-init', able to stay as the parent for all orphaned processes created by the started services. All SIGCHLD signals will be delivered to the service manager. Receiving SIGCHLD and doing wait() is in cases of a service-manager much preferred over any possible asynchronous notification about specific PIDs, because the service manager has full access to the child process data in /proc and the PID can not be re-used until the wait(), the service-manager itself is in charge of, has happened. As a side effect, the relevant parent PID information does not get lost by a double-fork, which results in a more elaborate process tree and 'ps' output: before: # ps afx 253 ? Ss 0:00 /bin/dbus-daemon --system --nofork 294 ? Sl 0:00 /usr/libexec/polkit-1/polkitd 328 ? S 0:00 /usr/sbin/modem-manager 608 ? Sl 0:00 /usr/libexec/colord 658 ? Sl 0:00 /usr/libexec/upowerd 819 ? Sl 0:00 /usr/libexec/imsettings-daemon 916 ? Sl 0:00 /usr/libexec/udisks-daemon 917 ? S 0:00 \_ udisks-daemon: not polling any devices after: # ps afx 294 ? Ss 0:00 /bin/dbus-daemon --system --nofork 426 ? Sl 0:00 \_ /usr/libexec/polkit-1/polkitd 449 ? S 0:00 \_ /usr/sbin/modem-manager 635 ? Sl 0:00 \_ /usr/libexec/colord 705 ? Sl 0:00 \_ /usr/libexec/upowerd 959 ? Sl 0:00 \_ /usr/libexec/udisks-daemon 960 ? S 0:00 | \_ udisks-daemon: not polling any devices 977 ? Sl 0:00 \_ /usr/libexec/packagekitd This prctl is orthogonal to PID namespaces. PID namespaces are isolated from each other, while a service management process usually requires the services to live in the same namespace, to be able to talk to each other. Users of this will be the systemd per-user instance, which provides init-like functionality for the user's login session and D-Bus, which activates bus services on-demand. Both need init-like capabilities to be able to properly keep track of the services they start. Many thanks to Oleg for several rounds of review and insights. [akpm@linux-foundation.org: fix comment layout and spelling] [akpm@linux-foundation.org: add lengthy code comment from Oleg] Reviewed-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Lennart Poettering <lennart@poettering.net> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Acked-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-23 22:01:54 +00:00
} else if (father->signal->has_child_subreaper) {
struct task_struct *reaper;
/*
* Find the first ancestor marked as child_subreaper.
* Note that the code below checks same_thread_group(reaper,
* pid_ns->child_reaper). This is what we need to DTRT in a
* PID namespace. However we still need the check above, see
* http://marc.info/?l=linux-kernel&m=131385460420380
*/
for (reaper = father->real_parent;
reaper != &init_task;
reaper = reaper->real_parent) {
if (same_thread_group(reaper, pid_ns->child_reaper))
break;
if (!reaper->signal->is_child_subreaper)
continue;
thread = reaper;
do {
if (!(thread->flags & PF_EXITING))
return reaper;
} while_each_thread(reaper, thread);
}
}
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
return pid_ns->child_reaper;
}
/*
* Any that need to be release_task'd are put on the @dead list.
*/
static void reparent_leader(struct task_struct *father, struct task_struct *p,
struct list_head *dead)
{
list_move_tail(&p->sibling, &p->real_parent->children);
if (p->exit_state == EXIT_DEAD)
return;
/*
* If this is a threaded reparent there is no need to
* notify anyone anything has happened.
*/
if (same_thread_group(p->real_parent, father))
return;
/* We don't want people slaying init. */
p->exit_signal = SIGCHLD;
/* If it has exited notify the new parent about this child's death. */
if (!p->ptrace &&
p->exit_state == EXIT_ZOMBIE && thread_group_empty(p)) {
if (do_notify_parent(p, p->exit_signal)) {
p->exit_state = EXIT_DEAD;
list_move_tail(&p->sibling, dead);
}
}
kill_orphaned_pgrp(p, father);
}
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
static void forget_original_parent(struct task_struct *father)
{
struct task_struct *p, *n, *reaper;
LIST_HEAD(dead_children);
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
write_lock_irq(&tasklist_lock);
/*
* Note that exit_ptrace() and find_new_reaper() might
* drop tasklist_lock and reacquire it.
*/
exit_ptrace(father);
reaper = find_new_reaper(father);
list_for_each_entry_safe(p, n, &father->children, sibling) {
struct task_struct *t = p;
do {
t->real_parent = reaper;
if (t->parent == father) {
BUG_ON(t->ptrace);
t->parent = t->real_parent;
}
if (t->pdeath_signal)
group_send_sig_info(t->pdeath_signal,
SEND_SIG_NOINFO, t);
} while_each_thread(p, t);
reparent_leader(father, p, &dead_children);
}
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
write_unlock_irq(&tasklist_lock);
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
BUG_ON(!list_empty(&father->children));
list_for_each_entry_safe(p, n, &dead_children, sibling) {
list_del_init(&p->sibling);
release_task(p);
}
}
/*
* Send signals to all our closest relatives so that they know
* to properly mourn us..
*/
static void exit_notify(struct task_struct *tsk, int group_dead)
{
bool autoreap;
/*
* This does two things:
*
* A. Make init inherit all the child processes
* B. Check to see if any process groups have become orphaned
* as a result of our exiting, and if they have any stopped
* jobs, send them a SIGHUP and then a SIGCONT. (POSIX 3.2.2.2)
*/
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
forget_original_parent(tsk);
exit_task_namespaces(tsk);
pid namespaces: rework forget_original_parent() A pid namespace is a "view" of a particular set of tasks on the system. They work in a similar way to filesystem namespaces. A file (or a process) can be accessed in multiple namespaces, but it may have a different name in each. In a filesystem, this name might be /etc/passwd in one namespace, but /chroot/etc/passwd in another. For processes, a process may have pid 1234 in one namespace, but be pid 1 in another. This allows new pid namespaces to have basically arbitrary pids, and not have to worry about what pids exist in other namespaces. This is essential for checkpoint/restart where a restarted process's pid might collide with an existing process on the system's pid. In this particular implementation, pid namespaces have a parent-child relationship, just like processes. A process in a pid namespace may see all of the processes in the same namespace, as well as all of the processes in all of the namespaces which are children of its namespace. Processes may not, however, see others which are in their parent's namespace, but not in their own. The same goes for sibling namespaces. The know issue to be solved in the nearest future is signal handling in the namespace boundary. That is, currently the namespace's init is treated like an ordinary task that can be killed from within an namespace. Ideally, the signal handling by the namespace's init should have two sides: when signaling the init from its namespace, the init should look like a real init task, i.e. receive only those signals, that is explicitly wants to; when signaling the init from one of the parent namespaces, init should look like an ordinary task, i.e. receive any signal, only taking the general permissions into account. The pid namespace was developed by Pavel Emlyanov and Sukadev Bhattiprolu and we eventually came to almost the same implementation, which differed in some details. This set is based on Pavel's patches, but it includes comments and patches that from Sukadev. Many thanks to Oleg, who reviewed the patches, pointed out many BUGs and made valuable advises on how to make this set cleaner. This patch: We have to call exit_task_namespaces() only after the exiting task has reparented all his children and is sure that no other threads will reparent theirs for it. Why this is needed is explained in appropriate patch. This one only reworks the forget_original_parent() so that after calling this a task cannot be/become parent of any other task. We check PF_EXITING instead of ->exit_state while choosing the new parent. Note that tasklits_lock acts as a barrier, everyone who takes tasklist after us (when forget_original_parent() drops it) must see PF_EXITING. The other changes are just cleanups. They just move some code from exit_notify to forget_original_parent(). It is a bit silly to declare ptrace_dead in exit_notify(), take tasklist, pass ptrace_dead to forget_original_parent(), unlock-lock-unlock tasklist, and then use ptrace_dead. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Cc: Paul Menage <menage@google.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:40:00 +00:00
write_lock_irq(&tasklist_lock);
if (group_dead)
kill_orphaned_pgrp(tsk->group_leader, NULL);
if (unlikely(tsk->ptrace)) {
int sig = thread_group_leader(tsk) &&
thread_group_empty(tsk) &&
!ptrace_reparented(tsk) ?
tsk->exit_signal : SIGCHLD;
autoreap = do_notify_parent(tsk, sig);
} else if (thread_group_leader(tsk)) {
autoreap = thread_group_empty(tsk) &&
do_notify_parent(tsk, tsk->exit_signal);
} else {
autoreap = true;
}
tsk->exit_state = autoreap ? EXIT_DEAD : EXIT_ZOMBIE;
/* mt-exec, de_thread() is waiting for group leader */
if (unlikely(tsk->signal->notify_count < 0))
wake_up_process(tsk->signal->group_exit_task);
write_unlock_irq(&tasklist_lock);
/* If the process is dead, release it - nobody will wait for it */
if (autoreap)
release_task(tsk);
}
#ifdef CONFIG_DEBUG_STACK_USAGE
static void check_stack_usage(void)
{
static DEFINE_SPINLOCK(low_water_lock);
static int lowest_to_date = THREAD_SIZE;
unsigned long free;
free = stack_not_used(current);
if (free >= lowest_to_date)
return;
spin_lock(&low_water_lock);
if (free < lowest_to_date) {
printk(KERN_WARNING "%s (%d) used greatest stack depth: "
"%lu bytes left\n",
current->comm, task_pid_nr(current), free);
lowest_to_date = free;
}
spin_unlock(&low_water_lock);
}
#else
static inline void check_stack_usage(void) {}
#endif
void do_exit(long code)
{
struct task_struct *tsk = current;
int group_dead;
profile_task_exit(tsk);
WARN_ON(blk_needs_flush_plug(tsk));
if (unlikely(in_interrupt()))
panic("Aiee, killing interrupt handler!");
if (unlikely(!tsk->pid))
panic("Attempted to kill the idle task!");
/*
* If do_exit is called because this processes oopsed, it's possible
* that get_fs() was left as KERNEL_DS, so reset it to USER_DS before
* continuing. Amongst other possible reasons, this is to prevent
* mm_release()->clear_child_tid() from writing to a user-controlled
* kernel address.
*/
set_fs(USER_DS);
ptrace_event(PTRACE_EVENT_EXIT, code);
validate_creds_for_do_exit(tsk);
/*
* We're taking recursive faults here in do_exit. Safest is to just
* leave this task alone and wait for reboot.
*/
if (unlikely(tsk->flags & PF_EXITING)) {
printk(KERN_ALERT
"Fixing recursive fault but reboot is needed!\n");
pi-futex: fix exit races and locking problems 1. New entries can be added to tsk->pi_state_list after task completed exit_pi_state_list(). The result is memory leakage and deadlocks. 2. handle_mm_fault() is called under spinlock. The result is obvious. 3. results in self-inflicted deadlock inside glibc. Sometimes futex_lock_pi returns -ESRCH, when it is not expected and glibc enters to for(;;) sleep() to simulate deadlock. This problem is quite obvious and I think the patch is right. Though it looks like each "if" in futex_lock_pi() got some stupid special case "else if". :-) 4. sometimes futex_lock_pi() returns -EDEADLK, when nobody has the lock. The reason is also obvious (see comment in the patch), but correct fix is far beyond my comprehension. I guess someone already saw this, the chunk: if (rt_mutex_trylock(&q.pi_state->pi_mutex)) ret = 0; is obviously from the same opera. But it does not work, because the rtmutex is really taken at this point: wake_futex_pi() of previous owner reassigned it to us. My fix works. But it looks very stupid. I would think about removal of shift of ownership in wake_futex_pi() and making all the work in context of process taking lock. From: Thomas Gleixner <tglx@linutronix.de> Fix 1) Avoid the tasklist lock variant of the exit race fix by adding an additional state transition to the exit code. This fixes also the issue, when a task with recursive segfaults is not able to release the futexes. Fix 2) Cleanup the lookup_pi_state() failure path and solve the -ESRCH problem finally. Fix 3) Solve the fixup_pi_state_owner() problem which needs to do the fixup in the lock protected section by using the in_atomic userspace access functions. This removes also the ugly lock drop / unqueue inside of fixup_pi_state() Fix 4) Fix a stale lock in the error path of futex_wake_pi() Added some error checks for verification. The -EDEADLK problem is solved by the rtmutex fixups. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Ingo Molnar <mingo@elte.hu> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-06-08 20:47:00 +00:00
/*
* We can do this unlocked here. The futex code uses
* this flag just to verify whether the pi state
* cleanup has been done or not. In the worst case it
* loops once more. We pretend that the cleanup was
* done as there is no way to return. Either the
* OWNER_DIED bit is set by now or we push the blocked
* task into the wait for ever nirwana as well.
*/
tsk->flags |= PF_EXITPIDONE;
set_current_state(TASK_UNINTERRUPTIBLE);
schedule();
}
exit_signals(tsk); /* sets PF_EXITING */
pi-futex: fix exit races and locking problems 1. New entries can be added to tsk->pi_state_list after task completed exit_pi_state_list(). The result is memory leakage and deadlocks. 2. handle_mm_fault() is called under spinlock. The result is obvious. 3. results in self-inflicted deadlock inside glibc. Sometimes futex_lock_pi returns -ESRCH, when it is not expected and glibc enters to for(;;) sleep() to simulate deadlock. This problem is quite obvious and I think the patch is right. Though it looks like each "if" in futex_lock_pi() got some stupid special case "else if". :-) 4. sometimes futex_lock_pi() returns -EDEADLK, when nobody has the lock. The reason is also obvious (see comment in the patch), but correct fix is far beyond my comprehension. I guess someone already saw this, the chunk: if (rt_mutex_trylock(&q.pi_state->pi_mutex)) ret = 0; is obviously from the same opera. But it does not work, because the rtmutex is really taken at this point: wake_futex_pi() of previous owner reassigned it to us. My fix works. But it looks very stupid. I would think about removal of shift of ownership in wake_futex_pi() and making all the work in context of process taking lock. From: Thomas Gleixner <tglx@linutronix.de> Fix 1) Avoid the tasklist lock variant of the exit race fix by adding an additional state transition to the exit code. This fixes also the issue, when a task with recursive segfaults is not able to release the futexes. Fix 2) Cleanup the lookup_pi_state() failure path and solve the -ESRCH problem finally. Fix 3) Solve the fixup_pi_state_owner() problem which needs to do the fixup in the lock protected section by using the in_atomic userspace access functions. This removes also the ugly lock drop / unqueue inside of fixup_pi_state() Fix 4) Fix a stale lock in the error path of futex_wake_pi() Added some error checks for verification. The -EDEADLK problem is solved by the rtmutex fixups. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Ingo Molnar <mingo@elte.hu> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-06-08 20:47:00 +00:00
/*
* tsk->flags are checked in the futex code to protect against
* an exiting task cleaning up the robust pi futexes.
