try make blake2bip work
This commit is contained in:
John Tromp
parent
1788096b83
commit
6473d85b1f
2
Makefile
2
Makefile
|
|
@ -28,7 +28,7 @@ hash1: equi.h dev_miner.h dev_miner.cpp blake2b/asm/zcblake2_avx2.o Makefile
|
|||
equidev: equi.h equi_dev_miner.h equi_dev_miner.cpp Makefile
|
||||
$(GPP) -DATOMIC equi_dev_miner.cpp blake/blake2b.cpp blake2-avx2/blake2bip.c -o equidev
|
||||
|
||||
equidev1: equi.h equi_dev_miner.h equi_dev_miner.cpp Makefile
|
||||
equidev1: equi.h equi_dev_miner.h equi_dev_miner.cpp blake2-avx2/blake2bip.c Makefile
|
||||
$(GPP) equi_dev_miner.cpp blake/blake2b.cpp blake2-avx2/blake2bip.c -o equidev1
|
||||
|
||||
eqcuda: equi_miner.cu equi.h blake2b.cu Makefile
|
||||
|
|
|
|||
|
|
@ -0,0 +1,121 @@
|
|||
Creative Commons Legal Code
|
||||
|
||||
CC0 1.0 Universal
|
||||
|
||||
CREATIVE COMMONS CORPORATION IS NOT A LAW FIRM AND DOES NOT PROVIDE
|
||||
LEGAL SERVICES. DISTRIBUTION OF THIS DOCUMENT DOES NOT CREATE AN
|
||||
ATTORNEY-CLIENT RELATIONSHIP. CREATIVE COMMONS PROVIDES THIS
|
||||
INFORMATION ON AN "AS-IS" BASIS. CREATIVE COMMONS MAKES NO WARRANTIES
|
||||
REGARDING THE USE OF THIS DOCUMENT OR THE INFORMATION OR WORKS
|
||||
PROVIDED HEREUNDER, AND DISCLAIMS LIABILITY FOR DAMAGES RESULTING FROM
|
||||
THE USE OF THIS DOCUMENT OR THE INFORMATION OR WORKS PROVIDED
|
||||
HEREUNDER.
|
||||
|
||||
Statement of Purpose
|
||||
|
||||
The laws of most jurisdictions throughout the world automatically confer
|
||||
exclusive Copyright and Related Rights (defined below) upon the creator
|
||||
and subsequent owner(s) (each and all, an "owner") of an original work of
|
||||
authorship and/or a database (each, a "Work").
|
||||
|
||||
Certain owners wish to permanently relinquish those rights to a Work for
|
||||
the purpose of contributing to a commons of creative, cultural and
|
||||
scientific works ("Commons") that the public can reliably and without fear
|
||||
of later claims of infringement build upon, modify, incorporate in other
|
||||
works, reuse and redistribute as freely as possible in any form whatsoever
|
||||
and for any purposes, including without limitation commercial purposes.
|
||||
These owners may contribute to the Commons to promote the ideal of a free
|
||||
culture and the further production of creative, cultural and scientific
|
||||
works, or to gain reputation or greater distribution for their Work in
|
||||
part through the use and efforts of others.
|
||||
|
||||
For these and/or other purposes and motivations, and without any
|
||||
expectation of additional consideration or compensation, the person
|
||||
associating CC0 with a Work (the "Affirmer"), to the extent that he or she
|
||||
is an owner of Copyright and Related Rights in the Work, voluntarily
|
||||
elects to apply CC0 to the Work and publicly distribute the Work under its
|
||||
terms, with knowledge of his or her Copyright and Related Rights in the
|
||||
Work and the meaning and intended legal effect of CC0 on those rights.
