diff options
| author | Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> | 2026-03-10 16:33:44 -0400 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2026-03-23 14:15:42 -0400 |
| commit | 3cf496080f539b69f069027cece82642b46a6cb5 (patch) | |
| tree | 0331696c0438ad4693d3c5e3e61ac0edb98786e0 /tools/perf/scripts/python/task-analyzer.py | |
| parent | 7d59465de38e8cbebf55f122b0514de10f52852c (diff) | |
drm/amd/display: Revert inbox0 lock for cursor due to deadlock
[Why]
A deadlock occurs when using inbox0 lock for cursor operations on
PSR-SU and Replays that does not when using the inbox1 locking path.
This is because of a priority inversion issue where inbox1 work
cannot be serviced while holding the HW lock from driver and sending
cursor notifications to DMUB.
Typically the lower priority of inbox1 for the lock command would
allow the PSR and Replay FSMs to complete their transition prior
to giving driver the lock but this is no longer the case with inbox0
having the highest priority in servicing.
[How]
This will reintroduce any synchronization bugs that were there
with Replay or PSR-SU touching the cursor at the same time as driver.
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Chuanyu Tseng <chuanyu.tseng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Diffstat (limited to 'tools/perf/scripts/python/task-analyzer.py')
0 files changed, 0 insertions, 0 deletions
