Re: [arch-general] [arch-dev-public] kernel 2.6.32-1

newer
[arch-general] [OT] What is wrong...

Emmanuel Benisty

5 Dec 2009 5 Dec '09

4:13 p.m.

On Sat, Dec 5, 2009 at 10:46 PM, Allan McRae wrote:

Thomas Bächler wrote:

...
Allan McRae schrieb:

...
That works for me. I initially thought it was strange that it came on but now I see why. extreme-tuxracer is giving me (>50%) more FPS with this update.

Intel DRM has been improved, I noticed that too - probably due to framebuffer compression.

KMS is enabled by default because the next intel driver version will require it to work.

With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine...

It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

Show replies by date

Thomas Bächler

5 Dec 5 Dec

4:24 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

Emmanuel Benisty schrieb:

...

...
With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine...

It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

What's your chipset? Allan's chipset? No such issues here on 945GM.

Emmanuel Benisty

4:29 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Sat, Dec 5, 2009 at 11:24 PM, Thomas Bächler wrote:

...

Emmanuel Benisty schrieb:

...
...
With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine...

It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

What's your chipset? Allan's chipset? No such issues here on 945GM.

Integrated Graphics Chipset: Intel(R) 965GM

Roman Kyrylych

4:39 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Sat, Dec 5, 2009 at 18:29, Emmanuel Benisty wrote:

...

On Sat, Dec 5, 2009 at 11:24 PM, Thomas Bächler wrote:

...
What's your chipset? Allan's chipset? No such issues here on 945GM.

Integrated Graphics Chipset: Intel(R) 965GM

I have the same chipset. No issues. -- Roman Kyrylych (Роман Кирилич)

Chris Bannister

5:02 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

Whats the plan re devtmpfs ? 2009/12/5 Roman Kyrylych :

...

On Sat, Dec 5, 2009 at 18:29, Emmanuel Benisty wrote:

...
On Sat, Dec 5, 2009 at 11:24 PM, Thomas Bächler wrote:

...
What's your chipset? Allan's chipset? No such issues here on 945GM.

Integrated Graphics Chipset: Intel(R) 965GM

I have the same chipset. No issues.

-- Roman Kyrylych (Роман Кирилич)

Allan McRae

6 Dec 6 Dec

12:12 a.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

Roman Kyrylych wrote:

...

On Sat, Dec 5, 2009 at 18:29, Emmanuel Benisty wrote:

...
On Sat, Dec 5, 2009 at 11:24 PM, Thomas Bächler wrote:

...
What's your chipset? Allan's chipset? No such issues here on 945GM. Integrated Graphics Chipset: Intel(R) 965GM

I have the same chipset. No issues.

Same chipset with issues

Emmanuel Benisty

7 Dec 7 Dec

1:42 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Sun, Dec 6, 2009 at 7:12 AM, Allan McRae wrote:

...

Roman Kyrylych wrote:

...
On Sat, Dec 5, 2009 at 18:29, Emmanuel Benisty wrote:

...
On Sat, Dec 5, 2009 at 11:24 PM, Thomas Bächler wrote:

...
What's your chipset? Allan's chipset? No such issues here on 945GM.

Integrated Graphics Chipset: Intel(R) 965GM

I have the same chipset. No issues.

Same chipset with issues

Just FTR, I have bisected the issue and found out this commit was to blame: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=... I've reverted it and rebuilt 2.6.32. Issue is gone.

Daenyth Blank

9:51 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Mon, Dec 7, 2009 at 08:42, Emmanuel Benisty wrote:

...

Just FTR, I have bisected the issue and found out this commit was to blame: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=...

I've reverted it and rebuilt 2.6.32. Issue is gone.

Did you report this as a fix upstream?

Thomas Bächler

10:01 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

Daenyth Blank schrieb:

...

On Mon, Dec 7, 2009 at 08:42, Emmanuel Benisty wrote:

...
Just FTR, I have bisected the issue and found out this commit was to blame: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=...

I've reverted it and rebuilt 2.6.32. Issue is gone.

Did you report this as a fix upstream?

This is on the intel-gfx list and has been replied to by one of the devs.

Daenyth Blank

10:02 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Mon, Dec 7, 2009 at 17:01, Thomas Bächler wrote:

...

This is on the intel-gfx list and has been replied to by one of the devs.

Sweet :D

Emmanuel Benisty

8 Dec 8 Dec

4:23 a.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Tue, Dec 8, 2009 at 5:02 AM, Daenyth Blank wrote:

...

On Mon, Dec 7, 2009 at 17:01, Thomas Bächler wrote:

...
This is on the intel-gfx list and has been replied to by one of the devs.

Sweet :D

actually, the thread is dead silent since my melt-your-cpu-git-bisecting result but I keep the faith :) in the meantime, attached is the patch to revert this commit if anyone is interested (it's not a proper bugfix, just a revert patch)

Alexander Lam

12 Dec 12 Dec

4:27 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

I have an Acer aspire one (945gse) and I don't have these problems....but perhaps a temporary workaround is to set the module parameter powersave = 0 i915.powersave=0 On Mon, Dec 7, 2009 at 11:23 PM, Emmanuel Benisty wrote:

...

On Tue, Dec 8, 2009 at 5:02 AM, Daenyth Blank > wrote:

...
On Mon, Dec 7, 2009 at 17:01, Thomas Bächler wrote:

...
This is on the intel-gfx list and has been replied to by one of the devs.

