Summary: | [945GM bisected] unredrawn borders | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | Carlo Segre <segre> | ||||||
Component: | Driver/intel | Assignee: | Carl Worth <cworth> | ||||||
Status: | RESOLVED FIXED | QA Contact: | Xorg Project Team <xorg-team> | ||||||
Severity: | major | ||||||||
Priority: | medium | CC: | ejsheldrake, kevmitch | ||||||
Version: | unspecified | ||||||||
Hardware: | x86-64 (AMD64) | ||||||||
OS: | Linux (All) | ||||||||
Whiteboard: | |||||||||
i915 platform: | i915 features: | ||||||||
Bug Depends on: | |||||||||
Bug Blocks: | 37233 | ||||||||
Attachments: |
|
Description
Carlo Segre
2010-09-30 21:22:58 UTC
Created attachment 39095 [details]
example of xmgrace self-corruption
Looks like an off-by-one computation in the damage region. Can you try the most recent xserver? (What's in experimental currently?) Hmm, I thought I heard such a bug discussed but can't see any pertinent commits. In which case the issue may be the driver not causing damage for itself (or rather the xserver failing to wrap an entry point and track damage...). Anyway can you verify that the current xserver is similarly broken? Does xfce4 wm perform compositing? Does that make a difference? I'm using openbox wm with no composting. I have bisected it down to the following pair of consecutive commits. My test case is moving a qiv window displaying an image over top of a lucid emacs window. I cannot perform this test on the earlier of the two commits listed because opening emacs crashes the xserver in this case. commit a2aa4c23f65b03c389c2ba8eac0298dfb205e67d Author: Dave Airlie <airlied@redhat.com> Date: Mon Jul 5 14:02:42 2010 +1000 uxa: oops typo in previous commit commit feff2ec80eeeba5074d19efcddf4867adf639b4f Author: Dave Airlie <airlied@redhat.com> Date: Mon Jul 5 09:07:08 2010 +1000 uxa: don't compare planemask with FB_ALLONES. planemask is an unsigned long initialised to ~0, on 64-bit this is not equal to an (unsigned int)-1. Use the macro provided to do this. Signed-off-by: Dave Airlie <airlied@redhat.com> Now that was a result I wasn't expecting. Hmm. I'm seeing the same kind of problem, but with nedit and xpdf. However, nedit rebuilt against current openmotif (rather than lesstif) doesn't display this issue. I'm using: GM45 chipset / x86_64 arch xorg-x11-drv-intel 2.13.0 xorg-x11-server-Xorg 1.9.1-2.fc14 mesa 7.9 libdrm 2.4.22 kernel 2.6.36 Just to be sure, can you retest with commit e0066e77e026b0dd0daa0c3765473c7d63aa6753 Author: Eric Anholt <eric@anholt.net> Date: Tue May 31 23:17:16 2011 -0700 uxa: Simplify Composite solid acceleration for spans by only clipping once. Actually, I'm going to close this one fixed, because I'm confident enough based on the original report that the bug I deleted was the one described here. Great, that fixes it. Thanks for all the hard work! |
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.