[PATCH] D156972: [DAG] Fix crash in replaceStoreOfInsertLoad

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/test/CodeGen/AMDGPU/replace-store-of-insert-load.ll:2
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
+; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck %s
+
----------------
Don't need -verify-machineinstrs


================
Comment at: llvm/test/CodeGen/AMDGPU/replace-store-of-insert-load.ll:2
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
+; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck %s
+
----------------
arsenm wrote:
> Don't need -verify-machineinstrs
should use a full triple since the alignment matters


================
Comment at: llvm/test/CodeGen/AMDGPU/replace-store-of-insert-load.ll:15
+; CHECK-NEXT:    s_setpc_b64 s[30:31]
+  %1 = load <4 x float>, ptr addrspace(1) %in
+  %2 = insertelement <4 x float> %1, float %arg, i64 3
----------------
Use named values


================
Comment at: llvm/test/CodeGen/AMDGPU/replace-store-of-insert-load.ll:16
+  %1 = load <4 x float>, ptr addrspace(1) %in
+  %2 = insertelement <4 x float> %1, float %arg, i64 3
+  store <4 x float> %2, ptr addrspace(1) %in
----------------
can you try a few different offsets and types?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D156972/new/

https://reviews.llvm.org/D156972