DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :cc:content-type:content-transfer-encoding;
        b=ui5mFLCsQQ8oGRvB/D+LO4b7qIUz6fL8FaUJcHCpmDbiA65nbPUeckJZPM1wt9ZGVr
         59ug7E6/JM7AmfQXsAFKqUbcGAvlX8mdi/TDrgN2tvWm70ad8LXrBTFuwC4CFuaxibE9
         vvd3vKGfx/zKsv9XTR2rjHzUzfMACO+KlJRVc=
MIME-Version: 1.0
In-Reply-To: <43e72e890912181155h444956a3tf228ddd8d281a9d3@mail.gmail.com>
References: <43e72e890912180926oad3b09fl6b7951864a836700@mail.gmail.com> 
	<200912181838.20204.bzolnier@gmail.com> <43e72e890912181119t53e01ec2y5dc8687d23f7a668@mail.gmail.com> 
	<43e72e890912181155h444956a3tf228ddd8d281a9d3@mail.gmail.com>
From: "Luis R. Rodriguez" <mcgrof@gmail.com>
Date: Fri, 18 Dec 2009 12:51:12 -0800
Message-ID: <43e72e890912181251q7ef1fb59n42da94c86e5c33c1@mail.gmail.com>
Subject: Re: git pull on linux-next makes my system crawl to its knees and beg 
	for mercy
To: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Cc: linux-kernel@vger.kernel.org, Stephen Rothwell <sfr@canb.auug.org.au>,
       Bob Copeland <me@bobcopeland.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 3651
Lines: 81

On Fri, Dec 18, 2009 at 11:55 AM, Luis R. Rodriguez <mcgrof@gmail.com> wrote:
> On Fri, Dec 18, 2009 at 11:19 AM, Luis R. Rodriguez <mcgrof@gmail.com> wrote:
>> On Fri, Dec 18, 2009 at 9:38 AM, Bartlomiej Zolnierkiewicz
>> <bzolnier@gmail.com> wrote:
>>> On Friday 18 December 2009 06:26:29 pm Luis R. Rodriguez wrote:
>>>
>>>> on my kernel logs. Bewildered with this issue I set out to prove to
>>>> myself this issue was not a 2.6.32 issue and booted other kernels,
>>>> including Ubuntu's distro kernel on 2.6.31 and then later my own built
>>>> fresh 2.6.27.41 kernel. The issue was reproducible on all three
>>>> kernels!
>>>>
>>>> This lead me to believe this was a system / hard drive issue and
>>>> embraced myself for a system fix. I yet needed to prove this was
>>>
>>> Just some hints for ruling out the system / hard drive problem.
>>>
>>> smartctl -a /dev/sdx is your friend for checking your disk (keep an eye
>>> on anything suspicious like re-allocated sector count going up etc.)
>>
>> Sweet thanks, here's my current output, I'll try later after I get
>> some day work done to pull linux-next and make it moan. Let me know if
>> you see anything fishy.
>
> <-- snip full log -->
>
>> Also available at:
>>
>> http://bombadil.infradead.org/~mcgrof/logs/2009/12/smart-ctl-sda2.txt
>>
>>> It could be also fs related issue that shows up only under specific
>>> conditions
>>
>> OK -- I see, I used a fresh new ext3, did not make the jump to ext4.
>>
>>> (i.e. almost full partition -- some file-systems starts to
>>> crawl when the amount of available free space gets low).
>>
>> Got it, thanks, so partition has a lot of room.
>>
>> mcgrof@tux ~ $ df -h
>> Filesystem            Size  Used Avail Use% Mounted on
>> /dev/sda2              91G   43G   44G  50% /
>>
>> Also ony have one partition.
>
> GSmartControl is very cool, just ran the short self test and it passed
> without issues. I'll now run the extended self tests. I'll not that
> right after the self test I had to checkout the 2.6.32.y branch on
> hpa's tree and noticed similar type of slow down as I did with pulling
> linux-next. Only thing with linux-next is it takes ages complete which
> just makes waiting unbearable. This all makes me suspect its something
> else. But lets seee what these results on the GSmartControl yield.

I tested the same exact git pull on the other T61 laptop I have and
was able to see the same crippling effects but not as bad as with my
main T61. Different between them is the one where I see the worst
issue has a Intel(R) Core(TM)2 Duo CPU     T8100  @ 1.80GHz while the
other one has the same CPU but at 2.10GHz. The only thing I see
different between linux-next and say wireless-testing is linux-next
will have a lot more newer objects and the pull will end with a git
merge that will fail and require you to 'git reset --hard origin'. The
later part shouldn't be taken into the equation there though as I see
the issue creeping up early on during the pull, while git is counting
objects and even later compressing.

I'm starting to glare at CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y
with suspicious looks. On both boxes the CPU kept itself @ 800 MHz
during most of the git pull, I did see the CPU idle hitting 0
frequently and the CPU wait time ~ 20 or 30.

My GSmartControl extensive test is almost done.

I'll test 2.6.33-rc1 once John gets it into his tree.

  Luis
--
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/