pi-futex: fix exit races and locking problems 1. New entries can be added to tsk->pi_state_list after task completed exit_pi_state_list(). The result is memory leakage and deadlocks. 2. handle_mm_fault() is called under spinlock. The result is obvious. 3. results in self-inflicted deadlock inside glibc. Sometimes futex_lock_pi returns -ESRCH, when it is not expected and glibc enters to for(;;) sleep() to simulate deadlock. This problem is quite obvious and I think the patch is right. Though it looks like each "if" in futex_lock_pi() got some stupid special case "else if". :-) 4. sometimes futex_lock_pi() returns -EDEADLK, when nobody has the lock. The reason is also obvious (see comment in the patch), but correct fix is far beyond my comprehension. I guess someone already saw this, the chunk: if (rt_mutex_trylock(&q.pi_state->pi_mutex)) ret = 0; is obviously from the same opera. But it does not work, because the rtmutex is really taken at this point: wake_futex_pi() of previous owner reassigned it to us. My fix works. But it looks very stupid. I would think about removal of shift of ownership in wake_futex_pi() and making all the work in context of process taking lock. From: Thomas Gleixner <tglx@linutronix.de> Fix 1) Avoid the tasklist lock variant of the exit race fix by adding an additional state transition to the exit code. This fixes also the issue, when a task with recursive segfaults is not able to release the futexes. Fix 2) Cleanup the lookup_pi_state() failure path and solve the -ESRCH problem finally. Fix 3) Solve the fixup_pi_state_owner() problem which needs to do the fixup in the lock protected section by using the in_atomic userspace access functions. This removes also the ugly lock drop / unqueue inside of fixup_pi_state() Fix 4) Fix a stale lock in the error path of futex_wake_pi() Added some error checks for verification. The -EDEADLK problem is solved by the rtmutex fixups. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Ingo Molnar <mingo@elte.hu> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-06-08 20:47:00 +00:00
*/
smp_mb();
raw_spin_unlock_wait(&tsk->pi_lock);
if (unlikely(in_atomic()))
printk(KERN_INFO "note: %s[%d] exited with preempt_count %d\n",
current->comm, task_pid_nr(current),
preempt_count());
acct_update_integrals(tsk);
/* sync mm's RSS info before statistics gathering */
if (tsk->mm)
sync_mm_rss(tsk->mm);
group_dead = atomic_dec_and_test(&tsk->signal->live);
if (group_dead) {
pi-futex: fix exit races and locking problems 1. New entries can be added to tsk->pi_state_list after task completed exit_pi_state_list(). The result is memory leakage and deadlocks. 2. handle_mm_fault() is called under spinlock. The result is obvious. 3. results in self-inflicted deadlock inside glibc. Sometimes futex_lock_pi returns -ESRCH, when it is not expected and glibc enters to for(;;) sleep() to simulate deadlock. This problem is quite obvious and I think the patch is right. Though it looks like each "if" in futex_lock_pi() got some stupid special case "else if". :-) 4. sometimes futex_lock_pi() returns -EDEADLK, when nobody has the lock. The reason is also obvious (see comment in the patch), but correct fix is far beyond my comprehension. I guess someone already saw this, the chunk: if (rt_mutex_trylock(&q.pi_state->pi_mutex)) ret = 0; is obviously from the same opera. But it does not work, because the rtmutex is really taken at this point: wake_futex_pi() of previous owner reassigned it to us. My fix works. But it looks very stupid. I would think about removal of shift of ownership in wake_futex_pi() and making all the work in context of process taking lock. From: Thomas Gleixner <tglx@linutronix.de> Fix 1) Avoid the tasklist lock variant of the exit race fix by adding an additional state transition to the exit code. This fixes also the issue, when a task with recursive segfaults is not able to release the futexes. Fix 2) Cleanup the lookup_pi_state() failure path and solve the -ESRCH problem finally. Fix 3) Solve the fixup_pi_state_owner() problem which needs to do the fixup in the lock protected section by using the in_atomic userspace access functions. This removes also the ugly lock drop / unqueue inside of fixup_pi_state() Fix 4) Fix a stale lock in the error path of futex_wake_pi() Added some error checks for verification. The -EDEADLK problem is solved by the rtmutex fixups. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Ingo Molnar <mingo@elte.hu> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-06-08 20:47:00 +00:00
hrtimer_cancel(&tsk->signal->real_timer);
exit_itimers(tsk->signal);
getrusage: fill ru_maxrss value Make ->ru_maxrss value in struct rusage filled accordingly to rss hiwater mark. This struct is filled as a parameter to getrusage syscall. ->ru_maxrss value is set to KBs which is the way it is done in BSD systems. /usr/bin/time (gnu time) application converts ->ru_maxrss to KBs which seems to be incorrect behavior. Maintainer of this util was notified by me with the patch which corrects it and cc'ed. To make this happen we extend struct signal_struct by two fields. The first one is ->maxrss which we use to store rss hiwater of the task. The second one is ->cmaxrss which we use to store highest rss hiwater of all task childs. These values are used in k_getrusage() to actually fill ->ru_maxrss. k_getrusage() uses current rss hiwater value directly if mm struct exists. Note: exec() clear mm->hiwater_rss, but doesn't clear sig->maxrss. it is intetionally behavior. *BSD getrusage have exec() inheriting. test programs ======================================================== getrusage.c =========== #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include <sys/types.h> #include <sys/wait.h> #include <unistd.h> #include <signal.h> #include <sys/mman.h> #include "common.h" #define err(str) perror(str), exit(1) int main(int argc, char** argv) { int status; printf("allocate 100MB\n"); consume(100); printf("testcase1: fork inherit? \n"); printf(" expect: initial.self ~= child.self\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase2: fork inherit? (cont.) \n"); printf(" expect: initial.children ~= 100MB, but child.children = 0\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("child"); _exit(0); } printf("\n"); printf("testcase3: fork + malloc \n"); printf(" expect: child.self ~= initial.self + 50MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { printf("allocate +50MB\n"); consume(50); show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase4: grandchild maxrss\n"); printf(" expect: post_wait.children ~= 300MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); show_rusage("post_wait"); } else { system("./child -n 0 -g 300"); _exit(0); } printf("\n"); printf("testcase5: zombie\n"); printf(" expect: pre_wait ~= initial, IOW the zombie process is not accounted.\n"); printf(" post_wait ~= 400MB, IOW wait() collect child's max_rss. \n"); show_rusage("initial"); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("pre_wait"); wait(&status); show_rusage("post_wait"); } else { system("./child -n 400"); _exit(0); } printf("\n"); printf("testcase6: SIG_IGN\n"); printf(" expect: initial ~= after_zombie (child's 500MB alloc should be ignored).\n"); show_rusage("initial"); signal(SIGCHLD, SIG_IGN); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("after_zombie"); } else { system("./child -n 500"); _exit(0); } printf("\n"); signal(SIGCHLD, SIG_DFL); printf("testcase7: exec (without fork) \n"); printf(" expect: initial ~= exec \n"); show_rusage("initial"); execl("./child", "child", "-v", NULL); return 0; } child.c ======= #include <sys/types.h> #include <unistd.h> #include <sys/types.h> #include <sys/wait.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include "common.h" int main(int argc, char** argv) { int status; int c; long consume_size = 0; long grandchild_consume_size = 0; int show = 0; while ((c = getopt(argc, argv, "n:g:v")) != -1) { switch (c) { case 'n': consume_size = atol(optarg); break; case 'v': show = 1; break; case 'g': grandchild_consume_size = atol(optarg); break; default: break; } } if (show) show_rusage("exec"); if (consume_size) { printf("child alloc %ldMB\n", consume_size); consume(consume_size); } if (grandchild_consume_size) { if (fork()) { wait(&status); } else { printf("grandchild alloc %ldMB\n", grandchild_consume_size); consume(grandchild_consume_size); exit(0); } } return 0; } common.c ======== #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include <sys/types.h> #include <sys/wait.h> #include <unistd.h> #include <signal.h> #include <sys/mman.h> #include "common.h" #define err(str) perror(str), exit(1) void show_rusage(char *prefix) { int err, err2; struct rusage rusage_self; struct rusage rusage_children; printf("%s: ", prefix); err = getrusage(RUSAGE_SELF, &rusage_self); if (!err) printf("self %ld ", rusage_self.ru_maxrss); err2 = getrusage(RUSAGE_CHILDREN, &rusage_children); if (!err2) printf("children %ld ", rusage_children.ru_maxrss); printf("\n"); } /* Some buggy OS need this worthless CPU waste. */ void make_pagefault(void) { void *addr; int size = getpagesize(); int i; for (i=0; i<1000; i++) { addr = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0); if (addr == MAP_FAILED) err("make_pagefault"); memset(addr, 0, size); munmap(addr, size); } } void consume(int mega) { size_t sz = mega * 1024 * 1024; void *ptr; ptr = malloc(sz); memset(ptr, 0, sz); make_pagefault(); } pid_t __fork(void) { pid_t pid; pid = fork(); make_pagefault(); return pid; } common.h ======== void show_rusage(char *prefix); void make_pagefault(void); void consume(int mega); pid_t __fork(void); FreeBSD result (expected result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 103492 children 0 fork child: self 103540 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 103540 children 103540 child: self 103564 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 103564 children 103564 allocate +50MB fork child: self 154860 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 103564 children 154860 grandchild alloc 300MB post_wait: self 103564 children 308720 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 103564 children 308720 child alloc 400MB pre_wait: self 103564 children 308720 post_wait: self 103564 children 411312 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 103564 children 411312 child alloc 500MB after_zombie: self 103624 children 411312 testcase7: exec (without fork) expect: initial ~= exec initial: self 103624 children 411312 exec: self 103624 children 411312 Linux result (actual test result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 102848 children 0 fork child: self 102572 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 102876 children 102644 child: self 102572 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 102876 children 102644 allocate +50MB fork child: self 153804 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 102876 children 153864 grandchild alloc 300MB post_wait: self 102876 children 307536 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 102876 children 307536 child alloc 400MB pre_wait: self 102876 children 307536 post_wait: self 102876 children 410076 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 102876 children 410076 child alloc 500MB after_zombie: self 102880 children 410076 testcase7: exec (without fork) expect: initial ~= exec initial: self 102880 children 410076 exec: self 102880 children 410076 Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-09-22 23:44:10 +00:00
if (tsk->mm)
setmax_mm_hiwater_rss(&tsk->signal->maxrss, tsk->mm);
}
acct_collect(code, group_dead);
Audit: add TTY input auditing Add TTY input auditing, used to audit system administrator's actions. This is required by various security standards such as DCID 6/3 and PCI to provide non-repudiation of administrator's actions and to allow a review of past actions if the administrator seems to overstep their duties or if the system becomes misconfigured for unknown reasons. These requirements do not make it necessary to audit TTY output as well. Compared to an user-space keylogger, this approach records TTY input using the audit subsystem, correlated with other audit events, and it is completely transparent to the user-space application (e.g. the console ioctls still work). TTY input auditing works on a higher level than auditing all system calls within the session, which would produce an overwhelming amount of mostly useless audit events. Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs by process with the attribute is sent to the audit subsystem by the kernel. The audit netlink interface is extended to allow modifying the audit_tty attribute, and to allow sending explanatory audit events from user-space (for example, a shell might send an event containing the final command, after the interactive command-line editing and history expansion is performed, which might be difficult to decipher from the TTY input alone). Because the "audit_tty" attribute is inherited across fork (), it would be set e.g. for sshd restarted within an audited session. To prevent this, the audit_tty attribute is cleared when a process with no open TTY file descriptors (e.g. after daemon startup) opens a TTY. See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a more detailed rationale document for an older version of this patch. [akpm@linux-foundation.org: build fix] Signed-off-by: Miloslav Trmac <mitr@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Paul Fulghum <paulkf@microgate.com> Cc: Casey Schaufler <casey@schaufler-ca.com> Cc: Steve Grubb <sgrubb@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 06:40:56 +00:00
if (group_dead)
tty_audit_exit();
audit_free(tsk);
tsk->exit_code = code;
taskstats_exit(tsk, group_dead);
exit_mm(tsk);
if (group_dead)
acct_process();
trace_sched_process_exit(tsk);
exit_sem(tsk);
ipc: introduce shm_rmid_forced sysctl Add support for the shm_rmid_forced sysctl. If set to 1, all shared memory objects in current ipc namespace will be automatically forced to use IPC_RMID. The POSIX way of handling shmem allows one to create shm objects and call shmdt(), leaving shm object associated with no process, thus consuming memory not counted via rlimits. With shm_rmid_forced=1 the shared memory object is counted at least for one process, so OOM killer may effectively kill the fat process holding the shared memory. It obviously breaks POSIX - some programs relying on the feature would stop working. So set shm_rmid_forced=1 only if you're sure nobody uses "orphaned" memory. Use shm_rmid_forced=0 by default for compatability reasons. The feature was previously impemented in -ow as a configure option. [akpm@linux-foundation.org: fix documentation, per Randy] [akpm@linux-foundation.org: fix warning] [akpm@linux-foundation.org: readability/conventionality tweaks] [akpm@linux-foundation.org: fix shm_rmid_forced/shm_forced_rmid confusion, use standard comment layout] Signed-off-by: Vasiliy Kulikov <segoon@openwall.com> Cc: Randy Dunlap <rdunlap@xenotime.net> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: "Serge E. Hallyn" <serge.hallyn@canonical.com> Cc: Daniel Lezcano <daniel.lezcano@free.fr> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Tejun Heo <tj@kernel.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Solar Designer <solar@openwall.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-07-26 23:08:48 +00:00
exit_shm(tsk);
exit_files(tsk);
exit_fs(tsk);
exit_task_work(tsk);
check_stack_usage();
exit_thread();
/*
* Flush inherited counters to the parent - before the parent
* gets woken up by child-exit notifications.
