faas-rs/faas
1
0
Fork 0
mirror of https://github.com/openfaas/faas.git synced 2025-06-22 14:53:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

Rename Makefile targets

Browse Source 
Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alex@openfaas.com>
This commit is contained in:
Alex Ellis (OpenFaaS Ltd)
2023-10-23 11:29:19 +01:00
parent 479285caf6
commit 9ba4a73d5d
350 changed files with 22981 additions and 3972 deletions
Download Patch File Download Diff File Expand all files Collapse all files

291
gateway/vendor/golang.org/x/crypto/blake2b/blake2b.go generated vendored Normal file
View File

@ -0,0 +1,291 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package blake2b implements the BLAKE2b hash algorithm defined by RFC 7693
// and the extendable output function (XOF) BLAKE2Xb.
//
// BLAKE2b is optimized for 64-bit platforms—including NEON-enabled ARMs—and
// produces digests of any size between 1 and 64 bytes.
// For a detailed specification of BLAKE2b see https://blake2.net/blake2.pdf
// and for BLAKE2Xb see https://blake2.net/blake2x.pdf
//
// If you aren't sure which function you need, use BLAKE2b (Sum512 or New512).
// If you need a secret-key MAC (message authentication code), use the New512
// function with a non-nil key.
//
// BLAKE2X is a construction to compute hash values larger than 64 bytes. It
// can produce hash values between 0 and 4 GiB.
package blake2b
import (
"encoding/binary"
"errors"
"hash"
)
const (
// The blocksize of BLAKE2b in bytes.
BlockSize = 128
// The hash size of BLAKE2b-512 in bytes.
Size = 64
// The hash size of BLAKE2b-384 in bytes.
Size384 = 48
// The hash size of BLAKE2b-256 in bytes.
Size256 = 32
)
var (
useAVX2 bool
useAVX bool
useSSE4 bool
)
var (
errKeySize = errors.New("blake2b: invalid key size")
errHashSize = errors.New("blake2b: invalid hash size")
)
var iv = [8]uint64{
0x6a09e667f3bcc908, 0xbb67ae8584caa73b, 0x3c6ef372fe94f82b, 0xa54ff53a5f1d36f1,
0x510e527fade682d1, 0x9b05688c2b3e6c1f, 0x1f83d9abfb41bd6b, 0x5be0cd19137e2179,
}
// Sum512 returns the BLAKE2b-512 checksum of the data.
func Sum512(data []byte) [Size]byte {
var sum [Size]byte
checkSum(&sum, Size, data)
return sum
}
// Sum384 returns the BLAKE2b-384 checksum of the data.
func Sum384(data []byte) [Size384]byte {
var sum [Size]byte
var sum384 [Size384]byte
checkSum(&sum, Size384, data)
copy(sum384[:], sum[:Size384])
return sum384
}
// Sum256 returns the BLAKE2b-256 checksum of the data.
func Sum256(data []byte) [Size256]byte {
var sum [Size]byte
var sum256 [Size256]byte
checkSum(&sum, Size256, data)
copy(sum256[:], sum[:Size256])
return sum256
}
// New512 returns a new hash.Hash computing the BLAKE2b-512 checksum. A non-nil
// key turns the hash into a MAC. The key must be between zero and 64 bytes long.
func New512(key []byte) (hash.Hash, error) { return newDigest(Size, key) }
// New384 returns a new hash.Hash computing the BLAKE2b-384 checksum. A non-nil
// key turns the hash into a MAC. The key must be between zero and 64 bytes long.
func New384(key []byte) (hash.Hash, error) { return newDigest(Size384, key) }
// New256 returns a new hash.Hash computing the BLAKE2b-256 checksum. A non-nil
// key turns the hash into a MAC. The key must be between zero and 64 bytes long.
func New256(key []byte) (hash.Hash, error) { return newDigest(Size256, key) }
// New returns a new hash.Hash computing the BLAKE2b checksum with a custom length.
// A non-nil key turns the hash into a MAC. The key must be between zero and 64 bytes long.
// The hash size can be a value between 1 and 64 but it is highly recommended to use
// values equal or greater than:
// - 32 if BLAKE2b is used as a hash function (The key is zero bytes long).
// - 16 if BLAKE2b is used as a MAC function (The key is at least 16 bytes long).
// When the key is nil, the returned hash.Hash implements BinaryMarshaler
// and BinaryUnmarshaler for state (de)serialization as documented by hash.Hash.
func New(size int, key []byte) (hash.Hash, error) { return newDigest(size, key) }
func newDigest(hashSize int, key []byte) (*digest, error) {
if hashSize < 1 || hashSize > Size {
return nil, errHashSize
}
if len(key) > Size {
return nil, errKeySize
}
d := &digest{
size: hashSize,
keyLen: len(key),
}
copy(d.key[:], key)
d.Reset()
return d, nil
}
func checkSum(sum *[Size]byte, hashSize int, data []byte) {
h := iv
h[0] ^= uint64(hashSize) | (1 << 16) | (1 << 24)
var c [2]uint64
if length := len(data); length > BlockSize {
n := length &^ (BlockSize - 1)
if length == n {
n -= BlockSize
}
hashBlocks(&h, &c, 0, data[:n])
data = data[n:]
}
var block [BlockSize]byte
offset := copy(block[:], data)
remaining := uint64(BlockSize - offset)
if c[0] < remaining {
c[1]--
}
c[0] -= remaining
hashBlocks(&h, &c, 0xFFFFFFFFFFFFFFFF, block[:])
for i, v := range h[:(hashSize+7)/8] {
binary.LittleEndian.PutUint64(sum[8*i:], v)
}
}
type digest struct {
h [8]uint64
c [2]uint64
size int
block [BlockSize]byte
offset int
key [BlockSize]byte
keyLen int
}
const (
magic = "b2b"
marshaledSize = len(magic) + 8*8 + 2*8 + 1 + BlockSize + 1
)
func (d *digest) MarshalBinary() ([]byte, error) {
if d.keyLen != 0 {
return nil, errors.New("crypto/blake2b: cannot marshal MACs")
}
b := make([]byte, 0, marshaledSize)
b = append(b, magic...)
for i := 0; i < 8; i++ {
b = appendUint64(b, d.h[i])
}
b = appendUint64(b, d.c[0])
b = appendUint64(b, d.c[1])
// Maximum value for size is 64
b = append(b, byte(d.size))
b = append(b, d.block[:]...)
b = append(b, byte(d.offset))
return b, nil
}
func (d *digest) UnmarshalBinary(b []byte) error {
if len(b) < len(magic) || string(b[:len(magic)]) != magic {
return errors.New("crypto/blake2b: invalid hash state identifier")
}
if len(b) != marshaledSize {
return errors.New("crypto/blake2b: invalid hash state size")
}
b = b[len(magic):]
for i := 0; i < 8; i++ {
b, d.h[i] = consumeUint64(b)
}
b, d.c[0] = consumeUint64(b)
b, d.c[1] = consumeUint64(b)
d.size = int(b[0])
b = b[1:]
copy(d.block[:], b[:BlockSize])
b = b[BlockSize:]
d.offset = int(b[0])
return nil
}
func (d *digest) BlockSize() int { return BlockSize }
func (d *digest) Size() int { return d.size }
func (d *digest) Reset() {
d.h = iv
d.h[0] ^= uint64(d.size) | (uint64(d.keyLen) << 8) | (1 << 16) | (1 << 24)
d.offset, d.c[0], d.c[1] = 0, 0, 0
if d.keyLen > 0 {
d.block = d.key
d.offset = BlockSize
}
}
func (d *digest) Write(p []byte) (n int, err error) {
n = len(p)
if d.offset > 0 {
remaining := BlockSize - d.offset
if n <= remaining {
d.offset += copy(d.block[d.offset:], p)
return
}
copy(d.block[d.offset:], p[:remaining])
hashBlocks(&d.h, &d.c, 0, d.block[:])
d.offset = 0
p = p[remaining:]
}
if length := len(p); length > BlockSize {
nn := length &^ (BlockSize - 1)
if length == nn {
nn -= BlockSize
}
hashBlocks(&d.h, &d.c, 0, p[:nn])
p = p[nn:]
}
if len(p) > 0 {
d.offset += copy(d.block[:], p)
}
return
}
func (d *digest) Sum(sum []byte) []byte {
var hash [Size]byte
d.finalize(&hash)
return append(sum, hash[:d.size]...)
}
func (d *digest) finalize(hash *[Size]byte) {
var block [BlockSize]byte
copy(block[:], d.block[:d.offset])
remaining := uint64(BlockSize - d.offset)
c := d.c
if c[0] < remaining {
c[1]--
}
c[0] -= remaining
h := d.h
hashBlocks(&h, &c, 0xFFFFFFFFFFFFFFFF, block[:])
for i, v := range h {
binary.LittleEndian.PutUint64(hash[8*i:], v)
}
}
func appendUint64(b []byte, x uint64) []byte {
var a [8]byte
binary.BigEndian.PutUint64(a[:], x)
return append(b, a[:]...)
}
func appendUint32(b []byte, x uint32) []byte {
var a [4]byte
binary.BigEndian.PutUint32(a[:], x)
return append(b, a[:]...)
}
func consumeUint64(b []byte) ([]byte, uint64) {
x := binary.BigEndian.Uint64(b)
return b[8:], x
}
func consumeUint32(b []byte) ([]byte, uint32) {
x := binary.BigEndian.Uint32(b)
return b[4:], x
}

38
gateway/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.go generated vendored Normal file
View File

@ -0,0 +1,38 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build go1.7 && amd64 && gc && !purego
// +build go1.7,amd64,gc,!purego
package blake2b
import "golang.org/x/sys/cpu"
func init() {
useAVX2 = cpu.X86.HasAVX2
useAVX = cpu.X86.HasAVX
useSSE4 = cpu.X86.HasSSE41
}
//go:noescape
func hashBlocksAVX2(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
//go:noescape
func hashBlocksAVX(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
//go:noescape
func hashBlocksSSE4(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
func hashBlocks(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) {
switch {
case useAVX2:
hashBlocksAVX2(h, c, flag, blocks)
case useAVX:
hashBlocksAVX(h, c, flag, blocks)
case useSSE4:
hashBlocksSSE4(h, c, flag, blocks)
default:
hashBlocksGeneric(h, c, flag, blocks)
}
}

745
gateway/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.s generated vendored Normal file
View File

@ -0,0 +1,745 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build go1.7 && amd64 && gc && !purego
// +build go1.7,amd64,gc,!purego
#include "textflag.h"
DATA ·AVX2_iv0<>+0x00(SB)/8, $0x6a09e667f3bcc908
DATA ·AVX2_iv0<>+0x08(SB)/8, $0xbb67ae8584caa73b
DATA ·AVX2_iv0<>+0x10(SB)/8, $0x3c6ef372fe94f82b
DATA ·AVX2_iv0<>+0x18(SB)/8, $0xa54ff53a5f1d36f1
GLOBL ·AVX2_iv0<>(SB), (NOPTR+RODATA), $32
DATA ·AVX2_iv1<>+0x00(SB)/8, $0x510e527fade682d1
DATA ·AVX2_iv1<>+0x08(SB)/8, $0x9b05688c2b3e6c1f
DATA ·AVX2_iv1<>+0x10(SB)/8, $0x1f83d9abfb41bd6b
DATA ·AVX2_iv1<>+0x18(SB)/8, $0x5be0cd19137e2179
GLOBL ·AVX2_iv1<>(SB), (NOPTR+RODATA), $32
DATA ·AVX2_c40<>+0x00(SB)/8, $0x0201000706050403
DATA ·AVX2_c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b
DATA ·AVX2_c40<>+0x10(SB)/8, $0x0201000706050403
DATA ·AVX2_c40<>+0x18(SB)/8, $0x0a09080f0e0d0c0b
GLOBL ·AVX2_c40<>(SB), (NOPTR+RODATA), $32
DATA ·AVX2_c48<>+0x00(SB)/8, $0x0100070605040302
DATA ·AVX2_c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a
DATA ·AVX2_c48<>+0x10(SB)/8, $0x0100070605040302
DATA ·AVX2_c48<>+0x18(SB)/8, $0x09080f0e0d0c0b0a
GLOBL ·AVX2_c48<>(SB), (NOPTR+RODATA), $32
DATA ·AVX_iv0<>+0x00(SB)/8, $0x6a09e667f3bcc908
DATA ·AVX_iv0<>+0x08(SB)/8, $0xbb67ae8584caa73b
GLOBL ·AVX_iv0<>(SB), (NOPTR+RODATA), $16
DATA ·AVX_iv1<>+0x00(SB)/8, $0x3c6ef372fe94f82b
DATA ·AVX_iv1<>+0x08(SB)/8, $0xa54ff53a5f1d36f1
GLOBL ·AVX_iv1<>(SB), (NOPTR+RODATA), $16
DATA ·AVX_iv2<>+0x00(SB)/8, $0x510e527fade682d1
DATA ·AVX_iv2<>+0x08(SB)/8, $0x9b05688c2b3e6c1f
GLOBL ·AVX_iv2<>(SB), (NOPTR+RODATA), $16
DATA ·AVX_iv3<>+0x00(SB)/8, $0x1f83d9abfb41bd6b
DATA ·AVX_iv3<>+0x08(SB)/8, $0x5be0cd19137e2179
GLOBL ·AVX_iv3<>(SB), (NOPTR+RODATA), $16
DATA ·AVX_c40<>+0x00(SB)/8, $0x0201000706050403
DATA ·AVX_c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b
GLOBL ·AVX_c40<>(SB), (NOPTR+RODATA), $16
DATA ·AVX_c48<>+0x00(SB)/8, $0x0100070605040302
DATA ·AVX_c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a
GLOBL ·AVX_c48<>(SB), (NOPTR+RODATA), $16
#define VPERMQ_0x39_Y1_Y1 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xc9; BYTE $0x39
#define VPERMQ_0x93_Y1_Y1 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xc9; BYTE $0x93
#define VPERMQ_0x4E_Y2_Y2 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xd2; BYTE $0x4e
#define VPERMQ_0x93_Y3_Y3 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xdb; BYTE $0x93
#define VPERMQ_0x39_Y3_Y3 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xdb; BYTE $0x39
#define ROUND_AVX2(m0, m1, m2, m3, t, c40, c48) \
VPADDQ m0, Y0, Y0; \
VPADDQ Y1, Y0, Y0; \
VPXOR Y0, Y3, Y3; \
VPSHUFD $-79, Y3, Y3; \
VPADDQ Y3, Y2, Y2; \
VPXOR Y2, Y1, Y1; \
VPSHUFB c40, Y1, Y1; \
VPADDQ m1, Y0, Y0; \
VPADDQ Y1, Y0, Y0; \
VPXOR Y0, Y3, Y3; \
VPSHUFB c48, Y3, Y3; \
VPADDQ Y3, Y2, Y2; \
VPXOR Y2, Y1, Y1; \
VPADDQ Y1, Y1, t; \
VPSRLQ $63, Y1, Y1; \
VPXOR t, Y1, Y1; \
VPERMQ_0x39_Y1_Y1; \
VPERMQ_0x4E_Y2_Y2; \
VPERMQ_0x93_Y3_Y3; \
VPADDQ m2, Y0, Y0; \
VPADDQ Y1, Y0, Y0; \
VPXOR Y0, Y3, Y3; \
VPSHUFD $-79, Y3, Y3; \
VPADDQ Y3, Y2, Y2; \
VPXOR Y2, Y1, Y1; \
VPSHUFB c40, Y1, Y1; \
VPADDQ m3, Y0, Y0; \
VPADDQ Y1, Y0, Y0; \
VPXOR Y0, Y3, Y3; \
VPSHUFB c48, Y3, Y3; \
VPADDQ Y3, Y2, Y2; \
VPXOR Y2, Y1, Y1; \
VPADDQ Y1, Y1, t; \
VPSRLQ $63, Y1, Y1; \
VPXOR t, Y1, Y1; \
VPERMQ_0x39_Y3_Y3; \
VPERMQ_0x4E_Y2_Y2; \
VPERMQ_0x93_Y1_Y1
#define VMOVQ_SI_X11_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x1E
#define VMOVQ_SI_X12_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x26
#define VMOVQ_SI_X13_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x2E
#define VMOVQ_SI_X14_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x36
#define VMOVQ_SI_X15_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x3E
#define VMOVQ_SI_X11(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x5E; BYTE $n
#define VMOVQ_SI_X12(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x66; BYTE $n
#define VMOVQ_SI_X13(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x6E; BYTE $n
#define VMOVQ_SI_X14(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x76; BYTE $n
#define VMOVQ_SI_X15(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x7E; BYTE $n
#define VPINSRQ_1_SI_X11_0 BYTE $0xC4; BYTE $0x63; BYTE $0xA1; BYTE $0x22; BYTE $0x1E; BYTE $0x01
#define VPINSRQ_1_SI_X12_0 BYTE $0xC4; BYTE $0x63; BYTE $0x99; BYTE $0x22; BYTE $0x26; BYTE $0x01
#define VPINSRQ_1_SI_X13_0 BYTE $0xC4; BYTE $0x63; BYTE $0x91; BYTE $0x22; BYTE $0x2E; BYTE $0x01
#define VPINSRQ_1_SI_X14_0 BYTE $0xC4; BYTE $0x63; BYTE $0x89; BYTE $0x22; BYTE $0x36; BYTE $0x01
#define VPINSRQ_1_SI_X15_0 BYTE $0xC4; BYTE $0x63; BYTE $0x81; BYTE $0x22; BYTE $0x3E; BYTE $0x01
#define VPINSRQ_1_SI_X11(n) BYTE $0xC4; BYTE $0x63; BYTE $0xA1; BYTE $0x22; BYTE $0x5E; BYTE $n; BYTE $0x01
#define VPINSRQ_1_SI_X12(n) BYTE $0xC4; BYTE $0x63; BYTE $0x99; BYTE $0x22; BYTE $0x66; BYTE $n; BYTE $0x01
#define VPINSRQ_1_SI_X13(n) BYTE $0xC4; BYTE $0x63; BYTE $0x91; BYTE $0x22; BYTE $0x6E; BYTE $n; BYTE $0x01
#define VPINSRQ_1_SI_X14(n) BYTE $0xC4; BYTE $0x63; BYTE $0x89; BYTE $0x22; BYTE $0x76; BYTE $n; BYTE $0x01
#define VPINSRQ_1_SI_X15(n) BYTE $0xC4; BYTE $0x63; BYTE $0x81; BYTE $0x22; BYTE $0x7E; BYTE $n; BYTE $0x01
#define VMOVQ_R8_X15 BYTE $0xC4; BYTE $0x41; BYTE $0xF9; BYTE $0x6E; BYTE $0xF8
#define VPINSRQ_1_R9_X15 BYTE $0xC4; BYTE $0x43; BYTE $0x81; BYTE $0x22; BYTE $0xF9; BYTE $0x01
// load msg: Y12 = (i0, i1, i2, i3)
// i0, i1, i2, i3 must not be 0
#define LOAD_MSG_AVX2_Y12(i0, i1, i2, i3) \
VMOVQ_SI_X12(i0*8); \
VMOVQ_SI_X11(i2*8); \
VPINSRQ_1_SI_X12(i1*8); \
VPINSRQ_1_SI_X11(i3*8); \
VINSERTI128 $1, X11, Y12, Y12
// load msg: Y13 = (i0, i1, i2, i3)
// i0, i1, i2, i3 must not be 0
#define LOAD_MSG_AVX2_Y13(i0, i1, i2, i3) \
VMOVQ_SI_X13(i0*8); \
VMOVQ_SI_X11(i2*8); \
VPINSRQ_1_SI_X13(i1*8); \
VPINSRQ_1_SI_X11(i3*8); \
VINSERTI128 $1, X11, Y13, Y13
// load msg: Y14 = (i0, i1, i2, i3)
// i0, i1, i2, i3 must not be 0
#define LOAD_MSG_AVX2_Y14(i0, i1, i2, i3) \
VMOVQ_SI_X14(i0*8); \
VMOVQ_SI_X11(i2*8); \
VPINSRQ_1_SI_X14(i1*8); \
VPINSRQ_1_SI_X11(i3*8); \
VINSERTI128 $1, X11, Y14, Y14
// load msg: Y15 = (i0, i1, i2, i3)
// i0, i1, i2, i3 must not be 0
#define LOAD_MSG_AVX2_Y15(i0, i1, i2, i3) \
VMOVQ_SI_X15(i0*8); \
VMOVQ_SI_X11(i2*8); \
VPINSRQ_1_SI_X15(i1*8); \
VPINSRQ_1_SI_X11(i3*8); \
VINSERTI128 $1, X11, Y15, Y15
#define LOAD_MSG_AVX2_0_2_4_6_1_3_5_7_8_10_12_14_9_11_13_15() \
VMOVQ_SI_X12_0; \
VMOVQ_SI_X11(4*8); \
VPINSRQ_1_SI_X12(2*8); \
VPINSRQ_1_SI_X11(6*8); \
VINSERTI128 $1, X11, Y12, Y12; \
LOAD_MSG_AVX2_Y13(1, 3, 5, 7); \
LOAD_MSG_AVX2_Y14(8, 10, 12, 14); \
LOAD_MSG_AVX2_Y15(9, 11, 13, 15)
#define LOAD_MSG_AVX2_14_4_9_13_10_8_15_6_1_0_11_5_12_2_7_3() \
LOAD_MSG_AVX2_Y12(14, 4, 9, 13); \
LOAD_MSG_AVX2_Y13(10, 8, 15, 6); \
VMOVQ_SI_X11(11*8); \
VPSHUFD $0x4E, 0*8(SI), X14; \
VPINSRQ_1_SI_X11(5*8); \
VINSERTI128 $1, X11, Y14, Y14; \
LOAD_MSG_AVX2_Y15(12, 2, 7, 3)
#define LOAD_MSG_AVX2_11_12_5_15_8_0_2_13_10_3_7_9_14_6_1_4() \
VMOVQ_SI_X11(5*8); \
VMOVDQU 11*8(SI), X12; \
VPINSRQ_1_SI_X11(15*8); \
VINSERTI128 $1, X11, Y12, Y12; \
VMOVQ_SI_X13(8*8); \
VMOVQ_SI_X11(2*8); \
VPINSRQ_1_SI_X13_0; \
VPINSRQ_1_SI_X11(13*8); \
VINSERTI128 $1, X11, Y13, Y13; \
LOAD_MSG_AVX2_Y14(10, 3, 7, 9); \
LOAD_MSG_AVX2_Y15(14, 6, 1, 4)
#define LOAD_MSG_AVX2_7_3_13_11_9_1_12_14_2_5_4_15_6_10_0_8() \
LOAD_MSG_AVX2_Y12(7, 3, 13, 11); \
LOAD_MSG_AVX2_Y13(9, 1, 12, 14); \
LOAD_MSG_AVX2_Y14(2, 5, 4, 15); \
VMOVQ_SI_X15(6*8); \
VMOVQ_SI_X11_0; \
VPINSRQ_1_SI_X15(10*8); \
VPINSRQ_1_SI_X11(8*8); \
VINSERTI128 $1, X11, Y15, Y15
#define LOAD_MSG_AVX2_9_5_2_10_0_7_4_15_14_11_6_3_1_12_8_13() \
LOAD_MSG_AVX2_Y12(9, 5, 2, 10); \
VMOVQ_SI_X13_0; \
VMOVQ_SI_X11(4*8); \
VPINSRQ_1_SI_X13(7*8); \
VPINSRQ_1_SI_X11(15*8); \
VINSERTI128 $1, X11, Y13, Y13; \
LOAD_MSG_AVX2_Y14(14, 11, 6, 3); \
LOAD_MSG_AVX2_Y15(1, 12, 8, 13)
#define LOAD_MSG_AVX2_2_6_0_8_12_10_11_3_4_7_15_1_13_5_14_9() \
VMOVQ_SI_X12(2*8); \
VMOVQ_SI_X11_0; \
VPINSRQ_1_SI_X12(6*8); \
VPINSRQ_1_SI_X11(8*8); \
VINSERTI128 $1, X11, Y12, Y12; \
LOAD_MSG_AVX2_Y13(12, 10, 11, 3); \
LOAD_MSG_AVX2_Y14(4, 7, 15, 1); \
LOAD_MSG_AVX2_Y15(13, 5, 14, 9)
#define LOAD_MSG_AVX2_12_1_14_4_5_15_13_10_0_6_9_8_7_3_2_11() \
LOAD_MSG_AVX2_Y12(12, 1, 14, 4); \
LOAD_MSG_AVX2_Y13(5, 15, 13, 10); \
VMOVQ_SI_X14_0; \
VPSHUFD $0x4E, 8*8(SI), X11; \
VPINSRQ_1_SI_X14(6*8); \
VINSERTI128 $1, X11, Y14, Y14; \
LOAD_MSG_AVX2_Y15(7, 3, 2, 11)
#define LOAD_MSG_AVX2_13_7_12_3_11_14_1_9_5_15_8_2_0_4_6_10() \
LOAD_MSG_AVX2_Y12(13, 7, 12, 3); \
LOAD_MSG_AVX2_Y13(11, 14, 1, 9); \
LOAD_MSG_AVX2_Y14(5, 15, 8, 2); \
VMOVQ_SI_X15_0; \
VMOVQ_SI_X11(6*8); \
VPINSRQ_1_SI_X15(4*8); \
VPINSRQ_1_SI_X11(10*8); \
VINSERTI128 $1, X11, Y15, Y15
#define LOAD_MSG_AVX2_6_14_11_0_15_9_3_8_12_13_1_10_2_7_4_5() \
VMOVQ_SI_X12(6*8); \
VMOVQ_SI_X11(11*8); \
VPINSRQ_1_SI_X12(14*8); \
VPINSRQ_1_SI_X11_0; \
VINSERTI128 $1, X11, Y12, Y12; \
LOAD_MSG_AVX2_Y13(15, 9, 3, 8); \
VMOVQ_SI_X11(1*8); \
VMOVDQU 12*8(SI), X14; \
VPINSRQ_1_SI_X11(10*8); \
VINSERTI128 $1, X11, Y14, Y14; \
VMOVQ_SI_X15(2*8); \
VMOVDQU 4*8(SI), X11; \
VPINSRQ_1_SI_X15(7*8); \
VINSERTI128 $1, X11, Y15, Y15
#define LOAD_MSG_AVX2_10_8_7_1_2_4_6_5_15_9_3_13_11_14_12_0() \
LOAD_MSG_AVX2_Y12(10, 8, 7, 1); \
VMOVQ_SI_X13(2*8); \
VPSHUFD $0x4E, 5*8(SI), X11; \
VPINSRQ_1_SI_X13(4*8); \
VINSERTI128 $1, X11, Y13, Y13; \
LOAD_MSG_AVX2_Y14(15, 9, 3, 13); \
VMOVQ_SI_X15(11*8); \
VMOVQ_SI_X11(12*8); \
VPINSRQ_1_SI_X15(14*8); \
VPINSRQ_1_SI_X11_0; \
VINSERTI128 $1, X11, Y15, Y15
// func hashBlocksAVX2(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
TEXT ·hashBlocksAVX2(SB), 4, $320-48 // frame size = 288 + 32 byte alignment
MOVQ h+0(FP), AX
MOVQ c+8(FP), BX
MOVQ flag+16(FP), CX
MOVQ blocks_base+24(FP), SI
MOVQ blocks_len+32(FP), DI
MOVQ SP, DX
ADDQ $31, DX
ANDQ $~31, DX
MOVQ CX, 16(DX)
XORQ CX, CX
MOVQ CX, 24(DX)
VMOVDQU ·AVX2_c40<>(SB), Y4
VMOVDQU ·AVX2_c48<>(SB), Y5
VMOVDQU 0(AX), Y8
VMOVDQU 32(AX), Y9
VMOVDQU ·AVX2_iv0<>(SB), Y6
VMOVDQU ·AVX2_iv1<>(SB), Y7
MOVQ 0(BX), R8
MOVQ 8(BX), R9
MOVQ R9, 8(DX)
loop:
ADDQ $128, R8
MOVQ R8, 0(DX)
CMPQ R8, $128
JGE noinc
INCQ R9
MOVQ R9, 8(DX)
noinc:
VMOVDQA Y8, Y0
VMOVDQA Y9, Y1
VMOVDQA Y6, Y2
VPXOR 0(DX), Y7, Y3
LOAD_MSG_AVX2_0_2_4_6_1_3_5_7_8_10_12_14_9_11_13_15()
VMOVDQA Y12, 32(DX)
VMOVDQA Y13, 64(DX)
VMOVDQA Y14, 96(DX)
VMOVDQA Y15, 128(DX)
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_14_4_9_13_10_8_15_6_1_0_11_5_12_2_7_3()
VMOVDQA Y12, 160(DX)
VMOVDQA Y13, 192(DX)
VMOVDQA Y14, 224(DX)
VMOVDQA Y15, 256(DX)
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_11_12_5_15_8_0_2_13_10_3_7_9_14_6_1_4()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_7_3_13_11_9_1_12_14_2_5_4_15_6_10_0_8()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_9_5_2_10_0_7_4_15_14_11_6_3_1_12_8_13()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_2_6_0_8_12_10_11_3_4_7_15_1_13_5_14_9()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_12_1_14_4_5_15_13_10_0_6_9_8_7_3_2_11()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_13_7_12_3_11_14_1_9_5_15_8_2_0_4_6_10()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_6_14_11_0_15_9_3_8_12_13_1_10_2_7_4_5()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
LOAD_MSG_AVX2_10_8_7_1_2_4_6_5_15_9_3_13_11_14_12_0()
ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5)
ROUND_AVX2(32(DX), 64(DX), 96(DX), 128(DX), Y10, Y4, Y5)
ROUND_AVX2(160(DX), 192(DX), 224(DX), 256(DX), Y10, Y4, Y5)
VPXOR Y0, Y8, Y8
VPXOR Y1, Y9, Y9
VPXOR Y2, Y8, Y8
VPXOR Y3, Y9, Y9
LEAQ 128(SI), SI
SUBQ $128, DI
JNE loop
MOVQ R8, 0(BX)
MOVQ R9, 8(BX)
VMOVDQU Y8, 0(AX)
VMOVDQU Y9, 32(AX)
VZEROUPPER
RET
#define VPUNPCKLQDQ_X2_X2_X15 BYTE $0xC5; BYTE $0x69; BYTE $0x6C; BYTE $0xFA
#define VPUNPCKLQDQ_X3_X3_X15 BYTE $0xC5; BYTE $0x61; BYTE $0x6C; BYTE $0xFB
#define VPUNPCKLQDQ_X7_X7_X15 BYTE $0xC5; BYTE $0x41; BYTE $0x6C; BYTE $0xFF
#define VPUNPCKLQDQ_X13_X13_X15 BYTE $0xC4; BYTE $0x41; BYTE $0x11; BYTE $0x6C; BYTE $0xFD
#define VPUNPCKLQDQ_X14_X14_X15 BYTE $0xC4; BYTE $0x41; BYTE $0x09; BYTE $0x6C; BYTE $0xFE
#define VPUNPCKHQDQ_X15_X2_X2 BYTE $0xC4; BYTE $0xC1; BYTE $0x69; BYTE $0x6D; BYTE $0xD7
#define VPUNPCKHQDQ_X15_X3_X3 BYTE $0xC4; BYTE $0xC1; BYTE $0x61; BYTE $0x6D; BYTE $0xDF
#define VPUNPCKHQDQ_X15_X6_X6 BYTE $0xC4; BYTE $0xC1; BYTE $0x49; BYTE $0x6D; BYTE $0xF7
#define VPUNPCKHQDQ_X15_X7_X7 BYTE $0xC4; BYTE $0xC1; BYTE $0x41; BYTE $0x6D; BYTE $0xFF
#define VPUNPCKHQDQ_X15_X3_X2 BYTE $0xC4; BYTE $0xC1; BYTE $0x61; BYTE $0x6D; BYTE $0xD7
#define VPUNPCKHQDQ_X15_X7_X6 BYTE $0xC4; BYTE $0xC1; BYTE $0x41; BYTE $0x6D; BYTE $0xF7
#define VPUNPCKHQDQ_X15_X13_X3 BYTE $0xC4; BYTE $0xC1; BYTE $0x11; BYTE $0x6D; BYTE $0xDF
#define VPUNPCKHQDQ_X15_X13_X7 BYTE $0xC4; BYTE $0xC1; BYTE $0x11; BYTE $0x6D; BYTE $0xFF
#define SHUFFLE_AVX() \
VMOVDQA X6, X13; \
VMOVDQA X2, X14; \
VMOVDQA X4, X6; \
VPUNPCKLQDQ_X13_X13_X15; \
VMOVDQA X5, X4; \
VMOVDQA X6, X5; \
VPUNPCKHQDQ_X15_X7_X6; \
VPUNPCKLQDQ_X7_X7_X15; \
VPUNPCKHQDQ_X15_X13_X7; \
VPUNPCKLQDQ_X3_X3_X15; \
VPUNPCKHQDQ_X15_X2_X2; \
VPUNPCKLQDQ_X14_X14_X15; \
VPUNPCKHQDQ_X15_X3_X3; \
#define SHUFFLE_AVX_INV() \
VMOVDQA X2, X13; \
VMOVDQA X4, X14; \
VPUNPCKLQDQ_X2_X2_X15; \
VMOVDQA X5, X4; \
VPUNPCKHQDQ_X15_X3_X2; \
VMOVDQA X14, X5; \
VPUNPCKLQDQ_X3_X3_X15; \
VMOVDQA X6, X14; \
VPUNPCKHQDQ_X15_X13_X3; \
VPUNPCKLQDQ_X7_X7_X15; \
VPUNPCKHQDQ_X15_X6_X6; \
VPUNPCKLQDQ_X14_X14_X15; \
VPUNPCKHQDQ_X15_X7_X7; \
#define HALF_ROUND_AVX(v0, v1, v2, v3, v4, v5, v6, v7, m0, m1, m2, m3, t0, c40, c48) \
VPADDQ m0, v0, v0; \
VPADDQ v2, v0, v0; \
VPADDQ m1, v1, v1; \
VPADDQ v3, v1, v1; \
VPXOR v0, v6, v6; \
VPXOR v1, v7, v7; \
VPSHUFD $-79, v6, v6; \
VPSHUFD $-79, v7, v7; \
VPADDQ v6, v4, v4; \
VPADDQ v7, v5, v5; \
VPXOR v4, v2, v2; \
VPXOR v5, v3, v3; \
VPSHUFB c40, v2, v2; \
VPSHUFB c40, v3, v3; \
VPADDQ m2, v0, v0; \
VPADDQ v2, v0, v0; \
VPADDQ m3, v1, v1; \
VPADDQ v3, v1, v1; \
VPXOR v0, v6, v6; \
VPXOR v1, v7, v7; \
VPSHUFB c48, v6, v6; \
VPSHUFB c48, v7, v7; \
VPADDQ v6, v4, v4; \
VPADDQ v7, v5, v5; \
VPXOR v4, v2, v2; \
VPXOR v5, v3, v3; \
VPADDQ v2, v2, t0; \
VPSRLQ $63, v2, v2; \
VPXOR t0, v2, v2; \
VPADDQ v3, v3, t0; \
VPSRLQ $63, v3, v3; \
VPXOR t0, v3, v3
// load msg: X12 = (i0, i1), X13 = (i2, i3), X14 = (i4, i5), X15 = (i6, i7)
// i0, i1, i2, i3, i4, i5, i6, i7 must not be 0
#define LOAD_MSG_AVX(i0, i1, i2, i3, i4, i5, i6, i7) \
VMOVQ_SI_X12(i0*8); \
VMOVQ_SI_X13(i2*8); \
VMOVQ_SI_X14(i4*8); \
VMOVQ_SI_X15(i6*8); \
VPINSRQ_1_SI_X12(i1*8); \
VPINSRQ_1_SI_X13(i3*8); \
VPINSRQ_1_SI_X14(i5*8); \
VPINSRQ_1_SI_X15(i7*8)
// load msg: X12 = (0, 2), X13 = (4, 6), X14 = (1, 3), X15 = (5, 7)
#define LOAD_MSG_AVX_0_2_4_6_1_3_5_7() \
VMOVQ_SI_X12_0; \
VMOVQ_SI_X13(4*8); \
VMOVQ_SI_X14(1*8); \
VMOVQ_SI_X15(5*8); \
VPINSRQ_1_SI_X12(2*8); \
VPINSRQ_1_SI_X13(6*8); \
VPINSRQ_1_SI_X14(3*8); \
VPINSRQ_1_SI_X15(7*8)
// load msg: X12 = (1, 0), X13 = (11, 5), X14 = (12, 2), X15 = (7, 3)
#define LOAD_MSG_AVX_1_0_11_5_12_2_7_3() \
VPSHUFD $0x4E, 0*8(SI), X12; \
VMOVQ_SI_X13(11*8); \
VMOVQ_SI_X14(12*8); \
VMOVQ_SI_X15(7*8); \
VPINSRQ_1_SI_X13(5*8); \
VPINSRQ_1_SI_X14(2*8); \
VPINSRQ_1_SI_X15(3*8)
// load msg: X12 = (11, 12), X13 = (5, 15), X14 = (8, 0), X15 = (2, 13)
#define LOAD_MSG_AVX_11_12_5_15_8_0_2_13() \
VMOVDQU 11*8(SI), X12; \
VMOVQ_SI_X13(5*8); \
VMOVQ_SI_X14(8*8); \
VMOVQ_SI_X15(2*8); \
VPINSRQ_1_SI_X13(15*8); \
VPINSRQ_1_SI_X14_0; \
VPINSRQ_1_SI_X15(13*8)
// load msg: X12 = (2, 5), X13 = (4, 15), X14 = (6, 10), X15 = (0, 8)
#define LOAD_MSG_AVX_2_5_4_15_6_10_0_8() \
VMOVQ_SI_X12(2*8); \
VMOVQ_SI_X13(4*8); \
VMOVQ_SI_X14(6*8); \
VMOVQ_SI_X15_0; \
VPINSRQ_1_SI_X12(5*8); \
VPINSRQ_1_SI_X13(15*8); \
VPINSRQ_1_SI_X14(10*8); \
VPINSRQ_1_SI_X15(8*8)
// load msg: X12 = (9, 5), X13 = (2, 10), X14 = (0, 7), X15 = (4, 15)
#define LOAD_MSG_AVX_9_5_2_10_0_7_4_15() \
VMOVQ_SI_X12(9*8); \
VMOVQ_SI_X13(2*8); \
VMOVQ_SI_X14_0; \
VMOVQ_SI_X15(4*8); \
VPINSRQ_1_SI_X12(5*8); \
VPINSRQ_1_SI_X13(10*8); \
VPINSRQ_1_SI_X14(7*8); \
VPINSRQ_1_SI_X15(15*8)
// load msg: X12 = (2, 6), X13 = (0, 8), X14 = (12, 10), X15 = (11, 3)
#define LOAD_MSG_AVX_2_6_0_8_12_10_11_3() \
VMOVQ_SI_X12(2*8); \
VMOVQ_SI_X13_0; \
VMOVQ_SI_X14(12*8); \
VMOVQ_SI_X15(11*8); \
VPINSRQ_1_SI_X12(6*8); \
VPINSRQ_1_SI_X13(8*8); \
VPINSRQ_1_SI_X14(10*8); \
VPINSRQ_1_SI_X15(3*8)
// load msg: X12 = (0, 6), X13 = (9, 8), X14 = (7, 3), X15 = (2, 11)
#define LOAD_MSG_AVX_0_6_9_8_7_3_2_11() \
MOVQ 0*8(SI), X12; \
VPSHUFD $0x4E, 8*8(SI), X13; \
MOVQ 7*8(SI), X14; \
MOVQ 2*8(SI), X15; \
VPINSRQ_1_SI_X12(6*8); \
VPINSRQ_1_SI_X14(3*8); \
VPINSRQ_1_SI_X15(11*8)
// load msg: X12 = (6, 14), X13 = (11, 0), X14 = (15, 9), X15 = (3, 8)
#define LOAD_MSG_AVX_6_14_11_0_15_9_3_8() \
MOVQ 6*8(SI), X12; \
MOVQ 11*8(SI), X13; \
MOVQ 15*8(SI), X14; \
MOVQ 3*8(SI), X15; \
VPINSRQ_1_SI_X12(14*8); \
VPINSRQ_1_SI_X13_0; \
VPINSRQ_1_SI_X14(9*8); \
VPINSRQ_1_SI_X15(8*8)
// load msg: X12 = (5, 15), X13 = (8, 2), X14 = (0, 4), X15 = (6, 10)
#define LOAD_MSG_AVX_5_15_8_2_0_4_6_10() \
MOVQ 5*8(SI), X12; \
MOVQ 8*8(SI), X13; \
MOVQ 0*8(SI), X14; \
MOVQ 6*8(SI), X15; \
VPINSRQ_1_SI_X12(15*8); \
VPINSRQ_1_SI_X13(2*8); \
VPINSRQ_1_SI_X14(4*8); \
VPINSRQ_1_SI_X15(10*8)
// load msg: X12 = (12, 13), X13 = (1, 10), X14 = (2, 7), X15 = (4, 5)
#define LOAD_MSG_AVX_12_13_1_10_2_7_4_5() \
VMOVDQU 12*8(SI), X12; \
MOVQ 1*8(SI), X13; \
MOVQ 2*8(SI), X14; \
VPINSRQ_1_SI_X13(10*8); \
VPINSRQ_1_SI_X14(7*8); \
VMOVDQU 4*8(SI), X15
// load msg: X12 = (15, 9), X13 = (3, 13), X14 = (11, 14), X15 = (12, 0)
#define LOAD_MSG_AVX_15_9_3_13_11_14_12_0() \
MOVQ 15*8(SI), X12; \
MOVQ 3*8(SI), X13; \
MOVQ 11*8(SI), X14; \
MOVQ 12*8(SI), X15; \
VPINSRQ_1_SI_X12(9*8); \
VPINSRQ_1_SI_X13(13*8); \
VPINSRQ_1_SI_X14(14*8); \
VPINSRQ_1_SI_X15_0
// func hashBlocksAVX(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
TEXT ·hashBlocksAVX(SB), 4, $288-48 // frame size = 272 + 16 byte alignment
MOVQ h+0(FP), AX
MOVQ c+8(FP), BX
MOVQ flag+16(FP), CX
MOVQ blocks_base+24(FP), SI
MOVQ blocks_len+32(FP), DI
MOVQ SP, R10
ADDQ $15, R10
ANDQ $~15, R10
VMOVDQU ·AVX_c40<>(SB), X0
VMOVDQU ·AVX_c48<>(SB), X1
VMOVDQA X0, X8
VMOVDQA X1, X9
VMOVDQU ·AVX_iv3<>(SB), X0
VMOVDQA X0, 0(R10)
XORQ CX, 0(R10) // 0(R10) = ·AVX_iv3 ^ (CX || 0)
VMOVDQU 0(AX), X10
VMOVDQU 16(AX), X11
VMOVDQU 32(AX), X2
VMOVDQU 48(AX), X3
MOVQ 0(BX), R8
MOVQ 8(BX), R9
loop:
ADDQ $128, R8
CMPQ R8, $128
JGE noinc
INCQ R9
noinc:
VMOVQ_R8_X15
VPINSRQ_1_R9_X15
VMOVDQA X10, X0
VMOVDQA X11, X1
VMOVDQU ·AVX_iv0<>(SB), X4
VMOVDQU ·AVX_iv1<>(SB), X5
VMOVDQU ·AVX_iv2<>(SB), X6
VPXOR X15, X6, X6
VMOVDQA 0(R10), X7
LOAD_MSG_AVX_0_2_4_6_1_3_5_7()
VMOVDQA X12, 16(R10)
VMOVDQA X13, 32(R10)
VMOVDQA X14, 48(R10)
VMOVDQA X15, 64(R10)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX(8, 10, 12, 14, 9, 11, 13, 15)
VMOVDQA X12, 80(R10)
VMOVDQA X13, 96(R10)
VMOVDQA X14, 112(R10)
VMOVDQA X15, 128(R10)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX(14, 4, 9, 13, 10, 8, 15, 6)
VMOVDQA X12, 144(R10)
VMOVDQA X13, 160(R10)
VMOVDQA X14, 176(R10)
VMOVDQA X15, 192(R10)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX_1_0_11_5_12_2_7_3()
VMOVDQA X12, 208(R10)
VMOVDQA X13, 224(R10)
VMOVDQA X14, 240(R10)
VMOVDQA X15, 256(R10)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX_11_12_5_15_8_0_2_13()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX(10, 3, 7, 9, 14, 6, 1, 4)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX(7, 3, 13, 11, 9, 1, 12, 14)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX_2_5_4_15_6_10_0_8()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX_9_5_2_10_0_7_4_15()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX(14, 11, 6, 3, 1, 12, 8, 13)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX_2_6_0_8_12_10_11_3()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX(4, 7, 15, 1, 13, 5, 14, 9)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX(12, 1, 14, 4, 5, 15, 13, 10)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX_0_6_9_8_7_3_2_11()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX(13, 7, 12, 3, 11, 14, 1, 9)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX_5_15_8_2_0_4_6_10()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX_6_14_11_0_15_9_3_8()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX_12_13_1_10_2_7_4_5()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
LOAD_MSG_AVX(10, 8, 7, 1, 2, 4, 6, 5)
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX()
LOAD_MSG_AVX_15_9_3_13_11_14_12_0()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9)
SHUFFLE_AVX_INV()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 16(R10), 32(R10), 48(R10), 64(R10), X15, X8, X9)
SHUFFLE_AVX()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 80(R10), 96(R10), 112(R10), 128(R10), X15, X8, X9)
SHUFFLE_AVX_INV()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 144(R10), 160(R10), 176(R10), 192(R10), X15, X8, X9)
SHUFFLE_AVX()
HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 208(R10), 224(R10), 240(R10), 256(R10), X15, X8, X9)
SHUFFLE_AVX_INV()
VMOVDQU 32(AX), X14
VMOVDQU 48(AX), X15
VPXOR X0, X10, X10
VPXOR X1, X11, X11
VPXOR X2, X14, X14
VPXOR X3, X15, X15
VPXOR X4, X10, X10
VPXOR X5, X11, X11
VPXOR X6, X14, X2
VPXOR X7, X15, X3
VMOVDQU X2, 32(AX)
VMOVDQU X3, 48(AX)
LEAQ 128(SI), SI
SUBQ $128, DI
JNE loop
VMOVDQU X10, 0(AX)
VMOVDQU X11, 16(AX)
MOVQ R8, 0(BX)
MOVQ R9, 8(BX)
VZEROUPPER
RET

