lkml.org 
[lkml]   [2006]   [Jul]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch] Reorganize the cpufreq cpu hotplug locking to not be totally bizare
From
Date
On Wed, 2006-07-26 at 23:35 +0100, Sanjoy Mahajan wrote:
> From: Andrew Morton <akpm@osdl.org>
> > We should delete lock_cpu_hotplug() and start again.
>
> Here is another example of possible lock_cpu_hotplug() problems. Is
> it worth tracking down, or should I just ignore the messages until a
> proper solution is figured out? The only problem is that it means S3
> suspend doesn't work.
>
> The hardware is a Thinkpad T60, T2400 dual-core, compiled with SMP and
> PREEMPT, hotpluggable CPUs, and it has a SATA drive. Kernel is
> 2.6.18-rc1. Suspend (UP) worked with 2.6.15-25-386 from Ubuntu using
> the same sleep.sh script. The messages below, including the large
> lockdep backtrace, occur after running sleep.sh (run by Fn-F4):
>
>
> [ 546.652000] Stopping tasks: ====================================================================================
> [ 566.848000] stopping tasks timed out after 20 seconds (8 tasks remaining):
> [ 566.848000] rt-test-0
> [ 566.848000] rt-test-1
> [ 566.848000] rt-test-2
> [ 566.848000] rt-test-3
> [ 566.848000] rt-test-4
> [ 566.848000] rt-test-5
> [ 566.848000] rt-test-6
> [ 566.848000] rt-test-7
>
this got fixed in -rc2

> The lockdep code also reported problems:
>
> [ 538.292000] ACPI: PCI interrupt for device 0000:02:00.0 disabled
> [ 546.144000] Freezing cpus ...
> [ 546.172000]
> [ 546.172000] =======================================================
> [ 546.172000] [ INFO: possible circular locking dependency detected ]
> [ 546.172000] -------------------------------------------------------
> [ 546.172000] sleep.sh/15184 is trying to acquire lock:
> [ 546.172000] (&policy->lock){--..}, at: [<c0310645>] mutex_lock+0x25/0x30
> [ 546.172000]
> [ 546.172000] but task is already holding lock:
> [ 546.172000] ((cpu_chain).rwsem){----}, at: [<c0133267>] blocking_notifier_call_chain+0x17/0x40
> [ 546.172000]

and afaik my patches in current git should fix all this up
if not please send a trace ;)

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2006-07-27 00:51    [W:0.036 / U:0.940 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site