*
* because of cgroup mode, must be called before cgroup_exit()
*/
perf_event_exit_task(tsk);
cgroup_exit(tsk, 1);
if (group_dead)
disassociate_ctty(1);
module_put(task_thread_info(tsk)->exec_domain->module);
proc_exit_connector(tsk);
hw-breakpoints: Rewrite the hw-breakpoints layer on top of perf events This patch rebase the implementation of the breakpoints API on top of perf events instances. Each breakpoints are now perf events that handle the register scheduling, thread/cpu attachment, etc.. The new layering is now made as follows: ptrace kgdb ftrace perf syscall \ | / / \ | / / / Core breakpoint API / / | / | / Breakpoints perf events | | Breakpoints PMU ---- Debug Register constraints handling (Part of core breakpoint API) | | Hardware debug registers Reasons of this rewrite: - Use the centralized/optimized pmu registers scheduling, implying an easier arch integration - More powerful register handling: perf attributes (pinned/flexible events, exclusive/non-exclusive, tunable period, etc...) Impact: - New perf ABI: the hardware breakpoints counters - Ptrace breakpoints setting remains tricky and still needs some per thread breakpoints references. Todo (in the order): - Support breakpoints perf counter events for perf tools (ie: implement perf_bpcounter_event()) - Support from perf tools Changes in v2: - Follow the perf "event " rename - The ptrace regression have been fixed (ptrace breakpoint perf events weren't released when a task ended) - Drop the struct hw_breakpoint and store generic fields in perf_event_attr. - Separate core and arch specific headers, drop asm-generic/hw_breakpoint.h and create linux/hw_breakpoint.h - Use new generic len/type for breakpoint - Handle off case: when breakpoints api is not supported by an arch Changes in v3: - Fix broken CONFIG_KVM, we need to propagate the breakpoint api changes to kvm when we exit the guest and restore the bp registers to the host. Changes in v4: - Drop the hw_breakpoint_restore() stub as it is only used by KVM - EXPORT_SYMBOL_GPL hw_breakpoint_restore() as KVM can be built as a module - Restore the breakpoints unconditionally on kvm guest exit: TIF_DEBUG_THREAD doesn't anymore cover every cases of running breakpoints and vcpu->arch.switch_db_regs might not always be set when the guest used debug registers. (Waiting for a reliable optimization) Changes in v5: - Split-up the asm-generic/hw-breakpoint.h moving to linux/hw_breakpoint.h into a separate patch - Optimize the breakpoints restoring while switching from kvm guest to host. We only want to restore the state if we have active breakpoints to the host, otherwise we don't care about messed-up address registers. - Add asm/hw_breakpoint.h to Kbuild - Fix bad breakpoint type in trace_selftest.c Changes in v6: - Fix wrong header inclusion in trace.h (triggered a build error with CONFIG_FTRACE_SELFTEST Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Prasad <prasad@linux.vnet.ibm.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jan Kiszka <jan.kiszka@web.de> Cc: Jiri Slaby <jirislaby@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Avi Kivity <avi@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Paul Mundt <lethal@linux-sh.org>
2009-09-09 17:22:48 +00:00
/*
* FIXME: do that only when needed, using sched_exit tracepoint
*/
ptrace_put_breakpoints(tsk);
exit_notify(tsk, group_dead);
#ifdef CONFIG_NUMA
cpuset,mm: fix no node to alloc memory when changing cpuset's mems Before applying this patch, cpuset updates task->mems_allowed and mempolicy by setting all new bits in the nodemask first, and clearing all old unallowed bits later. But in the way, the allocator may find that there is no node to alloc memory. The reason is that cpuset rebinds the task's mempolicy, it cleans the nodes which the allocater can alloc pages on, for example: (mpol: mempolicy) task1 task1's mpol task2 alloc page 1 alloc on node0? NO 1 1 change mems from 1 to 0 1 rebind task1's mpol 0-1 set new bits 0 clear disallowed bits alloc on node1? NO 0 ... can't alloc page goto oom This patch fixes this problem by expanding the nodes range first(set newly allowed bits) and shrink it lazily(clear newly disallowed bits). So we use a variable to tell the write-side task that read-side task is reading nodemask, and the write-side task clears newly disallowed nodes after read-side task ends the current memory allocation. [akpm@linux-foundation.org: fix spello] Signed-off-by: Miao Xie <miaox@cn.fujitsu.com> Cc: David Rientjes <rientjes@google.com> Cc: Nick Piggin <npiggin@suse.de> Cc: Paul Menage <menage@google.com> Cc: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk> Cc: Ravikiran Thirumalai <kiran@scalex86.org> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Christoph Lameter <cl@linux-foundation.org> Cc: Andi Kleen <andi@firstfloor.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2010-05-24 21:32:08 +00:00
task_lock(tsk);
mpol_put(tsk->mempolicy);
tsk->mempolicy = NULL;
cpuset,mm: fix no node to alloc memory when changing cpuset's mems Before applying this patch, cpuset updates task->mems_allowed and mempolicy by setting all new bits in the nodemask first, and clearing all old unallowed bits later. But in the way, the allocator may find that there is no node to alloc memory. The reason is that cpuset rebinds the task's mempolicy, it cleans the nodes which the allocater can alloc pages on, for example: (mpol: mempolicy) task1 task1's mpol task2 alloc page 1 alloc on node0? NO 1 1 change mems from 1 to 0 1 rebind task1's mpol 0-1 set new bits 0 clear disallowed bits alloc on node1? NO 0 ... can't alloc page goto oom This patch fixes this problem by expanding the nodes range first(set newly allowed bits) and shrink it lazily(clear newly disallowed bits). So we use a variable to tell the write-side task that read-side task is reading nodemask, and the write-side task clears newly disallowed nodes after read-side task ends the current memory allocation. [akpm@linux-foundation.org: fix spello] Signed-off-by: Miao Xie <miaox@cn.fujitsu.com> Cc: David Rientjes <rientjes@google.com> Cc: Nick Piggin <npiggin@suse.de> Cc: Paul Menage <menage@google.com> Cc: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk> Cc: Ravikiran Thirumalai <kiran@scalex86.org> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Christoph Lameter <cl@linux-foundation.org> Cc: Andi Kleen <andi@firstfloor.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2010-05-24 21:32:08 +00:00
task_unlock(tsk);
#endif
#ifdef CONFIG_FUTEX
if (unlikely(current->pi_state_cache))
kfree(current->pi_state_cache);
#endif
/*
* Make sure we are holding no locks:
*/
debug_check_no_locks_held(tsk);
pi-futex: fix exit races and locking problems 1. New entries can be added to tsk->pi_state_list after task completed exit_pi_state_list(). The result is memory leakage and deadlocks. 2. handle_mm_fault() is called under spinlock. The result is obvious. 3. results in self-inflicted deadlock inside glibc. Sometimes futex_lock_pi returns -ESRCH, when it is not expected and glibc enters to for(;;) sleep() to simulate deadlock. This problem is quite obvious and I think the patch is right. Though it looks like each "if" in futex_lock_pi() got some stupid special case "else if". :-) 4. sometimes futex_lock_pi() returns -EDEADLK, when nobody has the lock. The reason is also obvious (see comment in the patch), but correct fix is far beyond my comprehension. I guess someone already saw this, the chunk: if (rt_mutex_trylock(&q.pi_state->pi_mutex)) ret = 0; is obviously from the same opera. But it does not work, because the rtmutex is really taken at this point: wake_futex_pi() of previous owner reassigned it to us. My fix works. But it looks very stupid. I would think about removal of shift of ownership in wake_futex_pi() and making all the work in context of process taking lock. From: Thomas Gleixner <tglx@linutronix.de> Fix 1) Avoid the tasklist lock variant of the exit race fix by adding an additional state transition to the exit code. This fixes also the issue, when a task with recursive segfaults is not able to release the futexes. Fix 2) Cleanup the lookup_pi_state() failure path and solve the -ESRCH problem finally. Fix 3) Solve the fixup_pi_state_owner() problem which needs to do the fixup in the lock protected section by using the in_atomic userspace access functions. This removes also the ugly lock drop / unqueue inside of fixup_pi_state() Fix 4) Fix a stale lock in the error path of futex_wake_pi() Added some error checks for verification. The -EDEADLK problem is solved by the rtmutex fixups. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Ingo Molnar <mingo@elte.hu> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Eric Dumazet <dada1@cosmosbay.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-06-08 20:47:00 +00:00
/*
* We can do this unlocked here. The futex code uses this flag
* just to verify whether the pi state cleanup has been done
* or not. In the worst case it loops once more.
*/
tsk->flags |= PF_EXITPIDONE;
if (tsk->io_context)
exit_io_context(tsk);
if (tsk->splice_pipe)
__free_pipe_info(tsk->splice_pipe);
net: use a per task frag allocator We currently use a per socket order-0 page cache for tcp_sendmsg() operations. This page is used to build fragments for skbs. Its done to increase probability of coalescing small write() into single segments in skbs still in write queue (not yet sent) But it wastes a lot of memory for applications handling many mostly idle sockets, since each socket holds one page in sk->sk_sndmsg_page Its also quite inefficient to build TSO 64KB packets, because we need about 16 pages per skb on arches where PAGE_SIZE = 4096, so we hit page allocator more than wanted. This patch adds a per task frag allocator and uses bigger pages, if available. An automatic fallback is done in case of memory pressure. (up to 32768 bytes per frag, thats order-3 pages on x86) This increases TCP stream performance by 20% on loopback device, but also benefits on other network devices, since 8x less frags are mapped on transmit and unmapped on tx completion. Alexander Duyck mentioned a probable performance win on systems with IOMMU enabled. Its possible some SG enabled hardware cant cope with bigger fragments, but their ndo_start_xmit() should already handle this, splitting a fragment in sub fragments, since some arches have PAGE_SIZE=65536 Successfully tested on various ethernet devices. (ixgbe, igb, bnx2x, tg3, mellanox mlx4) Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Ben Hutchings <bhutchings@solarflare.com> Cc: Vijay Subramanian <subramanian.vijay@gmail.com> Cc: Alexander Duyck <alexander.h.duyck@intel.com> Tested-by: Vijay Subramanian <subramanian.vijay@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-09-23 23:04:42 +00:00
if (tsk->task_frag.page)
put_page(tsk->task_frag.page);
validate_creds_for_do_exit(tsk);
preempt_disable();
if (tsk->nr_dirtied)
__this_cpu_add(dirty_throttle_leaks, tsk->nr_dirtied);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-22 20:56:52 +00:00
exit_rcu();
sched: Fix ancient race in do_exit() try_to_wake_up() has a problem which may change status from TASK_DEAD to TASK_RUNNING in race condition with SMI or guest environment of virtual machine. As a result, exited task is scheduled() again and panic occurs. Here is the sequence how it occurs: ----------------------------------+----------------------------- | CPU A | CPU B ----------------------------------+----------------------------- TASK A calls exit().... do_exit() exit_mm() down_read(mm->mmap_sem); rwsem_down_failed_common() set TASK_UNINTERRUPTIBLE set waiter.task <= task A list_add to sem->wait_list : raw_spin_unlock_irq() (I/O interruption occured) __rwsem_do_wake(mmap_sem) list_del(&waiter->list); waiter->task = NULL wake_up_process(task A) try_to_wake_up() (task is still TASK_UNINTERRUPTIBLE) p->on_rq is still 1.) ttwu_do_wakeup() (*A) : (I/O interruption handler finished) if (!waiter.task) schedule() is not called due to waiter.task is NULL. tsk->state = TASK_RUNNING : check_preempt_curr(); : task->state = TASK_DEAD (*B) <--- set TASK_RUNNING (*C) schedule() (exit task is running again) BUG_ON() is called! -------------------------------------------------------- The execution time between (*A) and (*B) is usually very short, because the interruption is disabled, and setting TASK_RUNNING at (*C) must be executed before setting TASK_DEAD. HOWEVER, if SMI is interrupted between (*A) and (*B), (*C) is able to execute AFTER setting TASK_DEAD! Then, exited task is scheduled again, and BUG_ON() is called.... If the system works on guest system of virtual machine, the time between (*A) and (*B) may be also long due to scheduling of hypervisor, and same phenomenon can occur. By this patch, do_exit() waits for releasing task->pi_lock which is used in try_to_wake_up(). It guarantees the task becomes TASK_DEAD after waking up. Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com> Acked-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Link: http://lkml.kernel.org/r/20120117174031.3118.E1E9C6FF@jp.fujitsu.com Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-01-17 08:40:31 +00:00
/*
* The setting of TASK_RUNNING by try_to_wake_up() may be delayed
* when the following two conditions become true.
* - There is race condition of mmap_sem (It is acquired by
* exit_mm()), and
* - SMI occurs before setting TASK_RUNINNG.
* (or hypervisor of virtual machine switches to other guest)
* As a result, we may become TASK_RUNNING after becoming TASK_DEAD
*
* To avoid it, we have to wait for releasing tsk->pi_lock which
* is held by try_to_wake_up()
*/
smp_mb();
raw_spin_unlock_wait(&tsk->pi_lock);
/* causes final put_task_struct in finish_task_switch(). */
tsk->state = TASK_DEAD;
tsk->flags |= PF_NOFREEZE; /* tell freezer to ignore us */
schedule();
BUG();
/* Avoid "noreturn function does return". */
for (;;)
cpu_relax(); /* For when BUG is null */
}
EXPORT_SYMBOL_GPL(do_exit);
void complete_and_exit(struct completion *comp, long code)
{
if (comp)
complete(comp);
do_exit(code);
}
EXPORT_SYMBOL(complete_and_exit);
SYSCALL_DEFINE1(exit, int, error_code)
{
do_exit((error_code&0xff)<<8);
}
/*
* Take down every thread in the group. This is called by fatal signals
* as well as by sys_exit_group (below).
*/
void
do_group_exit(int exit_code)
{
struct signal_struct *sig = current->signal;
BUG_ON(exit_code & 0x80); /* core dumps don't get here */
if (signal_group_exit(sig))
exit_code = sig->group_exit_code;
else if (!thread_group_empty(current)) {
struct sighand_struct *const sighand = current->sighand;
spin_lock_irq(&sighand->siglock);
if (signal_group_exit(sig))
/* Another thread got here before we took the lock. */
exit_code = sig->group_exit_code;
else {
sig->group_exit_code = exit_code;
sig->flags = SIGNAL_GROUP_EXIT;
zap_other_threads(current);
}
spin_unlock_irq(&sighand->siglock);
}
do_exit(exit_code);
/* NOTREACHED */
}
/*
* this kills every thread in the thread group. Note that any externally
* wait4()-ing process will get the correct exit code - even if this
* thread is not the thread group leader.
*/
SYSCALL_DEFINE1(exit_group, int, error_code)
{
do_group_exit((error_code & 0xff) << 8);
/* NOTREACHED */
return 0;
}
struct wait_opts {
enum pid_type wo_type;
int wo_flags;
struct pid *wo_pid;
struct siginfo __user *wo_info;
int __user *wo_stat;
struct rusage __user *wo_rusage;
wait_queue_t child_wait;
int notask_error;
};
static inline
struct pid *task_pid_type(struct task_struct *task, enum pid_type type)
{
if (type != PIDTYPE_PID)
task = task->group_leader;
return task->pids[type].pid;
}
static int eligible_pid(struct wait_opts *wo, struct task_struct *p)
{
return wo->wo_type == PIDTYPE_MAX ||
task_pid_type(p, wo->wo_type) == wo->wo_pid;
}
static int eligible_child(struct wait_opts *wo, struct task_struct *p)
{
if (!eligible_pid(wo, p))
return 0;
/* Wait for all children (clone and not) if __WALL is set;
* otherwise, wait for clone children *only* if __WCLONE is
* set; otherwise, wait for non-clone children *only*. (Note:
* A "clone" child here is one that reports to its parent
* using a signal other than SIGCHLD.) */
if (((p->exit_signal != SIGCHLD) ^ !!(wo->wo_flags & __WCLONE))
&& !(wo->wo_flags & __WALL))
return 0;
return 1;
}
static int wait_noreap_copyout(struct wait_opts *wo, struct task_struct *p,
pid_t pid, uid_t uid, int why, int status)
{
struct siginfo __user *infop;
int retval = wo->wo_rusage
? getrusage(p, RUSAGE_BOTH, wo->wo_rusage) : 0;
put_task_struct(p);
infop = wo->wo_info;
if (infop) {
if (!retval)
retval = put_user(SIGCHLD, &infop->si_signo);
if (!retval)
retval = put_user(0, &infop->si_errno);
if (!retval)
retval = put_user((short)why, &infop->si_code);
if (!retval)
retval = put_user(pid, &infop->si_pid);
if (!retval)
retval = put_user(uid, &infop->si_uid);
if (!retval)
retval = put_user(status, &infop->si_status);
}
if (!retval)
retval = pid;
return retval;
}
/*
* Handle sys_wait4 work for one task in state EXIT_ZOMBIE. We hold
* read_lock(&tasklist_lock) on entry. If we return zero, we still hold
* the lock and this task is uninteresting. If we return nonzero, we have
* released the lock and the system call should return.
