Bug 38762 - [IVB bisected]3D demos like glxgears abort
[IVB bisected]3D demos like glxgears abort
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965
x86 (IA32) Linux (All)
: highest blocker
Assigned To: Kenneth Graunke
Depends on:
  Show dependency treegraph
Reported: 2011-06-28 23:49 UTC by meng
Modified: 2011-06-29 17:36 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:

Xorg.0.log (62.41 KB, text/plain)
2011-06-28 23:49 UTC, meng

Note You need to log in before you can comment on or make changes to this bug.
Description meng 2011-06-28 23:49:40 UTC
Created attachment 48540 [details]

System Environment:
Platform: IvyBridge
Mesa:     (7.11)bd40e7cebd94a6ec68b9f3a73e6d62d95bb4a5a8
Xserver:  (server-1.10-branch)xorg-server-
Cairo:    (master)ea645913ba8739377ee2e2b51480310befc19b76
Libva:    (master)5b5dfd0e196c53331fde8a21e942619ab5237a9a
Kernel:   (drm-intel-fixes) f01c22fd59aa10a3738ede20fd4b9b6fd1e2eac3

Bug detailed description:
3D demos like glxgears abort on IvyBridge.And compiz can't work in gnome-session.It's mesa regression.By bisected,show that c173541d9769d41a85cc899bc49699a3587df4bf
is first bad commit.Pls see Xorg.log.
In console when running glxgears:
Running synchronized to the vertical refresh.  The framerate should be
approximately the same as the monitor refresh rate.
glxgears: gen7_urb.c:81: upload_urb: Assertion `brw->gs.prog_active' failed.
Aborted (core dumped)

commit c173541d9769d41a85cc899bc49699a3587df4bf
Author: Eric Anholt <eric@anholt.net>
Date:   Wed Apr 27 13:33:10 2011 -0700

    i965: Use state streaming on programs, and state base address on gen5+.

    There will be a little bit of thrashing of the program cache BO as the
    cache warms up, but once the application is in steady state, this
    reduces relocations on gen5 and later.

    On my T420 laptop, cairogl firefox-talos-gfx performance improves 2.6%
    +/- 1.3% (n=6).  No statistically significant performance difference
    on nexuiz (n=5).

Reproduce steps:
Comment 1 Gordon Jin 2011-06-29 00:26:00 UTC
No problem on SNB or Ironlake?
Comment 2 meng 2011-06-29 01:10:33 UTC
SNB and Ironlake is OK.
Comment 3 Kenneth Graunke 2011-06-29 11:02:27 UTC
Should be fixed on master by the following commit:

commit 5ddc518401ae69ad92218643f00ef50617a0f11d
Author: Kenneth Graunke <kenneth@whitecape.org>
Date:   Wed Jun 29 10:53:51 2011 -0700

    i965/gen7: Add missing ! to brw->gs.prog_active assertion.
    A typo in commit c173541d9769 accidentally removed the !.
    It's supposed to assert that there is _not_ an active GS program.
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38762
    Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>

Also cherry-picked to 7.11 as a8d7f36d6598da452c9724d41b74feff4a0cdd88.  Normally I'd let it sit, but this was obviously wrong, trivial, and for unreleased hardware and software anyway.
Comment 4 meng 2011-06-29 17:36:29 UTC
Verified with 5ddc518401ae69ad92218643f00ef50617a0f11d,it works fine.