TEXT ·Syscall6(SB),NOSPLIT,$20-40
	MOVW	num+0(FP), R2	// syscall entry
	MOVW	a1+4(FP), R4
	MOVW	a2+8(FP), R5
	MOVW	a3+12(FP), R6
	MOVW	a4+16(FP), R7
	MOVW	a5+20(FP), R8
	MOVW	a6+24(FP), R9
	MOVW	R8, 16(R29)
	MOVW	R9, 20(R29)
	MOVW	R0, R3	// reset R3 to 0 as 1-ret SYSCALL keeps it
	SYSCALL
	BEQ	R7, ok
	MOVW	$-1, R1
	MOVW	R1, r1+28(FP)
	MOVW	R0, r2+32(FP)
	MOVW	R2, errno+36(FP)
	RET
ok:

TEXT ·Syscall6(SB),NOSPLIT,$0-40
	MOVW	num+0(FP), R7	// syscall entry
	MOVW	a1+4(FP), R0
	MOVW	a2+8(FP), R1
	MOVW	a3+12(FP), R2
	MOVW	a4+16(FP), R3
	MOVW	a5+20(FP), R4
	MOVW	a6+24(FP), R5
	SWI	$0
	MOVW	$0xfffff001, R6
	CMP	R6, R0
	BLS	ok
	MOVW	$-1, R1
	MOVW	R1, r1+28(FP)
	MOVW	$0, R2
	MOVW	R2, r2+32(FP)
	RSB	$0, R0, R0
	MOVW	R0, errno+36(FP)
	RET
ok:
	MOVW	R0, r1+28(FP)
	MOVW	R1, r2+32(FP)

//		return 0;
TEXT ·Cas(SB),NOSPLIT,$0-13
	MOVW	ptr+0(FP), R1
	MOVW	old+4(FP), R2
	MOVW	new+8(FP), R5
	SYNC
try_cas:
	MOVW	R5, R3
	LL	(R1), R4	// R4 = *R1
	BNE	R2, R4, cas_fail
	SC	R3, (R1)	// *R1 = R3
	BEQ	R3, try_cas
	SYNC
	MOVB	R3, ret+12(FP)
	RET
cas_fail:
	SYNC
	MOVB	R0, ret+12(FP)
	RET

TEXT ·Store(SB),NOSPLIT,$0-8
	MOVW	ptr+0(FP), R1
	MOVW	val+4(FP), R2
	SYNC
	MOVW	R2, 0(R1)
	SYNC

	MOVD	name+0(FP), R3
	MOVW	mode+8(FP), R4
	MOVW	perm+12(FP), R5
	SYSCALL	$SYS_open
	BVC	2(PC)
	MOVW	$-1, R3
	MOVW	R3, ret+16(FP)
	RET

TEXT runtime·closefd(SB),NOSPLIT|NOFRAME,$0-12
	MOVW	fd+0(FP), R3
	SYSCALL	$SYS_close
	BVC	2(PC)
	MOVW	$-1, R3
	MOVW	R3, ret+8(FP)
	RET

TEXT runtime·write1(SB),NOSPLIT|NOFRAME,$0-28
	MOVD	fd+0(FP), R3
	MOVD	p+8(FP), R4
	MOVW	n+16(FP), R5
	SYSCALL	$SYS_write
	BVC	2(PC)

	ORR	Rt2<<16, Rt0, Rt0	    ; \
	ORR	Rt1<<24, Rt0, Rt0	    ; \
	MOVW.P	Rt0, 4(Rw)		    ; \
	ADD	Rt0, Re, Re

	// tmp := w[(i-3)&0xf] ^ w[(i-8)&0xf] ^ w[(i-14)&0xf] ^ w[(i)&0xf]
	// w[i&0xf] = tmp<<1 | tmp>>(32-1)
	// e += w[i&0xf]
#define SHUFFLE(Re) \
	MOVW	(-16*4)(Rw), Rt0 ; \
	MOVW	(-14*4)(Rw), Rt1 ; \
	MOVW	(-8*4)(Rw), Rt2  ; \
	EOR	Rt0, Rt1, Rt0  ; \
	MOVW	(-3*4)(Rw), Rt1  ; \
	EOR	Rt2, Rt0, Rt0  ; \
	EOR	Rt0, Rt1, Rt0  ; \

