mirrors/ceph-csi
1
0
Fork 0
mirror of https://github.com/ceph/ceph-csi.git synced 2025-06-14 18:53:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

added vendors

Browse Source
This commit is contained in:
mickymiek
2018-12-19 15:29:25 +01:00
committed by Huamin Chen
parent 62d65ad0cb
commit 35561301b2
2952 changed files with 1124359 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

244
vendor/golang.org/x/crypto/blake2s/blake2s.go generated vendored Normal file
View File

@ -0,0 +1,244 @@
// 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 blake2s implements the BLAKE2s hash algorithm defined by RFC 7693
// and the extendable output function (XOF) BLAKE2Xs.
//
// For a detailed specification of BLAKE2s see https://blake2.net/blake2.pdf
// and for BLAKE2Xs see https://blake2.net/blake2x.pdf
//
// If you aren't sure which function you need, use BLAKE2s (Sum256 or New256).
// If you need a secret-key MAC (message authentication code), use the New256
// function with a non-nil key.
//
// BLAKE2X is a construction to compute hash values larger than 32 bytes. It
// can produce hash values between 0 and 65535 bytes.
package blake2s // import "golang.org/x/crypto/blake2s"
import (
"encoding/binary"
"errors"
"hash"
)
const (
// The blocksize of BLAKE2s in bytes.
BlockSize = 64
// The hash size of BLAKE2s-256 in bytes.
Size = 32
// The hash size of BLAKE2s-128 in bytes.
Size128 = 16
)
var errKeySize = errors.New("blake2s: invalid key size")
var iv = [8]uint32{
0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a,
0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19,
}
// Sum256 returns the BLAKE2s-256 checksum of the data.
func Sum256(data []byte) [Size]byte {
var sum [Size]byte
checkSum(&sum, Size, data)
return sum
}
// New256 returns a new hash.Hash computing the BLAKE2s-256 checksum. A non-nil
// key turns the hash into a MAC. The key must between zero and 32 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 New256(key []byte) (hash.Hash, error) { return newDigest(Size, key) }
// New128 returns a new hash.Hash computing the BLAKE2s-128 checksum given a
// non-empty key. Note that a 128-bit digest is too small to be secure as a
// cryptographic hash and should only be used as a MAC, thus the key argument
// is not optional.
func New128(key []byte) (hash.Hash, error) {
if len(key) == 0 {
return nil, errors.New("blake2s: a key is required for a 128-bit hash")
}
return newDigest(Size128, key)
}
func newDigest(hashSize int, key []byte) (*digest, error) {
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) {
var (
h [8]uint32
c [2]uint32
)
h = iv
h[0] ^= uint32(hashSize) | (1 << 16) | (1 << 24)
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 := uint32(BlockSize - offset)
if c[0] < remaining {
c[1]--
}
c[0] -= remaining
hashBlocks(&h, &c, 0xFFFFFFFF, block[:])
for i, v := range h {
binary.LittleEndian.PutUint32(sum[4*i:], v)
}
}
type digest struct {
h [8]uint32
c [2]uint32
size int
block [BlockSize]byte
offset int
key [BlockSize]byte
keyLen int
}
const (
magic = "b2s"
marshaledSize = len(magic) + 8*4 + 2*4 + 1 + BlockSize + 1
)
func (d *digest) MarshalBinary() ([]byte, error) {
if d.keyLen != 0 {
return nil, errors.New("crypto/blake2s: cannot marshal MACs")
}
b := make([]byte, 0, marshaledSize)
b = append(b, magic...)
for i := 0; i < 8; i++ {
b = appendUint32(b, d.h[i])
}
b = appendUint32(b, d.c[0])
b = appendUint32(b, d.c[1])
// Maximum value for size is 32
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/blake2s: invalid hash state identifier")
}
if len(b) != marshaledSize {
return errors.New("crypto/blake2s: invalid hash state size")
}
b = b[len(magic):]
for i := 0; i < 8; i++ {
b, d.h[i] = consumeUint32(b)
}
b, d.c[0] = consumeUint32(b)
b, d.c[1] = consumeUint32(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] ^= uint32(d.size) | (uint32(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:]
}
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
h := d.h
c := d.c
copy(block[:], d.block[:d.offset])
remaining := uint32(BlockSize - d.offset)
if c[0] < remaining {
c[1]--
}
c[0] -= remaining
hashBlocks(&h, &c, 0xFFFFFFFF, block[:])
for i, v := range h {
binary.LittleEndian.PutUint32(hash[4*i:], v)
}
}
func appendUint32(b []byte, x uint32) []byte {
var a [4]byte
binary.BigEndian.PutUint32(a[:], x)
return append(b, a[:]...)
}
func consumeUint32(b []byte) ([]byte, uint32) {
x := binary.BigEndian.Uint32(b)
return b[4:], x
}

32
vendor/golang.org/x/crypto/blake2s/blake2s_386.go generated vendored Normal file
View File

@ -0,0 +1,32 @@
// 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.
// +build 386,!gccgo,!appengine
package blake2s
import "golang.org/x/sys/cpu"
var (
useSSE4 = false
useSSSE3 = cpu.X86.HasSSSE3
useSSE2 = cpu.X86.HasSSE2
)
//go:noescape
func hashBlocksSSE2(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
//go:noescape
func hashBlocksSSSE3(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
func hashBlocks(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte) {
switch {
case useSSSE3:
hashBlocksSSSE3(h, c, flag, blocks)
case useSSE2:
hashBlocksSSE2(h, c, flag, blocks)
default:
hashBlocksGeneric(h, c, flag, blocks)
}
}

435
vendor/golang.org/x/crypto/blake2s/blake2s_386.s generated vendored Normal file
View File

