Bug 75291

Summary: [ILK Bisected]Piglit/shaders_glsl-bug-22603 fails
Product: Mesa Reporter: lu hua <huax.lu>
Component: Drivers/DRI/i965Assignee: Kenneth Graunke <kenneth>
Status: VERIFIED FIXED QA Contact: Intel 3D Bugs Mailing List <intel-3d-bugs>
Severity: major    
Priority: high CC: idr
Version: unspecified   
Hardware: All   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:

Description lu hua 2014-02-21 02:05:16 UTC 
System Environment:
--------------------------
Platform: Ironlake
Libdrm:		(master)libdrm-2.4.52-4-gc5de5abbd90333fe1359283fb3a5e457b0f389f3
Mesa:		(master)57405605a8c320f9d6ea389afd43ce6f013330a5
Xserver:	(master)xorg-server-1.15.0-627-gf34dc7fa96457ea6a0703493d74e63cca357712e
Xf86_video_intel:(master)2.99.910-47-gc91af569ee90a832899c9038badd84921e9a87fc
Cairo:		(master)4144307dbfbe7b297135d9ea4b080cae7e06b997
Libva:		(staging)fae9c44816a4c3cfc480d2879d1b4a0c1c3a1527
Libva_intel_driver:(staging)bd630edd844b88ea543a027654db296ff7da16cd
Kernel:	(drm-intel-nightly) 164a4cb4c1431a0689f85507868356fae24da638

Bug detailed description:
-------------------------
It fails on Ironlake with mesa master branch, works well on 10.1 branch.
Following piglit cases also fail with same bisect commit:
spec_ARB_depth_texture_fbo-depth-GL_DEPTH_COMPONENT24-drawpixels	
spec_ARB_depth_texture_fbo-depth-GL_DEPTH_COMPONENT32-drawpixels	
spec_EXT_framebuffer_object_fbo-alphatest-nocolor	
spec_EXT_framebuffer_object_fbo-alphatest-nocolor-ff	

Bisect shows: 09d9a8913e8c28fc4c1c60d7da85a2f093786894 is the first bad commit.
commit 09d9a8913e8c28fc4c1c60d7da85a2f093786894
Author:     Kenneth Graunke <kenneth@whitecape.org>
AuthorDate: Fri Feb 7 22:22:45 2014 -0800
Commit:     Kenneth Graunke <kenneth@whitecape.org>
CommitDate: Wed Feb 19 01:46:16 2014 -0800

    i965: Pull format conversion logic out of brw_depthbuffer_format.

    brw_depthbuffer_format is not very reusable at the moment, since it
    uses global state (ctx->DrawBuffer) to access a particular depth buffer.

    For HiZ on Broadwell, I need a function which simply converts the
    formats.  However, at least one existing user of brw_depthbuffer_format
    really wants the existing interface.  So, I've created a new function.

    Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
    Reviewed-by: Eric Anholt <eric@anholt.net>


output:
Probe color at (500,400)
  Expected: 0.300000 0.300000 0.300000
  Observed: 0.600000 0.600000 0.600000
Probe color at (220,300)
  Expected: 0.100000 0.100000 0.100000
  Observed: 0.800000 0.800000 0.800000
Probe color at (450,350)
  Expected: 0.300000 0.300000 0.300000
  Observed: 0.600000 0.600000 0.600000
Probe color at (270,350)
  Expected: 0.100000 0.100000 0.100000
  Observed: 0.729412 0.729412 0.729412
Probe color at (270,225)
  Expected: 0.010000 0.010000 0.010000
  Observed: 0.592157 0.592157 0.592157
Probe color at (450,250)
  Expected: 0.090000 0.090000 0.090000
  Observed: 0.600000 0.600000 0.600000
PIGLIT: {'result': 'fail' }


Reproduce steps:
-------------------------
1. xinit
2. ./bin/glsl-bug-22603 -fbo -auto
Comment 1 lu hua 2014-02-21 05:09:34 UTC 
Following cases also fail on Ironlake with same bisect commit:
spec_ARB_ES2_compatibility_FBO_blit_from_missing_attachment_(ES2_completeness_rules)	
spec_ARB_ES2_compatibility_FBO_blit_to_missing_attachment_(ES2_completeness_rules)	
spec_ARB_framebuffer_object_FBO_blit_from_missing_attachment	
spec_ARB_framebuffer_object_FBO_blit_to_missing_attachment	
spec_ARB_framebuffer_object_framebuffer-blit-levels_draw_depth	
spec_ARB_framebuffer_object_framebuffer-blit-levels_read_depth
Comment 2 Kenneth Graunke 2014-02-21 08:21:47 UTC 
Whoops, looks like I failed to rebase this properly after fixing the MESA_FORMAT X8 <-> S8 mixup.  Thanks for the report!

Patch on the mailing list:
http://lists.freedesktop.org/archives/mesa-dev/2014-February/054581.html
Comment 3 Kenneth Graunke 2014-02-24 02:11:19 UTC 
commit 82f9ad8c6011f5c1e8fd7065eeb6094f6e16dea9
Author: Kenneth Graunke <kenneth@whitecape.org>
Date:   Thu Feb 20 23:08:22 2014 -0800

    i965: Fix S8 and X8 reversal in brw_depthbuffer_format refactor.
    
    In commit 09d9a8913e8c28fc4c1c60d7da85a2f093786894, I accidentally
    botched the X8 and S8 cases.  (I wrote this patch before realizing that
    X8 and S8 had been swapped in the big MESA_FORMAT rename, and apparently
    didn't rebase it properly after fixing that...)
    
    Fixes regressions in 13 Piglit tests on Ironlake.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75291
    Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
    Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
    Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Comment 4 lu hua 2014-02-26 01:33:10 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.