[Intel-xe] [PATCH 1/2] drm/xe/selftests: consider multi-GT for eviction test
Matthew Auld
matthew.auld at intel.com
Fri Sep 1 14:28:25 UTC 2023
We need to sanitize and reset each GT, since xe_bo_evict_all() will
evict everything regardless of GT, which can leave other GTs in a broken
state.
Signed-off-by: Matthew Auld <matthew.auld at intel.com>
Cc: Nirmoy Das <nirmoy.das at intel.com>
---
drivers/gpu/drm/xe/tests/xe_bo.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/xe/tests/xe_bo.c b/drivers/gpu/drm/xe/tests/xe_bo.c
index b32a9068d76c..0e4ec22c5667 100644
--- a/drivers/gpu/drm/xe/tests/xe_bo.c
+++ b/drivers/gpu/drm/xe/tests/xe_bo.c
@@ -181,7 +181,8 @@ static int evict_test_run_gt(struct xe_device *xe, struct xe_gt *gt, struct kuni
XE_BO_CREATE_VRAM_IF_DGFX(gt_to_tile(gt));
struct xe_vm *vm = xe_migrate_get_vm(xe_device_get_root_tile(xe)->migrate);
struct ww_acquire_ctx ww;
- int err, i;
+ struct xe_gt *__gt;
+ int err, i, id;
kunit_info(test, "Testing device %s gt id %u vram id %u\n",
dev_name(xe->drm.dev), gt->info.id, gt_to_tile(gt)->id);
@@ -218,7 +219,8 @@ static int evict_test_run_gt(struct xe_device *xe, struct xe_gt *gt, struct kuni
goto cleanup_all;
}
- xe_gt_sanitize(gt);
+ for_each_gt(__gt, xe, id)
+ xe_gt_sanitize(__gt);
err = xe_bo_restore_kernel(xe);
/*
* Snapshotting the CTB and copying back a potentially old
@@ -231,8 +233,10 @@ static int evict_test_run_gt(struct xe_device *xe, struct xe_gt *gt, struct kuni
* however seems quite fragile not to also restart the GT. Try
* to do that here by triggering a GT reset.
*/
- xe_gt_reset_async(gt);
- flush_work(>->reset.worker);
+ for_each_gt(__gt, xe, id) {
+ xe_gt_reset_async(__gt);
+ flush_work(&__gt->reset.worker);
+ }
if (err) {
KUNIT_FAIL(test, "restore kernel err=%pe\n",
ERR_PTR(err));
--
2.41.0
More information about the Intel-xe
mailing list