@ -0,0 +1,435 @@
// 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.
// +build 386,!gccgo,!appengine
#include "textflag.h"
DATA iv0<>+0x00(SB)/4, $0x6a09e667
DATA iv0<>+0x04(SB)/4, $0xbb67ae85
DATA iv0<>+0x08(SB)/4, $0x3c6ef372
DATA iv0<>+0x0c(SB)/4, $0xa54ff53a
GLOBL iv0<>(SB), (NOPTR+RODATA), $16
DATA iv1<>+0x00(SB)/4, $0x510e527f
DATA iv1<>+0x04(SB)/4, $0x9b05688c
DATA iv1<>+0x08(SB)/4, $0x1f83d9ab
DATA iv1<>+0x0c(SB)/4, $0x5be0cd19
GLOBL iv1<>(SB), (NOPTR+RODATA), $16
DATA rol16<>+0x00(SB)/8, $0x0504070601000302
DATA rol16<>+0x08(SB)/8, $0x0D0C0F0E09080B0A
GLOBL rol16<>(SB), (NOPTR+RODATA), $16
DATA rol8<>+0x00(SB)/8, $0x0407060500030201
DATA rol8<>+0x08(SB)/8, $0x0C0F0E0D080B0A09
GLOBL rol8<>(SB), (NOPTR+RODATA), $16
DATA counter<>+0x00(SB)/8, $0x40
DATA counter<>+0x08(SB)/8, $0x0
GLOBL counter<>(SB), (NOPTR+RODATA), $16
#define ROTL_SSE2(n, t, v) \
MOVO v, t; \
PSLLL $n, t; \
PSRLL $(32-n), v; \
PXOR t, v
#define ROTL_SSSE3(c, v) \
PSHUFB c, v
#define ROUND_SSE2(v0, v1, v2, v3, m0, m1, m2, m3, t) \
PADDL m0, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(16, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m1, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(24, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v1, v1; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v3, v3; \
PADDL m2, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(16, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m3, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(24, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v3, v3; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v1, v1
#define ROUND_SSSE3(v0, v1, v2, v3, m0, m1, m2, m3, t, c16, c8) \
PADDL m0, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c16, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m1, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c8, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v1, v1; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v3, v3; \
PADDL m2, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c16, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m3, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c8, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v3, v3; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v1, v1
#define PRECOMPUTE(dst, off, src, t) \
MOVL 0*4(src), t; \
MOVL t, 0*4+off+0(dst); \
MOVL t, 9*4+off+64(dst); \
MOVL t, 5*4+off+128(dst); \
MOVL t, 14*4+off+192(dst); \
MOVL t, 4*4+off+256(dst); \
MOVL t, 2*4+off+320(dst); \
MOVL t, 8*4+off+384(dst); \
MOVL t, 12*4+off+448(dst); \
MOVL t, 3*4+off+512(dst); \
MOVL t, 15*4+off+576(dst); \
MOVL 1*4(src), t; \
MOVL t, 4*4+off+0(dst); \
MOVL t, 8*4+off+64(dst); \
MOVL t, 14*4+off+128(dst); \
MOVL t, 5*4+off+192(dst); \
MOVL t, 12*4+off+256(dst); \
MOVL t, 11*4+off+320(dst); \
MOVL t, 1*4+off+384(dst); \
MOVL t, 6*4+off+448(dst); \
MOVL t, 10*4+off+512(dst); \
MOVL t, 3*4+off+576(dst); \
MOVL 2*4(src), t; \
MOVL t, 1*4+off+0(dst); \
MOVL t, 13*4+off+64(dst); \
MOVL t, 6*4+off+128(dst); \
MOVL t, 8*4+off+192(dst); \
MOVL t, 2*4+off+256(dst); \
MOVL t, 0*4+off+320(dst); \
MOVL t, 14*4+off+384(dst); \
MOVL t, 11*4+off+448(dst); \
MOVL t, 12*4+off+512(dst); \
MOVL t, 4*4+off+576(dst); \
MOVL 3*4(src), t; \
MOVL t, 5*4+off+0(dst); \
MOVL t, 15*4+off+64(dst); \
MOVL t, 9*4+off+128(dst); \
MOVL t, 1*4+off+192(dst); \
MOVL t, 11*4+off+256(dst); \
MOVL t, 7*4+off+320(dst); \
MOVL t, 13*4+off+384(dst); \
MOVL t, 3*4+off+448(dst); \
MOVL t, 6*4+off+512(dst); \
MOVL t, 10*4+off+576(dst); \
MOVL 4*4(src), t; \
MOVL t, 2*4+off+0(dst); \
MOVL t, 1*4+off+64(dst); \
MOVL t, 15*4+off+128(dst); \
MOVL t, 10*4+off+192(dst); \
MOVL t, 6*4+off+256(dst); \
MOVL t, 8*4+off+320(dst); \
MOVL t, 3*4+off+384(dst); \
MOVL t, 13*4+off+448(dst); \
MOVL t, 14*4+off+512(dst); \
MOVL t, 5*4+off+576(dst); \
MOVL 5*4(src), t; \
MOVL t, 6*4+off+0(dst); \
MOVL t, 11*4+off+64(dst); \
MOVL t, 2*4+off+128(dst); \
MOVL t, 9*4+off+192(dst); \
MOVL t, 1*4+off+256(dst); \
MOVL t, 13*4+off+320(dst); \
MOVL t, 4*4+off+384(dst); \
MOVL t, 8*4+off+448(dst); \
MOVL t, 15*4+off+512(dst); \
MOVL t, 7*4+off+576(dst); \
MOVL 6*4(src), t; \
MOVL t, 3*4+off+0(dst); \
MOVL t, 7*4+off+64(dst); \
MOVL t, 13*4+off+128(dst); \
MOVL t, 12*4+off+192(dst); \
MOVL t, 10*4+off+256(dst); \
MOVL t, 1*4+off+320(dst); \
MOVL t, 9*4+off+384(dst); \
MOVL t, 14*4+off+448(dst); \
MOVL t, 0*4+off+512(dst); \
MOVL t, 6*4+off+576(dst); \
MOVL 7*4(src), t; \
MOVL t, 7*4+off+0(dst); \
MOVL t, 14*4+off+64(dst); \
MOVL t, 10*4+off+128(dst); \
MOVL t, 0*4+off+192(dst); \
MOVL t, 5*4+off+256(dst); \
MOVL t, 9*4+off+320(dst); \
MOVL t, 12*4+off+384(dst); \
MOVL t, 1*4+off+448(dst); \
MOVL t, 13*4+off+512(dst); \
MOVL t, 2*4+off+576(dst); \
MOVL 8*4(src), t; \
MOVL t, 8*4+off+0(dst); \
MOVL t, 5*4+off+64(dst); \
MOVL t, 4*4+off+128(dst); \
MOVL t, 15*4+off+192(dst); \
MOVL t, 14*4+off+256(dst); \
MOVL t, 3*4+off+320(dst); \
MOVL t, 11*4+off+384(dst); \
MOVL t, 10*4+off+448(dst); \
MOVL t, 7*4+off+512(dst); \
MOVL t, 1*4+off+576(dst); \
MOVL 9*4(src), t; \
MOVL t, 12*4+off+0(dst); \
MOVL t, 2*4+off+64(dst); \
MOVL t, 11*4+off+128(dst); \
MOVL t, 4*4+off+192(dst); \
MOVL t, 0*4+off+256(dst); \
MOVL t, 15*4+off+320(dst); \
MOVL t, 10*4+off+384(dst); \
MOVL t, 7*4+off+448(dst); \
MOVL t, 5*4+off+512(dst); \
MOVL t, 9*4+off+576(dst); \
MOVL 10*4(src), t; \