25
gateway/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.go generated vendored Normal file
View File

@ -0,0 +1,25 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !go1.7 && amd64 && gc && !purego
// +build !go1.7,amd64,gc,!purego
package blake2b
import "golang.org/x/sys/cpu"
func init() {
useSSE4 = cpu.X86.HasSSE41
}
//go:noescape
func hashBlocksSSE4(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
func hashBlocks(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) {
if useSSE4 {
hashBlocksSSE4(h, c, flag, blocks)
} else {
hashBlocksGeneric(h, c, flag, blocks)
}
}

279
gateway/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.s generated vendored Normal file
View File

@ -0,0 +1,279 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build amd64 && gc && !purego
// +build amd64,gc,!purego
#include "textflag.h"
DATA ·iv0<>+0x00(SB)/8, $0x6a09e667f3bcc908
DATA ·iv0<>+0x08(SB)/8, $0xbb67ae8584caa73b
GLOBL ·iv0<>(SB), (NOPTR+RODATA), $16
DATA ·iv1<>+0x00(SB)/8, $0x3c6ef372fe94f82b
DATA ·iv1<>+0x08(SB)/8, $0xa54ff53a5f1d36f1
GLOBL ·iv1<>(SB), (NOPTR+RODATA), $16
DATA ·iv2<>+0x00(SB)/8, $0x510e527fade682d1
DATA ·iv2<>+0x08(SB)/8, $0x9b05688c2b3e6c1f
GLOBL ·iv2<>(SB), (NOPTR+RODATA), $16
DATA ·iv3<>+0x00(SB)/8, $0x1f83d9abfb41bd6b
DATA ·iv3<>+0x08(SB)/8, $0x5be0cd19137e2179
GLOBL ·iv3<>(SB), (NOPTR+RODATA), $16
DATA ·c40<>+0x00(SB)/8, $0x0201000706050403
DATA ·c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b
GLOBL ·c40<>(SB), (NOPTR+RODATA), $16
DATA ·c48<>+0x00(SB)/8, $0x0100070605040302
DATA ·c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a
GLOBL ·c48<>(SB), (NOPTR+RODATA), $16
#define SHUFFLE(v2, v3, v4, v5, v6, v7, t1, t2) \
MOVO v4, t1; \
MOVO v5, v4; \
MOVO t1, v5; \
MOVO v6, t1; \
PUNPCKLQDQ v6, t2; \
PUNPCKHQDQ v7, v6; \
PUNPCKHQDQ t2, v6; \
PUNPCKLQDQ v7, t2; \
MOVO t1, v7; \
MOVO v2, t1; \
PUNPCKHQDQ t2, v7; \
PUNPCKLQDQ v3, t2; \
PUNPCKHQDQ t2, v2; \
PUNPCKLQDQ t1, t2; \
PUNPCKHQDQ t2, v3
#define SHUFFLE_INV(v2, v3, v4, v5, v6, v7, t1, t2) \
MOVO v4, t1; \
MOVO v5, v4; \
MOVO t1, v5; \
MOVO v2, t1; \
PUNPCKLQDQ v2, t2; \
PUNPCKHQDQ v3, v2; \
PUNPCKHQDQ t2, v2; \
PUNPCKLQDQ v3, t2; \
MOVO t1, v3; \
MOVO v6, t1; \
PUNPCKHQDQ t2, v3; \
PUNPCKLQDQ v7, t2; \
PUNPCKHQDQ t2, v6; \
PUNPCKLQDQ t1, t2; \
PUNPCKHQDQ t2, v7
#define HALF_ROUND(v0, v1, v2, v3, v4, v5, v6, v7, m0, m1, m2, m3, t0, c40, c48) \
PADDQ m0, v0; \
PADDQ m1, v1; \
PADDQ v2, v0; \
PADDQ v3, v1; \
PXOR v0, v6; \
PXOR v1, v7; \
PSHUFD $0xB1, v6, v6; \
PSHUFD $0xB1, v7, v7; \
PADDQ v6, v4; \
PADDQ v7, v5; \
PXOR v4, v2; \
PXOR v5, v3; \
PSHUFB c40, v2; \
PSHUFB c40, v3; \
PADDQ m2, v0; \
PADDQ m3, v1; \
PADDQ v2, v0; \
PADDQ v3, v1; \
PXOR v0, v6; \
PXOR v1, v7; \
PSHUFB c48, v6; \
PSHUFB c48, v7; \
PADDQ v6, v4; \
PADDQ v7, v5; \
PXOR v4, v2; \
PXOR v5, v3; \
MOVOU v2, t0; \
PADDQ v2, t0; \
PSRLQ $63, v2; \
PXOR t0, v2; \
MOVOU v3, t0; \
PADDQ v3, t0; \
PSRLQ $63, v3; \
PXOR t0, v3
#define LOAD_MSG(m0, m1, m2, m3, src, i0, i1, i2, i3, i4, i5, i6, i7) \
MOVQ i0*8(src), m0; \
PINSRQ $1, i1*8(src), m0; \
MOVQ i2*8(src), m1; \
PINSRQ $1, i3*8(src), m1; \
MOVQ i4*8(src), m2; \
PINSRQ $1, i5*8(src), m2; \
MOVQ i6*8(src), m3; \
PINSRQ $1, i7*8(src), m3
// func hashBlocksSSE4(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte)
TEXT ·hashBlocksSSE4(SB), 4, $288-48 // frame size = 272 + 16 byte alignment
MOVQ h+0(FP), AX
MOVQ c+8(FP), BX
MOVQ flag+16(FP), CX
MOVQ blocks_base+24(FP), SI
MOVQ blocks_len+32(FP), DI
MOVQ SP, R10
ADDQ $15, R10
ANDQ $~15, R10
MOVOU ·iv3<>(SB), X0
MOVO X0, 0(R10)
XORQ CX, 0(R10) // 0(R10) = ·iv3 ^ (CX || 0)
MOVOU ·c40<>(SB), X13
MOVOU ·c48<>(SB), X14
MOVOU 0(AX), X12
MOVOU 16(AX), X15
MOVQ 0(BX), R8
MOVQ 8(BX), R9
loop:
ADDQ $128, R8
CMPQ R8, $128
JGE noinc
INCQ R9
noinc:
MOVQ R8, X8
PINSRQ $1, R9, X8
MOVO X12, X0
MOVO X15, X1
MOVOU 32(AX), X2
MOVOU 48(AX), X3
MOVOU ·iv0<>(SB), X4
MOVOU ·iv1<>(SB), X5
MOVOU ·iv2<>(SB), X6
PXOR X8, X6
MOVO 0(R10), X7
LOAD_MSG(X8, X9, X10, X11, SI, 0, 2, 4, 6, 1, 3, 5, 7)
MOVO X8, 16(R10)
MOVO X9, 32(R10)
MOVO X10, 48(R10)
MOVO X11, 64(R10)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 8, 10, 12, 14, 9, 11, 13, 15)
MOVO X8, 80(R10)
MOVO X9, 96(R10)
MOVO X10, 112(R10)
MOVO X11, 128(R10)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 14, 4, 9, 13, 10, 8, 15, 6)
MOVO X8, 144(R10)
MOVO X9, 160(R10)
MOVO X10, 176(R10)
MOVO X11, 192(R10)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 1, 0, 11, 5, 12, 2, 7, 3)
MOVO X8, 208(R10)
MOVO X9, 224(R10)
MOVO X10, 240(R10)
MOVO X11, 256(R10)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 11, 12, 5, 15, 8, 0, 2, 13)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 10, 3, 7, 9, 14, 6, 1, 4)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 7, 3, 13, 11, 9, 1, 12, 14)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 2, 5, 4, 15, 6, 10, 0, 8)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 9, 5, 2, 10, 0, 7, 4, 15)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 14, 11, 6, 3, 1, 12, 8, 13)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 2, 6, 0, 8, 12, 10, 11, 3)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 4, 7, 15, 1, 13, 5, 14, 9)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 12, 1, 14, 4, 5, 15, 13, 10)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 0, 6, 9, 8, 7, 3, 2, 11)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 13, 7, 12, 3, 11, 14, 1, 9)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 5, 15, 8, 2, 0, 4, 6, 10)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 6, 14, 11, 0, 15, 9, 3, 8)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 12, 13, 1, 10, 2, 7, 4, 5)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 10, 8, 7, 1, 2, 4, 6, 5)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
LOAD_MSG(X8, X9, X10, X11, SI, 15, 9, 3, 13, 11, 14, 12, 0)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 16(R10), 32(R10), 48(R10), 64(R10), X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 80(R10), 96(R10), 112(R10), 128(R10), X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 144(R10), 160(R10), 176(R10), 192(R10), X11, X13, X14)
SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9)
HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 208(R10), 224(R10), 240(R10), 256(R10), X11, X13, X14)
SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9)
MOVOU 32(AX), X10
MOVOU 48(AX), X11
PXOR X0, X12
PXOR X1, X15
PXOR X2, X10
PXOR X3, X11
PXOR X4, X12
PXOR X5, X15
PXOR X6, X10
PXOR X7, X11
MOVOU X10, 32(AX)
MOVOU X11, 48(AX)
LEAQ 128(SI), SI
SUBQ $128, DI
JNE loop
MOVOU X12, 0(AX)
MOVOU X15, 16(AX)
MOVQ R8, 0(BX)
MOVQ R9, 8(BX)
RET

182
gateway/vendor/golang.org/x/crypto/blake2b/blake2b_generic.go generated vendored Normal file
View File

@ -0,0 +1,182 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package blake2b
import (
"encoding/binary"
"math/bits"
)
// the precomputed values for BLAKE2b
// there are 12 16-byte arrays - one for each round
// the entries are calculated from the sigma constants.
var precomputed = [12][16]byte{
{0, 2, 4, 6, 1, 3, 5, 7, 8, 10, 12, 14, 9, 11, 13, 15},
{14, 4, 9, 13, 10, 8, 15, 6, 1, 0, 11, 5, 12, 2, 7, 3},
{11, 12, 5, 15, 8, 0, 2, 13, 10, 3, 7, 9, 14, 6, 1, 4},
{7, 3, 13, 11, 9, 1, 12, 14, 2, 5, 4, 15, 6, 10, 0, 8},
{9, 5, 2, 10, 0, 7, 4, 15, 14, 11, 6, 3, 1, 12, 8, 13},
{2, 6, 0, 8, 12, 10, 11, 3, 4, 7, 15, 1, 13, 5, 14, 9},
{12, 1, 14, 4, 5, 15, 13, 10, 0, 6, 9, 8, 7, 3, 2, 11},
{13, 7, 12, 3, 11, 14, 1, 9, 5, 15, 8, 2, 0, 4, 6, 10},
{6, 14, 11, 0, 15, 9, 3, 8, 12, 13, 1, 10, 2, 7, 4, 5},
{10, 8, 7, 1, 2, 4, 6, 5, 15, 9, 3, 13, 11, 14, 12, 0},
{0, 2, 4, 6, 1, 3, 5, 7, 8, 10, 12, 14, 9, 11, 13, 15}, // equal to the first
{14, 4, 9, 13, 10, 8, 15, 6, 1, 0, 11, 5, 12, 2, 7, 3}, // equal to the second
}
func hashBlocksGeneric(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) {
var m [16]uint64
c0, c1 := c[0], c[1]
for i := 0; i < len(blocks); {
c0 += BlockSize
if c0 < BlockSize {
c1++
}
v0, v1, v2, v3, v4, v5, v6, v7 := h[0], h[1], h[2], h[3], h[4], h[5], h[6], h[7]
v8, v9, v10, v11, v12, v13, v14, v15 := iv[0], iv[1], iv[2], iv[3], iv[4], iv[5], iv[6], iv[7]
v12 ^= c0
v13 ^= c1
v14 ^= flag
for j := range m {
m[j] = binary.LittleEndian.Uint64(blocks[i:])
i += 8
}
for j := range precomputed {
s := &(precomputed[j])
v0 += m[s[0]]
v0 += v4
v12 ^= v0
v12 = bits.RotateLeft64(v12, -32)
v8 += v12
v4 ^= v8
v4 = bits.RotateLeft64(v4, -24)
v1 += m[s[1]]
v1 += v5
v13 ^= v1
v13 = bits.RotateLeft64(v13, -32)
v9 += v13
v5 ^= v9
v5 = bits.RotateLeft64(v5, -24)
v2 += m[s[2]]
v2 += v6
v14 ^= v2
v14 = bits.RotateLeft64(v14, -32)
v10 += v14
v6 ^= v10
v6 = bits.RotateLeft64(v6, -24)
v3 += m[s[3]]
v3 += v7
v15 ^= v3
v15 = bits.RotateLeft64(v15, -32)
v11 += v15
v7 ^= v11
v7 = bits.RotateLeft64(v7, -24)
v0 += m[s[4]]
v0 += v4
v12 ^= v0
v12 = bits.RotateLeft64(v12, -16)
v8 += v12
v4 ^= v8
v4 = bits.RotateLeft64(v4, -63)
v1 += m[s[5]]
v1 += v5
v13 ^= v1
v13 = bits.RotateLeft64(v13, -16)
v9 += v13
v5 ^= v9
v5 = bits.RotateLeft64(v5, -63)
v2 += m[s[6]]
v2 += v6
v14 ^= v2
v14 = bits.RotateLeft64(v14, -16)
v10 += v14
v6 ^= v10
v6 = bits.RotateLeft64(v6, -63)
v3 += m[s[7]]
v3 += v7
v15 ^= v3
v15 = bits.RotateLeft64(v15, -16)
v11 += v15
v7 ^= v11
v7 = bits.RotateLeft64(v7, -63)
v0 += m[s[8]]
v0 += v5
v15 ^= v0
v15 = bits.RotateLeft64(v15, -32)
v10 += v15
v5 ^= v10
v5 = bits.RotateLeft64(v5, -24)
v1 += m[s[9]]
v1 += v6
v12 ^= v1
v12 = bits.RotateLeft64(v12, -32)
v11 += v12
v6 ^= v11
v6 = bits.RotateLeft64(v6, -24)
v2 += m[s[10]]
v2 += v7
v13 ^= v2
v13 = bits.RotateLeft64(v13, -32)
v8 += v13
v7 ^= v8
v7 = bits.RotateLeft64(v7, -24)
v3 += m[s[11]]
v3 += v4
v14 ^= v3
v14 = bits.RotateLeft64(v14, -32)
v9 += v14
v4 ^= v9
v4 = bits.RotateLeft64(v4, -24)
v0 += m[s[12]]
v0 += v5
v15 ^= v0
v15 = bits.RotateLeft64(v15, -16)
v10 += v15
v5 ^= v10
v5 = bits.RotateLeft64(v5, -63)
v1 += m[s[13]]
v1 += v6
v12 ^= v1
v12 = bits.RotateLeft64(v12, -16)
v11 += v12
v6 ^= v11
v6 = bits.RotateLeft64(v6, -63)
v2 += m[s[14]]
v2 += v7
v13 ^= v2
v13 = bits.RotateLeft64(v13, -16)
v8 += v13
v7 ^= v8
v7 = bits.RotateLeft64(v7, -63)
v3 += m[s[15]]
v3 += v4
v14 ^= v3
v14 = bits.RotateLeft64(v14, -16)
v9 += v14
v4 ^= v9
v4 = bits.RotateLeft64(v4, -63)
}
h[0] ^= v0 ^ v8
h[1] ^= v1 ^ v9
h[2] ^= v2 ^ v10
h[3] ^= v3 ^ v11
h[4] ^= v4 ^ v12
h[5] ^= v5 ^ v13
h[6] ^= v6 ^ v14
h[7] ^= v7 ^ v15
}
c[0], c[1] = c0, c1
}

12
gateway/vendor/golang.org/x/crypto/blake2b/blake2b_ref.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !amd64 || purego || !gc
// +build !amd64 purego !gc
package blake2b
func hashBlocks(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) {
hashBlocksGeneric(h, c, flag, blocks)
}

177
gateway/vendor/golang.org/x/crypto/blake2b/blake2x.go generated vendored Normal file
View File

@ -0,0 +1,177 @@
// Copyright 2017 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package blake2b
import (
"encoding/binary"
"errors"
"io"
)
// XOF defines the interface to hash functions that
// support arbitrary-length output.
type XOF interface {
// Write absorbs more data into the hash's state. It panics if called
// after Read.
io.Writer
// Read reads more output from the hash. It returns io.EOF if the limit
// has been reached.
io.Reader
// Clone returns a copy of the XOF in its current state.
Clone() XOF
// Reset resets the XOF to its initial state.
Reset()
}
// OutputLengthUnknown can be used as the size argument to NewXOF to indicate
// the length of the output is not known in advance.
const OutputLengthUnknown = 0
// magicUnknownOutputLength is a magic value for the output size that indicates
// an unknown number of output bytes.
const magicUnknownOutputLength = (1 << 32) - 1
// maxOutputLength is the absolute maximum number of bytes to produce when the
// number of output bytes is unknown.
const maxOutputLength = (1 << 32) * 64
// NewXOF creates a new variable-output-length hash. The hash either produce a
// known number of bytes (1 <= size < 2**32-1), or an unknown number of bytes
// (size == OutputLengthUnknown). In the latter case, an absolute limit of
// 256GiB applies.
//
// A non-nil key turns the hash into a MAC. The key must between
// zero and 32 bytes long.
func NewXOF(size uint32, key []byte) (XOF, error) {
if len(key) > Size {
return nil, errKeySize
}
if size == magicUnknownOutputLength {
// 2^32-1 indicates an unknown number of bytes and thus isn't a
// valid length.
return nil, errors.New("blake2b: XOF length too large")
}
if size == OutputLengthUnknown {
size = magicUnknownOutputLength
}
x := &xof{
d: digest{
size: Size,
keyLen: len(key),
},
length: size,
}
copy(x.d.key[:], key)
x.Reset()
return x, nil
}
type xof struct {
d digest
length uint32
remaining uint64
cfg, root, block [Size]byte
offset int
nodeOffset uint32
readMode bool
}
func (x *xof) Write(p []byte) (n int, err error) {
if x.readMode {
panic("blake2b: write to XOF after read")
}
return x.d.Write(p)
}
func (x *xof) Clone() XOF {
clone := *x
return &clone
}
func (x *xof) Reset() {
x.cfg[0] = byte(Size)
binary.LittleEndian.PutUint32(x.cfg[4:], uint32(Size)) // leaf length
binary.LittleEndian.PutUint32(x.cfg[12:], x.length) // XOF length
x.cfg[17] = byte(Size) // inner hash size
x.d.Reset()
x.d.h[1] ^= uint64(x.length) << 32
x.remaining = uint64(x.length)
if x.remaining == magicUnknownOutputLength {
x.remaining = maxOutputLength
}
x.offset, x.nodeOffset = 0, 0
x.readMode = false
}
func (x *xof) Read(p []byte) (n int, err error) {
if !x.readMode {
x.d.finalize(&x.root)
x.readMode = true
}
if x.remaining == 0 {
return 0, io.EOF
}
n = len(p)
if uint64(n) > x.remaining {
n = int(x.remaining)
p = p[:n]
}
if x.offset > 0 {
blockRemaining := Size - x.offset
if n < blockRemaining {
x.offset += copy(p, x.block[x.offset:])
x.remaining -= uint64(n)
return
}
copy(p, x.block[x.offset:])
p = p[blockRemaining:]
x.offset = 0
x.remaining -= uint64(blockRemaining)
}
for len(p) >= Size {
binary.LittleEndian.PutUint32(x.cfg[8:], x.nodeOffset)
x.nodeOffset++
x.d.initConfig(&x.cfg)
x.d.Write(x.root[:])
x.d.finalize(&x.block)
copy(p, x.block[:])
p = p[Size:]
x.remaining -= uint64(Size)
}
if todo := len(p); todo > 0 {
if x.remaining < uint64(Size) {
x.cfg[0] = byte(x.remaining)
}
binary.LittleEndian.PutUint32(x.cfg[8:], x.nodeOffset)
x.nodeOffset++
x.d.initConfig(&x.cfg)
x.d.Write(x.root[:])
x.d.finalize(&x.block)
x.offset = copy(p, x.block[:todo])
x.remaining -= uint64(todo)
}
return
}
func (d *digest) initConfig(cfg *[Size]byte) {
d.offset, d.c[0], d.c[1] = 0, 0, 0
for i := range d.h {
d.h[i] = iv[i] ^ binary.LittleEndian.Uint64(cfg[i*8:])
}
}

33
gateway/vendor/golang.org/x/crypto/blake2b/register.go generated vendored Normal file
View File

@ -0,0 +1,33 @@
// Copyright 2017 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build go1.9
// +build go1.9
package blake2b
import (
"crypto"
"hash"
)
func init() {
newHash256 := func() hash.Hash {
h, _ := New256(nil)
return h
}
newHash384 := func() hash.Hash {
h, _ := New384(nil)
return h
}
newHash512 := func() hash.Hash {
h, _ := New512(nil)
return h
}
crypto.RegisterHash(crypto.BLAKE2b_256, newHash256)
crypto.RegisterHash(crypto.BLAKE2b_384, newHash384)
crypto.RegisterHash(crypto.BLAKE2b_512, newHash512)
}

59
gateway/vendor/golang.org/x/crypto/curve25519/curve25519.go generated vendored Normal file
View File

@ -0,0 +1,59 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package curve25519 provides an implementation of the X25519 function, which
// performs scalar multiplication on the elliptic curve known as Curve25519.
// See RFC 7748.
//
// Starting in Go 1.20, this package is a wrapper for the X25519 implementation
// in the crypto/ecdh package.
package curve25519 // import "golang.org/x/crypto/curve25519"
// ScalarMult sets dst to the product scalar * point.
//
// Deprecated: when provided a low-order point, ScalarMult will set dst to all
// zeroes, irrespective of the scalar. Instead, use the X25519 function, which
// will return an error.
func ScalarMult(dst, scalar, point *[32]byte) {
scalarMult(dst, scalar, point)
}
// ScalarBaseMult sets dst to the product scalar * base where base is the
// standard generator.
//
// It is recommended to use the X25519 function with Basepoint instead, as
// copying into fixed size arrays can lead to unexpected bugs.
func ScalarBaseMult(dst, scalar *[32]byte) {
scalarBaseMult(dst, scalar)
}
const (
// ScalarSize is the size of the scalar input to X25519.
ScalarSize = 32
// PointSize is the size of the point input to X25519.
PointSize = 32
)
// Basepoint is the canonical Curve25519 generator.
var Basepoint []byte
var basePoint = [32]byte{9}
func init() { Basepoint = basePoint[:] }
// X25519 returns the result of the scalar multiplication (scalar * point),
// according to RFC 7748, Section 5. scalar, point and the return value are
// slices of 32 bytes.
//
// scalar can be generated at random, for example with crypto/rand. point should
// be either Basepoint or the output of another X25519 call.
//
// If point is Basepoint (but not if it's a different slice with the same
// contents) a precomputed implementation might be used for performance.
func X25519(scalar, point []byte) ([]byte, error) {
// Outline the body of function, to let the allocation be inlined in the
// caller, and possibly avoid escaping to the heap.
var dst [32]byte
return x25519(&dst, scalar, point)
}

105
gateway/vendor/golang.org/x/crypto/curve25519/curve25519_compat.go generated vendored Normal file
View File

@ -0,0 +1,105 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !go1.20
package curve25519
import (
"crypto/subtle"
"errors"
"strconv"
"golang.org/x/crypto/curve25519/internal/field"
)
func scalarMult(dst, scalar, point *[32]byte) {
var e [32]byte
copy(e[:], scalar[:])
e[0] &= 248
e[31] &= 127
e[31] |= 64
var x1, x2, z2, x3, z3, tmp0, tmp1 field.Element
x1.SetBytes(point[:])
x2.One()
x3.Set(&x1)
z3.One()
swap := 0
for pos := 254; pos >= 0; pos-- {
b := e[pos/8] >> uint(pos&7)
b &= 1
swap ^= int(b)
x2.Swap(&x3, swap)
z2.Swap(&z3, swap)
swap = int(b)
tmp0.Subtract(&x3, &z3)
tmp1.Subtract(&x2, &z2)
x2.Add(&x2, &z2)
z2.Add(&x3, &z3)
z3.Multiply(&tmp0, &x2)
z2.Multiply(&z2, &tmp1)
tmp0.Square(&tmp1)
tmp1.Square(&x2)
x3.Add(&z3, &z2)
z2.Subtract(&z3, &z2)
x2.Multiply(&tmp1, &tmp0)
tmp1.Subtract(&tmp1, &tmp0)
z2.Square(&z2)
z3.Mult32(&tmp1, 121666)
x3.Square(&x3)
tmp0.Add(&tmp0, &z3)
z3.Multiply(&x1, &z2)
z2.Multiply(&tmp1, &tmp0)
}
x2.Swap(&x3, swap)
z2.Swap(&z3, swap)
z2.Invert(&z2)
x2.Multiply(&x2, &z2)
copy(dst[:], x2.Bytes())
}
func scalarBaseMult(dst, scalar *[32]byte) {
checkBasepoint()
scalarMult(dst, scalar, &basePoint)
}
func x25519(dst *[32]byte, scalar, point []byte) ([]byte, error) {
var in [32]byte
if l := len(scalar); l != 32 {
return nil, errors.New("bad scalar length: " + strconv.Itoa(l) + ", expected 32")
}
if l := len(point); l != 32 {
return nil, errors.New("bad point length: " + strconv.Itoa(l) + ", expected 32")
}
copy(in[:], scalar)
if &point[0] == &Basepoint[0] {
scalarBaseMult(dst, &in)
} else {
var base, zero [32]byte
copy(base[:], point)
scalarMult(dst, &in, &base)
if subtle.ConstantTimeCompare(dst[:], zero[:]) == 1 {
return nil, errors.New("bad input point: low order point")
}
}
return dst[:], nil
}
func checkBasepoint() {
if subtle.ConstantTimeCompare(Basepoint, []byte{
0x09, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
}) != 1 {
panic("curve25519: global Basepoint value was modified")
}
}

46
gateway/vendor/golang.org/x/crypto/curve25519/curve25519_go120.go generated vendored Normal file
View File

@ -0,0 +1,46 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build go1.20
package curve25519
import "crypto/ecdh"
func x25519(dst *[32]byte, scalar, point []byte) ([]byte, error) {
curve := ecdh.X25519()
pub, err := curve.NewPublicKey(point)
if err != nil {
return nil, err
}
priv, err := curve.NewPrivateKey(scalar)
if err != nil {
return nil, err
}
out, err := priv.ECDH(pub)
if err != nil {
return nil, err
}
copy(dst[:], out)
return dst[:], nil
}
func scalarMult(dst, scalar, point *[32]byte) {
if _, err := x25519(dst, scalar[:], point[:]); err != nil {
// The only error condition for x25519 when the inputs are 32 bytes long
// is if the output would have been the all-zero value.
for i := range dst {
dst[i] = 0
}
}
}
func scalarBaseMult(dst, scalar *[32]byte) {
curve := ecdh.X25519()
priv, err := curve.NewPrivateKey(scalar[:])
if err != nil {
panic("curve25519: internal error: scalarBaseMult was not 32 bytes")
}
copy(dst[:], priv.PublicKey().Bytes())
}

7
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/README generated vendored Normal file
View File

@ -0,0 +1,7 @@
This package is kept in sync with crypto/ed25519/internal/edwards25519/field in
the standard library.
If there are any changes in the standard library that need to be synced to this
package, run sync.sh. It will not overwrite any local changes made since the
previous sync, so it's ok to land changes in this package first, and then sync
to the standard library later.

