[Intel-gfx] Fencing on the i915

Chris Wilson chris at chris-wilson.co.uk
Mon Feb 9 11:54:37 CET 2009


Hello all,
I've spent a bit of time trying to improve the behaviour of my i915 when
using tiled buffers and fencing. I believe I've found a few critical
issues in the code, but alas I can still cause the chip to hang if I
write to the fence registers too frequently.

I've attempted to break down my patches into logically separate series -
debugging, error handling fixes, some miscellaneous and finally changes
to i915_gem_object_get_fence_reg(). The first set are attached, the
latter I will attach in reply as I finish tidying them up.

Thanks.
-ickle
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-drm-i915-Show-buffer-status-in-proc.patch
Type: text/x-patch
Size: 3809 bytes
Desc: 
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20090209/bca514c8/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0002-drm-i915-Display-fence-registers-in-proc.patch
Type: text/x-patch
Size: 2785 bytes
Desc: 
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20090209/bca514c8/attachment-0001.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0003-drm-i915-Show-ringbuffer-details-in-proc.patch
Type: text/x-patch
Size: 2584 bytes
Desc: 
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20090209/bca514c8/attachment-0002.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0004-drm-i915-Increase-debugging-detail-for-LRU.patch
Type: text/x-patch
Size: 3466 bytes
Desc: 
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20090209/bca514c8/attachment-0003.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0005-drm-i915-Add-WATCH_FENCE.patch
Type: text/x-patch
Size: 5296 bytes
Desc: 
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20090209/bca514c8/attachment-0004.bin>


More information about the Intel-gfx mailing list