Bug 67089 - GPU hang whilst running counter-strike 1.6 game
Summary: GPU hang whilst running counter-strike 1.6 game
Status: RESOLVED INVALID
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965 (show other bugs)
Version: 9.1
Hardware: All Linux (All)
: medium major
Assignee: Tapani Pälli
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-07-19 13:27 UTC by Arnas Milaševičius
Modified: 2017-02-10 22:39 UTC (History)
2 users (show)

See Also:
i915 platform:
i915 features:


Attachments
i915_error_state after the hang (2.23 MB, text/plain)
2013-07-19 13:27 UTC, Arnas Milaševičius
Details
glxinfo (13.96 KB, text/plain)
2013-07-19 13:33 UTC, Arnas Milaševičius
Details

Description Arnas Milaševičius 2013-07-19 13:27:28 UTC
Created attachment 82689 [details]
i915_error_state after the hang

When I'm playing Counter-Strike 1.6 game, from time to time the game just freezes for like 10-20 seconds and I can do nothing, I can't even minimize the game. After a freeze I checked dmesg and saw this: [ 4337.658522] [drm] capturing error event; look for more information in/sys/kernel/debug/dri/0/i915_error_state

Since i915_error_state is really big, I'll attach it.
Comment 1 Arnas Milaševičius 2013-07-19 13:33:23 UTC
Created attachment 82690 [details]
glxinfo
Comment 2 Ian Romanick 2013-07-19 15:38:30 UTC
This sounds like a GPU hang.  Can you provide the output of 'uname -r'?

Ken, can you try to reproduce this?
Comment 3 Arnas Milaševičius 2013-07-19 15:54:56 UTC
Ian, there you go:

3.9.9-1-ARCH
Comment 4 Tapani Pälli 2013-08-07 05:19:42 UTC
Got the game now, will try to reproduce with IVB
Comment 5 Tapani Pälli 2013-08-07 07:35:57 UTC
So far could not reproduce, I am running Steam on Ubuntu with 3.9.9 kernel and origin/9.1 branch of Mesa, no problems visible and no error state dumped by the driver, will try some more with older Mesa.
Comment 6 Ian Romanick 2013-08-08 01:54:03 UTC
(In reply to comment #4)
> Got the game now, will try to reproduce with IVB

The original report was on Sandybridge, so it may be specific to that hardware.
Comment 7 Tapani Pälli 2013-08-08 04:22:18 UTC
(In reply to comment #6)
> (In reply to comment #4)
> > Got the game now, will try to reproduce with IVB
> 
> The original report was on Sandybridge, so it may be specific to that
> hardware.

Yep, I wanted to try first with IVB as I have already working Steam setup there. I'm having some problems getting Steam to work on my SNB (steam under ubuntu chroot in fedora installation) but I'll try to get that working today.
Comment 8 Tapani Pälli 2013-08-08 09:35:42 UTC
Got it working, no problems seen with Sandybridge running kernel 3.9.11, mesa 9.1.5 (git-07c3e87), will test some more.
Comment 9 Tapani Pälli 2013-08-08 11:18:04 UTC
Tried also running with kernel 3.9.9-200 (from fedora 18 kernel builds), could not still reproduce the hang.
Comment 10 Ian Romanick 2013-08-08 18:12:33 UTC
Arnas,

Can you perhaps capture a timedemo that reproduces the hang?  Does it occur more often in any particular map?  Any clues? :)
Comment 11 Arnas Milaševičius 2013-08-14 12:31:40 UTC
Ian,

Sorry for not responding for so long, I didn't see an email. What I've noticed tho, that for the last like 4-5 days, sometimes the GPU hang doesn't occur for like 30 minutes, but sometimes there're like 2 LONG (like 30 sec) hangs in a row, like 1 hang occurs, then I move and after like 1 sec another long hang occurs.

P.S. what do you mean by timedemo? You want me to record a game how I'm playing and to see where the hang occurs or what?
Comment 12 Tapani Pälli 2013-08-15 07:57:22 UTC
Arnas, would you be willing and able to test with a different Mesa version? Your glxinfo states 9.1.5 but I think there is no way for me to know if I have exactly same 9.1.5 as you. Are you using Mesa from Arch repositories and does it offer you different versions? One difference I see is that I have SNB desktop while yours is mobile, otherwise we *should* have at least close to same kernel and mesa.
Comment 13 Ian Romanick 2013-08-15 21:49:09 UTC
Since we can't reproduce this, I'm going to take it off our release blocker list.  I'm not going to close it (yet).
Comment 14 Annie 2017-02-10 22:39:17 UTC
Dear Reporter,

This Mesa bug has been in the "NEEDINFO" status for over 60 days. I am closing this bug based on lack of response but feel free to reopen if resolution is still needed. Please ensure you're supplying the correct information as requested.

Thank you.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.