[clang] [llvm] Clang: don't unnecessarily convert inline-asm operands to x86mmx in IR. (PR #98273)

[Phoebe Wang via cfe-commits]

================
@@ -58173,8 +58173,8 @@ X86TargetLowering::getSingleConstraintMatchWeight(
       return CW_Invalid;
     // Any MMX reg
     case 'm':
-      if (Ty->isX86_MMXTy() && Subtarget.hasMMX())
-        return Wt;
+      if (Ty->getPrimitiveSizeInBits() == 64 && Subtarget.hasMMX())
----------------
phoebewang wrote:

Add a test case for "Ym"?

https://github.com/llvm/llvm-project/pull/98273