[llvm] AMDGPU: Use m_GTrunc in selector def check (PR #216511)

[via llvm-commits]

llvmorg-github-actions[bot] wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-amdgpu

Author: Matt Arsenault (arsenm)

<details>
<summary>Changes</summary>

Prefer mi_match over raw getVRegDef + opcode checks.

---
Full diff: https://github.com/llvm/llvm-project/pull/216511.diff


1 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp (+1-1) 


``````````diff
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp b/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp
index 677312f4f0f52..0a5dbca02fea2 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp
@@ -86,7 +86,7 @@ bool AMDGPUInstructionSelector::isVCC(Register Reg,
     if (!Ty.isValid() || Ty.getSizeInBits() != 1)
       return false;
     // G_TRUNC s1 result is never vcc.
-    return MRI.getVRegDef(Reg)->getOpcode() != AMDGPU::G_TRUNC &&
+    return !mi_match(Reg, MRI, m_GTrunc(m_Reg())) &&
            RC->hasSuperClassEq(TRI.getBoolRC());
   }
 

``````````

</details>


https://github.com/llvm/llvm-project/pull/216511