Messages in this thread | | | Date | Fri, 11 Jun 2010 20:06:06 +0200 | From | Oleg Nesterov <> | Subject | Re: [PATCH 6/5] fix the racy usage of thread_group_cputimer() in fastpath_timer_check() |
| |
On 06/11, Oleg Nesterov wrote: > > fastpath_timer_check()->thread_group_cputimer() is racy and > unneeded.
Just in case... this fix doesn't depend on other patches I sent.
> 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> > --- > > kernel/posix-cpu-timers.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > --- 35-rc2/kernel/posix-cpu-timers.c~6_FPTC_DONT_SET_RUNNING 2010-06-11 01:08:03.000000000 +0200 > +++ 35-rc2/kernel/posix-cpu-timers.c 2010-06-11 19:40:22.000000000 +0200 > @@ -1287,7 +1287,10 @@ static inline int fastpath_timer_check(s > if (sig->cputimer.running) { > struct task_cputime group_sample; > > - thread_group_cputimer(tsk, &group_sample); > + spin_lock(&sig->cputimer.lock); > + group_sample = sig->cputimer.cputime; > + spin_unlock(&sig->cputimer.lock); > + > if (task_cputime_expired(&group_sample, &sig->cputime_expires)) > return 1; > }
| |