[llvm] [AMDGPU][CodeGen] Do not backtrace invalid -regalloc param (PR #119687)

via llvm-commits llvm-commits at lists.llvm.org
Thu Dec 12 02:18:58 PST 2024


llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-amdgpu

Author: Akshat Oke (optimisan)

<details>
<summary>Changes</summary>

No need to generate a stack trace and a GitHub issue prompt on a wrongly set regalloc option.

---
Full diff: https://github.com/llvm/llvm-project/pull/119687.diff


1 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp (+2-2) 


``````````diff
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
index 34ad99dd980f27..c0dace31393fd7 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
@@ -1549,7 +1549,7 @@ static const char RegAllocOptNotSupportedMessage[] =
 
 bool GCNPassConfig::addRegAssignAndRewriteFast() {
   if (!usingDefaultRegAlloc())
-    report_fatal_error(RegAllocOptNotSupportedMessage);
+    report_fatal_error(RegAllocOptNotSupportedMessage, /*GenCrashDiag=*/false);
 
   addPass(&GCNPreRALongBranchRegID);
 
@@ -1575,7 +1575,7 @@ bool GCNPassConfig::addRegAssignAndRewriteFast() {
 
 bool GCNPassConfig::addRegAssignAndRewriteOptimized() {
   if (!usingDefaultRegAlloc())
-    report_fatal_error(RegAllocOptNotSupportedMessage);
+    report_fatal_error(RegAllocOptNotSupportedMessage, /*GenCrashDiag=*/false);
 
   addPass(&GCNPreRALongBranchRegID);
 

``````````

</details>


https://github.com/llvm/llvm-project/pull/119687


More information about the llvm-commits mailing list