Our fence management is lazy, very lazy. If the user marks an object as untiled, we do not immediately flush the fence but merely mark it as dirty. On the next use we have to remember to check and remove the fence, by which time we hope it is idle and we do not have to wait. v2: Throw away the old fence on the next ggtt_pin. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111468 Fixes: 1f7fd484fff1 ("drm/i915: Replace i915_vma_put_fence()") Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> Cc: Matthew Auld <matthew.auld@xxxxxxxxx> --- I'm just going to ignore the race with userspace installing a fence (set-tiling + mmap_gtt) while we pread/pwrite. The answer to that is the object-lock, which we currently take after we pin. Should be easy enough to rearrange... Hopefully. --- drivers/gpu/drm/i915/i915_gem.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index eb31b69a316a..8bf346c32b8f 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -1026,6 +1026,14 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj, return ERR_PTR(ret); } + if (vma->fence && !i915_gem_object_is_tiled(obj)) { + mutex_lock(&vma->vm->mutex); + ret = i915_vma_revoke_fence(vma); + mutex_unlock(&vma->vm->mutex); + if (ret) + return ERR_PTR(ret); + } + ret = i915_vma_pin(vma, size, alignment, flags | PIN_GLOBAL); if (ret) return ERR_PTR(ret); -- 2.23.0 _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
