[llvm] [X86] For inline memset and memcpy with minsize, use size for alignment, rather than actual alignment (PR #87003)

[via llvm-commits]

================
@@ -62,18 +62,31 @@ SDValue X86SelectionDAGInfo::EmitTargetCodeForMemset(
   const X86Subtarget &Subtarget =
       DAG.getMachineFunction().getSubtarget<X86Subtarget>();
 
+  // If we have minsize, then don't care about the alignment.
+  // On x86, the CPU doesn't care and neither should you.
+  // As long as the count is aligned, we can use the minimum number of
+  // instructions without always having to resort to stosb.
+  //
+  // Because this is a feature specific to x86, we must handle it here.
----------------
goldsteinn wrote:

Why do you need count aligned? Just use `rep stosb` (likewise with minsize we shouldn't try to promote `rep stosb` to 16/32/64 bit if we have minsize (and probably not if we have `ERMS` in general but that should be an independent patch)).

https://github.com/llvm/llvm-project/pull/87003