linux-kernel - Re: Reading EeePC900 battery info causes stalls

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]

Date:	Mon, 22 Sep 2008 07:24:51 +0100
From:	Sitsofe Wheeler <sitsofe@...oo.com>
To:	Steven Rostedt <rostedt@...dmis.org>
CC:	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Ingo Molnar <mingo@...e.hu>,
	Arjan van de Ven <arjan@...radead.org>,
	LKML <linux-kernel@...r.kernel.org>, lenb@...nel.org,
	astarikovskiy@...e.de
Subject: Re: Reading EeePC900 battery info causes stalls

Steven Rostedt wrote:
> On Sat, 20 Sep 2008, Sitsofe Wheeler wrote:
> 
>> Steven Rostedt wrote:
>>> OK, that is probably the known bug you are hitting. Simply disable the
>>> CONFIG_FTRACE_STARTUP_TEST and you should have the wakeup tracer. The bug is
>>> with the test, not the tracer, so it should not hurt you.
>> Thanks - this made the wakeup tracer appear a you said. I have put two wakeup
>> traces up:
>>
>> http://sucs.org/~sits/test/eeepc-debug/20080920/latency_trace.txt.gz
>> http://sucs.org/~sits/test/eeepc-debug/20080920/trace.txt.gz
>> (each file is around 6Mbytes uncompressed)
>>
>> Here's a small extract of latency_trace.txt:
>>
>>> # tracer: wakeup
>>> #
>>> wakeup latency trace v1.1.5 on 2.6.27-rc6skw-dirty
>>> --------------------------------------------------------------------
>>>  latency: 3232905 us, #65620/6180619, CPU#0 | (M:desktop VP:0, KP:0, SP:0
> 
> Peter, these times are crazy, mainly due to the cpu_clock. He probably 
> wants to use the sched_clock. Below is a patch to use it instead.
> 
> Sitsofe, I notice that the trace states "desktop". This means that you
> are running with CONFIG_PREEMPT_VOLUNTARY. You want 
> CONFIG_PREEMPT.
> 
> [...]
>> Is it intentional that the last event has a time earlier closer to that of the
>> first event?
>>
> 
> Change the config, and see what you get with this patch:
> 
> Note this is not compiled tested:

OK I've made the changes you suggested. Without preempt enabled the last 
event will have a stamp closer to the first event and the times are very 
high. With preempt enabled that beahviour has gone. Here are results 
with preempt enabled:

latency: 19657 us, #3268/3268, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0):
http://sucs.org/~sits/test/eeepc-debug/20080922/preempt/ath5k/latency_trace.txt 


latency: 104 us, #182/182, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0)
http://sucs.org/~sits/test/eeepc-debug/20080922/preempt/unplug/latency_trace.txt

latency: 95 us, #134/134, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0):
http://sucs.org/~sits/test/eeepc-debug/20080922/preempt/acpi/latency_trace.txt

latency: 91 us, #152/152, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0)
http://sucs.org/~sits/test/eeepc-debug/20080922/preempt/touchpad/latency_trace.txt

Are these the type of latencies to be expected with preemption on?

-- 
Sitsofe | http://sucs.org/~sits/
--
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/