Bug 67286 - [HSW]Piglit spec_OpenGL_1.1_read-front_samples=2 fails with compositer
Summary: [HSW]Piglit spec_OpenGL_1.1_read-front_samples=2 fails with compositer
Status: VERIFIED FIXED
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965 (show other bugs)
Version: unspecified
Hardware: All Linux (All)
: medium normal
Assignee: Ian Romanick
QA Contact: Intel 3D Bugs Mailing List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-07-25 02:41 UTC by lu hua
Modified: 2015-03-11 06:34 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:

Attachments

Description lu hua 2013-07-25 02:41:40 UTC 
System Environment:
--------------------------
Arch:         x86_64
Platform:     Haswell
Libdrm:		(master)libdrm-2.4.46-2-gfea5408098c3c3057958e85ea9d7146f0b08749e
Mesa:		(master)0acf3a8407fbeaa9d61731ce83fbe6de529e303c
Xserver:        (master)xorg-server-1.14.99.1-148-gd5ebe20f9ba9569351c4a41449866679fd60ba45
Xf86_video_intel:(master)2.21.12-46-g74dd10d252df2d826b6db5da7e53bd7c5d76dec0
Cairo:		(master)03c81d414d4edb710c91f96ddb7dbf73e5432583
Libva:		(staging)d4988142a3f2256e38c5c5cdcdfc1b4f5f3c1ea9
Libva_intel_driver:(staging)7783b945824c9c01b415ca7b72c2d9463c8ae40b
Kernel:	(drm-intel-nightly) ae5779ebae3a73ca4d18118a4b8600e4a36aaea7

Bug detailed description:
-------------------------
It fails with compositer, and works well with raw X.
It happens on haswell with mesa master branch and 9.2 branch. 

Following cases also has this issue:
spec_EGL_NOK_swap_region_basic	
spec_OpenGL_1.1_read-front		
spec_OpenGL_1.1_read-front_samples=4	
spec_OpenGL_1.1_read-front_samples=6	
spec_OpenGL_1.1_read-front_samples=8	

output:
Probe color at (0,0)
  Expected: 0.000000 0.000000 1.000000
  Observed: 0.078431 0.133333 0.219608
Probe color at (0,75)
  Expected: 0.000000 1.000000 0.000000
  Observed: 0.070588 0.125490 0.211765
PIGLIT: {'result': 'fail' }

Reproduce steps:
----------------------------
1. xinit
2. gnome-session
3. ./read-front -auto -samples=2
Comment 1 Iago Toral 2014-03-20 11:13:17 UTC 
This seems to be a duplicate of bug #74005. The patch I attached there for waffle seems to fix all the failing cases reported here too.
Comment 2 lu hua 2014-03-24 01:34:45 UTC 
It randomly fails on latest Mesa master branch.
./bin/read-front -samples=2 -auto 
Probe color at (0,0) 
Expected: 0.000000 0.000000 1.000000 
Observed: 0.003922 0.000000 0.133333 
Probe color at (0,75) 
Expected: 0.000000 1.000000 0.000000 
Observed: 0.003922 0.000000 0.062745 
PIGLIT: {'result': 'fail' }
Comment 3 Eric Anholt 2014-04-15 22:33:49 UTC 
The last few piglit runs since fixing the MCS problem haven't had any MSAA read-front failures for me.  This might be fixed.
Comment 4 Iago Toral 2015-03-10 08:45:47 UTC 
I can't reproduce this on Haswell, as Eric said, this is probably already fixed.
Comment 5 Matt Turner 2015-03-10 17:13:05 UTC 
lu hua, can you still reproduce?
Comment 6 lu hua 2015-03-11 06:34:16 UTC 
Run 30 cycles, it works well. Close it.
Comment 7 lu hua 2015-03-11 06:34:32 UTC 
Verified.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.