|
||||
|
||||
1. Copyright and Related Rights. A Work made available under CC0 may be
|
||||
protected by copyright and related or neighboring rights ("Copyright and
|
||||
Related Rights"). Copyright and Related Rights include, but are not
|
||||
limited to, the following:
|
||||
|
||||
i. the right to reproduce, adapt, distribute, perform, display,
|
||||
communicate, and translate a Work;
|
||||
ii. moral rights retained by the original author(s) and/or performer(s);
|
||||
iii. publicity and privacy rights pertaining to a person's image or
|
||||
likeness depicted in a Work;
|
||||
iv. rights protecting against unfair competition in regards to a Work,
|
||||
subject to the limitations in paragraph 4(a), below;
|
||||
v. rights protecting the extraction, dissemination, use and reuse of data
|
||||
in a Work;
|
||||
vi. database rights (such as those arising under Directive 96/9/EC of the
|
||||
European Parliament and of the Council of 11 March 1996 on the legal
|
||||
protection of databases, and under any national implementation
|
||||
thereof, including any amended or successor version of such
|
||||
directive); and
|
||||
vii. other similar, equivalent or corresponding rights throughout the
|
||||
world based on applicable law or treaty, and any national
|
||||
implementations thereof.
|
||||
|
||||
2. Waiver. To the greatest extent permitted by, but not in contravention
|
||||
of, applicable law, Affirmer hereby overtly, fully, permanently,
|
||||
irrevocably and unconditionally waives, abandons, and surrenders all of
|
||||
Affirmer's Copyright and Related Rights and associated claims and causes
|
||||
of action, whether now known or unknown (including existing as well as
|
||||
future claims and causes of action), in the Work (i) in all territories
|
||||
worldwide, (ii) for the maximum duration provided by applicable law or
|
||||
treaty (including future time extensions), (iii) in any current or future
|
||||
medium and for any number of copies, and (iv) for any purpose whatsoever,
|
||||
including without limitation commercial, advertising or promotional
|
||||
purposes (the "Waiver"). Affirmer makes the Waiver for the benefit of each
|
||||
member of the public at large and to the detriment of Affirmer's heirs and
|
||||
successors, fully intending that such Waiver shall not be subject to
|
||||
revocation, rescission, cancellation, termination, or any other legal or
|
||||
equitable action to disrupt the quiet enjoyment of the Work by the public
|
||||
as contemplated by Affirmer's express Statement of Purpose.
|
||||
|
||||
3. Public License Fallback. Should any part of the Waiver for any reason
|
||||
be judged legally invalid or ineffective under applicable law, then the
|
||||
Waiver shall be preserved to the maximum extent permitted taking into
|
||||
account Affirmer's express Statement of Purpose. In addition, to the
|
||||
extent the Waiver is so judged Affirmer hereby grants to each affected
|
||||
person a royalty-free, non transferable, non sublicensable, non exclusive,
|
||||
irrevocable and unconditional license to exercise Affirmer's Copyright and
|
||||
Related Rights in the Work (i) in all territories worldwide, (ii) for the
|
||||
maximum duration provided by applicable law or treaty (including future
|
||||
time extensions), (iii) in any current or future medium and for any number
|
||||
of copies, and (iv) for any purpose whatsoever, including without
|
||||
limitation commercial, advertising or promotional purposes (the
|
||||
"License"). The License shall be deemed effective as of the date CC0 was
|
||||
applied by Affirmer to the Work. Should any part of the License for any
|
||||
reason be judged legally invalid or ineffective under applicable law, such
|
||||
partial invalidity or ineffectiveness shall not invalidate the remainder
|
||||
of the License, and in such case Affirmer hereby affirms that he or she
|
||||
will not (i) exercise any of his or her remaining Copyright and Related
|
||||
Rights in the Work or (ii) assert any associated claims and causes of
|
||||
action with respect to the Work, in either case contrary to Affirmer's
|
||||
express Statement of Purpose.
|
||||
|
||||
4. Limitations and Disclaimers.
|
||||
|
||||
a. No trademark or patent rights held by Affirmer are waived, abandoned,
|
||||
surrendered, licensed or otherwise affected by this document.
|
||||
b. Affirmer offers the Work as-is and makes no representations or
|
||||
warranties of any kind concerning the Work, express, implied,
|
||||
statutory or otherwise, including without limitation warranties of
|
||||
title, merchantability, fitness for a particular purpose, non
|
||||
infringement, or the absence of latent or other defects, accuracy, or
|
||||
the present or absence of errors, whether or not discoverable, all to
|
||||
the greatest extent permissible under applicable law.