416
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe.go generated vendored Normal file
View File

@ -0,0 +1,416 @@
// Copyright (c) 2017 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package field implements fast arithmetic modulo 2^255-19.
package field
import (
"crypto/subtle"
"encoding/binary"
"math/bits"
)
// Element represents an element of the field GF(2^255-19). Note that this
// is not a cryptographically secure group, and should only be used to interact
// with edwards25519.Point coordinates.
//
// This type works similarly to math/big.Int, and all arguments and receivers
// are allowed to alias.
//
// The zero value is a valid zero element.
type Element struct {
// An element t represents the integer
// t.l0 + t.l1*2^51 + t.l2*2^102 + t.l3*2^153 + t.l4*2^204
//
// Between operations, all limbs are expected to be lower than 2^52.
l0 uint64
l1 uint64
l2 uint64
l3 uint64
l4 uint64
}
const maskLow51Bits uint64 = (1 << 51) - 1
var feZero = &Element{0, 0, 0, 0, 0}
// Zero sets v = 0, and returns v.
func (v *Element) Zero() *Element {
*v = *feZero
return v
}
var feOne = &Element{1, 0, 0, 0, 0}
// One sets v = 1, and returns v.
func (v *Element) One() *Element {
*v = *feOne
return v
}
// reduce reduces v modulo 2^255 - 19 and returns it.
func (v *Element) reduce() *Element {
v.carryPropagate()
// After the light reduction we now have a field element representation
// v < 2^255 + 2^13 * 19, but need v < 2^255 - 19.
// If v >= 2^255 - 19, then v + 19 >= 2^255, which would overflow 2^255 - 1,
// generating a carry. That is, c will be 0 if v < 2^255 - 19, and 1 otherwise.
c := (v.l0 + 19) >> 51
c = (v.l1 + c) >> 51
c = (v.l2 + c) >> 51
c = (v.l3 + c) >> 51
c = (v.l4 + c) >> 51
// If v < 2^255 - 19 and c = 0, this will be a no-op. Otherwise, it's
// effectively applying the reduction identity to the carry.
v.l0 += 19 * c
v.l1 += v.l0 >> 51
v.l0 = v.l0 & maskLow51Bits
v.l2 += v.l1 >> 51
v.l1 = v.l1 & maskLow51Bits
v.l3 += v.l2 >> 51
v.l2 = v.l2 & maskLow51Bits
v.l4 += v.l3 >> 51
v.l3 = v.l3 & maskLow51Bits
// no additional carry
v.l4 = v.l4 & maskLow51Bits
return v
}
// Add sets v = a + b, and returns v.
func (v *Element) Add(a, b *Element) *Element {
v.l0 = a.l0 + b.l0
v.l1 = a.l1 + b.l1
v.l2 = a.l2 + b.l2
v.l3 = a.l3 + b.l3
v.l4 = a.l4 + b.l4
// Using the generic implementation here is actually faster than the
// assembly. Probably because the body of this function is so simple that
// the compiler can figure out better optimizations by inlining the carry
// propagation. TODO
return v.carryPropagateGeneric()
}
// Subtract sets v = a - b, and returns v.
func (v *Element) Subtract(a, b *Element) *Element {
// We first add 2 * p, to guarantee the subtraction won't underflow, and
// then subtract b (which can be up to 2^255 + 2^13 * 19).
v.l0 = (a.l0 + 0xFFFFFFFFFFFDA) - b.l0
v.l1 = (a.l1 + 0xFFFFFFFFFFFFE) - b.l1
v.l2 = (a.l2 + 0xFFFFFFFFFFFFE) - b.l2
v.l3 = (a.l3 + 0xFFFFFFFFFFFFE) - b.l3
v.l4 = (a.l4 + 0xFFFFFFFFFFFFE) - b.l4
return v.carryPropagate()
}
// Negate sets v = -a, and returns v.
func (v *Element) Negate(a *Element) *Element {
return v.Subtract(feZero, a)
}
// Invert sets v = 1/z mod p, and returns v.
//
// If z == 0, Invert returns v = 0.
func (v *Element) Invert(z *Element) *Element {
// Inversion is implemented as exponentiation with exponent p − 2. It uses the
// same sequence of 255 squarings and 11 multiplications as [Curve25519].
var z2, z9, z11, z2_5_0, z2_10_0, z2_20_0, z2_50_0, z2_100_0, t Element
z2.Square(z) // 2
t.Square(&z2) // 4
t.Square(&t) // 8
z9.Multiply(&t, z) // 9
z11.Multiply(&z9, &z2) // 11
t.Square(&z11) // 22
z2_5_0.Multiply(&t, &z9) // 31 = 2^5 - 2^0
t.Square(&z2_5_0) // 2^6 - 2^1
for i := 0; i < 4; i++ {
t.Square(&t) // 2^10 - 2^5
}
z2_10_0.Multiply(&t, &z2_5_0) // 2^10 - 2^0
t.Square(&z2_10_0) // 2^11 - 2^1
for i := 0; i < 9; i++ {
t.Square(&t) // 2^20 - 2^10
}
z2_20_0.Multiply(&t, &z2_10_0) // 2^20 - 2^0
t.Square(&z2_20_0) // 2^21 - 2^1
for i := 0; i < 19; i++ {
t.Square(&t) // 2^40 - 2^20
}
t.Multiply(&t, &z2_20_0) // 2^40 - 2^0
t.Square(&t) // 2^41 - 2^1
for i := 0; i < 9; i++ {
t.Square(&t) // 2^50 - 2^10
}
z2_50_0.Multiply(&t, &z2_10_0) // 2^50 - 2^0
t.Square(&z2_50_0) // 2^51 - 2^1
for i := 0; i < 49; i++ {
t.Square(&t) // 2^100 - 2^50
}
z2_100_0.Multiply(&t, &z2_50_0) // 2^100 - 2^0
t.Square(&z2_100_0) // 2^101 - 2^1
for i := 0; i < 99; i++ {
t.Square(&t) // 2^200 - 2^100
}
t.Multiply(&t, &z2_100_0) // 2^200 - 2^0
t.Square(&t) // 2^201 - 2^1
for i := 0; i < 49; i++ {
t.Square(&t) // 2^250 - 2^50
}
t.Multiply(&t, &z2_50_0) // 2^250 - 2^0
t.Square(&t) // 2^251 - 2^1
t.Square(&t) // 2^252 - 2^2
t.Square(&t) // 2^253 - 2^3
t.Square(&t) // 2^254 - 2^4
t.Square(&t) // 2^255 - 2^5
return v.Multiply(&t, &z11) // 2^255 - 21
}
// Set sets v = a, and returns v.
func (v *Element) Set(a *Element) *Element {
*v = *a
return v
}
// SetBytes sets v to x, which must be a 32-byte little-endian encoding.
//
// Consistent with RFC 7748, the most significant bit (the high bit of the
// last byte) is ignored, and non-canonical values (2^255-19 through 2^255-1)
// are accepted. Note that this is laxer than specified by RFC 8032.
func (v *Element) SetBytes(x []byte) *Element {
if len(x) != 32 {
panic("edwards25519: invalid field element input size")
}
// Bits 0:51 (bytes 0:8, bits 0:64, shift 0, mask 51).
v.l0 = binary.LittleEndian.Uint64(x[0:8])
v.l0 &= maskLow51Bits
// Bits 51:102 (bytes 6:14, bits 48:112, shift 3, mask 51).
v.l1 = binary.LittleEndian.Uint64(x[6:14]) >> 3
v.l1 &= maskLow51Bits
// Bits 102:153 (bytes 12:20, bits 96:160, shift 6, mask 51).
v.l2 = binary.LittleEndian.Uint64(x[12:20]) >> 6
v.l2 &= maskLow51Bits
// Bits 153:204 (bytes 19:27, bits 152:216, shift 1, mask 51).
v.l3 = binary.LittleEndian.Uint64(x[19:27]) >> 1
v.l3 &= maskLow51Bits
// Bits 204:251 (bytes 24:32, bits 192:256, shift 12, mask 51).
// Note: not bytes 25:33, shift 4, to avoid overread.
v.l4 = binary.LittleEndian.Uint64(x[24:32]) >> 12
v.l4 &= maskLow51Bits
return v
}
// Bytes returns the canonical 32-byte little-endian encoding of v.
func (v *Element) Bytes() []byte {
// This function is outlined to make the allocations inline in the caller
// rather than happen on the heap.
var out [32]byte
return v.bytes(&out)
}
func (v *Element) bytes(out *[32]byte) []byte {
t := *v
t.reduce()
var buf [8]byte
for i, l := range [5]uint64{t.l0, t.l1, t.l2, t.l3, t.l4} {
bitsOffset := i * 51
binary.LittleEndian.PutUint64(buf[:], l<<uint(bitsOffset%8))
for i, bb := range buf {
off := bitsOffset/8 + i
if off >= len(out) {
break
}
out[off] |= bb
}
}
return out[:]
}
// Equal returns 1 if v and u are equal, and 0 otherwise.
func (v *Element) Equal(u *Element) int {
sa, sv := u.Bytes(), v.Bytes()
return subtle.ConstantTimeCompare(sa, sv)
}
// mask64Bits returns 0xffffffff if cond is 1, and 0 otherwise.
func mask64Bits(cond int) uint64 { return ^(uint64(cond) - 1) }
// Select sets v to a if cond == 1, and to b if cond == 0.
func (v *Element) Select(a, b *Element, cond int) *Element {
m := mask64Bits(cond)
v.l0 = (m & a.l0) | (^m & b.l0)
v.l1 = (m & a.l1) | (^m & b.l1)
v.l2 = (m & a.l2) | (^m & b.l2)
v.l3 = (m & a.l3) | (^m & b.l3)
v.l4 = (m & a.l4) | (^m & b.l4)
return v
}
// Swap swaps v and u if cond == 1 or leaves them unchanged if cond == 0, and returns v.
func (v *Element) Swap(u *Element, cond int) {
m := mask64Bits(cond)
t := m & (v.l0 ^ u.l0)
v.l0 ^= t
u.l0 ^= t
t = m & (v.l1 ^ u.l1)
v.l1 ^= t
u.l1 ^= t
t = m & (v.l2 ^ u.l2)
v.l2 ^= t
u.l2 ^= t
t = m & (v.l3 ^ u.l3)
v.l3 ^= t
u.l3 ^= t
t = m & (v.l4 ^ u.l4)
v.l4 ^= t
u.l4 ^= t
}
// IsNegative returns 1 if v is negative, and 0 otherwise.
func (v *Element) IsNegative() int {
return int(v.Bytes()[0] & 1)
}
// Absolute sets v to |u|, and returns v.
func (v *Element) Absolute(u *Element) *Element {
return v.Select(new(Element).Negate(u), u, u.IsNegative())
}
// Multiply sets v = x * y, and returns v.
func (v *Element) Multiply(x, y *Element) *Element {
feMul(v, x, y)
return v
}
// Square sets v = x * x, and returns v.
func (v *Element) Square(x *Element) *Element {
feSquare(v, x)
return v
}
// Mult32 sets v = x * y, and returns v.
func (v *Element) Mult32(x *Element, y uint32) *Element {
x0lo, x0hi := mul51(x.l0, y)
x1lo, x1hi := mul51(x.l1, y)
x2lo, x2hi := mul51(x.l2, y)
x3lo, x3hi := mul51(x.l3, y)
x4lo, x4hi := mul51(x.l4, y)
v.l0 = x0lo + 19*x4hi // carried over per the reduction identity
v.l1 = x1lo + x0hi
v.l2 = x2lo + x1hi
v.l3 = x3lo + x2hi
v.l4 = x4lo + x3hi
// The hi portions are going to be only 32 bits, plus any previous excess,
// so we can skip the carry propagation.
return v
}
// mul51 returns lo + hi * 2⁵¹ = a * b.
func mul51(a uint64, b uint32) (lo uint64, hi uint64) {
mh, ml := bits.Mul64(a, uint64(b))
lo = ml & maskLow51Bits
hi = (mh << 13) | (ml >> 51)
return
}
// Pow22523 set v = x^((p-5)/8), and returns v. (p-5)/8 is 2^252-3.
func (v *Element) Pow22523(x *Element) *Element {
var t0, t1, t2 Element
t0.Square(x) // x^2
t1.Square(&t0) // x^4
t1.Square(&t1) // x^8
t1.Multiply(x, &t1) // x^9
t0.Multiply(&t0, &t1) // x^11
t0.Square(&t0) // x^22
t0.Multiply(&t1, &t0) // x^31
t1.Square(&t0) // x^62
for i := 1; i < 5; i++ { // x^992
t1.Square(&t1)
}
t0.Multiply(&t1, &t0) // x^1023 -> 1023 = 2^10 - 1
t1.Square(&t0) // 2^11 - 2
for i := 1; i < 10; i++ { // 2^20 - 2^10
t1.Square(&t1)
}
t1.Multiply(&t1, &t0) // 2^20 - 1
t2.Square(&t1) // 2^21 - 2
for i := 1; i < 20; i++ { // 2^40 - 2^20
t2.Square(&t2)
}
t1.Multiply(&t2, &t1) // 2^40 - 1
t1.Square(&t1) // 2^41 - 2
for i := 1; i < 10; i++ { // 2^50 - 2^10
t1.Square(&t1)
}
t0.Multiply(&t1, &t0) // 2^50 - 1
t1.Square(&t0) // 2^51 - 2
for i := 1; i < 50; i++ { // 2^100 - 2^50
t1.Square(&t1)
}
t1.Multiply(&t1, &t0) // 2^100 - 1
t2.Square(&t1) // 2^101 - 2
for i := 1; i < 100; i++ { // 2^200 - 2^100
t2.Square(&t2)
}
t1.Multiply(&t2, &t1) // 2^200 - 1
t1.Square(&t1) // 2^201 - 2
for i := 1; i < 50; i++ { // 2^250 - 2^50
t1.Square(&t1)
}
t0.Multiply(&t1, &t0) // 2^250 - 1
t0.Square(&t0) // 2^251 - 2
t0.Square(&t0) // 2^252 - 4
return v.Multiply(&t0, x) // 2^252 - 3 -> x^(2^252-3)
}
// sqrtM1 is 2^((p-1)/4), which squared is equal to -1 by Euler's Criterion.
var sqrtM1 = &Element{1718705420411056, 234908883556509,
2233514472574048, 2117202627021982, 765476049583133}
// SqrtRatio sets r to the non-negative square root of the ratio of u and v.
//
// If u/v is square, SqrtRatio returns r and 1. If u/v is not square, SqrtRatio
// sets r according to Section 4.3 of draft-irtf-cfrg-ristretto255-decaf448-00,
// and returns r and 0.
func (r *Element) SqrtRatio(u, v *Element) (rr *Element, wasSquare int) {
var a, b Element
// r = (u * v3) * (u * v7)^((p-5)/8)
v2 := a.Square(v)
uv3 := b.Multiply(u, b.Multiply(v2, v))
uv7 := a.Multiply(uv3, a.Square(v2))
r.Multiply(uv3, r.Pow22523(uv7))
check := a.Multiply(v, a.Square(r)) // check = v * r^2
uNeg := b.Negate(u)
correctSignSqrt := check.Equal(u)
flippedSignSqrt := check.Equal(uNeg)
flippedSignSqrtI := check.Equal(uNeg.Multiply(uNeg, sqrtM1))
rPrime := b.Multiply(r, sqrtM1) // r_prime = SQRT_M1 * r
// r = CT_SELECT(r_prime IF flipped_sign_sqrt | flipped_sign_sqrt_i ELSE r)
r.Select(rPrime, r, flippedSignSqrt|flippedSignSqrtI)
r.Absolute(r) // Choose the nonnegative square root.
return r, correctSignSqrt | flippedSignSqrt
}

16
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_amd64.go generated vendored Normal file
View File

@ -0,0 +1,16 @@
// Code generated by command: go run fe_amd64_asm.go -out ../fe_amd64.s -stubs ../fe_amd64.go -pkg field. DO NOT EDIT.
//go:build amd64 && gc && !purego
// +build amd64,gc,!purego
package field
// feMul sets out = a * b. It works like feMulGeneric.
//
//go:noescape
func feMul(out *Element, a *Element, b *Element)
// feSquare sets out = a * a. It works like feSquareGeneric.
//
//go:noescape
func feSquare(out *Element, a *Element)

379
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_amd64.s generated vendored Normal file
View File

@ -0,0 +1,379 @@
// Code generated by command: go run fe_amd64_asm.go -out ../fe_amd64.s -stubs ../fe_amd64.go -pkg field. DO NOT EDIT.
//go:build amd64 && gc && !purego
// +build amd64,gc,!purego
#include "textflag.h"
// func feMul(out *Element, a *Element, b *Element)
TEXT ·feMul(SB), NOSPLIT, $0-24
MOVQ a+8(FP), CX
MOVQ b+16(FP), BX
// r0 = a0×b0
MOVQ (CX), AX
MULQ (BX)
MOVQ AX, DI
MOVQ DX, SI
// r0 += 19×a1×b4
MOVQ 8(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 32(BX)
ADDQ AX, DI
ADCQ DX, SI
// r0 += 19×a2×b3
MOVQ 16(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 24(BX)
ADDQ AX, DI
ADCQ DX, SI
// r0 += 19×a3×b2
MOVQ 24(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 16(BX)
ADDQ AX, DI
ADCQ DX, SI
// r0 += 19×a4×b1
MOVQ 32(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 8(BX)
ADDQ AX, DI
ADCQ DX, SI
// r1 = a0×b1
MOVQ (CX), AX
MULQ 8(BX)
MOVQ AX, R9
MOVQ DX, R8
// r1 += a1×b0
MOVQ 8(CX), AX
MULQ (BX)
ADDQ AX, R9
ADCQ DX, R8
// r1 += 19×a2×b4
MOVQ 16(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 32(BX)
ADDQ AX, R9
ADCQ DX, R8
// r1 += 19×a3×b3
MOVQ 24(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 24(BX)
ADDQ AX, R9
ADCQ DX, R8
// r1 += 19×a4×b2
MOVQ 32(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 16(BX)
ADDQ AX, R9
ADCQ DX, R8
// r2 = a0×b2
MOVQ (CX), AX
MULQ 16(BX)
MOVQ AX, R11
MOVQ DX, R10
// r2 += a1×b1
MOVQ 8(CX), AX
MULQ 8(BX)
ADDQ AX, R11
ADCQ DX, R10
// r2 += a2×b0
MOVQ 16(CX), AX
MULQ (BX)
ADDQ AX, R11
ADCQ DX, R10
// r2 += 19×a3×b4
MOVQ 24(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 32(BX)
ADDQ AX, R11
ADCQ DX, R10
// r2 += 19×a4×b3
MOVQ 32(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 24(BX)
ADDQ AX, R11
ADCQ DX, R10
// r3 = a0×b3
MOVQ (CX), AX
MULQ 24(BX)
MOVQ AX, R13
MOVQ DX, R12
// r3 += a1×b2
MOVQ 8(CX), AX
MULQ 16(BX)
ADDQ AX, R13
ADCQ DX, R12
// r3 += a2×b1
MOVQ 16(CX), AX
MULQ 8(BX)
ADDQ AX, R13
ADCQ DX, R12
// r3 += a3×b0
MOVQ 24(CX), AX
MULQ (BX)
ADDQ AX, R13
ADCQ DX, R12
// r3 += 19×a4×b4
MOVQ 32(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 32(BX)
ADDQ AX, R13
ADCQ DX, R12
// r4 = a0×b4
MOVQ (CX), AX
MULQ 32(BX)
MOVQ AX, R15
MOVQ DX, R14
// r4 += a1×b3
MOVQ 8(CX), AX
MULQ 24(BX)
ADDQ AX, R15
ADCQ DX, R14
// r4 += a2×b2
MOVQ 16(CX), AX
MULQ 16(BX)
ADDQ AX, R15
ADCQ DX, R14
// r4 += a3×b1
MOVQ 24(CX), AX
MULQ 8(BX)
ADDQ AX, R15
ADCQ DX, R14
// r4 += a4×b0
MOVQ 32(CX), AX
MULQ (BX)
ADDQ AX, R15
ADCQ DX, R14
// First reduction chain
MOVQ $0x0007ffffffffffff, AX
SHLQ $0x0d, DI, SI
SHLQ $0x0d, R9, R8
SHLQ $0x0d, R11, R10
SHLQ $0x0d, R13, R12
SHLQ $0x0d, R15, R14
ANDQ AX, DI
IMUL3Q $0x13, R14, R14
ADDQ R14, DI
ANDQ AX, R9
ADDQ SI, R9
ANDQ AX, R11
ADDQ R8, R11
ANDQ AX, R13
ADDQ R10, R13
ANDQ AX, R15
ADDQ R12, R15
// Second reduction chain (carryPropagate)
MOVQ DI, SI
SHRQ $0x33, SI
MOVQ R9, R8
SHRQ $0x33, R8
MOVQ R11, R10
SHRQ $0x33, R10
MOVQ R13, R12
SHRQ $0x33, R12
MOVQ R15, R14
SHRQ $0x33, R14
ANDQ AX, DI
IMUL3Q $0x13, R14, R14
ADDQ R14, DI
ANDQ AX, R9
ADDQ SI, R9
ANDQ AX, R11
ADDQ R8, R11
ANDQ AX, R13
ADDQ R10, R13
ANDQ AX, R15
ADDQ R12, R15
// Store output
MOVQ out+0(FP), AX
MOVQ DI, (AX)
MOVQ R9, 8(AX)
MOVQ R11, 16(AX)
MOVQ R13, 24(AX)
MOVQ R15, 32(AX)
RET
// func feSquare(out *Element, a *Element)
TEXT ·feSquare(SB), NOSPLIT, $0-16
MOVQ a+8(FP), CX
// r0 = l0×l0
MOVQ (CX), AX
MULQ (CX)
MOVQ AX, SI
MOVQ DX, BX
// r0 += 38×l1×l4
MOVQ 8(CX), AX
IMUL3Q $0x26, AX, AX
MULQ 32(CX)
ADDQ AX, SI
ADCQ DX, BX
// r0 += 38×l2×l3
MOVQ 16(CX), AX
IMUL3Q $0x26, AX, AX
MULQ 24(CX)
ADDQ AX, SI
ADCQ DX, BX
// r1 = 2×l0×l1
MOVQ (CX), AX
SHLQ $0x01, AX
MULQ 8(CX)
MOVQ AX, R8
MOVQ DX, DI
// r1 += 38×l2×l4
MOVQ 16(CX), AX
IMUL3Q $0x26, AX, AX
MULQ 32(CX)
ADDQ AX, R8
ADCQ DX, DI
// r1 += 19×l3×l3
MOVQ 24(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 24(CX)
ADDQ AX, R8
ADCQ DX, DI
// r2 = 2×l0×l2
MOVQ (CX), AX
SHLQ $0x01, AX
MULQ 16(CX)
MOVQ AX, R10
MOVQ DX, R9
// r2 += l1×l1
MOVQ 8(CX), AX
MULQ 8(CX)
ADDQ AX, R10
ADCQ DX, R9
// r2 += 38×l3×l4
MOVQ 24(CX), AX
IMUL3Q $0x26, AX, AX
MULQ 32(CX)
ADDQ AX, R10
ADCQ DX, R9
// r3 = 2×l0×l3
MOVQ (CX), AX
SHLQ $0x01, AX
MULQ 24(CX)
MOVQ AX, R12
MOVQ DX, R11
// r3 += 2×l1×l2
MOVQ 8(CX), AX
IMUL3Q $0x02, AX, AX
MULQ 16(CX)
ADDQ AX, R12
ADCQ DX, R11
// r3 += 19×l4×l4
MOVQ 32(CX), AX
IMUL3Q $0x13, AX, AX
MULQ 32(CX)
ADDQ AX, R12
ADCQ DX, R11
// r4 = 2×l0×l4
MOVQ (CX), AX
SHLQ $0x01, AX
MULQ 32(CX)
MOVQ AX, R14
MOVQ DX, R13
// r4 += 2×l1×l3
MOVQ 8(CX), AX
IMUL3Q $0x02, AX, AX
MULQ 24(CX)
ADDQ AX, R14
ADCQ DX, R13
// r4 += l2×l2
MOVQ 16(CX), AX
MULQ 16(CX)
ADDQ AX, R14
ADCQ DX, R13
// First reduction chain
MOVQ $0x0007ffffffffffff, AX
SHLQ $0x0d, SI, BX
SHLQ $0x0d, R8, DI
SHLQ $0x0d, R10, R9
SHLQ $0x0d, R12, R11
SHLQ $0x0d, R14, R13
ANDQ AX, SI
IMUL3Q $0x13, R13, R13
ADDQ R13, SI
ANDQ AX, R8
ADDQ BX, R8
ANDQ AX, R10
ADDQ DI, R10
ANDQ AX, R12
ADDQ R9, R12
ANDQ AX, R14
ADDQ R11, R14
// Second reduction chain (carryPropagate)
MOVQ SI, BX
SHRQ $0x33, BX
MOVQ R8, DI
SHRQ $0x33, DI
MOVQ R10, R9
SHRQ $0x33, R9
MOVQ R12, R11
SHRQ $0x33, R11
MOVQ R14, R13
SHRQ $0x33, R13
ANDQ AX, SI
IMUL3Q $0x13, R13, R13
ADDQ R13, SI
ANDQ AX, R8
ADDQ BX, R8
ANDQ AX, R10
ADDQ DI, R10
ANDQ AX, R12
ADDQ R9, R12
ANDQ AX, R14
ADDQ R11, R14
// Store output
MOVQ out+0(FP), AX
MOVQ SI, (AX)
MOVQ R8, 8(AX)
MOVQ R10, 16(AX)
MOVQ R12, 24(AX)
MOVQ R14, 32(AX)
RET

12
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_amd64_noasm.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright (c) 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !amd64 || !gc || purego
// +build !amd64 !gc purego
package field
func feMul(v, x, y *Element) { feMulGeneric(v, x, y) }
func feSquare(v, x *Element) { feSquareGeneric(v, x) }

16
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_arm64.go generated vendored Normal file
View File

@ -0,0 +1,16 @@
// Copyright (c) 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build arm64 && gc && !purego
// +build arm64,gc,!purego
package field
//go:noescape
func carryPropagate(v *Element)
func (v *Element) carryPropagate() *Element {
carryPropagate(v)
return v
}

43
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_arm64.s generated vendored Normal file
View File

@ -0,0 +1,43 @@
// Copyright (c) 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build arm64 && gc && !purego
// +build arm64,gc,!purego
#include "textflag.h"
// carryPropagate works exactly like carryPropagateGeneric and uses the
// same AND, ADD, and LSR+MADD instructions emitted by the compiler, but
// avoids loading R0-R4 twice and uses LDP and STP.
//
// See https://golang.org/issues/43145 for the main compiler issue.
//
// func carryPropagate(v *Element)
TEXT ·carryPropagate(SB),NOFRAME|NOSPLIT,$0-8
MOVD v+0(FP), R20
LDP 0(R20), (R0, R1)
LDP 16(R20), (R2, R3)
MOVD 32(R20), R4
AND $0x7ffffffffffff, R0, R10
AND $0x7ffffffffffff, R1, R11
AND $0x7ffffffffffff, R2, R12
AND $0x7ffffffffffff, R3, R13
AND $0x7ffffffffffff, R4, R14
ADD R0>>51, R11, R11
ADD R1>>51, R12, R12
ADD R2>>51, R13, R13
ADD R3>>51, R14, R14
// R4>>51 * 19 + R10 -> R10
LSR $51, R4, R21
MOVD $19, R22
MADD R22, R10, R21, R10
STP (R10, R11), 0(R20)
STP (R12, R13), 16(R20)
MOVD R14, 32(R20)
RET

12
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_arm64_noasm.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright (c) 2021 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !arm64 || !gc || purego
// +build !arm64 !gc purego
package field
func (v *Element) carryPropagate() *Element {
return v.carryPropagateGeneric()
}

264
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/fe_generic.go generated vendored Normal file
View File

@ -0,0 +1,264 @@
// Copyright (c) 2017 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package field
import "math/bits"
// uint128 holds a 128-bit number as two 64-bit limbs, for use with the
// bits.Mul64 and bits.Add64 intrinsics.
type uint128 struct {
lo, hi uint64
}
// mul64 returns a * b.
func mul64(a, b uint64) uint128 {
hi, lo := bits.Mul64(a, b)
return uint128{lo, hi}
}
// addMul64 returns v + a * b.
func addMul64(v uint128, a, b uint64) uint128 {
hi, lo := bits.Mul64(a, b)
lo, c := bits.Add64(lo, v.lo, 0)
hi, _ = bits.Add64(hi, v.hi, c)
return uint128{lo, hi}
}
// shiftRightBy51 returns a >> 51. a is assumed to be at most 115 bits.
func shiftRightBy51(a uint128) uint64 {
return (a.hi << (64 - 51)) | (a.lo >> 51)
}
func feMulGeneric(v, a, b *Element) {
a0 := a.l0
a1 := a.l1
a2 := a.l2
a3 := a.l3
a4 := a.l4
b0 := b.l0
b1 := b.l1
b2 := b.l2
b3 := b.l3
b4 := b.l4
// Limb multiplication works like pen-and-paper columnar multiplication, but
// with 51-bit limbs instead of digits.
//
// a4 a3 a2 a1 a0 x
// b4 b3 b2 b1 b0 =
// ------------------------
// a4b0 a3b0 a2b0 a1b0 a0b0 +
// a4b1 a3b1 a2b1 a1b1 a0b1 +
// a4b2 a3b2 a2b2 a1b2 a0b2 +
// a4b3 a3b3 a2b3 a1b3 a0b3 +
// a4b4 a3b4 a2b4 a1b4 a0b4 =
// ----------------------------------------------
// r8 r7 r6 r5 r4 r3 r2 r1 r0
//
// We can then use the reduction identity (a * 2²⁵⁵ + b = a * 19 + b) to
// reduce the limbs that would overflow 255 bits. r5 * 2²⁵⁵ becomes 19 * r5,
// r6 * 2³⁰⁶ becomes 19 * r6 * 2⁵¹, etc.
//
// Reduction can be carried out simultaneously to multiplication. For
// example, we do not compute r5: whenever the result of a multiplication
// belongs to r5, like a1b4, we multiply it by 19 and add the result to r0.
//
// a4b0 a3b0 a2b0 a1b0 a0b0 +
// a3b1 a2b1 a1b1 a0b1 19×a4b1 +
// a2b2 a1b2 a0b2 19×a4b2 19×a3b2 +
// a1b3 a0b3 19×a4b3 19×a3b3 19×a2b3 +
// a0b4 19×a4b4 19×a3b4 19×a2b4 19×a1b4 =
// --------------------------------------
// r4 r3 r2 r1 r0
//
// Finally we add up the columns into wide, overlapping limbs.
a1_19 := a1 * 19
a2_19 := a2 * 19
a3_19 := a3 * 19
a4_19 := a4 * 19
// r0 = a0×b0 + 19×(a1×b4 + a2×b3 + a3×b2 + a4×b1)
r0 := mul64(a0, b0)
r0 = addMul64(r0, a1_19, b4)
r0 = addMul64(r0, a2_19, b3)
r0 = addMul64(r0, a3_19, b2)
r0 = addMul64(r0, a4_19, b1)
// r1 = a0×b1 + a1×b0 + 19×(a2×b4 + a3×b3 + a4×b2)
r1 := mul64(a0, b1)
r1 = addMul64(r1, a1, b0)
r1 = addMul64(r1, a2_19, b4)
r1 = addMul64(r1, a3_19, b3)
r1 = addMul64(r1, a4_19, b2)
// r2 = a0×b2 + a1×b1 + a2×b0 + 19×(a3×b4 + a4×b3)
r2 := mul64(a0, b2)
r2 = addMul64(r2, a1, b1)
r2 = addMul64(r2, a2, b0)
r2 = addMul64(r2, a3_19, b4)
r2 = addMul64(r2, a4_19, b3)
// r3 = a0×b3 + a1×b2 + a2×b1 + a3×b0 + 19×a4×b4
r3 := mul64(a0, b3)
r3 = addMul64(r3, a1, b2)
r3 = addMul64(r3, a2, b1)
r3 = addMul64(r3, a3, b0)
r3 = addMul64(r3, a4_19, b4)
// r4 = a0×b4 + a1×b3 + a2×b2 + a3×b1 + a4×b0
r4 := mul64(a0, b4)
r4 = addMul64(r4, a1, b3)
r4 = addMul64(r4, a2, b2)
r4 = addMul64(r4, a3, b1)
r4 = addMul64(r4, a4, b0)
// After the multiplication, we need to reduce (carry) the five coefficients
// to obtain a result with limbs that are at most slightly larger than 2⁵¹,
// to respect the Element invariant.
//
// Overall, the reduction works the same as carryPropagate, except with
// wider inputs: we take the carry for each coefficient by shifting it right
// by 51, and add it to the limb above it. The top carry is multiplied by 19
// according to the reduction identity and added to the lowest limb.
//
// The largest coefficient (r0) will be at most 111 bits, which guarantees
// that all carries are at most 111 - 51 = 60 bits, which fits in a uint64.
//
// r0 = a0×b0 + 19×(a1×b4 + a2×b3 + a3×b2 + a4×b1)
// r0 < 2⁵²×2⁵² + 19×(2⁵²×2⁵² + 2⁵²×2⁵² + 2⁵²×2⁵² + 2⁵²×2⁵²)
// r0 < (1 + 19 × 4) × 2⁵² × 2⁵²
// r0 < 2⁷ × 2⁵² × 2⁵²
// r0 < 2¹¹¹
//
// Moreover, the top coefficient (r4) is at most 107 bits, so c4 is at most
// 56 bits, and c4 * 19 is at most 61 bits, which again fits in a uint64 and
// allows us to easily apply the reduction identity.
//
// r4 = a0×b4 + a1×b3 + a2×b2 + a3×b1 + a4×b0
// r4 < 5 × 2⁵² × 2⁵²
// r4 < 2¹⁰⁷
//
c0 := shiftRightBy51(r0)
c1 := shiftRightBy51(r1)
c2 := shiftRightBy51(r2)
c3 := shiftRightBy51(r3)
c4 := shiftRightBy51(r4)
rr0 := r0.lo&maskLow51Bits + c4*19
rr1 := r1.lo&maskLow51Bits + c0
rr2 := r2.lo&maskLow51Bits + c1
rr3 := r3.lo&maskLow51Bits + c2
rr4 := r4.lo&maskLow51Bits + c3
// Now all coefficients fit into 64-bit registers but are still too large to
// be passed around as a Element. We therefore do one last carry chain,
// where the carries will be small enough to fit in the wiggle room above 2⁵¹.
*v = Element{rr0, rr1, rr2, rr3, rr4}
v.carryPropagate()
}
func feSquareGeneric(v, a *Element) {
l0 := a.l0
l1 := a.l1
l2 := a.l2
l3 := a.l3
l4 := a.l4
// Squaring works precisely like multiplication above, but thanks to its
// symmetry we get to group a few terms together.
//
// l4 l3 l2 l1 l0 x
// l4 l3 l2 l1 l0 =
// ------------------------
// l4l0 l3l0 l2l0 l1l0 l0l0 +
// l4l1 l3l1 l2l1 l1l1 l0l1 +
// l4l2 l3l2 l2l2 l1l2 l0l2 +
// l4l3 l3l3 l2l3 l1l3 l0l3 +
// l4l4 l3l4 l2l4 l1l4 l0l4 =
// ----------------------------------------------
// r8 r7 r6 r5 r4 r3 r2 r1 r0
//
// l4l0 l3l0 l2l0 l1l0 l0l0 +
// l3l1 l2l1 l1l1 l0l1 19×l4l1 +
// l2l2 l1l2 l0l2 19×l4l2 19×l3l2 +
// l1l3 l0l3 19×l4l3 19×l3l3 19×l2l3 +
// l0l4 19×l4l4 19×l3l4 19×l2l4 19×l1l4 =
// --------------------------------------
// r4 r3 r2 r1 r0
//
// With precomputed 2×, 19×, and 2×19× terms, we can compute each limb with
// only three Mul64 and four Add64, instead of five and eight.
l0_2 := l0 * 2
l1_2 := l1 * 2
l1_38 := l1 * 38
l2_38 := l2 * 38
l3_38 := l3 * 38
l3_19 := l3 * 19
l4_19 := l4 * 19
// r0 = l0×l0 + 19×(l1×l4 + l2×l3 + l3×l2 + l4×l1) = l0×l0 + 19×2×(l1×l4 + l2×l3)
r0 := mul64(l0, l0)
r0 = addMul64(r0, l1_38, l4)
r0 = addMul64(r0, l2_38, l3)
// r1 = l0×l1 + l1×l0 + 19×(l2×l4 + l3×l3 + l4×l2) = 2×l0×l1 + 19×2×l2×l4 + 19×l3×l3
r1 := mul64(l0_2, l1)
r1 = addMul64(r1, l2_38, l4)
r1 = addMul64(r1, l3_19, l3)
// r2 = l0×l2 + l1×l1 + l2×l0 + 19×(l3×l4 + l4×l3) = 2×l0×l2 + l1×l1 + 19×2×l3×l4
r2 := mul64(l0_2, l2)
r2 = addMul64(r2, l1, l1)
r2 = addMul64(r2, l3_38, l4)
// r3 = l0×l3 + l1×l2 + l2×l1 + l3×l0 + 19×l4×l4 = 2×l0×l3 + 2×l1×l2 + 19×l4×l4
r3 := mul64(l0_2, l3)
r3 = addMul64(r3, l1_2, l2)
r3 = addMul64(r3, l4_19, l4)
// r4 = l0×l4 + l1×l3 + l2×l2 + l3×l1 + l4×l0 = 2×l0×l4 + 2×l1×l3 + l2×l2
r4 := mul64(l0_2, l4)
r4 = addMul64(r4, l1_2, l3)
r4 = addMul64(r4, l2, l2)
c0 := shiftRightBy51(r0)
c1 := shiftRightBy51(r1)
c2 := shiftRightBy51(r2)
c3 := shiftRightBy51(r3)
c4 := shiftRightBy51(r4)
rr0 := r0.lo&maskLow51Bits + c4*19
rr1 := r1.lo&maskLow51Bits + c0
rr2 := r2.lo&maskLow51Bits + c1
rr3 := r3.lo&maskLow51Bits + c2
rr4 := r4.lo&maskLow51Bits + c3
*v = Element{rr0, rr1, rr2, rr3, rr4}
v.carryPropagate()
}
// carryPropagateGeneric brings the limbs below 52 bits by applying the reduction
// identity (a * 2²⁵⁵ + b = a * 19 + b) to the l4 carry. TODO inline
func (v *Element) carryPropagateGeneric() *Element {
c0 := v.l0 >> 51
c1 := v.l1 >> 51
c2 := v.l2 >> 51
c3 := v.l3 >> 51
c4 := v.l4 >> 51
v.l0 = v.l0&maskLow51Bits + c4*19
v.l1 = v.l1&maskLow51Bits + c0
v.l2 = v.l2&maskLow51Bits + c1
v.l3 = v.l3&maskLow51Bits + c2
v.l4 = v.l4&maskLow51Bits + c3
return v
}

1
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/sync.checkpoint generated vendored Normal file
View File

@ -0,0 +1 @@
b0c49ae9f59d233526f8934262c5bbbe14d4358d

19
gateway/vendor/golang.org/x/crypto/curve25519/internal/field/sync.sh generated vendored Normal file
View File

@ -0,0 +1,19 @@
#! /bin/bash
set -euo pipefail
cd "$(git rev-parse --show-toplevel)"
STD_PATH=src/crypto/ed25519/internal/edwards25519/field
LOCAL_PATH=curve25519/internal/field
LAST_SYNC_REF=$(cat $LOCAL_PATH/sync.checkpoint)
git fetch https://go.googlesource.com/go master
if git diff --quiet $LAST_SYNC_REF:$STD_PATH FETCH_HEAD:$STD_PATH; then
echo "No changes."
else
NEW_REF=$(git rev-parse FETCH_HEAD | tee $LOCAL_PATH/sync.checkpoint)
echo "Applying changes from $LAST_SYNC_REF to $NEW_REF..."
git diff $LAST_SYNC_REF:$STD_PATH FETCH_HEAD:$STD_PATH | \
git apply -3 --directory=$LOCAL_PATH
fi

32
gateway/vendor/golang.org/x/crypto/internal/alias/alias.go generated vendored Normal file
View File

@ -0,0 +1,32 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !purego
// +build !purego
// Package alias implements memory aliasing tests.
package alias
import "unsafe"
// AnyOverlap reports whether x and y share memory at any (not necessarily
// corresponding) index. The memory beyond the slice length is ignored.
func AnyOverlap(x, y []byte) bool {
return len(x) > 0 && len(y) > 0 &&
uintptr(unsafe.Pointer(&x[0])) <= uintptr(unsafe.Pointer(&y[len(y)-1])) &&
uintptr(unsafe.Pointer(&y[0])) <= uintptr(unsafe.Pointer(&x[len(x)-1]))
}
// InexactOverlap reports whether x and y share memory at any non-corresponding
// index. The memory beyond the slice length is ignored. Note that x and y can
// have different lengths and still not have any inexact overlap.
//
// InexactOverlap can be used to implement the requirements of the crypto/cipher
// AEAD, Block, BlockMode and Stream interfaces.
func InexactOverlap(x, y []byte) bool {
if len(x) == 0 || len(y) == 0 || &x[0] == &y[0] {
return false
}
return AnyOverlap(x, y)
}

35
gateway/vendor/golang.org/x/crypto/internal/alias/alias_purego.go generated vendored Normal file
View File

@ -0,0 +1,35 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build purego
// +build purego
// Package alias implements memory aliasing tests.
package alias
// This is the Google App Engine standard variant based on reflect
// because the unsafe package and cgo are disallowed.
import "reflect"
// AnyOverlap reports whether x and y share memory at any (not necessarily
// corresponding) index. The memory beyond the slice length is ignored.
func AnyOverlap(x, y []byte) bool {
return len(x) > 0 && len(y) > 0 &&
reflect.ValueOf(&x[0]).Pointer() <= reflect.ValueOf(&y[len(y)-1]).Pointer() &&
reflect.ValueOf(&y[0]).Pointer() <= reflect.ValueOf(&x[len(x)-1]).Pointer()
}
// InexactOverlap reports whether x and y share memory at any non-corresponding
// index. The memory beyond the slice length is ignored. Note that x and y can
// have different lengths and still not have any inexact overlap.
//
// InexactOverlap can be used to implement the requirements of the crypto/cipher
// AEAD, Block, BlockMode and Stream interfaces.
func InexactOverlap(x, y []byte) bool {
if len(x) == 0 || len(y) == 0 || &x[0] == &y[0] {
return false
}
return AnyOverlap(x, y)
}

40
gateway/vendor/golang.org/x/crypto/internal/poly1305/bits_compat.go generated vendored Normal file
View File

@ -0,0 +1,40 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !go1.13
// +build !go1.13
package poly1305
// Generic fallbacks for the math/bits intrinsics, copied from
// src/math/bits/bits.go. They were added in Go 1.12, but Add64 and Sum64 had
// variable time fallbacks until Go 1.13.
func bitsAdd64(x, y, carry uint64) (sum, carryOut uint64) {
sum = x + y + carry
carryOut = ((x & y) | ((x | y) &^ sum)) >> 63
return
}
func bitsSub64(x, y, borrow uint64) (diff, borrowOut uint64) {
diff = x - y - borrow
borrowOut = ((^x & y) | (^(x ^ y) & diff)) >> 63
return
}
func bitsMul64(x, y uint64) (hi, lo uint64) {
const mask32 = 1<<32 - 1
x0 := x & mask32
x1 := x >> 32
y0 := y & mask32
y1 := y >> 32
w0 := x0 * y0
t := x1*y0 + w0>>32
w1 := t & mask32
w2 := t >> 32
w1 += x0 * y1
hi = x1*y1 + w2 + w1>>32
lo = x * y
return
}

22
gateway/vendor/golang.org/x/crypto/internal/poly1305/bits_go1.13.go generated vendored Normal file
View File

