Bug 16190 - Compiz triggers abort in dri_fake_emit_reloc
Summary: Compiz triggers abort in dri_fake_emit_reloc
Status: RESOLVED DUPLICATE of bug 15639
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965 (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: Eric Anholt
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-06-01 20:07 UTC by Ben Gamari
Modified: 2008-08-05 21:06 UTC (History)
3 users (show)

See Also:
i915 platform:
i915 features:


Attachments
xserver backtrace (2.06 KB, text/plain)
2008-06-01 20:07 UTC, Ben Gamari
Details
Log of running Xorg session (215.58 KB, text/plain)
2008-06-19 07:40 UTC, Ben Gamari
Details

Description Ben Gamari 2008-06-01 20:07:09 UTC
Created attachment 16861 [details]
xserver backtrace

Sporatically while running compiz, the xserver will crash on abort(target_fake->is_static || target_fake->size_accounted) in dri_fake_emit_reloc() while drawing a primitive array. Happens several times a day although I haven't found a reliable means of reproducing. Backtrace attached
Comment 1 Ben Gamari 2008-06-10 08:15:37 UTC
Does anyone have any ideas on this? It still seems to occur on today's git and quite frequently at that (just crashed twice in 30 minutes).
Comment 2 Michael Fu 2008-06-18 19:29:23 UTC
Ben, would you please kindly provide us a detailed list of your environment(HW/SW)? the http://www.intellinuxgraphics.org/how_to_report_bug.html has some hints. thanks!
Comment 3 Ben Gamari 2008-06-19 07:39:56 UTC
(In reply to comment #2)
> Ben, would you please kindly provide us a detailed list of your
> environment(HW/SW)? the
> http://www.intellinuxgraphics.org/how_to_report_bug.html has some hints.
> thanks!
> 

i965 in Dell D830 with Core 2 Duo. Gentoo. Running xserver, drm, mesa, xf86-video-intel from git. 
Comment 4 Ben Gamari 2008-06-19 07:40:26 UTC
Created attachment 17229 [details]
Log of running Xorg session
Comment 5 Ben Gamari 2008-06-19 08:03:25 UTC
Actually, my bad. I'm presently running drm from the 2.6.26-rc5 kernel due to issues with vblank rework in git.

From dmesg:

Jun 18 10:32:16 mercury [drm] Initialized drm 1.1.0 20060810
Jun 18 10:32:16 mercury ACPI: PCI Interrupt 0000:00:02.0[A] -> GSI 16 (level, low) -> IRQ 16
Jun 18 10:32:16 mercury PCI: Setting latency timer of device 0000:00:02.0 to 64
Jun 18 10:32:16 mercury [drm] Initialized i915 1.6.0 20060119 on minor 0
Comment 6 Ben Gamari 2008-06-23 17:35:58 UTC
Am I the only one who thinks that bugs #16316 and #16190 sharing dri_fake_emit_reloc is a little more than coincidence?
Comment 7 Dave Airlie 2008-07-13 01:48:18 UTC
any chance we could retest with mesa current master? I checked in a fix for one reloc emit bug.
Comment 8 Ben Gamari 2008-07-13 16:09:28 UTC
It's actually been a while since I've seen this crash. I've been tracking master for some time so I suspect it might be gone. Are there any recent commits that might have fixed it?
Comment 9 Brian Rogers 2008-08-04 14:57:35 UTC
I was getting this assert 100% of the time when starting my session while running the drm-gem branches of drm, mesa, and the intel driver.

But the master branch is running fine.
Comment 10 Hedayat Vatankhah 2008-08-05 13:10:50 UTC
I see the same error message when running rcssserver3d application (sf.net/projects/sserver). I use Fedora 9 with latest updates for mesa* packages and intel driver.
Comment 11 Michael Fu 2008-08-05 21:06:47 UTC

*** This bug has been marked as a duplicate of bug 15639 ***


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.