[Xtensa] Xtensa ABI 128bit arg alignment

- Forces 128bit arguments to have 128bit alignment as per the Xtensa ABI
  in LLVM & Clang.
- Adds a check in the Xtensa calling convention to ensure 128bit aligned
  arguments are always passed as the first argument _or_ passed via the
  stack.

Author: MabezDev

clang/lib/Basic/Targets/Xtensa.h

@@ -50,7 +50,7 @@ class LLVM_LIBRARY_VISIBILITY XtensaTargetInfo : public TargetInfo {
     WIntType = UnsignedInt;
     UseZeroLengthBitfieldAlignment = true;
     MaxAtomicPromoteWidth = MaxAtomicInlineWidth = 32;
-    resetDataLayout("e-m:e-p:32:32-i8:8:32-i16:16:32-i64:64-n32");
+    resetDataLayout("e-m:e-p:32:32-i8:8:32-i16:16:32-i64:64-i128:128-n32");
   }
 
   void getTargetDefines(const LangOptions &Opts,

llvm/lib/Target/Xtensa/XtensaISelLowering.cpp

@@ -755,14 +755,19 @@ static bool CC_Xtensa_Custom(unsigned ValNo, MVT ValVT, MVT LocVT,
   unsigned Reg;
 
   Align OrigAlign = ArgFlags.getNonZeroOrigAlign();
-  bool isI64 = (ValVT == MVT::i32 && OrigAlign == Align(8));
+  bool needs64BitAlign = (ValVT == MVT::i32 && OrigAlign == Align(8));
+  bool needs128BitAlign = (ValVT == MVT::i32 && OrigAlign == Align(16));
 
   if (ValVT == MVT::i32 || ValVT == MVT::f32) {
     Reg = State.AllocateReg(IntRegs);
     // If this is the first part of an i64 arg,
     // the allocated register must be either A2, A4 or A6.
-    if (isI64 && (Reg == Xtensa::A3 || Reg == Xtensa::A5 || Reg == Xtensa::A7))
+    if (needs64BitAlign && (Reg == Xtensa::A3 || Reg == Xtensa::A5 || Reg == Xtensa::A7))
       Reg = State.AllocateReg(IntRegs);
+    // arguments with 16byte alignment must be passed in the first register or passed via stack
+    if (needs128BitAlign && Reg != Xtensa::A2)
+      while ( (Reg = State.AllocateReg(IntRegs)) ) {
+      }
     LocVT = MVT::i32;
   } else if (ValVT == MVT::f64) {
     // Allocate int register and shadow next int register.

llvm/lib/Target/Xtensa/XtensaTargetMachine.cpp

@@ -33,7 +33,7 @@ extern "C" LLVM_EXTERNAL_VISIBILITY void LLVMInitializeXtensaTarget() {
 static std::string computeDataLayout(const Triple &TT, StringRef CPU,
                                      const TargetOptions &Options,
                                      bool IsLittle) {
-  std::string Ret = "e-m:e-p:32:32-i8:8:32-i16:16:32-i64:64-n32";
+  std::string Ret = "e-m:e-p:32:32-i8:8:32-i16:16:32-i64:64-i128:128-n32";
 
   return Ret;
 }