Commit 4221a9918e "Add RCU check for
find_task_by_vpid()" introduced rcu_lockdep_assert to find_task_by_pid_ns.
Add rcu_read_lock/rcu_read_unlock to call find_task_by_vpid.
Tetsuo Handa wrote:
| Quoting from one of posts in that thead
| http://kerneltrap.org/mailarchive/linux-kernel/2010/2/8/4536388
|
|| Usually tasklist gives enough protection, but if copy_process() fails
|| it calls free_pid() lockless and does call_rcu(delayed_put_pid().
|| This means, without rcu lock find_pid_ns() can't scan the hash table
|| safely.
Thomas Gleixner wrote:
| We can remove the tasklist_lock while at it. rcu_read_lock is enough.
Patch also replaces thread_group_leader with has_group_leader_pid
in accordance to comment by Oleg Nesterov:
| ... thread_group_leader() check is not relaible without
| tasklist. If we race with de_thread() find_task_by_vpid() can find
| the new leader before it updates its ->group_leader.
|
| perhaps it makes sense to change posix_cpu_timer_create() to use
| has_group_leader_pid() instead, just to make this code not look racy
| and avoid adding new problems.
Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
LKML-Reference: <20101103165256.GD30053@swordfish.minsk.epam.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
* 'writable_limits' of git://decibel.fi.muni.cz/~xslaby/linux:
unistd: add __NR_prlimit64 syscall numbers
rlimits: implement prlimit64 syscall
rlimits: switch more rlimit syscalls to do_prlimit
rlimits: redo do_setrlimit to more generic do_prlimit
rlimits: add rlimit64 structure
rlimits: do security check under task_lock
rlimits: allow setrlimit to non-current tasks
rlimits: split sys_setrlimit
rlimits: selinux, do rlimits changes under task_lock
rlimits: make sure ->rlim_max never grows in sys_setrlimit
rlimits: add task_struct to update_rlimit_cpu
rlimits: security, add task_struct to setrlimit
Fix up various system call number conflicts. We not only added fanotify
system calls in the meantime, but asm-generic/unistd.h added a wait4
along with a range of reserved per-architecture system calls.
Add task_struct as a parameter to update_rlimit_cpu to be able to set
rlimit_cpu of different task than current.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
fastpath_timer_check()->thread_group_cputimer() is racy and
unneeded.
It is racy because another thread can clear ->running before
thread_group_cputimer() takes cputimer->lock. In this case
thread_group_cputimer() will set ->running = true again and call
thread_group_cputime(). But since we do not hold tasklist or
siglock, we can race with fork/exit and copy the wrong results
into cputimer->cputime.
It is unneeded because if ->running == true we can just use
the numbers in cputimer->cputime we already have.
Change fastpath_timer_check() to copy cputimer->cputime into
the local variable under cputimer->lock. We do not re-check
->running under cputimer->lock, run_posix_cpu_timers() does
this check later.
Note: we can add more optimizations on top of this change.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100611180446.GA13025@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
run_posix_cpu_timers() doesn't work if current has already passed
exit_notify(). This was needed to prevent the races with do_wait().
Since ea6d290c ->signal is always valid and can't go away. We can
remove the "tsk->exit_state == 0" in fastpath_timer_check() and
convert run_posix_cpu_timers() to use lock_task_sighand().
Note: it makes sense to take group_leader's sighand instead, the
sub-thread still uses CPU after release_task(). But we need more
changes to do this.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100610231018.GA25942@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
thread_group_cputime() looks as if it is rcu-safe, but in fact this
was wrong until ea6d290c which pins task->signal to task_struct.
It checks ->sighand != NULL under rcu, but this can't help if ->signal
can go away. Fortunately the caller either holds ->siglock, or it is
fastpath_timer_check() which uses current and checks exit_state == 0.
- Since ea6d290c commit tsk->signal is stable, we can read it first
and avoid the initialization from INIT_CPUTIME.
