https://github.com/jhuber6 created 
https://github.com/llvm/llvm-project/pull/209317

Summary:
This made some cases use the wrong triple while we are migrating to
using `amdgpu` instead.

Now, this might have some side-effects if someone is still trying to use
`amdgcn`. Honestly, it might be best to just make CMake auto-update this
if we want this to be smooth.


>From b9cc61a276683e18eb56578614fb10fe3bbf68d2 Mon Sep 17 00:00:00 2001
From: Joseph Huber <[email protected]>
Date: Mon, 13 Jul 2026 16:43:07 -0500
Subject: [PATCH] [AMDGPU] Replace old reference to `amdgcn-amd-amdhsa' triple

Summary:
This made some cases use the wrong triple while we are migrating to
using `amdgpu` instead.

Now, this might have some side-effects if someone is still trying to use
`amdgcn`. Honestly, it might be best to just make CMake auto-update this
if we want this to be smooth.
---
 clang/lib/Basic/OffloadArch.cpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/clang/lib/Basic/OffloadArch.cpp b/clang/lib/Basic/OffloadArch.cpp
index 5d12f257f3680..d7e5e5470f886 100644
--- a/clang/lib/Basic/OffloadArch.cpp
+++ b/clang/lib/Basic/OffloadArch.cpp
@@ -173,7 +173,8 @@ llvm::Triple OffloadArchToTriple(const llvm::Triple 
&DefaultToolchainTriple,
   }
 
   if (IsAMDOffloadArch(ID))
-    return llvm::Triple("amdgcn-amd-amdhsa");
+    return llvm::Triple(llvm::Triple::amdgpu, llvm::Triple::NoSubArch,
+                        llvm::Triple::AMD, llvm::Triple::AMDHSA);
 
   return {};
 }

_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to