Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#23599 - [kernel26] Suspend regressions on intel graphics

Attached to Project: Arch Linux
Opened by Adrian C. (anrxc) - Tuesday, 05 April 2011, 21:13 GMT
Last edited by Andrea Scarpino (BaSh) - Saturday, 17 September 2011, 09:16 GMT
Task Type Bug Report
Category Packages: Testing
Status Closed
Assigned To Tobias Powalowski (tpowa)
Thomas Bächler (brain0)
Architecture i686
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Latest kernel26 in testing broke suspend on intel graphics (yet again, how do they do it). If X is running at the time of pm-suspend it returns broken on resume. Terminal emulators lag, black and white stripes appear sometimes etc.

Additional info:
* kernel26 2.6.38.2-1
* dmesg excerpt
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -11 (awaiting 55 at 44, next 56)
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -11 (awaiting 60 at 44, next 61)
[drm:i915_reset] *ERROR* GPU hanging too fast, declaring wedged!
[drm:i915_reset] *ERROR* Failed to reset chip.
* Xorg.0.log excerpt
[ 9400.923] (EE) intel(0): Detected a hung GPU, disabling acceleration.
* Hardware
Intel GMA4500HD
Intel GMA5700MHD

Steps to reproduce:
/usr/sbin/pm-suspend
- using either uswsusp or kernel


There are also known problems with pm-hibernate. Which I did not dare to test on my hardware and risk file-system corruption.
* Bugzilla
https://bugzilla.kernel.org/show_bug.cgi?id=32652
This task depends upon

Closed by  Andrea Scarpino (BaSh)
Saturday, 17 September 2011, 09:16 GMT
Reason for closing:  Not a bug
Additional comments about closing:  seems fixed
Comment by Adrian C. (anrxc) - Tuesday, 12 April 2011, 18:25 GMT
There was a commit in xf86-video-intel that is not perfecly clear to me, but suggests the hangs might be caused by the userspace driver:

"Tell users to grab i915_error_state on gpu hangs"
http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=fb40bf2b33a


There was also a commit that fixes some hangs, but all three of my chips are
newer than the supposedly affected versions:

"Turn relaxed-fencing off by default for older (pre-G33) chipsets"
http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=686018f283f
Comment by Jelle van der Waa (jelly) - Friday, 16 September 2011, 21:03 GMT
status?
Comment by Adrian C. (anrxc) - Friday, 16 September 2011, 21:30 GMT
Still broken but in another way - every 2nd or 3rd suspend cycle machine will freeze. Still a piece of sh* as usual, but this particular bug no longer reproducable. Thank you.

Loading...