@ -0,0 +1,22 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build go1.13
// +build go1.13
package poly1305
import "math/bits"
func bitsAdd64(x, y, carry uint64) (sum, carryOut uint64) {
return bits.Add64(x, y, carry)
}
func bitsSub64(x, y, borrow uint64) (diff, borrowOut uint64) {
return bits.Sub64(x, y, borrow)
}
func bitsMul64(x, y uint64) (hi, lo uint64) {
return bits.Mul64(x, y)
}

10
gateway/vendor/golang.org/x/crypto/internal/poly1305/mac_noasm.go generated vendored Normal file
View File

@ -0,0 +1,10 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build (!amd64 && !ppc64le && !s390x) || !gc || purego
// +build !amd64,!ppc64le,!s390x !gc purego
package poly1305
type mac struct{ macGeneric }

99
gateway/vendor/golang.org/x/crypto/internal/poly1305/poly1305.go generated vendored Normal file
View File

@ -0,0 +1,99 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package poly1305 implements Poly1305 one-time message authentication code as
// specified in https://cr.yp.to/mac/poly1305-20050329.pdf.
//
// Poly1305 is a fast, one-time authentication function. It is infeasible for an
// attacker to generate an authenticator for a message without the key. However, a
// key must only be used for a single message. Authenticating two different
// messages with the same key allows an attacker to forge authenticators for other
// messages with the same key.
//
// Poly1305 was originally coupled with AES in order to make Poly1305-AES. AES was
// used with a fixed key in order to generate one-time keys from an nonce.
// However, in this package AES isn't used and the one-time key is specified
// directly.
package poly1305
import "crypto/subtle"
// TagSize is the size, in bytes, of a poly1305 authenticator.
const TagSize = 16
// Sum generates an authenticator for msg using a one-time key and puts the
// 16-byte result into out. Authenticating two different messages with the same
// key allows an attacker to forge messages at will.
func Sum(out *[16]byte, m []byte, key *[32]byte) {
h := New(key)
h.Write(m)
h.Sum(out[:0])
}
// Verify returns true if mac is a valid authenticator for m with the given key.
func Verify(mac *[16]byte, m []byte, key *[32]byte) bool {
var tmp [16]byte
Sum(&tmp, m, key)
return subtle.ConstantTimeCompare(tmp[:], mac[:]) == 1
}
// New returns a new MAC computing an authentication
// tag of all data written to it with the given key.
// This allows writing the message progressively instead
// of passing it as a single slice. Common users should use
// the Sum function instead.
//
// The key must be unique for each message, as authenticating
// two different messages with the same key allows an attacker
// to forge messages at will.
func New(key *[32]byte) *MAC {
m := &MAC{}
initialize(key, &m.macState)
return m
}
// MAC is an io.Writer computing an authentication tag
// of the data written to it.
//
// MAC cannot be used like common hash.Hash implementations,
// because using a poly1305 key twice breaks its security.
// Therefore writing data to a running MAC after calling
// Sum or Verify causes it to panic.
type MAC struct {
mac // platform-dependent implementation
finalized bool
}
// Size returns the number of bytes Sum will return.
func (h *MAC) Size() int { return TagSize }
// Write adds more data to the running message authentication code.
// It never returns an error.
//
// It must not be called after the first call of Sum or Verify.
func (h *MAC) Write(p []byte) (n int, err error) {
if h.finalized {
panic("poly1305: write to MAC after Sum or Verify")
}
return h.mac.Write(p)
}
// Sum computes the authenticator of all data written to the
// message authentication code.
func (h *MAC) Sum(b []byte) []byte {
var mac [TagSize]byte
h.mac.Sum(&mac)
h.finalized = true
return append(b, mac[:]...)
}
// Verify returns whether the authenticator of all data written to
// the message authentication code matches the expected value.
func (h *MAC) Verify(expected []byte) bool {
var mac [TagSize]byte
h.mac.Sum(&mac)
h.finalized = true
return subtle.ConstantTimeCompare(expected, mac[:]) == 1
}

48
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_amd64.go generated vendored Normal file
View File

@ -0,0 +1,48 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc && !purego
// +build gc,!purego
package poly1305
//go:noescape
func update(state *macState, msg []byte)
// mac is a wrapper for macGeneric that redirects calls that would have gone to
// updateGeneric to update.
//
// Its Write and Sum methods are otherwise identical to the macGeneric ones, but
// using function pointers would carry a major performance cost.
type mac struct{ macGeneric }
func (h *mac) Write(p []byte) (int, error) {
nn := len(p)
if h.offset > 0 {
n := copy(h.buffer[h.offset:], p)
if h.offset+n < TagSize {
h.offset += n
return nn, nil
}
p = p[n:]
h.offset = 0
update(&h.macState, h.buffer[:])
}
if n := len(p) - (len(p) % TagSize); n > 0 {
update(&h.macState, p[:n])
p = p[n:]
}
if len(p) > 0 {
h.offset += copy(h.buffer[h.offset:], p)
}
return nn, nil
}
func (h *mac) Sum(out *[16]byte) {
state := h.macState
if h.offset > 0 {
update(&state, h.buffer[:h.offset])
}
finalize(out, &state.h, &state.s)
}

109
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_amd64.s generated vendored Normal file
View File

@ -0,0 +1,109 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc && !purego
// +build gc,!purego
#include "textflag.h"
#define POLY1305_ADD(msg, h0, h1, h2) \
ADDQ 0(msg), h0; \
ADCQ 8(msg), h1; \
ADCQ $1, h2; \
LEAQ 16(msg), msg
#define POLY1305_MUL(h0, h1, h2, r0, r1, t0, t1, t2, t3) \
MOVQ r0, AX; \
MULQ h0; \
MOVQ AX, t0; \
MOVQ DX, t1; \
MOVQ r0, AX; \
MULQ h1; \
ADDQ AX, t1; \
ADCQ $0, DX; \
MOVQ r0, t2; \
IMULQ h2, t2; \
ADDQ DX, t2; \
\
MOVQ r1, AX; \
MULQ h0; \
ADDQ AX, t1; \
ADCQ $0, DX; \
MOVQ DX, h0; \
MOVQ r1, t3; \
IMULQ h2, t3; \
MOVQ r1, AX; \
MULQ h1; \
ADDQ AX, t2; \
ADCQ DX, t3; \
ADDQ h0, t2; \
ADCQ $0, t3; \
\
MOVQ t0, h0; \
MOVQ t1, h1; \
MOVQ t2, h2; \
ANDQ $3, h2; \
MOVQ t2, t0; \
ANDQ $0xFFFFFFFFFFFFFFFC, t0; \
ADDQ t0, h0; \
ADCQ t3, h1; \
ADCQ $0, h2; \
SHRQ $2, t3, t2; \
SHRQ $2, t3; \
ADDQ t2, h0; \
ADCQ t3, h1; \
ADCQ $0, h2
// func update(state *[7]uint64, msg []byte)
TEXT ·update(SB), $0-32
MOVQ state+0(FP), DI
MOVQ msg_base+8(FP), SI
MOVQ msg_len+16(FP), R15
MOVQ 0(DI), R8 // h0
MOVQ 8(DI), R9 // h1
MOVQ 16(DI), R10 // h2
MOVQ 24(DI), R11 // r0
MOVQ 32(DI), R12 // r1
CMPQ R15, $16
JB bytes_between_0_and_15
loop:
POLY1305_ADD(SI, R8, R9, R10)
multiply:
POLY1305_MUL(R8, R9, R10, R11, R12, BX, CX, R13, R14)
SUBQ $16, R15
CMPQ R15, $16
JAE loop
bytes_between_0_and_15:
TESTQ R15, R15
JZ done
MOVQ $1, BX
XORQ CX, CX
XORQ R13, R13
ADDQ R15, SI
flush_buffer:
SHLQ $8, BX, CX
SHLQ $8, BX
MOVB -1(SI), R13
XORQ R13, BX
DECQ SI
DECQ R15
JNZ flush_buffer
ADDQ BX, R8
ADCQ CX, R9
ADCQ $0, R10
MOVQ $16, R15
JMP multiply
done:
MOVQ R8, 0(DI)
MOVQ R9, 8(DI)
MOVQ R10, 16(DI)
RET

309
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_generic.go generated vendored Normal file
View File

@ -0,0 +1,309 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// This file provides the generic implementation of Sum and MAC. Other files
// might provide optimized assembly implementations of some of this code.
package poly1305
import "encoding/binary"
// Poly1305 [RFC 7539] is a relatively simple algorithm: the authentication tag
// for a 64 bytes message is approximately
//
// s + m[0:16] * r⁴ + m[16:32] * r³ + m[32:48] * r² + m[48:64] * r mod 2¹³⁰ - 5
//
// for some secret r and s. It can be computed sequentially like
//
// for len(msg) > 0:
// h += read(msg, 16)
// h *= r
// h %= 2¹³⁰ - 5
// return h + s
//
// All the complexity is about doing performant constant-time math on numbers
// larger than any available numeric type.
func sumGeneric(out *[TagSize]byte, msg []byte, key *[32]byte) {
h := newMACGeneric(key)
h.Write(msg)
h.Sum(out)
}
func newMACGeneric(key *[32]byte) macGeneric {
m := macGeneric{}
initialize(key, &m.macState)
return m
}
// macState holds numbers in saturated 64-bit little-endian limbs. That is,
// the value of [x0, x1, x2] is x[0] + x[1] * 2⁶⁴ + x[2] * 2¹²⁸.
type macState struct {
// h is the main accumulator. It is to be interpreted modulo 2¹³⁰ - 5, but
// can grow larger during and after rounds. It must, however, remain below
// 2 * (2¹³⁰ - 5).
h [3]uint64
// r and s are the private key components.
r [2]uint64
s [2]uint64
}
type macGeneric struct {
macState
buffer [TagSize]byte
offset int
}
// Write splits the incoming message into TagSize chunks, and passes them to
// update. It buffers incomplete chunks.
func (h *macGeneric) Write(p []byte) (int, error) {
nn := len(p)
if h.offset > 0 {
n := copy(h.buffer[h.offset:], p)
if h.offset+n < TagSize {
h.offset += n
return nn, nil
}
p = p[n:]
h.offset = 0
updateGeneric(&h.macState, h.buffer[:])
}
if n := len(p) - (len(p) % TagSize); n > 0 {
updateGeneric(&h.macState, p[:n])
p = p[n:]
}
if len(p) > 0 {
h.offset += copy(h.buffer[h.offset:], p)
}
return nn, nil
}
// Sum flushes the last incomplete chunk from the buffer, if any, and generates
// the MAC output. It does not modify its state, in order to allow for multiple
// calls to Sum, even if no Write is allowed after Sum.
func (h *macGeneric) Sum(out *[TagSize]byte) {
state := h.macState
if h.offset > 0 {
updateGeneric(&state, h.buffer[:h.offset])
}
finalize(out, &state.h, &state.s)
}
// [rMask0, rMask1] is the specified Poly1305 clamping mask in little-endian. It
// clears some bits of the secret coefficient to make it possible to implement
// multiplication more efficiently.
const (
rMask0 = 0x0FFFFFFC0FFFFFFF
rMask1 = 0x0FFFFFFC0FFFFFFC
)
// initialize loads the 256-bit key into the two 128-bit secret values r and s.
func initialize(key *[32]byte, m *macState) {
m.r[0] = binary.LittleEndian.Uint64(key[0:8]) & rMask0
m.r[1] = binary.LittleEndian.Uint64(key[8:16]) & rMask1
m.s[0] = binary.LittleEndian.Uint64(key[16:24])
m.s[1] = binary.LittleEndian.Uint64(key[24:32])
}
// uint128 holds a 128-bit number as two 64-bit limbs, for use with the
// bits.Mul64 and bits.Add64 intrinsics.
type uint128 struct {
lo, hi uint64
}
func mul64(a, b uint64) uint128 {
hi, lo := bitsMul64(a, b)
return uint128{lo, hi}
}
func add128(a, b uint128) uint128 {
lo, c := bitsAdd64(a.lo, b.lo, 0)
hi, c := bitsAdd64(a.hi, b.hi, c)
if c != 0 {
panic("poly1305: unexpected overflow")
}
return uint128{lo, hi}
}
func shiftRightBy2(a uint128) uint128 {
a.lo = a.lo>>2 | (a.hi&3)<<62
a.hi = a.hi >> 2
return a
}
// updateGeneric absorbs msg into the state.h accumulator. For each chunk m of
// 128 bits of message, it computes
//
// h₊ = (h + m) * r mod 2¹³⁰ - 5
//
// If the msg length is not a multiple of TagSize, it assumes the last
// incomplete chunk is the final one.
func updateGeneric(state *macState, msg []byte) {
h0, h1, h2 := state.h[0], state.h[1], state.h[2]
r0, r1 := state.r[0], state.r[1]
for len(msg) > 0 {
var c uint64
// For the first step, h + m, we use a chain of bits.Add64 intrinsics.
// The resulting value of h might exceed 2¹³⁰ - 5, but will be partially
// reduced at the end of the multiplication below.
//
// The spec requires us to set a bit just above the message size, not to
// hide leading zeroes. For full chunks, that's 1 << 128, so we can just
// add 1 to the most significant (2¹²⁸) limb, h2.
if len(msg) >= TagSize {
h0, c = bitsAdd64(h0, binary.LittleEndian.Uint64(msg[0:8]), 0)
h1, c = bitsAdd64(h1, binary.LittleEndian.Uint64(msg[8:16]), c)
h2 += c + 1
msg = msg[TagSize:]
} else {
var buf [TagSize]byte
copy(buf[:], msg)
buf[len(msg)] = 1
h0, c = bitsAdd64(h0, binary.LittleEndian.Uint64(buf[0:8]), 0)
h1, c = bitsAdd64(h1, binary.LittleEndian.Uint64(buf[8:16]), c)
h2 += c
msg = nil
}
// Multiplication of big number limbs is similar to elementary school
// columnar multiplication. Instead of digits, there are 64-bit limbs.
//
// We are multiplying a 3 limbs number, h, by a 2 limbs number, r.
//
// h2 h1 h0 x
// r1 r0 =
// ----------------
// h2r0 h1r0 h0r0 <-- individual 128-bit products
// + h2r1 h1r1 h0r1
// ------------------------
// m3 m2 m1 m0 <-- result in 128-bit overlapping limbs
// ------------------------
// m3.hi m2.hi m1.hi m0.hi <-- carry propagation
// + m3.lo m2.lo m1.lo m0.lo
// -------------------------------
// t4 t3 t2 t1 t0 <-- final result in 64-bit limbs
//
// The main difference from pen-and-paper multiplication is that we do
// carry propagation in a separate step, as if we wrote two digit sums
// at first (the 128-bit limbs), and then carried the tens all at once.
h0r0 := mul64(h0, r0)
h1r0 := mul64(h1, r0)
h2r0 := mul64(h2, r0)
h0r1 := mul64(h0, r1)
h1r1 := mul64(h1, r1)
h2r1 := mul64(h2, r1)
// Since h2 is known to be at most 7 (5 + 1 + 1), and r0 and r1 have their
// top 4 bits cleared by rMask{0,1}, we know that their product is not going
// to overflow 64 bits, so we can ignore the high part of the products.
//
// This also means that the product doesn't have a fifth limb (t4).
if h2r0.hi != 0 {
panic("poly1305: unexpected overflow")
}
if h2r1.hi != 0 {
panic("poly1305: unexpected overflow")
}
m0 := h0r0
m1 := add128(h1r0, h0r1) // These two additions don't overflow thanks again
m2 := add128(h2r0, h1r1) // to the 4 masked bits at the top of r0 and r1.
m3 := h2r1
t0 := m0.lo
t1, c := bitsAdd64(m1.lo, m0.hi, 0)
t2, c := bitsAdd64(m2.lo, m1.hi, c)
t3, _ := bitsAdd64(m3.lo, m2.hi, c)
// Now we have the result as 4 64-bit limbs, and we need to reduce it
// modulo 2¹³⁰ - 5. The special shape of this Crandall prime lets us do
// a cheap partial reduction according to the reduction identity
//
// c * 2¹³⁰ + n = c * 5 + n mod 2¹³⁰ - 5
//
// because 2¹³⁰ = 5 mod 2¹³⁰ - 5. Partial reduction since the result is
// likely to be larger than 2¹³⁰ - 5, but still small enough to fit the
// assumptions we make about h in the rest of the code.
//
// See also https://speakerdeck.com/gtank/engineering-prime-numbers?slide=23
// We split the final result at the 2¹³⁰ mark into h and cc, the carry.
// Note that the carry bits are effectively shifted left by 2, in other
// words, cc = c * 4 for the c in the reduction identity.
h0, h1, h2 = t0, t1, t2&maskLow2Bits
cc := uint128{t2 & maskNotLow2Bits, t3}
// To add c * 5 to h, we first add cc = c * 4, and then add (cc >> 2) = c.
h0, c = bitsAdd64(h0, cc.lo, 0)
h1, c = bitsAdd64(h1, cc.hi, c)
h2 += c
cc = shiftRightBy2(cc)
h0, c = bitsAdd64(h0, cc.lo, 0)
h1, c = bitsAdd64(h1, cc.hi, c)
h2 += c
// h2 is at most 3 + 1 + 1 = 5, making the whole of h at most
//
// 5 * 2¹²⁸ + (2¹²⁸ - 1) = 6 * 2¹²⁸ - 1
}
state.h[0], state.h[1], state.h[2] = h0, h1, h2
}
const (
maskLow2Bits uint64 = 0x0000000000000003
maskNotLow2Bits uint64 = ^maskLow2Bits
)
// select64 returns x if v == 1 and y if v == 0, in constant time.
func select64(v, x, y uint64) uint64 { return ^(v-1)&x | (v-1)&y }
// [p0, p1, p2] is 2¹³⁰ - 5 in little endian order.
const (
p0 = 0xFFFFFFFFFFFFFFFB
p1 = 0xFFFFFFFFFFFFFFFF
p2 = 0x0000000000000003
)
// finalize completes the modular reduction of h and computes
//
// out = h + s mod 2¹²⁸
func finalize(out *[TagSize]byte, h *[3]uint64, s *[2]uint64) {
h0, h1, h2 := h[0], h[1], h[2]
// After the partial reduction in updateGeneric, h might be more than
// 2¹³⁰ - 5, but will be less than 2 * (2¹³⁰ - 5). To complete the reduction
// in constant time, we compute t = h - (2¹³⁰ - 5), and select h as the
// result if the subtraction underflows, and t otherwise.
hMinusP0, b := bitsSub64(h0, p0, 0)
hMinusP1, b := bitsSub64(h1, p1, b)
_, b = bitsSub64(h2, p2, b)
// h = h if h < p else h - p
h0 = select64(b, h0, hMinusP0)
h1 = select64(b, h1, hMinusP1)
// Finally, we compute the last Poly1305 step
//
// tag = h + s mod 2¹²⁸
//
// by just doing a wide addition with the 128 low bits of h and discarding
// the overflow.
h0, c := bitsAdd64(h0, s[0], 0)
h1, _ = bitsAdd64(h1, s[1], c)
binary.LittleEndian.PutUint64(out[0:8], h0)
binary.LittleEndian.PutUint64(out[8:16], h1)
}

48
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_ppc64le.go generated vendored Normal file
View File

@ -0,0 +1,48 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc && !purego
// +build gc,!purego
package poly1305
//go:noescape
func update(state *macState, msg []byte)
// mac is a wrapper for macGeneric that redirects calls that would have gone to
// updateGeneric to update.
//
// Its Write and Sum methods are otherwise identical to the macGeneric ones, but
// using function pointers would carry a major performance cost.
type mac struct{ macGeneric }
func (h *mac) Write(p []byte) (int, error) {
nn := len(p)
if h.offset > 0 {
n := copy(h.buffer[h.offset:], p)
if h.offset+n < TagSize {
h.offset += n
return nn, nil
}
p = p[n:]
h.offset = 0
update(&h.macState, h.buffer[:])
}
if n := len(p) - (len(p) % TagSize); n > 0 {
update(&h.macState, p[:n])
p = p[n:]
}
if len(p) > 0 {
h.offset += copy(h.buffer[h.offset:], p)
}
return nn, nil
}
func (h *mac) Sum(out *[16]byte) {
state := h.macState
if h.offset > 0 {
update(&state, h.buffer[:h.offset])
}
finalize(out, &state.h, &state.s)
}

182
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_ppc64le.s generated vendored Normal file
View File

@ -0,0 +1,182 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc && !purego
// +build gc,!purego
#include "textflag.h"
// This was ported from the amd64 implementation.
#define POLY1305_ADD(msg, h0, h1, h2, t0, t1, t2) \
MOVD (msg), t0; \
MOVD 8(msg), t1; \
MOVD $1, t2; \
ADDC t0, h0, h0; \
ADDE t1, h1, h1; \
ADDE t2, h2; \
ADD $16, msg
#define POLY1305_MUL(h0, h1, h2, r0, r1, t0, t1, t2, t3, t4, t5) \
MULLD r0, h0, t0; \
MULLD r0, h1, t4; \
MULHDU r0, h0, t1; \
MULHDU r0, h1, t5; \
ADDC t4, t1, t1; \
MULLD r0, h2, t2; \
ADDZE t5; \
MULHDU r1, h0, t4; \
MULLD r1, h0, h0; \
ADD t5, t2, t2; \
ADDC h0, t1, t1; \
MULLD h2, r1, t3; \
ADDZE t4, h0; \
MULHDU r1, h1, t5; \
MULLD r1, h1, t4; \
ADDC t4, t2, t2; \
ADDE t5, t3, t3; \
ADDC h0, t2, t2; \
MOVD $-4, t4; \
MOVD t0, h0; \
MOVD t1, h1; \
ADDZE t3; \
ANDCC $3, t2, h2; \
AND t2, t4, t0; \
ADDC t0, h0, h0; \
ADDE t3, h1, h1; \
SLD $62, t3, t4; \
SRD $2, t2; \
ADDZE h2; \
OR t4, t2, t2; \
SRD $2, t3; \
ADDC t2, h0, h0; \
ADDE t3, h1, h1; \
ADDZE h2
DATA ·poly1305Mask<>+0x00(SB)/8, $0x0FFFFFFC0FFFFFFF
DATA ·poly1305Mask<>+0x08(SB)/8, $0x0FFFFFFC0FFFFFFC
GLOBL ·poly1305Mask<>(SB), RODATA, $16
// func update(state *[7]uint64, msg []byte)
TEXT ·update(SB), $0-32
MOVD state+0(FP), R3
MOVD msg_base+8(FP), R4
MOVD msg_len+16(FP), R5
MOVD 0(R3), R8 // h0
MOVD 8(R3), R9 // h1
MOVD 16(R3), R10 // h2
MOVD 24(R3), R11 // r0
MOVD 32(R3), R12 // r1
CMP R5, $16
BLT bytes_between_0_and_15
loop:
POLY1305_ADD(R4, R8, R9, R10, R20, R21, R22)
multiply:
POLY1305_MUL(R8, R9, R10, R11, R12, R16, R17, R18, R14, R20, R21)
ADD $-16, R5
CMP R5, $16
BGE loop
bytes_between_0_and_15:
CMP R5, $0
BEQ done
MOVD $0, R16 // h0
MOVD $0, R17 // h1
flush_buffer:
CMP R5, $8
BLE just1
MOVD $8, R21
SUB R21, R5, R21
// Greater than 8 -- load the rightmost remaining bytes in msg
// and put into R17 (h1)
MOVD (R4)(R21), R17
MOVD $16, R22
// Find the offset to those bytes
SUB R5, R22, R22
SLD $3, R22
// Shift to get only the bytes in msg
SRD R22, R17, R17
// Put 1 at high end
MOVD $1, R23
SLD $3, R21
SLD R21, R23, R23
OR R23, R17, R17
// Remainder is 8
MOVD $8, R5
just1:
CMP R5, $8
BLT less8
// Exactly 8
MOVD (R4), R16
CMP R17, $0
// Check if we've already set R17; if not
// set 1 to indicate end of msg.
BNE carry
MOVD $1, R17
BR carry
less8:
MOVD $0, R16 // h0
MOVD $0, R22 // shift count
CMP R5, $4
BLT less4
MOVWZ (R4), R16
ADD $4, R4
ADD $-4, R5
MOVD $32, R22
less4:
CMP R5, $2
BLT less2
MOVHZ (R4), R21
SLD R22, R21, R21
OR R16, R21, R16
ADD $16, R22
ADD $-2, R5
ADD $2, R4
less2:
CMP R5, $0
BEQ insert1
MOVBZ (R4), R21
SLD R22, R21, R21
OR R16, R21, R16
ADD $8, R22
insert1:
// Insert 1 at end of msg
MOVD $1, R21
SLD R22, R21, R21
OR R16, R21, R16
carry:
// Add new values to h0, h1, h2
ADDC R16, R8
ADDE R17, R9
ADDZE R10, R10
MOVD $16, R5
ADD R5, R4
BR multiply
done:
// Save h0, h1, h2 in state
MOVD R8, 0(R3)
MOVD R9, 8(R3)
MOVD R10, 16(R3)
RET

77
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_s390x.go generated vendored Normal file
View File

@ -0,0 +1,77 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc && !purego
// +build gc,!purego
package poly1305
import (
"golang.org/x/sys/cpu"
)
// updateVX is an assembly implementation of Poly1305 that uses vector
// instructions. It must only be called if the vector facility (vx) is
// available.
//
//go:noescape
func updateVX(state *macState, msg []byte)
// mac is a replacement for macGeneric that uses a larger buffer and redirects
// calls that would have gone to updateGeneric to updateVX if the vector
// facility is installed.
//
// A larger buffer is required for good performance because the vector
// implementation has a higher fixed cost per call than the generic
// implementation.
type mac struct {
macState
buffer [16 * TagSize]byte // size must be a multiple of block size (16)
offset int
}
func (h *mac) Write(p []byte) (int, error) {
nn := len(p)
if h.offset > 0 {
n := copy(h.buffer[h.offset:], p)
if h.offset+n < len(h.buffer) {
h.offset += n
return nn, nil
}
p = p[n:]
h.offset = 0
if cpu.S390X.HasVX {
updateVX(&h.macState, h.buffer[:])
} else {
updateGeneric(&h.macState, h.buffer[:])
}
}
tail := len(p) % len(h.buffer) // number of bytes to copy into buffer
body := len(p) - tail // number of bytes to process now
if body > 0 {
if cpu.S390X.HasVX {
updateVX(&h.macState, p[:body])
} else {
updateGeneric(&h.macState, p[:body])
}
}
h.offset = copy(h.buffer[:], p[body:]) // copy tail bytes - can be 0
return nn, nil
}
func (h *mac) Sum(out *[TagSize]byte) {
state := h.macState
remainder := h.buffer[:h.offset]
// Use the generic implementation if we have 2 or fewer blocks left
// to sum. The vector implementation has a higher startup time.
if cpu.S390X.HasVX && len(remainder) > 2*TagSize {
updateVX(&state, remainder)
} else if len(remainder) > 0 {
updateGeneric(&state, remainder)
}
finalize(out, &state.h, &state.s)
}

504
gateway/vendor/golang.org/x/crypto/internal/poly1305/sum_s390x.s generated vendored Normal file
View File

@ -0,0 +1,504 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc && !purego
// +build gc,!purego
#include "textflag.h"
// This implementation of Poly1305 uses the vector facility (vx)
// to process up to 2 blocks (32 bytes) per iteration using an
// algorithm based on the one described in:
//
// NEON crypto, Daniel J. Bernstein & Peter Schwabe
// https://cryptojedi.org/papers/neoncrypto-20120320.pdf
//
// This algorithm uses 5 26-bit limbs to represent a 130-bit
// value. These limbs are, for the most part, zero extended and
// placed into 64-bit vector register elements. Each vector
// register is 128-bits wide and so holds 2 of these elements.
// Using 26-bit limbs allows us plenty of headroom to accommodate
// accumulations before and after multiplication without
// overflowing either 32-bits (before multiplication) or 64-bits
// (after multiplication).
//
// In order to parallelise the operations required to calculate
// the sum we use two separate accumulators and then sum those
// in an extra final step. For compatibility with the generic
// implementation we perform this summation at the end of every
// updateVX call.
//
// To use two accumulators we must multiply the message blocks
// by r² rather than r. Only the final message block should be
// multiplied by r.
//
// Example:
//
// We want to calculate the sum (h) for a 64 byte message (m):
//
// h = m[0:16]r⁴ + m[16:32]r³ + m[32:48]r² + m[48:64]r
//
// To do this we split the calculation into the even indices
// and odd indices of the message. These form our SIMD 'lanes':
//
// h = m[ 0:16]r⁴ + m[32:48]r² + <- lane 0
// m[16:32]r³ + m[48:64]r <- lane 1
//
// To calculate this iteratively we refactor so that both lanes
// are written in terms of r² and r:
//
// h = (m[ 0:16]r² + m[32:48])r² + <- lane 0
// (m[16:32]r² + m[48:64])r <- lane 1
// ^ ^
// | coefficients for second iteration
// coefficients for first iteration
//
// So in this case we would have two iterations. In the first
// both lanes are multiplied by r². In the second only the
// first lane is multiplied by r² and the second lane is
// instead multiplied by r. This gives use the odd and even
// powers of r that we need from the original equation.
//
// Notation:
//
// h - accumulator
// r - key
// m - message
//
// [a, b] - SIMD register holding two 64-bit values
// [a, b, c, d] - SIMD register holding four 32-bit values
// xᵢ[n] - limb n of variable x with bit width i
//
// Limbs are expressed in little endian order, so for 26-bit
// limbs x₂₆[4] will be the most significant limb and x₂₆[0]
// will be the least significant limb.
// masking constants
#define MOD24 V0 // [0x0000000000ffffff, 0x0000000000ffffff] - mask low 24-bits
#define MOD26 V1 // [0x0000000003ffffff, 0x0000000003ffffff] - mask low 26-bits
// expansion constants (see EXPAND macro)
#define EX0 V2
#define EX1 V3
#define EX2 V4
// key (r², r or 1 depending on context)
#define R_0 V5
#define R_1 V6
#define R_2 V7
#define R_3 V8
#define R_4 V9
// precalculated coefficients (5r², 5r or 0 depending on context)
#define R5_1 V10
#define R5_2 V11
#define R5_3 V12
#define R5_4 V13
// message block (m)
#define M_0 V14
#define M_1 V15
#define M_2 V16
#define M_3 V17
#define M_4 V18
// accumulator (h)
#define H_0 V19
#define H_1 V20
#define H_2 V21
#define H_3 V22
#define H_4 V23
// temporary registers (for short-lived values)
#define T_0 V24
#define T_1 V25
#define T_2 V26
#define T_3 V27
#define T_4 V28
GLOBL ·constants<>(SB), RODATA, $0x30
// EX0
DATA ·constants<>+0x00(SB)/8, $0x0006050403020100
DATA ·constants<>+0x08(SB)/8, $0x1016151413121110
// EX1
DATA ·constants<>+0x10(SB)/8, $0x060c0b0a09080706
DATA ·constants<>+0x18(SB)/8, $0x161c1b1a19181716
// EX2
DATA ·constants<>+0x20(SB)/8, $0x0d0d0d0d0d0f0e0d
DATA ·constants<>+0x28(SB)/8, $0x1d1d1d1d1d1f1e1d
// MULTIPLY multiplies each lane of f and g, partially reduced
// modulo 2¹³⁰ - 5. The result, h, consists of partial products
// in each lane that need to be reduced further to produce the
// final result.
//
// h₁₃₀ = (f₁₃₀g₁₃₀) % 2¹³⁰ + (5f₁₃₀g₁₃₀) / 2¹³⁰
//
// Note that the multiplication by 5 of the high bits is
// achieved by precalculating the multiplication of four of the
// g coefficients by 5. These are g51-g54.
#define MULTIPLY(f0, f1, f2, f3, f4, g0, g1, g2, g3, g4, g51, g52, g53, g54, h0, h1, h2, h3, h4) \
VMLOF f0, g0, h0 \
VMLOF f0, g3, h3 \
VMLOF f0, g1, h1 \
VMLOF f0, g4, h4 \
VMLOF f0, g2, h2 \
VMLOF f1, g54, T_0 \
VMLOF f1, g2, T_3 \
VMLOF f1, g0, T_1 \
VMLOF f1, g3, T_4 \
VMLOF f1, g1, T_2 \
VMALOF f2, g53, h0, h0 \
VMALOF f2, g1, h3, h3 \
VMALOF f2, g54, h1, h1 \
VMALOF f2, g2, h4, h4 \
VMALOF f2, g0, h2, h2 \
VMALOF f3, g52, T_0, T_0 \
VMALOF f3, g0, T_3, T_3 \
VMALOF f3, g53, T_1, T_1 \
VMALOF f3, g1, T_4, T_4 \
VMALOF f3, g54, T_2, T_2 \
VMALOF f4, g51, h0, h0 \
VMALOF f4, g54, h3, h3 \
VMALOF f4, g52, h1, h1 \
VMALOF f4, g0, h4, h4 \
VMALOF f4, g53, h2, h2 \
VAG T_0, h0, h0 \
VAG T_3, h3, h3 \
VAG T_1, h1, h1 \
VAG T_4, h4, h4 \
VAG T_2, h2, h2
// REDUCE performs the following carry operations in four
// stages, as specified in Bernstein & Schwabe:
//
// 1: h₂₆[0]->h₂₆[1] h₂₆[3]->h₂₆[4]
// 2: h₂₆[1]->h₂₆[2] h₂₆[4]->h₂₆[0]
// 3: h₂₆[0]->h₂₆[1] h₂₆[2]->h₂₆[3]
// 4: h₂₆[3]->h₂₆[4]
//
// The result is that all of the limbs are limited to 26-bits
// except for h₂₆[1] and h₂₆[4] which are limited to 27-bits.
//
// Note that although each limb is aligned at 26-bit intervals
// they may contain values that exceed 2²⁶ - 1, hence the need
// to carry the excess bits in each limb.
#define REDUCE(h0, h1, h2, h3, h4) \
VESRLG $26, h0, T_0 \
VESRLG $26, h3, T_1 \
VN MOD26, h0, h0 \
VN MOD26, h3, h3 \
VAG T_0, h1, h1 \
VAG T_1, h4, h4 \
VESRLG $26, h1, T_2 \
VESRLG $26, h4, T_3 \
VN MOD26, h1, h1 \
VN MOD26, h4, h4 \
VESLG $2, T_3, T_4 \
VAG T_3, T_4, T_4 \
VAG T_2, h2, h2 \
VAG T_4, h0, h0 \
VESRLG $26, h2, T_0 \
VESRLG $26, h0, T_1 \
VN MOD26, h2, h2 \
VN MOD26, h0, h0 \
VAG T_0, h3, h3 \
VAG T_1, h1, h1 \
VESRLG $26, h3, T_2 \
VN MOD26, h3, h3 \
VAG T_2, h4, h4
// EXPAND splits the 128-bit little-endian values in0 and in1
// into 26-bit big-endian limbs and places the results into
// the first and second lane of d₂₆[0:4] respectively.
//
// The EX0, EX1 and EX2 constants are arrays of byte indices
// for permutation. The permutation both reverses the bytes
// in the input and ensures the bytes are copied into the
// destination limb ready to be shifted into their final
// position.
#define EXPAND(in0, in1, d0, d1, d2, d3, d4) \
VPERM in0, in1, EX0, d0 \
VPERM in0, in1, EX1, d2 \
VPERM in0, in1, EX2, d4 \
VESRLG $26, d0, d1 \
VESRLG $30, d2, d3 \
VESRLG $4, d2, d2 \
VN MOD26, d0, d0 \ // [in0₂₆[0], in1₂₆[0]]
VN MOD26, d3, d3 \ // [in0₂₆[3], in1₂₆[3]]
VN MOD26, d1, d1 \ // [in0₂₆[1], in1₂₆[1]]
VN MOD24, d4, d4 \ // [in0₂₆[4], in1₂₆[4]]
VN MOD26, d2, d2 // [in0₂₆[2], in1₂₆[2]]
// func updateVX(state *macState, msg []byte)
TEXT ·updateVX(SB), NOSPLIT, $0
MOVD state+0(FP), R1
LMG msg+8(FP), R2, R3 // R2=msg_base, R3=msg_len
// load EX0, EX1 and EX2
MOVD $·constants<>(SB), R5
VLM (R5), EX0, EX2
// generate masks
VGMG $(64-24), $63, MOD24 // [0x00ffffff, 0x00ffffff]
VGMG $(64-26), $63, MOD26 // [0x03ffffff, 0x03ffffff]
// load h (accumulator) and r (key) from state
VZERO T_1 // [0, 0]
VL 0(R1), T_0 // [h₆₄[0], h₆₄[1]]
VLEG $0, 16(R1), T_1 // [h₆₄[2], 0]
VL 24(R1), T_2 // [r₆₄[0], r₆₄[1]]
VPDI $0, T_0, T_2, T_3 // [h₆₄[0], r₆₄[0]]
VPDI $5, T_0, T_2, T_4 // [h₆₄[1], r₆₄[1]]
// unpack h and r into 26-bit limbs
// note: h₆₄[2] may have the low 3 bits set, so h₂₆[4] is a 27-bit value
VN MOD26, T_3, H_0 // [h₂₆[0], r₂₆[0]]
VZERO H_1 // [0, 0]
VZERO H_3 // [0, 0]
VGMG $(64-12-14), $(63-12), T_0 // [0x03fff000, 0x03fff000] - 26-bit mask with low 12 bits masked out
VESLG $24, T_1, T_1 // [h₆₄[2]<<24, 0]
VERIMG $-26&63, T_3, MOD26, H_1 // [h₂₆[1], r₂₆[1]]
VESRLG $+52&63, T_3, H_2 // [h₂₆[2], r₂₆[2]] - low 12 bits only
VERIMG $-14&63, T_4, MOD26, H_3 // [h₂₆[1], r₂₆[1]]
VESRLG $40, T_4, H_4 // [h₂₆[4], r₂₆[4]] - low 24 bits only
VERIMG $+12&63, T_4, T_0, H_2 // [h₂₆[2], r₂₆[2]] - complete
VO T_1, H_4, H_4 // [h₂₆[4], r₂₆[4]] - complete
// replicate r across all 4 vector elements
VREPF $3, H_0, R_0 // [r₂₆[0], r₂₆[0], r₂₆[0], r₂₆[0]]
VREPF $3, H_1, R_1 // [r₂₆[1], r₂₆[1], r₂₆[1], r₂₆[1]]
VREPF $3, H_2, R_2 // [r₂₆[2], r₂₆[2], r₂₆[2], r₂₆[2]]
VREPF $3, H_3, R_3 // [r₂₆[3], r₂₆[3], r₂₆[3], r₂₆[3]]
VREPF $3, H_4, R_4 // [r₂₆[4], r₂₆[4], r₂₆[4], r₂₆[4]]
// zero out lane 1 of h
VLEIG $1, $0, H_0 // [h₂₆[0], 0]
VLEIG $1, $0, H_1 // [h₂₆[1], 0]
VLEIG $1, $0, H_2 // [h₂₆[2], 0]
VLEIG $1, $0, H_3 // [h₂₆[3], 0]
VLEIG $1, $0, H_4 // [h₂₆[4], 0]
// calculate 5r (ignore least significant limb)
VREPIF $5, T_0
VMLF T_0, R_1, R5_1 // [5r₂₆[1], 5r₂₆[1], 5r₂₆[1], 5r₂₆[1]]
VMLF T_0, R_2, R5_2 // [5r₂₆[2], 5r₂₆[2], 5r₂₆[2], 5r₂₆[2]]
VMLF T_0, R_3, R5_3 // [5r₂₆[3], 5r₂₆[3], 5r₂₆[3], 5r₂₆[3]]
VMLF T_0, R_4, R5_4 // [5r₂₆[4], 5r₂₆[4], 5r₂₆[4], 5r₂₆[4]]
// skip r² calculation if we are only calculating one block
CMPBLE R3, $16, skip
// calculate r²
MULTIPLY(R_0, R_1, R_2, R_3, R_4, R_0, R_1, R_2, R_3, R_4, R5_1, R5_2, R5_3, R5_4, M_0, M_1, M_2, M_3, M_4)
REDUCE(M_0, M_1, M_2, M_3, M_4)
VGBM $0x0f0f, T_0
VERIMG $0, M_0, T_0, R_0 // [r₂₆[0], r²₂₆[0], r₂₆[0], r²₂₆[0]]
VERIMG $0, M_1, T_0, R_1 // [r₂₆[1], r²₂₆[1], r₂₆[1], r²₂₆[1]]
VERIMG $0, M_2, T_0, R_2 // [r₂₆[2], r²₂₆[2], r₂₆[2], r²₂₆[2]]
VERIMG $0, M_3, T_0, R_3 // [r₂₆[3], r²₂₆[3], r₂₆[3], r²₂₆[3]]
VERIMG $0, M_4, T_0, R_4 // [r₂₆[4], r²₂₆[4], r₂₆[4], r²₂₆[4]]
// calculate 5r² (ignore least significant limb)
VREPIF $5, T_0
VMLF T_0, R_1, R5_1 // [5r₂₆[1], 5r²₂₆[1], 5r₂₆[1], 5r²₂₆[1]]
VMLF T_0, R_2, R5_2 // [5r₂₆[2], 5r²₂₆[2], 5r₂₆[2], 5r²₂₆[2]]
VMLF T_0, R_3, R5_3 // [5r₂₆[3], 5r²₂₆[3], 5r₂₆[3], 5r²₂₆[3]]
VMLF T_0, R_4, R5_4 // [5r₂₆[4], 5r²₂₆[4], 5r₂₆[4], 5r²₂₆[4]]
loop:
CMPBLE R3, $32, b2 // 2 or fewer blocks remaining, need to change key coefficients
// load next 2 blocks from message
VLM (R2), T_0, T_1
// update message slice
SUB $32, R3
MOVD $32(R2), R2
// unpack message blocks into 26-bit big-endian limbs
EXPAND(T_0, T_1, M_0, M_1, M_2, M_3, M_4)
// add 2¹²⁸ to each message block value
VLEIB $4, $1, M_4
VLEIB $12, $1, M_4
multiply:
// accumulate the incoming message
VAG H_0, M_0, M_0
VAG H_3, M_3, M_3
VAG H_1, M_1, M_1
VAG H_4, M_4, M_4
VAG H_2, M_2, M_2
// multiply the accumulator by the key coefficient
MULTIPLY(M_0, M_1, M_2, M_3, M_4, R_0, R_1, R_2, R_3, R_4, R5_1, R5_2, R5_3, R5_4, H_0, H_1, H_2, H_3, H_4)
// carry and partially reduce the partial products
REDUCE(H_0, H_1, H_2, H_3, H_4)
CMPBNE R3, $0, loop
finish:
// sum lane 0 and lane 1 and put the result in lane 1
VZERO T_0
VSUMQG H_0, T_0, H_0
VSUMQG H_3, T_0, H_3
VSUMQG H_1, T_0, H_1
VSUMQG H_4, T_0, H_4
VSUMQG H_2, T_0, H_2
// reduce again after summation
// TODO(mundaym): there might be a more efficient way to do this
// now that we only have 1 active lane. For example, we could
// simultaneously pack the values as we reduce them.
REDUCE(H_0, H_1, H_2, H_3, H_4)
// carry h[1] through to h[4] so that only h[4] can exceed 2²⁶ - 1
// TODO(mundaym): in testing this final carry was unnecessary.
// Needs a proof before it can be removed though.
VESRLG $26, H_1, T_1
VN MOD26, H_1, H_1
VAQ T_1, H_2, H_2
VESRLG $26, H_2, T_2
VN MOD26, H_2, H_2
VAQ T_2, H_3, H_3
VESRLG $26, H_3, T_3
VN MOD26, H_3, H_3
VAQ T_3, H_4, H_4
// h is now < 2(2¹³⁰ - 5)
// Pack each lane in h₂₆[0:4] into h₁₂₈[0:1].
VESLG $26, H_1, H_1
VESLG $26, H_3, H_3
VO H_0, H_1, H_0
VO H_2, H_3, H_2
VESLG $4, H_2, H_2
VLEIB $7, $48, H_1
VSLB H_1, H_2, H_2
VO H_0, H_2, H_0
VLEIB $7, $104, H_1
VSLB H_1, H_4, H_3
VO H_3, H_0, H_0
VLEIB $7, $24, H_1
VSRLB H_1, H_4, H_1
// update state
VSTEG $1, H_0, 0(R1)
VSTEG $0, H_0, 8(R1)
VSTEG $1, H_1, 16(R1)
RET
b2: // 2 or fewer blocks remaining
CMPBLE R3, $16, b1
// Load the 2 remaining blocks (17-32 bytes remaining).
MOVD $-17(R3), R0 // index of final byte to load modulo 16
VL (R2), T_0 // load full 16 byte block
VLL R0, 16(R2), T_1 // load final (possibly partial) block and pad with zeros to 16 bytes
// The Poly1305 algorithm requires that a 1 bit be appended to
// each message block. If the final block is less than 16 bytes
// long then it is easiest to insert the 1 before the message
// block is split into 26-bit limbs. If, on the other hand, the
// final message block is 16 bytes long then we append the 1 bit
// after expansion as normal.
MOVBZ $1, R0
MOVD $-16(R3), R3 // index of byte in last block to insert 1 at (could be 16)
CMPBEQ R3, $16, 2(PC) // skip the insertion if the final block is 16 bytes long
VLVGB R3, R0, T_1 // insert 1 into the byte at index R3
// Split both blocks into 26-bit limbs in the appropriate lanes.
EXPAND(T_0, T_1, M_0, M_1, M_2, M_3, M_4)
// Append a 1 byte to the end of the second to last block.
VLEIB $4, $1, M_4
// Append a 1 byte to the end of the last block only if it is a
// full 16 byte block.
CMPBNE R3, $16, 2(PC)
VLEIB $12, $1, M_4
// Finally, set up the coefficients for the final multiplication.
// We have previously saved r and 5r in the 32-bit even indexes
// of the R_[0-4] and R5_[1-4] coefficient registers.
//
// We want lane 0 to be multiplied by r² so that can be kept the
// same. We want lane 1 to be multiplied by r so we need to move
// the saved r value into the 32-bit odd index in lane 1 by
// rotating the 64-bit lane by 32.
VGBM $0x00ff, T_0 // [0, 0xffffffffffffffff] - mask lane 1 only
VERIMG $32, R_0, T_0, R_0 // [_, r²₂₆[0], _, r₂₆[0]]
VERIMG $32, R_1, T_0, R_1 // [_, r²₂₆[1], _, r₂₆[1]]
VERIMG $32, R_2, T_0, R_2 // [_, r²₂₆[2], _, r₂₆[2]]
VERIMG $32, R_3, T_0, R_3 // [_, r²₂₆[3], _, r₂₆[3]]
VERIMG $32, R_4, T_0, R_4 // [_, r²₂₆[4], _, r₂₆[4]]
VERIMG $32, R5_1, T_0, R5_1 // [_, 5r²₂₆[1], _, 5r₂₆[1]]
VERIMG $32, R5_2, T_0, R5_2 // [_, 5r²₂₆[2], _, 5r₂₆[2]]
VERIMG $32, R5_3, T_0, R5_3 // [_, 5r²₂₆[3], _, 5r₂₆[3]]
VERIMG $32, R5_4, T_0, R5_4 // [_, 5r²₂₆[4], _, 5r₂₆[4]]
MOVD $0, R3
BR multiply
skip:
CMPBEQ R3, $0, finish
b1: // 1 block remaining
// Load the final block (1-16 bytes). This will be placed into
// lane 0.
MOVD $-1(R3), R0
VLL R0, (R2), T_0 // pad to 16 bytes with zeros
// The Poly1305 algorithm requires that a 1 bit be appended to
// each message block. If the final block is less than 16 bytes
// long then it is easiest to insert the 1 before the message
// block is split into 26-bit limbs. If, on the other hand, the
// final message block is 16 bytes long then we append the 1 bit
// after expansion as normal.
MOVBZ $1, R0
CMPBEQ R3, $16, 2(PC)
VLVGB R3, R0, T_0
// Set the message block in lane 1 to the value 0 so that it
// can be accumulated without affecting the final result.
VZERO T_1
// Split the final message block into 26-bit limbs in lane 0.
// Lane 1 will be contain 0.
EXPAND(T_0, T_1, M_0, M_1, M_2, M_3, M_4)
// Append a 1 byte to the end of the last block only if it is a
// full 16 byte block.
CMPBNE R3, $16, 2(PC)
VLEIB $4, $1, M_4
// We have previously saved r and 5r in the 32-bit even indexes
// of the R_[0-4] and R5_[1-4] coefficient registers.
//
// We want lane 0 to be multiplied by r so we need to move the
// saved r value into the 32-bit odd index in lane 0. We want
// lane 1 to be set to the value 1. This makes multiplication
// a no-op. We do this by setting lane 1 in every register to 0
// and then just setting the 32-bit index 3 in R_0 to 1.
VZERO T_0
MOVD $0, R0
MOVD $0x10111213, R12
VLVGP R12, R0, T_1 // [_, 0x10111213, _, 0x00000000]
VPERM T_0, R_0, T_1, R_0 // [_, r₂₆[0], _, 0]
VPERM T_0, R_1, T_1, R_1 // [_, r₂₆[1], _, 0]
VPERM T_0, R_2, T_1, R_2 // [_, r₂₆[2], _, 0]
VPERM T_0, R_3, T_1, R_3 // [_, r₂₆[3], _, 0]
VPERM T_0, R_4, T_1, R_4 // [_, r₂₆[4], _, 0]
VPERM T_0, R5_1, T_1, R5_1 // [_, 5r₂₆[1], _, 0]
VPERM T_0, R5_2, T_1, R5_2 // [_, 5r₂₆[2], _, 0]
VPERM T_0, R5_3, T_1, R5_3 // [_, 5r₂₆[3], _, 0]
VPERM T_0, R5_4, T_1, R5_4 // [_, 5r₂₆[4], _, 0]
// Set the value of lane 1 to be 1.
VLEIF $3, $1, R_0 // [_, r₂₆[0], _, 1]
MOVD $0, R3
BR multiply

