lkml.org 
[lkml]   [2008]   [Oct]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: RCU hang on cpu re-hotplug with 2.6.27rc8
On Tue, Oct 07, 2008 at 11:09:47PM +0200, Andi Kleen wrote:
> On Tue, Oct 07, 2008 at 09:34:01AM -0700, Paul E. McKenney wrote:
> > Thank you! Hmmm, classic RCU, worked just fine in 2.6.27-rc7 with
> > Thomas's patch. I was doing random onlines and offlines in a loop,
> > with about 3 seconds between each operation continuously for more than
> > ten hours, both x86 and Power. So could you please try 2.6.27-rc7 with
> > Thomas's patch as follows?
> >
> > http://www.rdrop.com/users/paulmck/patches/2.6.27-rc7-tglx-timer-1.patch
>
> Same effect. Hung on the first try
>
> bash D 00000000ffff25c1 0 4755 4742
> ffff88027b127bf8 0000000000000086 ffff88027b127c18 0000000000000296
> ffff88027c80b330 ffff8804be488b90 ffff88027c80b578 0000000300000296
> ffff88027b127c18 ffffffff808cbd18 ffff88002805d600 ffff88027d182098
> Call Trace:
> [<ffffffff805c318d>] schedule_timeout+0x22/0xb4
> [<ffffffff8020a029>] ? __switch_to+0x320/0x330
> [<ffffffff8025fa65>] ? cpupri_set+0xc5/0xd8
> [<ffffffff805c2fe7>] wait_for_common+0xcd/0x131
> [<ffffffff8022d297>] ? default_wake_function+0x0/0xf
> [<ffffffff805c30d5>] wait_for_completion+0x18/0x1a
> [<ffffffff8024374b>] synchronize_rcu+0x35/0x3c
> [<ffffffff802437ca>] ? wakeme_after_rcu+0x0/0x12
> [<ffffffff8022e435>] partition_sched_domains+0x9b/0x1dd
> [<ffffffff8022d2c3>] ? wake_up_process+0x10/0x12
> [<ffffffff8022e5a5>] update_sched_domains+0x2e/0x35
> [<ffffffff805c6bb2>] notifier_call_chain+0x33/0x5b
> [<ffffffff80248a29>] __raw_notifier_call_chain+0x9/0xb
> [<ffffffff80248a3a>] raw_notifier_call_chain+0xf/0x11
> [<ffffffff805c06e6>] _cpu_up+0xd3/0x10c
> [<ffffffff805c0776>] cpu_up+0x57/0x67
> [<ffffffff805a4ab7>] store_online+0x4d/0x75
> [<ffffffff803e5d0b>] sysdev_store+0x1b/0x1d
> [<ffffffff802cc49c>] sysfs_write_file+0xe0/0x11c
> [<ffffffff8028ad1d>] vfs_write+0xae/0x137
> [<ffffffff8028b1c6>] sys_write+0x47/0x6f
> [<ffffffff8020b36b>] system_call_fastpath+0x16/0x1b

Thus far, as usual, I cannot reproduce, either on x86 or Power. You are
running on hyperthreaded machines? If so, what happens if you disable
CONFIG_SCHED_SMT and CONFIG_SCHED_MC?

You are running on a 16-CPU x86-64 box?

Thanx, Paul


\
 
 \ /
  Last update: 2008-10-07 23:27    [W:0.329 / U:0.276 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site