|
||||
c. Affirmer disclaims responsibility for clearing rights of other persons
|
||||
that may apply to the Work or any use thereof, including without
|
||||
limitation any person's Copyright and Related Rights in the Work.
|
||||
Further, Affirmer disclaims responsibility for obtaining any necessary
|
||||
consents, permissions or other rights required for any use of the
|
||||
Work.
|
||||
d. Affirmer understands and acknowledges that Creative Commons is not a
|
||||
party to this document and has no duty or obligation with respect to
|
||||
this CC0 or use of the Work.
|
||||
|
|
@ -0,0 +1,48 @@
|
|||
CFLAGS=-std=c89 -Ofast -march=haswell -Wall -Wextra -pedantic -Wfatal-errors -Wno-long-long
|
||||
|
||||
all: check
|
||||
|
||||
bench: check
|
||||
$(SHELL) bench.sh
|
||||
|
||||
check: blake2b-check-1.exe blake2b-check-2.exe blake2b-check-3.exe blake2bp-check-1.exe blake2bp-check-2.exe blake2bp-check-3.exe blake2sp-check-1.exe blake2sp-check-2.exe blake2sp-check-3.exe
|
||||
|
||||
blake2b-check-1.exe: blake2b.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2B_SELFTEST -DPERMUTE_WITH_NOTHING -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2b-check-2.exe: blake2b.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2B_SELFTEST -DPERMUTE_WITH_SHUFFLES -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2b-check-3.exe: blake2b.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2B_SELFTEST -DPERMUTE_WITH_GATHER -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2bp-check-1.exe: blake2bp.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2BP_SELFTEST -DPERMUTE_WITH_NOTHING -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2bp-check-2.exe: blake2bp.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2BP_SELFTEST -DPERMUTE_WITH_SHUFFLES -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2bp-check-3.exe: blake2bp.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2BP_SELFTEST -DPERMUTE_WITH_GATHER -o $@ $<
|
||||
./$@
|
||||
|
||||
|
||||
blake2sp-check-1.exe: blake2sp.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2SP_SELFTEST -DPERMUTE_WITH_NOTHING -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2sp-check-2.exe: blake2sp.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2SP_SELFTEST -DPERMUTE_WITH_SHUFFLES -o $@ $<
|
||||
./$@
|
||||
|
||||
blake2sp-check-3.exe: blake2sp.c
|
||||
$(CC) $(CFLAGS) -DBLAKE2SP_SELFTEST -DPERMUTE_WITH_GATHER -o $@ $<
|
||||
./$@
|
||||
|
||||
clean:
|
||||
rm -f *.exe
|
||||
|
|
@ -0,0 +1,9 @@
|
|||
# BLAKE2 AVX2 implementations
|
||||
|
||||
This is **experimental** code implementing [BLAKE2](https://blake2.net/) using the AVX2 instruction set present in the Intel [Haswell](https://en.wikipedia.org/wiki/Haswell_%28microarchitecture%29) and later microarchitectures.
|
||||
|
||||
It currently implements BLAKE2b, BLAKE2bp, and BLAKE2sp using 3 similar but slightly different approaches: one lets the compiler choose how to permute the message, another one does it manually, and the final one uses the gather instructions introduced with AVX2. Current recorded speeds for long messages are:
|
||||
|
||||
- 3.19 cycles per byte on Haswell for BLAKE2b;
|
||||
- 1.45 cycles per byte on Haswell for BLAKE2bp;
|
||||
- 1.56 cycles per byte on Haswell for BLAKE2sp.
|
||||
|
|
@ -0,0 +1,156 @@
|
|||
/*
|
||||
BLAKE2 reference source code package - optimized C implementations
|
||||
|
||||
Written in 2012 by Samuel Neves <sneves@dei.uc.pt>
|
||||
|
||||
To the extent possible under law, the author(s) have dedicated all copyright
|
||||
and related and neighboring rights to this software to the public domain
|
||||
worldwide. This software is distributed without any warranty.
|
||||
|
||||
You should have received a copy of the CC0 Public Domain Dedication along with
|
||||
this software. If not, see <http://creativecommons.org/publicdomain/zero/1.0/>.