Sweet :D

actually, the thread is dead silent since my melt-your-cpu-git-bisecting result but I keep the faith :) in the meantime, attached is the patch to revert this commit if anyone is interested (it's not a proper bugfix, just a revert patch)

-- Alexander Lam

Alexander Lam

10:08 p.m.

New subject: [arch-general] Fwd: [arch-dev-public] kernel 2.6.32-1

I have an Acer aspire one (945gse) and I don't have these problems....but perhaps a temporary workaround is to set the module parameter powersave = 0 i915.powersave=0 ---------- Forwarded message ---------- From: Emmanuel Benisty Date: Mon, Dec 7, 2009 at 11:23 PM Subject: Re: [arch-general] [arch-dev-public] kernel 2.6.32-1 To: General Discusson about Arch Linux On Tue, Dec 8, 2009 at 5:02 AM, Daenyth Blank > wrote:

...

On Mon, Dec 7, 2009 at 17:01, Thomas Bächler wrote:

...
This is on the intel-gfx list and has been replied to by one of the devs.

Sweet :D

Xavier

5 Dec 5 Dec

7:28 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Sat, Dec 5, 2009 at 5:13 PM, Emmanuel Benisty wrote:

...

On Sat, Dec 5, 2009 at 10:46 PM, Allan McRae wrote:

...
Thomas Bächler wrote:

...
Allan McRae schrieb:

...
That works for me. I initially thought it was strange that it came on but now I see why. extreme-tuxracer is giving me (>50%) more FPS with this update.

Intel DRM has been improved, I noticed that too - probably due to framebuffer compression.

KMS is enabled by default because the next intel driver version will require it to work.

With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine...

It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

Does upstream know about this ? If you have no idea, make sure they know :) If the issue happens before starting X, just check your kernel log / dmesg. If it happens after X, check both dmesg and Xorg.0.log

Emmanuel Benisty

7:47 p.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Sun, Dec 6, 2009 at 2:28 AM, Xavier wrote:

...

On Sat, Dec 5, 2009 at 5:13 PM, Emmanuel Benisty wrote:

...
On Sat, Dec 5, 2009 at 10:46 PM, Allan McRae wrote:

...
Thomas Bächler wrote:

...
Allan McRae schrieb:

...
That works for me. I initially thought it was strange that it came on but now I see why. extreme-tuxracer is giving me (>50%) more FPS with this update.

Intel DRM has been improved, I noticed that too - probably due to framebuffer compression.

KMS is enabled by default because the next intel driver version will require it to work.

With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine...

It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

Does upstream know about this ? If you have no idea, make sure they know :)

As I had/still have no idea, I've started a thread on LKML earlier today, which has been CC'd to intel-gfx since.

...

If the issue happens before starting X, just check your kernel log / dmesg. If it happens after X, check both dmesg and Xorg.0.log

It happens in X only and I didn't notice anything suspicious in Xorg.0.log or dmesg...

Allan McRae

6 Dec 6 Dec

12:16 a.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

Emmanuel Benisty wrote:

...

On Sun, Dec 6, 2009 at 2:28 AM, Xavier wrote:

...
On Sat, Dec 5, 2009 at 5:13 PM, Emmanuel Benisty wrote:

...
On Sat, Dec 5, 2009 at 10:46 PM, Allan McRae wrote:

...
Thomas Bächler wrote:

...
...
That works for me. I initially thought it was strange that it came on but now I see why. extreme-tuxracer is giving me (>50%) more FPS with this update. Intel DRM has been improved, I noticed that too - probably due to

Allan McRae schrieb: framebuffer compression.

KMS is enabled by default because the next intel driver version will require it to work. With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine... It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

Does upstream know about this ? If you have no idea, make sure they know :)

As I had/still have no idea, I've started a thread on LKML earlier today, which has been CC'd to intel-gfx since.

Can I have a link please?

Emmanuel Benisty

3:02 a.m.

New subject: [arch-general] [arch-dev-public] kernel 2.6.32-1

On Sun, Dec 6, 2009 at 7:16 AM, Allan McRae wrote:

...

Emmanuel Benisty wrote:

...
On Sun, Dec 6, 2009 at 2:28 AM, Xavier wrote:

...
On Sat, Dec 5, 2009 at 5:13 PM, Emmanuel Benisty wrote:

...
On Sat, Dec 5, 2009 at 10:46 PM, Allan McRae wrote:

...
Thomas Bächler wrote:

...
Allan McRae schrieb: > > That works for me. I initially thought it was strange that it came > on > but now I see why. extreme-tuxracer is giving me (>50%) more FPS > with this > update.

Intel DRM has been improved, I noticed that too - probably due to framebuffer compression.

KMS is enabled by default because the next intel driver version will require it to work.

With KMS on, my screen flickers every 20sec or so. Tried with no xorg.conf and the same thing. I even got some sort of yellow/orange screen of death. I did not use KMS with 2.6.31 so this may not be kernel related. I'll look into it later but disabling kms from the grub prompt works fine...

It's definitely a kernel regression, I'm having the exact same issue with .32 only (using KMS since few kernel releases)

Does upstream know about this ? If you have no idea, make sure they know :)

As I had/still have no idea, I've started a thread on LKML earlier today, which has been CC'd to intel-gfx since.

Can I have a link please?

sure, it's there: http://lkml.org/lkml/2009/12/5/4 and there http://lists.freedesktop.org/archives/intel-gfx/2009-December/005077.html

5288

Age (days ago)

5295

Last active (days ago)

List overview

Download

16 comments

8 participants

participants (8)

Alexander Lam
Allan McRae
Chris Bannister
Daenyth Blank
Emmanuel Benisty
Roman Kyrylych
Thomas Bächler
Xavier