[Bug 70594] New: Graphics performance drops + GPU hangs in ubuntu 13.10 (linux 3.11)

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Fri Oct 18 04:15:13 CEST 2013


https://bugs.freedesktop.org/show_bug.cgi?id=70594

          Priority: medium
            Bug ID: 70594
                CC: intel-gfx-bugs at lists.freedesktop.org
          Assignee: intel-gfx-bugs at lists.freedesktop.org
           Summary: Graphics performance drops + GPU hangs in ubuntu 13.10
                    (linux 3.11)
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
          Severity: critical
    Classification: Unclassified
                OS: All
          Reporter: shuhao at shuhaowu.com
          Hardware: All
            Status: NEW
           Version: unspecified
         Component: DRM/Intel
           Product: DRI

Created attachment 87809
  --> https://bugs.freedesktop.org/attachment.cgi?id=87809&action=edit
i915_error_state

I'm experiencing some really bad GPU hangs + graphics overall slowdown while
playing games such as Kerbal Space Program.

This might be a temperature dependent thing, but right now the system is not
cooling itself via lowering the cpu clock when reaching ~95C like it did with
Ubuntu 13.04, but just mysteriously decreasing the performance (games become
very laggy) at ~86C (lm-sensor say that is the "high temperature") in Ubuntu
13.10. The laptop in question is a Thinkpad T420. 

Here's some dmesg logs:

[ 3737.594116] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3737.594123] [drm] capturing error event; look for more information in
/sys/kernel/debug/dri/0/i915_error_state
[ 3737.606201] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x16358000 ctx 2) at 0x163587e4
[ 3759.646704] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3759.646763] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x283c6000 ctx 2) at 0x283c6718
[ 3777.660257] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3777.660316] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28486000 ctx 2) at 0x284861c8
[ 3802.679082] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3802.679152] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x16425000 ctx 2) at 0x164251c8
[ 3810.685111] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3810.685179] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x16506000 ctx 2) at 0x16506260
[ 3838.706195] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3838.706252] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28416000 ctx 2) at 0x284161c8
[ 3838.706256] [drm:i915_set_reset_status] *ERROR* render ring hung flushing bo
(0x7ee6000 ctx 2) at 0x284161c8
[ 3845.893990] hda-intel: IRQ timing workaround is activated for card #0.
Suggest a bigger bdl_pos_adj.
[ 3847.700964] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3847.701496] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x7ef0000 ctx 2) at 0x7ef01c8
[ 3863.725016] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3863.725060] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x170e0000 ctx 2) at 0x170e051c
[ 3863.738137] [drm:__gen6_gt_force_wake_get] *ERROR* Timed out waiting for
forcewake old ack to clear.
[ 3866.978023] hda-intel 0000:00:1b.0: Unstable LPIB (14104 >= 1764); disabling
LPIB delay counting
[ 3891.710082] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3891.710171] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28cec000 ctx 2) at 0x28cec51c
[ 3910.748409] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3910.748494] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28d60000 ctx 2) at 0x28d601c8
[ 3928.749952] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 3928.750016] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x7eb4000 ctx 2) at 0x7eb41c8
[ 4020.843251] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4020.843325] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x283f6000 ctx 2) at 0x283f6534
[ 4033.841035] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4033.841115] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28cdf000 ctx 2) at 0x28cdf51c
[ 4071.845631] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4071.845731] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28f79000 ctx 2) at 0x28f7951c
[ 4071.859475] [drm:__gen6_gt_force_wake_get] *ERROR* Timed out waiting for
forcewake old ack to clear.
[ 4134.917096] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4134.917173] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x28ed1000 ctx 2) at 0x28ed151c
[ 4165.928426] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4165.928481] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x283b6000 ctx 2) at 0x283b61c8
[ 4181.940477] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4181.940547] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x283a6000 ctx 2) at 0x283a6ae4
[ 4216.990852] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4216.990918] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x2af9b000 ctx 2) at 0x2af9b51c
[ 4226.998387] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4226.998463] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x2ae72000 ctx 2) at 0x2ae721c8
[ 4234.992398] [drm:i915_hangcheck_elapsed] *ERROR* stuck on render ring
[ 4234.992479] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo
(0x2afba000 ctx 2) at 0x2afba1c8

Here's the error state:

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are on the CC list for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20131018/feb1c38f/attachment-0001.html>


More information about the intel-gfx-bugs mailing list