MOVUPS X0, 28(SP)
	MOVUPS X1, 44(SP)
	MOVUPS X2, 60(SP)
	MOVUPS X3, 76(SP)
	MOVUPS X4, 92(SP)
	MOVUPS X5, 108(SP)
	MOVUPS X6, 124(SP)
	MOVUPS X7, 140(SP)
	#endif
	CALL ·asyncPreempt2(SB)
	#ifndef GO386_softfloat
	MOVUPS 140(SP), X7
	MOVUPS 124(SP), X6
	MOVUPS 108(SP), X5
	MOVUPS 92(SP), X4
	MOVUPS 76(SP), X3
	MOVUPS 60(SP), X2
	MOVUPS 44(SP), X1
	MOVUPS 28(SP), X0
	#endif
	MOVL 24(SP), DI

	VPGATHERQQ Y2, (BP)(X7*2), Y2   // ERROR "mask, index, and destination registers should be distinct"
	VPGATHERQQ Y7, (BP)(X2*2), Y2   // ERROR "mask, index, and destination registers should be distinct"
	VPGATHERDQ X2, 664(X2*8), X2    // ERROR "mask, index, and destination registers should be distinct"
	VPGATHERDQ X2, 664(X2*8), X7    // ERROR "mask, index, and destination registers should be distinct"

		x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13)
		x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14)
		x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15)

		// Column round.
		x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15)
		x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12)
		x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13)
		x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14)
	}

	_ = out[31] // bounds check elimination hint

	x6, x5 = bits.Mul64(x4, arg2[3])
	var x7 uint64
	var x8 uint64
	x8, x7 = bits.Mul64(x4, arg2[2])
	var x9 uint64
	var x10 uint64
	x10, x9 = bits.Mul64(x4, arg2[1])
	var x11 uint64
	var x12 uint64
	x12, x11 = bits.Mul64(x4, arg2[0])
	var x13 uint64
	var x14 uint64
	x13, x14 = bits.Add64(x12, x9, uint64(0x0))
	var x15 uint64
	var x16 uint64
	x15, x16 = bits.Add64(x10, x7, uint64(p256Uint1(x14)))
	var x17 uint64

	QR(X1, X5, X9, X13, X15)
	MOVOU (15*16)(BX), X15 // reload X15; temp now X4
	QR(X2, X6, X10, X14, X4)
	QR(X3, X7, X11, X15, X4)

	QR(X0, X5, X10, X15, X4)
	MOVOU X15, (15*16)(BX) // save X15
	QR(X1, X6, X11, X12, X4)
	MOVOU (4*16)(BX), X4  // reload X4; temp now X15
	QR(X2, X7, X8, X13, X15)
	QR(X3, X4, X9, X14, X15)

	DECL DX
	JNZ loop

	// Store interlaced blocks back to output buffer,

// cycles via functions

var x4 = x5
var x5 /* ERROR "initialization cycle for x5" */ = f1()
func f1() int { return x5*10 }

var x6, x7 /* ERROR "initialization cycle" */ = f2()
var x8 = x7
func f2() (int, int) { return f3() + f3(), 0 }
func f3() int { return x8 }

// cycles via function literals

		x2 := byteorder.LeUint32(q[4*0x2:])
		x3 := byteorder.LeUint32(q[4*0x3:])
		x4 := byteorder.LeUint32(q[4*0x4:])
		x5 := byteorder.LeUint32(q[4*0x5:])
		x6 := byteorder.LeUint32(q[4*0x6:])
		x7 := byteorder.LeUint32(q[4*0x7:])
		x8 := byteorder.LeUint32(q[4*0x8:])
		x9 := byteorder.LeUint32(q[4*0x9:])
		xa := byteorder.LeUint32(q[4*0xa:])
		xb := byteorder.LeUint32(q[4*0xb:])
		xc := byteorder.LeUint32(q[4*0xc:])

	x6, x5 = bits.Mul64(x4, arg2[3])
	var x7 uint64
	var x8 uint64
	x8, x7 = bits.Mul64(x4, arg2[2])
	var x9 uint64
	var x10 uint64
	x10, x9 = bits.Mul64(x4, arg2[1])
	var x11 uint64
	var x12 uint64
	x12, x11 = bits.Mul64(x4, arg2[0])
	var x13 uint64
	var x14 uint64
	x13, x14 = bits.Add64(x12, x9, uint64(0x0))
	var x15 uint64
	var x16 uint64
	x15, x16 = bits.Add64(x10, x7, uint64(p224Uint1(x14)))
	var x17 uint64

	JE 2(PC)
	#endif
	VZEROUPPER
	#endif
	MOVUPS X0, 112(SP)
	MOVUPS X1, 128(SP)
	MOVUPS X2, 144(SP)
	MOVUPS X3, 160(SP)
	MOVUPS X4, 176(SP)
	MOVUPS X5, 192(SP)
	MOVUPS X6, 208(SP)
	MOVUPS X7, 224(SP)
	MOVUPS X8, 240(SP)
	MOVUPS X9, 256(SP)
	MOVUPS X10, 272(SP)
	MOVUPS X11, 288(SP)
	MOVUPS X12, 304(SP)
	MOVUPS X13, 320(SP)
	MOVUPS X14, 336(SP)
	MOVUPS X15, 352(SP)
	CALL ·asyncPreempt2(SB)

	MOVOU yout(16*0), X2
	MOVOU yout(16*1), X3
	MOVOU zout(16*0), X4
	MOVOU zout(16*1), X5

	MOVL BX, X6
	MOVL CX, X7

	PXOR X8, X8
	PCMPEQL X9, X9

	PSHUFD $0, X6, X6
	PSHUFD $0, X7, X7

	PCMPEQL X8, X6
	PCMPEQL X8, X7

	MOVOU X6, X15
	PANDN X9, X15

	MOVOU x1in(16*0), X9
	MOVOU x1in(16*1), X10
	MOVOU y1in(16*0), X11