[PATCH] D147157: [AMDGPU] Do not fix implicit vcc operand on INLINEASM

[Jay Foad via Phabricator via llvm-commits]

foad created this revision.
foad added a reviewer: AMDGPU.
Herald added subscribers: kosarev, StephenFan, kerbowa, hiraditya, tpr, dstuttard, yaxunl, jvesely, kzhuravl, arsenm.
Herald added a project: All.
foad requested review of this revision.
Herald added subscribers: llvm-commits, wdng.
Herald added a project: LLVM.

An INLINEASM can have an implicit def of vcc. It is not appropriate for
fixImplicitOperands to change this to vcc_lo on wave32.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D147157

Files:
  llvm/lib/Target/AMDGPU/SIInstrInfo.cpp


Index: llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -7877,6 +7877,9 @@
   if (!ST.isWave32())
     return;
 
+  if (MI.isInlineAsm())
+    return;
+
   for (auto &Op : MI.implicit_operands()) {
     if (Op.isReg() && Op.getReg() == AMDGPU::VCC)
       Op.setReg(AMDGPU::VCC_LO);


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D147157.509357.patch
Type: text/x-patch
Size: 430 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230329/b48a737c/attachment.bin>