- Even if tsk->signal is always valid, we still have to check it
is safe to use next_thread() under rcu_read_lock(). Currently
the code checks ->sighand != NULL, change it to use pid_alive()
which is commonly used to ensure the task wasn't unhashed before
we take rcu_read_lock().
Add the comment to explain this check.
- Change the main loop to use the while_each_thread() helper.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100610230956.GA25921@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Preparation to make task->signal immutable, no functional changes.
posix-cpu-timers.c checks task->signal != NULL to ensure this task is
alive and didn't pass __exit_signal(). This is correct but we are going
to change the lifetime rules for ->signal and never reset this pointer.
Change the code to check ->sighand instead, it doesn't matter which
pointer we check under tasklist, they both are cleared simultaneously.
As Roland pointed out, some of these changes are not strictly needed and
probably it makes sense to revert them later, when ->signal will be pinned
to task_struct. But this patch tries to ensure the subsequent changes in
fork/exit can't make any visible impact on posix cpu timers.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Acked-by: Roland McGrath <roland@redhat.com>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We can optimize and simplify things taking into account signal->cputimer
is always running when we have configured any process wide cpu timer.
In check_process_timers(), we don't have to check if new updated value of
signal->cputime_expires is smaller, since we maintain new first expiration
time ({prof,virt,sched}_expires) in code flow and all other writes to
expiration cache are protected by sighand->siglock .
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
time: Fix accumulation bug triggered by long delay.
posix-cpu-timers: Reset expire cache when no timer is running
timer stats: Fix del_timer_sync() and try_to_del_timer_sync()
clockevents: Sanitize min_delta_ns adjustment and prevent overflows
Spread p->sighand->siglock locking scope to make sure that
fastpath_timer_check() never iterates over all threads. Without
locking there is small possibility that signal->cputimer will stop
running while we write values to signal->cputime_expires.
Calling thread_group_cputime() from fastpath_timer_check() is not only
bad because it is slow, also it is racy with __exit_signal() which can
lead to invalid signal->{s,u}time values.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
When user sets up a timer without associated signal and process does
not use any other cpu timers and does not exit, tsk->signal->cputimer
is enabled and running forever.
Avoid running the timer for no reason.
I used below program to check patch does not break current user space
visible behavior.
#include <sys/time.h>
#include <signal.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>
#include <assert.h>
void consume_cpu(void)
{
int i = 0;
int count = 0;
for(i=0; i<100000000; i++)
count++;
}
int main(void)
{
int i;
struct sigaction act;
struct sigevent evt = { };
timer_t tid;
struct itimerspec spec = { };
evt.sigev_notify = SIGEV_NONE;
assert(timer_create(CLOCK_PROCESS_CPUTIME_ID, &evt, &tid) == 0);
spec.it_value.tv_sec = 10;
assert(timer_settime(tid, 0, &spec, NULL) == 0);
for (i = 0; i < 30; i++) {
consume_cpu();
memset(&spec, 0, sizeof(spec));
assert(timer_gettime(tid, &spec) == 0);
printf("%lu.%09lu\n",
(unsigned long) spec.it_value.tv_sec,
(unsigned long) spec.it_value.tv_nsec);
}
assert(timer_delete(tid) == 0);
return 0;
}
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Let always set signal->cputime_expires expiration cache when setting
new itimer, POSIX 1.b timer, and RLIMIT_CPU. Since we are
initializing prof_exp expiration cache during fork(), this allows to
remove "RLIMIT_CPU != inf" check from fastpath_timer_check() and do
some other cleanups.
Checked against regression using test cases from:
http://marc.info/?l=linux-kernel&m=123749066504641&w=4http://marc.info/?l=linux-kernel&m=123811277916642&w=2
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
When a process deletes cpu timer or a timer expires we do not clear
the expiration cache sig->cputimer_expires.
As a result the fastpath_timer_check() which prevents us to loop over
all threads in case no timer is active is not working and we run the
slow path needlessly on every tick.
