drm/radeon: Avoid double gpu reset by adding a timeout on IB ring tests.
authorMatthew Dawson <matthew@mjdsystems.ca>
Sun, 7 Feb 2016 21:51:12 +0000 (16:51 -0500)
committerAlex Deucher <alexander.deucher@amd.com>
Wed, 10 Feb 2016 19:17:15 +0000 (14:17 -0500)
commit04db4caf5c836c211977a54c9218f2cdee14897f
tree99460f045362a3d7bae6b9d1c64843f85cd19107
parent6e9821b26dd45b4a3e30c4219be2abd802563379
drm/radeon: Avoid double gpu reset by adding a timeout on IB ring tests.

When the radeon driver resets a gpu, it attempts to test whether all the
rings can successfully handle an IB.  If these rings fail to respond, the
process will wait forever.  Another gpu reset can't happen at this point,
as the current reset holds a lock required to do so.  Instead, make all
the IB tests run with a timeout, so the system can attempt to recover
in this case.

While this doesn't fix the underlying issue with card resets failing, it
gives the system a higher chance of recovering.  These timeouts have been
confirmed to help both a Tathi and Hawaii card recover after a gpu reset.

This also adds a new function, radeon_fence_wait_timeout, that behaves like
fence_wait_timeout.  It is used instead of fence_wait_timeout as it continues
to work during a reset.  radeon_fence_wait is changed to be implemented
using this function.

V2:
 - Changed the timeout to 1s, as the default 10s from radeon_wait_timeout was
too long.  A timeout of 100ms was tested and found to be too short.
 - Changed radeon_fence_wait_timeout to behave more like fence_wait_timeout.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Matthew Dawson <matthew@mjdsystems.ca>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/radeon/cik.c
drivers/gpu/drm/radeon/cik_sdma.c
drivers/gpu/drm/radeon/r100.c
drivers/gpu/drm/radeon/r600.c
drivers/gpu/drm/radeon/r600_dma.c
drivers/gpu/drm/radeon/radeon.h
drivers/gpu/drm/radeon/radeon_fence.c
drivers/gpu/drm/radeon/radeon_vce.c
drivers/gpu/drm/radeon/uvd_v1_0.c