2008-07-23 23:22:34

by Mikhail Kshevetskiy

[permalink] [raw]
Subject: BUG: 2.6.26-git8: spinlock lockup in c1e_idle()

BUG: spinlock lockup on CPU#0, swapper/0, ffffffff805d7e60
Pid: 0, comm: swapper Not tainted 2.6.26-git8 #6

Call Trace:
[<ffffffff803471f1>] _raw_spin_lock+0xdc/0x107
[<ffffffff80480170>] _spin_lock+0x55/0x62
[<ffffffff8025a303>] clockevents_notify+0x19/0x7a
[<ffffffff80213d40>] c1e_idle+0xf8/0x113
[<ffffffff80256770>] ? atomic_notifier_call_chain+0xf/0x11
[<ffffffff8020b2a7>] cpu_idle+0x88/0xee
[<ffffffff8046c392>] rest_init+0x86/0x8a
[<ffffffff80624e01>] start_kernel+0x399/0x3a4
[<ffffffff8062429f>] x86_64_start_reservations+0xa5/0xa9
[<ffffffff806243af>] x86_64_start_kernel+0xe9/0xf8

Full dmesg output and my kernel configuration are attached.

Mikhail Kshevetskiy


Attachments:
(No filename) (686.00 B)
config-2.6.26-git8 (74.72 kB)
dmesg (35.63 kB)
Download all attachments

2008-07-28 14:40:55

by Thomas Gleixner

[permalink] [raw]
Subject: Re: BUG: 2.6.26-git8: spinlock lockup in c1e_idle()

On Thu, 24 Jul 2008, Mikhail Kshevetskiy wrote:

> BUG: spinlock lockup on CPU#0, swapper/0, ffffffff805d7e60
> Pid: 0, comm: swapper Not tainted 2.6.26-git8 #6
>
> Call Trace:
> [<ffffffff803471f1>] _raw_spin_lock+0xdc/0x107
> [<ffffffff80480170>] _spin_lock+0x55/0x62
> [<ffffffff8025a303>] clockevents_notify+0x19/0x7a
> [<ffffffff80213d40>] c1e_idle+0xf8/0x113
> [<ffffffff80256770>] ? atomic_notifier_call_chain+0xf/0x11
> [<ffffffff8020b2a7>] cpu_idle+0x88/0xee
> [<ffffffff8046c392>] rest_init+0x86/0x8a
> [<ffffffff80624e01>] start_kernel+0x399/0x3a4
> [<ffffffff8062429f>] x86_64_start_reservations+0xa5/0xa9
> [<ffffffff806243af>] x86_64_start_kernel+0xe9/0xf8

Mikhail,

we merged a patch which might be fixing that problem after
2.6.26-git8. Can you give Linus latest a try please ?

Thanks,

tglx

2008-07-28 23:15:36

by Mikhail Kshevetskiy

[permalink] [raw]
Subject: Re: BUG: 2.6.26-git8: spinlock lockup in c1e_idle()

On Mon, 28 Jul 2008 16:39:45 +0200 (CEST)
Thomas Gleixner <[email protected]> wrote:

> On Thu, 24 Jul 2008, Mikhail Kshevetskiy wrote:
>
> > BUG: spinlock lockup on CPU#0, swapper/0, ffffffff805d7e60
> > Pid: 0, comm: swapper Not tainted 2.6.26-git8 #6
> >
> > Call Trace:
> > [<ffffffff803471f1>] _raw_spin_lock+0xdc/0x107
> > [<ffffffff80480170>] _spin_lock+0x55/0x62
> > [<ffffffff8025a303>] clockevents_notify+0x19/0x7a
> > [<ffffffff80213d40>] c1e_idle+0xf8/0x113
> > [<ffffffff80256770>] ? atomic_notifier_call_chain+0xf/0x11
> > [<ffffffff8020b2a7>] cpu_idle+0x88/0xee
> > [<ffffffff8046c392>] rest_init+0x86/0x8a
> > [<ffffffff80624e01>] start_kernel+0x399/0x3a4
> > [<ffffffff8062429f>] x86_64_start_reservations+0xa5/0xa9
> > [<ffffffff806243af>] x86_64_start_kernel+0xe9/0xf8
>
> Mikhail,
>
> we merged a patch which might be fixing that problem after
> 2.6.26-git8. Can you give Linus latest a try please ?
>
> Thanks,
>
> tglx

This does not help. I try 2.6.26-git17 and the problem still exist.
Please do not trust to kernel timestamp in dmesg. In reality it takes
2-3 times more time before spinlock lockup will be detected.

Here is the summary of kernel parameters I tried:
1) noapictimer -- boot ok, tickless disabled
2) nohpet -- boot ok, (i suppose tickless disabled)
3) clocksource=tsc -- the same as with "nohpet"
4) idle=halt -- need to press buttons to boot
5) none of above -- the kernel freeze after switch to high resolution
mode on CPU0 and detect spinlock lockup 5 minutes
later


Attachments:
(No filename) (1.56 kB)
dmesg-2.6.26-git17 (41.42 kB)
config-2.6.26-git17 (72.45 kB)
Download all attachments