MOVL t, 9*4+off+0(dst); \
MOVL t, 4*4+off+64(dst); \
MOVL t, 8*4+off+128(dst); \
MOVL t, 13*4+off+192(dst); \
MOVL t, 3*4+off+256(dst); \
MOVL t, 5*4+off+320(dst); \
MOVL t, 7*4+off+384(dst); \
MOVL t, 15*4+off+448(dst); \
MOVL t, 11*4+off+512(dst); \
MOVL t, 0*4+off+576(dst); \
MOVL 11*4(src), t; \
MOVL t, 13*4+off+0(dst); \
MOVL t, 10*4+off+64(dst); \
MOVL t, 0*4+off+128(dst); \
MOVL t, 3*4+off+192(dst); \
MOVL t, 9*4+off+256(dst); \
MOVL t, 6*4+off+320(dst); \
MOVL t, 15*4+off+384(dst); \
MOVL t, 4*4+off+448(dst); \
MOVL t, 2*4+off+512(dst); \
MOVL t, 12*4+off+576(dst); \
MOVL 12*4(src), t; \
MOVL t, 10*4+off+0(dst); \
MOVL t, 12*4+off+64(dst); \
MOVL t, 1*4+off+128(dst); \
MOVL t, 6*4+off+192(dst); \
MOVL t, 13*4+off+256(dst); \
MOVL t, 4*4+off+320(dst); \
MOVL t, 0*4+off+384(dst); \
MOVL t, 2*4+off+448(dst); \
MOVL t, 8*4+off+512(dst); \
MOVL t, 14*4+off+576(dst); \
MOVL 13*4(src), t; \
MOVL t, 14*4+off+0(dst); \
MOVL t, 3*4+off+64(dst); \
MOVL t, 7*4+off+128(dst); \
MOVL t, 2*4+off+192(dst); \
MOVL t, 15*4+off+256(dst); \
MOVL t, 12*4+off+320(dst); \
MOVL t, 6*4+off+384(dst); \
MOVL t, 0*4+off+448(dst); \
MOVL t, 9*4+off+512(dst); \
MOVL t, 11*4+off+576(dst); \
MOVL 14*4(src), t; \
MOVL t, 11*4+off+0(dst); \
MOVL t, 0*4+off+64(dst); \
MOVL t, 12*4+off+128(dst); \
MOVL t, 7*4+off+192(dst); \
MOVL t, 8*4+off+256(dst); \
MOVL t, 14*4+off+320(dst); \
MOVL t, 2*4+off+384(dst); \
MOVL t, 5*4+off+448(dst); \
MOVL t, 1*4+off+512(dst); \
MOVL t, 13*4+off+576(dst); \
MOVL 15*4(src), t; \
MOVL t, 15*4+off+0(dst); \
MOVL t, 6*4+off+64(dst); \
MOVL t, 3*4+off+128(dst); \
MOVL t, 11*4+off+192(dst); \
MOVL t, 7*4+off+256(dst); \
MOVL t, 10*4+off+320(dst); \
MOVL t, 5*4+off+384(dst); \
MOVL t, 9*4+off+448(dst); \
MOVL t, 4*4+off+512(dst); \
MOVL t, 8*4+off+576(dst)
// func hashBlocksSSE2(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
TEXT ·hashBlocksSSE2(SB), 0, $672-24 // frame = 656 + 16 byte alignment
MOVL h+0(FP), AX
MOVL c+4(FP), BX
MOVL flag+8(FP), CX
MOVL blocks_base+12(FP), SI
MOVL blocks_len+16(FP), DX
MOVL SP, BP
MOVL SP, DI
ADDL $15, DI
ANDL $~15, DI
MOVL DI, SP
MOVL CX, 8(SP)
MOVL 0(BX), CX
MOVL CX, 0(SP)
MOVL 4(BX), CX
MOVL CX, 4(SP)
XORL CX, CX
MOVL CX, 12(SP)
MOVOU 0(AX), X0
MOVOU 16(AX), X1
MOVOU counter<>(SB), X2
loop:
MOVO X0, X4
MOVO X1, X5
MOVOU iv0<>(SB), X6
MOVOU iv1<>(SB), X7
MOVO 0(SP), X3
PADDQ X2, X3
PXOR X3, X7
MOVO X3, 0(SP)
PRECOMPUTE(SP, 16, SI, CX)
ROUND_SSE2(X4, X5, X6, X7, 16(SP), 32(SP), 48(SP), 64(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+64(SP), 32+64(SP), 48+64(SP), 64+64(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+128(SP), 32+128(SP), 48+128(SP), 64+128(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+192(SP), 32+192(SP), 48+192(SP), 64+192(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+256(SP), 32+256(SP), 48+256(SP), 64+256(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+320(SP), 32+320(SP), 48+320(SP), 64+320(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+384(SP), 32+384(SP), 48+384(SP), 64+384(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+448(SP), 32+448(SP), 48+448(SP), 64+448(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+512(SP), 32+512(SP), 48+512(SP), 64+512(SP), X3)
ROUND_SSE2(X4, X5, X6, X7, 16+576(SP), 32+576(SP), 48+576(SP), 64+576(SP), X3)
PXOR X4, X0
PXOR X5, X1
PXOR X6, X0
PXOR X7, X1
LEAL 64(SI), SI
SUBL $64, DX
JNE loop
MOVL 0(SP), CX
MOVL CX, 0(BX)
MOVL 4(SP), CX
MOVL CX, 4(BX)
MOVOU X0, 0(AX)
MOVOU X1, 16(AX)
MOVL BP, SP
RET
// func hashBlocksSSSE3(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
TEXT ·hashBlocksSSSE3(SB), 0, $704-24 // frame = 688 + 16 byte alignment
MOVL h+0(FP), AX
MOVL c+4(FP), BX
MOVL flag+8(FP), CX
MOVL blocks_base+12(FP), SI
MOVL blocks_len+16(FP), DX
MOVL SP, BP
MOVL SP, DI
ADDL $15, DI
ANDL $~15, DI
MOVL DI, SP
MOVL CX, 8(SP)
MOVL 0(BX), CX
MOVL CX, 0(SP)
MOVL 4(BX), CX
MOVL CX, 4(SP)
XORL CX, CX
MOVL CX, 12(SP)
MOVOU 0(AX), X0
MOVOU 16(AX), X1
MOVOU counter<>(SB), X2
loop:
MOVO X0, 656(SP)
MOVO X1, 672(SP)
MOVO X0, X4
MOVO X1, X5
MOVOU iv0<>(SB), X6
MOVOU iv1<>(SB), X7
MOVO 0(SP), X3
PADDQ X2, X3
PXOR X3, X7
MOVO X3, 0(SP)
MOVOU rol16<>(SB), X0
MOVOU rol8<>(SB), X1
PRECOMPUTE(SP, 16, SI, CX)
ROUND_SSSE3(X4, X5, X6, X7, 16(SP), 32(SP), 48(SP), 64(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+64(SP), 32+64(SP), 48+64(SP), 64+64(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+128(SP), 32+128(SP), 48+128(SP), 64+128(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+192(SP), 32+192(SP), 48+192(SP), 64+192(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+256(SP), 32+256(SP), 48+256(SP), 64+256(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+320(SP), 32+320(SP), 48+320(SP), 64+320(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+384(SP), 32+384(SP), 48+384(SP), 64+384(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+448(SP), 32+448(SP), 48+448(SP), 64+448(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+512(SP), 32+512(SP), 48+512(SP), 64+512(SP), X3, X0, X1)
ROUND_SSSE3(X4, X5, X6, X7, 16+576(SP), 32+576(SP), 48+576(SP), 64+576(SP), X3, X0, X1)
MOVO 656(SP), X0
MOVO 672(SP), X1
PXOR X4, X0
PXOR X5, X1
PXOR X6, X0
PXOR X7, X1
LEAL 64(SI), SI
SUBL $64, DX
JNE loop
MOVL 0(SP), CX
MOVL CX, 0(BX)
MOVL 4(SP), CX
MOVL CX, 4(BX)
MOVOU X0, 0(AX)
MOVOU X1, 16(AX)
MOVL BP, SP
RET

