2009-12-09 12:00:29 |
Alberto Milone |
bug |
|
|
added bug |
2009-12-09 12:00:29 |
Alberto Milone |
attachment added |
|
fix_sync_to vblank_when_vga_is_switched_off.patch http://launchpadlibrarian.net/36620772/fix_sync_to%20vblank_when_vga_is_switched_off.patch |
|
2009-12-09 12:00:51 |
Alberto Milone |
linux (Ubuntu): status |
New |
Confirmed |
|
2009-12-09 12:00:54 |
Alberto Milone |
linux (Ubuntu): importance |
Undecided |
High |
|
2009-12-09 12:36:47 |
Alberto Milone |
linux (Ubuntu): assignee |
|
Alberto Milone (albertomilone) |
|
2009-12-09 12:37:33 |
Alberto Milone |
description |
This is a tracking bug for an SRU proposal for Karmic.
This problem was reported in a private OEM bug report for Dell and has been solved by the upstream.
Description:
drm/i915: Fix sync to vblank when VGA output is turned off
In current vblank-wait implementation, if we turn off VGA output,
drm_wait_vblank will still wait on the disabled pipe until timeout,
because vblank on the pipe is assumed be enabled. This would cause
slow system response on some system such as moblin.
Steps to reproduce:
The problem can be reproduced with modern Intel gpus such as i945, i965, etc. and affects only VGA connectors. Typing the following commands will enable and disable an external screen through VGA thus triggering the bug and causing a massive slowdown:
xrandr --output VGA1 --auto
xrandr --output VGA1 --off
The bug is a regression in the drm code.
The patch is in the drm-intel branch and in the linux-next branch. Here's a link to the commit in drm-intel:
http://git.kernel.org/?p=linux/kernel/git/anholt/drm-intel.git;a=commit;h=778c902640530371a169ad1c03566e7c51b09874
I have applied (and slightly adapted, as it didn't apply cleanly) and tested the attached patch successfully.
I deem it very important for both OEM projects and for desktop users to have this patch included in Karmic and Lucid ASAP. |
This is a tracking bug for an SRU proposal for Karmic.
This problem was reported in a private OEM bug report for Dell and has been solved by the upstream.
Description:
drm/i915: Fix sync to vblank when VGA output is turned off
In current vblank-wait implementation, if we turn off VGA output,
drm_wait_vblank will still wait on the disabled pipe until timeout,
because vblank on the pipe is assumed be enabled. This would cause
slow system response on some system such as moblin.
Steps to reproduce:
The problem can be reproduced with modern Intel gpus such as i945, i965, etc. and affects only VGA connectors. Typing the following commands will enable and disable an external screen through VGA thus triggering the bug and causing a massive slowdown:
xrandr --output VGA1 --auto
xrandr --output VGA1 --off
The bug is a regression in the drm code.
The patch is in the drm-intel branch and in the linux-next branch. Here's a link to the commit in drm-intel:
http://git.kernel.org/?p=linux/kernel/git/anholt/drm-intel.git;a=commit;h=778c902640530371a169ad1c03566e7c51b09874
I have applied (and slightly adapted, as it didn't apply cleanly) and tested the attached patch successfully.
I deem it very important for both OEM projects and for the Ubuntu desktop that we include this patch in Karmic and Lucid ASAP.
|
|
2009-12-09 12:48:51 |
Alberto Milone |
description |
This is a tracking bug for an SRU proposal for Karmic.
This problem was reported in a private OEM bug report for Dell and has been solved by the upstream.
Description:
drm/i915: Fix sync to vblank when VGA output is turned off
In current vblank-wait implementation, if we turn off VGA output,
drm_wait_vblank will still wait on the disabled pipe until timeout,
because vblank on the pipe is assumed be enabled. This would cause
slow system response on some system such as moblin.
Steps to reproduce:
The problem can be reproduced with modern Intel gpus such as i945, i965, etc. and affects only VGA connectors. Typing the following commands will enable and disable an external screen through VGA thus triggering the bug and causing a massive slowdown:
xrandr --output VGA1 --auto
xrandr --output VGA1 --off
The bug is a regression in the drm code.
The patch is in the drm-intel branch and in the linux-next branch. Here's a link to the commit in drm-intel:
http://git.kernel.org/?p=linux/kernel/git/anholt/drm-intel.git;a=commit;h=778c902640530371a169ad1c03566e7c51b09874
I have applied (and slightly adapted, as it didn't apply cleanly) and tested the attached patch successfully.
I deem it very important for both OEM projects and for the Ubuntu desktop that we include this patch in Karmic and Lucid ASAP.
|
This is a tracking bug for an SRU proposal for Karmic.
This problem was reported in a private OEM bug report for Dell and has been solved by upstream.
Description:
drm/i915: Fix sync to vblank when VGA output is turned off
In current vblank-wait implementation, if we turn off VGA output,
drm_wait_vblank will still wait on the disabled pipe until timeout,
because vblank on the pipe is assumed be enabled. This would cause
slow system response on some system such as moblin.
Steps to reproduce:
The problem can be reproduced with modern Intel gpus such as i945, i965, etc. and affects only VGA connectors. Typing the following commands will enable and disable an external screen through VGA thus triggering the bug and causing a massive slowdown:
xrandr --output VGA1 --auto
xrandr --output VGA1 --off
The bug is a regression in the drm code.
The patch is in the drm-intel branch and in the linux-next branch. Here's a link to the commit in drm-intel:
http://git.kernel.org/?p=linux/kernel/git/anholt/drm-intel.git;a=commit;h=778c902640530371a169ad1c03566e7c51b09874
I have applied (and slightly adapted, as it didn't apply cleanly) and tested the attached patch successfully.
I deem it very important for both OEM projects and for the Ubuntu desktop that we include this patch in Karmic and Lucid ASAP.
|
|
2009-12-09 17:21:47 |
Andy Whitcroft |
tags |
|
kernel-series-unknown |
|
2009-12-11 15:52:29 |
Alberto Milone |
tags |
kernel-series-unknown |
2.6.32 karmic lucid |
|
2009-12-16 16:59:10 |
Stefan Bader |
nominated for series |
|
Ubuntu Karmic |
|
2009-12-16 16:59:10 |
Stefan Bader |
bug task added |
|
linux (Ubuntu Karmic) |
|
2009-12-16 17:01:02 |
Stefan Bader |
linux (Ubuntu Karmic): importance |
Undecided |
High |
|
2009-12-16 17:01:02 |
Stefan Bader |
linux (Ubuntu Karmic): status |
New |
Fix Committed |
|
2009-12-17 17:00:08 |
Launchpad Janitor |
linux (Ubuntu): status |
Confirmed |
Fix Released |
|
2010-01-08 14:54:50 |
Martin Pitt |
tags |
2.6.32 karmic lucid |
2.6.32 karmic lucid verification-needed |
|
2010-02-09 12:28:47 |
Launchpad Janitor |
branch linked |
|
lp:ubuntu/linux-fsl-imx51 |
|
2010-03-04 08:42:32 |
Launchpad Janitor |
linux (Ubuntu Karmic): status |
Fix Committed |
Fix Released |
|