linux-kernel - Re: 3.15-rc2: longhaul cpufreq stalls tasks for 120s+

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <535A198F.3040009@linux.vnet.ibm.com>
Date:	Fri, 25 Apr 2014 13:45:11 +0530
From:	"Srivatsa S. Bhat" <srivatsa.bhat@...ux.vnet.ibm.com>
To:	Viresh Kumar <viresh.kumar@...aro.org>
CC:	Meelis Roos <mroos@...ux.ee>,
	"Rafael J. Wysocki" <rjw@...ysocki.net>,
	"cpufreq@...r.kernel.org" <cpufreq@...r.kernel.org>,
	"linux-pm@...r.kernel.org" <linux-pm@...r.kernel.org>,
	Linux Kernel list <linux-kernel@...r.kernel.org>
Subject: Re: 3.15-rc2: longhaul cpufreq stalls tasks for 120s+

On 04/25/2014 10:11 AM, Viresh Kumar wrote:
> On 25 April 2014 00:33, Meelis Roos <mroos@...ux.ee> wrote:
> 
>> [  240.140176] INFO: task kworker/0:1:116 blocked for more than 120 seconds.
>> [  240.140353]       Not tainted 3.15.0-rc2-dirty #37
>> [  240.140485] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>> [  240.140687] kworker/0:1     D cf6afd50     0   116      2 0x00000000
>> [  240.140938] Workqueue: events od_dbs_timer
>> [  240.141103]  cf6afd98 00000082 00000002 cf6afd50 c1040d91 cf6affec cf6ad310 cf6ad310
>> [  240.142479]  c1286dcb 00000002 cf6afd70 c1040f14 00000000 ce460b30 00000282 00000046
>> [  240.143011]  00000282 ce460b30 cf6afd78 c1040f39 cf6afd88 00000282 cf6afdb0 ce460b30
>> [  240.143544] Call Trace:
>> [  240.143706]  [<c1040d91>] ? mark_held_locks+0x4b/0x61
>> [  240.143883]  [<c1286dcb>] ? _raw_spin_unlock_irqrestore+0x33/0x3f
>> [  240.144043]  [<c1040f14>] ? trace_hardirqs_on_caller+0x16d/0x187
>> [  240.144203]  [<c1040f39>] ? trace_hardirqs_on+0xb/0xd
>> [  240.144358]  [<c1284145>] schedule+0x5d/0x5f
>> [  240.144527]  [<c11eaa67>] cpufreq_freq_transition_begin+0x4a/0x9d
>> [  240.144687]  [<c103e1b0>] ? __wake_up_sync+0x14/0x14
>> [  240.144860]  [<d06722d5>] longhaul_setstate+0x88/0x2f1 [longhaul]
>> [  240.145023]  [<c1036d43>] ? srcu_notifier_call_chain+0x1a/0x1c
>> [  240.145186]  [<c11eaab2>] ? cpufreq_freq_transition_begin+0x95/0x9d
>> [  240.145350]  [<d06725ba>] longhaul_target+0x7c/0x8b [longhaul]
>> [  240.145511]  [<c11eae78>] __cpufreq_driver_target+0xfe/0x148
> 
> Am I reading it correctly? It looks like we are starting another transition
> from notifier chain, but I couldn't figure out how from code.
> 

Indeed, its a case of double invocation of the _begin() and _end()
notifiers. I developed a patchset to fix this in longhaul, powernow-k6 and k7
drivers, before seeing your patchset that does the same thing. However, looking
closer, I don't completely agree with the approach you used to fix the issue,
so I'll post my patches as well (which have a different design).

Regards,
Srivatsa S. Bhat

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