	MOVW	$0, R11
	MOVW	R11, (g_sched+gobuf_lr)(g)

	// Switch to m->g0 & its stack, call fn.
	MOVW	g, R1
	MOVW	g_m(g), R8
	MOVW	m_g0(R8), R0
	BL	setg<>(SB)
	CMP	g, R1
	B.NE	2(PC)
	B	runtime·badmcall(SB)
	MOVW	fn+0(FP), R0
	MOVW	(g_sched+gobuf_sp)(g), R13
	SUB	$8, R13
	MOVW	R1, 4(R13)
	MOVW	R0, R7
	MOVW	0(R0), R0
	BL	(R0)
	B	runtime·badmcall2(SB)
	RET

	MOVW	R28, 96(R29)
	// R29 is SP.
	// R30 is g.
	// R31 is LR, which was saved by the prologue.

	CALL	runtime·wbBufFlush(SB)

	MOVW	4(R29), R20
	MOVW	8(R29), R21
	MOVW	12(R29), R3
	MOVW	16(R29), R4
	MOVW	20(R29), R5
	MOVW	24(R29), R6
	MOVW	28(R29), R7
	MOVW	32(R29), R8
	MOVW	36(R29), R9
	MOVW	40(R29), R10
	MOVW	44(R29), R11
	MOVW	48(R29), R12
	MOVW	52(R29), R13
	MOVW	56(R29), R14

	// amd64:"MOVQ",-"MOVL",-"SHRQ"
	// arm64:"MOVD",-"MOVW",-"LSR"
	// ppc64le:"MOVD",-"MOVW",-"SRD"
	p.a = uint32(x)
	// amd64:-"MOVL",-"SHRQ"
	// arm64:-"MOVW",-"LSR"
	// ppc64le:-"MOVW",-"SRD"
	p.b = uint32(x >> 32)
}
func store32be(p *struct{ a, b uint32 }, x uint64) {
	// ppc64:"MOVD",-"MOVW",-"SRD"
	// s390x:"MOVD",-"MOVW",-"SRD"
	p.a = uint32(x >> 32)
	// ppc64:-"MOVW",-"SRD"
	// s390x:-"MOVW",-"SRD"

TEXT ·Or32(SB), NOSPLIT, $0-20
	MOVD	ptr+0(FP), R3
	MOVW	val+8(FP), R4
	LWSYNC
again:
	LWAR	(R3), R6
	OR	R4, R6, R7
	STWCCC	R7, (R3)
	BNE	again
	MOVW	R6, ret+16(FP)
	RET

// func And32(addr *uint32, v uint32) old uint32
TEXT ·And32(SB), NOSPLIT, $0-20
	MOVD	ptr+0(FP), R3
	MOVW	val+8(FP), R4
	LWSYNC
again:
	LWAR	(R3),R6
	AND	R4, R6, R7
	STWCCC	R7, (R3)

TEXT ·And(SB), NOSPLIT, $0-12
	MOV	ptr+0(FP), A0
	MOVW	val+8(FP), A1
	AMOANDW	A1, (A0), ZERO
	RET

// func Or(ptr *uint32, val uint32)
TEXT ·Or(SB), NOSPLIT, $0-12
	MOV	ptr+0(FP), A0
	MOVW	val+8(FP), A1
	AMOORW	A1, (A0), ZERO
	RET

// func Or32(ptr *uint32, val uint32) uint32
TEXT ·Or32(SB), NOSPLIT, $0-20
	MOV	ptr+0(FP), A0
	MOVW	val+8(FP), A1
	AMOORW	A1, (A0), A2
	MOVW	A2, ret+16(FP)