182
gateway/vendor/golang.org/x/crypto/nacl/box/box.go generated vendored Normal file
View File

@ -0,0 +1,182 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
/*
Package box authenticates and encrypts small messages using public-key cryptography.
Box uses Curve25519, XSalsa20 and Poly1305 to encrypt and authenticate
messages. The length of messages is not hidden.
It is the caller's responsibility to ensure the uniqueness of nonces—for
example, by using nonce 1 for the first message, nonce 2 for the second
message, etc. Nonces are long enough that randomly generated nonces have
negligible risk of collision.
Messages should be small because:
1. The whole message needs to be held in memory to be processed.
2. Using large messages pressures implementations on small machines to decrypt
and process plaintext before authenticating it. This is very dangerous, and
this API does not allow it, but a protocol that uses excessive message sizes
might present some implementations with no other choice.
3. Fixed overheads will be sufficiently amortised by messages as small as 8KB.
4. Performance may be improved by working with messages that fit into data caches.
Thus large amounts of data should be chunked so that each message is small.
(Each message still needs a unique nonce.) If in doubt, 16KB is a reasonable
chunk size.
This package is interoperable with NaCl: https://nacl.cr.yp.to/box.html.
Anonymous sealing/opening is an extension of NaCl defined by and interoperable
with libsodium:
https://libsodium.gitbook.io/doc/public-key_cryptography/sealed_boxes.
*/
package box // import "golang.org/x/crypto/nacl/box"
import (
cryptorand "crypto/rand"
"io"
"golang.org/x/crypto/blake2b"
"golang.org/x/crypto/curve25519"
"golang.org/x/crypto/nacl/secretbox"
"golang.org/x/crypto/salsa20/salsa"
)
const (
// Overhead is the number of bytes of overhead when boxing a message.
Overhead = secretbox.Overhead
// AnonymousOverhead is the number of bytes of overhead when using anonymous
// sealed boxes.
AnonymousOverhead = Overhead + 32
)
// GenerateKey generates a new public/private key pair suitable for use with
// Seal and Open.
func GenerateKey(rand io.Reader) (publicKey, privateKey *[32]byte, err error) {
publicKey = new([32]byte)
privateKey = new([32]byte)
_, err = io.ReadFull(rand, privateKey[:])
if err != nil {
publicKey = nil
privateKey = nil
return
}
curve25519.ScalarBaseMult(publicKey, privateKey)
return
}
var zeros [16]byte
// Precompute calculates the shared key between peersPublicKey and privateKey
// and writes it to sharedKey. The shared key can be used with
// OpenAfterPrecomputation and SealAfterPrecomputation to speed up processing
// when using the same pair of keys repeatedly.
func Precompute(sharedKey, peersPublicKey, privateKey *[32]byte) {
curve25519.ScalarMult(sharedKey, privateKey, peersPublicKey)
salsa.HSalsa20(sharedKey, &zeros, sharedKey, &salsa.Sigma)
}
// Seal appends an encrypted and authenticated copy of message to out, which
// will be Overhead bytes longer than the original and must not overlap it. The
// nonce must be unique for each distinct message for a given pair of keys.
func Seal(out, message []byte, nonce *[24]byte, peersPublicKey, privateKey *[32]byte) []byte {
var sharedKey [32]byte
Precompute(&sharedKey, peersPublicKey, privateKey)
return secretbox.Seal(out, message, nonce, &sharedKey)
}
// SealAfterPrecomputation performs the same actions as Seal, but takes a
// shared key as generated by Precompute.
func SealAfterPrecomputation(out, message []byte, nonce *[24]byte, sharedKey *[32]byte) []byte {
return secretbox.Seal(out, message, nonce, sharedKey)
}
// Open authenticates and decrypts a box produced by Seal and appends the
// message to out, which must not overlap box. The output will be Overhead
// bytes smaller than box.
func Open(out, box []byte, nonce *[24]byte, peersPublicKey, privateKey *[32]byte) ([]byte, bool) {
var sharedKey [32]byte
Precompute(&sharedKey, peersPublicKey, privateKey)
return secretbox.Open(out, box, nonce, &sharedKey)
}
// OpenAfterPrecomputation performs the same actions as Open, but takes a
// shared key as generated by Precompute.
func OpenAfterPrecomputation(out, box []byte, nonce *[24]byte, sharedKey *[32]byte) ([]byte, bool) {
return secretbox.Open(out, box, nonce, sharedKey)
}
// SealAnonymous appends an encrypted and authenticated copy of message to out,
// which will be AnonymousOverhead bytes longer than the original and must not
// overlap it. This differs from Seal in that the sender is not required to
// provide a private key.
func SealAnonymous(out, message []byte, recipient *[32]byte, rand io.Reader) ([]byte, error) {
if rand == nil {
rand = cryptorand.Reader
}
ephemeralPub, ephemeralPriv, err := GenerateKey(rand)
if err != nil {
return nil, err
}
var nonce [24]byte
if err := sealNonce(ephemeralPub, recipient, &nonce); err != nil {
return nil, err
}
if total := len(out) + AnonymousOverhead + len(message); cap(out) < total {
original := out
out = make([]byte, 0, total)
out = append(out, original...)
}
out = append(out, ephemeralPub[:]...)
return Seal(out, message, &nonce, recipient, ephemeralPriv), nil
}
// OpenAnonymous authenticates and decrypts a box produced by SealAnonymous and
// appends the message to out, which must not overlap box. The output will be
// AnonymousOverhead bytes smaller than box.
func OpenAnonymous(out, box []byte, publicKey, privateKey *[32]byte) (message []byte, ok bool) {
if len(box) < AnonymousOverhead {
return nil, false
}
var ephemeralPub [32]byte
copy(ephemeralPub[:], box[:32])
var nonce [24]byte
if err := sealNonce(&ephemeralPub, publicKey, &nonce); err != nil {
return nil, false
}
return Open(out, box[32:], &nonce, &ephemeralPub, privateKey)
}
// sealNonce generates a 24 byte nonce that is a blake2b digest of the
// ephemeral public key and the receiver's public key.
func sealNonce(ephemeralPub, peersPublicKey *[32]byte, nonce *[24]byte) error {
h, err := blake2b.New(24, nil)
if err != nil {
return err
}
if _, err = h.Write(ephemeralPub[:]); err != nil {
return err
}
if _, err = h.Write(peersPublicKey[:]); err != nil {
return err
}
h.Sum(nonce[:0])
return nil
}

173
gateway/vendor/golang.org/x/crypto/nacl/secretbox/secretbox.go generated vendored Normal file
View File

@ -0,0 +1,173 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
/*
Package secretbox encrypts and authenticates small messages.
Secretbox uses XSalsa20 and Poly1305 to encrypt and authenticate messages with
secret-key cryptography. The length of messages is not hidden.
It is the caller's responsibility to ensure the uniqueness of nonces—for
example, by using nonce 1 for the first message, nonce 2 for the second
message, etc. Nonces are long enough that randomly generated nonces have
negligible risk of collision.
Messages should be small because:
1. The whole message needs to be held in memory to be processed.
2. Using large messages pressures implementations on small machines to decrypt
and process plaintext before authenticating it. This is very dangerous, and
this API does not allow it, but a protocol that uses excessive message sizes
might present some implementations with no other choice.
3. Fixed overheads will be sufficiently amortised by messages as small as 8KB.
4. Performance may be improved by working with messages that fit into data caches.
Thus large amounts of data should be chunked so that each message is small.
(Each message still needs a unique nonce.) If in doubt, 16KB is a reasonable
chunk size.
This package is interoperable with NaCl: https://nacl.cr.yp.to/secretbox.html.
*/
package secretbox // import "golang.org/x/crypto/nacl/secretbox"
import (
"golang.org/x/crypto/internal/alias"
"golang.org/x/crypto/internal/poly1305"
"golang.org/x/crypto/salsa20/salsa"
)
// Overhead is the number of bytes of overhead when boxing a message.
const Overhead = poly1305.TagSize
// setup produces a sub-key and Salsa20 counter given a nonce and key.
func setup(subKey *[32]byte, counter *[16]byte, nonce *[24]byte, key *[32]byte) {
// We use XSalsa20 for encryption so first we need to generate a
// key and nonce with HSalsa20.
var hNonce [16]byte
copy(hNonce[:], nonce[:])
salsa.HSalsa20(subKey, &hNonce, key, &salsa.Sigma)
// The final 8 bytes of the original nonce form the new nonce.
copy(counter[:], nonce[16:])
}
// sliceForAppend takes a slice and a requested number of bytes. It returns a
// slice with the contents of the given slice followed by that many bytes and a
// second slice that aliases into it and contains only the extra bytes. If the
// original slice has sufficient capacity then no allocation is performed.
func sliceForAppend(in []byte, n int) (head, tail []byte) {
if total := len(in) + n; cap(in) >= total {
head = in[:total]
} else {
head = make([]byte, total)
copy(head, in)
}
tail = head[len(in):]
return
}
// Seal appends an encrypted and authenticated copy of message to out, which
// must not overlap message. The key and nonce pair must be unique for each
// distinct message and the output will be Overhead bytes longer than message.
func Seal(out, message []byte, nonce *[24]byte, key *[32]byte) []byte {
var subKey [32]byte
var counter [16]byte
setup(&subKey, &counter, nonce, key)
// The Poly1305 key is generated by encrypting 32 bytes of zeros. Since
// Salsa20 works with 64-byte blocks, we also generate 32 bytes of
// keystream as a side effect.
var firstBlock [64]byte
salsa.XORKeyStream(firstBlock[:], firstBlock[:], &counter, &subKey)
var poly1305Key [32]byte
copy(poly1305Key[:], firstBlock[:])
ret, out := sliceForAppend(out, len(message)+poly1305.TagSize)
if alias.AnyOverlap(out, message) {
panic("nacl: invalid buffer overlap")
}
// We XOR up to 32 bytes of message with the keystream generated from
// the first block.
firstMessageBlock := message
if len(firstMessageBlock) > 32 {
firstMessageBlock = firstMessageBlock[:32]
}
tagOut := out
out = out[poly1305.TagSize:]
for i, x := range firstMessageBlock {
out[i] = firstBlock[32+i] ^ x
}
message = message[len(firstMessageBlock):]
ciphertext := out
out = out[len(firstMessageBlock):]
// Now encrypt the rest.
counter[8] = 1
salsa.XORKeyStream(out, message, &counter, &subKey)
var tag [poly1305.TagSize]byte
poly1305.Sum(&tag, ciphertext, &poly1305Key)
copy(tagOut, tag[:])
return ret
}
// Open authenticates and decrypts a box produced by Seal and appends the
// message to out, which must not overlap box. The output will be Overhead
// bytes smaller than box.
func Open(out, box []byte, nonce *[24]byte, key *[32]byte) ([]byte, bool) {
if len(box) < Overhead {
return nil, false
}
var subKey [32]byte
var counter [16]byte
setup(&subKey, &counter, nonce, key)
// The Poly1305 key is generated by encrypting 32 bytes of zeros. Since
// Salsa20 works with 64-byte blocks, we also generate 32 bytes of
// keystream as a side effect.
var firstBlock [64]byte
salsa.XORKeyStream(firstBlock[:], firstBlock[:], &counter, &subKey)
var poly1305Key [32]byte
copy(poly1305Key[:], firstBlock[:])
var tag [poly1305.TagSize]byte
copy(tag[:], box)
if !poly1305.Verify(&tag, box[poly1305.TagSize:], &poly1305Key) {
return nil, false
}
ret, out := sliceForAppend(out, len(box)-Overhead)
if alias.AnyOverlap(out, box) {
panic("nacl: invalid buffer overlap")
}
// We XOR up to 32 bytes of box with the keystream generated from
// the first block.
box = box[Overhead:]
firstMessageBlock := box
if len(firstMessageBlock) > 32 {
firstMessageBlock = firstMessageBlock[:32]
}
for i, x := range firstMessageBlock {
out[i] = firstBlock[32+i] ^ x
}
box = box[len(firstMessageBlock):]
out = out[len(firstMessageBlock):]
// Now decrypt the rest.
counter[8] = 1
salsa.XORKeyStream(out, box, &counter, &subKey)
return ret, true
}

146
gateway/vendor/golang.org/x/crypto/salsa20/salsa/hsalsa20.go generated vendored Normal file
View File

@ -0,0 +1,146 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package salsa provides low-level access to functions in the Salsa family.
package salsa // import "golang.org/x/crypto/salsa20/salsa"
import "math/bits"
// Sigma is the Salsa20 constant for 256-bit keys.
var Sigma = [16]byte{'e', 'x', 'p', 'a', 'n', 'd', ' ', '3', '2', '-', 'b', 'y', 't', 'e', ' ', 'k'}
// HSalsa20 applies the HSalsa20 core function to a 16-byte input in, 32-byte
// key k, and 16-byte constant c, and puts the result into the 32-byte array
// out.
func HSalsa20(out *[32]byte, in *[16]byte, k *[32]byte, c *[16]byte) {
x0 := uint32(c[0]) | uint32(c[1])<<8 | uint32(c[2])<<16 | uint32(c[3])<<24
x1 := uint32(k[0]) | uint32(k[1])<<8 | uint32(k[2])<<16 | uint32(k[3])<<24
x2 := uint32(k[4]) | uint32(k[5])<<8 | uint32(k[6])<<16 | uint32(k[7])<<24
x3 := uint32(k[8]) | uint32(k[9])<<8 | uint32(k[10])<<16 | uint32(k[11])<<24
x4 := uint32(k[12]) | uint32(k[13])<<8 | uint32(k[14])<<16 | uint32(k[15])<<24
x5 := uint32(c[4]) | uint32(c[5])<<8 | uint32(c[6])<<16 | uint32(c[7])<<24
x6 := uint32(in[0]) | uint32(in[1])<<8 | uint32(in[2])<<16 | uint32(in[3])<<24
x7 := uint32(in[4]) | uint32(in[5])<<8 | uint32(in[6])<<16 | uint32(in[7])<<24
x8 := uint32(in[8]) | uint32(in[9])<<8 | uint32(in[10])<<16 | uint32(in[11])<<24
x9 := uint32(in[12]) | uint32(in[13])<<8 | uint32(in[14])<<16 | uint32(in[15])<<24
x10 := uint32(c[8]) | uint32(c[9])<<8 | uint32(c[10])<<16 | uint32(c[11])<<24
x11 := uint32(k[16]) | uint32(k[17])<<8 | uint32(k[18])<<16 | uint32(k[19])<<24
x12 := uint32(k[20]) | uint32(k[21])<<8 | uint32(k[22])<<16 | uint32(k[23])<<24
x13 := uint32(k[24]) | uint32(k[25])<<8 | uint32(k[26])<<16 | uint32(k[27])<<24
x14 := uint32(k[28]) | uint32(k[29])<<8 | uint32(k[30])<<16 | uint32(k[31])<<24
x15 := uint32(c[12]) | uint32(c[13])<<8 | uint32(c[14])<<16 | uint32(c[15])<<24
for i := 0; i < 20; i += 2 {
u := x0 + x12
x4 ^= bits.RotateLeft32(u, 7)
u = x4 + x0
x8 ^= bits.RotateLeft32(u, 9)
u = x8 + x4
x12 ^= bits.RotateLeft32(u, 13)
u = x12 + x8
x0 ^= bits.RotateLeft32(u, 18)
u = x5 + x1
x9 ^= bits.RotateLeft32(u, 7)
u = x9 + x5
x13 ^= bits.RotateLeft32(u, 9)
u = x13 + x9
x1 ^= bits.RotateLeft32(u, 13)
u = x1 + x13
x5 ^= bits.RotateLeft32(u, 18)
u = x10 + x6
x14 ^= bits.RotateLeft32(u, 7)
u = x14 + x10
x2 ^= bits.RotateLeft32(u, 9)
u = x2 + x14
x6 ^= bits.RotateLeft32(u, 13)
u = x6 + x2
x10 ^= bits.RotateLeft32(u, 18)
u = x15 + x11
x3 ^= bits.RotateLeft32(u, 7)
u = x3 + x15
x7 ^= bits.RotateLeft32(u, 9)
u = x7 + x3
x11 ^= bits.RotateLeft32(u, 13)
u = x11 + x7
x15 ^= bits.RotateLeft32(u, 18)
u = x0 + x3
x1 ^= bits.RotateLeft32(u, 7)
u = x1 + x0
x2 ^= bits.RotateLeft32(u, 9)
u = x2 + x1
x3 ^= bits.RotateLeft32(u, 13)
u = x3 + x2
x0 ^= bits.RotateLeft32(u, 18)
u = x5 + x4
x6 ^= bits.RotateLeft32(u, 7)
u = x6 + x5
x7 ^= bits.RotateLeft32(u, 9)
u = x7 + x6
x4 ^= bits.RotateLeft32(u, 13)
u = x4 + x7
x5 ^= bits.RotateLeft32(u, 18)
u = x10 + x9
x11 ^= bits.RotateLeft32(u, 7)
u = x11 + x10
x8 ^= bits.RotateLeft32(u, 9)
u = x8 + x11
x9 ^= bits.RotateLeft32(u, 13)
u = x9 + x8
x10 ^= bits.RotateLeft32(u, 18)
u = x15 + x14
x12 ^= bits.RotateLeft32(u, 7)
u = x12 + x15
x13 ^= bits.RotateLeft32(u, 9)
u = x13 + x12
x14 ^= bits.RotateLeft32(u, 13)
u = x14 + x13
x15 ^= bits.RotateLeft32(u, 18)
}
out[0] = byte(x0)
out[1] = byte(x0 >> 8)
out[2] = byte(x0 >> 16)
out[3] = byte(x0 >> 24)
out[4] = byte(x5)
out[5] = byte(x5 >> 8)
out[6] = byte(x5 >> 16)
out[7] = byte(x5 >> 24)
out[8] = byte(x10)
out[9] = byte(x10 >> 8)
out[10] = byte(x10 >> 16)
out[11] = byte(x10 >> 24)
out[12] = byte(x15)
out[13] = byte(x15 >> 8)
out[14] = byte(x15 >> 16)
out[15] = byte(x15 >> 24)
out[16] = byte(x6)
out[17] = byte(x6 >> 8)
out[18] = byte(x6 >> 16)
out[19] = byte(x6 >> 24)
out[20] = byte(x7)
out[21] = byte(x7 >> 8)
out[22] = byte(x7 >> 16)
out[23] = byte(x7 >> 24)
out[24] = byte(x8)
out[25] = byte(x8 >> 8)
out[26] = byte(x8 >> 16)
out[27] = byte(x8 >> 24)
out[28] = byte(x9)
out[29] = byte(x9 >> 8)
out[30] = byte(x9 >> 16)
out[31] = byte(x9 >> 24)
}

201
gateway/vendor/golang.org/x/crypto/salsa20/salsa/salsa208.go generated vendored Normal file
View File

@ -0,0 +1,201 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package salsa
import "math/bits"
// Core208 applies the Salsa20/8 core function to the 64-byte array in and puts
// the result into the 64-byte array out. The input and output may be the same array.
func Core208(out *[64]byte, in *[64]byte) {
j0 := uint32(in[0]) | uint32(in[1])<<8 | uint32(in[2])<<16 | uint32(in[3])<<24
j1 := uint32(in[4]) | uint32(in[5])<<8 | uint32(in[6])<<16 | uint32(in[7])<<24
j2 := uint32(in[8]) | uint32(in[9])<<8 | uint32(in[10])<<16 | uint32(in[11])<<24
j3 := uint32(in[12]) | uint32(in[13])<<8 | uint32(in[14])<<16 | uint32(in[15])<<24
j4 := uint32(in[16]) | uint32(in[17])<<8 | uint32(in[18])<<16 | uint32(in[19])<<24
j5 := uint32(in[20]) | uint32(in[21])<<8 | uint32(in[22])<<16 | uint32(in[23])<<24
j6 := uint32(in[24]) | uint32(in[25])<<8 | uint32(in[26])<<16 | uint32(in[27])<<24
j7 := uint32(in[28]) | uint32(in[29])<<8 | uint32(in[30])<<16 | uint32(in[31])<<24
j8 := uint32(in[32]) | uint32(in[33])<<8 | uint32(in[34])<<16 | uint32(in[35])<<24
j9 := uint32(in[36]) | uint32(in[37])<<8 | uint32(in[38])<<16 | uint32(in[39])<<24
j10 := uint32(in[40]) | uint32(in[41])<<8 | uint32(in[42])<<16 | uint32(in[43])<<24
j11 := uint32(in[44]) | uint32(in[45])<<8 | uint32(in[46])<<16 | uint32(in[47])<<24
j12 := uint32(in[48]) | uint32(in[49])<<8 | uint32(in[50])<<16 | uint32(in[51])<<24
j13 := uint32(in[52]) | uint32(in[53])<<8 | uint32(in[54])<<16 | uint32(in[55])<<24
j14 := uint32(in[56]) | uint32(in[57])<<8 | uint32(in[58])<<16 | uint32(in[59])<<24
j15 := uint32(in[60]) | uint32(in[61])<<8 | uint32(in[62])<<16 | uint32(in[63])<<24
x0, x1, x2, x3, x4, x5, x6, x7, x8 := j0, j1, j2, j3, j4, j5, j6, j7, j8
x9, x10, x11, x12, x13, x14, x15 := j9, j10, j11, j12, j13, j14, j15
for i := 0; i < 8; i += 2 {
u := x0 + x12
x4 ^= bits.RotateLeft32(u, 7)
u = x4 + x0
x8 ^= bits.RotateLeft32(u, 9)
u = x8 + x4
x12 ^= bits.RotateLeft32(u, 13)
u = x12 + x8
x0 ^= bits.RotateLeft32(u, 18)
u = x5 + x1
x9 ^= bits.RotateLeft32(u, 7)
u = x9 + x5
x13 ^= bits.RotateLeft32(u, 9)
u = x13 + x9
x1 ^= bits.RotateLeft32(u, 13)
u = x1 + x13
x5 ^= bits.RotateLeft32(u, 18)
u = x10 + x6
x14 ^= bits.RotateLeft32(u, 7)
u = x14 + x10
x2 ^= bits.RotateLeft32(u, 9)
u = x2 + x14
x6 ^= bits.RotateLeft32(u, 13)
u = x6 + x2
x10 ^= bits.RotateLeft32(u, 18)
u = x15 + x11
x3 ^= bits.RotateLeft32(u, 7)
u = x3 + x15
x7 ^= bits.RotateLeft32(u, 9)
u = x7 + x3
x11 ^= bits.RotateLeft32(u, 13)
u = x11 + x7
x15 ^= bits.RotateLeft32(u, 18)
u = x0 + x3
x1 ^= bits.RotateLeft32(u, 7)
u = x1 + x0
x2 ^= bits.RotateLeft32(u, 9)
u = x2 + x1
x3 ^= bits.RotateLeft32(u, 13)
u = x3 + x2
x0 ^= bits.RotateLeft32(u, 18)
u = x5 + x4
x6 ^= bits.RotateLeft32(u, 7)
u = x6 + x5
x7 ^= bits.RotateLeft32(u, 9)
u = x7 + x6
x4 ^= bits.RotateLeft32(u, 13)
u = x4 + x7
x5 ^= bits.RotateLeft32(u, 18)
u = x10 + x9
x11 ^= bits.RotateLeft32(u, 7)
u = x11 + x10
x8 ^= bits.RotateLeft32(u, 9)
u = x8 + x11
x9 ^= bits.RotateLeft32(u, 13)
u = x9 + x8
x10 ^= bits.RotateLeft32(u, 18)
u = x15 + x14
x12 ^= bits.RotateLeft32(u, 7)
u = x12 + x15
x13 ^= bits.RotateLeft32(u, 9)
u = x13 + x12
x14 ^= bits.RotateLeft32(u, 13)
u = x14 + x13
x15 ^= bits.RotateLeft32(u, 18)
}
x0 += j0
x1 += j1
x2 += j2
x3 += j3
x4 += j4
x5 += j5
x6 += j6
x7 += j7
x8 += j8
x9 += j9
x10 += j10
x11 += j11
x12 += j12
x13 += j13
x14 += j14
x15 += j15
out[0] = byte(x0)
out[1] = byte(x0 >> 8)
out[2] = byte(x0 >> 16)
out[3] = byte(x0 >> 24)
out[4] = byte(x1)
out[5] = byte(x1 >> 8)
out[6] = byte(x1 >> 16)
out[7] = byte(x1 >> 24)
out[8] = byte(x2)
out[9] = byte(x2 >> 8)
out[10] = byte(x2 >> 16)
out[11] = byte(x2 >> 24)
out[12] = byte(x3)
out[13] = byte(x3 >> 8)
out[14] = byte(x3 >> 16)
out[15] = byte(x3 >> 24)
out[16] = byte(x4)
out[17] = byte(x4 >> 8)
out[18] = byte(x4 >> 16)
out[19] = byte(x4 >> 24)
out[20] = byte(x5)
out[21] = byte(x5 >> 8)
out[22] = byte(x5 >> 16)
out[23] = byte(x5 >> 24)
out[24] = byte(x6)
out[25] = byte(x6 >> 8)
out[26] = byte(x6 >> 16)
out[27] = byte(x6 >> 24)
out[28] = byte(x7)
out[29] = byte(x7 >> 8)
out[30] = byte(x7 >> 16)
out[31] = byte(x7 >> 24)
out[32] = byte(x8)
out[33] = byte(x8 >> 8)
out[34] = byte(x8 >> 16)
out[35] = byte(x8 >> 24)
out[36] = byte(x9)
out[37] = byte(x9 >> 8)
out[38] = byte(x9 >> 16)
out[39] = byte(x9 >> 24)
out[40] = byte(x10)
out[41] = byte(x10 >> 8)
out[42] = byte(x10 >> 16)
out[43] = byte(x10 >> 24)
out[44] = byte(x11)
out[45] = byte(x11 >> 8)
out[46] = byte(x11 >> 16)
out[47] = byte(x11 >> 24)
out[48] = byte(x12)
out[49] = byte(x12 >> 8)
out[50] = byte(x12 >> 16)
out[51] = byte(x12 >> 24)
out[52] = byte(x13)
out[53] = byte(x13 >> 8)
out[54] = byte(x13 >> 16)
out[55] = byte(x13 >> 24)
out[56] = byte(x14)
out[57] = byte(x14 >> 8)
out[58] = byte(x14 >> 16)
out[59] = byte(x14 >> 24)
out[60] = byte(x15)
out[61] = byte(x15 >> 8)
out[62] = byte(x15 >> 16)
out[63] = byte(x15 >> 24)
}

24
gateway/vendor/golang.org/x/crypto/salsa20/salsa/salsa20_amd64.go generated vendored Normal file
View File

@ -0,0 +1,24 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build amd64 && !purego && gc
// +build amd64,!purego,gc
package salsa
//go:noescape
// salsa2020XORKeyStream is implemented in salsa20_amd64.s.
func salsa2020XORKeyStream(out, in *byte, n uint64, nonce, key *byte)
// XORKeyStream crypts bytes from in to out using the given key and counters.
// In and out must overlap entirely or not at all. Counter
// contains the raw salsa20 counter bytes (both nonce and block counter).
func XORKeyStream(out, in []byte, counter *[16]byte, key *[32]byte) {
if len(in) == 0 {
return
}
_ = out[len(in)-1]
salsa2020XORKeyStream(&out[0], &in[0], uint64(len(in)), &counter[0], &key[0])
}

881
gateway/vendor/golang.org/x/crypto/salsa20/salsa/salsa20_amd64.s generated vendored Normal file
View File

