================
@@ -1,10 +1,22 @@
; RUN: opt -S -dxil-op-lower -mtriple=dxil-pc-shadermodel6.3-compute %s |
FileCheck %s
-define noundef <4 x i32> @wave_ballot_simple(i1 noundef %p1) {
+%dx.types.fouri32 = type { i32, i32, i32, i32 }
+
+define <4 x i32> @wave_ballot_simple(i1 noundef %p1) {
entry:
-; CHECK: call <4 x i32> @dx.op.waveBallot.void(i32 118, i1 %p1)
- %ret = call <4 x i32> @llvm.dx.wave.ballot(i1 %p1)
- ret <4 x i32> %ret
+; CHECK: call %dx.types.fouri32 @dx.op.waveActiveBallot(i32 116, i1 %p1)
+; CHECK-NOT: ret %dx.types.fouri32
+; CHECK: ret <4 x i32>
----------------
inbelic wrote:
Maybe this is just a confusion on my part but. I would not have expected us to
return a vector in the final dxil output? IIUC that is the point of the custom
struct type.
Ah wait, the insertelements will go away when we run the scalarizer.
Can we add a test that runs with the scalarizer to ensure we get the expected
output in that case?
edit: Although I see in the `dxil-dis` we still have a vector around, why is
that?
https://github.com/llvm/llvm-project/pull/175105
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