37
vendor/golang.org/x/crypto/blake2s/blake2s_amd64.go generated vendored Normal file
View File

@ -0,0 +1,37 @@
// 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.
// +build amd64,!gccgo,!appengine
package blake2s
import "golang.org/x/sys/cpu"
var (
useSSE4 = cpu.X86.HasSSE41
useSSSE3 = cpu.X86.HasSSSE3
useSSE2 = cpu.X86.HasSSE2
)
//go:noescape
func hashBlocksSSE2(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
//go:noescape
func hashBlocksSSSE3(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
//go:noescape
func hashBlocksSSE4(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
func hashBlocks(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte) {
switch {
case useSSE4:
hashBlocksSSE4(h, c, flag, blocks)
case useSSSE3:
hashBlocksSSSE3(h, c, flag, blocks)
case useSSE2:
hashBlocksSSE2(h, c, flag, blocks)
default:
hashBlocksGeneric(h, c, flag, blocks)
}
}

438
vendor/golang.org/x/crypto/blake2s/blake2s_amd64.s generated vendored Normal file
View File

@ -0,0 +1,438 @@
// 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.
// +build amd64,!gccgo,!appengine
#include "textflag.h"
DATA iv0<>+0x00(SB)/4, $0x6a09e667
DATA iv0<>+0x04(SB)/4, $0xbb67ae85
DATA iv0<>+0x08(SB)/4, $0x3c6ef372
DATA iv0<>+0x0c(SB)/4, $0xa54ff53a
GLOBL iv0<>(SB), (NOPTR+RODATA), $16
DATA iv1<>+0x00(SB)/4, $0x510e527f
DATA iv1<>+0x04(SB)/4, $0x9b05688c
DATA iv1<>+0x08(SB)/4, $0x1f83d9ab
DATA iv1<>+0x0c(SB)/4, $0x5be0cd19
GLOBL iv1<>(SB), (NOPTR+RODATA), $16
DATA rol16<>+0x00(SB)/8, $0x0504070601000302
DATA rol16<>+0x08(SB)/8, $0x0D0C0F0E09080B0A
GLOBL rol16<>(SB), (NOPTR+RODATA), $16
DATA rol8<>+0x00(SB)/8, $0x0407060500030201
DATA rol8<>+0x08(SB)/8, $0x0C0F0E0D080B0A09
GLOBL rol8<>(SB), (NOPTR+RODATA), $16
DATA counter<>+0x00(SB)/8, $0x40
DATA counter<>+0x08(SB)/8, $0x0
GLOBL counter<>(SB), (NOPTR+RODATA), $16
#define ROTL_SSE2(n, t, v) \
MOVO v, t; \
PSLLL $n, t; \
PSRLL $(32-n), v; \
PXOR t, v
#define ROTL_SSSE3(c, v) \
PSHUFB c, v
#define ROUND_SSE2(v0, v1, v2, v3, m0, m1, m2, m3, t) \
PADDL m0, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(16, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m1, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(24, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v1, v1; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v3, v3; \
PADDL m2, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(16, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m3, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSE2(24, t, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v3, v3; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v1, v1
#define ROUND_SSSE3(v0, v1, v2, v3, m0, m1, m2, m3, t, c16, c8) \
PADDL m0, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c16, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m1, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c8, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v1, v1; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v3, v3; \
PADDL m2, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c16, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(20, t, v1); \
PADDL m3, v0; \
PADDL v1, v0; \
PXOR v0, v3; \
ROTL_SSSE3(c8, v3); \
PADDL v3, v2; \
PXOR v2, v1; \
ROTL_SSE2(25, t, v1); \
PSHUFL $0x39, v3, v3; \
PSHUFL $0x4E, v2, v2; \
PSHUFL $0x93, v1, v1
#define LOAD_MSG_SSE4(m0, m1, m2, m3, src, i0, i1, i2, i3, i4, i5, i6, i7, i8, i9, i10, i11, i12, i13, i14, i15) \
MOVL i0*4(src), m0; \
PINSRD $1, i1*4(src), m0; \
PINSRD $2, i2*4(src), m0; \
PINSRD $3, i3*4(src), m0; \
MOVL i4*4(src), m1; \
PINSRD $1, i5*4(src), m1; \
PINSRD $2, i6*4(src), m1; \
PINSRD $3, i7*4(src), m1; \
MOVL i8*4(src), m2; \
PINSRD $1, i9*4(src), m2; \
PINSRD $2, i10*4(src), m2; \
PINSRD $3, i11*4(src), m2; \
MOVL i12*4(src), m3; \
PINSRD $1, i13*4(src), m3; \
PINSRD $2, i14*4(src), m3; \
PINSRD $3, i15*4(src), m3
#define PRECOMPUTE_MSG(dst, off, src, R8, R9, R10, R11, R12, R13, R14, R15) \
MOVQ 0*4(src), R8; \
MOVQ 2*4(src), R9; \
MOVQ 4*4(src), R10; \
MOVQ 6*4(src), R11; \
MOVQ 8*4(src), R12; \
MOVQ 10*4(src), R13; \
MOVQ 12*4(src), R14; \
MOVQ 14*4(src), R15; \
\
MOVL R8, 0*4+off+0(dst); \
MOVL R8, 9*4+off+64(dst); \
MOVL R8, 5*4+off+128(dst); \
MOVL R8, 14*4+off+192(dst); \
MOVL R8, 4*4+off+256(dst); \
MOVL R8, 2*4+off+320(dst); \
MOVL R8, 8*4+off+384(dst); \
MOVL R8, 12*4+off+448(dst); \
MOVL R8, 3*4+off+512(dst); \
MOVL R8, 15*4+off+576(dst); \
SHRQ $32, R8; \
MOVL R8, 4*4+off+0(dst); \
MOVL R8, 8*4+off+64(dst); \
MOVL R8, 14*4+off+128(dst); \
MOVL R8, 5*4+off+192(dst); \
MOVL R8, 12*4+off+256(dst); \
MOVL R8, 11*4+off+320(dst); \
MOVL R8, 1*4+off+384(dst); \
