[clang] [llvm] [LLVMABI][AARCH64] Implement general aggregate classification (PR #230631)

[Madhur Amilkanthwar via llvm-commits]

================
@@ -437,6 +437,7 @@ CodeGenModule::getLLVMABITargetInfo(llvm::abi::TypeBuilder &TB) {
     Opts.IsAndroidOrOHOS = T.isAndroid() || T.isOHOSFamily();
     Opts.IsWindowsArm64EC = T.isWindowsArm64EC();
     Opts.IsMicrosoftCXXABI = getTarget().getCXXABI().isMicrosoft();
+    Opts.IsBigEndian = T.getArch() == llvm::Triple::aarch64_be;
----------------
madhur13490 wrote:

Small thing: could we use !T.isLittleEndian() here instead of matching aarch64_be directly? That's closer to what Clang's lowering does (it asks the DataLayout), and it won't drift if another big-endian triple shows up.

https://github.com/llvm/llvm-project/pull/230631