// of the args of the ANDQ needs to be saved so it can be used as the arg to TESTQ.
func andWithUse(x, y int) int {
	z := x & y
	// amd64:`TESTQ\s(AX, AX|BX, BX|CX, CX|DX, DX|SI, SI|DI, DI|R8, R8|R9, R9|R10, R10|R11, R11|R12, R12|R13, R13|R15, R15)`
	if z == 0 {
		return 77
	}
	// use z by returning it
	return z
}

// Verify (OR x (NOT y)) rewrites to (ORN x y) where supported
func ornot(x, y int) int {
	// ppc64x:"ORN"
	z := x | ^y

	print("r8      ", hex(r.r8), "\n")
	print("r9      ", hex(r.r9), "\n")
	print("r10     ", hex(r.r10), "\n")
	print("r11     ", hex(r.r11), "\n")
	print("r12     ", hex(r.r12), "\n")
	print("r13     ", hex(r.r13), "\n")
	print("r14     ", hex(r.r14), "\n")
	print("r15     ", hex(r.r15), "\n")
	print("rip     ", hex(r.rip), "\n")
	print("rflags  ", hex(r.eflags), "\n")
	print("cs      ", hex(r.segcs), "\n")

	MOVQ	a4+32(FP), R10
	MOVQ	a5+40(FP), R8
	MOVQ	a6+48(FP), R9
	MOVQ	a7+56(FP), R11
	MOVQ	a8+64(FP), R12
	MOVQ	a9+72(FP), R13

	// only the first 6 arguments can be passed in registers,
	// the last three should be placed at the top of the stack.
	MOVQ	R11, 8(SP)	// arg 7
	MOVQ	R12, 16(SP)	// arg 8
	MOVQ	R13, 24(SP)	// arg 9

	SYSCALL
	JCC	ok9
	MOVQ	$-1, r1+80(FP)	// r1
	MOVQ	$0, r2+88(FP)	// r2
	MOVQ	AX, err+96(FP)	// errno

	// amd64:`MOVQ\t\$-2401018187971961171, R11`, `MOVQ\t\$-2401018187971961171, R12`, `MOVQ\t\$-2401018187971961171, R13`
	// amd64:-`MOVQ\t\$-2401018187971961171, BP` // frame pointer is not clobbered
	StackArgsCall([10]int{a, b, c})
	// amd64:`MOVQ\t\$-2401018187971961171, R12`, `MOVQ\t\$-2401018187971961171, R13`, `MOVQ\t\$-2401018187971961171, DX`

EXT(crosscall1):
	pushq %rbx
	pushq %rbp
	pushq %r12
	pushq %r13
	pushq %r14
	pushq %r15

#if defined(_WIN64)
	movq %r8, %rdi	/* arg of setg_gcc */
	call *%rdx	/* setg_gcc */
	call *%rcx	/* fn */
#else
	movq %rdi, %rbx
	movq %rdx, %rdi	/* arg of setg_gcc */
	call *%rsi	/* setg_gcc */
	call *%rbx	/* fn */
#endif

	popq %r15
	popq %r14
	popq %r13
	popq %r12
	popq %rbp
	popq %rbx
	ret

.file "gcc_s390x.S"

/*
 * void crosscall_s390x(void (*fn)(void), void *g)
 *
 * Calling into the go tool chain, where all registers are caller save.
 * Called from standard s390x C ABI, where r6-r13, r15, and f8-f15 are
 * callee-save, so they must be saved explicitly.
 */
.globl crosscall_s390x
crosscall_s390x:
	/* save r6-r15 in the register save area of the calling function */
	stmg    %r6, %r15, 48(%r15)

	MOVQ AX, 0(SP)
	MOVQ CX, 8(SP)
	MOVQ DX, 16(SP)
	MOVQ BX, 24(SP)
	MOVQ SI, 32(SP)
	MOVQ DI, 40(SP)
	MOVQ R8, 48(SP)
	MOVQ R9, 56(SP)
	MOVQ R10, 64(SP)
	MOVQ R11, 72(SP)
	MOVQ R12, 80(SP)
	MOVQ R13, 88(SP)
	MOVQ R14, 96(SP)
	MOVQ R15, 104(SP)
	#ifdef GOOS_darwin
	#ifndef hasAVX
	CMPB internal∕cpu·X86+const_offsetX86HasAVX(SB), $0
	JE 2(PC)
	#endif
	VZEROUPPER
	#endif
	MOVUPS X0, 112(SP)

	// load_g will clobber R0, Save R0
	MOVD	R0, R13
	load_g
	// restore R0
	MOVD	R13, R0
	MOVD	g_m(g), R13
	MOVD	m_g0(R13), R14
	CMP	R14, g
	BEQ	noswitch	// branch if already on g0
	MOVD	R14, g

	MOVD	R0, 8(RSP)	// func arg
	MOVD	R1, 16(RSP)	// func arg
	BL	runtime·racecallback(SB)

	// All registers are smashed after Go code, reload.
	MOVD	g_m(g), R13
	MOVD	m_curg(R13), g	// g = m->curg
ret:

	R8L:  "R8",
	R9L:  "R9",
	R10L: "R10",
	R11L: "R11",
	R12L: "R12",
	R13L: "R13",
	R14L: "R14",
	R15L: "R15",
	RAX:  "AX",
	RCX:  "CX",
	RDX:  "DX",
	RBX:  "BX",
	RSP:  "SP",
	RBP:  "BP",
	RSI:  "SI",
	RDI:  "DI",
	R8:   "R8",
	R9:   "R9",
	R10:  "R10",
	R11:  "R11",
	R12:  "R12",
	R13:  "R13",
	R14:  "R14",
	R15:  "R15",
	IP:   "IP",
	EIP:  "IP",
	RIP:  "IP",

	INCL CX
	MOVL CX, 8(SP)
	INCL CX
	MOVL CX, 12(SP)
	MOVOU 0(SP), X12

	// Load seed words into next two rows and into DI, SI, R8..R13
	SEED(0, DI, X4)
	SEED(1, SI, X5)
	SEED(2, R8, X6)
	SEED(3, R9, X7)
	SEED(4, R10, X8)
	SEED(5, R11, X9)
	SEED(6, R12, X10)
	SEED(7, R13, X11)

	// Zeros for remaining two matrix entries.
	// We have just enough XMM registers to hold the state,