Remove the inaccurately named ABI_PushAllCalleeSavedRegsAndAdjustStack (it didn't preserve FPRs!) and replace with ABI_PushRegistersAndAdjustStack.

To avoid FPRs being pushed unnecessarily, I checked the uses: DSPEmitter
doesn't use FPRs, and VertexLoader doesn't use anything but RAX, so I
specified the register list accordingly.  The regular JIT, however, does
use FPRs, and as far as I can tell, it was incorrect not to save them in
the outer routine.  Since the dispatcher loop is only exited when
pausing or stopping, this should have no noticeable performance impact.

Source/Core/Common/x64ABI.cpp

@@ -453,63 +453,3 @@ void XEmitter::ABI_CallFunctionA(void *func, const Gen::OpArg &arg1)
 	ABI_RestoreStack(0);
 }
 
-#ifdef _WIN32
-// Win64 Specific Code
-
-void XEmitter::ABI_PushAllCalleeSavedRegsAndAdjustStack()
-{
-	//we only want to do this once
-	PUSH(RBP);
-	MOV(64, R(RBP), R(RSP));
-	PUSH(RBX);
-	PUSH(RSI);
-	PUSH(RDI);
-	PUSH(R12);
-	PUSH(R13);
-	PUSH(R14);
-	PUSH(R15);
-	SUB(64, R(RSP), Imm8(0x28));
-	//TODO: Also preserve XMM0-3?
-}
-
-void XEmitter::ABI_PopAllCalleeSavedRegsAndAdjustStack()
-{
-	ADD(64, R(RSP), Imm8(0x28));
-	POP(R15);
-	POP(R14);
-	POP(R13);
-	POP(R12);
-	POP(RDI);
-	POP(RSI);
-	POP(RBX);
-	POP(RBP);
-}
-
-#else
-// Unix64 Specific Code
-
-void XEmitter::ABI_PushAllCalleeSavedRegsAndAdjustStack()
-{
-	PUSH(RBP);
-	MOV(64, R(RBP), R(RSP));
-	PUSH(RBX);
-	PUSH(R12);
-	PUSH(R13);
-	PUSH(R14);
-	PUSH(R15);
-	SUB(64, R(RSP), Imm8(8));
-}
-
-void XEmitter::ABI_PopAllCalleeSavedRegsAndAdjustStack()
-{
-	ADD(64, R(RSP), Imm8(8));
-	POP(R15);
-	POP(R14);
-	POP(R13);
-	POP(R12);
-	POP(RBX);
-	POP(RBP);
-}
-
-#endif // WIN32
-

Source/Core/Core/DSP/DSPEmitter.cpp

@@ -384,7 +384,9 @@ const u8 *DSPEmitter::CompileStub()
 void DSPEmitter::CompileDispatcher()
 {
 	enterDispatcher = AlignCode16();
-	ABI_PushAllCalleeSavedRegsAndAdjustStack();
+	// We don't use floating point (high 16 bits).
+	u32 registers_used = ABI_ALL_CALLEE_SAVED & 0xffff;
+	ABI_PushRegistersAndAdjustStack(registers_used, 8);
 
 	const u8 *dispatcherLoop = GetCodePtr();
 
@@ -419,6 +421,6 @@ void DSPEmitter::CompileDispatcher()
 		SetJumpTarget(exceptionExit);
 	}
 	//MOV(32, M(&cyclesLeft), Imm32(0));
-	ABI_PopAllCalleeSavedRegsAndAdjustStack();
+	ABI_PopRegistersAndAdjustStack(registers_used, 8);
 	RET();
 }

Source/Core/Core/PowerPC/Jit64/JitAsm.cpp

@@ -16,7 +16,7 @@ using namespace Gen;
 void Jit64AsmRoutineManager::Generate()
 {
 	enterCode = AlignCode16();
-	ABI_PushAllCalleeSavedRegsAndAdjustStack();
+	ABI_PushRegistersAndAdjustStack(ABI_ALL_CALLEE_SAVED, 8);
 
 	// Two statically allocated registers.
 	MOV(64, R(RMEM), Imm64((u64)Memory::base));
@@ -39,7 +39,7 @@ void Jit64AsmRoutineManager::Generate()
 				ABI_CallFunction(reinterpret_cast<void *>(&PowerPC::CheckBreakPoints));
 				TEST(32, M((void*)PowerPC::GetStatePtr()), Imm32(0xFFFFFFFF));
 				FixupBranch noBreakpoint = J_CC(CC_Z);
-				ABI_PopAllCalleeSavedRegsAndAdjustStack();
+				ABI_PopRegistersAndAdjustStack(ABI_ALL_CALLEE_SAVED, 8);
 				RET();
 				SetJumpTarget(noBreakpoint);
 				SetJumpTarget(notStepping);
@@ -126,7 +126,7 @@ void Jit64AsmRoutineManager::Generate()
 		J_CC(CC_Z, outerLoop);
 
 	//Landing pad for drec space
-	ABI_PopAllCalleeSavedRegsAndAdjustStack();
+	ABI_PopRegistersAndAdjustStack(ABI_ALL_CALLEE_SAVED, 8);
 	RET();
 
 	GenerateCommon();

Source/Core/VideoCommon/VertexLoader.cpp

@@ -584,7 +584,8 @@ void VertexLoader::CompileVertexTranslator()
 		PanicAlert("Trying to recompile a vertex translator");
 
 	m_compiledCode = GetCodePtr();
-	ABI_PushRegistersAndAdjustStack(ABI_ALL_CALLEE_SAVED, 8);
+	// We don't use any callee saved registers or anything but RAX.
+	ABI_PushRegistersAndAdjustStack(0, 8);
 
 	// Start loop here
 	const u8 *loop_start = GetCodePtr();
@@ -845,7 +846,7 @@ void VertexLoader::CompileVertexTranslator()
 	SUB(32, MatR(RAX), Imm8(1));
 
 	J_CC(CC_NZ, loop_start);
-	ABI_PopRegistersAndAdjustStack(ABI_ALL_CALLEE_SAVED, 8);
+	ABI_PopRegistersAndAdjustStack(0, 8);
 	RET();
 #endif
 }