Bug 51429 - [IVB SNB regression]Piglit glean/api2 fails
Summary: [IVB SNB regression]Piglit glean/api2 fails
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: All Linux (All)
: high major
Assignee: Daniel Vetter
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-06-25 20:46 UTC by lu hua
Modified: 2017-10-06 14:49 UTC (History)
5 users (show)

See Also:
i915 platform:
i915 features:


Attachments
output (3.37 KB, text/plain)
2012-06-25 20:46 UTC, lu hua
no flags Details
add CS_STALL to render cache flush (2.50 KB, patch)
2012-06-26 03:40 UTC, Daniel Vetter
no flags Details | Splinter Review

Description lu hua 2012-06-25 20:46:19 UTC
Created attachment 63473 [details]
output

System Environment:
--------------------------
Arch:             i386
Platform:         Sandybridge
Libdrm:		(master)libdrm-2.4.35-12-gae137f4669ccdbc615d18facebdb804a9af9846b
Mesa:		(master)37d699a296ac1d63b9276224847df4b645b53fe2
Xserver:	(master)xorg-server-1.12.0-239-g3ef3ce069d52dcfa932c90ccd30854a8d9daa15a
Xf86_video_intel:(master)2.19.0-334-gfa10005ce31483827547b7f71eae066899f0026c
Libva:		(staging)f12f80371fb534e6bbf248586b3c17c298a31f4e
Libva_intel_driver:(staging)82fa52510a37ab645daaa3bb7091ff5096a20d0b
Kernel:	(drm-intel-next-queued) ff049b6ce21d2814451afd4a116d001712e0116b

Bug detailed description:
-----------------------------
It fails on ivybridge and sandybridge with drm-intel-next-queued kernel.It doesn't happen on drm-intel-fixes kernel.
The last known good commit: e3f33d46fd917747e966f8e6d25f2940223ad1ee
The last known bad commit: ff049b6ce21d2814451afd4a116d001712e0116b


Reproduce steps:
----------------
1. start x
2. ./bin/glean -r /tmp/glean/api2 -o -v -v -v -t +api2
Comment 1 Daniel Vetter 2012-06-26 03:40:17 UTC
Created attachment 63481 [details] [review]
add CS_STALL to render cache flush

Please test this patch.
Comment 2 lu hua 2012-06-26 23:06:55 UTC
commit cc889e0f6ce6a63c62db17d702ecfed86d58083f is the first bad commit.
commit cc889e0f6ce6a63c62db17d702ecfed86d58083f
Author:     Daniel Vetter <daniel.vetter@ffwll.ch>
AuthorDate: Wed Jun 13 20:45:19 2012 +0200
Commit:     Daniel Vetter <daniel.vetter@ffwll.ch>
CommitDate: Wed Jun 20 13:54:28 2012 +0200


Many Oglconform_31, Piglit and Ogles2conform cases also fail with this commit.

Add CS_STALL patch, this issue still exists.
Comment 3 Daniel Vetter 2012-06-27 02:38:12 UTC
(In reply to comment #2)
> Add CS_STALL patch, this issue still exists.

This is strange, I've retested this on both my ivb and snb machines today again, and without that patch the latest dinq kernel fails the test. With the patch it works rock-solid (i.e. I've run it in a loop a few hundred times, no failure).

I've pushed out the exact kernel I've tested this on to the for-QA branch in the drm-intel repository.

Can you please double-check whether this glean test really still fails for you on both ivb/snb on this kernel?
Comment 4 lu hua 2012-06-28 00:21:39 UTC
Retest this on the queued kernel (commit:3e52259264832dc8c) with SC_STALL patch.
Issue goes away.
Comment 5 Daniel Vetter 2012-06-28 12:08:54 UTC
Fix merged to -queued as

commit 97f209bcfc0c5db08d9badf8cbafd489f22a6e44
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Jun 28 09:48:42 2012 +0200

    drm/i915: "Flush Me Harder" required on gen6+
    

Please check whether this really fixes all issues introduced with cc889e0f6ce6a63c62db17d702ecfed86d58083f
Comment 6 lu hua 2012-07-01 23:47:15 UTC
Verified.It has been fixed on drm-intel-next-queued kernel(commit 146937e5828ede495e11ba3a6f4a01b36b7166dc)
Comment 7 Elizabeth 2017-10-06 14:49:29 UTC
Closing old verified.


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.