On Mon, 2018-09-10 at 13:04 -0500, Jason Ekstrand wrote:
> ---
> src/compiler/spirv/spirv_to_nir.c | 170 ++++++++++++++++++--------
> ----
> 1 file changed, 103 insertions(+), 67 deletions(-)
>
(...)
> @@ -2516,6 +2600,9 @@ vtn_handle_image(struct vtn_builder *b, SpvOp
> opcode,
> } else {
> nir_builder_instr_insert(&b->nb, &intrin->instr);
> }
> +
> + if (opcode == SpvOpAtomicStore)
> + vtn_emit_memory_barrier(b, scope, semantics);
I think we need to emit the barrier for all "write" opcodes, not just
AtomicStore, right?
Iago
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
