#define polyMulStage3 IMULQ acc2, t3; ADDQ acc0, t2; ADCQ DX, t3
#define polyMulReduceStage MOVQ t0, acc0; MOVQ t1, acc1; MOVQ t2, acc2; ANDQ $3, acc2; MOVQ t2, t0; ANDQ $-4, t0; MOVQ t3, t1; SHRQ $2, t3, t2; SHRQ $2, t3; ADDQ t0, acc0; ADCQ t1, acc1; ADCQ $0, acc2; ADDQ t2, acc0; ADCQ t3, acc1; ADCQ $0, acc2

// Lowering shifts
// Unsigned shifts need to return 0 if shift amount is >= width of shifted value.
//   result = (arg << shift) & (shift >= argbits ? 0 : 0xffffffffffffffff)
(Lsh64x(64|32|16|8) <t> x y) && !shiftIsBounded(v) => (ANDQ (SHLQ <t> x y) (SBBQcarrymask <t> (CMP(Q|L|W|B)const y [64])))
(Lsh32x(64|32|16|8) <t> x y) && !shiftIsBounded(v) => (ANDL (SHLL <t> x y) (SBBLcarrymask <t> (CMP(Q|L|W|B)const y [32])))

	MOVQ	g_m(R14), R13
	// Switch to g0 stack.
	MOVQ	SP, R12		// callee-saved, preserved across the CALL
	MOVQ	m_g0(R13), R10
	CMPQ	R10, R14
	JE	call	// already on g0
	MOVQ	(g_sched+gobuf_sp)(R10), SP
call:
	ANDQ	$~15, SP	// alignment for gcc ABI
	CALL	AX
	MOVQ	R12, SP
	// Back to Go world, set special registers.
	// The g register (R14) is preserved in C.
	XORPS	X15, X15
	RET

	SBBQ $0, mul0

	MOVQ acc4, acc0
	MOVQ acc5, acc1
	MOVQ acc6, acc2
	MOVQ acc7, acc3

	ADDQ $-1, acc4
	ADCQ p256const0<>(SB), acc5
	ADCQ $0, acc6
	ADCQ p256const1<>(SB), acc7
	ANDQ $1, mul0

	CMOVQEQ acc0, acc4
	CMOVQEQ acc1, acc5
	CMOVQEQ acc2, acc6
	CMOVQEQ acc3, acc7

	RET
/* ---------------------------------------*/
TEXT p256MulInternal(SB),NOSPLIT,$8

	// amd64:`TESTL`,-`ANDL`
	c0 := a&b < 0
	// arm:`CMN`,-`ADD`
	// arm64:`CMNW`,-`ADD`
	c1 := a+b < 0
	// arm:`TEQ`,-`XOR`
	c2 := a^b < 0
	// arm64:`TST`,-`AND`
	// amd64:`TESTQ`,-`ANDQ`
	c3 := e&f < 0
	// arm64:`CMN`,-`ADD`
	c4 := e+f < 0
	// not optimized to single CMNW/CMN due to further use of b+d
	// arm64:`ADD`,-`CMNW`
	// arm:`ADD`,-`CMN`
	c5 := b+d == 0

	CALL	libc_kill(SB)
	RET

TEXT runtime·sigfwd(SB),NOSPLIT,$0-32
	MOVQ	fn+0(FP),    AX
	MOVL	sig+8(FP),   DI
	MOVQ	info+16(FP), SI
	MOVQ	ctx+24(FP),  DX
	MOVQ	SP, BX		// callee-saved
	ANDQ	$~15, SP	// alignment for x86_64 ABI
	CALL	AX
	MOVQ	BX, SP
	RET

// This is the function registered during sigaction and is invoked when
// a signal is received. It just redirects to the Go function sigtrampgo.

			v.AuxInt = int32ToAuxInt(int32(c))
			v.AddArg(x)
			return true
		}
		break
	}
	// match: (ANDQ x x)
	// result: x
	for {
		x := v_0
		if x != v_1 {
			break
		}
		v.copyOf(x)
		return true
	}
	// match: (ANDQ x l:(MOVQload [off] {sym} ptr mem))
	// cond: canMergeLoadClobber(v, l, x) && clobber(l)
	// result: (ANDQload x [off] {sym} ptr mem)
	for {

				{2, 49151}, // AX CX DX BX SP BP SI DI R8 R9 R10 R11 R12 R13 R15
			},
			outputs: []outputInfo{
				{0, 1}, // AX
				{1, 4}, // DX
			},
		},
	},
	{
		name:         "ANDQ",
		argLen:       2,
		commutative:  true,
		resultInArg0: true,
		clobberFlags: true,
		asm:          x86.AANDQ,
		reg: regInfo{
			inputs: []inputInfo{