*/
static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p)
{
unsigned long state;
wait_task_zombie: fix 2/3 races vs forget_original_parent() Two threads, T1 and T2. T2 ptraces P, and P is not a child of ptracer's thread group. P exits and goes to TASK_ZOMBIE. T1 does wait_task_zombie(P): P->exit_state = TASK_DEAD; ... read_unlock(&tasklist_lock); T2 does exit(), takes tasklist, forget_original_parent() does __ptrace_unlink(P) but doesn't call do_notify_parent(P) because p->exit_state == EXIT_DEAD. Now, P is not visible to our process: __ptrace_unlink() removed it from ->children. We should send notification to P->parent and release P if and only if SIGCHLD is ignored. And we have 3 bugs: 1. P->parent does do_wait() and gets -ECHILD (P is on ->parent->children, but its state is TASK_DEAD). 2. // wait_task_zombie() continues if (put_user(...)) { // TODO: is this safe? p->exit_state = EXIT_ZOMBIE; return; } we return without notification/release, task_struct leaked. Solution: ignore -EFAULT and proceed. It is an application's bug if we can't fill infop/stat_addr (in case of VM_FAULT_OOM we have much more problems). 3. // wait_task_zombie() continues if (p->real_parent != p->parent) { // Not taken, it was untraced'ed ... } release_task(p); we released the task which we shouldn't. Solution: check ->real_parent != ->parent before, under tasklist_lock, but use ptrace_unlink() instead of __ptrace_unlink() to check ->ptrace. This patch hopefully solves 2 and 3, the 1st bug will be fixed later, we need some cleanups in forget_original_parent/reparent_thread. However, the first race is very unlikely and not critical, so I hope it makes sense to fix 1 and 2 for now. 4. Small cleanup: don't "restore" EXIT_ZOMBIE unless we know we are not going to realease the child. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Cc: Roland McGrath <roland@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:26:58 +00:00
int retval, status, traced;
pid_t pid = task_pid_vnr(p);
uid_t uid = from_kuid_munged(current_user_ns(), task_uid(p));
struct siginfo __user *infop;
if (!likely(wo->wo_flags & WEXITED))
return 0;
if (unlikely(wo->wo_flags & WNOWAIT)) {
int exit_code = p->exit_code;
int why;
get_task_struct(p);
read_unlock(&tasklist_lock);
if ((exit_code & 0x7f) == 0) {
why = CLD_EXITED;
status = exit_code >> 8;
} else {
why = (exit_code & 0x80) ? CLD_DUMPED : CLD_KILLED;
status = exit_code & 0x7f;
}
return wait_noreap_copyout(wo, p, pid, uid, why, status);
}
/*
* Try to move the task's state to DEAD
* only one thread is allowed to do this:
*/
state = xchg(&p->exit_state, EXIT_DEAD);
if (state != EXIT_ZOMBIE) {
BUG_ON(state != EXIT_DEAD);
return 0;
}
traced = ptrace_reparented(p);
/*
* It can be ptraced but not reparented, check
* thread_group_leader() to filter out sub-threads.
*/
if (likely(!traced) && thread_group_leader(p)) {
struct signal_struct *psig;
struct signal_struct *sig;
getrusage: fill ru_maxrss value Make ->ru_maxrss value in struct rusage filled accordingly to rss hiwater mark. This struct is filled as a parameter to getrusage syscall. ->ru_maxrss value is set to KBs which is the way it is done in BSD systems. /usr/bin/time (gnu time) application converts ->ru_maxrss to KBs which seems to be incorrect behavior. Maintainer of this util was notified by me with the patch which corrects it and cc'ed. To make this happen we extend struct signal_struct by two fields. The first one is ->maxrss which we use to store rss hiwater of the task. The second one is ->cmaxrss which we use to store highest rss hiwater of all task childs. These values are used in k_getrusage() to actually fill ->ru_maxrss. k_getrusage() uses current rss hiwater value directly if mm struct exists. Note: exec() clear mm->hiwater_rss, but doesn't clear sig->maxrss. it is intetionally behavior. *BSD getrusage have exec() inheriting. test programs ======================================================== getrusage.c =========== #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include <sys/types.h> #include <sys/wait.h> #include <unistd.h> #include <signal.h> #include <sys/mman.h> #include "common.h" #define err(str) perror(str), exit(1) int main(int argc, char** argv) { int status; printf("allocate 100MB\n"); consume(100); printf("testcase1: fork inherit? \n"); printf(" expect: initial.self ~= child.self\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase2: fork inherit? (cont.) \n"); printf(" expect: initial.children ~= 100MB, but child.children = 0\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("child"); _exit(0); } printf("\n"); printf("testcase3: fork + malloc \n"); printf(" expect: child.self ~= initial.self + 50MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { printf("allocate +50MB\n"); consume(50); show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase4: grandchild maxrss\n"); printf(" expect: post_wait.children ~= 300MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); show_rusage("post_wait"); } else { system("./child -n 0 -g 300"); _exit(0); } printf("\n"); printf("testcase5: zombie\n"); printf(" expect: pre_wait ~= initial, IOW the zombie process is not accounted.\n"); printf(" post_wait ~= 400MB, IOW wait() collect child's max_rss. \n"); show_rusage("initial"); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("pre_wait"); wait(&status); show_rusage("post_wait"); } else { system("./child -n 400"); _exit(0); } printf("\n"); printf("testcase6: SIG_IGN\n"); printf(" expect: initial ~= after_zombie (child's 500MB alloc should be ignored).\n"); show_rusage("initial"); signal(SIGCHLD, SIG_IGN); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("after_zombie"); } else { system("./child -n 500"); _exit(0); } printf("\n"); signal(SIGCHLD, SIG_DFL); printf("testcase7: exec (without fork) \n"); printf(" expect: initial ~= exec \n"); show_rusage("initial"); execl("./child", "child", "-v", NULL); return 0; } child.c ======= #include <sys/types.h> #include <unistd.h> #include <sys/types.h> #include <sys/wait.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include "common.h" int main(int argc, char** argv) { int status; int c; long consume_size = 0; long grandchild_consume_size = 0; int show = 0; while ((c = getopt(argc, argv, "n:g:v")) != -1) { switch (c) { case 'n': consume_size = atol(optarg); break; case 'v': show = 1; break; case 'g': grandchild_consume_size = atol(optarg); break; default: break; } } if (show) show_rusage("exec"); if (consume_size) { printf("child alloc %ldMB\n", consume_size); consume(consume_size); } if (grandchild_consume_size) { if (fork()) { wait(&status); } else { printf("grandchild alloc %ldMB\n", grandchild_consume_size); consume(grandchild_consume_size); exit(0); } } return 0; } common.c ======== #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include <sys/types.h> #include <sys/wait.h> #include <unistd.h> #include <signal.h> #include <sys/mman.h> #include "common.h" #define err(str) perror(str), exit(1) void show_rusage(char *prefix) { int err, err2; struct rusage rusage_self; struct rusage rusage_children; printf("%s: ", prefix); err = getrusage(RUSAGE_SELF, &rusage_self); if (!err) printf("self %ld ", rusage_self.ru_maxrss); err2 = getrusage(RUSAGE_CHILDREN, &rusage_children); if (!err2) printf("children %ld ", rusage_children.ru_maxrss); printf("\n"); } /* Some buggy OS need this worthless CPU waste. */ void make_pagefault(void) { void *addr; int size = getpagesize(); int i; for (i=0; i<1000; i++) { addr = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0); if (addr == MAP_FAILED) err("make_pagefault"); memset(addr, 0, size); munmap(addr, size); } } void consume(int mega) { size_t sz = mega * 1024 * 1024; void *ptr; ptr = malloc(sz); memset(ptr, 0, sz); make_pagefault(); } pid_t __fork(void) { pid_t pid; pid = fork(); make_pagefault(); return pid; } common.h ======== void show_rusage(char *prefix); void make_pagefault(void); void consume(int mega); pid_t __fork(void); FreeBSD result (expected result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 103492 children 0 fork child: self 103540 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 103540 children 103540 child: self 103564 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 103564 children 103564 allocate +50MB fork child: self 154860 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 103564 children 154860 grandchild alloc 300MB post_wait: self 103564 children 308720 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 103564 children 308720 child alloc 400MB pre_wait: self 103564 children 308720 post_wait: self 103564 children 411312 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 103564 children 411312 child alloc 500MB after_zombie: self 103624 children 411312 testcase7: exec (without fork) expect: initial ~= exec initial: self 103624 children 411312 exec: self 103624 children 411312 Linux result (actual test result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 102848 children 0 fork child: self 102572 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 102876 children 102644 child: self 102572 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 102876 children 102644 allocate +50MB fork child: self 153804 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 102876 children 153864 grandchild alloc 300MB post_wait: self 102876 children 307536 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 102876 children 307536 child alloc 400MB pre_wait: self 102876 children 307536 post_wait: self 102876 children 410076 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 102876 children 410076 child alloc 500MB after_zombie: self 102880 children 410076 testcase7: exec (without fork) expect: initial ~= exec initial: self 102880 children 410076 exec: self 102880 children 410076 Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-09-22 23:44:10 +00:00
unsigned long maxrss;
sched, cputime: Introduce thread_group_times() This is a real fix for problem of utime/stime values decreasing described in the thread: http://lkml.org/lkml/2009/11/3/522 Now cputime is accounted in the following way: - {u,s}time in task_struct are increased every time when the thread is interrupted by a tick (timer interrupt). - When a thread exits, its {u,s}time are added to signal->{u,s}time, after adjusted by task_times(). - When all threads in a thread_group exits, accumulated {u,s}time (and also c{u,s}time) in signal struct are added to c{u,s}time in signal struct of the group's parent. So {u,s}time in task struct are "raw" tick count, while {u,s}time and c{u,s}time in signal struct are "adjusted" values. And accounted values are used by: - task_times(), to get cputime of a thread: This function returns adjusted values that originates from raw {u,s}time and scaled by sum_exec_runtime that accounted by CFS. - thread_group_cputime(), to get cputime of a thread group: This function returns sum of all {u,s}time of living threads in the group, plus {u,s}time in the signal struct that is sum of adjusted cputimes of all exited threads belonged to the group. The problem is the return value of thread_group_cputime(), because it is mixed sum of "raw" value and "adjusted" value: group's {u,s}time = foreach(thread){{u,s}time} + exited({u,s}time) This misbehavior can break {u,s}time monotonicity. Assume that if there is a thread that have raw values greater than adjusted values (e.g. interrupted by 1000Hz ticks 50 times but only runs 45ms) and if it exits, cputime will decrease (e.g. -5ms). To fix this, we could do: group's {u,s}time = foreach(t){task_times(t)} + exited({u,s}time) But task_times() contains hard divisions, so applying it for every thread should be avoided. This patch fixes the above problem in the following way: - Modify thread's exit (= __exit_signal()) not to use task_times(). It means {u,s}time in signal struct accumulates raw values instead of adjusted values. As the result it makes thread_group_cputime() to return pure sum of "raw" values. - Introduce a new function thread_group_times(*task, *utime, *stime) that converts "raw" values of thread_group_cputime() to "adjusted" values, in same calculation procedure as task_times(). - Modify group's exit (= wait_task_zombie()) to use this introduced thread_group_times(). It make c{u,s}time in signal struct to have adjusted values like before this patch. - Replace some thread_group_cputime() by thread_group_times(). This replacements are only applied where conveys the "adjusted" cputime to users, and where already uses task_times() near by it. (i.e. sys_times(), getrusage(), and /proc/<PID>/stat.) This patch have a positive side effect: - Before this patch, if a group contains many short-life threads (e.g. runs 0.9ms and not interrupted by ticks), the group's cputime could be invisible since thread's cputime was accumulated after adjusted: imagine adjustment function as adj(ticks, runtime), {adj(0, 0.9) + adj(0, 0.9) + ....} = {0 + 0 + ....} = 0. After this patch it will not happen because the adjustment is applied after accumulated. v2: - remove if()s, put new variables into signal_struct. Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com> Acked-by: Peter Zijlstra <peterz@infradead.org> Cc: Spencer Candland <spencer@bluehost.com> Cc: Americo Wang <xiyou.wangcong@gmail.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Stanislaw Gruszka <sgruszka@redhat.com> LKML-Reference: <4B162517.8040909@jp.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-12-02 08:28:07 +00:00
cputime_t tgutime, tgstime;
/*
* The resource counters for the group leader are in its
* own task_struct. Those for dead threads in the group
* are in its signal_struct, as are those for the child
* processes it has previously reaped. All these
* accumulate in the parent's signal_struct c* fields.
*
* We don't bother to take a lock here to protect these
* p->signal fields, because they are only touched by
* __exit_signal, which runs with tasklist_lock
* write-locked anyway, and so is excluded here. We do
* need to protect the access to parent->signal fields,
* as other threads in the parent group can be right
* here reaping other children at the same time.
sched, cputime: Introduce thread_group_times() This is a real fix for problem of utime/stime values decreasing described in the thread: http://lkml.org/lkml/2009/11/3/522 Now cputime is accounted in the following way: - {u,s}time in task_struct are increased every time when the thread is interrupted by a tick (timer interrupt). - When a thread exits, its {u,s}time are added to signal->{u,s}time, after adjusted by task_times(). - When all threads in a thread_group exits, accumulated {u,s}time (and also c{u,s}time) in signal struct are added to c{u,s}time in signal struct of the group's parent. So {u,s}time in task struct are "raw" tick count, while {u,s}time and c{u,s}time in signal struct are "adjusted" values. And accounted values are used by: - task_times(), to get cputime of a thread: This function returns adjusted values that originates from raw {u,s}time and scaled by sum_exec_runtime that accounted by CFS. - thread_group_cputime(), to get cputime of a thread group: This function returns sum of all {u,s}time of living threads in the group, plus {u,s}time in the signal struct that is sum of adjusted cputimes of all exited threads belonged to the group. The problem is the return value of thread_group_cputime(), because it is mixed sum of "raw" value and "adjusted" value: group's {u,s}time = foreach(thread){{u,s}time} + exited({u,s}time) This misbehavior can break {u,s}time monotonicity. Assume that if there is a thread that have raw values greater than adjusted values (e.g. interrupted by 1000Hz ticks 50 times but only runs 45ms) and if it exits, cputime will decrease (e.g. -5ms). To fix this, we could do: group's {u,s}time = foreach(t){task_times(t)} + exited({u,s}time) But task_times() contains hard divisions, so applying it for every thread should be avoided. This patch fixes the above problem in the following way: - Modify thread's exit (= __exit_signal()) not to use task_times(). It means {u,s}time in signal struct accumulates raw values instead of adjusted values. As the result it makes thread_group_cputime() to return pure sum of "raw" values. - Introduce a new function thread_group_times(*task, *utime, *stime) that converts "raw" values of thread_group_cputime() to "adjusted" values, in same calculation procedure as task_times(). - Modify group's exit (= wait_task_zombie()) to use this introduced thread_group_times(). It make c{u,s}time in signal struct to have adjusted values like before this patch. - Replace some thread_group_cputime() by thread_group_times(). This replacements are only applied where conveys the "adjusted" cputime to users, and where already uses task_times() near by it. (i.e. sys_times(), getrusage(), and /proc/<PID>/stat.) This patch have a positive side effect: - Before this patch, if a group contains many short-life threads (e.g. runs 0.9ms and not interrupted by ticks), the group's cputime could be invisible since thread's cputime was accumulated after adjusted: imagine adjustment function as adj(ticks, runtime), {adj(0, 0.9) + adj(0, 0.9) + ....} = {0 + 0 + ....} = 0. After this patch it will not happen because the adjustment is applied after accumulated. v2: - remove if()s, put new variables into signal_struct. Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com> Acked-by: Peter Zijlstra <peterz@infradead.org> Cc: Spencer Candland <spencer@bluehost.com> Cc: Americo Wang <xiyou.wangcong@gmail.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Stanislaw Gruszka <sgruszka@redhat.com> LKML-Reference: <4B162517.8040909@jp.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-12-02 08:28:07 +00:00
*
* We use thread_group_times() to get times for the thread
* group, which consolidates times for all threads in the
* group including the group leader.
