summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/nouveau/nouveau_fence.c
Commit message (Expand)AuthorAgeFilesLines
* drm/nouveau/fence: fix a race where fence->channel can disappearBen Skeggs2012-05-241-5/+9
* drm/nouveau/fence: convert to exec engine, and improve channel syncBen Skeggs2012-05-241-412/+62
* drm/nouveau/fence: minor api changes for an upcoming reworkBen Skeggs2012-05-241-130/+85
* drm/nouveau/fence: make ttm interfaces wrap ours, not the other way aroundBen Skeggs2012-05-241-23/+13
* drm/nouveau: create real execution engine for software object classBen Skeggs2012-05-241-5/+1
* drm/nouveau: base fence timeout on time of emissionMarcin Slusarz2012-05-241-1/+4
* drm/nouveau: use the same packet header macros as userspaceBen Skeggs2012-05-241-13/+13
* nouveau: add PRIME supportDave Airlie2012-05-231-1/+1
* drm/nouveau: move fence sequence check to start of loopBen Skeggs2012-03-221-6/+5
* drm/nouveau: remove subchannel names from places where it doesn't matterBen Skeggs2012-03-221-12/+12
* drm/nouveau: initialize chan->fence.lock before useMarcin Slusarz2011-11-101-2/+0
* drm/nouveau: workaround semaphore hw bug causing unnecessary interruptsBen Skeggs2011-09-201-1/+1
* drm/nouveau: fix oops on pre-semaphore hardwareBen Skeggs2011-09-091-1/+2
* drm/nouveau: remove 'chan' argument from nouveau_bo_newBen Skeggs2011-06-231-1/+1
* drm/nv84-nvc0: explicitly map semaphore buffer into channel vmBen Skeggs2011-06-231-16/+13
* drm/nouveau: fix assumption that semaphore dmaobj is valid in x-chan syncBen Skeggs2011-06-181-34/+25
* drm/nouveau: fix __nouveau_fence_wait performanceMarcin Slusarz2011-03-141-3/+12
* drm/nouveau: remove no_vm/mappable flags from nouveau_boBen Skeggs2011-02-251-1/+1
* drm/nv50: drop explicit yields in favour of smaller PFIFO timesliceBen Skeggs2011-02-251-46/+6
* drm/nv84: use vm offsets for semaphoresBen Skeggs2011-02-251-5/+17
* drm/nv50: 0x50 needs semaphore yields tooBen Skeggs2011-02-251-4/+17
* drm/nvc0: implement semaphores for inter-channel syncBen Skeggs2011-02-251-12/+40
* drm/nv84: switch to new-style semaphoresBen Skeggs2011-02-251-39/+80
* drm/nvc0: reserve only subc 0 for kernel useBen Skeggs2010-12-301-1/+1
* drm/nvc0: implement fencingBen Skeggs2010-12-211-8/+20
* drm/nouveau: Avoid potential race between nouveau_fence_update() and context ...Francisco Jerez2010-12-081-8/+11
* drm/nouveau: fix use of drm_mm_node in semaphore objectBen Skeggs2010-12-081-3/+2
* drm/nouveau: implicitly insert non-DMA objects into RAMHTBen Skeggs2010-12-081-6/+1
* drm/nouveau: Spin for a bit in nouveau_fence_wait() before yielding the CPU.Francisco Jerez2010-12-081-1/+2
* drm/nouveau: tidy up and extend dma object creation interfacesBen Skeggs2010-12-031-2/+2
* drm/nouveau: fix annoying nouveau_fence type issueMarcin Slusarz2010-12-031-11/+11
* drm/nouveau: use object class structs more extensivelyBen Skeggs2010-12-031-1/+1
* drm/nouveau: Take fence spinlock in nouveau_fence_channel_fini().Francisco Jerez2010-12-031-0/+4
* drm/nouveau: Avoid race in the interchannel sync code.Francisco Jerez2010-12-031-9/+15
* drm/nouveau: Make fences take a weak channel reference.Francisco Jerez2010-12-031-1/+2
* drm/nouveau: Use lazy fence waits when doing software interchannel sync.Francisco Jerez2010-12-031-2/+2
* drm/nouveau: fix thinko in channel locking in semaphore pathBen Skeggs2010-12-031-3/+3
* drm/nouveau: add more fine-grained locking to channel list + structuresBen Skeggs2010-12-031-0/+10
* drm/nouveau: Fix sleep while atomic in the semaphore code.Francisco Jerez2010-11-181-1/+6
* drm/nv50: Fix large 3D performance regression caused by the interchannel sync...Francisco Jerez2010-10-051-7/+39
* drm/nouveau: Use semaphores to handle inter-channel sync in hardware.Francisco Jerez2010-10-051-1/+194
* drm/nouveau: Provide a means to have arbitrary work run on fence completion.Francisco Jerez2010-10-051-0/+30
* drm/nouveau: Minor refactoring/cleanup of the fence code.Francisco Jerez2010-10-051-17/+46
* drm/nouveau: Take fence spinlock before reading the last sequence.Francisco Jerez2010-09-031-2/+4
* drm/nouveau: set TASK_(UN)INTERRUPTIBLE before schedule_timeout()Kulikov Vasiliy2010-08-021-2/+2
* drm/nouveau: reduce usage of fence spinlock to when absolutely necessaryBen Skeggs2010-07-131-25/+6
* drm/nouveau: Use the software object for fencing.Francisco Jerez2010-01-111-1/+1
* drm/nouveau: remove use of -ERESTARTBen Skeggs2009-12-151-1/+1
* drm/nouveau: Add DRM driver for NVIDIA GPUsBen Skeggs2009-12-111-0/+262