Bug 44146

Summary: [bisected SNB]Oglc pxtrans-draw(advanced.stencilTest) regressed
Product: Mesa Reporter: fangxun <xunx.fang>
Component: Drivers/DRI/i965Assignee: Ian Romanick <idr>
Status: VERIFIED FIXED QA Contact:
Severity: normal    
Priority: medium CC: brianp
Version: git   
Hardware: All   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Bug Depends on:    
Bug Blocks: 42993    

Description fangxun 2011-12-25 22:37:20 UTC 
System Environment:
--------------------------
Arch:           i386
Platform:       huronriver
Libdrm:         (master)2.4.29-3-gef20301a11afae50bfe127002913dbd0b81ddccc
Mesa:           (master)ec4851253bbf7fd7d11c5570f19f9733a885e471
Xf86_video_intel:(master)2.17.0-228-gb117f65520919f4ba36010cfe913a8c53166bf23
Xserver:     (master)xorg-server-1.11.99.2-16-gfcda98c48610fd507ca0b89c6006a5497d9dc1c9
Kernel:  (drm-intel-next)097354eb14fa94d31a09c64d640643f58e4a5a9a

Bug detailed description:
------------------------- 
It regressed from pass to fail on sandybridge. 
Bisect shows e66858fb67477db22139fe38d4f23530b7a142a7 is the first bad commit.
commit e66858fb67477db22139fe38d4f23530b7a142a7
Author:     Brian Paul <brianp@vmware.com>
AuthorDate: Sat Dec 24 08:54:26 2011 -0700
Commit:     Brian Paul <brianp@vmware.com>
CommitDate: Sat Dec 24 08:59:54 2011 -0700

    swrast: move swrast_render_start/finish() call in drawpixels code

    We don't want to call these functions where we'll be using
    Map/UnmapRenderbuffer().  So push them further down in the drawpixels
    cases so that we can switch over to Map/UnmapRenderbuffer() step by step.

    Reviewed-by: Eric Anholt <eric@anholt.net>

Reproduce steps:
----------------
1. start X
2. ./oglconform -z -s -suite all -v 2 -test pxtrans-draw advanced.stencilTest
Comment 1 Ian Romanick 2012-02-01 08:37:51 UTC 
With current master (f09910f) and current 8.0 (7f5d3f7), I get the same results as with e66858f^.  Note that the test fails on all visuals without alpha.
Comment 2 fangxun 2012-02-07 02:33:22 UTC 
It still fails on SNB with current mesa master and 8.0 tip.
Comment 3 Ian Romanick 2012-02-07 10:32:02 UTC 
Re-read my last comment.  It fails now in the same what that it failed before the bisect.  We've previously determined that the alpha problem is a bug in the test.  Do not reopen this bug.
Comment 4 Gordon Jin 2012-04-01 00:33:33 UTC 
I'm confused here.

Xun, can you confirm it fails as before?

Ian, do you have follow-up with the test issue?
Comment 5 fangxun 2012-04-01 01:06:40 UTC 
It passes now. Bisect shows it is fixed by following commit.
commit e1f9820b47e3f124c49cd2ab4e09328e0cc3e638
Author: Chad Versace 
Date: Thu Jan 26 11:01:36 2012 -0800

i965: Rewrite the HiZ op

This is a combination of 4 commits. The first commit rewrites the HiZ op,
and remaining three fix bugs introduced by the rewrite.

======== commit 1 ========

i965: Rewrite the HiZ op

The HiZ op was implemented as a meta-op. This patch reimplements it by
emitting a special HiZ batch. This fixes several known bugs, and likely

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.