|
||||
*/
|
||||
#pragma once
|
||||
#ifndef __BLAKE2_H__
|
||||
#define __BLAKE2_H__
|
||||
|
||||
#include <stddef.h>
|
||||
#include <stdint.h>
|
||||
|
||||
#if defined(_MSC_VER)
|
||||
#define ALIGN(x) __declspec(align(x))
|
||||
#else
|
||||
#define ALIGN(x) __attribute__ ((__aligned__(x)))
|
||||
#endif
|
||||
|
||||
#if defined(__cplusplus)
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
enum blake2s_constant
|
||||
{
|
||||
BLAKE2S_BLOCKBYTES = 64,
|
||||
BLAKE2S_OUTBYTES = 32,
|
||||
BLAKE2S_KEYBYTES = 32,
|
||||
BLAKE2S_SALTBYTES = 8,
|
||||
BLAKE2S_PERSONALBYTES = 8
|
||||
};
|
||||
|
||||
enum blake2b_constant
|
||||
{
|
||||
BLAKE2B_BLOCKBYTES = 128,
|
||||
BLAKE2B_OUTBYTES = 64,
|
||||
BLAKE2B_KEYBYTES = 64,
|
||||
BLAKE2B_SALTBYTES = 16,
|
||||
BLAKE2B_PERSONALBYTES = 16
|
||||
};
|
||||
|
||||
#pragma pack(push, 1)
|
||||
typedef struct __blake2s_param
|
||||
{
|
||||
uint8_t digest_length; // 1
|
||||
uint8_t key_length; // 2
|
||||
uint8_t fanout; // 3
|
||||
uint8_t depth; // 4
|
||||
uint32_t leaf_length; // 8
|
||||
uint8_t node_offset[6];// 14
|
||||
uint8_t node_depth; // 15
|
||||
uint8_t inner_length; // 16
|
||||
// uint8_t reserved[0];
|
||||
uint8_t salt[BLAKE2S_SALTBYTES]; // 24
|
||||
uint8_t personal[BLAKE2S_PERSONALBYTES]; // 32
|
||||
} blake2s_param;
|
||||
|
||||
ALIGN( 64 ) typedef struct __blake2s_state
|
||||
{
|
||||
uint32_t h[8];
|
||||
uint32_t t[2];
|
||||
uint32_t f[2];
|
||||
uint8_t buf[2 * BLAKE2S_BLOCKBYTES];
|
||||
size_t buflen;
|
||||
uint8_t last_node;
|
||||
} blake2s_state;
|
||||
|
||||
typedef struct __blake2b_param
|
||||
{
|
||||
uint8_t digest_length; // 1
|
||||
uint8_t key_length; // 2
|
||||
uint8_t fanout; // 3
|
||||
uint8_t depth; // 4
|
||||
uint32_t leaf_length; // 8
|
||||
uint64_t node_offset; // 16
|
||||
uint8_t node_depth; // 17
|
||||
uint8_t inner_length; // 18
|
||||
uint8_t reserved[14]; // 32
|
||||
uint8_t salt[BLAKE2B_SALTBYTES]; // 48
|
||||
uint8_t personal[BLAKE2B_PERSONALBYTES]; // 64
|
||||
} blake2b_param;
|
||||
|
||||
ALIGN( 64 ) typedef struct __blake2b_state
|
||||
{
|
||||
uint64_t h[8];
|
||||
uint8_t buf[BLAKE2B_BLOCKBYTES];
|
||||
uint16_t counter;
|
||||
uint8_t buflen;
|
||||
uint8_t lastblock;
|
||||
} blake2b_state;
|
||||
|
||||
ALIGN( 64 ) typedef struct __blake2sp_state
|
||||
{
|
||||
blake2s_state S[8][1];
|
||||
blake2s_state R[1];
|
||||
uint8_t buf[8 * BLAKE2S_BLOCKBYTES];
|
||||
size_t buflen;
|
||||
} blake2sp_state;
|
||||
|
||||
ALIGN( 64 ) typedef struct __blake2bp_state
|
||||
{
|
||||
blake2b_state S[4][1];
|
||||
blake2b_state R[1];
|
||||
uint8_t buf[4 * BLAKE2B_BLOCKBYTES];
|
||||
size_t buflen;
|
||||
} blake2bp_state;
|
||||
#pragma pack(pop)
|
||||
|
||||
// Streaming API
|
||||
int blake2s_init( blake2s_state *S, const uint8_t outlen );
|
||||
int blake2s_init_key( blake2s_state *S, const uint8_t outlen, const void *key, const uint8_t keylen );
|
||||
