in  In
	out Out
}

type testCase2[In1, In2, Out any] struct {
	in1 In1
	in2 In2
	out Out
}

func test1[In, Out any](t *testing.T, tests []testCase1[In, Out], name string, f func(In) Out) {
	t.Helper()
	for _, tt := range tests {
		if out := f(tt.in); !reflect.DeepEqual(out, tt.out) {
			t.Errorf("%s(%v) = %v, want %v", name, tt.in, out, tt.out)
		}
	}
}

	LXVD2X (INP)(R10), VIN3
	ADD    $0x40, INP, INP
#ifdef GOARCH_ppc64le
	VPERM  IN0, IN0, LEMASK, IN0
	VPERM  IN1, IN1, LEMASK, IN1
	VPERM  IN2, IN2, LEMASK, IN2
	VPERM  IN3, IN3, LEMASK, IN3
#endif

	VXOR IN0, XL, XH

	VPMSUMD IN1, H3L, XL1
	VPMSUMD IN1, H3, XM1
	VPMSUMD IN1, H3H, XH1

	VPERM   H2, H, HIPERM, H21L
	VPERM   IN2, IN3, LOPERM, T0
	VPERM   H2, H, LOPERM, H21H

	VSPLTW	$3, IN0, KEY        // vspltw 3,1,3
	VSLDOI	$12, ZERO, IN1, TMP // vsldoi 6,0,2,12
	VSBOX	KEY, KEY            // vsbox 3,3

	VXOR	IN1, TMP, IN1       // vxor 2,2,6
	VSLDOI	$12, ZERO, TMP, TMP // vsldoi 6,0,6,12
	VXOR	IN1, TMP, IN1       // vxor 2,2,6
	VSLDOI	$12, ZERO, TMP, TMP // vsldoi 6,0,6,12
	VXOR	IN1, TMP, IN1       // vxor 2,2,6

	VXOR	IN1, KEY, IN1 // vxor 2,2,3
	JMP	loop256       // b .Loop256

done:

  Output in0 = ops::Placeholder(on_gpu.WithOpName("in0"), DT_FLOAT);
  Output in1 = ops::Placeholder(on_gpu.WithOpName("in1"), DT_FLOAT);

  Output perm = ops::Const(on_cpu.WithOpName("perm"), {3, 1, 2, 0});

  {
    Output tr0 = ops::Transpose(on_gpu.WithOpName("tr0"), in0, perm);
    Output tr1 = ops::Transpose(on_gpu.WithOpName("tr1"), in1, perm);
  }

  std::unique_ptr<Graph> result;

	tests := []struct {
		desc   string
		in1    string
		in2    string
		expect bool
	}{
		{
			desc:   "yaml-equal",
			in1:    `foo: bar`,
			in2:    `foo: bar`,
			expect: true,
		},
		{
			desc:   "bad-yaml-1",
			in1:    "O#JF*()#",
			in2:    `foo: bar`,
			expect: false,
		},
		{
			desc:   "bad-yaml-2",
			in1:    `foo: bar`,
			in2:    "#OHJ*#()F",

// license that can be found in the LICENSE file.

//go:build (amd64 || arm64) && !purego

package nistec

import "errors"

// Montgomery multiplication modulo org(G). Sets res = in1 * in2 * R⁻¹.
//
//go:noescape
func p256OrdMul(res, in1, in2 *p256OrdElement)

// Montgomery square modulo org(G), repeated n times (n >= 1).
//
//go:noescape
func p256OrdSqr(res, in *p256OrdElement, n int)

// If zero is 0, sets res = in2. If sel is 0, sets res = in1.
// If sign is not 0, sets res = in1 + -in2. Otherwise, sets res = in1 + in2
//
//go:noescape
func p256PointAddAffineAsm(res, in1 *P256Point, in2 *p256AffinePoint, sign, sel, zero int)

// Point addition. Sets res = in1 + in2. Returns one if the two input points
// were equal and zero otherwise. If in1 or in2 are the point at infinity, res

	VEXT	$8, i2.B16, i1.B16, V7.B16 \
	VADD	V5.D2, i3.D2, i3.D2 \

#define SHA512ROUND(i0, i1, i2, i3, i4, rc0, rc1, in0, in1, in2, in3, in4) \
	VLD1.P	16(R4), [rc1.D2] \
	SHA512TRANS(i0, i1, i2, i3, i4, rc0, in0) \
	VEXT	$8, in4.B16, in3.B16, V5.B16 \
	SHA512SU0	in1.D2, in0.D2 \
	SHA512H	V7.D2, V6, i3 \
	SHA512SU1	V5.D2, in2.D2, in0.D2 \
	VADD	i3.D2, i1.D2, i4.D2 \
	SHA512H2	i0.D2, i1, i3

#define Y0    V2
#define Y1    V3
#define T0    V4
#define T1    V5

// Constants
#define P0    V30
#define P1    V31
// func p256MulAsm(res, in1, in2 *p256Element)
TEXT ·p256Mul(SB), NOSPLIT, $0-24
	MOVD res+0(FP), res_ptr
	MOVD in1+8(FP), x_ptr
	MOVD in2+16(FP), y_ptr
	MOVD $16, R16
	MOVD $32, R17

	MOVD $p256mul<>+0x00(SB), CPOOL


	LXVD2X (R0)(x_ptr), X0

	MOVD	y3, x3
	CBNZ	b_ptr, sqrLoop

	STP	(y0, y1), 0*16(res_ptr)
	STP	(y2, y3), 1*16(res_ptr)
	RET
/* ---------------------------------------*/
// func p256Mul(res, in1, in2 *p256Element)
TEXT ·p256Mul(SB),NOSPLIT,$0
	MOVD	res+0(FP), res_ptr
	MOVD	in1+8(FP), a_ptr
	MOVD	in2+16(FP), b_ptr

	MOVD	p256const0<>(SB), const0
	MOVD	p256const1<>(SB), const1

	LDP	0*16(a_ptr), (x0, x1)
	LDP	1*16(a_ptr), (x2, x3)