linux/drivers/gpu
Amber Lin f3820e9d35 drm/amdkfd: Tie UNMAP_LATENCY to queue_preemption
When KFD asks CP to preempt queues, other than preempt CP queues, CP
also requests SDMA to preempt SDMA queues with UNMAP_LATENCY timeout.
Currently queue_preemption_timeout_ms is 9000 ms by default but can be
configured via module parameter. KFD_UNMAP_LATENCY_MS is hard coded as
4000 ms though. This patch ties KFD_UNMAP_LATENCY_MS to
queue_preemption_timeout_ms so in a slow system such as emulator, both
CP and SDMA slowness are taken into account.

Signed-off-by: Amber Lin <Amber.Lin@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
History 2025-08-27 13:57:49 -04:00
..
drm drm/amdkfd: Tie UNMAP_LATENCY to queue_preemption 2025-08-27 13:57:49 -04:00
host1x gpu: host1x: Use for_each_available_child_of_node_scoped() 2025-05-07 18:09:30 +02:00
ipu-v3 gpu: ipu-v3: Switch to irq_find_mapping() 2025-05-16 21:06:12 +02:00
nova-core gpu: nova-core: fix bounds check in PmuLookupTableEntry::new 2025-07-17 14:10:58 +09:00
trace gpu/trace: make TRACE_GPU_MEM configurable 2025-07-08 09:34:47 -07:00
vga drm: Do delayed switcheroo in drm_lastclose() 2024-08-13 16:21:07 +02:00
Makefile gpu: nova-core: add initial driver stub 2025-03-09 19:24:27 +01:00