MOVL R8, 6*4+off+448(dst); \
MOVL R8, 10*4+off+512(dst); \
MOVL R8, 3*4+off+576(dst); \
\
MOVL R9, 1*4+off+0(dst); \
MOVL R9, 13*4+off+64(dst); \
MOVL R9, 6*4+off+128(dst); \
MOVL R9, 8*4+off+192(dst); \
MOVL R9, 2*4+off+256(dst); \
MOVL R9, 0*4+off+320(dst); \
MOVL R9, 14*4+off+384(dst); \
MOVL R9, 11*4+off+448(dst); \
MOVL R9, 12*4+off+512(dst); \
MOVL R9, 4*4+off+576(dst); \
SHRQ $32, R9; \
MOVL R9, 5*4+off+0(dst); \
MOVL R9, 15*4+off+64(dst); \
MOVL R9, 9*4+off+128(dst); \
MOVL R9, 1*4+off+192(dst); \
MOVL R9, 11*4+off+256(dst); \
MOVL R9, 7*4+off+320(dst); \
MOVL R9, 13*4+off+384(dst); \
MOVL R9, 3*4+off+448(dst); \
MOVL R9, 6*4+off+512(dst); \
MOVL R9, 10*4+off+576(dst); \
\
MOVL R10, 2*4+off+0(dst); \
MOVL R10, 1*4+off+64(dst); \
MOVL R10, 15*4+off+128(dst); \
MOVL R10, 10*4+off+192(dst); \
MOVL R10, 6*4+off+256(dst); \
MOVL R10, 8*4+off+320(dst); \
MOVL R10, 3*4+off+384(dst); \
MOVL R10, 13*4+off+448(dst); \
MOVL R10, 14*4+off+512(dst); \
MOVL R10, 5*4+off+576(dst); \
SHRQ $32, R10; \
MOVL R10, 6*4+off+0(dst); \
MOVL R10, 11*4+off+64(dst); \
MOVL R10, 2*4+off+128(dst); \
MOVL R10, 9*4+off+192(dst); \
MOVL R10, 1*4+off+256(dst); \
MOVL R10, 13*4+off+320(dst); \
MOVL R10, 4*4+off+384(dst); \
MOVL R10, 8*4+off+448(dst); \
MOVL R10, 15*4+off+512(dst); \
MOVL R10, 7*4+off+576(dst); \
\
MOVL R11, 3*4+off+0(dst); \
MOVL R11, 7*4+off+64(dst); \
MOVL R11, 13*4+off+128(dst); \
MOVL R11, 12*4+off+192(dst); \
MOVL R11, 10*4+off+256(dst); \
MOVL R11, 1*4+off+320(dst); \
MOVL R11, 9*4+off+384(dst); \
MOVL R11, 14*4+off+448(dst); \
MOVL R11, 0*4+off+512(dst); \
MOVL R11, 6*4+off+576(dst); \
SHRQ $32, R11; \
MOVL R11, 7*4+off+0(dst); \
MOVL R11, 14*4+off+64(dst); \
MOVL R11, 10*4+off+128(dst); \
MOVL R11, 0*4+off+192(dst); \
MOVL R11, 5*4+off+256(dst); \
MOVL R11, 9*4+off+320(dst); \
MOVL R11, 12*4+off+384(dst); \
MOVL R11, 1*4+off+448(dst); \
MOVL R11, 13*4+off+512(dst); \
MOVL R11, 2*4+off+576(dst); \
\
MOVL R12, 8*4+off+0(dst); \
MOVL R12, 5*4+off+64(dst); \
MOVL R12, 4*4+off+128(dst); \
MOVL R12, 15*4+off+192(dst); \
MOVL R12, 14*4+off+256(dst); \
MOVL R12, 3*4+off+320(dst); \
MOVL R12, 11*4+off+384(dst); \
MOVL R12, 10*4+off+448(dst); \
MOVL R12, 7*4+off+512(dst); \
MOVL R12, 1*4+off+576(dst); \
SHRQ $32, R12; \
MOVL R12, 12*4+off+0(dst); \
MOVL R12, 2*4+off+64(dst); \
MOVL R12, 11*4+off+128(dst); \
MOVL R12, 4*4+off+192(dst); \
MOVL R12, 0*4+off+256(dst); \
MOVL R12, 15*4+off+320(dst); \
MOVL R12, 10*4+off+384(dst); \
MOVL R12, 7*4+off+448(dst); \
MOVL R12, 5*4+off+512(dst); \
MOVL R12, 9*4+off+576(dst); \
\
MOVL R13, 9*4+off+0(dst); \
MOVL R13, 4*4+off+64(dst); \
MOVL R13, 8*4+off+128(dst); \
MOVL R13, 13*4+off+192(dst); \
MOVL R13, 3*4+off+256(dst); \
MOVL R13, 5*4+off+320(dst); \
MOVL R13, 7*4+off+384(dst); \
MOVL R13, 15*4+off+448(dst); \
MOVL R13, 11*4+off+512(dst); \
MOVL R13, 0*4+off+576(dst); \
SHRQ $32, R13; \
MOVL R13, 13*4+off+0(dst); \
MOVL R13, 10*4+off+64(dst); \
MOVL R13, 0*4+off+128(dst); \
MOVL R13, 3*4+off+192(dst); \
MOVL R13, 9*4+off+256(dst); \
MOVL R13, 6*4+off+320(dst); \
MOVL R13, 15*4+off+384(dst); \
MOVL R13, 4*4+off+448(dst); \
MOVL R13, 2*4+off+512(dst); \
MOVL R13, 12*4+off+576(dst); \
\
MOVL R14, 10*4+off+0(dst); \
MOVL R14, 12*4+off+64(dst); \
MOVL R14, 1*4+off+128(dst); \
MOVL R14, 6*4+off+192(dst); \
MOVL R14, 13*4+off+256(dst); \
MOVL R14, 4*4+off+320(dst); \
MOVL R14, 0*4+off+384(dst); \
MOVL R14, 2*4+off+448(dst); \
MOVL R14, 8*4+off+512(dst); \
MOVL R14, 14*4+off+576(dst); \
SHRQ $32, R14; \
MOVL R14, 14*4+off+0(dst); \
MOVL R14, 3*4+off+64(dst); \
MOVL R14, 7*4+off+128(dst); \
MOVL R14, 2*4+off+192(dst); \
MOVL R14, 15*4+off+256(dst); \
MOVL R14, 12*4+off+320(dst); \
MOVL R14, 6*4+off+384(dst); \
MOVL R14, 0*4+off+448(dst); \
MOVL R14, 9*4+off+512(dst); \
MOVL R14, 11*4+off+576(dst); \
\
MOVL R15, 11*4+off+0(dst); \
MOVL R15, 0*4+off+64(dst); \
MOVL R15, 12*4+off+128(dst); \
MOVL R15, 7*4+off+192(dst); \
MOVL R15, 8*4+off+256(dst); \
MOVL R15, 14*4+off+320(dst); \
MOVL R15, 2*4+off+384(dst); \
MOVL R15, 5*4+off+448(dst); \
MOVL R15, 1*4+off+512(dst); \
MOVL R15, 13*4+off+576(dst); \
SHRQ $32, R15; \
MOVL R15, 15*4+off+0(dst); \
MOVL R15, 6*4+off+64(dst); \
MOVL R15, 3*4+off+128(dst); \
MOVL R15, 11*4+off+192(dst); \
MOVL R15, 7*4+off+256(dst); \
MOVL R15, 10*4+off+320(dst); \
MOVL R15, 5*4+off+384(dst); \
MOVL R15, 9*4+off+448(dst); \
MOVL R15, 4*4+off+512(dst); \
MOVL R15, 8*4+off+576(dst)
#define BLAKE2s_SSE2() \
PRECOMPUTE_MSG(SP, 16, SI, R8, R9, R10, R11, R12, R13, R14, R15); \
ROUND_SSE2(X4, X5, X6, X7, 16(SP), 32(SP), 48(SP), 64(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+64(SP), 32+64(SP), 48+64(SP), 64+64(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+128(SP), 32+128(SP), 48+128(SP), 64+128(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+192(SP), 32+192(SP), 48+192(SP), 64+192(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+256(SP), 32+256(SP), 48+256(SP), 64+256(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+320(SP), 32+320(SP), 48+320(SP), 64+320(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+384(SP), 32+384(SP), 48+384(SP), 64+384(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+448(SP), 32+448(SP), 48+448(SP), 64+448(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+512(SP), 32+512(SP), 48+512(SP), 64+512(SP), X8); \
ROUND_SSE2(X4, X5, X6, X7, 16+576(SP), 32+576(SP), 48+576(SP), 64+576(SP), X8)
#define BLAKE2s_SSSE3() \
PRECOMPUTE_MSG(SP, 16, SI, R8, R9, R10, R11, R12, R13, R14, R15); \
ROUND_SSSE3(X4, X5, X6, X7, 16(SP), 32(SP), 48(SP), 64(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+64(SP), 32+64(SP), 48+64(SP), 64+64(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+128(SP), 32+128(SP), 48+128(SP), 64+128(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+192(SP), 32+192(SP), 48+192(SP), 64+192(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+256(SP), 32+256(SP), 48+256(SP), 64+256(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+320(SP), 32+320(SP), 48+320(SP), 64+320(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+384(SP), 32+384(SP), 48+384(SP), 64+384(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+448(SP), 32+448(SP), 48+448(SP), 64+448(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+512(SP), 32+512(SP), 48+512(SP), 64+512(SP), X8, X13, X14); \
ROUND_SSSE3(X4, X5, X6, X7, 16+576(SP), 32+576(SP), 48+576(SP), 64+576(SP), X8, X13, X14)
#define BLAKE2s_SSE4() \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 0, 2, 4, 6, 1, 3, 5, 7, 8, 10, 12, 14, 9, 11, 13, 15); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 14, 4, 9, 13, 10, 8, 15, 6, 1, 0, 11, 5, 12, 2, 7, 3); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 11, 12, 5, 15, 8, 0, 2, 13, 10, 3, 7, 9, 14, 6, 1, 4); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 7, 3, 13, 11, 9, 1, 12, 14, 2, 5, 4, 15, 6, 10, 0, 8); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 9, 5, 2, 10, 0, 7, 4, 15, 14, 11, 6, 3, 1, 12, 8, 13); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 2, 6, 0, 8, 12, 10, 11, 3, 4, 7, 15, 1, 13, 5, 14, 9); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 12, 1, 14, 4, 5, 15, 13, 10, 0, 6, 9, 8, 7, 3, 2, 11); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 13, 7, 12, 3, 11, 14, 1, 9, 5, 15, 8, 2, 0, 4, 6, 10); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 6, 14, 11, 0, 15, 9, 3, 8, 12, 13, 1, 10, 2, 7, 4, 5); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14); \
LOAD_MSG_SSE4(X8, X9, X10, X11, SI, 10, 8, 7, 1, 2, 4, 6, 5, 15, 9, 3, 13, 11, 14, 12, 0); \
ROUND_SSSE3(X4, X5, X6, X7, X8, X9, X10, X11, X8, X13, X14)
#define HASH_BLOCKS(h, c, flag, blocks_base, blocks_len, BLAKE2s_FUNC) \
MOVQ h, AX; \
MOVQ c, BX; \
MOVL flag, CX; \
MOVQ blocks_base, SI; \
MOVQ blocks_len, DX; \
\
MOVQ SP, BP; \
MOVQ SP, R9; \
ADDQ $15, R9; \
ANDQ $~15, R9; \
MOVQ R9, SP; \
\
MOVQ 0(BX), R9; \
MOVQ R9, 0(SP); \
XORQ R9, R9; \
MOVQ R9, 8(SP); \
MOVL CX, 8(SP); \
\
MOVOU 0(AX), X0; \
MOVOU 16(AX), X1; \
MOVOU iv0<>(SB), X2; \
MOVOU iv1<>(SB), X3 \
\
MOVOU counter<>(SB), X12; \
MOVOU rol16<>(SB), X13; \
MOVOU rol8<>(SB), X14; \
MOVO 0(SP), X15; \
\
loop: \
MOVO X0, X4; \
MOVO X1, X5; \
MOVO X2, X6; \
MOVO X3, X7; \
\
PADDQ X12, X15; \
PXOR X15, X7; \
\
BLAKE2s_FUNC(); \
\
PXOR X4, X0; \
PXOR X5, X1; \
PXOR X6, X0; \
PXOR X7, X1; \
\
LEAQ 64(SI), SI; \
SUBQ $64, DX; \
JNE loop; \
\
MOVO X15, 0(SP); \
MOVQ 0(SP), R9; \
MOVQ R9, 0(BX); \
\
MOVOU X0, 0(AX); \
MOVOU X1, 16(AX); \
\
MOVQ BP, SP
// func hashBlocksSSE2(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
TEXT ·hashBlocksSSE2(SB), 0, $672-48 // frame = 656 + 16 byte alignment
HASH_BLOCKS(h+0(FP), c+8(FP), flag+16(FP), blocks_base+24(FP), blocks_len+32(FP), BLAKE2s_SSE2)
RET
// func hashBlocksSSSE3(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
TEXT ·hashBlocksSSSE3(SB), 0, $672-48 // frame = 656 + 16 byte alignment
HASH_BLOCKS(h+0(FP), c+8(FP), flag+16(FP), blocks_base+24(FP), blocks_len+32(FP), BLAKE2s_SSSE3)
RET
// func hashBlocksSSE4(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte)
TEXT ·hashBlocksSSE4(SB), 0, $32-48 // frame = 16 + 16 byte alignment
HASH_BLOCKS(h+0(FP), c+8(FP), flag+16(FP), blocks_base+24(FP), blocks_len+32(FP), BLAKE2s_SSE4)
RET

174
vendor/golang.org/x/crypto/blake2s/blake2s_generic.go generated vendored Normal file
View File

@ -0,0 +1,174 @@
// 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 blake2s
// the precomputed values for BLAKE2s
// there are 10 16-byte arrays - one for each round
// the entries are calculated from the sigma constants.
var precomputed = [10][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},
}
func hashBlocksGeneric(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte) {
var m [16]uint32
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] = uint32(blocks[i]) | uint32(blocks[i+1])<<8 | uint32(blocks[i+2])<<16 | uint32(blocks[i+3])<<24
i += 4
}
for k := range precomputed {
s := &(precomputed[k])
v0 += m[s[0]]
v0 += v4
v12 ^= v0
v12 = v12<<(32-16) | v12>>16
v8 += v12
v4 ^= v8
v4 = v4<<(32-12) | v4>>12
v1 += m[s[1]]
v1 += v5
v13 ^= v1
v13 = v13<<(32-16) | v13>>16
v9 += v13
v5 ^= v9
v5 = v5<<(32-12) | v5>>12
v2 += m[s[2]]
v2 += v6
v14 ^= v2
v14 = v14<<(32-16) | v14>>16
v10 += v14
v6 ^= v10
v6 = v6<<(32-12) | v6>>12
v3 += m[s[3]]
v3 += v7
v15 ^= v3
v15 = v15<<(32-16) | v15>>16
v11 += v15
v7 ^= v11
v7 = v7<<(32-12) | v7>>12
v0 += m[s[4]]
v0 += v4
v12 ^= v0
v12 = v12<<(32-8) | v12>>8
v8 += v12
v4 ^= v8
v4 = v4<<(32-7) | v4>>7
v1 += m[s[5]]
v1 += v5
v13 ^= v1
v13 = v13<<(32-8) | v13>>8
v9 += v13
v5 ^= v9
v5 = v5<<(32-7) | v5>>7
v2 += m[s[6]]
v2 += v6
v14 ^= v2
v14 = v14<<(32-8) | v14>>8
v10 += v14
v6 ^= v10
v6 = v6<<(32-7) | v6>>7
v3 += m[s[7]]
v3 += v7
v15 ^= v3
v15 = v15<<(32-8) | v15>>8
v11 += v15
v7 ^= v11
v7 = v7<<(32-7) | v7>>7
v0 += m[s[8]]
v0 += v5
v15 ^= v0
v15 = v15<<(32-16) | v15>>16
v10 += v15
v5 ^= v10
v5 = v5<<(32-12) | v5>>12
v1 += m[s[9]]
v1 += v6
v12 ^= v1
v12 = v12<<(32-16) | v12>>16
v11 += v12
v6 ^= v11
v6 = v6<<(32-12) | v6>>12
v2 += m[s[10]]
v2 += v7
v13 ^= v2
v13 = v13<<(32-16) | v13>>16
v8 += v13
v7 ^= v8
v7 = v7<<(32-12) | v7>>12
v3 += m[s[11]]
v3 += v4
v14 ^= v3
v14 = v14<<(32-16) | v14>>16
v9 += v14
v4 ^= v9
v4 = v4<<(32-12) | v4>>12
v0 += m[s[12]]
v0 += v5
v15 ^= v0
v15 = v15<<(32-8) | v15>>8
v10 += v15
v5 ^= v10
v5 = v5<<(32-7) | v5>>7
v1 += m[s[13]]
v1 += v6
v12 ^= v1
v12 = v12<<(32-8) | v12>>8
v11 += v12
v6 ^= v11
v6 = v6<<(32-7) | v6>>7
v2 += m[s[14]]
v2 += v7
v13 ^= v2
v13 = v13<<(32-8) | v13>>8
v8 += v13
v7 ^= v8
v7 = v7<<(32-7) | v7>>7
v3 += m[s[15]]
v3 += v4
v14 ^= v3
v14 = v14<<(32-8) | v14>>8
v9 += v14
v4 ^= v9
v4 = v4<<(32-7) | v4>>7
}
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
}

17
vendor/golang.org/x/crypto/blake2s/blake2s_ref.go generated vendored Normal file
View File

@ -0,0 +1,17 @@
// 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.
// +build !amd64,!386 gccgo appengine
package blake2s
var (
useSSE4 = false
useSSSE3 = false
useSSE2 = false
)
func hashBlocks(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte) {
hashBlocksGeneric(h, c, flag, blocks)
}

1050
vendor/golang.org/x/crypto/blake2s/blake2s_test.go generated vendored Normal file
View File

File diff suppressed because it is too large Load Diff

178
vendor/golang.org/x/crypto/blake2s/blake2x.go generated vendored Normal file
View File

@ -0,0 +1,178 @@
// 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 blake2s
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 = 65535
// maxOutputLength is the absolute maximum number of bytes to produce when the
// number of output bytes is unknown.
const maxOutputLength = (1 << 32) * 32
// NewXOF creates a new variable-output-length hash. The hash either produce a
// known number of bytes (1 <= size < 65535), or an unknown number of bytes
// (size == OutputLengthUnknown). In the latter case, an absolute limit of
// 128GiB applies.
//
// A non-nil key turns the hash into a MAC. The key must between
// zero and 32 bytes long.
func NewXOF(size uint16, key []byte) (XOF, error) {
if len(key) > Size {
return nil, errKeySize
}
if size == magicUnknownOutputLength {
// 2^16-1 indicates an unknown number of bytes and thus isn't a
// valid length.
return nil, errors.New("blake2s: 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 uint16
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("blake2s: 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.PutUint16(x.cfg[12:], x.length) // XOF length
x.cfg[15] = byte(Size) // inner hash size
x.d.Reset()
x.d.h[3] ^= uint32(x.length)
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.Uint32(cfg[i*4:])
}
}

21
vendor/golang.org/x/crypto/blake2s/register.go generated vendored Normal file
View File

@ -0,0 +1,21 @@
// 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.
// +build go1.9
package blake2s
import (
"crypto"
"hash"
)
func init() {
newHash256 := func() hash.Hash {
h, _ := New256(nil)
return h
}
crypto.RegisterHash(crypto.BLAKE2s_256, newHash256)
}