FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Debian > Debian Kernel

 
 
LinkBack Thread Tools
 
Old 02-05-2012, 05:10 AM
Jonathan Nieder
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Hi Csaba,

Csaba GAAL wrote:

> At first it seemed that I have found a bug in GNU Parted
[...]
> “Unable to determine the size and length
> of /dev/sd[a-b][0-6]”
[...]

You are taking a long time to get to the point. ;-) I suppose after
we deal with this speed regression, it will be useful to revisit that
message and see if it can be eliminated.

[...]
> http://lists.debian.org/debian-user/2011/12/msg00153.html
[...]
> With returning to the kernel in linux-image-2.6.32-5-amd64 also did the speed.
> This also confirms that the speed degradation with 3.1.0-1-amd64 kernel has
> nothing to do with
[...]

If I understand correctly, you are reporting a speed regression
between 2.6.32.y and 3.1.8-2. Thanks for finding it.

Could you try versions 2.6.34-1~experimental.2 and
2.6.35~rc4-1~experimental.1 from <http://snapshot.debian.org/>, to
hopefully narrow down the regression range?

Thanks,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120205061017.GA1870@burratino">http://lists.debian.org/20120205061017.GA1870@burratino
 
Old 02-05-2012, 02:11 PM
Jonathan Nieder
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

merge 658670 658727
quit

Csaba GAAL wrote:

> [To: submit@bugs.debian.org]

For the future, you can followup by sending email to
658670@bugs.debian.org (for example using your mailer's "reply to all"
feature), or, if you don't have a working email setup, by using
"reportbug -N 658670". Submitting a separate new bug with each
message just makes for more work on this end...

[...]
> Just a preview of what's coming: still linux-image-2.6.39-2-amd64_2.6.39-3_amd64
> is free from this speed regression problem, however for the record I am going to
> file separate reports for:
> 1) linux-image-2.6.35-rc5-amd64_2.6.35~rc5-1~experimental.1_amd64
> 2) linux-image-2.6.36-rc5-amd64_2.6.36~rc5-1~experimental.1_amd64
> 3) linux-image-2.6.37-2-amd64_2.6.37-2_amd64
> 4) linux-image-2.6.38-2-amd64_2.6.38-5_amd64
> 5) linux-image-2.6.39-2-amd64_2.6.39-3_amd64

No thank you. So, which is the first version with the speed degredation?

Thanks for the quick followup.

Sincerely,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120205151156.GB32584@burratino">http://lists.debian.org/20120205151156.GB32584@burratino
 
Old 02-05-2012, 02:47 PM
Csaba GAL
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Dear Jonathan,

first of all thank You for the prompt reply, the merging, the proper
command to directly report on this bug and sorry for the
inconvenience. However the #612712 kicked in ("UnicodeEncodeError:
'ascii' codec can't encode character u'u201c' in position 734:
ordinal not in range(128)") therefore I'm going to use the webmail
interface.

I feel that the most important is finding where the regression is
introduced so fast forward to linux-image-3.0.0-2-amd64_3.0.0-6_amd64
and the only thing I can tell is that it is still free from this
problem.

I keep You updated.

Yours truly,

Csaba GAL

2012/2/5 Jonathan Nieder <jrnieder@gmail.com>
>
> merge 658670 658727
> quit
>
> Csaba GAAL wrote:
>
> > [To: submit@bugs.debian.org]
>
> For the future, you can followup by sending email to
> 658670@bugs.debian.org (for example using your mailer's "reply to all"
> feature), or, if you don't have a working email setup, by using
> "reportbug -N 658670". *Submitting a separate new bug with each
> message just makes for more work on this end...
>
> [...]
> > Just a preview of what's coming: still linux-image-2.6.39-2-amd64_2.6.39-3_amd64
> > is free from this speed regression problem, however for the record I am going to
> > file separate reports for:
> > 1) linux-image-2.6.35-rc5-amd64_2.6.35~rc5-1~experimental.1_amd64
> > 2) linux-image-2.6.36-rc5-amd64_2.6.36~rc5-1~experimental.1_amd64
> > 3) linux-image-2.6.37-2-amd64_2.6.37-2_amd64
> > 4) linux-image-2.6.38-2-amd64_2.6.38-5_amd64
> > 5) linux-image-2.6.39-2-amd64_2.6.39-3_amd64
>
> No thank you. *So, which is the first version with the speed degredation?
>
> Thanks for the quick followup.
>
> Sincerely,
> Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAG4jq7GbStfsKd2d-k3i8UZXhD6ENMCg-AJrz1WbQYPvaUmD7w@mail.gmail.com">http://lists.debian.org/CAG4jq7GbStfsKd2d-k3i8UZXhD6ENMCg-AJrz1WbQYPvaUmD7w@mail.gmail.com
 
Old 02-05-2012, 02:53 PM
Jonathan Nieder
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Csaba GAL wrote:

> I feel that the most important is finding where the regression is
> introduced so fast forward to linux-image-3.0.0-2-amd64_3.0.0-6_amd64
> and the only thing I can tell is that it is still free from this
> problem.

Thanks.

Pro tip: if you do a "bisection search" (i.e., try the version halfway
between the newest known ok version and oldest known broken version at
each step) then it will require fewer tests than just running through
the list.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120205155327.GA962@burratino">http://lists.debian.org/20120205155327.GA962@burratino
 
Old 02-05-2012, 03:05 PM
Csaba GAL
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Dear Jonathan,

the speed regression was introduced between
linux-image-3.0.0-2-amd64_3.0.0-6_amd64 and
linux-image-3.1.0-rc4-amd64_3.1.0~rc4-1~experimental.1_amd64. Hope
that this information helps.


