Bug 45673

This is more clipped PolyLine fail. Not surprising as the original fb code is buggy in this regard as well :( On the plus side at least we detect it rather than corrupt random memory...

I'm pretty sure that this isn't the last bug here...

commit b7e3aaf773f05ce82405e135c0f99b40b3c2f434
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Mon Feb 6 18:08:19 2012 +0000

    sna: Use the clipped end-point for recomputing segment length after clipping
    
    References: https://bugs.freedesktop.org/show_bug.cgi?id=45673
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Created attachment 57303 [details]
debug log of the crash with debug full. (92mb unpacked)

(In reply to comment #2)
> I'm pretty sure that this isn't the last bug here...
So true.

I'm attaching another full-debug trace for (very very probably, there was an update pushed while I was reproducing this) xf86-video-intel revision 7d147c7462ba183f5316e7e8dafeca62add6f97c .

Additionally I'd like to note that when compiling the driver without debugging, although the xserver won't crash, it will stall for several seconds at a time, making it hard to work with the software in question. The following appears in dmesg (once for every 'freeze'):
---8<---------
Feb 20 10:50:45 localhost kernel: [28738.880739] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Feb 20 10:50:45 localhost kernel: [28738.880800] [drm] capturing error event; look for more information in /debug/dri/0/i915_error_state
Feb 20 10:50:45 localhost kernel: [28738.896369] [drm:i915_wait_request] *ERROR* i915_wait_request returns -11 (awaiting 122079 at 122075, next 122080)
Feb 20 10:50:53 localhost kernel: [28746.146411] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Feb 20 10:50:53 localhost kernel: [28746.146536] [drm:i915_wait_request] *ERROR* i915_wait_request returns -11 (awaiting 122934 at 122932, next 122935)
Feb 20 10:50:59 localhost kernel: [28752.996223] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Feb 20 10:50:59 localhost kernel: [28752.996348] [drm:i915_wait_request] *ERROR* i915_wait_request returns -11 (awaiting 123584 at 123579, next 123585)
--->8---------

Created attachment 57304 [details]
i915 error state captured after the crash (but after restarting X)

A couple more steps:

commit bbb6794a3b97b1fcf72c8712ab0ec591683b128b
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Mon Feb 20 12:25:31 2012 +0000

    sna: Trim clipped lines to end within bounds
    
    References: https://bugs.freedesktop.org/show_bug.cgi?id=46261
    References: https://bugs.freedesktop.org/show_bug.cgi?id=45673
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

commit 805bc3310cd0a13eab8e48e7615bdd42638cfa33
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Mon Feb 20 12:09:19 2012 +0000

    sna: When reversing line-drawing direction, use the clipped endpoint
    
    Make sure we take the clipping into account if we choose to reverse the
    draw direction (to maintain left-to-right inside the box emission).
    
    References: https://bugs.freedesktop.org/show_bug.cgi?id=46261
    References: https://bugs.freedesktop.org/show_bug.cgi?id=45673
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

And with

commit 1a65e2b8a2ebfb4d736efb7631515babad75faf2
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Tue Feb 21 08:09:52 2012 +0000

    sna: Split up/down edge walking in order to handle endpoint clipping
    
    In order to prevent walking upwards off the top of the pixmap when
    rendering a clipped vertical edge, we need to tweak the boundary
    conditions for the vertical edge walker.
    
    Reported-by: Clemens Eisserer <linuxhippy@gmail.com>
    References: https://bugs.freedesktop.org/show_bug.cgi?id=46261
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

I hereby proclaim everything to be perfect! ;)

Thanks Da Fox, please keep reporting any issues you find.