*/
sched, cputime: Introduce thread_group_times() This is a real fix for problem of utime/stime values decreasing described in the thread: http://lkml.org/lkml/2009/11/3/522 Now cputime is accounted in the following way: - {u,s}time in task_struct are increased every time when the thread is interrupted by a tick (timer interrupt). - When a thread exits, its {u,s}time are added to signal->{u,s}time, after adjusted by task_times(). - When all threads in a thread_group exits, accumulated {u,s}time (and also c{u,s}time) in signal struct are added to c{u,s}time in signal struct of the group's parent. So {u,s}time in task struct are "raw" tick count, while {u,s}time and c{u,s}time in signal struct are "adjusted" values. And accounted values are used by: - task_times(), to get cputime of a thread: This function returns adjusted values that originates from raw {u,s}time and scaled by sum_exec_runtime that accounted by CFS. - thread_group_cputime(), to get cputime of a thread group: This function returns sum of all {u,s}time of living threads in the group, plus {u,s}time in the signal struct that is sum of adjusted cputimes of all exited threads belonged to the group. The problem is the return value of thread_group_cputime(), because it is mixed sum of "raw" value and "adjusted" value: group's {u,s}time = foreach(thread){{u,s}time} + exited({u,s}time) This misbehavior can break {u,s}time monotonicity. Assume that if there is a thread that have raw values greater than adjusted values (e.g. interrupted by 1000Hz ticks 50 times but only runs 45ms) and if it exits, cputime will decrease (e.g. -5ms). To fix this, we could do: group's {u,s}time = foreach(t){task_times(t)} + exited({u,s}time) But task_times() contains hard divisions, so applying it for every thread should be avoided. This patch fixes the above problem in the following way: - Modify thread's exit (= __exit_signal()) not to use task_times(). It means {u,s}time in signal struct accumulates raw values instead of adjusted values. As the result it makes thread_group_cputime() to return pure sum of "raw" values. - Introduce a new function thread_group_times(*task, *utime, *stime) that converts "raw" values of thread_group_cputime() to "adjusted" values, in same calculation procedure as task_times(). - Modify group's exit (= wait_task_zombie()) to use this introduced thread_group_times(). It make c{u,s}time in signal struct to have adjusted values like before this patch. - Replace some thread_group_cputime() by thread_group_times(). This replacements are only applied where conveys the "adjusted" cputime to users, and where already uses task_times() near by it. (i.e. sys_times(), getrusage(), and /proc/<PID>/stat.) This patch have a positive side effect: - Before this patch, if a group contains many short-life threads (e.g. runs 0.9ms and not interrupted by ticks), the group's cputime could be invisible since thread's cputime was accumulated after adjusted: imagine adjustment function as adj(ticks, runtime), {adj(0, 0.9) + adj(0, 0.9) + ....} = {0 + 0 + ....} = 0. After this patch it will not happen because the adjustment is applied after accumulated. v2: - remove if()s, put new variables into signal_struct. Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com> Acked-by: Peter Zijlstra <peterz@infradead.org> Cc: Spencer Candland <spencer@bluehost.com> Cc: Americo Wang <xiyou.wangcong@gmail.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Stanislaw Gruszka <sgruszka@redhat.com> LKML-Reference: <4B162517.8040909@jp.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-12-02 08:28:07 +00:00
thread_group_times(p, &tgutime, &tgstime);
spin_lock_irq(&p->real_parent->sighand->siglock);
psig = p->real_parent->signal;
sig = p->signal;
psig->cutime += tgutime + sig->cutime;
psig->cstime += tgstime + sig->cstime;
psig->cgtime += p->gtime + sig->gtime + sig->cgtime;
psig->cmin_flt +=
p->min_flt + sig->min_flt + sig->cmin_flt;
psig->cmaj_flt +=
p->maj_flt + sig->maj_flt + sig->cmaj_flt;
psig->cnvcsw +=
p->nvcsw + sig->nvcsw + sig->cnvcsw;
psig->cnivcsw +=
p->nivcsw + sig->nivcsw + sig->cnivcsw;
psig->cinblock +=
task_io_get_inblock(p) +
sig->inblock + sig->cinblock;
psig->coublock +=
task_io_get_oublock(p) +
sig->oublock + sig->coublock;
getrusage: fill ru_maxrss value Make ->ru_maxrss value in struct rusage filled accordingly to rss hiwater mark. This struct is filled as a parameter to getrusage syscall. ->ru_maxrss value is set to KBs which is the way it is done in BSD systems. /usr/bin/time (gnu time) application converts ->ru_maxrss to KBs which seems to be incorrect behavior. Maintainer of this util was notified by me with the patch which corrects it and cc'ed. To make this happen we extend struct signal_struct by two fields. The first one is ->maxrss which we use to store rss hiwater of the task. The second one is ->cmaxrss which we use to store highest rss hiwater of all task childs. These values are used in k_getrusage() to actually fill ->ru_maxrss. k_getrusage() uses current rss hiwater value directly if mm struct exists. Note: exec() clear mm->hiwater_rss, but doesn't clear sig->maxrss. it is intetionally behavior. *BSD getrusage have exec() inheriting. test programs ======================================================== getrusage.c =========== #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include <sys/types.h> #include <sys/wait.h> #include <unistd.h> #include <signal.h> #include <sys/mman.h> #include "common.h" #define err(str) perror(str), exit(1) int main(int argc, char** argv) { int status; printf("allocate 100MB\n"); consume(100); printf("testcase1: fork inherit? \n"); printf(" expect: initial.self ~= child.self\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase2: fork inherit? (cont.) \n"); printf(" expect: initial.children ~= 100MB, but child.children = 0\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("child"); _exit(0); } printf("\n"); printf("testcase3: fork + malloc \n"); printf(" expect: child.self ~= initial.self + 50MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { printf("allocate +50MB\n"); consume(50); show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase4: grandchild maxrss\n"); printf(" expect: post_wait.children ~= 300MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); show_rusage("post_wait"); } else { system("./child -n 0 -g 300"); _exit(0); } printf("\n"); printf("testcase5: zombie\n"); printf(" expect: pre_wait ~= initial, IOW the zombie process is not accounted.\n"); printf(" post_wait ~= 400MB, IOW wait() collect child's max_rss. \n"); show_rusage("initial"); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("pre_wait"); wait(&status); show_rusage("post_wait"); } else { system("./child -n 400"); _exit(0); } printf("\n"); printf("testcase6: SIG_IGN\n"); printf(" expect: initial ~= after_zombie (child's 500MB alloc should be ignored).\n"); show_rusage("initial"); signal(SIGCHLD, SIG_IGN); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("after_zombie"); } else { system("./child -n 500"); _exit(0); } printf("\n"); signal(SIGCHLD, SIG_DFL); printf("testcase7: exec (without fork) \n"); printf(" expect: initial ~= exec \n"); show_rusage("initial"); execl("./child", "child", "-v", NULL); return 0; } child.c ======= #include <sys/types.h> #include <unistd.h> #include <sys/types.h> #include <sys/wait.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include "common.h" int main(int argc, char** argv) { int status; int c; long consume_size = 0; long grandchild_consume_size = 0; int show = 0; while ((c = getopt(argc, argv, "n:g:v")) != -1) { switch (c) { case 'n': consume_size = atol(optarg); break; case 'v': show = 1; break; case 'g': grandchild_consume_size = atol(optarg); break; default: break; } } if (show) show_rusage("exec"); if (consume_size) { printf("child alloc %ldMB\n", consume_size); consume(consume_size); } if (grandchild_consume_size) { if (fork()) { wait(&status); } else { printf("grandchild alloc %ldMB\n", grandchild_consume_size); consume(grandchild_consume_size); exit(0); } } return 0; } common.c ======== #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/time.h> #include <sys/resource.h> #include <sys/types.h> #include <sys/wait.h> #include <unistd.h> #include <signal.h> #include <sys/mman.h> #include "common.h" #define err(str) perror(str), exit(1) void show_rusage(char *prefix) { int err, err2; struct rusage rusage_self; struct rusage rusage_children; printf("%s: ", prefix); err = getrusage(RUSAGE_SELF, &rusage_self); if (!err) printf("self %ld ", rusage_self.ru_maxrss); err2 = getrusage(RUSAGE_CHILDREN, &rusage_children); if (!err2) printf("children %ld ", rusage_children.ru_maxrss); printf("\n"); } /* Some buggy OS need this worthless CPU waste. */ void make_pagefault(void) { void *addr; int size = getpagesize(); int i; for (i=0; i<1000; i++) { addr = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0); if (addr == MAP_FAILED) err("make_pagefault"); memset(addr, 0, size); munmap(addr, size); } } void consume(int mega) { size_t sz = mega * 1024 * 1024; void *ptr; ptr = malloc(sz); memset(ptr, 0, sz); make_pagefault(); } pid_t __fork(void) { pid_t pid; pid = fork(); make_pagefault(); return pid; } common.h ======== void show_rusage(char *prefix); void make_pagefault(void); void consume(int mega); pid_t __fork(void); FreeBSD result (expected result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 103492 children 0 fork child: self 103540 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 103540 children 103540 child: self 103564 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 103564 children 103564 allocate +50MB fork child: self 154860 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 103564 children 154860 grandchild alloc 300MB post_wait: self 103564 children 308720 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 103564 children 308720 child alloc 400MB pre_wait: self 103564 children 308720 post_wait: self 103564 children 411312 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 103564 children 411312 child alloc 500MB after_zombie: self 103624 children 411312 testcase7: exec (without fork) expect: initial ~= exec initial: self 103624 children 411312 exec: self 103624 children 411312 Linux result (actual test result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 102848 children 0 fork child: self 102572 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 102876 children 102644 child: self 102572 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 102876 children 102644 allocate +50MB fork child: self 153804 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 102876 children 153864 grandchild alloc 300MB post_wait: self 102876 children 307536 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 102876 children 307536 child alloc 400MB pre_wait: self 102876 children 307536 post_wait: self 102876 children 410076 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 102876 children 410076 child alloc 500MB after_zombie: self 102880 children 410076 testcase7: exec (without fork) expect: initial ~= exec initial: self 102880 children 410076 exec: self 102880 children 410076 Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-09-22 23:44:10 +00:00
maxrss = max(sig->maxrss, sig->cmaxrss);
if (psig->cmaxrss < maxrss)
psig->cmaxrss = maxrss;
task_io_accounting_add(&psig->ioac, &p->ioac);
task_io_accounting_add(&psig->ioac, &sig->ioac);
spin_unlock_irq(&p->real_parent->sighand->siglock);
}
/*
* Now we are sure this task is interesting, and no other
* thread can reap it because we set its state to EXIT_DEAD.
*/
read_unlock(&tasklist_lock);
retval = wo->wo_rusage
? getrusage(p, RUSAGE_BOTH, wo->wo_rusage) : 0;
status = (p->signal->flags & SIGNAL_GROUP_EXIT)
? p->signal->group_exit_code : p->exit_code;
if (!retval && wo->wo_stat)
retval = put_user(status, wo->wo_stat);
infop = wo->wo_info;
if (!retval && infop)
retval = put_user(SIGCHLD, &infop->si_signo);
if (!retval && infop)
retval = put_user(0, &infop->si_errno);
if (!retval && infop) {
int why;
if ((status & 0x7f) == 0) {
why = CLD_EXITED;
status >>= 8;
} else {
why = (status & 0x80) ? CLD_DUMPED : CLD_KILLED;
status &= 0x7f;
}
retval = put_user((short)why, &infop->si_code);
if (!retval)
retval = put_user(status, &infop->si_status);
}
if (!retval && infop)
retval = put_user(pid, &infop->si_pid);
if (!retval && infop)
retval = put_user(uid, &infop->si_uid);
wait_task_zombie: fix 2/3 races vs forget_original_parent() Two threads, T1 and T2. T2 ptraces P, and P is not a child of ptracer's thread group. P exits and goes to TASK_ZOMBIE. T1 does wait_task_zombie(P): P->exit_state = TASK_DEAD; ... read_unlock(&tasklist_lock); T2 does exit(), takes tasklist, forget_original_parent() does __ptrace_unlink(P) but doesn't call do_notify_parent(P) because p->exit_state == EXIT_DEAD. Now, P is not visible to our process: __ptrace_unlink() removed it from ->children. We should send notification to P->parent and release P if and only if SIGCHLD is ignored. And we have 3 bugs: 1. P->parent does do_wait() and gets -ECHILD (P is on ->parent->children, but its state is TASK_DEAD). 2. // wait_task_zombie() continues if (put_user(...)) { // TODO: is this safe? p->exit_state = EXIT_ZOMBIE; return; } we return without notification/release, task_struct leaked. Solution: ignore -EFAULT and proceed. It is an application's bug if we can't fill infop/stat_addr (in case of VM_FAULT_OOM we have much more problems). 3. // wait_task_zombie() continues if (p->real_parent != p->parent) { // Not taken, it was untraced'ed ... } release_task(p); we released the task which we shouldn't. Solution: check ->real_parent != ->parent before, under tasklist_lock, but use ptrace_unlink() instead of __ptrace_unlink() to check ->ptrace. This patch hopefully solves 2 and 3, the 1st bug will be fixed later, we need some cleanups in forget_original_parent/reparent_thread. However, the first race is very unlikely and not critical, so I hope it makes sense to fix 1 and 2 for now. 4. Small cleanup: don't "restore" EXIT_ZOMBIE unless we know we are not going to realease the child. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Cc: Roland McGrath <roland@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:26:58 +00:00
if (!retval)
retval = pid;
wait_task_zombie: fix 2/3 races vs forget_original_parent() Two threads, T1 and T2. T2 ptraces P, and P is not a child of ptracer's thread group. P exits and goes to TASK_ZOMBIE. T1 does wait_task_zombie(P): P->exit_state = TASK_DEAD; ... read_unlock(&tasklist_lock); T2 does exit(), takes tasklist, forget_original_parent() does __ptrace_unlink(P) but doesn't call do_notify_parent(P) because p->exit_state == EXIT_DEAD. Now, P is not visible to our process: __ptrace_unlink() removed it from ->children. We should send notification to P->parent and release P if and only if SIGCHLD is ignored. And we have 3 bugs: 1. P->parent does do_wait() and gets -ECHILD (P is on ->parent->children, but its state is TASK_DEAD). 2. // wait_task_zombie() continues if (put_user(...)) { // TODO: is this safe? p->exit_state = EXIT_ZOMBIE; return; } we return without notification/release, task_struct leaked. Solution: ignore -EFAULT and proceed. It is an application's bug if we can't fill infop/stat_addr (in case of VM_FAULT_OOM we have much more problems). 3. // wait_task_zombie() continues if (p->real_parent != p->parent) { // Not taken, it was untraced'ed ... } release_task(p); we released the task which we shouldn't. Solution: check ->real_parent != ->parent before, under tasklist_lock, but use ptrace_unlink() instead of __ptrace_unlink() to check ->ptrace. This patch hopefully solves 2 and 3, the 1st bug will be fixed later, we need some cleanups in forget_original_parent/reparent_thread. However, the first race is very unlikely and not critical, so I hope it makes sense to fix 1 and 2 for now. 4. Small cleanup: don't "restore" EXIT_ZOMBIE unless we know we are not going to realease the child. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Cc: Roland McGrath <roland@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:26:58 +00:00
if (traced) {
write_lock_irq(&tasklist_lock);
wait_task_zombie: fix 2/3 races vs forget_original_parent() Two threads, T1 and T2. T2 ptraces P, and P is not a child of ptracer's thread group. P exits and goes to TASK_ZOMBIE. T1 does wait_task_zombie(P): P->exit_state = TASK_DEAD; ... read_unlock(&tasklist_lock); T2 does exit(), takes tasklist, forget_original_parent() does __ptrace_unlink(P) but doesn't call do_notify_parent(P) because p->exit_state == EXIT_DEAD. Now, P is not visible to our process: __ptrace_unlink() removed it from ->children. We should send notification to P->parent and release P if and only if SIGCHLD is ignored. And we have 3 bugs: 1. P->parent does do_wait() and gets -ECHILD (P is on ->parent->children, but its state is TASK_DEAD). 2. // wait_task_zombie() continues if (put_user(...)) { // TODO: is this safe? p->exit_state = EXIT_ZOMBIE; return; } we return without notification/release, task_struct leaked. Solution: ignore -EFAULT and proceed. It is an application's bug if we can't fill infop/stat_addr (in case of VM_FAULT_OOM we have much more problems). 3. // wait_task_zombie() continues if (p->real_parent != p->parent) { // Not taken, it was untraced'ed ... } release_task(p); we released the task which we shouldn't. Solution: check ->real_parent != ->parent before, under tasklist_lock, but use ptrace_unlink() instead of __ptrace_unlink() to check ->ptrace. This patch hopefully solves 2 and 3, the 1st bug will be fixed later, we need some cleanups in forget_original_parent/reparent_thread. However, the first race is very unlikely and not critical, so I hope it makes sense to fix 1 and 2 for now. 4. Small cleanup: don't "restore" EXIT_ZOMBIE unless we know we are not going to realease the child. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Cc: Roland McGrath <roland@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:26:58 +00:00
/* We dropped tasklist, ptracer could die and untrace */
ptrace_unlink(p);
/*
* If this is not a sub-thread, notify the parent.
