Apparently we do try and attach a fence to an invalid vma (during
execbuf) so we cannot simply assert it never happens and report EINVAL
instead.
Fixes: dec9cf9ee8cb ("drm/i915/gt: Pull restoration of GGTT fences underneath
the GT")
Signed-off-by: Chris Wilson <[email protected]>
Reviewed-by: Mika Kuoppala <[email protected]>
---
drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
index b6ba68c42546..225970f4a4ef 100644
--- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
+++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
@@ -233,10 +233,12 @@ static int fence_update(struct i915_fence_reg *fence,
int ret;
if (vma) {
- GEM_BUG_ON(!i915_vma_is_map_and_fenceable(vma));
GEM_BUG_ON(!i915_gem_object_get_stride(vma->obj) ||
!i915_gem_object_get_tiling(vma->obj));
+ if (!i915_vma_is_map_and_fenceable(vma))
+ return -EINVAL;
+
ret = i915_vma_sync(vma);
if (ret)
return ret;
--
2.20.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
