virgl: Request setting the atomic offset in the range_base

With that NTT can encode the array base of atomic arrays separately
so that the host driver can address the arrays correctly.

Fixes GL-CTS: KHR-Single-GL43.arrays_of_arrays_gl.AtomicUsage

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19980>
This commit is contained in:
Gert Wollny
2022-11-24 17:09:42 +01:00
committed by Marge Bot
parent 994cf0e995
commit d59e5aa08f
2 changed files with 1 additions and 1 deletions
@@ -51,7 +51,6 @@ KHR-GL43.shader_storage_buffer_object.basic-atomic-case1,Fail
KHR-GL43.shader_storage_buffer_object.basic-atomic-case2,Fail
KHR-GL43.shader_subroutine.ssbo_atomic_image_load_store,Fail
KHR-Single-GL43.arrays_of_arrays_gl.AtomicUsage,Fail
KHR-Single-GL43.arrays_of_arrays_gl.InteractionArgumentAliasing1,Fail
KHR-Single-GL43.arrays_of_arrays_gl.InteractionArgumentAliasing2,Fail
KHR-Single-GL43.arrays_of_arrays_gl.InteractionArgumentAliasing3,Fail
+1
View File
@@ -1181,6 +1181,7 @@ virgl_create_screen(struct virgl_winsys *vws, const struct pipe_screen_config *c
screen->compiler_options.lower_ffma32 = true;
screen->compiler_options.fuse_ffma32 = false;
screen->compiler_options.lower_image_offset_to_range_base = true;
screen->compiler_options.lower_atomic_offset_to_range_base = true;
slab_create_parent(&screen->transfer_pool, sizeof(struct virgl_transfer), 16);