p.Add(p, base)

		// Convert p to affine coordinates.
		p256Inverse(zInv, &p.z)
		p256Sqr(zInvSq, zInv, 1)
		p256Mul(zInv, zInv, zInvSq)

		p256Mul(&p.x, &p.x, zInvSq)
		p256Mul(&p.y, &p.y, zInv)
		p.z = p256One

		if p256Equal(&table[j].x, &p.x) != 1 || p256Equal(&table[j].y, &p.y) != 1 {
			t.Fatalf("incorrect table entry at index %d", j)
		}
	}

	"unsafe"
)

// p256Element is a P-256 base field element in [0, P-1] in the Montgomery
// domain (with R 2²⁵⁶) as four limbs in little-endian order value.
type p256Element [4]uint64

// p256One is one in the Montgomery domain.
var p256One = p256Element{0x0000000000000001, 0xffffffff00000000,
	0xffffffffffffffff, 0x00000000fffffffe}

var p256Zero = p256Element{}

// p256P is 2²⁵⁶ - 2²²⁴ + 2¹⁹² + 2⁹⁶ - 1 in the Montgomery domain.

DATA p256ordK0<>+0x00(SB)/8, $0xccd1c8aaee00bc4f
DATA p256ord<>+0x00(SB)/8, $0xf3b9cac2fc632551
DATA p256ord<>+0x08(SB)/8, $0xbce6faada7179e84
DATA p256ord<>+0x10(SB)/8, $0xffffffffffffffff
DATA p256ord<>+0x18(SB)/8, $0xffffffff00000000
DATA p256one<>+0x00(SB)/8, $0x0000000000000001
DATA p256one<>+0x08(SB)/8, $0xffffffff00000000
DATA p256one<>+0x10(SB)/8, $0xffffffffffffffff
DATA p256one<>+0x18(SB)/8, $0x00000000fffffffe

DATA p256ordK0<>+0x00(SB)/8, $0xccd1c8aaee00bc4f
DATA p256ord<>+0x00(SB)/8, $0xf3b9cac2fc632551
DATA p256ord<>+0x08(SB)/8, $0xbce6faada7179e84
DATA p256ord<>+0x10(SB)/8, $0xffffffffffffffff
DATA p256ord<>+0x18(SB)/8, $0xffffffff00000000
DATA p256one<>+0x00(SB)/8, $0x0000000000000001
DATA p256one<>+0x08(SB)/8, $0xffffffff00000000
DATA p256one<>+0x10(SB)/8, $0xffffffffffffffff
DATA p256one<>+0x18(SB)/8, $0x00000000fffffffe

			newPoint: func() *nistec.P224Point { return nistec.NewP224Point() },
		}
		precomputeParams(_p224, elliptic.P224())
	})
	return _p224
}

var p256Once sync.Once
var _p256 *nistCurve[*nistec.P256Point]

func p256() *nistCurve[*nistec.P256Point] {
	p256Once.Do(func() {
		_p256 = &nistCurve[*nistec.P256Point]{
			newPoint: func() *nistec.P256Point { return nistec.NewP256Point() },
		}

//go:build !purego

#include "textflag.h"
#include "go_asm.h"

DATA p256ordK0<>+0x00(SB)/4, $0xee00bc4f
DATA p256ord<>+0x00(SB)/8, $0xffffffff00000000
DATA p256ord<>+0x08(SB)/8, $0xffffffffffffffff
DATA p256ord<>+0x10(SB)/8, $0xbce6faada7179e84
DATA p256ord<>+0x18(SB)/8, $0xf3b9cac2fc632551
DATA p256<>+0x00(SB)/8, $0xffffffff00000001 // P256
DATA p256<>+0x08(SB)/8, $0x0000000000000000 // P256

DATA p256mul<>+0x90(SB)/8, $0x00000000fffffffe // (1*2^256)%P256
DATA p256mul<>+0x98(SB)/8, $0xffffffffffffffff // (1*2^256)%P256

// External declarations for constants
GLOBL p256ord<>(SB), 8, $32
GLOBL p256<>(SB), 8, $80
GLOBL p256mul<>(SB), 8, $160

// The following macros are used to implement the ppc64le
// equivalent function from the corresponding s390x