* If parent wants a zombie, don't release it now.
wait_task_zombie: fix 2/3 races vs forget_original_parent() Two threads, T1 and T2. T2 ptraces P, and P is not a child of ptracer's thread group. P exits and goes to TASK_ZOMBIE. T1 does wait_task_zombie(P): P->exit_state = TASK_DEAD; ... read_unlock(&tasklist_lock); T2 does exit(), takes tasklist, forget_original_parent() does __ptrace_unlink(P) but doesn't call do_notify_parent(P) because p->exit_state == EXIT_DEAD. Now, P is not visible to our process: __ptrace_unlink() removed it from ->children. We should send notification to P->parent and release P if and only if SIGCHLD is ignored. And we have 3 bugs: 1. P->parent does do_wait() and gets -ECHILD (P is on ->parent->children, but its state is TASK_DEAD). 2. // wait_task_zombie() continues if (put_user(...)) { // TODO: is this safe? p->exit_state = EXIT_ZOMBIE; return; } we return without notification/release, task_struct leaked. Solution: ignore -EFAULT and proceed. It is an application's bug if we can't fill infop/stat_addr (in case of VM_FAULT_OOM we have much more problems). 3. // wait_task_zombie() continues if (p->real_parent != p->parent) { // Not taken, it was untraced'ed ... } release_task(p); we released the task which we shouldn't. Solution: check ->real_parent != ->parent before, under tasklist_lock, but use ptrace_unlink() instead of __ptrace_unlink() to check ->ptrace. This patch hopefully solves 2 and 3, the 1st bug will be fixed later, we need some cleanups in forget_original_parent/reparent_thread. However, the first race is very unlikely and not critical, so I hope it makes sense to fix 1 and 2 for now. 4. Small cleanup: don't "restore" EXIT_ZOMBIE unless we know we are not going to realease the child. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Cc: Roland McGrath <roland@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:26:58 +00:00
*/
if (thread_group_leader(p) &&
!do_notify_parent(p, p->exit_signal)) {
p->exit_state = EXIT_ZOMBIE;
p = NULL;
}
write_unlock_irq(&tasklist_lock);
}
if (p != NULL)
release_task(p);
wait_task_zombie: fix 2/3 races vs forget_original_parent() Two threads, T1 and T2. T2 ptraces P, and P is not a child of ptracer's thread group. P exits and goes to TASK_ZOMBIE. T1 does wait_task_zombie(P): P->exit_state = TASK_DEAD; ... read_unlock(&tasklist_lock); T2 does exit(), takes tasklist, forget_original_parent() does __ptrace_unlink(P) but doesn't call do_notify_parent(P) because p->exit_state == EXIT_DEAD. Now, P is not visible to our process: __ptrace_unlink() removed it from ->children. We should send notification to P->parent and release P if and only if SIGCHLD is ignored. And we have 3 bugs: 1. P->parent does do_wait() and gets -ECHILD (P is on ->parent->children, but its state is TASK_DEAD). 2. // wait_task_zombie() continues if (put_user(...)) { // TODO: is this safe? p->exit_state = EXIT_ZOMBIE; return; } we return without notification/release, task_struct leaked. Solution: ignore -EFAULT and proceed. It is an application's bug if we can't fill infop/stat_addr (in case of VM_FAULT_OOM we have much more problems). 3. // wait_task_zombie() continues if (p->real_parent != p->parent) { // Not taken, it was untraced'ed ... } release_task(p); we released the task which we shouldn't. Solution: check ->real_parent != ->parent before, under tasklist_lock, but use ptrace_unlink() instead of __ptrace_unlink() to check ->ptrace. This patch hopefully solves 2 and 3, the 1st bug will be fixed later, we need some cleanups in forget_original_parent/reparent_thread. However, the first race is very unlikely and not critical, so I hope it makes sense to fix 1 and 2 for now. 4. Small cleanup: don't "restore" EXIT_ZOMBIE unless we know we are not going to realease the child. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Cc: Roland McGrath <roland@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:26:58 +00:00
return retval;
}
do_wait: fix waiting for the group stop with the dead leader do_wait(WSTOPPED) assumes that p->state must be == TASK_STOPPED, this is not true if the leader is already dead. Check SIGNAL_STOP_STOPPED instead and use signal->group_exit_code. Trivial test-case: void *tfunc(void *arg) { pause(); return NULL; } int main(void) { pthread_t thr; pthread_create(&thr, NULL, tfunc, NULL); pthread_exit(NULL); return 0; } It doesn't react to ^Z (and then to ^C or ^\). The task is stopped, but bash can't see this. The bug is very old, and it was reported multiple times. This patch was sent more than a year ago (http://marc.info/?t=119713920000003) but it was ignored. This change also fixes other oddities (but not all) in this area. For example, before this patch: $ sleep 100 ^Z [1]+ Stopped sleep 100 $ strace -p `pidof sleep` Process 11442 attached - interrupt to quit strace hangs in do_wait(), because ->exit_code was already consumed by bash. After this patch, strace happily proceeds: --- SIGTSTP (Stopped) @ 0 (0) --- restart_syscall(<... resuming interrupted call ...> To me, this looks much more "natural" and correct. Another example. Let's suppose we have the main thread M and sub-thread T, the process is stopped, and its parent did wait(WSTOPPED). Now we can ptrace T but not M. This looks at least strange to me. Imho, do_wait() should not confuse the per-thread ptrace stops with the per-process job control stops. Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Jan Kratochvil <jan.kratochvil@redhat.com> Cc: Kaz Kylheku <kkylheku@gmail.com> Cc: Michael Kerrisk <mtk.manpages@googlemail.com> Cc: Roland McGrath <roland@redhat.com> Cc: Ulrich Drepper <drepper@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:57:58 +00:00
static int *task_stopped_code(struct task_struct *p, bool ptrace)
{
if (ptrace) {
ptrace: implement PTRACE_LISTEN The previous patch implemented async notification for ptrace but it only worked while trace is running. This patch introduces PTRACE_LISTEN which is suggested by Oleg Nestrov. It's allowed iff tracee is in STOP trap and puts tracee into quasi-running state - tracee never really runs but wait(2) and ptrace(2) consider it to be running. While ptracer is listening, tracee is allowed to re-enter STOP to notify an async event. Listening state is cleared on the first notification. Ptracer can also clear it by issuing INTERRUPT - tracee will re-trap into STOP with listening state cleared. This allows ptracer to monitor group stop state without running tracee - use INTERRUPT to put tracee into STOP trap, issue LISTEN and then wait(2) to wait for the next group stop event. When it happens, PTRACE_GETSIGINFO provides information to determine the current state. Test program follows. #define PTRACE_SEIZE 0x4206 #define PTRACE_INTERRUPT 0x4207 #define PTRACE_LISTEN 0x4208 #define PTRACE_SEIZE_DEVEL 0x80000000 static const struct timespec ts1s = { .tv_sec = 1 }; int main(int argc, char **argv) { pid_t tracee, tracer; int i; tracee = fork(); if (!tracee) while (1) pause(); tracer = fork(); if (!tracer) { siginfo_t si; ptrace(PTRACE_SEIZE, tracee, NULL, (void *)(unsigned long)PTRACE_SEIZE_DEVEL); ptrace(PTRACE_INTERRUPT, tracee, NULL, NULL); repeat: waitid(P_PID, tracee, NULL, WSTOPPED); ptrace(PTRACE_GETSIGINFO, tracee, NULL, &si); if (!si.si_code) { printf("tracer: SIG %d\n", si.si_signo); ptrace(PTRACE_CONT, tracee, NULL, (void *)(unsigned long)si.si_signo); goto repeat; } printf("tracer: stopped=%d signo=%d\n", si.si_signo != SIGTRAP, si.si_signo); if (si.si_signo != SIGTRAP) ptrace(PTRACE_LISTEN, tracee, NULL, NULL); else ptrace(PTRACE_CONT, tracee, NULL, NULL); goto repeat; } for (i = 0; i < 3; i++) { nanosleep(&ts1s, NULL); printf("mother: SIGSTOP\n"); kill(tracee, SIGSTOP); nanosleep(&ts1s, NULL); printf("mother: SIGCONT\n"); kill(tracee, SIGCONT); } nanosleep(&ts1s, NULL); kill(tracer, SIGKILL); kill(tracee, SIGKILL); return 0; } This is identical to the program to test TRAP_NOTIFY except that tracee is PTRACE_LISTEN'd instead of PTRACE_CONT'd when group stopped. This allows ptracer to monitor when group stop ends without running tracee. # ./test-listen tracer: stopped=0 signo=5 mother: SIGSTOP tracer: SIG 19 tracer: stopped=1 signo=19 mother: SIGCONT tracer: stopped=0 signo=5 tracer: SIG 18 mother: SIGSTOP tracer: SIG 19 tracer: stopped=1 signo=19 mother: SIGCONT tracer: stopped=0 signo=5 tracer: SIG 18 mother: SIGSTOP tracer: SIG 19 tracer: stopped=1 signo=19 mother: SIGCONT tracer: stopped=0 signo=5 tracer: SIG 18 -v2: Moved JOBCTL_LISTENING check in wait_task_stopped() into task_stopped_code() as suggested by Oleg. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com>
2011-06-14 09:20:18 +00:00
if (task_is_stopped_or_traced(p) &&
!(p->jobctl & JOBCTL_LISTENING))
do_wait: fix waiting for the group stop with the dead leader do_wait(WSTOPPED) assumes that p->state must be == TASK_STOPPED, this is not true if the leader is already dead. Check SIGNAL_STOP_STOPPED instead and use signal->group_exit_code. Trivial test-case: void *tfunc(void *arg) { pause(); return NULL; } int main(void) { pthread_t thr; pthread_create(&thr, NULL, tfunc, NULL); pthread_exit(NULL); return 0; } It doesn't react to ^Z (and then to ^C or ^\). The task is stopped, but bash can't see this. The bug is very old, and it was reported multiple times. This patch was sent more than a year ago (http://marc.info/?t=119713920000003) but it was ignored. This change also fixes other oddities (but not all) in this area. For example, before this patch: $ sleep 100 ^Z [1]+ Stopped sleep 100 $ strace -p `pidof sleep` Process 11442 attached - interrupt to quit strace hangs in do_wait(), because ->exit_code was already consumed by bash. After this patch, strace happily proceeds: --- SIGTSTP (Stopped) @ 0 (0) --- restart_syscall(<... resuming interrupted call ...> To me, this looks much more "natural" and correct. Another example. Let's suppose we have the main thread M and sub-thread T, the process is stopped, and its parent did wait(WSTOPPED). Now we can ptrace T but not M. This looks at least strange to me. Imho, do_wait() should not confuse the per-thread ptrace stops with the per-process job control stops. Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Jan Kratochvil <jan.kratochvil@redhat.com> Cc: Kaz Kylheku <kkylheku@gmail.com> Cc: Michael Kerrisk <mtk.manpages@googlemail.com> Cc: Roland McGrath <roland@redhat.com> Cc: Ulrich Drepper <drepper@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:57:58 +00:00
return &p->exit_code;
} else {
if (p->signal->flags & SIGNAL_STOP_STOPPED)
return &p->signal->group_exit_code;
}
return NULL;
}
/**
* wait_task_stopped - Wait for %TASK_STOPPED or %TASK_TRACED
* @wo: wait options
* @ptrace: is the wait for ptrace
* @p: task to wait for
*
* Handle sys_wait4() work for %p in state %TASK_STOPPED or %TASK_TRACED.
*
* CONTEXT:
* read_lock(&tasklist_lock), which is released if return value is
* non-zero. Also, grabs and releases @p->sighand->siglock.
*
* RETURNS:
* 0 if wait condition didn't exist and search for other wait conditions
* should continue. Non-zero return, -errno on failure and @p's pid on
* success, implies that tasklist_lock is released and wait condition
* search should terminate.
*/
static int wait_task_stopped(struct wait_opts *wo,
int ptrace, struct task_struct *p)
{
struct siginfo __user *infop;
do_wait: fix waiting for the group stop with the dead leader do_wait(WSTOPPED) assumes that p->state must be == TASK_STOPPED, this is not true if the leader is already dead. Check SIGNAL_STOP_STOPPED instead and use signal->group_exit_code. Trivial test-case: void *tfunc(void *arg) { pause(); return NULL; } int main(void) { pthread_t thr; pthread_create(&thr, NULL, tfunc, NULL); pthread_exit(NULL); return 0; } It doesn't react to ^Z (and then to ^C or ^\). The task is stopped, but bash can't see this. The bug is very old, and it was reported multiple times. This patch was sent more than a year ago (http://marc.info/?t=119713920000003) but it was ignored. This change also fixes other oddities (but not all) in this area. For example, before this patch: $ sleep 100 ^Z [1]+ Stopped sleep 100 $ strace -p `pidof sleep` Process 11442 attached - interrupt to quit strace hangs in do_wait(), because ->exit_code was already consumed by bash. After this patch, strace happily proceeds: --- SIGTSTP (Stopped) @ 0 (0) --- restart_syscall(<... resuming interrupted call ...> To me, this looks much more "natural" and correct. Another example. Let's suppose we have the main thread M and sub-thread T, the process is stopped, and its parent did wait(WSTOPPED). Now we can ptrace T but not M. This looks at least strange to me. Imho, do_wait() should not confuse the per-thread ptrace stops with the per-process job control stops. Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Jan Kratochvil <jan.kratochvil@redhat.com> Cc: Kaz Kylheku <kkylheku@gmail.com> Cc: Michael Kerrisk <mtk.manpages@googlemail.com> Cc: Roland McGrath <roland@redhat.com> Cc: Ulrich Drepper <drepper@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:57:58 +00:00
int retval, exit_code, *p_code, why;
uid_t uid = 0; /* unneeded, required by compiler */
pid_t pid;
/*
* Traditionally we see ptrace'd stopped tasks regardless of options.
