[llvm] [AMDGPU] Fix iterator type in fixVALUMaskWriteHazard for bundled instr (PR #178195)

[Matt Arsenault via llvm-commits]

================
@@ -0,0 +1,22 @@
+; RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize64 < %s | FileCheck %s
+;
+; This test validates that the VALU mask write hazard pass correctly handles
+; bundled instructions.
+
+; CHECK-LABEL: main:
+; CHECK: s_getpc_b64
+; CHECK: s_waitcnt_depctr
+
+define amdgpu_cs i32 @main(i32 %arg, i32 %arg1) {
----------------
arsenm wrote:

Include issue number in function name / file name?

https://github.com/llvm/llvm-project/pull/178195