@ -0,0 +1,881 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build amd64 && !purego && gc
// +build amd64,!purego,gc
// This code was translated into a form compatible with 6a from the public
// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html
// func salsa2020XORKeyStream(out, in *byte, n uint64, nonce, key *byte)
// This needs up to 64 bytes at 360(R12); hence the non-obvious frame size.
TEXT ·salsa2020XORKeyStream(SB),0,$456-40 // frame = 424 + 32 byte alignment
MOVQ out+0(FP),DI
MOVQ in+8(FP),SI
MOVQ n+16(FP),DX
MOVQ nonce+24(FP),CX
MOVQ key+32(FP),R8
MOVQ SP,R12
ADDQ $31, R12
ANDQ $~31, R12
MOVQ DX,R9
MOVQ CX,DX
MOVQ R8,R10
CMPQ R9,$0
JBE DONE
START:
MOVL 20(R10),CX
MOVL 0(R10),R8
MOVL 0(DX),AX
MOVL 16(R10),R11
MOVL CX,0(R12)
MOVL R8, 4 (R12)
MOVL AX, 8 (R12)
MOVL R11, 12 (R12)
MOVL 8(DX),CX
MOVL 24(R10),R8
MOVL 4(R10),AX
MOVL 4(DX),R11
MOVL CX,16(R12)
MOVL R8, 20 (R12)
MOVL AX, 24 (R12)
MOVL R11, 28 (R12)
MOVL 12(DX),CX
MOVL 12(R10),DX
MOVL 28(R10),R8
MOVL 8(R10),AX
MOVL DX,32(R12)
MOVL CX, 36 (R12)
MOVL R8, 40 (R12)
MOVL AX, 44 (R12)
MOVQ $1634760805,DX
MOVQ $857760878,CX
MOVQ $2036477234,R8
MOVQ $1797285236,AX
MOVL DX,48(R12)
MOVL CX, 52 (R12)
MOVL R8, 56 (R12)
MOVL AX, 60 (R12)
CMPQ R9,$256
JB BYTESBETWEEN1AND255
MOVOA 48(R12),X0
PSHUFL $0X55,X0,X1
PSHUFL $0XAA,X0,X2
PSHUFL $0XFF,X0,X3
PSHUFL $0X00,X0,X0
MOVOA X1,64(R12)
MOVOA X2,80(R12)
MOVOA X3,96(R12)
MOVOA X0,112(R12)
MOVOA 0(R12),X0
PSHUFL $0XAA,X0,X1
PSHUFL $0XFF,X0,X2
PSHUFL $0X00,X0,X3
PSHUFL $0X55,X0,X0
MOVOA X1,128(R12)
MOVOA X2,144(R12)
MOVOA X3,160(R12)
MOVOA X0,176(R12)
MOVOA 16(R12),X0
PSHUFL $0XFF,X0,X1
PSHUFL $0X55,X0,X2
PSHUFL $0XAA,X0,X0
MOVOA X1,192(R12)
MOVOA X2,208(R12)
MOVOA X0,224(R12)
MOVOA 32(R12),X0
PSHUFL $0X00,X0,X1
PSHUFL $0XAA,X0,X2
PSHUFL $0XFF,X0,X0
MOVOA X1,240(R12)
MOVOA X2,256(R12)
MOVOA X0,272(R12)
BYTESATLEAST256:
MOVL 16(R12),DX
MOVL 36 (R12),CX
MOVL DX,288(R12)
MOVL CX,304(R12)
SHLQ $32,CX
ADDQ CX,DX
ADDQ $1,DX
MOVQ DX,CX
SHRQ $32,CX
MOVL DX, 292 (R12)
MOVL CX, 308 (R12)
ADDQ $1,DX
MOVQ DX,CX
SHRQ $32,CX
MOVL DX, 296 (R12)
MOVL CX, 312 (R12)
ADDQ $1,DX
MOVQ DX,CX
SHRQ $32,CX
MOVL DX, 300 (R12)
MOVL CX, 316 (R12)
ADDQ $1,DX
MOVQ DX,CX
SHRQ $32,CX
MOVL DX,16(R12)
MOVL CX, 36 (R12)
MOVQ R9,352(R12)
MOVQ $20,DX
MOVOA 64(R12),X0
MOVOA 80(R12),X1
MOVOA 96(R12),X2
MOVOA 256(R12),X3
MOVOA 272(R12),X4
MOVOA 128(R12),X5
MOVOA 144(R12),X6
MOVOA 176(R12),X7
MOVOA 192(R12),X8
MOVOA 208(R12),X9
MOVOA 224(R12),X10
MOVOA 304(R12),X11
MOVOA 112(R12),X12
MOVOA 160(R12),X13
MOVOA 240(R12),X14
MOVOA 288(R12),X15
MAINLOOP1:
MOVOA X1,320(R12)
MOVOA X2,336(R12)
MOVOA X13,X1
PADDL X12,X1
MOVOA X1,X2
PSLLL $7,X1
PXOR X1,X14
PSRLL $25,X2
PXOR X2,X14
MOVOA X7,X1
PADDL X0,X1
MOVOA X1,X2
PSLLL $7,X1
PXOR X1,X11
PSRLL $25,X2
PXOR X2,X11
MOVOA X12,X1
PADDL X14,X1
MOVOA X1,X2
PSLLL $9,X1
PXOR X1,X15
PSRLL $23,X2
PXOR X2,X15
MOVOA X0,X1
PADDL X11,X1
MOVOA X1,X2
PSLLL $9,X1
PXOR X1,X9
PSRLL $23,X2
PXOR X2,X9
MOVOA X14,X1
PADDL X15,X1
MOVOA X1,X2
PSLLL $13,X1
PXOR X1,X13
PSRLL $19,X2
PXOR X2,X13
MOVOA X11,X1
PADDL X9,X1
MOVOA X1,X2
PSLLL $13,X1
PXOR X1,X7
PSRLL $19,X2
PXOR X2,X7
MOVOA X15,X1
PADDL X13,X1
MOVOA X1,X2
PSLLL $18,X1
PXOR X1,X12
PSRLL $14,X2
PXOR X2,X12
MOVOA 320(R12),X1
MOVOA X12,320(R12)
MOVOA X9,X2
PADDL X7,X2
MOVOA X2,X12
PSLLL $18,X2
PXOR X2,X0
PSRLL $14,X12
PXOR X12,X0
MOVOA X5,X2
PADDL X1,X2
MOVOA X2,X12
PSLLL $7,X2
PXOR X2,X3
PSRLL $25,X12
PXOR X12,X3
MOVOA 336(R12),X2
MOVOA X0,336(R12)
MOVOA X6,X0
PADDL X2,X0
MOVOA X0,X12
PSLLL $7,X0
PXOR X0,X4
PSRLL $25,X12
PXOR X12,X4
MOVOA X1,X0
PADDL X3,X0
MOVOA X0,X12
PSLLL $9,X0
PXOR X0,X10
PSRLL $23,X12
PXOR X12,X10
MOVOA X2,X0
PADDL X4,X0
MOVOA X0,X12
PSLLL $9,X0
PXOR X0,X8
PSRLL $23,X12
PXOR X12,X8
MOVOA X3,X0
PADDL X10,X0
MOVOA X0,X12
PSLLL $13,X0
PXOR X0,X5
PSRLL $19,X12
PXOR X12,X5
MOVOA X4,X0
PADDL X8,X0
MOVOA X0,X12
PSLLL $13,X0
PXOR X0,X6
PSRLL $19,X12
PXOR X12,X6
MOVOA X10,X0
PADDL X5,X0
MOVOA X0,X12
PSLLL $18,X0
PXOR X0,X1
PSRLL $14,X12
PXOR X12,X1
MOVOA 320(R12),X0
MOVOA X1,320(R12)
MOVOA X4,X1
PADDL X0,X1
MOVOA X1,X12
PSLLL $7,X1
PXOR X1,X7
PSRLL $25,X12
PXOR X12,X7
MOVOA X8,X1
PADDL X6,X1
MOVOA X1,X12
PSLLL $18,X1
PXOR X1,X2
PSRLL $14,X12
PXOR X12,X2
MOVOA 336(R12),X12
MOVOA X2,336(R12)
MOVOA X14,X1
PADDL X12,X1
MOVOA X1,X2
PSLLL $7,X1
PXOR X1,X5
PSRLL $25,X2
PXOR X2,X5
MOVOA X0,X1
PADDL X7,X1
MOVOA X1,X2
PSLLL $9,X1
PXOR X1,X10
PSRLL $23,X2
PXOR X2,X10
MOVOA X12,X1
PADDL X5,X1
MOVOA X1,X2
PSLLL $9,X1
PXOR X1,X8
PSRLL $23,X2
PXOR X2,X8
MOVOA X7,X1
PADDL X10,X1
MOVOA X1,X2
PSLLL $13,X1
PXOR X1,X4
PSRLL $19,X2
PXOR X2,X4
MOVOA X5,X1
PADDL X8,X1
MOVOA X1,X2
PSLLL $13,X1
PXOR X1,X14
PSRLL $19,X2
PXOR X2,X14
MOVOA X10,X1
PADDL X4,X1
MOVOA X1,X2
PSLLL $18,X1
PXOR X1,X0
PSRLL $14,X2
PXOR X2,X0
MOVOA 320(R12),X1
MOVOA X0,320(R12)
MOVOA X8,X0
PADDL X14,X0
MOVOA X0,X2
PSLLL $18,X0
PXOR X0,X12
PSRLL $14,X2
PXOR X2,X12
MOVOA X11,X0
PADDL X1,X0
MOVOA X0,X2
PSLLL $7,X0
PXOR X0,X6
PSRLL $25,X2
PXOR X2,X6
MOVOA 336(R12),X2
MOVOA X12,336(R12)
MOVOA X3,X0
PADDL X2,X0
MOVOA X0,X12
PSLLL $7,X0
PXOR X0,X13
PSRLL $25,X12
PXOR X12,X13
MOVOA X1,X0
PADDL X6,X0
MOVOA X0,X12
PSLLL $9,X0
PXOR X0,X15
PSRLL $23,X12
PXOR X12,X15
MOVOA X2,X0
PADDL X13,X0
MOVOA X0,X12
PSLLL $9,X0
PXOR X0,X9
PSRLL $23,X12
PXOR X12,X9
MOVOA X6,X0
PADDL X15,X0
MOVOA X0,X12
PSLLL $13,X0
PXOR X0,X11
PSRLL $19,X12
PXOR X12,X11
MOVOA X13,X0
PADDL X9,X0
MOVOA X0,X12
PSLLL $13,X0
PXOR X0,X3
PSRLL $19,X12
PXOR X12,X3
MOVOA X15,X0
PADDL X11,X0
MOVOA X0,X12
PSLLL $18,X0
PXOR X0,X1
PSRLL $14,X12
PXOR X12,X1
MOVOA X9,X0
PADDL X3,X0
MOVOA X0,X12
PSLLL $18,X0
PXOR X0,X2
PSRLL $14,X12
PXOR X12,X2
MOVOA 320(R12),X12
MOVOA 336(R12),X0
SUBQ $2,DX
JA MAINLOOP1
PADDL 112(R12),X12
PADDL 176(R12),X7
PADDL 224(R12),X10
PADDL 272(R12),X4
MOVD X12,DX
MOVD X7,CX
MOVD X10,R8
MOVD X4,R9
PSHUFL $0X39,X12,X12
PSHUFL $0X39,X7,X7
PSHUFL $0X39,X10,X10
PSHUFL $0X39,X4,X4
XORL 0(SI),DX
XORL 4(SI),CX
XORL 8(SI),R8
XORL 12(SI),R9
MOVL DX,0(DI)
MOVL CX,4(DI)
MOVL R8,8(DI)
MOVL R9,12(DI)
MOVD X12,DX
MOVD X7,CX
MOVD X10,R8
MOVD X4,R9
PSHUFL $0X39,X12,X12
PSHUFL $0X39,X7,X7
PSHUFL $0X39,X10,X10
PSHUFL $0X39,X4,X4
XORL 64(SI),DX
XORL 68(SI),CX
XORL 72(SI),R8
XORL 76(SI),R9
MOVL DX,64(DI)
MOVL CX,68(DI)
MOVL R8,72(DI)
MOVL R9,76(DI)
MOVD X12,DX
MOVD X7,CX
MOVD X10,R8
MOVD X4,R9
PSHUFL $0X39,X12,X12
PSHUFL $0X39,X7,X7
PSHUFL $0X39,X10,X10
PSHUFL $0X39,X4,X4
XORL 128(SI),DX
XORL 132(SI),CX
XORL 136(SI),R8
XORL 140(SI),R9
MOVL DX,128(DI)
MOVL CX,132(DI)
MOVL R8,136(DI)
MOVL R9,140(DI)
MOVD X12,DX
MOVD X7,CX
MOVD X10,R8
MOVD X4,R9
XORL 192(SI),DX
XORL 196(SI),CX
XORL 200(SI),R8
XORL 204(SI),R9
MOVL DX,192(DI)
MOVL CX,196(DI)
MOVL R8,200(DI)
MOVL R9,204(DI)
PADDL 240(R12),X14
PADDL 64(R12),X0
PADDL 128(R12),X5
PADDL 192(R12),X8
MOVD X14,DX
MOVD X0,CX
MOVD X5,R8
MOVD X8,R9
PSHUFL $0X39,X14,X14
PSHUFL $0X39,X0,X0
PSHUFL $0X39,X5,X5
PSHUFL $0X39,X8,X8
XORL 16(SI),DX
XORL 20(SI),CX
XORL 24(SI),R8
XORL 28(SI),R9
MOVL DX,16(DI)
MOVL CX,20(DI)
MOVL R8,24(DI)
MOVL R9,28(DI)
MOVD X14,DX
MOVD X0,CX
MOVD X5,R8
MOVD X8,R9
PSHUFL $0X39,X14,X14
PSHUFL $0X39,X0,X0
PSHUFL $0X39,X5,X5
PSHUFL $0X39,X8,X8
XORL 80(SI),DX
XORL 84(SI),CX
XORL 88(SI),R8
XORL 92(SI),R9
MOVL DX,80(DI)
MOVL CX,84(DI)
MOVL R8,88(DI)
MOVL R9,92(DI)
MOVD X14,DX
MOVD X0,CX
MOVD X5,R8
MOVD X8,R9
PSHUFL $0X39,X14,X14
PSHUFL $0X39,X0,X0
PSHUFL $0X39,X5,X5
PSHUFL $0X39,X8,X8
XORL 144(SI),DX
XORL 148(SI),CX
XORL 152(SI),R8
XORL 156(SI),R9
MOVL DX,144(DI)
MOVL CX,148(DI)
MOVL R8,152(DI)
MOVL R9,156(DI)
MOVD X14,DX
MOVD X0,CX
MOVD X5,R8
MOVD X8,R9
XORL 208(SI),DX
XORL 212(SI),CX
XORL 216(SI),R8
XORL 220(SI),R9
MOVL DX,208(DI)
MOVL CX,212(DI)
MOVL R8,216(DI)
MOVL R9,220(DI)
PADDL 288(R12),X15
PADDL 304(R12),X11
PADDL 80(R12),X1
PADDL 144(R12),X6
MOVD X15,DX
MOVD X11,CX
MOVD X1,R8
MOVD X6,R9
PSHUFL $0X39,X15,X15
PSHUFL $0X39,X11,X11
PSHUFL $0X39,X1,X1
PSHUFL $0X39,X6,X6
XORL 32(SI),DX
XORL 36(SI),CX
XORL 40(SI),R8
XORL 44(SI),R9
MOVL DX,32(DI)
MOVL CX,36(DI)
MOVL R8,40(DI)
MOVL R9,44(DI)
MOVD X15,DX
MOVD X11,CX
MOVD X1,R8
MOVD X6,R9
PSHUFL $0X39,X15,X15
PSHUFL $0X39,X11,X11
PSHUFL $0X39,X1,X1
PSHUFL $0X39,X6,X6
XORL 96(SI),DX
XORL 100(SI),CX
XORL 104(SI),R8
XORL 108(SI),R9
MOVL DX,96(DI)
MOVL CX,100(DI)
MOVL R8,104(DI)
MOVL R9,108(DI)
MOVD X15,DX
MOVD X11,CX
MOVD X1,R8
MOVD X6,R9
PSHUFL $0X39,X15,X15
PSHUFL $0X39,X11,X11
PSHUFL $0X39,X1,X1
PSHUFL $0X39,X6,X6
XORL 160(SI),DX
XORL 164(SI),CX
XORL 168(SI),R8
XORL 172(SI),R9
MOVL DX,160(DI)
MOVL CX,164(DI)
MOVL R8,168(DI)
MOVL R9,172(DI)
MOVD X15,DX
MOVD X11,CX
MOVD X1,R8
MOVD X6,R9
XORL 224(SI),DX
XORL 228(SI),CX
XORL 232(SI),R8
XORL 236(SI),R9
MOVL DX,224(DI)
MOVL CX,228(DI)
MOVL R8,232(DI)
MOVL R9,236(DI)
PADDL 160(R12),X13
PADDL 208(R12),X9
PADDL 256(R12),X3
PADDL 96(R12),X2
MOVD X13,DX
MOVD X9,CX
MOVD X3,R8
MOVD X2,R9
PSHUFL $0X39,X13,X13
PSHUFL $0X39,X9,X9
PSHUFL $0X39,X3,X3
PSHUFL $0X39,X2,X2
XORL 48(SI),DX
XORL 52(SI),CX
XORL 56(SI),R8
XORL 60(SI),R9
MOVL DX,48(DI)
MOVL CX,52(DI)
MOVL R8,56(DI)
MOVL R9,60(DI)
MOVD X13,DX
MOVD X9,CX
MOVD X3,R8
MOVD X2,R9
PSHUFL $0X39,X13,X13
PSHUFL $0X39,X9,X9
PSHUFL $0X39,X3,X3
PSHUFL $0X39,X2,X2
XORL 112(SI),DX
XORL 116(SI),CX
XORL 120(SI),R8
XORL 124(SI),R9
MOVL DX,112(DI)
MOVL CX,116(DI)
MOVL R8,120(DI)
MOVL R9,124(DI)
MOVD X13,DX
MOVD X9,CX
MOVD X3,R8
MOVD X2,R9
PSHUFL $0X39,X13,X13
PSHUFL $0X39,X9,X9
PSHUFL $0X39,X3,X3
PSHUFL $0X39,X2,X2
XORL 176(SI),DX
XORL 180(SI),CX
XORL 184(SI),R8
XORL 188(SI),R9
MOVL DX,176(DI)
MOVL CX,180(DI)
MOVL R8,184(DI)
MOVL R9,188(DI)
MOVD X13,DX
MOVD X9,CX
MOVD X3,R8
MOVD X2,R9
XORL 240(SI),DX
XORL 244(SI),CX
XORL 248(SI),R8
XORL 252(SI),R9
MOVL DX,240(DI)
MOVL CX,244(DI)
MOVL R8,248(DI)
MOVL R9,252(DI)
MOVQ 352(R12),R9
SUBQ $256,R9
ADDQ $256,SI
ADDQ $256,DI
CMPQ R9,$256
JAE BYTESATLEAST256
CMPQ R9,$0
JBE DONE
BYTESBETWEEN1AND255:
CMPQ R9,$64
JAE NOCOPY
MOVQ DI,DX
LEAQ 360(R12),DI
MOVQ R9,CX
REP; MOVSB
LEAQ 360(R12),DI
LEAQ 360(R12),SI
NOCOPY:
MOVQ R9,352(R12)
MOVOA 48(R12),X0
MOVOA 0(R12),X1
MOVOA 16(R12),X2
MOVOA 32(R12),X3
MOVOA X1,X4
MOVQ $20,CX
MAINLOOP2:
PADDL X0,X4
MOVOA X0,X5
MOVOA X4,X6
PSLLL $7,X4
PSRLL $25,X6
PXOR X4,X3
PXOR X6,X3
PADDL X3,X5
MOVOA X3,X4
MOVOA X5,X6
PSLLL $9,X5
PSRLL $23,X6
PXOR X5,X2
PSHUFL $0X93,X3,X3
PXOR X6,X2
PADDL X2,X4
MOVOA X2,X5
MOVOA X4,X6
PSLLL $13,X4
PSRLL $19,X6
PXOR X4,X1
PSHUFL $0X4E,X2,X2
PXOR X6,X1
PADDL X1,X5
MOVOA X3,X4
MOVOA X5,X6
PSLLL $18,X5
PSRLL $14,X6
PXOR X5,X0
PSHUFL $0X39,X1,X1
PXOR X6,X0
PADDL X0,X4
MOVOA X0,X5
MOVOA X4,X6
PSLLL $7,X4
PSRLL $25,X6
PXOR X4,X1
PXOR X6,X1
PADDL X1,X5
MOVOA X1,X4
MOVOA X5,X6
PSLLL $9,X5
PSRLL $23,X6
PXOR X5,X2
PSHUFL $0X93,X1,X1
PXOR X6,X2
PADDL X2,X4
MOVOA X2,X5
MOVOA X4,X6
PSLLL $13,X4
PSRLL $19,X6
PXOR X4,X3
PSHUFL $0X4E,X2,X2
PXOR X6,X3
PADDL X3,X5
MOVOA X1,X4
MOVOA X5,X6
PSLLL $18,X5
PSRLL $14,X6
PXOR X5,X0
PSHUFL $0X39,X3,X3
PXOR X6,X0
PADDL X0,X4
MOVOA X0,X5
MOVOA X4,X6
PSLLL $7,X4
PSRLL $25,X6
PXOR X4,X3
PXOR X6,X3
PADDL X3,X5
MOVOA X3,X4
MOVOA X5,X6
PSLLL $9,X5
PSRLL $23,X6
PXOR X5,X2
PSHUFL $0X93,X3,X3
PXOR X6,X2
PADDL X2,X4
MOVOA X2,X5
MOVOA X4,X6
PSLLL $13,X4
PSRLL $19,X6
PXOR X4,X1
PSHUFL $0X4E,X2,X2
PXOR X6,X1
PADDL X1,X5
MOVOA X3,X4
MOVOA X5,X6
PSLLL $18,X5
PSRLL $14,X6
PXOR X5,X0
PSHUFL $0X39,X1,X1
PXOR X6,X0
PADDL X0,X4
MOVOA X0,X5
MOVOA X4,X6
PSLLL $7,X4
PSRLL $25,X6
PXOR X4,X1
PXOR X6,X1
PADDL X1,X5
MOVOA X1,X4
MOVOA X5,X6
PSLLL $9,X5
PSRLL $23,X6
PXOR X5,X2
PSHUFL $0X93,X1,X1
PXOR X6,X2
PADDL X2,X4
MOVOA X2,X5
MOVOA X4,X6
PSLLL $13,X4
PSRLL $19,X6
PXOR X4,X3
PSHUFL $0X4E,X2,X2
PXOR X6,X3
SUBQ $4,CX
PADDL X3,X5
MOVOA X1,X4
MOVOA X5,X6
PSLLL $18,X5
PXOR X7,X7
PSRLL $14,X6
PXOR X5,X0
PSHUFL $0X39,X3,X3
PXOR X6,X0
JA MAINLOOP2
PADDL 48(R12),X0
PADDL 0(R12),X1
PADDL 16(R12),X2
PADDL 32(R12),X3
MOVD X0,CX
MOVD X1,R8
MOVD X2,R9
MOVD X3,AX
PSHUFL $0X39,X0,X0
PSHUFL $0X39,X1,X1
PSHUFL $0X39,X2,X2
PSHUFL $0X39,X3,X3
XORL 0(SI),CX
XORL 48(SI),R8
XORL 32(SI),R9
XORL 16(SI),AX
MOVL CX,0(DI)
MOVL R8,48(DI)
MOVL R9,32(DI)
MOVL AX,16(DI)
MOVD X0,CX
MOVD X1,R8
MOVD X2,R9
MOVD X3,AX
PSHUFL $0X39,X0,X0
PSHUFL $0X39,X1,X1
PSHUFL $0X39,X2,X2
PSHUFL $0X39,X3,X3
XORL 20(SI),CX
XORL 4(SI),R8
XORL 52(SI),R9
XORL 36(SI),AX
MOVL CX,20(DI)
MOVL R8,4(DI)
MOVL R9,52(DI)
MOVL AX,36(DI)
MOVD X0,CX
MOVD X1,R8
MOVD X2,R9
MOVD X3,AX
PSHUFL $0X39,X0,X0
PSHUFL $0X39,X1,X1
PSHUFL $0X39,X2,X2
PSHUFL $0X39,X3,X3
XORL 40(SI),CX
XORL 24(SI),R8
XORL 8(SI),R9
XORL 56(SI),AX
MOVL CX,40(DI)
MOVL R8,24(DI)
MOVL R9,8(DI)
MOVL AX,56(DI)
MOVD X0,CX
MOVD X1,R8
MOVD X2,R9
MOVD X3,AX
XORL 60(SI),CX
XORL 44(SI),R8
XORL 28(SI),R9
XORL 12(SI),AX
MOVL CX,60(DI)
MOVL R8,44(DI)
MOVL R9,28(DI)
MOVL AX,12(DI)
MOVQ 352(R12),R9
MOVL 16(R12),CX
MOVL 36 (R12),R8
ADDQ $1,CX
SHLQ $32,R8
ADDQ R8,CX
MOVQ CX,R8
SHRQ $32,R8
MOVL CX,16(R12)
MOVL R8, 36 (R12)
CMPQ R9,$64
JA BYTESATLEAST65
JAE BYTESATLEAST64
MOVQ DI,SI
MOVQ DX,DI
MOVQ R9,CX
REP; MOVSB
BYTESATLEAST64:
DONE:
RET
BYTESATLEAST65:
SUBQ $64,R9
ADDQ $64,DI
ADDQ $64,SI
JMP BYTESBETWEEN1AND255

15
gateway/vendor/golang.org/x/crypto/salsa20/salsa/salsa20_noasm.go generated vendored Normal file
View File

@ -0,0 +1,15 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !amd64 || purego || !gc
// +build !amd64 purego !gc
package salsa
// XORKeyStream crypts bytes from in to out using the given key and counters.
// In and out must overlap entirely or not at all. Counter
// contains the raw salsa20 counter bytes (both nonce and block counter).
func XORKeyStream(out, in []byte, counter *[16]byte, key *[32]byte) {
genericXORKeyStream(out, in, counter, key)
}

233
gateway/vendor/golang.org/x/crypto/salsa20/salsa/salsa20_ref.go generated vendored Normal file
View File

@ -0,0 +1,233 @@
// Copyright 2012 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package salsa
import "math/bits"
const rounds = 20
// core applies the Salsa20 core function to 16-byte input in, 32-byte key k,
// and 16-byte constant c, and puts the result into 64-byte array out.
func core(out *[64]byte, in *[16]byte, k *[32]byte, c *[16]byte) {
j0 := uint32(c[0]) | uint32(c[1])<<8 | uint32(c[2])<<16 | uint32(c[3])<<24
j1 := uint32(k[0]) | uint32(k[1])<<8 | uint32(k[2])<<16 | uint32(k[3])<<24
j2 := uint32(k[4]) | uint32(k[5])<<8 | uint32(k[6])<<16 | uint32(k[7])<<24
j3 := uint32(k[8]) | uint32(k[9])<<8 | uint32(k[10])<<16 | uint32(k[11])<<24
j4 := uint32(k[12]) | uint32(k[13])<<8 | uint32(k[14])<<16 | uint32(k[15])<<24
j5 := uint32(c[4]) | uint32(c[5])<<8 | uint32(c[6])<<16 | uint32(c[7])<<24
j6 := uint32(in[0]) | uint32(in[1])<<8 | uint32(in[2])<<16 | uint32(in[3])<<24
j7 := uint32(in[4]) | uint32(in[5])<<8 | uint32(in[6])<<16 | uint32(in[7])<<24
j8 := uint32(in[8]) | uint32(in[9])<<8 | uint32(in[10])<<16 | uint32(in[11])<<24
j9 := uint32(in[12]) | uint32(in[13])<<8 | uint32(in[14])<<16 | uint32(in[15])<<24
j10 := uint32(c[8]) | uint32(c[9])<<8 | uint32(c[10])<<16 | uint32(c[11])<<24
j11 := uint32(k[16]) | uint32(k[17])<<8 | uint32(k[18])<<16 | uint32(k[19])<<24
j12 := uint32(k[20]) | uint32(k[21])<<8 | uint32(k[22])<<16 | uint32(k[23])<<24
j13 := uint32(k[24]) | uint32(k[25])<<8 | uint32(k[26])<<16 | uint32(k[27])<<24
j14 := uint32(k[28]) | uint32(k[29])<<8 | uint32(k[30])<<16 | uint32(k[31])<<24
j15 := uint32(c[12]) | uint32(c[13])<<8 | uint32(c[14])<<16 | uint32(c[15])<<24
x0, x1, x2, x3, x4, x5, x6, x7, x8 := j0, j1, j2, j3, j4, j5, j6, j7, j8
x9, x10, x11, x12, x13, x14, x15 := j9, j10, j11, j12, j13, j14, j15
for i := 0; i < rounds; i += 2 {
u := x0 + x12
x4 ^= bits.RotateLeft32(u, 7)
u = x4 + x0
x8 ^= bits.RotateLeft32(u, 9)
u = x8 + x4
x12 ^= bits.RotateLeft32(u, 13)
u = x12 + x8
x0 ^= bits.RotateLeft32(u, 18)
u = x5 + x1
x9 ^= bits.RotateLeft32(u, 7)
u = x9 + x5
x13 ^= bits.RotateLeft32(u, 9)
u = x13 + x9
x1 ^= bits.RotateLeft32(u, 13)
u = x1 + x13
x5 ^= bits.RotateLeft32(u, 18)
u = x10 + x6
x14 ^= bits.RotateLeft32(u, 7)
u = x14 + x10
x2 ^= bits.RotateLeft32(u, 9)
u = x2 + x14
x6 ^= bits.RotateLeft32(u, 13)
u = x6 + x2
x10 ^= bits.RotateLeft32(u, 18)
u = x15 + x11
x3 ^= bits.RotateLeft32(u, 7)
u = x3 + x15
x7 ^= bits.RotateLeft32(u, 9)
u = x7 + x3
x11 ^= bits.RotateLeft32(u, 13)
u = x11 + x7
x15 ^= bits.RotateLeft32(u, 18)
u = x0 + x3
x1 ^= bits.RotateLeft32(u, 7)
u = x1 + x0
x2 ^= bits.RotateLeft32(u, 9)
u = x2 + x1
x3 ^= bits.RotateLeft32(u, 13)
u = x3 + x2
x0 ^= bits.RotateLeft32(u, 18)
u = x5 + x4
x6 ^= bits.RotateLeft32(u, 7)
u = x6 + x5
x7 ^= bits.RotateLeft32(u, 9)
u = x7 + x6
x4 ^= bits.RotateLeft32(u, 13)
u = x4 + x7
x5 ^= bits.RotateLeft32(u, 18)
u = x10 + x9
x11 ^= bits.RotateLeft32(u, 7)
u = x11 + x10
x8 ^= bits.RotateLeft32(u, 9)
u = x8 + x11
x9 ^= bits.RotateLeft32(u, 13)
u = x9 + x8
x10 ^= bits.RotateLeft32(u, 18)
u = x15 + x14
x12 ^= bits.RotateLeft32(u, 7)
u = x12 + x15
x13 ^= bits.RotateLeft32(u, 9)
u = x13 + x12
x14 ^= bits.RotateLeft32(u, 13)
u = x14 + x13
x15 ^= bits.RotateLeft32(u, 18)
}
x0 += j0
x1 += j1
x2 += j2
x3 += j3
x4 += j4
x5 += j5
x6 += j6
x7 += j7
x8 += j8
x9 += j9
x10 += j10
x11 += j11
x12 += j12
x13 += j13
x14 += j14
x15 += j15
out[0] = byte(x0)
out[1] = byte(x0 >> 8)
out[2] = byte(x0 >> 16)
out[3] = byte(x0 >> 24)
out[4] = byte(x1)
out[5] = byte(x1 >> 8)
out[6] = byte(x1 >> 16)
out[7] = byte(x1 >> 24)
out[8] = byte(x2)
out[9] = byte(x2 >> 8)
out[10] = byte(x2 >> 16)
out[11] = byte(x2 >> 24)
out[12] = byte(x3)
out[13] = byte(x3 >> 8)
out[14] = byte(x3 >> 16)
out[15] = byte(x3 >> 24)
out[16] = byte(x4)
out[17] = byte(x4 >> 8)
out[18] = byte(x4 >> 16)
out[19] = byte(x4 >> 24)
out[20] = byte(x5)
out[21] = byte(x5 >> 8)
out[22] = byte(x5 >> 16)
out[23] = byte(x5 >> 24)
out[24] = byte(x6)
out[25] = byte(x6 >> 8)
out[26] = byte(x6 >> 16)
out[27] = byte(x6 >> 24)
out[28] = byte(x7)
out[29] = byte(x7 >> 8)
out[30] = byte(x7 >> 16)
out[31] = byte(x7 >> 24)
out[32] = byte(x8)
out[33] = byte(x8 >> 8)
out[34] = byte(x8 >> 16)
out[35] = byte(x8 >> 24)
out[36] = byte(x9)
out[37] = byte(x9 >> 8)
out[38] = byte(x9 >> 16)
out[39] = byte(x9 >> 24)
out[40] = byte(x10)
out[41] = byte(x10 >> 8)
out[42] = byte(x10 >> 16)
out[43] = byte(x10 >> 24)
out[44] = byte(x11)
out[45] = byte(x11 >> 8)
out[46] = byte(x11 >> 16)
out[47] = byte(x11 >> 24)
out[48] = byte(x12)
out[49] = byte(x12 >> 8)
out[50] = byte(x12 >> 16)
out[51] = byte(x12 >> 24)
out[52] = byte(x13)
out[53] = byte(x13 >> 8)
out[54] = byte(x13 >> 16)
out[55] = byte(x13 >> 24)
out[56] = byte(x14)
out[57] = byte(x14 >> 8)
out[58] = byte(x14 >> 16)
out[59] = byte(x14 >> 24)
out[60] = byte(x15)
out[61] = byte(x15 >> 8)
out[62] = byte(x15 >> 16)
out[63] = byte(x15 >> 24)
}
// genericXORKeyStream is the generic implementation of XORKeyStream to be used
// when no assembly implementation is available.
func genericXORKeyStream(out, in []byte, counter *[16]byte, key *[32]byte) {
var block [64]byte
var counterCopy [16]byte
copy(counterCopy[:], counter[:])
for len(in) >= 64 {
core(&block, &counterCopy, key, &Sigma)
for i, x := range block {
out[i] = in[i] ^ x
}
u := uint32(1)
for i := 8; i < 16; i++ {
u += uint32(counterCopy[i])
counterCopy[i] = byte(u)
u >>= 8
}
in = in[64:]
out = out[64:]
}
if len(in) > 0 {
core(&block, &counterCopy, key, &Sigma)
for i, v := range in {
out[i] = v ^ block[i]
}
}
}

9
gateway/vendor/golang.org/x/sync/singleflight/singleflight.go generated vendored
View File

@ -31,6 +31,15 @@ func (p *panicError) Error() string {
return fmt.Sprintf("%v\n\n%s", p.value, p.stack)
}
func (p *panicError) Unwrap() error {
err, ok := p.value.(error)
if !ok {
return nil
}
return err
}
func newPanicError(v interface{}) error {
stack := debug.Stack()

18
gateway/vendor/golang.org/x/sys/cpu/asm_aix_ppc64.s generated vendored Normal file
View File

@ -0,0 +1,18 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc
// +build gc
#include "textflag.h"
//
// System calls for ppc64, AIX are implemented in runtime/syscall_aix.go
//
TEXT ·syscall6(SB),NOSPLIT,$0-88
JMP syscall·syscall6(SB)
TEXT ·rawSyscall6(SB),NOSPLIT,$0-88
JMP syscall·rawSyscall6(SB)

66
gateway/vendor/golang.org/x/sys/cpu/byteorder.go generated vendored Normal file
View File

@ -0,0 +1,66 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import (
"runtime"
)
// byteOrder is a subset of encoding/binary.ByteOrder.
type byteOrder interface {
Uint32([]byte) uint32
Uint64([]byte) uint64
}
type littleEndian struct{}
type bigEndian struct{}
func (littleEndian) Uint32(b []byte) uint32 {
_ = b[3] // bounds check hint to compiler; see golang.org/issue/14808
return uint32(b[0]) | uint32(b[1])<<8 | uint32(b[2])<<16 | uint32(b[3])<<24
}
func (littleEndian) Uint64(b []byte) uint64 {
_ = b[7] // bounds check hint to compiler; see golang.org/issue/14808
return uint64(b[0]) | uint64(b[1])<<8 | uint64(b[2])<<16 | uint64(b[3])<<24 |
uint64(b[4])<<32 | uint64(b[5])<<40 | uint64(b[6])<<48 | uint64(b[7])<<56
}
func (bigEndian) Uint32(b []byte) uint32 {
_ = b[3] // bounds check hint to compiler; see golang.org/issue/14808
return uint32(b[3]) | uint32(b[2])<<8 | uint32(b[1])<<16 | uint32(b[0])<<24
}
func (bigEndian) Uint64(b []byte) uint64 {
_ = b[7] // bounds check hint to compiler; see golang.org/issue/14808
return uint64(b[7]) | uint64(b[6])<<8 | uint64(b[5])<<16 | uint64(b[4])<<24 |
uint64(b[3])<<32 | uint64(b[2])<<40 | uint64(b[1])<<48 | uint64(b[0])<<56
}
// hostByteOrder returns littleEndian on little-endian machines and
// bigEndian on big-endian machines.
func hostByteOrder() byteOrder {
switch runtime.GOARCH {
case "386", "amd64", "amd64p32",
"alpha",
"arm", "arm64",
"loong64",
"mipsle", "mips64le", "mips64p32le",
"nios2",
"ppc64le",
"riscv", "riscv64",
"sh":
return littleEndian{}
case "armbe", "arm64be",
"m68k",
"mips", "mips64", "mips64p32",
"ppc", "ppc64",
"s390", "s390x",
"shbe",
"sparc", "sparc64":
return bigEndian{}
}
panic("unknown architecture")
}

290
gateway/vendor/golang.org/x/sys/cpu/cpu.go generated vendored Normal file
View File