*/
if (!ptrace && !(wo->wo_flags & WUNTRACED))
return 0;
if (!task_stopped_code(p, ptrace))
return 0;
exit_code = 0;
spin_lock_irq(&p->sighand->siglock);
do_wait: fix waiting for the group stop with the dead leader do_wait(WSTOPPED) assumes that p->state must be == TASK_STOPPED, this is not true if the leader is already dead. Check SIGNAL_STOP_STOPPED instead and use signal->group_exit_code. Trivial test-case: void *tfunc(void *arg) { pause(); return NULL; } int main(void) { pthread_t thr; pthread_create(&thr, NULL, tfunc, NULL); pthread_exit(NULL); return 0; } It doesn't react to ^Z (and then to ^C or ^\). The task is stopped, but bash can't see this. The bug is very old, and it was reported multiple times. This patch was sent more than a year ago (http://marc.info/?t=119713920000003) but it was ignored. This change also fixes other oddities (but not all) in this area. For example, before this patch: $ sleep 100 ^Z [1]+ Stopped sleep 100 $ strace -p `pidof sleep` Process 11442 attached - interrupt to quit strace hangs in do_wait(), because ->exit_code was already consumed by bash. After this patch, strace happily proceeds: --- SIGTSTP (Stopped) @ 0 (0) --- restart_syscall(<... resuming interrupted call ...> To me, this looks much more "natural" and correct. Another example. Let's suppose we have the main thread M and sub-thread T, the process is stopped, and its parent did wait(WSTOPPED). Now we can ptrace T but not M. This looks at least strange to me. Imho, do_wait() should not confuse the per-thread ptrace stops with the per-process job control stops. Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Jan Kratochvil <jan.kratochvil@redhat.com> Cc: Kaz Kylheku <kkylheku@gmail.com> Cc: Michael Kerrisk <mtk.manpages@googlemail.com> Cc: Roland McGrath <roland@redhat.com> Cc: Ulrich Drepper <drepper@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:57:58 +00:00
p_code = task_stopped_code(p, ptrace);
if (unlikely(!p_code))
goto unlock_sig;
do_wait: fix waiting for the group stop with the dead leader do_wait(WSTOPPED) assumes that p->state must be == TASK_STOPPED, this is not true if the leader is already dead. Check SIGNAL_STOP_STOPPED instead and use signal->group_exit_code. Trivial test-case: void *tfunc(void *arg) { pause(); return NULL; } int main(void) { pthread_t thr; pthread_create(&thr, NULL, tfunc, NULL); pthread_exit(NULL); return 0; } It doesn't react to ^Z (and then to ^C or ^\). The task is stopped, but bash can't see this. The bug is very old, and it was reported multiple times. This patch was sent more than a year ago (http://marc.info/?t=119713920000003) but it was ignored. This change also fixes other oddities (but not all) in this area. For example, before this patch: $ sleep 100 ^Z [1]+ Stopped sleep 100 $ strace -p `pidof sleep` Process 11442 attached - interrupt to quit strace hangs in do_wait(), because ->exit_code was already consumed by bash. After this patch, strace happily proceeds: --- SIGTSTP (Stopped) @ 0 (0) --- restart_syscall(<... resuming interrupted call ...> To me, this looks much more "natural" and correct. Another example. Let's suppose we have the main thread M and sub-thread T, the process is stopped, and its parent did wait(WSTOPPED). Now we can ptrace T but not M. This looks at least strange to me. Imho, do_wait() should not confuse the per-thread ptrace stops with the per-process job control stops. Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Jan Kratochvil <jan.kratochvil@redhat.com> Cc: Kaz Kylheku <kkylheku@gmail.com> Cc: Michael Kerrisk <mtk.manpages@googlemail.com> Cc: Roland McGrath <roland@redhat.com> Cc: Ulrich Drepper <drepper@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:57:58 +00:00
exit_code = *p_code;
if (!exit_code)
goto unlock_sig;
if (!unlikely(wo->wo_flags & WNOWAIT))
do_wait: fix waiting for the group stop with the dead leader do_wait(WSTOPPED) assumes that p->state must be == TASK_STOPPED, this is not true if the leader is already dead. Check SIGNAL_STOP_STOPPED instead and use signal->group_exit_code. Trivial test-case: void *tfunc(void *arg) { pause(); return NULL; } int main(void) { pthread_t thr; pthread_create(&thr, NULL, tfunc, NULL); pthread_exit(NULL); return 0; } It doesn't react to ^Z (and then to ^C or ^\). The task is stopped, but bash can't see this. The bug is very old, and it was reported multiple times. This patch was sent more than a year ago (http://marc.info/?t=119713920000003) but it was ignored. This change also fixes other oddities (but not all) in this area. For example, before this patch: $ sleep 100 ^Z [1]+ Stopped sleep 100 $ strace -p `pidof sleep` Process 11442 attached - interrupt to quit strace hangs in do_wait(), because ->exit_code was already consumed by bash. After this patch, strace happily proceeds: --- SIGTSTP (Stopped) @ 0 (0) --- restart_syscall(<... resuming interrupted call ...> To me, this looks much more "natural" and correct. Another example. Let's suppose we have the main thread M and sub-thread T, the process is stopped, and its parent did wait(WSTOPPED). Now we can ptrace T but not M. This looks at least strange to me. Imho, do_wait() should not confuse the per-thread ptrace stops with the per-process job control stops. Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Jan Kratochvil <jan.kratochvil@redhat.com> Cc: Kaz Kylheku <kkylheku@gmail.com> Cc: Michael Kerrisk <mtk.manpages@googlemail.com> Cc: Roland McGrath <roland@redhat.com> Cc: Ulrich Drepper <drepper@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:57:58 +00:00
*p_code = 0;
uid = from_kuid_munged(current_user_ns(), task_uid(p));
unlock_sig:
spin_unlock_irq(&p->sighand->siglock);
if (!exit_code)
return 0;
/*
* Now we are pretty sure this task is interesting.
* Make sure it doesn't get reaped out from under us while we
* give up the lock and then examine it below. We don't want to
* keep holding onto the tasklist_lock while we call getrusage and
* possibly take page faults for user memory.
*/
get_task_struct(p);
pid = task_pid_vnr(p);
why = ptrace ? CLD_TRAPPED : CLD_STOPPED;
read_unlock(&tasklist_lock);
if (unlikely(wo->wo_flags & WNOWAIT))
return wait_noreap_copyout(wo, p, pid, uid, why, exit_code);
retval = wo->wo_rusage
? getrusage(p, RUSAGE_BOTH, wo->wo_rusage) : 0;
if (!retval && wo->wo_stat)
retval = put_user((exit_code << 8) | 0x7f, wo->wo_stat);
infop = wo->wo_info;
if (!retval && infop)
retval = put_user(SIGCHLD, &infop->si_signo);
if (!retval && infop)
retval = put_user(0, &infop->si_errno);
if (!retval && infop)
retval = put_user((short)why, &infop->si_code);
if (!retval && infop)
retval = put_user(exit_code, &infop->si_status);
if (!retval && infop)
retval = put_user(pid, &infop->si_pid);
if (!retval && infop)
retval = put_user(uid, &infop->si_uid);
if (!retval)
retval = pid;
put_task_struct(p);
BUG_ON(!retval);
return retval;
}
/*
* Handle do_wait work for one task in a live, non-stopped state.
* read_lock(&tasklist_lock) on entry. If we return zero, we still hold
* the lock and this task is uninteresting. If we return nonzero, we have
* released the lock and the system call should return.
*/
static int wait_task_continued(struct wait_opts *wo, struct task_struct *p)
{
int retval;
pid_t pid;
uid_t uid;
if (!unlikely(wo->wo_flags & WCONTINUED))
return 0;
if (!(p->signal->flags & SIGNAL_STOP_CONTINUED))
return 0;
spin_lock_irq(&p->sighand->siglock);
/* Re-check with the lock held. */
if (!(p->signal->flags & SIGNAL_STOP_CONTINUED)) {
spin_unlock_irq(&p->sighand->siglock);
return 0;
}
if (!unlikely(wo->wo_flags & WNOWAIT))
p->signal->flags &= ~SIGNAL_STOP_CONTINUED;
uid = from_kuid_munged(current_user_ns(), task_uid(p));
spin_unlock_irq(&p->sighand->siglock);
pid = task_pid_vnr(p);
get_task_struct(p);
read_unlock(&tasklist_lock);
if (!wo->wo_info) {
retval = wo->wo_rusage
? getrusage(p, RUSAGE_BOTH, wo->wo_rusage) : 0;
put_task_struct(p);
if (!retval && wo->wo_stat)
retval = put_user(0xffff, wo->wo_stat);
if (!retval)
retval = pid;
} else {
retval = wait_noreap_copyout(wo, p, pid, uid,
CLD_CONTINUED, SIGCONT);
BUG_ON(retval == 0);
}
return retval;
}
/*
* Consider @p for a wait by @parent.
*
* -ECHILD should be in ->notask_error before the first call.
* Returns nonzero for a final return, when we have unlocked tasklist_lock.
* Returns zero if the search for a child should continue;
* then ->notask_error is 0 if @p is an eligible child,
* or another error from security_task_wait(), or still -ECHILD.
*/
static int wait_consider_task(struct wait_opts *wo, int ptrace,
struct task_struct *p)
{
int ret = eligible_child(wo, p);
if (!ret)
return ret;
ret = security_task_wait(p);
if (unlikely(ret < 0)) {
/*
* If we have not yet seen any eligible child,
* then let this error code replace -ECHILD.
* A permission error will give the user a clue
* to look for security policy problems, rather
* than for mysterious wait bugs.
*/
if (wo->notask_error)
wo->notask_error = ret;
return 0;
}
/* dead body doesn't have much to contribute */
ptrace: partially fix the do_wait(WEXITED) vs EXIT_DEAD->EXIT_ZOMBIE race Test-case: int main(void) { int pid, status; pid = fork(); if (!pid) { for (;;) { if (!fork()) return 0; if (waitpid(-1, &status, 0) < 0) { printf("ERR!! wait: %m\n"); return 0; } } } assert(ptrace(PTRACE_ATTACH, pid, 0,0) == 0); assert(waitpid(-1, NULL, 0) == pid); assert(ptrace(PTRACE_SETOPTIONS, pid, 0, PTRACE_O_TRACEFORK) == 0); do { ptrace(PTRACE_CONT, pid, 0, 0); pid = waitpid(-1, NULL, 0); } while (pid > 0); return 1; } It fails because ->real_parent sees its child in EXIT_DEAD state while the tracer is going to change the state back to EXIT_ZOMBIE in wait_task_zombie(). The offending commit is 823b018e which moved the EXIT_DEAD check, but in fact we should not blame it. The original code was not correct as well because it didn't take ptrace_reparented() into account and because we can't really trust ->ptrace. This patch adds the additional check to close this particular race but it doesn't solve the whole problem. We simply can't rely on ->ptrace in this case, it can be cleared if the tracer is multithreaded by the exiting ->parent. I think we should kill EXIT_DEAD altogether, we should always remove the soon-to-be-reaped child from ->children or at least we should never do the DEAD->ZOMBIE transition. But this is too complex for 3.2. Reported-and-tested-by: Denys Vlasenko <vda.linux@googlemail.com> Tested-by: Lukasz Michalik <lmi@ift.uni.wroc.pl> Acked-by: Tejun Heo <tj@kernel.org> Cc: <stable@kernel.org> [3.0+] Signed-off-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-04 16:29:02 +00:00
if (unlikely(p->exit_state == EXIT_DEAD)) {
/*
* But do not ignore this task until the tracer does
* wait_task_zombie()->do_notify_parent().
*/
if (likely(!ptrace) && unlikely(ptrace_reparented(p)))
wo->notask_error = 0;
return 0;
ptrace: partially fix the do_wait(WEXITED) vs EXIT_DEAD->EXIT_ZOMBIE race Test-case: int main(void) { int pid, status; pid = fork(); if (!pid) { for (;;) { if (!fork()) return 0; if (waitpid(-1, &status, 0) < 0) { printf("ERR!! wait: %m\n"); return 0; } } } assert(ptrace(PTRACE_ATTACH, pid, 0,0) == 0); assert(waitpid(-1, NULL, 0) == pid); assert(ptrace(PTRACE_SETOPTIONS, pid, 0, PTRACE_O_TRACEFORK) == 0); do { ptrace(PTRACE_CONT, pid, 0, 0); pid = waitpid(-1, NULL, 0); } while (pid > 0); return 1; } It fails because ->real_parent sees its child in EXIT_DEAD state while the tracer is going to change the state back to EXIT_ZOMBIE in wait_task_zombie(). The offending commit is 823b018e which moved the EXIT_DEAD check, but in fact we should not blame it. The original code was not correct as well because it didn't take ptrace_reparented() into account and because we can't really trust ->ptrace. This patch adds the additional check to close this particular race but it doesn't solve the whole problem. We simply can't rely on ->ptrace in this case, it can be cleared if the tracer is multithreaded by the exiting ->parent. I think we should kill EXIT_DEAD altogether, we should always remove the soon-to-be-reaped child from ->children or at least we should never do the DEAD->ZOMBIE transition. But this is too complex for 3.2. Reported-and-tested-by: Denys Vlasenko <vda.linux@googlemail.com> Tested-by: Lukasz Michalik <lmi@ift.uni.wroc.pl> Acked-by: Tejun Heo <tj@kernel.org> Cc: <stable@kernel.org> [3.0+] Signed-off-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-04 16:29:02 +00:00
}
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
/* slay zombie? */
if (p->exit_state == EXIT_ZOMBIE) {
/*
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
* A zombie ptracee is only visible to its ptracer.
* Notification and reaping will be cascaded to the real
* parent when the ptracer detaches.
