Bug 17196

Summary:	[GEM] a lot of glean cases run failed
Product:	Mesa	Reporter:	Shuang He <shuang.he>
Component:	Drivers/DRI/i915	Assignee:	Eric Anholt <eric>
Status:	VERIFIED INVALID	QA Contact:
Severity:	normal
Priority:	high	CC:	dri-devel
Version:	7.1
Hardware:	Other
OS:	Linux (All)
Whiteboard:
i915 platform:		i915 features:
Attachments:	xorg conf xorg log

Description Shuang He 2008-08-18 22:31:20 UTC

System Environment:
--------------------------

--Platform: 945G Q965

--Architecture(32-bit,64-bit,compatiblity):32bit 

Kernel: anholt's linux-2.6 tree

Libdrm: a5381cac55e54a535acf752970886b659948563c
Mesa:   c5d25c85093f089cc58c520785844fc021a09612
Xserver:99583b43a9a202d047ff417d47485e4c0e0c9670
2D driver:76b5a7ff5c943cfaa254d2fd94b6397c91d1f8e6






Bug detailed description:
-------------------------
startx ,run glean cases, 
for Q965, it get pass/fail=17/19
For 945G, it get pass/fail=17/21
They both have these cases failed: makeCurrent,blendFunc,logicOp,maskedClear,pixelFormats,readPixSanity,scissor,texCombine, texCube,texEnv,texgen,texRect,vertProg1,
945G has extra 2 failures: fexceptions,fragProg1
Q965 has extra 2 failures: paths,pointSprite




Current result:
----------------
a lot of cases run failed 


Expected result:
----------------
run normally

Comment 1 Shuang He 2008-08-18 22:33:27 UTC

Created attachment 18369 [details]
xorg conf

Comment 2 Shuang He 2008-08-18 22:34:06 UTC

Created attachment 18370 [details]
xorg log

Comment 3 Eric Anholt 2008-08-25 23:16:16 UTC

Fixed an issue with G965 in 0793456076f8356f1b7fb99e0b76b8e3d1eeb74a in the kernel that would have manifested in drawpix failing.

Also, please include kernel sha1s, as I'm having to push history rewrites as part of the kernel tree being for merging upstream.

Comment 4 Shuang He 2008-08-26 00:54:12 UTC

(In reply to comment #3)
> Fixed an issue with G965 in 0793456076f8356f1b7fb99e0b76b8e3d1eeb74a in the
> kernel that would have manifested in drawpix failing.
> 
> Also, please include kernel sha1s, as I'm having to push history rewrites as
> part of the kernel tree being for merging upstream.
> 

It doesn't change much with kernel commit 0793456076f8356f1b7fb99e0b76b8e3d1eeb74a.
Q965 pass/fail=17/21
945G pass/fail=16/22, test case 'paths' failed this time

Comment 5 Shuang He 2008-09-01 00:30:46 UTC

(In reply to comment #4)
> (In reply to comment #3)
> > Fixed an issue with G965 in 0793456076f8356f1b7fb99e0b76b8e3d1eeb74a in the
> > kernel that would have manifested in drawpix failing.
> > 
> > Also, please include kernel sha1s, as I'm having to push history rewrites as
> > part of the kernel tree being for merging upstream.
> > 
> 
> It doesn't change much with kernel commit
> 0793456076f8356f1b7fb99e0b76b8e3d1eeb74a.
> Q965 pass/fail=17/21
> 945G pass/fail=16/22, test case 'paths' failed this time
> 
This is fixed with latest anholt's linux-2.6 commit ef8556545627f57c1d69fdfcb2d90805cdb9a2c3
Now, we get 
Q965 pass/fail=32/9
945G pass/fail=31/10

Comment 6 Shuang He 2008-09-01 00:31:07 UTC

Verified, Thanks

Comment 7 Shuang He 2008-09-05 00:07:45 UTC

We still get a lot of failures on G33, pass/fail=15/21.

Comment 8 Eric Anholt 2008-09-10 15:27:52 UTC

Please don't repurpose a bug in comment #7.

File a new bug with the full information: logs, sha1s, readpix results, etc.

Comment 9 Shuang He 2008-09-10 19:09:27 UTC

(In reply to comment #8)
> Please don't repurpose a bug in comment #7.
> 
> File a new bug with the full information: logs, sha1s, readpix results, etc.
> 

ok, 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.