Thank You once again for the prompt replies and the guidance.

Yours truly,

Csaba GAL

2012/2/5 Jonathan Nieder <jrnieder@gmail.com>:
> Csaba GAL wrote:
>
>> I feel that the most important is finding where the regression is
>> introduced so fast forward to linux-image-3.0.0-2-amd64_3.0.0-6_amd64
>> and the only thing I can tell is that it is still free from this
>> problem.
>
> Thanks.
>
> Pro tip: if you do a "bisection search" (i.e., try the version halfway
> between the newest known ok version and oldest known broken version at
> each step) then it will require fewer tests than just running through
> the list.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAG4jq7EVjAzBE-7qqAetJBcW3FeKQ7BY_bKbwdYvWUe6p+E94w@mail.gmail.co m">http://lists.debian.org/CAG4jq7EVjAzBE-7qqAetJBcW3FeKQ7BY_bKbwdYvWUe6p+E94w@mail.gmail.co m
 
Old 02-05-2012, 03:11 PM
"Russell L. Carter"
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Greetings,
Thank you for cc'ing me.

On 02/05/2012 09:05 AM, Csaba GAL wrote:
> Dear Jonathan,
>
> the speed regression was introduced between
> linux-image-3.0.0-2-amd64_3.0.0-6_amd64 and
> linux-image-3.1.0-rc4-amd64_3.1.0~rc4-1~experimental.1_amd64. Hope
> that this information helps.

I can confirm these and 3.2~rc7-1~experimental.1 as well.
I never saw it in any 2.6.xx series kernel.

Best,
Russell
(I'm new to the debian process and would like to get better at working
within it so if you see an obvious way I can improve please let me know)

>
> Thank You once again for the prompt replies and the guidance.
>
> Yours truly,
>
> Csaba GAL
>
> 2012/2/5 Jonathan Nieder <jrnieder@gmail.com>:
>> Csaba GAL wrote:
>>
>>> I feel that the most important is finding where the regression is
>>> introduced so fast forward to linux-image-3.0.0-2-amd64_3.0.0-6_amd64
>>> and the only thing I can tell is that it is still free from this
>>> problem.
>>
>> Thanks.
>>
>> Pro tip: if you do a "bisection search" (i.e., try the version halfway
>> between the newest known ok version and oldest known broken version at
>> each step) then it will require fewer tests than just running through
>> the list.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4F2EAA3C.8080306@pinyon.org">http://lists.debian.org/4F2EAA3C.8080306@pinyon.org
 
Old 02-05-2012, 03:34 PM
Jonathan Nieder
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Russell L. Carter wrote:
> On 02/05/2012 09:05 AM, Csaba GAL wrote:

>> the speed regression was introduced between
>> linux-image-3.0.0-2-amd64_3.0.0-6_amd64 and
>> linux-image-3.1.0-rc4-amd64_3.1.0~rc4-1~experimental.1_amd64. Hope
>> that this information helps.
>
> I can confirm these and 3.2~rc7-1~experimental.1 as well.

Thanks, both. I'll try to get back to you later today with
information on the next step. (Feel free to ping me if I don't.)



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120205163439.GB2731@burratino">http://lists.debian.org/20120205163439.GB2731@burratino
 
Old 02-05-2012, 03:45 PM
Csaba GAL
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

Dear Sirs,

thank You for the prompt response.

Actually and not so surprisingly I can confirm that the speed
regression is also included in
linux-image-3.2.0-1-amd64_3.2.2-1_amd64.

I can also confirm that this is not an XFS related bug as on my
notebook using linux-image-3.1.0-1-amd64 with exclusively XFS there is
no such speed regression problem.


Yours truly,

Csaba GAL


2012/2/5 Jonathan Nieder <jrnieder@gmail.com>:
> Russell L. Carter wrote:
>> On 02/05/2012 09:05 AM, Csaba GAL wrote:
>
>>> the speed regression was introduced between
>>> linux-image-3.0.0-2-amd64_3.0.0-6_amd64 and
>>> linux-image-3.1.0-rc4-amd64_3.1.0~rc4-1~experimental.1_amd64. Hope
>>> that this information helps.
>>
>> I can confirm these and 3.2~rc7-1~experimental.1 as well.
>
> Thanks, both. *I'll try to get back to you later today with
> information on the next step. *(Feel free to ping me if I don't.)



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAG4jq7E1O9Kikc3MiLy64Ek0WF28niTf2fDfr7DhYZNU7M5RQ Q@mail.gmail.com">http://lists.debian.org/CAG4jq7E1O9Kikc3MiLy64Ek0WF28niTf2fDfr7DhYZNU7M5RQ Q@mail.gmail.com
 
Old 04-01-2012, 01:07 PM
Solomon Peachy
 
Default Bug#658670: 3w-9xxx abnormally slows the system down

For what it's worth, this bug is not debian-specific, nor does it have
anything to do with XFS. I'm seeing a similar regression on Fedora
[15|16] systems. The bug was not present in the 3.0.x (or older)
kernels, but is still present in 3.1.x and 3.2.x. I haven't had a
chance to test 3.3.x yet.

My particular setup has a pair of 3Ware 9550SX cards plugged into PCI-X
slots on an Asus M2N-LR motherboard. I can provide more details if
necessary.

I haven't reported this problem upstream to Fedora or the kernel.org due
to an inability to save the kernel messages due to the system being
unusable, but in my case the kernel log is full of scsi command timeout
messages.

- Solomon
--
Solomon Peachy pizza at shaftnet dot org
Melbourne, FL ^^ (mail/jabber/gtalk) ^^
Quidquid latine dictum sit, altum viditur.
 

Thread Tools




All times are GMT. The time now is 01:03 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org