[clang][AArch64] multilib: fix deduction of "-march=" option

[domin144]

The deduced "-march=" option always started with aarch64, which is not a valid value. There was also no way to distinguish between armv8-r and armv8-a. After this commit, the deduced "-march=" option will start with greatest available "armv*-a" value or "armv8-r".

[llvmbot]

@llvm/pr-subscribers-clang-driver

Author: Dominik Wójt (domin144)

Changes

The deduced "-march=" option always started with aarch64, which is not a valid value. There was also no way to distinguish between armv8-r and armv8-a. After this commit, the deduced "-march=" option will start with greatest available "armv*-a" value or "armv8-r".

---

Full diff: https://github.com/llvm/llvm-project/pull/81474.diff

1 Files Affected:

• (modified) clang/lib/Driver/ToolChain.cpp (+5-1)

diff --git a/clang/lib/Driver/ToolChain.cpp b/clang/lib/Driver/ToolChain.cpp
index 388030592b4836..bfc80ac79df1e3 100644
--- a/clang/lib/Driver/ToolChain.cpp
+++ b/clang/lib/Driver/ToolChain.cpp
@@ -191,7 +191,11 @@ static void getAArch64MultilibFlags(const Driver &D,
   for (const auto &Ext : AArch64::Extensions)
     if (FeatureSet.contains(Ext.NegFeature))
       MArch.push_back(("no" + Ext.Name).str());
-  MArch.insert(MArch.begin(), ("-march=" + Triple.getArchName()).str());
+  StringRef ArchName;
+  for (const auto &ArchInfo : AArch64::ArchInfos)
+    if (FeatureSet.contains(ArchInfo->ArchFeature))
+      ArchName = ArchInfo->Name;
+  MArch.insert(MArch.begin(), ("-march=" + ArchName).str());
   Result.push_back(llvm::join(MArch, "+"));
 }
 

[llvmbot]

@llvm/pr-subscribers-clang

Author: Dominik Wójt (domin144)

Changes

The deduced "-march=" option always started with aarch64, which is not a valid value. There was also no way to distinguish between armv8-r and armv8-a. After this commit, the deduced "-march=" option will start with greatest available "armv*-a" value or "armv8-r".

---

Full diff: https://github.com/llvm/llvm-project/pull/81474.diff

1 Files Affected:

• (modified) clang/lib/Driver/ToolChain.cpp (+5-1)

diff --git a/clang/lib/Driver/ToolChain.cpp b/clang/lib/Driver/ToolChain.cpp
index 388030592b4836..bfc80ac79df1e3 100644
--- a/clang/lib/Driver/ToolChain.cpp
+++ b/clang/lib/Driver/ToolChain.cpp
@@ -191,7 +191,11 @@ static void getAArch64MultilibFlags(const Driver &D,
   for (const auto &Ext : AArch64::Extensions)
     if (FeatureSet.contains(Ext.NegFeature))
       MArch.push_back(("no" + Ext.Name).str());
-  MArch.insert(MArch.begin(), ("-march=" + Triple.getArchName()).str());
+  StringRef ArchName;
+  for (const auto &ArchInfo : AArch64::ArchInfos)
+    if (FeatureSet.contains(ArchInfo->ArchFeature))
+      ArchName = ArchInfo->Name;
+  MArch.insert(MArch.begin(), ("-march=" + ArchName).str());
   Result.push_back(llvm::join(MArch, "+"));
 }
 

[domin144]

@MaskRay @petrhosek @smithp35 @mplatings
Looking at previous reviews for multilib related changes you might be interested.
(ignore the buildkite failure for now. It seems unrelated.)

[mplatings]

Definitely an improvement, thanks.

[domin144]

@mplatings Can you merge for me? I don't have needed rights.
I rebased hoping the check will pass now, but there are still some unrelated failures on windows.