@ -0,0 +1,290 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package cpu implements processor feature detection for
// various CPU architectures.
package cpu
import (
"os"
"strings"
)
// Initialized reports whether the CPU features were initialized.
//
// For some GOOS/GOARCH combinations initialization of the CPU features depends
// on reading an operating specific file, e.g. /proc/self/auxv on linux/arm
// Initialized will report false if reading the file fails.
var Initialized bool
// CacheLinePad is used to pad structs to avoid false sharing.
type CacheLinePad struct{ _ [cacheLineSize]byte }
// X86 contains the supported CPU features of the
// current X86/AMD64 platform. If the current platform
// is not X86/AMD64 then all feature flags are false.
//
// X86 is padded to avoid false sharing. Further the HasAVX
// and HasAVX2 are only set if the OS supports XMM and YMM
// registers in addition to the CPUID feature bit being set.
var X86 struct {
_ CacheLinePad
HasAES bool // AES hardware implementation (AES NI)
HasADX bool // Multi-precision add-carry instruction extensions
HasAVX bool // Advanced vector extension
HasAVX2 bool // Advanced vector extension 2
HasAVX512 bool // Advanced vector extension 512
HasAVX512F bool // Advanced vector extension 512 Foundation Instructions
HasAVX512CD bool // Advanced vector extension 512 Conflict Detection Instructions
HasAVX512ER bool // Advanced vector extension 512 Exponential and Reciprocal Instructions
HasAVX512PF bool // Advanced vector extension 512 Prefetch Instructions
HasAVX512VL bool // Advanced vector extension 512 Vector Length Extensions
HasAVX512BW bool // Advanced vector extension 512 Byte and Word Instructions
HasAVX512DQ bool // Advanced vector extension 512 Doubleword and Quadword Instructions
HasAVX512IFMA bool // Advanced vector extension 512 Integer Fused Multiply Add
HasAVX512VBMI bool // Advanced vector extension 512 Vector Byte Manipulation Instructions
HasAVX5124VNNIW bool // Advanced vector extension 512 Vector Neural Network Instructions Word variable precision
HasAVX5124FMAPS bool // Advanced vector extension 512 Fused Multiply Accumulation Packed Single precision
HasAVX512VPOPCNTDQ bool // Advanced vector extension 512 Double and quad word population count instructions
HasAVX512VPCLMULQDQ bool // Advanced vector extension 512 Vector carry-less multiply operations
HasAVX512VNNI bool // Advanced vector extension 512 Vector Neural Network Instructions
HasAVX512GFNI bool // Advanced vector extension 512 Galois field New Instructions
HasAVX512VAES bool // Advanced vector extension 512 Vector AES instructions
HasAVX512VBMI2 bool // Advanced vector extension 512 Vector Byte Manipulation Instructions 2
HasAVX512BITALG bool // Advanced vector extension 512 Bit Algorithms
HasAVX512BF16 bool // Advanced vector extension 512 BFloat16 Instructions
HasAMXTile bool // Advanced Matrix Extension Tile instructions
HasAMXInt8 bool // Advanced Matrix Extension Int8 instructions
HasAMXBF16 bool // Advanced Matrix Extension BFloat16 instructions
HasBMI1 bool // Bit manipulation instruction set 1
HasBMI2 bool // Bit manipulation instruction set 2
HasCX16 bool // Compare and exchange 16 Bytes
HasERMS bool // Enhanced REP for MOVSB and STOSB
HasFMA bool // Fused-multiply-add instructions
HasOSXSAVE bool // OS supports XSAVE/XRESTOR for saving/restoring XMM registers.
HasPCLMULQDQ bool // PCLMULQDQ instruction - most often used for AES-GCM
HasPOPCNT bool // Hamming weight instruction POPCNT.
HasRDRAND bool // RDRAND instruction (on-chip random number generator)
HasRDSEED bool // RDSEED instruction (on-chip random number generator)
HasSSE2 bool // Streaming SIMD extension 2 (always available on amd64)
HasSSE3 bool // Streaming SIMD extension 3
HasSSSE3 bool // Supplemental streaming SIMD extension 3
HasSSE41 bool // Streaming SIMD extension 4 and 4.1
HasSSE42 bool // Streaming SIMD extension 4 and 4.2
_ CacheLinePad
}
// ARM64 contains the supported CPU features of the
// current ARMv8(aarch64) platform. If the current platform
// is not arm64 then all feature flags are false.
var ARM64 struct {
_ CacheLinePad
HasFP bool // Floating-point instruction set (always available)
HasASIMD bool // Advanced SIMD (always available)
HasEVTSTRM bool // Event stream support
HasAES bool // AES hardware implementation
HasPMULL bool // Polynomial multiplication instruction set
HasSHA1 bool // SHA1 hardware implementation
HasSHA2 bool // SHA2 hardware implementation
HasCRC32 bool // CRC32 hardware implementation
HasATOMICS bool // Atomic memory operation instruction set
HasFPHP bool // Half precision floating-point instruction set
HasASIMDHP bool // Advanced SIMD half precision instruction set
HasCPUID bool // CPUID identification scheme registers
HasASIMDRDM bool // Rounding double multiply add/subtract instruction set
HasJSCVT bool // Javascript conversion from floating-point to integer
HasFCMA bool // Floating-point multiplication and addition of complex numbers
HasLRCPC bool // Release Consistent processor consistent support
HasDCPOP bool // Persistent memory support
HasSHA3 bool // SHA3 hardware implementation
HasSM3 bool // SM3 hardware implementation
HasSM4 bool // SM4 hardware implementation
HasASIMDDP bool // Advanced SIMD double precision instruction set
HasSHA512 bool // SHA512 hardware implementation
HasSVE bool // Scalable Vector Extensions
HasASIMDFHM bool // Advanced SIMD multiplication FP16 to FP32
_ CacheLinePad
}
// ARM contains the supported CPU features of the current ARM (32-bit) platform.
// All feature flags are false if:
// 1. the current platform is not arm, or
// 2. the current operating system is not Linux.
var ARM struct {
_ CacheLinePad
HasSWP bool // SWP instruction support
HasHALF bool // Half-word load and store support
HasTHUMB bool // ARM Thumb instruction set
Has26BIT bool // Address space limited to 26-bits
HasFASTMUL bool // 32-bit operand, 64-bit result multiplication support
HasFPA bool // Floating point arithmetic support
HasVFP bool // Vector floating point support
HasEDSP bool // DSP Extensions support
HasJAVA bool // Java instruction set
HasIWMMXT bool // Intel Wireless MMX technology support
HasCRUNCH bool // MaverickCrunch context switching and handling
HasTHUMBEE bool // Thumb EE instruction set
HasNEON bool // NEON instruction set
HasVFPv3 bool // Vector floating point version 3 support
HasVFPv3D16 bool // Vector floating point version 3 D8-D15
HasTLS bool // Thread local storage support
HasVFPv4 bool // Vector floating point version 4 support
HasIDIVA bool // Integer divide instruction support in ARM mode
HasIDIVT bool // Integer divide instruction support in Thumb mode
HasVFPD32 bool // Vector floating point version 3 D15-D31
HasLPAE bool // Large Physical Address Extensions
HasEVTSTRM bool // Event stream support
HasAES bool // AES hardware implementation
HasPMULL bool // Polynomial multiplication instruction set
HasSHA1 bool // SHA1 hardware implementation
HasSHA2 bool // SHA2 hardware implementation
HasCRC32 bool // CRC32 hardware implementation
_ CacheLinePad
}
// MIPS64X contains the supported CPU features of the current mips64/mips64le
// platforms. If the current platform is not mips64/mips64le or the current
// operating system is not Linux then all feature flags are false.
var MIPS64X struct {
_ CacheLinePad
HasMSA bool // MIPS SIMD architecture
_ CacheLinePad
}
// PPC64 contains the supported CPU features of the current ppc64/ppc64le platforms.
// If the current platform is not ppc64/ppc64le then all feature flags are false.
//
// For ppc64/ppc64le, it is safe to check only for ISA level starting on ISA v3.00,
// since there are no optional categories. There are some exceptions that also
// require kernel support to work (DARN, SCV), so there are feature bits for
// those as well. The struct is padded to avoid false sharing.
var PPC64 struct {
_ CacheLinePad
HasDARN bool // Hardware random number generator (requires kernel enablement)
HasSCV bool // Syscall vectored (requires kernel enablement)
IsPOWER8 bool // ISA v2.07 (POWER8)
IsPOWER9 bool // ISA v3.00 (POWER9), implies IsPOWER8
_ CacheLinePad
}
// S390X contains the supported CPU features of the current IBM Z
// (s390x) platform. If the current platform is not IBM Z then all
// feature flags are false.
//
// S390X is padded to avoid false sharing. Further HasVX is only set
// if the OS supports vector registers in addition to the STFLE
// feature bit being set.
var S390X struct {
_ CacheLinePad
HasZARCH bool // z/Architecture mode is active [mandatory]
HasSTFLE bool // store facility list extended
HasLDISP bool // long (20-bit) displacements
HasEIMM bool // 32-bit immediates
HasDFP bool // decimal floating point
HasETF3EH bool // ETF-3 enhanced
HasMSA bool // message security assist (CPACF)
HasAES bool // KM-AES{128,192,256} functions
HasAESCBC bool // KMC-AES{128,192,256} functions
HasAESCTR bool // KMCTR-AES{128,192,256} functions
HasAESGCM bool // KMA-GCM-AES{128,192,256} functions
HasGHASH bool // KIMD-GHASH function
HasSHA1 bool // K{I,L}MD-SHA-1 functions
HasSHA256 bool // K{I,L}MD-SHA-256 functions
HasSHA512 bool // K{I,L}MD-SHA-512 functions
HasSHA3 bool // K{I,L}MD-SHA3-{224,256,384,512} and K{I,L}MD-SHAKE-{128,256} functions
HasVX bool // vector facility
HasVXE bool // vector-enhancements facility 1
_ CacheLinePad
}
func init() {
archInit()
initOptions()
processOptions()
}
// options contains the cpu debug options that can be used in GODEBUG.
// Options are arch dependent and are added by the arch specific initOptions functions.
// Features that are mandatory for the specific GOARCH should have the Required field set
// (e.g. SSE2 on amd64).
var options []option
// Option names should be lower case. e.g. avx instead of AVX.
type option struct {
Name string
Feature *bool
Specified bool // whether feature value was specified in GODEBUG
Enable bool // whether feature should be enabled
Required bool // whether feature is mandatory and can not be disabled
}
func processOptions() {
env := os.Getenv("GODEBUG")
field:
for env != "" {
field := ""
i := strings.IndexByte(env, ',')
if i < 0 {
field, env = env, ""
} else {
field, env = env[:i], env[i+1:]
}
if len(field) < 4 || field[:4] != "cpu." {
continue
}
i = strings.IndexByte(field, '=')
if i < 0 {
print("GODEBUG sys/cpu: no value specified for \"", field, "\"\n")
continue
}
key, value := field[4:i], field[i+1:] // e.g. "SSE2", "on"
var enable bool
switch value {
case "on":
enable = true
case "off":
enable = false
default:
print("GODEBUG sys/cpu: value \"", value, "\" not supported for cpu option \"", key, "\"\n")
continue field
}
if key == "all" {
for i := range options {
options[i].Specified = true
options[i].Enable = enable || options[i].Required
}
continue field
}
for i := range options {
if options[i].Name == key {
options[i].Specified = true
options[i].Enable = enable
continue field
}
}
print("GODEBUG sys/cpu: unknown cpu feature \"", key, "\"\n")
}
for _, o := range options {
if !o.Specified {
continue
}
if o.Enable && !*o.Feature {
print("GODEBUG sys/cpu: can not enable \"", o.Name, "\", missing CPU support\n")
continue
}
if !o.Enable && o.Required {
print("GODEBUG sys/cpu: can not disable \"", o.Name, "\", required CPU feature\n")
continue
}
*o.Feature = o.Enable
}
}

34
gateway/vendor/golang.org/x/sys/cpu/cpu_aix.go generated vendored Normal file
View File

@ -0,0 +1,34 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build aix
// +build aix
package cpu
const (
// getsystemcfg constants
_SC_IMPL = 2
_IMPL_POWER8 = 0x10000
_IMPL_POWER9 = 0x20000
)
func archInit() {
impl := getsystemcfg(_SC_IMPL)
if impl&_IMPL_POWER8 != 0 {
PPC64.IsPOWER8 = true
}
if impl&_IMPL_POWER9 != 0 {
PPC64.IsPOWER8 = true
PPC64.IsPOWER9 = true
}
Initialized = true
}
func getsystemcfg(label int) (n uint64) {
r0, _ := callgetsystemcfg(label)
n = uint64(r0)
return
}

73
gateway/vendor/golang.org/x/sys/cpu/cpu_arm.go generated vendored Normal file
View File

@ -0,0 +1,73 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
const cacheLineSize = 32
// HWCAP/HWCAP2 bits.
// These are specific to Linux.
const (
hwcap_SWP = 1 << 0
hwcap_HALF = 1 << 1
hwcap_THUMB = 1 << 2
hwcap_26BIT = 1 << 3
hwcap_FAST_MULT = 1 << 4
hwcap_FPA = 1 << 5
hwcap_VFP = 1 << 6
hwcap_EDSP = 1 << 7
hwcap_JAVA = 1 << 8
hwcap_IWMMXT = 1 << 9
hwcap_CRUNCH = 1 << 10
hwcap_THUMBEE = 1 << 11
hwcap_NEON = 1 << 12
hwcap_VFPv3 = 1 << 13
hwcap_VFPv3D16 = 1 << 14
hwcap_TLS = 1 << 15
hwcap_VFPv4 = 1 << 16
hwcap_IDIVA = 1 << 17
hwcap_IDIVT = 1 << 18
hwcap_VFPD32 = 1 << 19
hwcap_LPAE = 1 << 20
hwcap_EVTSTRM = 1 << 21
hwcap2_AES = 1 << 0
hwcap2_PMULL = 1 << 1
hwcap2_SHA1 = 1 << 2
hwcap2_SHA2 = 1 << 3
hwcap2_CRC32 = 1 << 4
)
func initOptions() {
options = []option{
{Name: "pmull", Feature: &ARM.HasPMULL},
{Name: "sha1", Feature: &ARM.HasSHA1},
{Name: "sha2", Feature: &ARM.HasSHA2},
{Name: "swp", Feature: &ARM.HasSWP},
{Name: "thumb", Feature: &ARM.HasTHUMB},
{Name: "thumbee", Feature: &ARM.HasTHUMBEE},
{Name: "tls", Feature: &ARM.HasTLS},
{Name: "vfp", Feature: &ARM.HasVFP},
{Name: "vfpd32", Feature: &ARM.HasVFPD32},
{Name: "vfpv3", Feature: &ARM.HasVFPv3},
{Name: "vfpv3d16", Feature: &ARM.HasVFPv3D16},
{Name: "vfpv4", Feature: &ARM.HasVFPv4},
{Name: "half", Feature: &ARM.HasHALF},
{Name: "26bit", Feature: &ARM.Has26BIT},
{Name: "fastmul", Feature: &ARM.HasFASTMUL},
{Name: "fpa", Feature: &ARM.HasFPA},
{Name: "edsp", Feature: &ARM.HasEDSP},
{Name: "java", Feature: &ARM.HasJAVA},
{Name: "iwmmxt", Feature: &ARM.HasIWMMXT},
{Name: "crunch", Feature: &ARM.HasCRUNCH},
{Name: "neon", Feature: &ARM.HasNEON},
{Name: "idivt", Feature: &ARM.HasIDIVT},
{Name: "idiva", Feature: &ARM.HasIDIVA},
{Name: "lpae", Feature: &ARM.HasLPAE},
{Name: "evtstrm", Feature: &ARM.HasEVTSTRM},
{Name: "aes", Feature: &ARM.HasAES},
{Name: "crc32", Feature: &ARM.HasCRC32},
}
}

172
gateway/vendor/golang.org/x/sys/cpu/cpu_arm64.go generated vendored Normal file
View File

@ -0,0 +1,172 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import "runtime"
// cacheLineSize is used to prevent false sharing of cache lines.
// We choose 128 because Apple Silicon, a.k.a. M1, has 128-byte cache line size.
// It doesn't cost much and is much more future-proof.
const cacheLineSize = 128
func initOptions() {
options = []option{
{Name: "fp", Feature: &ARM64.HasFP},
{Name: "asimd", Feature: &ARM64.HasASIMD},
{Name: "evstrm", Feature: &ARM64.HasEVTSTRM},
{Name: "aes", Feature: &ARM64.HasAES},
{Name: "fphp", Feature: &ARM64.HasFPHP},
{Name: "jscvt", Feature: &ARM64.HasJSCVT},
{Name: "lrcpc", Feature: &ARM64.HasLRCPC},
{Name: "pmull", Feature: &ARM64.HasPMULL},
{Name: "sha1", Feature: &ARM64.HasSHA1},
{Name: "sha2", Feature: &ARM64.HasSHA2},
{Name: "sha3", Feature: &ARM64.HasSHA3},
{Name: "sha512", Feature: &ARM64.HasSHA512},
{Name: "sm3", Feature: &ARM64.HasSM3},
{Name: "sm4", Feature: &ARM64.HasSM4},
{Name: "sve", Feature: &ARM64.HasSVE},
{Name: "crc32", Feature: &ARM64.HasCRC32},
{Name: "atomics", Feature: &ARM64.HasATOMICS},
{Name: "asimdhp", Feature: &ARM64.HasASIMDHP},
{Name: "cpuid", Feature: &ARM64.HasCPUID},
{Name: "asimrdm", Feature: &ARM64.HasASIMDRDM},
{Name: "fcma", Feature: &ARM64.HasFCMA},
{Name: "dcpop", Feature: &ARM64.HasDCPOP},
{Name: "asimddp", Feature: &ARM64.HasASIMDDP},
{Name: "asimdfhm", Feature: &ARM64.HasASIMDFHM},
}
}
func archInit() {
switch runtime.GOOS {
case "freebsd":
readARM64Registers()
case "linux", "netbsd", "openbsd":
doinit()
default:
// Many platforms don't seem to allow reading these registers.
setMinimalFeatures()
}
}
// setMinimalFeatures fakes the minimal ARM64 features expected by
// TestARM64minimalFeatures.
func setMinimalFeatures() {
ARM64.HasASIMD = true
ARM64.HasFP = true
}
func readARM64Registers() {
Initialized = true
parseARM64SystemRegisters(getisar0(), getisar1(), getpfr0())
}
func parseARM64SystemRegisters(isar0, isar1, pfr0 uint64) {
// ID_AA64ISAR0_EL1
switch extractBits(isar0, 4, 7) {
case 1:
ARM64.HasAES = true
case 2:
ARM64.HasAES = true
ARM64.HasPMULL = true
}
switch extractBits(isar0, 8, 11) {
case 1:
ARM64.HasSHA1 = true
}
switch extractBits(isar0, 12, 15) {
case 1:
ARM64.HasSHA2 = true
case 2:
ARM64.HasSHA2 = true
ARM64.HasSHA512 = true
}
switch extractBits(isar0, 16, 19) {
case 1:
ARM64.HasCRC32 = true
}
switch extractBits(isar0, 20, 23) {
case 2:
ARM64.HasATOMICS = true
}
switch extractBits(isar0, 28, 31) {
case 1:
ARM64.HasASIMDRDM = true
}
switch extractBits(isar0, 32, 35) {
case 1:
ARM64.HasSHA3 = true
}
switch extractBits(isar0, 36, 39) {
case 1:
ARM64.HasSM3 = true
}
switch extractBits(isar0, 40, 43) {
case 1:
ARM64.HasSM4 = true
}
switch extractBits(isar0, 44, 47) {
case 1:
ARM64.HasASIMDDP = true
}
// ID_AA64ISAR1_EL1
switch extractBits(isar1, 0, 3) {
case 1:
ARM64.HasDCPOP = true
}
switch extractBits(isar1, 12, 15) {
case 1:
ARM64.HasJSCVT = true
}
switch extractBits(isar1, 16, 19) {
case 1:
ARM64.HasFCMA = true
}
switch extractBits(isar1, 20, 23) {
case 1:
ARM64.HasLRCPC = true
}
// ID_AA64PFR0_EL1
switch extractBits(pfr0, 16, 19) {
case 0:
ARM64.HasFP = true
case 1:
ARM64.HasFP = true
ARM64.HasFPHP = true
}
switch extractBits(pfr0, 20, 23) {
case 0:
ARM64.HasASIMD = true
case 1:
ARM64.HasASIMD = true
ARM64.HasASIMDHP = true
}
switch extractBits(pfr0, 32, 35) {
case 1:
ARM64.HasSVE = true
}
}
func extractBits(data uint64, start, end uint) uint {
return (uint)(data>>start) & ((1 << (end - start + 1)) - 1)
}

32
gateway/vendor/golang.org/x/sys/cpu/cpu_arm64.s generated vendored Normal file
View File

@ -0,0 +1,32 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc
// +build gc
#include "textflag.h"
// func getisar0() uint64
TEXT ·getisar0(SB),NOSPLIT,$0-8
// get Instruction Set Attributes 0 into x0
// mrs x0, ID_AA64ISAR0_EL1 = d5380600
WORD $0xd5380600
MOVD R0, ret+0(FP)
RET
// func getisar1() uint64
TEXT ·getisar1(SB),NOSPLIT,$0-8
// get Instruction Set Attributes 1 into x0
// mrs x0, ID_AA64ISAR1_EL1 = d5380620
WORD $0xd5380620
MOVD R0, ret+0(FP)
RET
// func getpfr0() uint64
TEXT ·getpfr0(SB),NOSPLIT,$0-8
// get Processor Feature Register 0 into x0
// mrs x0, ID_AA64PFR0_EL1 = d5380400
WORD $0xd5380400
MOVD R0, ret+0(FP)
RET

12
gateway/vendor/golang.org/x/sys/cpu/cpu_gc_arm64.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc
// +build gc
package cpu
func getisar0() uint64
func getisar1() uint64
func getpfr0() uint64

22
gateway/vendor/golang.org/x/sys/cpu/cpu_gc_s390x.go generated vendored Normal file
View File

@ -0,0 +1,22 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc
// +build gc
package cpu
// haveAsmFunctions reports whether the other functions in this file can
// be safely called.
func haveAsmFunctions() bool { return true }
// The following feature detection functions are defined in cpu_s390x.s.
// They are likely to be expensive to call so the results should be cached.
func stfle() facilityList
func kmQuery() queryResult
func kmcQuery() queryResult
func kmctrQuery() queryResult
func kmaQuery() queryResult
func kimdQuery() queryResult
func klmdQuery() queryResult

17
gateway/vendor/golang.org/x/sys/cpu/cpu_gc_x86.go generated vendored Normal file
View File

@ -0,0 +1,17 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build (386 || amd64 || amd64p32) && gc
// +build 386 amd64 amd64p32
// +build gc
package cpu
// cpuid is implemented in cpu_x86.s for gc compiler
// and in cpu_gccgo.c for gccgo.
func cpuid(eaxArg, ecxArg uint32) (eax, ebx, ecx, edx uint32)
// xgetbv with ecx = 0 is implemented in cpu_x86.s for gc compiler
// and in cpu_gccgo.c for gccgo.
func xgetbv() (eax, edx uint32)

12
gateway/vendor/golang.org/x/sys/cpu/cpu_gccgo_arm64.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gccgo
// +build gccgo
package cpu
func getisar0() uint64 { return 0 }
func getisar1() uint64 { return 0 }
func getpfr0() uint64 { return 0 }

23
gateway/vendor/golang.org/x/sys/cpu/cpu_gccgo_s390x.go generated vendored Normal file
View File

@ -0,0 +1,23 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gccgo
// +build gccgo
package cpu
// haveAsmFunctions reports whether the other functions in this file can
// be safely called.
func haveAsmFunctions() bool { return false }
// TODO(mundaym): the following feature detection functions are currently
// stubs. See https://golang.org/cl/162887 for how to fix this.
// They are likely to be expensive to call so the results should be cached.
func stfle() facilityList { panic("not implemented for gccgo") }
func kmQuery() queryResult { panic("not implemented for gccgo") }
func kmcQuery() queryResult { panic("not implemented for gccgo") }
func kmctrQuery() queryResult { panic("not implemented for gccgo") }
func kmaQuery() queryResult { panic("not implemented for gccgo") }
func kimdQuery() queryResult { panic("not implemented for gccgo") }
func klmdQuery() queryResult { panic("not implemented for gccgo") }

39
gateway/vendor/golang.org/x/sys/cpu/cpu_gccgo_x86.c generated vendored Normal file
View File

@ -0,0 +1,39 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build (386 || amd64 || amd64p32) && gccgo
// +build 386 amd64 amd64p32
// +build gccgo
#include <cpuid.h>
#include <stdint.h>
#include <x86intrin.h>
// Need to wrap __get_cpuid_count because it's declared as static.
int
gccgoGetCpuidCount(uint32_t leaf, uint32_t subleaf,
uint32_t *eax, uint32_t *ebx,
uint32_t *ecx, uint32_t *edx)
{
return __get_cpuid_count(leaf, subleaf, eax, ebx, ecx, edx);
}
#pragma GCC diagnostic ignored "-Wunknown-pragmas"
#pragma GCC push_options
#pragma GCC target("xsave")
#pragma clang attribute push (__attribute__((target("xsave"))), apply_to=function)
// xgetbv reads the contents of an XCR (Extended Control Register)
// specified in the ECX register into registers EDX:EAX.
// Currently, the only supported value for XCR is 0.
void
gccgoXgetbv(uint32_t *eax, uint32_t *edx)
{
uint64_t v = _xgetbv(0);
*eax = v & 0xffffffff;
*edx = v >> 32;
}
#pragma clang attribute pop
#pragma GCC pop_options

33
gateway/vendor/golang.org/x/sys/cpu/cpu_gccgo_x86.go generated vendored Normal file
View File

@ -0,0 +1,33 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build (386 || amd64 || amd64p32) && gccgo
// +build 386 amd64 amd64p32
// +build gccgo
package cpu
//extern gccgoGetCpuidCount
func gccgoGetCpuidCount(eaxArg, ecxArg uint32, eax, ebx, ecx, edx *uint32)
func cpuid(eaxArg, ecxArg uint32) (eax, ebx, ecx, edx uint32) {
var a, b, c, d uint32
gccgoGetCpuidCount(eaxArg, ecxArg, &a, &b, &c, &d)
return a, b, c, d
}
//extern gccgoXgetbv
func gccgoXgetbv(eax, edx *uint32)
func xgetbv() (eax, edx uint32) {
var a, d uint32
gccgoXgetbv(&a, &d)
return a, d
}
// gccgo doesn't build on Darwin, per:
// https://github.com/Homebrew/homebrew-core/blob/HEAD/Formula/gcc.rb#L76
func darwinSupportsAVX512() bool {
return false
}

16
gateway/vendor/golang.org/x/sys/cpu/cpu_linux.go generated vendored Normal file
View File

@ -0,0 +1,16 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !386 && !amd64 && !amd64p32 && !arm64
// +build !386,!amd64,!amd64p32,!arm64
package cpu
func archInit() {
if err := readHWCAP(); err != nil {
return
}
doinit()
Initialized = true
}

39
gateway/vendor/golang.org/x/sys/cpu/cpu_linux_arm.go generated vendored Normal file
View File

@ -0,0 +1,39 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
func doinit() {
ARM.HasSWP = isSet(hwCap, hwcap_SWP)
ARM.HasHALF = isSet(hwCap, hwcap_HALF)
ARM.HasTHUMB = isSet(hwCap, hwcap_THUMB)
ARM.Has26BIT = isSet(hwCap, hwcap_26BIT)
ARM.HasFASTMUL = isSet(hwCap, hwcap_FAST_MULT)
ARM.HasFPA = isSet(hwCap, hwcap_FPA)
ARM.HasVFP = isSet(hwCap, hwcap_VFP)
ARM.HasEDSP = isSet(hwCap, hwcap_EDSP)
ARM.HasJAVA = isSet(hwCap, hwcap_JAVA)
ARM.HasIWMMXT = isSet(hwCap, hwcap_IWMMXT)
ARM.HasCRUNCH = isSet(hwCap, hwcap_CRUNCH)
ARM.HasTHUMBEE = isSet(hwCap, hwcap_THUMBEE)
ARM.HasNEON = isSet(hwCap, hwcap_NEON)
ARM.HasVFPv3 = isSet(hwCap, hwcap_VFPv3)
ARM.HasVFPv3D16 = isSet(hwCap, hwcap_VFPv3D16)
ARM.HasTLS = isSet(hwCap, hwcap_TLS)
ARM.HasVFPv4 = isSet(hwCap, hwcap_VFPv4)
ARM.HasIDIVA = isSet(hwCap, hwcap_IDIVA)
ARM.HasIDIVT = isSet(hwCap, hwcap_IDIVT)
ARM.HasVFPD32 = isSet(hwCap, hwcap_VFPD32)
ARM.HasLPAE = isSet(hwCap, hwcap_LPAE)
ARM.HasEVTSTRM = isSet(hwCap, hwcap_EVTSTRM)
ARM.HasAES = isSet(hwCap2, hwcap2_AES)
ARM.HasPMULL = isSet(hwCap2, hwcap2_PMULL)
ARM.HasSHA1 = isSet(hwCap2, hwcap2_SHA1)
ARM.HasSHA2 = isSet(hwCap2, hwcap2_SHA2)
ARM.HasCRC32 = isSet(hwCap2, hwcap2_CRC32)
}
func isSet(hwc uint, value uint) bool {
return hwc&value != 0
}

111
gateway/vendor/golang.org/x/sys/cpu/cpu_linux_arm64.go generated vendored Normal file
View File

@ -0,0 +1,111 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import (
"strings"
"syscall"
)
// HWCAP/HWCAP2 bits. These are exposed by Linux.
const (
hwcap_FP = 1 << 0
hwcap_ASIMD = 1 << 1
hwcap_EVTSTRM = 1 << 2
hwcap_AES = 1 << 3
hwcap_PMULL = 1 << 4
hwcap_SHA1 = 1 << 5
hwcap_SHA2 = 1 << 6
hwcap_CRC32 = 1 << 7
hwcap_ATOMICS = 1 << 8
hwcap_FPHP = 1 << 9
hwcap_ASIMDHP = 1 << 10
hwcap_CPUID = 1 << 11
hwcap_ASIMDRDM = 1 << 12
hwcap_JSCVT = 1 << 13
hwcap_FCMA = 1 << 14
hwcap_LRCPC = 1 << 15
hwcap_DCPOP = 1 << 16
hwcap_SHA3 = 1 << 17
hwcap_SM3 = 1 << 18
hwcap_SM4 = 1 << 19
hwcap_ASIMDDP = 1 << 20
hwcap_SHA512 = 1 << 21
hwcap_SVE = 1 << 22
hwcap_ASIMDFHM = 1 << 23
)
// linuxKernelCanEmulateCPUID reports whether we're running
// on Linux 4.11+. Ideally we'd like to ask the question about
// whether the current kernel contains
// https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=77c97b4ee21290f5f083173d957843b615abbff2
// but the version number will have to do.
func linuxKernelCanEmulateCPUID() bool {
var un syscall.Utsname
syscall.Uname(&un)
var sb strings.Builder
for _, b := range un.Release[:] {
if b == 0 {
break
}
sb.WriteByte(byte(b))
}
major, minor, _, ok := parseRelease(sb.String())
return ok && (major > 4 || major == 4 && minor >= 11)
}
func doinit() {
if err := readHWCAP(); err != nil {
// We failed to read /proc/self/auxv. This can happen if the binary has
// been given extra capabilities(7) with /bin/setcap.
//
// When this happens, we have two options. If the Linux kernel is new
// enough (4.11+), we can read the arm64 registers directly which'll
// trap into the kernel and then return back to userspace.
//
// But on older kernels, such as Linux 4.4.180 as used on many Synology
// devices, calling readARM64Registers (specifically getisar0) will
// cause a SIGILL and we'll die. So for older kernels, parse /proc/cpuinfo
// instead.
//
// See golang/go#57336.
if linuxKernelCanEmulateCPUID() {
readARM64Registers()
} else {
readLinuxProcCPUInfo()
}
return
}
// HWCAP feature bits
ARM64.HasFP = isSet(hwCap, hwcap_FP)
ARM64.HasASIMD = isSet(hwCap, hwcap_ASIMD)
ARM64.HasEVTSTRM = isSet(hwCap, hwcap_EVTSTRM)
ARM64.HasAES = isSet(hwCap, hwcap_AES)
ARM64.HasPMULL = isSet(hwCap, hwcap_PMULL)
ARM64.HasSHA1 = isSet(hwCap, hwcap_SHA1)
ARM64.HasSHA2 = isSet(hwCap, hwcap_SHA2)
ARM64.HasCRC32 = isSet(hwCap, hwcap_CRC32)
ARM64.HasATOMICS = isSet(hwCap, hwcap_ATOMICS)
ARM64.HasFPHP = isSet(hwCap, hwcap_FPHP)
ARM64.HasASIMDHP = isSet(hwCap, hwcap_ASIMDHP)
ARM64.HasCPUID = isSet(hwCap, hwcap_CPUID)
ARM64.HasASIMDRDM = isSet(hwCap, hwcap_ASIMDRDM)
ARM64.HasJSCVT = isSet(hwCap, hwcap_JSCVT)
ARM64.HasFCMA = isSet(hwCap, hwcap_FCMA)
ARM64.HasLRCPC = isSet(hwCap, hwcap_LRCPC)
ARM64.HasDCPOP = isSet(hwCap, hwcap_DCPOP)
ARM64.HasSHA3 = isSet(hwCap, hwcap_SHA3)
ARM64.HasSM3 = isSet(hwCap, hwcap_SM3)
ARM64.HasSM4 = isSet(hwCap, hwcap_SM4)
ARM64.HasASIMDDP = isSet(hwCap, hwcap_ASIMDDP)
ARM64.HasSHA512 = isSet(hwCap, hwcap_SHA512)
ARM64.HasSVE = isSet(hwCap, hwcap_SVE)
ARM64.HasASIMDFHM = isSet(hwCap, hwcap_ASIMDFHM)
}
func isSet(hwc uint, value uint) bool {
return hwc&value != 0
}

24
gateway/vendor/golang.org/x/sys/cpu/cpu_linux_mips64x.go generated vendored Normal file
View File

@ -0,0 +1,24 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build linux && (mips64 || mips64le)
// +build linux
// +build mips64 mips64le
package cpu
// HWCAP bits. These are exposed by the Linux kernel 5.4.
const (
// CPU features
hwcap_MIPS_MSA = 1 << 1
)
func doinit() {
// HWCAP feature bits
MIPS64X.HasMSA = isSet(hwCap, hwcap_MIPS_MSA)
}
func isSet(hwc uint, value uint) bool {
return hwc&value != 0
}

10
gateway/vendor/golang.org/x/sys/cpu/cpu_linux_noinit.go generated vendored Normal file
View File

@ -0,0 +1,10 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build linux && !arm && !arm64 && !mips64 && !mips64le && !ppc64 && !ppc64le && !s390x
// +build linux,!arm,!arm64,!mips64,!mips64le,!ppc64,!ppc64le,!s390x
package cpu
func doinit() {}

32
gateway/vendor/golang.org/x/sys/cpu/cpu_linux_ppc64x.go generated vendored Normal file
View File

@ -0,0 +1,32 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build linux && (ppc64 || ppc64le)
// +build linux
// +build ppc64 ppc64le
package cpu
// HWCAP/HWCAP2 bits. These are exposed by the kernel.
const (
// ISA Level
_PPC_FEATURE2_ARCH_2_07 = 0x80000000
_PPC_FEATURE2_ARCH_3_00 = 0x00800000
// CPU features
_PPC_FEATURE2_DARN = 0x00200000
_PPC_FEATURE2_SCV = 0x00100000
)
func doinit() {
// HWCAP2 feature bits
PPC64.IsPOWER8 = isSet(hwCap2, _PPC_FEATURE2_ARCH_2_07)
PPC64.IsPOWER9 = isSet(hwCap2, _PPC_FEATURE2_ARCH_3_00)
PPC64.HasDARN = isSet(hwCap2, _PPC_FEATURE2_DARN)
PPC64.HasSCV = isSet(hwCap2, _PPC_FEATURE2_SCV)
}
func isSet(hwc uint, value uint) bool {
return hwc&value != 0
}

40
gateway/vendor/golang.org/x/sys/cpu/cpu_linux_s390x.go generated vendored Normal file
View File

@ -0,0 +1,40 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
const (
// bit mask values from /usr/include/bits/hwcap.h
hwcap_ZARCH = 2
hwcap_STFLE = 4
hwcap_MSA = 8
hwcap_LDISP = 16
hwcap_EIMM = 32
hwcap_DFP = 64
hwcap_ETF3EH = 256
hwcap_VX = 2048
hwcap_VXE = 8192
)
func initS390Xbase() {
// test HWCAP bit vector
has := func(featureMask uint) bool {
return hwCap&featureMask == featureMask
}
// mandatory
S390X.HasZARCH = has(hwcap_ZARCH)
// optional
S390X.HasSTFLE = has(hwcap_STFLE)
S390X.HasLDISP = has(hwcap_LDISP)
S390X.HasEIMM = has(hwcap_EIMM)
S390X.HasETF3EH = has(hwcap_ETF3EH)
S390X.HasDFP = has(hwcap_DFP)
S390X.HasMSA = has(hwcap_MSA)
S390X.HasVX = has(hwcap_VX)
if S390X.HasVX {
S390X.HasVXE = has(hwcap_VXE)
}
}

13
gateway/vendor/golang.org/x/sys/cpu/cpu_loong64.go generated vendored Normal file
View File

@ -0,0 +1,13 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build loong64
// +build loong64
package cpu
const cacheLineSize = 64
func initOptions() {
}

16
gateway/vendor/golang.org/x/sys/cpu/cpu_mips64x.go generated vendored Normal file
View File

@ -0,0 +1,16 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build mips64 || mips64le
// +build mips64 mips64le
package cpu
const cacheLineSize = 32
func initOptions() {
options = []option{
{Name: "msa", Feature: &MIPS64X.HasMSA},
}
}

12
gateway/vendor/golang.org/x/sys/cpu/cpu_mipsx.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build mips || mipsle
// +build mips mipsle
package cpu
const cacheLineSize = 32
func initOptions() {}

173
gateway/vendor/golang.org/x/sys/cpu/cpu_netbsd_arm64.go generated vendored Normal file
View File

@ -0,0 +1,173 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import (
"syscall"
"unsafe"
)
// Minimal copy of functionality from x/sys/unix so the cpu package can call
// sysctl without depending on x/sys/unix.
const (
_CTL_QUERY = -2
_SYSCTL_VERS_1 = 0x1000000
)
var _zero uintptr
func sysctl(mib []int32, old *byte, oldlen *uintptr, new *byte, newlen uintptr) (err error) {
var _p0 unsafe.Pointer
if len(mib) > 0 {
_p0 = unsafe.Pointer(&mib[0])
} else {
_p0 = unsafe.Pointer(&_zero)
}
_, _, errno := syscall.Syscall6(
syscall.SYS___SYSCTL,
uintptr(_p0),
uintptr(len(mib)),
uintptr(unsafe.Pointer(old)),
uintptr(unsafe.Pointer(oldlen)),
uintptr(unsafe.Pointer(new)),
uintptr(newlen))
if errno != 0 {
return errno
}
return nil
}
type sysctlNode struct {
Flags uint32
Num int32
Name [32]int8
Ver uint32
__rsvd uint32
Un [16]byte
_sysctl_size [8]byte
_sysctl_func [8]byte
_sysctl_parent [8]byte
_sysctl_desc [8]byte
}
func sysctlNodes(mib []int32) ([]sysctlNode, error) {
var olen uintptr
// Get a list of all sysctl nodes below the given MIB by performing
// a sysctl for the given MIB with CTL_QUERY appended.
mib = append(mib, _CTL_QUERY)
qnode := sysctlNode{Flags: _SYSCTL_VERS_1}
qp := (*byte)(unsafe.Pointer(&qnode))
sz := unsafe.Sizeof(qnode)
if err := sysctl(mib, nil, &olen, qp, sz); err != nil {
return nil, err
}
// Now that we know the size, get the actual nodes.
nodes := make([]sysctlNode, olen/sz)
np := (*byte)(unsafe.Pointer(&nodes[0]))
if err := sysctl(mib, np, &olen, qp, sz); err != nil {
return nil, err
}
return nodes, nil
}
func nametomib(name string) ([]int32, error) {
// Split name into components.
var parts []string
last := 0
for i := 0; i < len(name); i++ {
if name[i] == '.' {
parts = append(parts, name[last:i])
last = i + 1
}
}
parts = append(parts, name[last:])
mib := []int32{}
// Discover the nodes and construct the MIB OID.
for partno, part := range parts {
nodes, err := sysctlNodes(mib)
if err != nil {
return nil, err
}
for _, node := range nodes {
n := make([]byte, 0)
for i := range node.Name {
if node.Name[i] != 0 {
n = append(n, byte(node.Name[i]))
}
}
if string(n) == part {
mib = append(mib, int32(node.Num))
break
}
}
if len(mib) != partno+1 {
return nil, err
}
}
return mib, nil
}
// aarch64SysctlCPUID is struct aarch64_sysctl_cpu_id from NetBSD's <aarch64/armreg.h>
type aarch64SysctlCPUID struct {
midr uint64 /* Main ID Register */
revidr uint64 /* Revision ID Register */
mpidr uint64 /* Multiprocessor Affinity Register */
aa64dfr0 uint64 /* A64 Debug Feature Register 0 */
aa64dfr1 uint64 /* A64 Debug Feature Register 1 */
aa64isar0 uint64 /* A64 Instruction Set Attribute Register 0 */
aa64isar1 uint64 /* A64 Instruction Set Attribute Register 1 */
aa64mmfr0 uint64 /* A64 Memory Model Feature Register 0 */
aa64mmfr1 uint64 /* A64 Memory Model Feature Register 1 */
aa64mmfr2 uint64 /* A64 Memory Model Feature Register 2 */
aa64pfr0 uint64 /* A64 Processor Feature Register 0 */
aa64pfr1 uint64 /* A64 Processor Feature Register 1 */
aa64zfr0 uint64 /* A64 SVE Feature ID Register 0 */
mvfr0 uint32 /* Media and VFP Feature Register 0 */
mvfr1 uint32 /* Media and VFP Feature Register 1 */
mvfr2 uint32 /* Media and VFP Feature Register 2 */
pad uint32
clidr uint64 /* Cache Level ID Register */
ctr uint64 /* Cache Type Register */
}
func sysctlCPUID(name string) (*aarch64SysctlCPUID, error) {
mib, err := nametomib(name)
if err != nil {
return nil, err
}
out := aarch64SysctlCPUID{}
n := unsafe.Sizeof(out)
_, _, errno := syscall.Syscall6(
syscall.SYS___SYSCTL,
uintptr(unsafe.Pointer(&mib[0])),
uintptr(len(mib)),
uintptr(unsafe.Pointer(&out)),
uintptr(unsafe.Pointer(&n)),
uintptr(0),
uintptr(0))
if errno != 0 {
return nil, errno
}
return &out, nil
}
func doinit() {
cpuid, err := sysctlCPUID("machdep.cpu0.cpu_id")
if err != nil {
setMinimalFeatures()
return
}
parseARM64SystemRegisters(cpuid.aa64isar0, cpuid.aa64isar1, cpuid.aa64pfr0)
Initialized = true
}

