[llvm] [BOLT][AArch64] Enabling Inlining for Memcpy for AArch64 in BOLT (PR #154929)

[Sjoerd Meijer via llvm-commits]

================
@@ -1866,8 +1866,35 @@ Error InlineMemcpy::runOnFunctions(BinaryContext &BC) {
         const bool IsMemcpy8 = (CalleeSymbol->getName() == "_memcpy8");
         const bool IsTailCall = BC.MIB->isTailCall(Inst);
 
+        // Extract size from preceding instructions (AArch64 only).
----------------
sjoerdmeijer wrote:

I may have added to the confusion. But yeah, I like this new implementation, because before we had a large AArch64 specific blob of code inlined in code that was shared with X86. My only remaining question is the `++II` below. Ideally we fold that the target hook (and don't think we need the isAArch64 check), but I am actually unsure why we need it in the first place. 

https://github.com/llvm/llvm-project/pull/154929