The reason we were doing this was to ensure that the kernel did the appropriate cross-ring synchronization and flushing. However, the kernel only looks at EXEC_OBJECT_WRITE to determine whether or not to insert a fence. It only cares about the domain for determining whether or not it needs to clflush the BO before using it for scanout but the domain automatically gets set to RENDER internally by the kernel if EXEC_OBJECT_WRITE is set.
Cc: Chris Wilson <[email protected]> --- src/intel/vulkan/anv_batch_chain.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/src/intel/vulkan/anv_batch_chain.c b/src/intel/vulkan/anv_batch_chain.c index 9def174..9776a45 100644 --- a/src/intel/vulkan/anv_batch_chain.c +++ b/src/intel/vulkan/anv_batch_chain.c @@ -148,9 +148,6 @@ anv_reloc_list_add(struct anv_reloc_list *list, struct drm_i915_gem_relocation_entry *entry; int index; - const uint32_t domain = - (target_bo->flags & EXEC_OBJECT_WRITE) ? I915_GEM_DOMAIN_RENDER : 0; - VkResult result = anv_reloc_list_grow(list, alloc, 1); if (result != VK_SUCCESS) return result; @@ -163,8 +160,8 @@ anv_reloc_list_add(struct anv_reloc_list *list, entry->delta = delta; entry->offset = offset; entry->presumed_offset = target_bo->offset; - entry->read_domains = domain; - entry->write_domain = domain; + entry->read_domains = 0; + entry->write_domain = 0; VG(VALGRIND_CHECK_MEM_IS_DEFINED(entry, sizeof(*entry))); return VK_SUCCESS; -- 2.5.0.400.gff86faf _______________________________________________ mesa-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/mesa-dev
