Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] drm/xe/vm: Use for_each_tlb_inval() to calculate invalidation fences
@ 2025-11-18 20:14 Matt Roper
  2025-11-18 20:20 ` ✗ CI.KUnit: failure for " Patchwork
                   ` (6 more replies)
  0 siblings, 7 replies; 10+ messages in thread
From: Matt Roper @ 2025-11-18 20:14 UTC (permalink / raw)
  To: intel-xe; +Cc: matthew.d.roper

ops_execute() calculates the size of a fence array based on
XE_MAX_GT_PER_TILE, while the code that actually fills in the fence
array uses a for_each_tlb_inval() iterator.  This works out okay today
since both approaches come up with the same number of invalidation
fences (2: primary GT invalidation + media GT invalidation), but could
be problematic in the future if there isn't a 1:1 relationship between
TLBs needing invalidation and potential GTs on the tile.

Adjust the allocation code to use the same for_each_tlb_inval()
counting logic as the code that fills the array to future-proof the
code.

Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
---
 drivers/gpu/drm/xe/xe_vm.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
index 7cac646bdf1c..6794f38b9340 100644
--- a/drivers/gpu/drm/xe/xe_vm.c
+++ b/drivers/gpu/drm/xe/xe_vm.c
@@ -3111,12 +3111,12 @@ static struct dma_fence *ops_execute(struct xe_vm *vm,
 	if (number_tiles == 0)
 		return ERR_PTR(-ENODATA);
 
-	if (vops->flags & XE_VMA_OPS_FLAG_SKIP_TLB_WAIT) {
-		for_each_tile(tile, vm->xe, id)
-			++n_fence;
-	} else {
-		for_each_tile(tile, vm->xe, id)
-			n_fence += (1 + XE_MAX_GT_PER_TILE);
+	for_each_tile(tile, vm->xe, id) {
+		++n_fence;
+
+		if (!(vops->flags & XE_VMA_OPS_FLAG_SKIP_TLB_WAIT))
+			for_each_tlb_inval(i)
+				++n_fence;
 	}
 
 	fences = kmalloc_array(n_fence, sizeof(*fences), GFP_KERNEL);
-- 
2.51.1


^ permalink raw reply related	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2025-11-19 15:27 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-11-18 20:14 [PATCH] drm/xe/vm: Use for_each_tlb_inval() to calculate invalidation fences Matt Roper
2025-11-18 20:20 ` ✗ CI.KUnit: failure for " Patchwork
2025-11-18 20:26 ` [PATCH] " Matt Roper
2025-11-19  0:24   ` Matthew Brost
2025-11-18 20:33 ` ✓ CI.KUnit: success for drm/xe/vm: Use for_each_tlb_inval() to calculate invalidation fences (rev2) Patchwork
2025-11-18 22:51 ` [PATCH] drm/xe/vm: Use for_each_tlb_inval() to calculate invalidation fences Matthew Brost
2025-11-19  5:49 ` ✓ CI.KUnit: success for drm/xe/vm: Use for_each_tlb_inval() to calculate invalidation fences (rev3) Patchwork
2025-11-19  6:58 ` ✓ Xe.CI.BAT: " Patchwork
2025-11-19  8:15 ` ✓ Xe.CI.Full: " Patchwork
2025-11-19 15:26   ` Matt Roper

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