65
gateway/vendor/golang.org/x/sys/cpu/cpu_openbsd_arm64.go generated vendored Normal file
View File

@ -0,0 +1,65 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import (
"syscall"
"unsafe"
)
// Minimal copy of functionality from x/sys/unix so the cpu package can call
// sysctl without depending on x/sys/unix.
const (
// From OpenBSD's sys/sysctl.h.
_CTL_MACHDEP = 7
// From OpenBSD's machine/cpu.h.
_CPU_ID_AA64ISAR0 = 2
_CPU_ID_AA64ISAR1 = 3
)
// Implemented in the runtime package (runtime/sys_openbsd3.go)
func syscall_syscall6(fn, a1, a2, a3, a4, a5, a6 uintptr) (r1, r2 uintptr, err syscall.Errno)
//go:linkname syscall_syscall6 syscall.syscall6
func sysctl(mib []uint32, old *byte, oldlen *uintptr, new *byte, newlen uintptr) (err error) {
_, _, errno := syscall_syscall6(libc_sysctl_trampoline_addr, uintptr(unsafe.Pointer(&mib[0])), uintptr(len(mib)), uintptr(unsafe.Pointer(old)), uintptr(unsafe.Pointer(oldlen)), uintptr(unsafe.Pointer(new)), uintptr(newlen))
if errno != 0 {
return errno
}
return nil
}
var libc_sysctl_trampoline_addr uintptr
//go:cgo_import_dynamic libc_sysctl sysctl "libc.so"
func sysctlUint64(mib []uint32) (uint64, bool) {
var out uint64
nout := unsafe.Sizeof(out)
if err := sysctl(mib, (*byte)(unsafe.Pointer(&out)), &nout, nil, 0); err != nil {
return 0, false
}
return out, true
}
func doinit() {
setMinimalFeatures()
// Get ID_AA64ISAR0 and ID_AA64ISAR1 from sysctl.
isar0, ok := sysctlUint64([]uint32{_CTL_MACHDEP, _CPU_ID_AA64ISAR0})
if !ok {
return
}
isar1, ok := sysctlUint64([]uint32{_CTL_MACHDEP, _CPU_ID_AA64ISAR1})
if !ok {
return
}
parseARM64SystemRegisters(isar0, isar1, 0)
Initialized = true
}

11
gateway/vendor/golang.org/x/sys/cpu/cpu_openbsd_arm64.s generated vendored Normal file
View File

@ -0,0 +1,11 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
#include "textflag.h"
TEXT libc_sysctl_trampoline<>(SB),NOSPLIT,$0-0
JMP libc_sysctl(SB)
GLOBL ·libc_sysctl_trampoline_addr(SB), RODATA, $8
DATA ·libc_sysctl_trampoline_addr(SB)/8, $libc_sysctl_trampoline<>(SB)

10
gateway/vendor/golang.org/x/sys/cpu/cpu_other_arm.go generated vendored Normal file
View File

@ -0,0 +1,10 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !linux && arm
// +build !linux,arm
package cpu
func archInit() {}

10
gateway/vendor/golang.org/x/sys/cpu/cpu_other_arm64.go generated vendored Normal file
View File

@ -0,0 +1,10 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !linux && !netbsd && !openbsd && arm64
// +build !linux,!netbsd,!openbsd,arm64
package cpu
func doinit() {}

13
gateway/vendor/golang.org/x/sys/cpu/cpu_other_mips64x.go generated vendored Normal file
View File

@ -0,0 +1,13 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !linux && (mips64 || mips64le)
// +build !linux
// +build mips64 mips64le
package cpu
func archInit() {
Initialized = true
}

15
gateway/vendor/golang.org/x/sys/cpu/cpu_other_ppc64x.go generated vendored Normal file
View File

@ -0,0 +1,15 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !aix && !linux && (ppc64 || ppc64le)
// +build !aix
// +build !linux
// +build ppc64 ppc64le
package cpu
func archInit() {
PPC64.IsPOWER8 = true
Initialized = true
}

12
gateway/vendor/golang.org/x/sys/cpu/cpu_other_riscv64.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build !linux && riscv64
// +build !linux,riscv64
package cpu
func archInit() {
Initialized = true
}

17
gateway/vendor/golang.org/x/sys/cpu/cpu_ppc64x.go generated vendored Normal file
View File

@ -0,0 +1,17 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build ppc64 || ppc64le
// +build ppc64 ppc64le
package cpu
const cacheLineSize = 128
func initOptions() {
options = []option{
{Name: "darn", Feature: &PPC64.HasDARN},
{Name: "scv", Feature: &PPC64.HasSCV},
}
}

12
gateway/vendor/golang.org/x/sys/cpu/cpu_riscv64.go generated vendored Normal file
View File

@ -0,0 +1,12 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build riscv64
// +build riscv64
package cpu
const cacheLineSize = 64
func initOptions() {}

172
gateway/vendor/golang.org/x/sys/cpu/cpu_s390x.go generated vendored Normal file
View File

@ -0,0 +1,172 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
const cacheLineSize = 256
func initOptions() {
options = []option{
{Name: "zarch", Feature: &S390X.HasZARCH, Required: true},
{Name: "stfle", Feature: &S390X.HasSTFLE, Required: true},
{Name: "ldisp", Feature: &S390X.HasLDISP, Required: true},
{Name: "eimm", Feature: &S390X.HasEIMM, Required: true},
{Name: "dfp", Feature: &S390X.HasDFP},
{Name: "etf3eh", Feature: &S390X.HasETF3EH},
{Name: "msa", Feature: &S390X.HasMSA},
{Name: "aes", Feature: &S390X.HasAES},
{Name: "aescbc", Feature: &S390X.HasAESCBC},
{Name: "aesctr", Feature: &S390X.HasAESCTR},
{Name: "aesgcm", Feature: &S390X.HasAESGCM},
{Name: "ghash", Feature: &S390X.HasGHASH},
{Name: "sha1", Feature: &S390X.HasSHA1},
{Name: "sha256", Feature: &S390X.HasSHA256},
{Name: "sha3", Feature: &S390X.HasSHA3},
{Name: "sha512", Feature: &S390X.HasSHA512},
{Name: "vx", Feature: &S390X.HasVX},
{Name: "vxe", Feature: &S390X.HasVXE},
}
}
// bitIsSet reports whether the bit at index is set. The bit index
// is in big endian order, so bit index 0 is the leftmost bit.
func bitIsSet(bits []uint64, index uint) bool {
return bits[index/64]&((1<<63)>>(index%64)) != 0
}
// facility is a bit index for the named facility.
type facility uint8
const (
// mandatory facilities
zarch facility = 1 // z architecture mode is active
stflef facility = 7 // store-facility-list-extended
ldisp facility = 18 // long-displacement
eimm facility = 21 // extended-immediate
// miscellaneous facilities
dfp facility = 42 // decimal-floating-point
etf3eh facility = 30 // extended-translation 3 enhancement
// cryptography facilities
msa facility = 17 // message-security-assist
msa3 facility = 76 // message-security-assist extension 3
msa4 facility = 77 // message-security-assist extension 4
msa5 facility = 57 // message-security-assist extension 5
msa8 facility = 146 // message-security-assist extension 8
msa9 facility = 155 // message-security-assist extension 9
// vector facilities
vx facility = 129 // vector facility
vxe facility = 135 // vector-enhancements 1
vxe2 facility = 148 // vector-enhancements 2
)
// facilityList contains the result of an STFLE call.
// Bits are numbered in big endian order so the
// leftmost bit (the MSB) is at index 0.
type facilityList struct {
bits [4]uint64
}
// Has reports whether the given facilities are present.
func (s *facilityList) Has(fs ...facility) bool {
if len(fs) == 0 {
panic("no facility bits provided")
}
for _, f := range fs {
if !bitIsSet(s.bits[:], uint(f)) {
return false
}
}
return true
}
// function is the code for the named cryptographic function.
type function uint8
const (
// KM{,A,C,CTR} function codes
aes128 function = 18 // AES-128
aes192 function = 19 // AES-192
aes256 function = 20 // AES-256
// K{I,L}MD function codes
sha1 function = 1 // SHA-1
sha256 function = 2 // SHA-256
sha512 function = 3 // SHA-512
sha3_224 function = 32 // SHA3-224
sha3_256 function = 33 // SHA3-256
sha3_384 function = 34 // SHA3-384
sha3_512 function = 35 // SHA3-512
shake128 function = 36 // SHAKE-128
shake256 function = 37 // SHAKE-256
// KLMD function codes
ghash function = 65 // GHASH
)
// queryResult contains the result of a Query function
// call. Bits are numbered in big endian order so the
// leftmost bit (the MSB) is at index 0.
type queryResult struct {
bits [2]uint64
}
// Has reports whether the given functions are present.
func (q *queryResult) Has(fns ...function) bool {
if len(fns) == 0 {
panic("no function codes provided")
}
for _, f := range fns {
if !bitIsSet(q.bits[:], uint(f)) {
return false
}
}
return true
}
func doinit() {
initS390Xbase()
// We need implementations of stfle, km and so on
// to detect cryptographic features.
if !haveAsmFunctions() {
return
}
// optional cryptographic functions
if S390X.HasMSA {
aes := []function{aes128, aes192, aes256}
// cipher message
km, kmc := kmQuery(), kmcQuery()
S390X.HasAES = km.Has(aes...)
S390X.HasAESCBC = kmc.Has(aes...)
if S390X.HasSTFLE {
facilities := stfle()
if facilities.Has(msa4) {
kmctr := kmctrQuery()
S390X.HasAESCTR = kmctr.Has(aes...)
}
if facilities.Has(msa8) {
kma := kmaQuery()
S390X.HasAESGCM = kma.Has(aes...)
}
}
// compute message digest
kimd := kimdQuery() // intermediate (no padding)
klmd := klmdQuery() // last (padding)
S390X.HasSHA1 = kimd.Has(sha1) && klmd.Has(sha1)
S390X.HasSHA256 = kimd.Has(sha256) && klmd.Has(sha256)
S390X.HasSHA512 = kimd.Has(sha512) && klmd.Has(sha512)
S390X.HasGHASH = kimd.Has(ghash) // KLMD-GHASH does not exist
sha3 := []function{
sha3_224, sha3_256, sha3_384, sha3_512,
shake128, shake256,
}
S390X.HasSHA3 = kimd.Has(sha3...) && klmd.Has(sha3...)
}
}

58
gateway/vendor/golang.org/x/sys/cpu/cpu_s390x.s generated vendored Normal file
View File

@ -0,0 +1,58 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build gc
// +build gc
#include "textflag.h"
// func stfle() facilityList
TEXT ·stfle(SB), NOSPLIT|NOFRAME, $0-32
MOVD $ret+0(FP), R1
MOVD $3, R0 // last doubleword index to store
XC $32, (R1), (R1) // clear 4 doublewords (32 bytes)
WORD $0xb2b01000 // store facility list extended (STFLE)
RET
// func kmQuery() queryResult
TEXT ·kmQuery(SB), NOSPLIT|NOFRAME, $0-16
MOVD $0, R0 // set function code to 0 (KM-Query)
MOVD $ret+0(FP), R1 // address of 16-byte return value
WORD $0xB92E0024 // cipher message (KM)
RET
// func kmcQuery() queryResult
TEXT ·kmcQuery(SB), NOSPLIT|NOFRAME, $0-16
MOVD $0, R0 // set function code to 0 (KMC-Query)
MOVD $ret+0(FP), R1 // address of 16-byte return value
WORD $0xB92F0024 // cipher message with chaining (KMC)
RET
// func kmctrQuery() queryResult
TEXT ·kmctrQuery(SB), NOSPLIT|NOFRAME, $0-16
MOVD $0, R0 // set function code to 0 (KMCTR-Query)
MOVD $ret+0(FP), R1 // address of 16-byte return value
WORD $0xB92D4024 // cipher message with counter (KMCTR)
RET
// func kmaQuery() queryResult
TEXT ·kmaQuery(SB), NOSPLIT|NOFRAME, $0-16
MOVD $0, R0 // set function code to 0 (KMA-Query)
MOVD $ret+0(FP), R1 // address of 16-byte return value
WORD $0xb9296024 // cipher message with authentication (KMA)
RET
// func kimdQuery() queryResult
TEXT ·kimdQuery(SB), NOSPLIT|NOFRAME, $0-16
MOVD $0, R0 // set function code to 0 (KIMD-Query)
MOVD $ret+0(FP), R1 // address of 16-byte return value
WORD $0xB93E0024 // compute intermediate message digest (KIMD)
RET
// func klmdQuery() queryResult
TEXT ·klmdQuery(SB), NOSPLIT|NOFRAME, $0-16
MOVD $0, R0 // set function code to 0 (KLMD-Query)
MOVD $ret+0(FP), R1 // address of 16-byte return value
WORD $0xB93F0024 // compute last message digest (KLMD)
RET

18
gateway/vendor/golang.org/x/sys/cpu/cpu_wasm.go generated vendored Normal file
View File

@ -0,0 +1,18 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build wasm
// +build wasm
package cpu
// We're compiling the cpu package for an unknown (software-abstracted) CPU.
// Make CacheLinePad an empty struct and hope that the usual struct alignment
// rules are good enough.
const cacheLineSize = 0
func initOptions() {}
func archInit() {}

152
gateway/vendor/golang.org/x/sys/cpu/cpu_x86.go generated vendored Normal file
View File

@ -0,0 +1,152 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build 386 || amd64 || amd64p32
// +build 386 amd64 amd64p32
package cpu
import "runtime"
const cacheLineSize = 64
func initOptions() {
options = []option{
{Name: "adx", Feature: &X86.HasADX},
{Name: "aes", Feature: &X86.HasAES},
{Name: "avx", Feature: &X86.HasAVX},
{Name: "avx2", Feature: &X86.HasAVX2},
{Name: "avx512", Feature: &X86.HasAVX512},
{Name: "avx512f", Feature: &X86.HasAVX512F},
{Name: "avx512cd", Feature: &X86.HasAVX512CD},
{Name: "avx512er", Feature: &X86.HasAVX512ER},
{Name: "avx512pf", Feature: &X86.HasAVX512PF},
{Name: "avx512vl", Feature: &X86.HasAVX512VL},
{Name: "avx512bw", Feature: &X86.HasAVX512BW},
{Name: "avx512dq", Feature: &X86.HasAVX512DQ},
{Name: "avx512ifma", Feature: &X86.HasAVX512IFMA},
{Name: "avx512vbmi", Feature: &X86.HasAVX512VBMI},
{Name: "avx512vnniw", Feature: &X86.HasAVX5124VNNIW},
{Name: "avx5124fmaps", Feature: &X86.HasAVX5124FMAPS},
{Name: "avx512vpopcntdq", Feature: &X86.HasAVX512VPOPCNTDQ},
{Name: "avx512vpclmulqdq", Feature: &X86.HasAVX512VPCLMULQDQ},
{Name: "avx512vnni", Feature: &X86.HasAVX512VNNI},
{Name: "avx512gfni", Feature: &X86.HasAVX512GFNI},
{Name: "avx512vaes", Feature: &X86.HasAVX512VAES},
{Name: "avx512vbmi2", Feature: &X86.HasAVX512VBMI2},
{Name: "avx512bitalg", Feature: &X86.HasAVX512BITALG},
{Name: "avx512bf16", Feature: &X86.HasAVX512BF16},
{Name: "amxtile", Feature: &X86.HasAMXTile},
{Name: "amxint8", Feature: &X86.HasAMXInt8},
{Name: "amxbf16", Feature: &X86.HasAMXBF16},
{Name: "bmi1", Feature: &X86.HasBMI1},
{Name: "bmi2", Feature: &X86.HasBMI2},
{Name: "cx16", Feature: &X86.HasCX16},
{Name: "erms", Feature: &X86.HasERMS},
{Name: "fma", Feature: &X86.HasFMA},
{Name: "osxsave", Feature: &X86.HasOSXSAVE},
{Name: "pclmulqdq", Feature: &X86.HasPCLMULQDQ},
{Name: "popcnt", Feature: &X86.HasPOPCNT},
{Name: "rdrand", Feature: &X86.HasRDRAND},
{Name: "rdseed", Feature: &X86.HasRDSEED},
{Name: "sse3", Feature: &X86.HasSSE3},
{Name: "sse41", Feature: &X86.HasSSE41},
{Name: "sse42", Feature: &X86.HasSSE42},
{Name: "ssse3", Feature: &X86.HasSSSE3},
// These capabilities should always be enabled on amd64:
{Name: "sse2", Feature: &X86.HasSSE2, Required: runtime.GOARCH == "amd64"},
}
}
func archInit() {
Initialized = true
maxID, _, _, _ := cpuid(0, 0)
if maxID < 1 {
return
}
_, _, ecx1, edx1 := cpuid(1, 0)
X86.HasSSE2 = isSet(26, edx1)
X86.HasSSE3 = isSet(0, ecx1)
X86.HasPCLMULQDQ = isSet(1, ecx1)
X86.HasSSSE3 = isSet(9, ecx1)
X86.HasFMA = isSet(12, ecx1)
X86.HasCX16 = isSet(13, ecx1)
X86.HasSSE41 = isSet(19, ecx1)
X86.HasSSE42 = isSet(20, ecx1)
X86.HasPOPCNT = isSet(23, ecx1)
X86.HasAES = isSet(25, ecx1)
X86.HasOSXSAVE = isSet(27, ecx1)
X86.HasRDRAND = isSet(30, ecx1)
var osSupportsAVX, osSupportsAVX512 bool
// For XGETBV, OSXSAVE bit is required and sufficient.
if X86.HasOSXSAVE {
eax, _ := xgetbv()
// Check if XMM and YMM registers have OS support.
osSupportsAVX = isSet(1, eax) && isSet(2, eax)
if runtime.GOOS == "darwin" {
// Darwin doesn't save/restore AVX-512 mask registers correctly across signal handlers.
// Since users can't rely on mask register contents, let's not advertise AVX-512 support.
// See issue 49233.
osSupportsAVX512 = false
} else {
// Check if OPMASK and ZMM registers have OS support.
osSupportsAVX512 = osSupportsAVX && isSet(5, eax) && isSet(6, eax) && isSet(7, eax)
}
}
X86.HasAVX = isSet(28, ecx1) && osSupportsAVX
if maxID < 7 {
return
}
_, ebx7, ecx7, edx7 := cpuid(7, 0)
X86.HasBMI1 = isSet(3, ebx7)
X86.HasAVX2 = isSet(5, ebx7) && osSupportsAVX
X86.HasBMI2 = isSet(8, ebx7)
X86.HasERMS = isSet(9, ebx7)
X86.HasRDSEED = isSet(18, ebx7)
X86.HasADX = isSet(19, ebx7)
X86.HasAVX512 = isSet(16, ebx7) && osSupportsAVX512 // Because avx-512 foundation is the core required extension
if X86.HasAVX512 {
X86.HasAVX512F = true
X86.HasAVX512CD = isSet(28, ebx7)
X86.HasAVX512ER = isSet(27, ebx7)
X86.HasAVX512PF = isSet(26, ebx7)
X86.HasAVX512VL = isSet(31, ebx7)
X86.HasAVX512BW = isSet(30, ebx7)
X86.HasAVX512DQ = isSet(17, ebx7)
X86.HasAVX512IFMA = isSet(21, ebx7)
X86.HasAVX512VBMI = isSet(1, ecx7)
X86.HasAVX5124VNNIW = isSet(2, edx7)
X86.HasAVX5124FMAPS = isSet(3, edx7)
X86.HasAVX512VPOPCNTDQ = isSet(14, ecx7)
X86.HasAVX512VPCLMULQDQ = isSet(10, ecx7)
X86.HasAVX512VNNI = isSet(11, ecx7)
X86.HasAVX512GFNI = isSet(8, ecx7)
X86.HasAVX512VAES = isSet(9, ecx7)
X86.HasAVX512VBMI2 = isSet(6, ecx7)
X86.HasAVX512BITALG = isSet(12, ecx7)
eax71, _, _, _ := cpuid(7, 1)
X86.HasAVX512BF16 = isSet(5, eax71)
}
X86.HasAMXTile = isSet(24, edx7)
X86.HasAMXInt8 = isSet(25, edx7)
X86.HasAMXBF16 = isSet(22, edx7)
}
func isSet(bitpos uint, value uint32) bool {
return value&(1<<bitpos) != 0
}

28
gateway/vendor/golang.org/x/sys/cpu/cpu_x86.s generated vendored Normal file
View File

@ -0,0 +1,28 @@
// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build (386 || amd64 || amd64p32) && gc
// +build 386 amd64 amd64p32
// +build gc
#include "textflag.h"
// func cpuid(eaxArg, ecxArg uint32) (eax, ebx, ecx, edx uint32)
TEXT ·cpuid(SB), NOSPLIT, $0-24
MOVL eaxArg+0(FP), AX
MOVL ecxArg+4(FP), CX
CPUID
MOVL AX, eax+8(FP)
MOVL BX, ebx+12(FP)
MOVL CX, ecx+16(FP)
MOVL DX, edx+20(FP)
RET
// func xgetbv() (eax, edx uint32)
TEXT ·xgetbv(SB),NOSPLIT,$0-8
MOVL $0, CX
XGETBV
MOVL AX, eax+0(FP)
MOVL DX, edx+4(FP)
RET

10
gateway/vendor/golang.org/x/sys/cpu/cpu_zos.go generated vendored Normal file
View File

@ -0,0 +1,10 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
func archInit() {
doinit()
Initialized = true
}

25
gateway/vendor/golang.org/x/sys/cpu/cpu_zos_s390x.go generated vendored Normal file
View File

@ -0,0 +1,25 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
func initS390Xbase() {
// get the facilities list
facilities := stfle()
// mandatory
S390X.HasZARCH = facilities.Has(zarch)
S390X.HasSTFLE = facilities.Has(stflef)
S390X.HasLDISP = facilities.Has(ldisp)
S390X.HasEIMM = facilities.Has(eimm)
// optional
S390X.HasETF3EH = facilities.Has(etf3eh)
S390X.HasDFP = facilities.Has(dfp)
S390X.HasMSA = facilities.Has(msa)
S390X.HasVX = facilities.Has(vx)
if S390X.HasVX {
S390X.HasVXE = facilities.Has(vxe)
}
}

11
gateway/vendor/golang.org/x/sys/cpu/endian_big.go generated vendored Normal file
View File

@ -0,0 +1,11 @@
// Copyright 2023 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build armbe || arm64be || m68k || mips || mips64 || mips64p32 || ppc || ppc64 || s390 || s390x || shbe || sparc || sparc64
// +build armbe arm64be m68k mips mips64 mips64p32 ppc ppc64 s390 s390x shbe sparc sparc64
package cpu
// IsBigEndian records whether the GOARCH's byte order is big endian.
const IsBigEndian = true

11
gateway/vendor/golang.org/x/sys/cpu/endian_little.go generated vendored Normal file
View File

@ -0,0 +1,11 @@
// Copyright 2023 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build 386 || amd64 || amd64p32 || alpha || arm || arm64 || loong64 || mipsle || mips64le || mips64p32le || nios2 || ppc64le || riscv || riscv64 || sh || wasm
// +build 386 amd64 amd64p32 alpha arm arm64 loong64 mipsle mips64le mips64p32le nios2 ppc64le riscv riscv64 sh wasm
package cpu
// IsBigEndian records whether the GOARCH's byte order is big endian.
const IsBigEndian = false

71
gateway/vendor/golang.org/x/sys/cpu/hwcap_linux.go generated vendored Normal file
View File

@ -0,0 +1,71 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import (
"os"
)
const (
_AT_HWCAP = 16
_AT_HWCAP2 = 26
procAuxv = "/proc/self/auxv"
uintSize = int(32 << (^uint(0) >> 63))
)
// For those platforms don't have a 'cpuid' equivalent we use HWCAP/HWCAP2
// These are initialized in cpu_$GOARCH.go
// and should not be changed after they are initialized.
var hwCap uint
var hwCap2 uint
func readHWCAP() error {
// For Go 1.21+, get auxv from the Go runtime.
if a := getAuxv(); len(a) > 0 {
for len(a) >= 2 {
tag, val := a[0], uint(a[1])
a = a[2:]
switch tag {
case _AT_HWCAP:
hwCap = val
case _AT_HWCAP2:
hwCap2 = val
}
}
return nil
}
buf, err := os.ReadFile(procAuxv)
if err != nil {
// e.g. on android /proc/self/auxv is not accessible, so silently
// ignore the error and leave Initialized = false. On some
// architectures (e.g. arm64) doinit() implements a fallback
// readout and will set Initialized = true again.
return err
}
bo := hostByteOrder()
for len(buf) >= 2*(uintSize/8) {
var tag, val uint
switch uintSize {
case 32:
tag = uint(bo.Uint32(buf[0:]))
val = uint(bo.Uint32(buf[4:]))
buf = buf[8:]
case 64:
tag = uint(bo.Uint64(buf[0:]))
val = uint(bo.Uint64(buf[8:]))
buf = buf[16:]
}
switch tag {
case _AT_HWCAP:
hwCap = val
case _AT_HWCAP2:
hwCap2 = val
}
}
return nil
}

43
gateway/vendor/golang.org/x/sys/cpu/parse.go generated vendored Normal file
View File

@ -0,0 +1,43 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
import "strconv"
// parseRelease parses a dot-separated version number. It follows the semver
// syntax, but allows the minor and patch versions to be elided.
//
// This is a copy of the Go runtime's parseRelease from
// https://golang.org/cl/209597.
func parseRelease(rel string) (major, minor, patch int, ok bool) {
// Strip anything after a dash or plus.
for i := 0; i < len(rel); i++ {
if rel[i] == '-' || rel[i] == '+' {
rel = rel[:i]
break
}
}
next := func() (int, bool) {
for i := 0; i < len(rel); i++ {
if rel[i] == '.' {
ver, err := strconv.Atoi(rel[:i])
rel = rel[i+1:]
return ver, err == nil
}
}
ver, err := strconv.Atoi(rel)
rel = ""
return ver, err == nil
}
if major, ok = next(); !ok || rel == "" {
return
}
if minor, ok = next(); !ok || rel == "" {
return
}
patch, ok = next()
return
}

54
gateway/vendor/golang.org/x/sys/cpu/proc_cpuinfo_linux.go generated vendored Normal file
View File

@ -0,0 +1,54 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build linux && arm64
// +build linux,arm64
package cpu
import (
"errors"
"io"
"os"
"strings"
)
func readLinuxProcCPUInfo() error {
f, err := os.Open("/proc/cpuinfo")
if err != nil {
return err
}
defer f.Close()
var buf [1 << 10]byte // enough for first CPU
n, err := io.ReadFull(f, buf[:])
if err != nil && err != io.ErrUnexpectedEOF {
return err
}
in := string(buf[:n])
const features = "\nFeatures : "
i := strings.Index(in, features)
if i == -1 {
return errors.New("no CPU features found")
}
in = in[i+len(features):]
if i := strings.Index(in, "\n"); i != -1 {
in = in[:i]
}
m := map[string]*bool{}
initOptions() // need it early here; it's harmless to call twice
for _, o := range options {
m[o.Name] = o.Feature
}
// The EVTSTRM field has alias "evstrm" in Go, but Linux calls it "evtstrm".
m["evtstrm"] = &ARM64.HasEVTSTRM
for _, f := range strings.Fields(in) {
if p, ok := m[f]; ok {
*p = true
}
}
return nil
}

16
gateway/vendor/golang.org/x/sys/cpu/runtime_auxv.go generated vendored Normal file
View File

@ -0,0 +1,16 @@
// Copyright 2023 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package cpu
// getAuxvFn is non-nil on Go 1.21+ (via runtime_auxv_go121.go init)
// on platforms that use auxv.
var getAuxvFn func() []uintptr
func getAuxv() []uintptr {
if getAuxvFn == nil {
return nil
}
return getAuxvFn()
}

19
gateway/vendor/golang.org/x/sys/cpu/runtime_auxv_go121.go generated vendored Normal file
View File

@ -0,0 +1,19 @@
// Copyright 2023 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build go1.21
// +build go1.21
package cpu
import (
_ "unsafe" // for linkname
)
//go:linkname runtime_getAuxv runtime.getAuxv
func runtime_getAuxv() []uintptr
func init() {
getAuxvFn = runtime_getAuxv
}

27
gateway/vendor/golang.org/x/sys/cpu/syscall_aix_gccgo.go generated vendored Normal file
View File

@ -0,0 +1,27 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Recreate a getsystemcfg syscall handler instead of
// using the one provided by x/sys/unix to avoid having
// the dependency between them. (See golang.org/issue/32102)
// Moreover, this file will be used during the building of
// gccgo's libgo and thus must not used a CGo method.
//go:build aix && gccgo
// +build aix,gccgo
package cpu
import (
"syscall"
)
//extern getsystemcfg
func gccgoGetsystemcfg(label uint32) (r uint64)
func callgetsystemcfg(label int) (r1 uintptr, e1 syscall.Errno) {
r1 = uintptr(gccgoGetsystemcfg(uint32(label)))
e1 = syscall.GetErrno()
return
}

36
gateway/vendor/golang.org/x/sys/cpu/syscall_aix_ppc64_gc.go generated vendored Normal file
View File

@ -0,0 +1,36 @@
// Copyright 2019 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Minimal copy of x/sys/unix so the cpu package can make a
// system call on AIX without depending on x/sys/unix.
// (See golang.org/issue/32102)
//go:build aix && ppc64 && gc
// +build aix,ppc64,gc
package cpu
import (
"syscall"
"unsafe"
)
//go:cgo_import_dynamic libc_getsystemcfg getsystemcfg "libc.a/shr_64.o"
//go:linkname libc_getsystemcfg libc_getsystemcfg
type syscallFunc uintptr
var libc_getsystemcfg syscallFunc
type errno = syscall.Errno
// Implemented in runtime/syscall_aix.go.
func rawSyscall6(trap, nargs, a1, a2, a3, a4, a5, a6 uintptr) (r1, r2 uintptr, err errno)
func syscall6(trap, nargs, a1, a2, a3, a4, a5, a6 uintptr) (r1, r2 uintptr, err errno)
func callgetsystemcfg(label int) (r1 uintptr, e1 errno) {
r1, _, e1 = syscall6(uintptr(unsafe.Pointer(&libc_getsystemcfg)), 1, uintptr(label), 0, 0, 0, 0, 0)
return
}

30
gateway/vendor/golang.org/x/sys/internal/unsafeheader/unsafeheader.go generated vendored
View File

@ -1,30 +0,0 @@
// Copyright 2020 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package unsafeheader contains header declarations for the Go runtime's
// slice and string implementations.
//
// This package allows x/sys to use types equivalent to
// reflect.SliceHeader and reflect.StringHeader without introducing
// a dependency on the (relatively heavy) "reflect" package.
package unsafeheader
import (
"unsafe"
)
// Slice is the runtime representation of a slice.
// It cannot be used safely or portably and its representation may change in a later release.
type Slice struct {
Data unsafe.Pointer
Len int
Cap int
}
// String is the runtime representation of a string.
// It cannot be used safely or portably and its representation may change in a later release.
type String struct {
Data unsafe.Pointer
Len int
}

2
gateway/vendor/golang.org/x/sys/unix/mkall.sh generated vendored
View File

@ -50,7 +50,7 @@ if [[ "$GOOS" = "linux" ]]; then
# Use the Docker-based build system
# Files generated through docker (use $cmd so you can Ctl-C the build or run)
$cmd docker build --tag generate:$GOOS $GOOS
$cmd docker run --interactive --tty --volume $(cd -- "$(dirname -- "$0")/.." && /bin/pwd):/build generate:$GOOS
$cmd docker run --interactive --tty --volume $(cd -- "$(dirname -- "$0")/.." && pwd):/build generate:$GOOS
exit
fi

11
gateway/vendor/golang.org/x/sys/unix/mkerrors.sh generated vendored
View File

@ -519,7 +519,7 @@ ccflags="$@"
$2 ~ /^LOCK_(SH|EX|NB|UN)$/ ||
$2 ~ /^LO_(KEY|NAME)_SIZE$/ ||
$2 ~ /^LOOP_(CLR|CTL|GET|SET)_/ ||
$2 ~ /^(AF|SOCK|SO|SOL|IPPROTO|IP|IPV6|TCP|MCAST|EVFILT|NOTE|SHUT|PROT|MAP|MFD|T?PACKET|MSG|SCM|MCL|DT|MADV|PR|LOCAL|TCPOPT|UDP)_/ ||
$2 ~ /^(AF|SOCK|SO|SOL|IPPROTO|IP|IPV6|TCP|MCAST|EVFILT|NOTE|SHUT|PROT|MAP|MREMAP|MFD|T?PACKET|MSG|SCM|MCL|DT|MADV|PR|LOCAL|TCPOPT|UDP)_/ ||
$2 ~ /^NFC_(GENL|PROTO|COMM|RF|SE|DIRECTION|LLCP|SOCKPROTO)_/ ||
$2 ~ /^NFC_.*_(MAX)?SIZE$/ ||
$2 ~ /^RAW_PAYLOAD_/ ||
@ -583,6 +583,7 @@ ccflags="$@"
$2 ~ /^PERF_/ ||
$2 ~ /^SECCOMP_MODE_/ ||
$2 ~ /^SEEK_/ ||
$2 ~ /^SCHED_/ ||
$2 ~ /^SPLICE_/ ||
$2 ~ /^SYNC_FILE_RANGE_/ ||
$2 !~ /IOC_MAGIC/ &&
@ -624,7 +625,7 @@ ccflags="$@"
$2 ~ /^MEM/ ||
$2 ~ /^WG/ ||
$2 ~ /^FIB_RULE_/ ||
$2 ~ /^BLK[A-Z]*(GET$|SET$|BUF$|PART$|SIZE)/ {printf("\t%s = C.%s\n", $2, $2)}
$2 ~ /^BLK[A-Z]*(GET$|SET$|BUF$|PART$|SIZE|IOMIN$|IOOPT$|ALIGNOFF$|DISCARD|ROTATIONAL$|ZEROOUT$|GETDISKSEQ$)/ {printf("\t%s = C.%s\n", $2, $2)}
$2 ~ /^__WCOREFLAG$/ {next}
$2 ~ /^__W[A-Z0-9]+$/ {printf("\t%s = C.%s\n", substr($2,3), $2)}
@ -741,7 +742,8 @@ main(void)
e = errors[i].num;
if(i > 0 && errors[i-1].num == e)
continue;
strcpy(buf, strerror(e));
strncpy(buf, strerror(e), sizeof(buf) - 1);
buf[sizeof(buf) - 1] = '\0';
// lowercase first letter: Bad -> bad, but STREAM -> STREAM.
if(A <= buf[0] && buf[0] <= Z && a <= buf[1] && buf[1] <= z)
buf[0] += a - A;
@ -760,7 +762,8 @@ main(void)
e = signals[i].num;
if(i > 0 && signals[i-1].num == e)
continue;
strcpy(buf, strsignal(e));
strncpy(buf, strsignal(e), sizeof(buf) - 1);
buf[sizeof(buf) - 1] = '\0';
// lowercase first letter: Bad -> bad, but STREAM -> STREAM.
if(A <= buf[0] && buf[0] <= Z && a <= buf[1] && buf[1] <= z)
buf[0] += a - A;

14
gateway/vendor/golang.org/x/sys/unix/mmap_nomremap.go generated vendored Normal file
View File

@ -0,0 +1,14 @@
// Copyright 2023 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build aix || darwin || dragonfly || freebsd || openbsd || solaris
// +build aix darwin dragonfly freebsd openbsd solaris
package unix
var mapper = &mmapper{
active: make(map[*byte][]byte),
mmap: mmap,
munmap: munmap,
}

53
gateway/vendor/golang.org/x/sys/unix/mremap.go generated vendored Normal file
View File

@ -0,0 +1,53 @@
// Copyright 2023 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build linux || netbsd
// +build linux netbsd
package unix
import "unsafe"
type mremapMmapper struct {
mmapper
mremap func(oldaddr uintptr, oldlength uintptr, newlength uintptr, flags int, newaddr uintptr) (xaddr uintptr, err error)
}
var mapper = &mremapMmapper{
mmapper: mmapper{
active: make(map[*byte][]byte),
mmap: mmap,
munmap: munmap,
},
mremap: mremap,
}
func (m *mremapMmapper) Mremap(oldData []byte, newLength int, flags int) (data []byte, err error) {
if newLength <= 0 || len(oldData) == 0 || len(oldData) != cap(oldData) || flags&mremapFixed != 0 {
return nil, EINVAL
}
pOld := &oldData[cap(oldData)-1]
m.Lock()
defer m.Unlock()
bOld := m.active[pOld]
if bOld == nil || &bOld[0] != &oldData[0] {
return nil, EINVAL
}
newAddr, errno := m.mremap(uintptr(unsafe.Pointer(&bOld[0])), uintptr(len(bOld)), uintptr(newLength), flags, 0)
if errno != nil {
return nil, errno
}
bNew := unsafe.Slice((*byte)(unsafe.Pointer(newAddr)), newLength)
pNew := &bNew[cap(bNew)-1]
if flags&mremapDontunmap == 0 {
delete(m.active, pOld)
}
m.active[pNew] = bNew
return bNew, nil
}
func Mremap(oldData []byte, newLength int, flags int) (data []byte, err error) {
return mapper.Mremap(oldData, newLength, flags)
}

Some files were not shown because too many files have changed in this diff Show More