[PATCH] D83584: [AMDGPU] Move LowerSwitch pass to CodeGenPrepare.

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp:790
     addPass(createLoadStoreVectorizerPass());
+
+  addPass(createLowerSwitchPass());
----------------
Can you add a comment for why this is here


================
Comment at: llvm/test/CodeGen/AMDGPU/switch-default-block-unreachable.ll:6
+  ; follows the LowerSwitch. Otherwise, this testcase will crash soonafter
+  ; the instruction selection due to the incomplete PHI node in an MBB whose
+  ; incoming values were never codegenerated.
----------------
typo soonafter


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D83584/new/

https://reviews.llvm.org/D83584