Zero sig->cputimer_expires in stop_process_timers().
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Spencer Candland <spencer@bluehost.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Make sure compiler won't do weird things with limits. E.g. fetching them
twice may return 2 different values after writable limits are implemented.
I.e. either use rlimit helpers added in commit 3e10e716ab ("resource:
add helpers for fetching rlimits") or ACCESS_ONCE if not applicable.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: john stultz <johnstul@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fetch rlimit (both hard and soft) values only once and work on them. It
removes many accesses through sig structure and makes the code cleaner.
Mostly a preparation for writable resource limits support.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: john stultz <johnstul@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We have already new_timer initialized to all-zeros hence in function
initializations are not needed. Document function expectation about
new_timer argument as well.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: johnstul@us.ibm.com
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Add tracepoints for all itimer variants: ITIMER_REAL, ITIMER_VIRTUAL
and ITIMER_PROF.
[ tglx: Fixed comments and made the output more readable, parseable
and consistent. Replaced pid_vnr by pid_nr because the hrtimer
callback can happen in any namespace ]
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Cc: Anton Blanchard <anton@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Zhaolei <zhaolei@cn.fujitsu.com>
LKML-Reference: <4A7F8B6E.2010109@cn.fujitsu.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
When the process exits we don't have to run new cputimer nor
use running one (as it not accounts when tsk->exit_state != 0)
to get process CPU times. As there is only one thread we can
just use CPU times fields from task and signal structs.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Roland McGrath <roland@redhat.com>
Cc: Vitaly Mayatskikh <vmayatsk@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
For powerpc with CONFIG_VIRT_CPU_ACCOUNTING
jiffies_to_cputime(1) is not compile time constant and run time
calculations are quite expensive. To optimize we use
precomputed value. For all other architectures is is
preprocessor definition.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
LKML-Reference: <1248862529-6063-5-git-send-email-sgruszka@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Don't update values in expiration cache when new ones are
equal. Add expire_le() and expire_gt() helpers to simplify the
code.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
LKML-Reference: <1248862529-6063-4-git-send-email-sgruszka@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Measure ITIMER_PROF and ITIMER_VIRT timers interval error
between real ticks and requested by user. Take it into account
when scheduling next tick.
This patch introduce possibility where time between two
consecutive tics is smaller then requested interval, it
preserve however dependency that n tick is generated not
earlier than n*interval time - counting from the beginning of
periodic signal generation.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
LKML-Reference: <1248862529-6063-3-git-send-email-sgruszka@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Both cpu itimers have same data flow in the few places, this
patch make unification of code related with VIRT and PROF
itimers.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
LKML-Reference: <1248862529-6063-2-git-send-email-sgruszka@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Sparse reports the following in kernel/posix-cpu-timers.c:
warning: symbol 'firing' shadows an earlier one
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Subrata Modak <subrata@linux.vnet.ibm.com>
LKML-Reference: <BD79186B4FD85F4B8E60E381CAEE1909016C1AFE@mi8nycmail19.Mi8.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: do not count frozen tasks toward load
sched: refresh MAINTAINERS entry
sched: Print sched_group::__cpu_power in sched_domain_debug
cpuacct: add per-cgroup utime/stime statistics
posixtimers, sched: Fix posix clock monotonicity
sched_rt: don't allocate cpumask in fastpath
cpuacct: make cpuacct hierarchy walk in cpuacct_charge() safe when rcupreempt is used -v2
update_rlimit_cpu() tries to optimize out set_process_cpu_timer() in case
when we already have CPUCLOCK_PROF timer which should expire first. But it
uses cputime_lt() instead of cputime_gt().
