summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2013-03-11convert another lock to i915_mutex_lock_interruptibleJonathan Gray
2013-03-11handle error completions the same way as FreeBSDJonathan Gray
2013-03-10bring i915_gem_object_bind_to_gtt closer to Linux 3.8.2Mark Kettenis
2013-03-10some random bits reducing diffs with linuxMark Kettenis
2013-03-10move inteldrm_purge_obj to i915_gem.c and rename to i915_gem_object_truncateMark Kettenis
2013-03-10Revert f7940bd8bd7da9bbfd38e41d99edb378dd9b4df7. Now that we we rebindMark Kettenis
objects after resume we can reinitialize the GATT again.
2013-03-10Make i915_gem_gtt_rebind_object() less stupid and move it into i915_gem_gtt.c.Mark Kettenis
Use it to properly implement i915_gem_resetore_gtt_mappings().
2013-03-09Adjust timeout loops in the reset functions to match the 500ms timeout thatMark Kettenis
Linux uses.
2013-03-09Add agp_bus_dma_rebind(), and interface to re-enter mappings into the GART/GTTMark Kettenis
when we resume or want to change for example the cachability of the mapping.
2013-03-09use i915_mutex_lock_interruptible() where appropriateMark Kettenis
2013-03-09partly revert b1eebaa3c34e225793793b38f963d94c4e14f961 as it breaksMark Kettenis
suspend/resume on my x220. See https://bugzilla.kernel.org/show_bug.cgi?id=47941
2013-03-09Do the proper callback dance when switching screensMark Kettenis
2013-03-09Enable multiple screensMark Kettenis
Since proper VT support is still lacking this will make VTs share a single framebuffer in emulation mode. As a result the display will be messed up a bit if you run anything on VTs 1-8. This can be mitigated by disabling the apropriate lines in /etc/ttys. This does allow VT switching between X and the framebuffer console and makes suspend/resume do the right thing when in X.
2013-03-09Enable framebuffer consoleMark Kettenis
2013-03-09handle llc and semaphore paramsJonathan Gray
2013-03-09add some more paramsJonathan Gray
2013-03-09add support for setting the addressing modeJonathan Gray
2013-03-09In intel_user_framebuffer_create_handle(), add a reference to the object sinceMark Kettenis
drm_handle_create() consumes a reference unlike drm_gem_handle_create() which current Linux uses.
2013-03-09inteldrm_chipset_flush -> i915_gem_chipset_flushJonathan Gray
2013-03-09move the retire work handler to i915_gem.cJonathan Gray
2013-03-08Implement i915_mutex_lock_interruptibleMark Kettenis
2013-03-08i915_gem_object_move_to_inactive_locked needs to clear the write domainJonathan Gray
2013-03-08move i915_gem_object_set_to_{cpu,gtt}_domain closer to 3.8.2Jonathan Gray
2013-03-08don't clear sync_seqno in i915_gem_retire_requests_ringJonathan Gray
2013-03-08match 3.8.2 and preallocate the next seqno before touching the ringJonathan Gray
2013-03-08use i915_gem_check_olr in a few more placesJonathan Gray
2013-03-08seqno wrapping will be handled elsewhere soonJonathan Gray
2013-03-08remove some now unused codeJonathan Gray
2013-03-08move the domain/flush code closer to 3.8.2Jonathan Gray
2013-03-08remove flushing_listJonathan Gray
2013-03-08remove some no longer needed flush codeJonathan Gray
2013-03-08remove gpu_write_listJonathan Gray
2013-03-08decrement of the seqno for semaphore ring sync moved into the ring codeJonathan Gray
2013-03-07missed a file in a previous commitJonathan Gray
2013-03-07sync the shared modesetting code with 3.8.2Jonathan Gray
2013-03-07sync intel modesetting code with 3.8.2Jonathan Gray
2013-03-06remove the now unused dispatch wrapperJonathan Gray
2013-03-06directly call the dispatch execbuffer callbackJonathan Gray
2013-03-06our request retiring is still done inside wait_seqno, drop surplus callsJonathan Gray
2013-03-06expand locking in wait_seqno to include the first get_seqno callJonathan Gray
2013-03-06disable powersaving as all the code is not ported yetJonathan Gray
2013-03-06convert some commented mdelay callsJonathan Gray
2013-03-06drop the force wake delays back down to their original valuesJonathan Gray
2013-03-06unconditionally do the workaround to force correct ordering on gen6+Jonathan Gray
seems to fix the laggyness with firefox (but not the hangs)
2013-03-06add a few more i915_gem_next_request_seqno calls 3.4.33 makesJonathan Gray
2013-03-06enable using gpu semaphores for ring syncJonathan Gray
2013-03-06add some missing locking in the throttle funcJonathan Gray
2013-03-05unstub the ring throttle ioctlJonathan Gray
2013-03-04move i915_gem_fault a bit closer to Linux 3.4.33Mark Kettenis
2013-03-05claim to support page flippingJonathan Gray
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-21 22:01:38 +0000