[llvm] [AMDGPU] Remove unnecessary and broken sign/zero-extension (PR #203436)

[Jay Foad via llvm-commits]

================
@@ -1,11 +1,12 @@
-; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6
 ; RUN: opt -S -mtriple=amdgcn-- -mcpu=tahiti -amdgpu-codegenprepare -amdgpu-bypass-slow-div=0 %s | FileCheck %s
 ; RUN: llc -mtriple=amdgcn-- -mcpu=tahiti -amdgpu-bypass-slow-div=0 < %s | FileCheck -check-prefix=GFX6 %s
 ; RUN: llc -mtriple=amdgcn-- -mcpu=gfx900 -amdgpu-bypass-slow-div=0 < %s | FileCheck -check-prefix=GFX9 %s
 
 define amdgpu_kernel void @udiv_i32(ptr addrspace(1) %out, i32 %x, i32 %y) {
-; CHECK-LABEL: @udiv_i32(
-; CHECK-NEXT:    [[TMP1:%.*]] = uitofp fast i32 [[Y:%.*]] to float
+; CHECK-LABEL: define amdgpu_kernel void @udiv_i32(
+; CHECK-SAME: ptr addrspace(1) [[OUT:%.*]], i32 [[X:%.*]], i32 [[Y:%.*]]) #[[ATTR0:[0-9]+]] {
+; CHECK-NEXT:    [[TMP1:%.*]] = uitofp fast i32 [[Y]] to float
----------------
jayfoad wrote:

Are some of these diffs just from rerunning the update script? Can you commit that separately as an NFC?

https://github.com/llvm/llvm-project/pull/203436