 Marcin Slusarz
|
bd35fe5a79
drm/nouveau: fix __nouveau_fence_wait performance
|
14 years ago |
 Ben Skeggs
|
d550c41e4f
drm/nouveau: remove no_vm/mappable flags from nouveau_bo
|
14 years ago |
|
Ben Skeggs
|
ec23802d61
drm/nv50: drop explicit yields in favour of smaller PFIFO timeslice
|
14 years ago |
|
Ben Skeggs
|
e3b7ed5e99
drm/nv84: use vm offsets for semaphores
|
14 years ago |
|
Ben Skeggs
|
fc772ec48d
drm/nv50: 0x50 needs semaphore yields too
|
14 years ago |
|
Ben Skeggs
|
cb1d771aa0
drm/nvc0: implement semaphores for inter-channel sync
|
14 years ago |
|
Ben Skeggs
|
c3b90a7d4c
drm/nv84: switch to new-style semaphores
|
14 years ago |
|
Ben Skeggs
|
2a55c9a7ff
drm/nvc0: reserve only subc 0 for kernel use
|
14 years ago |
|
Ben Skeggs
|
529c495912
drm/nvc0: implement fencing
|
14 years ago |
 Francisco Jerez
|
937c3471cc
drm/nouveau: Avoid potential race between nouveau_fence_update() and context takedown.
|
14 years ago |
|
Ben Skeggs
|
a8b214f007
drm/nouveau: fix use of drm_mm_node in semaphore object
|
14 years ago |
|
Ben Skeggs
|
ceac30999d
drm/nouveau: implicitly insert non-DMA objects into RAMHT
|
14 years ago |
|
Francisco Jerez
|
395a31ec7e
drm/nouveau: Spin for a bit in nouveau_fence_wait() before yielding the CPU.
|
14 years ago |
|
Ben Skeggs
|
7f4a195fcb
drm/nouveau: tidy up and extend dma object creation interfaces
|
14 years ago |
|
Marcin Slusarz
|
382d62e524
drm/nouveau: fix annoying nouveau_fence type issue
|
14 years ago |
|
Ben Skeggs
|
a6a1a38075
drm/nouveau: use object class structs more extensively
|
14 years ago |
|
Francisco Jerez
|
889fa93dc0
drm/nouveau: Take fence spinlock in nouveau_fence_channel_fini().
|
14 years ago |
|
Francisco Jerez
|
2b478addc0
drm/nouveau: Avoid race in the interchannel sync code.
|
14 years ago |
|
Francisco Jerez
|
2a6789ae5e
drm/nouveau: Make fences take a weak channel reference.
|
14 years ago |
|
Francisco Jerez
|
fcccab2e4e
drm/nouveau: Use lazy fence waits when doing software interchannel sync.
|
14 years ago |
|
Ben Skeggs
|
08cd3d4311
drm/nouveau: fix thinko in channel locking in semaphore path
|
14 years ago |
|
Ben Skeggs
|
cff5c13324
drm/nouveau: add more fine-grained locking to channel list + structures
|
14 years ago |
|
Francisco Jerez
|
907af60b93
drm/nouveau: Fix sleep while atomic in the semaphore code.
|
14 years ago |
|
Francisco Jerez
|
8af29ccd79
drm/nv50: Fix large 3D performance regression caused by the interchannel sync patches.
|
14 years ago |
|
Francisco Jerez
|
0c6c1c2fb8
drm/nouveau: Use semaphores to handle inter-channel sync in hardware.
|
14 years ago |
|
Francisco Jerez
|
8ac3891b48
drm/nouveau: Provide a means to have arbitrary work run on fence completion.
|
14 years ago |
|
Francisco Jerez
|
2730723bbc
drm/nouveau: Minor refactoring/cleanup of the fence code.
|
14 years ago |
|
Francisco Jerez
|
3ba6462355
drm/nouveau: Take fence spinlock before reading the last sequence.
|
15 years ago |
 Kulikov Vasiliy
|
05991110cf
drm/nouveau: set TASK_(UN)INTERRUPTIBLE before schedule_timeout()
|
15 years ago |
|
Ben Skeggs
|
047d1d3cae
drm/nouveau: reduce usage of fence spinlock to when absolutely necessary
|
15 years ago |