Test case:
int main(void)
{
struct itimerval it = {
.it_value = { .tv_sec = 1000 },
};
assert(!setitimer(ITIMER_PROF, &it, NULL));
struct rlimit rl = {
.rlim_cur = 1,
.rlim_max = 1,
};
assert(!setrlimit(RLIMIT_CPU, &rl));
for (;;)
;
return 0;
}
Without this patch, the task is not killed as RLIMIT_CPU demands.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Peter Lojkin <ia6432@inbox.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: stable@kernel.org
LKML-Reference: <20090327000610.GA10108@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: Regression fix (against clock_gettime() backwarding bug)
This patch re-introduces a couple of functions, task_sched_runtime
and thread_group_sched_runtime, which was once removed at the
time of 2.6.28-rc1.
These functions protect the sampling of thread/process clock with
rq lock. This rq lock is required not to update rq->clock during
the sampling.
i.e.
The clock_gettime() may return
((accounted runtime before update) + (delta after update))
that is less than what it should be.
v2 -> v3:
- Rename static helper function __task_delta_exec()
to do_task_delta_exec() since -tip tree already has
a __task_delta_exec() of different version.
v1 -> v2:
- Revises comments of function and patch description.
- Add note about accuracy of thread group's runtime.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: stable@kernel.org [2.6.28.x][2.6.29.x]
LKML-Reference: <49D1CC93.4080401@jp.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
See http://bugzilla.kernel.org/show_bug.cgi?id=12911
copy_signal() copies signal->rlim, but RLIMIT_CPU is "lost". Because
posix_cpu_timers_init_group() sets cputime_expires.prof_exp = 0 and thus
fastpath_timer_check() returns false unless we have other cpu timers.
This is the minimal fix for 2.6.29 (tested) and 2.6.28. The patch is not
optimal, we need further cleanups here. With this patch update_rlimit_cpu()
is not really needed, but I don't think it should be removed.
The proper fix (I think) is:
- set_process_cpu_timer() should just start the cputimer->running
logic (it does), no need to change cputime_expires.xxx_exp
- posix_cpu_timers_init_group() should set ->running when needed
- fastpath_timer_check() can check ->running instead of
task_cputime_zero(signal->cputime_expires)
Reported-by: Peter Lojkin <ia6432@inbox.ru>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Roland McGrath <roland@redhat.com>
Cc: <stable@kernel.org> [for 2.6.29.x]
LKML-Reference: <20090323193411.GA17514@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
While reviewing the manpages, I noticed I'd missed some clock vs timer sites.
Make sure that all timer functions call cpu_timer_sample_group() and not
cpu_clock_sample_group(). This ensures that we enable the process wide timer
in time, and therefore pay the O(n) thread group cost from the syscall.
Not doing it here, will result in the first jiffy tick after setting the timer
doing this, resulting in a very expensive tick (but only once) and a delay in
actually starting the timer.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The POSIX timer interface allows for absolute time expiry values through the
TIMER_ABSTIME flag, therefore we have to synchronize the timer to the clock
every time we start it.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
To decrease the chance of a missed enable, always enable the timer when we
sample it, we'll always disable it when we find that there are no active timers
in the jiffy tick.
This fixes a flood of warnings reported by Mike Galbraith.
Reported-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Change the process wide cpu timers/clocks so that we:
1) don't mess up the kernel with too many threads,
2) don't have a per-cpu allocation for each process,
3) have no impact when not used.
In order to accomplish this we're going to split it into two parts:
- clocks; which can take all the time they want since they run
from user context -- ie. sys_clock_gettime(CLOCK_PROCESS_CPUTIME_ID)
- timers; which need constant time sampling but since they're
explicity used, the user can pay the overhead.
The clock readout will go back to a full sum of the thread group, while the
timers will run of a global 'clock' that only runs when needed, so only
programs that make use of the facility pay the price.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Either we bounce once cacheline per cpu per tick, yielding n^2 bounces
or we just bounce a single..
Also, using per-cpu allocations for the thread-groups complicates the
per-cpu allocator in that its currently aimed to be a fixed sized
allocator and the only possible extention to that would be vmap based,
which is seriously constrained on 32 bit archs.
