Bug 16334 - [DRM 945GM] i915 Oops (i915_driver_lastclose) when shutdown X
Summary: [DRM 945GM] i915 Oops (i915_driver_lastclose) when shutdown X
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: x86 (IA32) Linux (All)
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2008-06-13 01:34 UTC by Jie Luo
Modified: 2017-07-24 23:11 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg (47.83 KB, text/plain)
2008-06-13 01:34 UTC, Jie Luo
no flags Details

Description Jie Luo 2008-06-13 01:34:01 UTC
Created attachment 17100 [details]
dmesg

The i915 drm module cause a oops when I try to shutdown X, i915_driver_lastclose seems to be the cause of problem.

And this problem also exists in the drm git master.


Kernel: 2.6.25-gentoo-r4
xserver: git master commit d15b3790307053587df8daed1936ff6923881b63
mesa: drm-gem branch commit 407ce3da3c53c9ebba0fbf827d7b0f610122d44b
libdrm: drm-gem branch commit 100d6f3f83b559be000f69c04f9e5596cdc17c77
drm: drm-gem branch commit 100d6f3f83b559be000f69c04f9e5596cdc17c77
xf86-video-intel: drm-gem branch commit 02d7141ac395446613edba0f2e2a951e750e4e9f
Comment 1 Eric Anholt 2008-06-13 09:43:25 UTC
Removing the GEM from subject since reporter said it was in master, too.
Comment 2 Keith Packard 2008-06-13 20:37:18 UTC
I've just rewritten the leavevt_ioctl for GEM, having found several crashes along that path. That should fix the drm-gem branch piece of this bug.
Comment 3 Jie Luo 2008-06-14 15:49:32 UTC
(In reply to comment #2)
> I've just rewritten the leavevt_ioctl for GEM, having found several crashes
> along that path. That should fix the drm-gem branch piece of this bug.
> 

I tried the latest drm, and I have some problem with it. There are a lot of "[drm:i915_wait_request] *ERROR* i915_wait_request returns -512" message in my kern.log. This used to occur when MSI is enabled, and disappear when MSI is disabled. But now it appears no matter I enable or disable MSI in kernel. Is there anything addition I should do to get it works fine.
Comment 4 Keith Packard 2008-06-14 17:21:04 UTC
If you're using the latest drm_gem branch (3e48e144992fb11b31875989d45bc8a7c041cdef or newer), that message should have (awaiting %d at %d) appended to it.
Comment 5 Jie Luo 2008-06-15 11:40:59 UTC
(In reply to comment #4)
> If you're using the latest drm_gem branch
> (3e48e144992fb11b31875989d45bc8a7c041cdef or newer), that message should have
> (awaiting %d at %d) appended to it.
> 

Yes, it looks like this.

Jun 15 07:44:17 Fates [  922.770832] [drm:i915_wait_request] *ERROR* i915_wait_request returns -512 (awaiting 44794 at 44759)
Jun 15 07:44:17 Fates [  922.792122] [drm:i915_wait_request] *ERROR* i915_wait_request returns -512 (awaiting 44838 at 44814)
Jun 15 07:44:17 Fates [  922.814410] [drm:i915_wait_request] *ERROR* i915_wait_request returns -512 (awaiting 44882 at 44853)
Jun 15 07:44:17 Fates [  922.847440] [drm:i915_wait_request] *ERROR* i915_wait_request returns -512 (awaiting 44926 at 44902)

And I also got a lot of "intel_bufmgr_gem.c:591: Error setting memory domains 9 (00000001 00000000): Interrupted system call ." messages in my gdm.log.

Currently I didn't see the oops, but I did have hard lock when running glxgears with current drm-gem. The screen is stay still, mouse and keyboard have no response, and I can't generate ACPI event to shutdown.
Comment 6 Johannes Engel 2008-07-08 02:23:22 UTC
The problem still persists for me using drm-gem of yesterday. Is there any kind of information I can provide to help chasing it down?
Comment 7 Gordon Jin 2009-02-05 00:15:13 UTC
Does this still exist with the latest code?
Comment 8 Johannes Engel 2009-02-05 01:01:23 UTC
For me it is gone.
Comment 9 Jie Luo 2009-02-07 18:47:34 UTC
(In reply to comment #7)
> Does this still exist with the latest code?
> 

Didn't see it for a long time, assume it was fixed.


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.