MOVV	savedm-8(SP), R12
	BNE	R12, droppedm

	// Skip dropm to reuse it in the next call, when a pthread key has been created.
	MOVV	_cgo_pthread_key_created(SB), R12
	// It means cgo is disabled when _cgo_pthread_key_created is a nil pointer, need dropm.
	BEQ	R12, dropm
	MOVV    (R12), R12
	BNE	R12, droppedm

dropm:
	MOVV	$runtime·dropm(SB), R4
	JAL	(R4)

 * Called from standard x86-64 ABI, where %rbx, %rbp, %r12-%r15
 * are callee-save so they must be saved explicitly.
 * The standard x86-64 ABI passes the three arguments m, g, fn
 * in %rdi, %rsi, %rdx.
 */
.globl EXT(crosscall1)
EXT(crosscall1):
	pushq %rbx
	pushq %rbp
	pushq %r12
	pushq %r13
	pushq %r14
	pushq %r15

#if defined(_WIN64)

	MVC	$4, 12(R8), 12(R1)  // set chain value
	MVC	$16, (R8), 64(R1)   // set initial counter value
	XC	$32, 16(R1), 16(R1) // set hash subkey and tag
	SLD	$3, R7, R12
	MOVD	R12, 48(R1)         // set total AAD length
	SLD	$3, R5, R12
	MOVD	R12, 56(R1)         // set total plaintext/ciphertext length

	LMG	key+8(FP), R8, R9   // R8=base R9=len
	MVC	$16, (R8), 80(R1)   // set key
	CMPBEQ	R9, $16, kma

	MOVQ	$__msan_memmove(SB), AX
	JMP	msancall<>(SB)

// Switches SP to g0 stack and calls (AX). Arguments already set.
TEXT	msancall<>(SB), NOSPLIT, $0-0
	get_tls(R12)
	MOVQ	g(R12), R14
	MOVQ	SP, R12		// callee-saved, preserved across the CALL
	CMPQ	R14, $0
	JE	call	// no g; still on a system stack

	MOVQ	g_m(R14), R13
	// Switch to g0 stack.
	MOVQ	m_g0(R13), R10
	CMPQ	R10, R14

			"\tMOVBLZX	(off*4)(R12),	reg \\",
			"\tADDB	reg,		DX",
			"",
			"LOAD(8, AX)",
		),
		"\n.\n.MOVBLZX.(.8.*.4.).(.R12.).,.AX.\n.ADDB.AX.,.DX.\n",
	},
	{
		"nested multiline macro",
		lines(
			"#define KEYROUND(xmm, load, off, r1, r2, index) \\",
			"\tMOVBLZX	(BP)(DX*4),	R8 \\",
			"\tload((off+1), r2) \\",
			"\tMOVB	R8,		(off*4)(R12) \\",
			"\tPINSRW	$index, (BP)(R8*4), xmm",

	BEQ     loopexit
scalar:	
	ADD     $-1, R8, R10
	SLD     $3, R10
	MOVD    (R6)(R10),R11
	SRD     R9, R11, R11    // x[len(z)-2] >> s
	SLD     $3, R8, R12
	MOVD    (R6)(R12), R12
	SLD     R5, R12, R12    // x[len(z)-1]<<ŝ
	OR      R12, R11, R11   // x[len(z)-2]>>s | x[len(z)-1]<<ŝ
	MOVD    R11, (R3)(R10)  // z[len(z)-2]=x[len(z)-2]>>s | x[len(z)-1]<<ŝ
loopexit:
	ADD     $-1, R4
	SLD     $3, R4

	print("r6    ", hex(u.r6), "\n")
	print("r7    ", hex(u.r7), "\n")
	print("r8    ", hex(u.r8), "\n")
	print("r9    ", hex(u.r9), "\n")
	print("r10   ", hex(u.r10), "\n")
	print("r11   ", hex(u.r11), "\n")
	print("r12   ", hex(u.r12), "\n")
	print("sp    ", hex(u.sp), "\n")
	print("link  ", hex(u.link), "\n")
	print("pc    ", hex(u.pc), "\n")
	print("psr   ", hex(u.psr), "\n")
}

	MOVD z+0(FP), R2
	MOVD x+8(FP), R8
	MOVD y+16(FP), R9

	MOVD $0, R1 // i*8 = 0
	MOVD $0, R7 // i = 0
	MOVD $0, R0 // make sure it's zero
	MOVD $0, R4 // c = 0

	MOVD   R5, R12
	AND    $-2, R12
	CMPBGE R5, $2, A6
	BR     E6

A6:
	MOVD   (R8)(R1*1), R6
	MULHDU R9, R6
	MOVD   (R2)(R1*1), R10
	ADDC   R10, R11        // add to low order bits
	ADDE   R0, R6
	ADDC   R4, R11
	ADDE   R0, R6

	print("bp    ", hex(u.bp), "\n")
	print("sp    ", hex(u.sp), "\n")
	print("r8    ", hex(u.r8), "\n")
	print("r9    ", hex(u.r9), "\n")
	print("r10   ", hex(u.r10), "\n")
	print("r11   ", hex(u.r11), "\n")
	print("r12   ", hex(u.r12), "\n")
	print("r13   ", hex(u.r13), "\n")
	print("r14   ", hex(u.r14), "\n")
	print("r15   ", hex(u.r15), "\n")
	print("ip    ", hex(u.ip), "\n")
	print("flags ", hex(u.flags), "\n")

	STMG	R2, R3, 8(R15)
	MOVD	R0, 24(R15)
	// R1 already saved.
	// R4 already saved.
	STMG	R5, R12, 32(R15) // save R5 - R12
	// R13 is g.
	// R14 is LR.
	// R15 is SP.

	CALL	runtime·wbBufFlush(SB)

	LMG	8(R15), R2, R3   // restore R2 - R3
	MOVD	24(R15), R0      // restore R0
	LMG	32(R15), R5, R12 // restore R5 - R12
	JMP	retry

TEXT runtime·gcWriteBarrier1<ABIInternal>(SB),NOSPLIT,$0
	MOVD	$8, R9