So making the per-cpu memory requirement depend on the number of
processes is an issue.
Lastly, it didn't deal with cpu-hotplug, although admittedly that might
be fixable.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Since CLOCK_PROCESS_CPUTIME_ID is in fact translated to -6, the switch
statement in cpu_clock_sample_group() must first mask off the irrelevant
bits, similar to cpu_clock_sample().
Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
--
posix-cpu-timers.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Impact: simplify the code
thread_group_cputime() is called by current when it must have the valid
->signal, or under ->siglock, or under tasklist_lock after the ->signal
check, or the caller is wait_task_zombie() which reaps the child. In any
case ->signal can't be NULL.
But the point of this patch is not optimization. If it is possible to call
thread_group_cputime() when ->signal == NULL we are doing something wrong,
and we should not mask the problem. thread_group_cputime() fills *times
and the caller will use it, if we silently use task_struct->*times* we
report the wrong values.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: fix potential NULL dereference
Contrary to ad474caca3 changelog, other
acct_group_xxx() helpers can be called after exit_notify() by timer tick.
Thanks to Roland for pointing out this. Somehow I missed this simple fact
when I read the original patch, and I am afraid I confused Frank during
the discussion. Sorry.
Fortunately, these helpers work with current, we can check ->exit_state
to ensure that ->signal can't go away under us.
Also, add the comment and compiler barrier to account_group_exec_runtime(),
to make sure we load ->signal only once.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This is the second resubmission of the posix timer rework patch, posted
a few days ago.
This includes the changes from the previous resubmittion, which addressed
Oleg Nesterov's comments, removing the RCU stuff from the patch and
un-inlining the thread_group_cputime() function for SMP.
In addition, per Ingo Molnar it simplifies the UP code, consolidating much
of it with the SMP version and depending on lower-level SMP/UP handling to
take care of the differences.
It also cleans up some UP compile errors, moves the scheduler stats-related
macros into kernel/sched_stats.h, cleans up a merge error in
kernel/fork.c and has a few other minor fixes and cleanups as suggested
by Oleg and Ingo. Thanks for the review, guys.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Overview
This patch reworks the handling of POSIX CPU timers, including the
ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together
with the help of Roland McGrath, the owner and original writer of this code.
The problem we ran into, and the reason for this rework, has to do with using
a profiling timer in a process with a large number of threads. It appears
that the performance of the old implementation of run_posix_cpu_timers() was
at least O(n*3) (where "n" is the number of threads in a process) or worse.
Everything is fine with an increasing number of threads until the time taken
for that routine to run becomes the same as or greater than the tick time, at
which point things degrade rather quickly.
This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF."
Code Changes
This rework corrects the implementation of run_posix_cpu_timers() to make it
run in constant time for a particular machine. (Performance may vary between
one machine and another depending upon whether the kernel is built as single-
or multiprocessor and, in the latter case, depending upon the number of
running processors.) To do this, at each tick we now update fields in
signal_struct as well as task_struct. The run_posix_cpu_timers() function
uses those fields to make its decisions.
We define a new structure, "task_cputime," to contain user, system and
scheduler times and use these in appropriate places:
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
This is included in the structure "thread_group_cputime," which is a new
substructure of signal_struct and which varies for uniprocessor versus
multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as
a simple substructure, while for multiprocessor kernels it is a pointer:
struct thread_group_cputime {
struct task_cputime totals;
};
struct thread_group_cputime {
struct task_cputime *totals;
};
We also add a new task_cputime substructure directly to signal_struct, to
cache the earliest expiration of process-wide timers, and task_cputime also
replaces the it_*_expires fields of task_struct (used for earliest expiration
of thread timers). The "thread_group_cputime" structure contains process-wide
timers that are updated via account_user_time() and friends. In the non-SMP
case the structure is a simple aggregator; unfortunately in the SMP case that
simplicity was not achievable due to cache-line contention between CPUs (in
one measured case performance was actually _worse_ on a 16-cpu system than
the same test on a 4-cpu system, due to this contention). For SMP, the
thread_group_cputime counters are maintained as a per-cpu structure allocated
using alloc_percpu(). The timer functions update only the timer field in
the structure corresponding to the running CPU, obtained using per_cpu_ptr().
