Mesa (master): radeonsi: work around an LLVM crash when using llvm.amdgcn.icmp.i64.i1

GitLab Mirror gitlab-mirror at kemper.freedesktop.org
Wed Jan 15 21:03:58 UTC 2020


Module: Mesa
Branch: master
Commit: 8ff8e68e422a5c6b9ab26eaa56eff5c27a642d2e
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=8ff8e68e422a5c6b9ab26eaa56eff5c27a642d2e

Author: Marek Olšák <marek.olsak at amd.com>
Date:   Wed Jan  8 21:52:26 2020 -0500

radeonsi: work around an LLVM crash when using llvm.amdgcn.icmp.i64.i1

Cc: 19.2 19.3 <mesa-stable at lists.freedesktop.org>
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer at amd.com>
Tested-by: Marge Bot <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3338>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3338>

---

 src/gallium/drivers/radeonsi/si_compute_prim_discard.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/gallium/drivers/radeonsi/si_compute_prim_discard.c b/src/gallium/drivers/radeonsi/si_compute_prim_discard.c
index 4e691d088c2..2fcec436bb0 100644
--- a/src/gallium/drivers/radeonsi/si_compute_prim_discard.c
+++ b/src/gallium/drivers/radeonsi/si_compute_prim_discard.c
@@ -718,6 +718,7 @@ void si_build_prim_discard_compute_shader(struct si_shader_context *ctx)
 				 ac_get_arg(&ctx->ac, param_smallprim_precision),
 				 &options);
 
+	ac_build_optimization_barrier(&ctx->ac, &accepted);
 	LLVMValueRef accepted_threadmask = ac_get_i1_sgpr_mask(&ctx->ac, accepted);
 
 	/* Count the number of active threads by doing bitcount(accepted). */



More information about the mesa-commit mailing list