Igt-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v3 i-g-t 0/1] tests/intel/xe_compute_preempt: Compute preemption check
@ 2025-09-10 17:19 Sobin Thomas
  2025-09-10 17:19 ` [PATCH v3 i-g-t 1/1] " Sobin Thomas
                   ` (4 more replies)
  0 siblings, 5 replies; 9+ messages in thread
From: Sobin Thomas @ 2025-09-10 17:19 UTC (permalink / raw)
  To: igt-dev; +Cc: zbigniew.kempczynski, kamil.konieczny, Sobin Thomas

On platforms like PVC that lack walker mid-thread preemption, running
tests in forked threads can cause inconsistent states due to mid-
execution igt_skip calls. This patch adds a kernel preemption check
at the start of each subtest to ensure consistent behavior. If the
GPU lacks required preemption, the test is skipped gracefully.

v2: Moved compatibility check to igt_fixture per review.
v3: - Added thread group preemption and WMTP checks;
    - Introduced enum flags for preemption types. [Zbigniew]
Signed-off-by: Sobin Thomas <sobin.thomas@intel.com>

Sobin Thomas (1):
  tests/intel/xe_compute_preempt: Compute preemption check

 lib/intel_compute.c              | 22 ++++++++++++++++++++++
 lib/intel_compute.h              |  7 +++++++
 tests/intel/xe_compute_preempt.c |  6 ++++++
 3 files changed, 35 insertions(+)

-- 
2.34.1


^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2025-09-11 15:05 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-09-10 17:19 [PATCH v3 i-g-t 0/1] tests/intel/xe_compute_preempt: Compute preemption check Sobin Thomas
2025-09-10 17:19 ` [PATCH v3 i-g-t 1/1] " Sobin Thomas
2025-09-11  6:49   ` Dandamudi, Priyanka
2025-09-11 11:58     ` Thomas, Sobin
2025-09-11  7:16   ` Zbigniew Kempczyński
2025-09-10 18:08 ` ✓ Xe.CI.BAT: success for " Patchwork
2025-09-10 18:09 ` ✓ i915.CI.BAT: " Patchwork
2025-09-11  0:57 ` ✗ Xe.CI.Full: failure " Patchwork
2025-09-11 15:05 ` ✓ i915.CI.Full: success " Patchwork

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox