[PATCH] D16988: AMDGPU: Remove bfi and bfm intrinsics

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Mon Feb 8 08:55:22 PST 2016


arsenm added inline comments.

================
Comment at: test/CodeGen/AMDGPU/llvm.AMDGPU.bfm.ll:43-60
@@ -42,19 @@
-
-; FUNC-LABEL: {{^}}bfm_pattern:
-; SI: s_bfm_b32 {{s[0-9]+}}, {{s[0-9]+}}, {{s[0-9]+}}
-define void @bfm_pattern(i32 addrspace(1)* %out, i32 %x, i32 %y) {
-  %a = shl i32 1, %x
-  %b = sub i32 %a, 1
-  %c = shl i32 %b, %y
-  store i32 %c, i32 addrspace(1)* %out
-  ret void
-}
-
-; FUNC-LABEL: {{^}}bfm_pattern_simple:
-; SI: s_bfm_b32 {{s[0-9]+}}, {{s[0-9]+}}, 0
-define void @bfm_pattern_simple(i32 addrspace(1)* %out, i32 %x) {
-  %a = shl i32 1, %x
-  %b = sub i32 %a, 1
-  store i32 %b, i32 addrspace(1)* %out
-  ret void
-}
----------------
tstellarAMD wrote:
> These last two tests should be kept.
I extracted these to a new file but forgot to add it


http://reviews.llvm.org/D16988





More information about the llvm-commits mailing list