*/
if (likely(!ptrace) && unlikely(p->ptrace)) {
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
/* it will become visible, clear notask_error */
wo->notask_error = 0;
return 0;
}
job control: Fix ptracer wait(2) hang and explain notask_error clearing wait(2) and friends allow access to stopped/continued states through zombies, which is required as the states are process-wide and should be accessible whether the leader task is alive or undead. wait_consider_task() implements this by always clearing notask_error and going through wait_task_stopped/continued() for unreaped zombies. However, while ptraced, the stopped state is per-task and as such if the ptracee became a zombie, there's no further stopped event to listen to and wait(2) and friends should return -ECHILD on the tracee. Fix it by clearing notask_error only if WCONTINUED | WEXITED is set for ptraced zombies. While at it, document why clearing notask_error is safe for each case. Test case follows. #include <stdio.h> #include <unistd.h> #include <pthread.h> #include <time.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> static void *nooper(void *arg) { pause(); return NULL; } int main(void) { const struct timespec ts1s = { .tv_sec = 1 }; pid_t tracee, tracer; siginfo_t si; tracee = fork(); if (tracee == 0) { pthread_t thr; pthread_create(&thr, NULL, nooper, NULL); nanosleep(&ts1s, NULL); printf("tracee exiting\n"); pthread_exit(NULL); /* let subthread run */ } tracer = fork(); if (tracer == 0) { ptrace(PTRACE_ATTACH, tracee, NULL, NULL); while (1) { if (waitid(P_PID, tracee, &si, WSTOPPED) < 0) { perror("waitid"); break; } ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } return 0; } waitid(P_PID, tracer, &si, WEXITED); kill(tracee, SIGKILL); return 0; } Before the patch, after the tracee becomes a zombie, the tracer's waitid(WSTOPPED) never returns and the program doesn't terminate. tracee exiting ^C After the patch, tracee exiting triggers waitid() to fail. tracee exiting waitid: No child processes -v2: Oleg pointed out that exited in addition to continued can happen for ptraced dead group leader. Clear notask_error for ptraced child on WEXITED too. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
/* we don't reap group leaders with subthreads */
if (!delay_group_leader(p))
return wait_task_zombie(wo, p);
/*
job control: Fix ptracer wait(2) hang and explain notask_error clearing wait(2) and friends allow access to stopped/continued states through zombies, which is required as the states are process-wide and should be accessible whether the leader task is alive or undead. wait_consider_task() implements this by always clearing notask_error and going through wait_task_stopped/continued() for unreaped zombies. However, while ptraced, the stopped state is per-task and as such if the ptracee became a zombie, there's no further stopped event to listen to and wait(2) and friends should return -ECHILD on the tracee. Fix it by clearing notask_error only if WCONTINUED | WEXITED is set for ptraced zombies. While at it, document why clearing notask_error is safe for each case. Test case follows. #include <stdio.h> #include <unistd.h> #include <pthread.h> #include <time.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> static void *nooper(void *arg) { pause(); return NULL; } int main(void) { const struct timespec ts1s = { .tv_sec = 1 }; pid_t tracee, tracer; siginfo_t si; tracee = fork(); if (tracee == 0) { pthread_t thr; pthread_create(&thr, NULL, nooper, NULL); nanosleep(&ts1s, NULL); printf("tracee exiting\n"); pthread_exit(NULL); /* let subthread run */ } tracer = fork(); if (tracer == 0) { ptrace(PTRACE_ATTACH, tracee, NULL, NULL); while (1) { if (waitid(P_PID, tracee, &si, WSTOPPED) < 0) { perror("waitid"); break; } ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } return 0; } waitid(P_PID, tracer, &si, WEXITED); kill(tracee, SIGKILL); return 0; } Before the patch, after the tracee becomes a zombie, the tracer's waitid(WSTOPPED) never returns and the program doesn't terminate. tracee exiting ^C After the patch, tracee exiting triggers waitid() to fail. tracee exiting waitid: No child processes -v2: Oleg pointed out that exited in addition to continued can happen for ptraced dead group leader. Clear notask_error for ptraced child on WEXITED too. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
* Allow access to stopped/continued state via zombie by
* falling through. Clearing of notask_error is complex.
*
* When !@ptrace:
*
* If WEXITED is set, notask_error should naturally be
* cleared. If not, subset of WSTOPPED|WCONTINUED is set,
* so, if there are live subthreads, there are events to
* wait for. If all subthreads are dead, it's still safe
* to clear - this function will be called again in finite
* amount time once all the subthreads are released and
* will then return without clearing.
*
* When @ptrace:
*
* Stopped state is per-task and thus can't change once the
* target task dies. Only continued and exited can happen.
* Clear notask_error if WCONTINUED | WEXITED.
*/
if (likely(!ptrace) || (wo->wo_flags & (WCONTINUED | WEXITED)))
wo->notask_error = 0;
} else {
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
/*
* If @p is ptraced by a task in its real parent's group,
* hide group stop/continued state when looking at @p as
* the real parent; otherwise, a single stop can be
* reported twice as group and ptrace stops.
*
* If a ptracer wants to distinguish the two events for its
* own children, it should create a separate process which
* takes the role of real parent.
*/
if (likely(!ptrace) && p->ptrace && !ptrace_reparented(p))
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
return 0;
job control: Fix ptracer wait(2) hang and explain notask_error clearing wait(2) and friends allow access to stopped/continued states through zombies, which is required as the states are process-wide and should be accessible whether the leader task is alive or undead. wait_consider_task() implements this by always clearing notask_error and going through wait_task_stopped/continued() for unreaped zombies. However, while ptraced, the stopped state is per-task and as such if the ptracee became a zombie, there's no further stopped event to listen to and wait(2) and friends should return -ECHILD on the tracee. Fix it by clearing notask_error only if WCONTINUED | WEXITED is set for ptraced zombies. While at it, document why clearing notask_error is safe for each case. Test case follows. #include <stdio.h> #include <unistd.h> #include <pthread.h> #include <time.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> static void *nooper(void *arg) { pause(); return NULL; } int main(void) { const struct timespec ts1s = { .tv_sec = 1 }; pid_t tracee, tracer; siginfo_t si; tracee = fork(); if (tracee == 0) { pthread_t thr; pthread_create(&thr, NULL, nooper, NULL); nanosleep(&ts1s, NULL); printf("tracee exiting\n"); pthread_exit(NULL); /* let subthread run */ } tracer = fork(); if (tracer == 0) { ptrace(PTRACE_ATTACH, tracee, NULL, NULL); while (1) { if (waitid(P_PID, tracee, &si, WSTOPPED) < 0) { perror("waitid"); break; } ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } return 0; } waitid(P_PID, tracer, &si, WEXITED); kill(tracee, SIGKILL); return 0; } Before the patch, after the tracee becomes a zombie, the tracer's waitid(WSTOPPED) never returns and the program doesn't terminate. tracee exiting ^C After the patch, tracee exiting triggers waitid() to fail. tracee exiting waitid: No child processes -v2: Oleg pointed out that exited in addition to continued can happen for ptraced dead group leader. Clear notask_error for ptraced child on WEXITED too. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
/*
* @p is alive and it's gonna stop, continue or exit, so
* there always is something to wait for.
*/
wo->notask_error = 0;
}
/*
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
* Wait for stopped. Depending on @ptrace, different stopped state
* is used and the two don't interact with each other.
*/
ret = wait_task_stopped(wo, ptrace, p);
if (ret)
return ret;
/*
job control: Allow access to job control events through ptracees Currently a real parent can't access job control stopped/continued events through a ptraced child. This utterly breaks job control when the children are ptraced. For example, if a program is run from an interactive shell and then strace(1) attaches to it, pressing ^Z would send SIGTSTP and strace(1) would notice it but the shell has no way to tell whether the child entered job control stop and thus can't tell when to take over the terminal - leading to awkward lone ^Z on the terminal. Because the job control and ptrace stopped states are independent, there is no reason to prevent real parents from accessing the stopped state regardless of ptrace. The continued state isn't separate but ptracers don't have any use for them as ptracees can never resume without explicit command from their ptracers, so as long as ptracers don't consume it, it should be fine. Although this is a behavior change, because the previous behavior is utterly broken when viewed from real parents and the change is only visible to real parents, I don't think it's necessary to make this behavior optional. One situation to be careful about is when a task from the real parent's group is ptracing. The parent group is the recipient of both ptrace and job control stop events and one stop can be reported as both job control and ptrace stops. As this can break the current ptrace users, suppress job control stopped events for these cases. If a real parent ptracer wants to know about both job control and ptrace stops, it can create a separate process to serve the role of real parent. Note that this only updates wait(2) side of things. The real parent can access the states via wait(2) but still is not properly notified (woken up and delivered signal). Test case polls wait(2) with WNOHANG to work around. Notification will be updated by future patches. Test case follows. #include <stdio.h> #include <unistd.h> #include <time.h> #include <errno.h> #include <sys/types.h> #include <sys/ptrace.h> #include <sys/wait.h> int main(void) { const struct timespec ts100ms = { .tv_nsec = 100000000 }; pid_t tracee, tracer; siginfo_t si; int i; tracee = fork(); if (tracee == 0) { while (1) { printf("tracee: SIGSTOP\n"); raise(SIGSTOP); nanosleep(&ts100ms, NULL); printf("tracee: SIGCONT\n"); raise(SIGCONT); nanosleep(&ts100ms, NULL); } } waitid(P_PID, tracee, &si, WSTOPPED | WNOHANG | WNOWAIT); tracer = fork(); if (tracer == 0) { nanosleep(&ts100ms, NULL); ptrace(PTRACE_ATTACH, tracee, NULL, NULL); for (i = 0; i < 11; i++) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED); if (si.si_pid && si.si_code == CLD_TRAPPED) ptrace(PTRACE_CONT, tracee, NULL, (void *)(long)si.si_status); } printf("tracer: EXITING\n"); return 0; } while (1) { si.si_pid = 0; waitid(P_PID, tracee, &si, WSTOPPED | WCONTINUED | WEXITED | WNOHANG); if (si.si_pid) printf("mommy : WAIT status=%02d code=%02d\n", si.si_status, si.si_code); nanosleep(&ts100ms, NULL); } return 0; } Before the patch, while ptraced, the parent can't see any job control events. tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracee: SIGCONT tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C After the patch, tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP mommy : WAIT status=19 code=05 tracee: SIGCONT mommy : WAIT status=18 code=06 tracee: SIGSTOP tracer: EXITING mommy : WAIT status=19 code=05 ^C -v2: Oleg pointed out that wait(2) should be suppressed for the real parent's group instead of only the real parent task itself. Updated accordingly. Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Oleg Nesterov <oleg@redhat.com>
2011-03-23 09:37:01 +00:00
* Wait for continued. There's only one continued state and the
* ptracer can consume it which can confuse the real parent. Don't
* use WCONTINUED from ptracer. You don't need or want it.
*/
return wait_task_continued(wo, p);
}
/*
* Do the work of do_wait() for one thread in the group, @tsk.
*
* -ECHILD should be in ->notask_error before the first call.
* Returns nonzero for a final return, when we have unlocked tasklist_lock.
* Returns zero if the search for a child should continue; then
* ->notask_error is 0 if there were any eligible children,
* or another error from security_task_wait(), or still -ECHILD.
*/
static int do_wait_thread(struct wait_opts *wo, struct task_struct *tsk)
{
struct task_struct *p;
list_for_each_entry(p, &tsk->children, sibling) {
int ret = wait_consider_task(wo, 0, p);
if (ret)
return ret;
}
return 0;
}
static int ptrace_do_wait(struct wait_opts *wo, struct task_struct *tsk)
{
struct task_struct *p;
list_for_each_entry(p, &tsk->ptraced, ptrace_entry) {
int ret = wait_consider_task(wo, 1, p);
if (ret)
return ret;
}
return 0;
}
static int child_wait_callback(wait_queue_t *wait, unsigned mode,
int sync, void *key)
{
struct wait_opts *wo = container_of(wait, struct wait_opts,
child_wait);
struct task_struct *p = key;
if (!eligible_pid(wo, p))
return 0;
if ((wo->wo_flags & __WNOTHREAD) && wait->private != p->parent)
return 0;
return default_wake_function(wait, mode, sync, key);
}
void __wake_up_parent(struct task_struct *p, struct task_struct *parent)
{
__wake_up_sync_key(&parent->signal->wait_chldexit,
TASK_INTERRUPTIBLE, 1, p);
}
static long do_wait(struct wait_opts *wo)
{
struct task_struct *tsk;
int retval;
trace_sched_process_wait(wo->wo_pid);
init_waitqueue_func_entry(&wo->child_wait, child_wait_callback);
wo->child_wait.private = current;
add_wait_queue(&current->signal->wait_chldexit, &wo->child_wait);
repeat:
/*
* If there is nothing that can match our critiera just get out.
* We will clear ->notask_error to zero if we see any child that
* might later match our criteria, even if we are not able to reap
* it yet.
*/
wo->notask_error = -ECHILD;
if ((wo->wo_type < PIDTYPE_MAX) &&
(!wo->wo_pid || hlist_empty(&wo->wo_pid->tasks[wo->wo_type])))
goto notask;
set_current_state(TASK_INTERRUPTIBLE);
read_lock(&tasklist_lock);
tsk = current;
do {
retval = do_wait_thread(wo, tsk);
if (retval)
goto end;
retval = ptrace_do_wait(wo, tsk);
if (retval)
goto end;
if (wo->wo_flags & __WNOTHREAD)
break;
} while_each_thread(current, tsk);
read_unlock(&tasklist_lock);
notask:
retval = wo->notask_error;
if (!retval && !(wo->wo_flags & WNOHANG)) {
retval = -ERESTARTSYS;
if (!signal_pending(current)) {
schedule();
goto repeat;
}
}
end:
__set_current_state(TASK_RUNNING);
remove_wait_queue(&current->signal->wait_chldexit, &wo->child_wait);
return retval;
}
SYSCALL_DEFINE5(waitid, int, which, pid_t, upid, struct siginfo __user *,
infop, int, options, struct rusage __user *, ru)
{
struct wait_opts wo;
struct pid *pid = NULL;
enum pid_type type;
long ret;
if (options & ~(WNOHANG|WNOWAIT|WEXITED|WSTOPPED|WCONTINUED))
return -EINVAL;
if (!(options & (WEXITED|WSTOPPED|WCONTINUED)))
return -EINVAL;
switch (which) {
case P_ALL:
type = PIDTYPE_MAX;
break;
case P_PID:
type = PIDTYPE_PID;
if (upid <= 0)
return -EINVAL;
break;
case P_PGID:
type = PIDTYPE_PGID;
if (upid <= 0)
return -EINVAL;
break;
default:
return -EINVAL;
}
if (type < PIDTYPE_MAX)
pid = find_get_pid(upid);
wo.wo_type = type;
wo.wo_pid = pid;
wo.wo_flags = options;
wo.wo_info = infop;
wo.wo_stat = NULL;
wo.wo_rusage = ru;
ret = do_wait(&wo);
if (ret > 0) {
ret = 0;
} else if (infop) {
/*
* For a WNOHANG return, clear out all the fields
* we would set so the user can easily tell the
* difference.
*/
if (!ret)
ret = put_user(0, &infop->si_signo);
if (!ret)
ret = put_user(0, &infop->si_errno);
if (!ret)
ret = put_user(0, &infop->si_code);
if (!ret)
ret = put_user(0, &infop->si_pid);
if (!ret)
ret = put_user(0, &infop->si_uid);
if (!ret)
ret = put_user(0, &infop->si_status);
}
put_pid(pid);
/* avoid REGPARM breakage on x86: */
asmlinkage_protect(5, ret, which, upid, infop, options, ru);
return ret;
}
SYSCALL_DEFINE4(wait4, pid_t, upid, int __user *, stat_addr,
int, options, struct rusage __user *, ru)
{
struct wait_opts wo;
struct pid *pid = NULL;
enum pid_type type;
long ret;
if (options & ~(WNOHANG|WUNTRACED|WCONTINUED|
__WNOTHREAD|__WCLONE|__WALL))
return -EINVAL;
if (upid == -1)
type = PIDTYPE_MAX;
else if (upid < 0) {
type = PIDTYPE_PGID;
pid = find_get_pid(-upid);
} else if (upid == 0) {
type = PIDTYPE_PGID;
pid = get_task_pid(current, PIDTYPE_PGID);
} else /* upid > 0 */ {
type = PIDTYPE_PID;
pid = find_get_pid(upid);
}
wo.wo_type = type;
wo.wo_pid = pid;
wo.wo_flags = options | WEXITED;
wo.wo_info = NULL;
wo.wo_stat = stat_addr;
wo.wo_rusage = ru;
ret = do_wait(&wo);
put_pid(pid);
/* avoid REGPARM breakage on x86: */
asmlinkage_protect(4, ret, upid, stat_addr, options, ru);
return ret;
}
#ifdef __ARCH_WANT_SYS_WAITPID
/*
* sys_waitpid() remains for compatibility. waitpid() should be
* implemented by calling sys_wait4() from libc.a.
*/
SYSCALL_DEFINE3(waitpid, pid_t, pid, int __user *, stat_addr, int, options)
{
return sys_wait4(pid, stat_addr, options, NULL);
}
#endif