We define a set of inline functions in sched.h that we use to maintain the
thread_group_cputime structure and hide the differences between UP and SMP
implementations from the rest of the kernel. The thread_group_cputime_init()
function initializes the thread_group_cputime structure for the given task.
The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the
out-of-line function thread_group_cputime_alloc_smp() to allocate and fill
in the per-cpu structures and fields. The thread_group_cputime_free()
function, also a no-op for UP, in SMP frees the per-cpu structures. The
thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls
thread_group_cputime_alloc() if the per-cpu structures haven't yet been
allocated. The thread_group_cputime() function fills the task_cputime
structure it is passed with the contents of the thread_group_cputime fields;
in UP it's that simple but in SMP it must also safely check that tsk->signal
is non-NULL (if it is it just uses the appropriate fields of task_struct) and,
if so, sums the per-cpu values for each online CPU. Finally, the three
functions account_group_user_time(), account_group_system_time() and
account_group_exec_runtime() are used by timer functions to update the
respective fields of the thread_group_cputime structure.
Non-SMP operation is trivial and will not be mentioned further.
The per-cpu structure is always allocated when a task creates its first new
thread, via a call to thread_group_cputime_clone_thread() from copy_signal().
It is freed at process exit via a call to thread_group_cputime_free() from
cleanup_signal().
All functions that formerly summed utime/stime/sum_sched_runtime values from
from all threads in the thread group now use thread_group_cputime() to
snapshot the values in the thread_group_cputime structure or the values in
the task structure itself if the per-cpu structure hasn't been allocated.
Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit.
The run_posix_cpu_timers() function has been split into a fast path and a
slow path; the former safely checks whether there are any expired thread
timers and, if not, just returns, while the slow path does the heavy lifting.
With the dedicated thread group fields, timers are no longer "rebalanced" and
the process_timer_rebalance() function and related code has gone away. All
summing loops are gone and all code that used them now uses the
thread_group_cputime() inline. When process-wide timers are set, the new
task_cputime structure in signal_struct is used to cache the earliest
expiration; this is checked in the fast path.
Performance
The fix appears not to add significant overhead to existing operations. It
generally performs the same as the current code except in two cases, one in
which it performs slightly worse (Case 5 below) and one in which it performs
very significantly better (Case 2 below). Overall it's a wash except in those
two cases.
I've since done somewhat more involved testing on a dual-core Opteron system.
Case 1: With no itimer running, for a test with 100,000 threads, the fixed
kernel took 1428.5 seconds, 513 seconds more than the unfixed system,
all of which was spent in the system. There were twice as many
voluntary context switches with the fix as without it.
Case 2: With an itimer running at .01 second ticks and 4000 threads (the most
an unmodified kernel can handle), the fixed kernel ran the test in
eight percent of the time (5.8 seconds as opposed to 70 seconds) and
had better tick accuracy (.012 seconds per tick as opposed to .023
seconds per tick).
Case 3: A 4000-thread test with an initial timer tick of .01 second and an
interval of 10,000 seconds (i.e. a timer that ticks only once) had
very nearly the same performance in both cases: 6.3 seconds elapsed
for the fixed kernel versus 5.5 seconds for the unfixed kernel.
With fewer threads (eight in these tests), the Case 1 test ran in essentially
the same time on both the modified and unmodified kernels (5.2 seconds versus
5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds
versus 5.4 seconds but again with much better tick accuracy, .013 seconds per
tick versus .025 seconds per tick for the unmodified kernel.
Since the fix affected the rlimit code, I also tested soft and hard CPU limits.
Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer
running), the modified kernel was very slightly favored in that while
it killed the process in 19.997 seconds of CPU time (5.002 seconds of
wall time), only .003 seconds of that was system time, the rest was
user time. The unmodified kernel killed the process in 20.001 seconds
of CPU (5.014 seconds of wall time) of which .016 seconds was system
time. Really, though, the results were too close to call. The results
were essentially the same with no itimer running.
Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds
(where the hard limit would never be reached) and an itimer running,
the modified kernel exhibited worse tick accuracy than the unmodified
kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise,
performance was almost indistinguishable. With no itimer running this
test exhibited virtually identical behavior and times in both cases.
In times past I did some limited performance testing. those results are below.
On a four-cpu Opteron system without this fix, a sixteen-thread test executed
in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On
the same system with the fix, user and elapsed time were about the same, but
system time dropped to 0.007 seconds. Performance with eight, four and one
thread were comparable. Interestingly, the timer ticks with the fix seemed
more accurate: The sixteen-thread test with the fix received 149543 ticks
for 0.024 seconds per tick, while the same test without the fix received 58720
for 0.061 seconds per tick. Both cases were configured for an interval of
0.01 seconds. Again, the other tests were comparable. Each thread in this
test computed the primes up to 25,000,000.
I also did a test with a large number of threads, 100,000 threads, which is
impossible without the fix. In this case each thread computed the primes only
up to 10,000 (to make the runtime manageable). System time dominated, at
1546.968 seconds out of a total 2176.906 seconds (giving a user time of
629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite
accurate. There is obviously no comparable test without the fix.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It's useful to detect which process is killed by RT watchdog.
Signed-off-by: Hiroshi Shimamoto <h-shimamoto@ct.jp.nec.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
x86 is the only arch right now, which provides an optimized for
div_long_long_rem and it has the downside that one has to be very careful that
the divide doesn't overflow.
The API is a little akward, as the arguments for the unsigned divide are
signed. The signed version also doesn't handle a negative divisor and
produces worse code on 64bit archs.
There is little incentive to keep this API alive, so this converts the few
users to the new API.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix sparse warnings like this:
kernel/posix-cpu-timers.c:1090:25: warning: symbol 't' shadows an earlier one
kernel/posix-cpu-timers.c:1058:21: originally declared here
Signed-off-by: WANG Cong <xiyou.wangcong@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
All the functions that need to lookup a task by pid in posix timers obtain
this pid from a user space, and thus this value refers to a task in the same
namespace, as the current task lives in.
So the proper behavior is to call find_task_by_vpid() here.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove the curious logic to set it_sched_expires in the future. It useless
because rt.timeout wouldn't be incremented anyway.
Explicity check for RLIM_INFINITY as a test programm that had a 1s soft limit
and a inf hard limit would SIGKILL at 1s. This is because RLIM_INFINITY+d-1
is d-2.
Signed-off-by: Peter Zijlsta <a.p.zijlstra@chello.nl>
CC: Michal Schmidt <mschmidt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Introduce a new rlimit that allows the user to set a runtime timeout on
real-time tasks their slice. Once this limit is exceeded the task will receive
SIGXCPU.
So it measures runtime since the last sleep.
Input and ideas by Thomas Gleixner and Lennart Poettering.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
CC: Lennart Poettering <mzxreary@0pointer.de>
CC: Michael Kerrisk <mtk.manpages@googlemail.com>
CC: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Sergey Senozhatsky
Linus Torvalds
Jiri Slaby
Oleg Nesterov
Stanislaw Gruszka
Thomas Gleixner
Jiri Slaby
Xiao Guangrong
H Hartley Sweeten
Ingo Molnar
Hidetoshi Seto
Peter Zijlstra
Peter Zijlstra
Petr Tesarik
Frank Mayhar
Hiroshi Shimamoto
Roman Zippel
WANG Cong
Pavel Emelyanov