On Thu 27-08-15 00:03:13, Tetsuo Handa wrote: > Reposting updated version as it turned out that we can call do_send_sig_info() > with task_lock held. ;-) ( http://marc.info/?l=linux-mm&m=144059948628905&w=2 ) > ---------------------------------------- > >From 69489b37171d9a84ccbb00e7441f8f73bc526e11 Mon Sep 17 00:00:00 2001 > From: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx> > Date: Wed, 26 Aug 2015 23:54:12 +0900 > Subject: [PATCH] mm,oom: Reverse the order of setting TIF_MEMDIE and sending > SIGKILL. > > It is observed that a multi-threaded program can deplete memory reserves > using time lag between the OOM killer sets TIF_MEMDIE and the OOM killer > sends SIGKILL. This is because a thread group leader who gets TIF_MEMDIE > received SIGKILL too lately when there is a lot of child threads sharing > the same memory due to doing a lot of printk() inside for_each_process() > loop which can take many seconds. > > Before starting oom-depleter process: > > Node 0 DMA: 3*4kB (UM) 6*8kB (U) 4*16kB (UEM) 0*32kB 0*64kB 1*128kB (M) 2*256kB (EM) 2*512kB (UE) 2*1024kB (EM) 1*2048kB (E) 1*4096kB (M) = 9980kB > Node 0 DMA32: 31*4kB (UEM) 27*8kB (UE) 32*16kB (UE) 13*32kB (UE) 14*64kB (UM) 7*128kB (UM) 8*256kB (UM) 8*512kB (UM) 3*1024kB (U) 4*2048kB (UM) 362*4096kB (UM) = 1503220kB > > As of invoking the OOM killer: > > Node 0 DMA: 11*4kB (UE) 8*8kB (UEM) 6*16kB (UE) 2*32kB (EM) 0*64kB 1*128kB (U) 3*256kB (UEM) 2*512kB (UE) 3*1024kB (UEM) 1*2048kB (U) 0*4096kB = 7308kB > Node 0 DMA32: 1049*4kB (UEM) 507*8kB (UE) 151*16kB (UE) 53*32kB (UEM) 83*64kB (UEM) 52*128kB (EM) 25*256kB (UEM) 11*512kB (M) 6*1024kB (UM) 1*2048kB (M) 0*4096kB = 44556kB > > Between the thread group leader got TIF_MEMDIE and receives SIGKILL: > > Node 0 DMA: 0*4kB 0*8kB 0*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 0kB > Node 0 DMA32: 0*4kB 0*8kB 0*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 0kB > > The oom-depleter's thread group leader which got TIF_MEMDIE started > memset() in user space after the OOM killer set TIF_MEMDIE, and it > was free to abuse ALLOC_NO_WATERMARKS by TIF_MEMDIE for memset() > in user space until SIGKILL is delivered. If SIGKILL is delivered > before TIF_MEMDIE is set, the oom-depleter can terminate without > touching memory reserves. The comment could have been a bit better because it doesn't exmplain why but this is much much better. > Signed-off-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx> Acked-by: Michal Hocko <mhocko@xxxxxxxx> I guess you want to add Andrew to the CC because he might miss it this way easily. I would also consider marking this for stable. Thanks! > --- > mm/oom_kill.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/mm/oom_kill.c b/mm/oom_kill.c > index 5249e7e..408aa8e 100644 > --- a/mm/oom_kill.c > +++ b/mm/oom_kill.c > @@ -555,6 +555,8 @@ void oom_kill_process(struct oom_control *oc, struct task_struct *p, > /* Get a reference to safely compare mm after task_unlock(victim) */ > mm = victim->mm; > atomic_inc(&mm->mm_users); > + /* Send SIGKILL before setting TIF_MEMDIE. */ > + do_send_sig_info(SIGKILL, SEND_SIG_FORCED, victim, true); > mark_oom_victim(victim); > pr_err("Killed process %d (%s) total-vm:%lukB, anon-rss:%lukB, file-rss:%lukB\n", > task_pid_nr(victim), victim->comm, K(victim->mm->total_vm), > @@ -586,7 +588,6 @@ void oom_kill_process(struct oom_control *oc, struct task_struct *p, > } > rcu_read_unlock(); > > - do_send_sig_info(SIGKILL, SEND_SIG_FORCED, victim, true); > mmput(mm); > put_task_struct(victim); > } > -- > 1.8.3.1 -- Michal Hocko SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>