lkml.org 
[lkml]   [2008]   [Sep]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [ath9k-devel] ath9k: massive unexplained latency in 2.6.27 (rc5, rc6, probably others)
On Sat, Sep 20, 2008 at 8:21 AM, Steven Noonan <steven@uplinklabs.net> wrote:
> On Fri, Sep 19, 2008 at 8:29 PM, Steven Noonan <steven@uplinklabs.net> wrote:
>> And here's the change as I unload and reload ath9k:
>>
>> [...]
>> 20:21:38 sum 20138.83
>> 20:21:40 sum 21576.09
>> 20:21:41 sum 17678.38
>> 20:21:42 sum 19834.34
>> 20:21:43 sum 19845.10
>> 20:21:44 sum 19470.00
>> 20:21:45 sum 9273.00 <--- ath9k unloaded/reloaded
>> 20:21:46 sum 187.00
>> 20:21:47 sum 1066.00
>> 20:21:48 sum 610.00
>> 20:21:49 sum 106.93
>> 20:21:50 sum 670.00
>> 20:21:51 sum 742.00
>> 20:21:52 sum 781.00
>> Average: sum 19562.56
>>
>
> I think it might help if you can see the .sar file yourself. I've
> uploaded a gzipped copy (great compression ratio in this case for
> merely using gzip!).
>
> http://www.uplinklabs.net/~tycho/linux/irqlog-storm.sar.gz

mcgrof@tesla ~/tmp/irq $ sar -f irqlog-storm.sar -A
Invalid system activity file: irqlog-storm.sar

Perhaps we its a different format than what my sar can read...

So here's what we need to do -- try to zero on in what causes this IRQ
storm. Most likely its some strange event that triggers it, perhaps
some sort of noise which makes the device try to adapt and hits a path
we are not handling correctly yet. Can you provide the initial log of
when ath9k gets loaded? This tells us what card it is, like AR5416 or
AR9280, etc.

Luis


\
 
 \ /
  Last update: 2008-09-20 23:01    [W:0.109 / U:0.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site