int blake2s_init_param( blake2s_state *S, const blake2s_param *P );
|
||||
int blake2s_update( blake2s_state *S, const uint8_t *in, uint64_t inlen );
|
||||
int blake2s_final( blake2s_state *S, uint8_t *out, uint8_t outlen );
|
||||
|
||||
int blake2b_init( blake2b_state *S, const uint8_t outlen );
|
||||
int blake2b_init_key( blake2b_state *S, const uint8_t outlen, const void *key, const uint8_t keylen );
|
||||
int blake2b_init_param( blake2b_state *S, const blake2b_param *P );
|
||||
int blake2b_update( blake2b_state *S, const uint8_t *in, uint64_t inlen );
|
||||
int blake2b_final( blake2b_state *S, uint8_t *out, uint8_t outlen );
|
||||
|
||||
int blake2sp_init( blake2sp_state *S, const uint8_t outlen );
|
||||
int blake2sp_init_key( blake2sp_state *S, const uint8_t outlen, const void *key, const uint8_t keylen );
|
||||
int blake2sp_update( blake2sp_state *S, const uint8_t *in, uint64_t inlen );
|
||||
int blake2sp_final( blake2sp_state *S, uint8_t *out, uint8_t outlen );
|
||||
|
||||
int blake2bp_init( blake2bp_state *S, const uint8_t outlen );
|
||||
int blake2bp_init_key( blake2bp_state *S, const uint8_t outlen, const void *key, const uint8_t keylen );
|
||||
int blake2bp_update( blake2bp_state *S, const uint8_t *in, uint64_t inlen );
|
||||
int blake2bp_final( blake2bp_state *S, uint8_t *out, uint8_t outlen );
|
||||
|
||||
// Simple API
|
||||
int blake2s( uint8_t *out, const void *in, const void *key, const uint8_t outlen, const uint64_t inlen, uint8_t keylen );
|
||||
int blake2b( uint8_t *out, const void *in, const void *key, const uint8_t outlen, const uint64_t inlen, uint8_t keylen );
|
||||
int blake2b_long(uint8_t *out, const void *in, const uint32_t outlen, const uint64_t inlen);
|
||||
|
||||
int blake2sp( uint8_t *out, const void *in, const void *key, const uint8_t outlen, const uint64_t inlen, uint8_t keylen );
|
||||
int blake2bp( uint8_t *out, const void *in, const void *key, const uint8_t outlen, const uint64_t inlen, uint8_t keylen );
|
||||
|
||||
static inline int blake2( uint8_t *out, const void *in, const void *key, const uint8_t outlen, const uint64_t inlen, uint8_t keylen )
|
||||
{
|
||||
return blake2b( out, in, key, outlen, inlen, keylen );
|
||||
}
|
||||
|
||||
#if defined(__cplusplus)
|
||||
}
|
||||
#endif
|
||||
|
||||
#endif
|
||||
|
||||
|
|
@ -0,0 +1,60 @@
|
|||
#ifndef BLAKE2_AVX2_BLAKE2B_COMMON_H
|
||||
#define BLAKE2_AVX2_BLAKE2B_COMMON_H
|
||||
|
||||
#include <stddef.h>
|
||||
#include <stdint.h>
|
||||
#include <string.h>
|
||||
|
||||
#include <immintrin.h>
|
||||
|
||||
#include "blake2.h"
|
||||
|
||||
#define LOAD128(p) _mm_load_si128( (__m128i *)(p) )
|
||||
#define STORE128(p,r) _mm_store_si128((__m128i *)(p), r)
|
||||
|
||||
#define LOADU128(p) _mm_loadu_si128( (__m128i *)(p) )
|
||||
#define STOREU128(p,r) _mm_storeu_si128((__m128i *)(p), r)
|
||||
|
||||
#define LOAD(p) _mm256_load_si256( (__m256i *)(p) )
|
||||
#define STORE(p,r) _mm256_store_si256((__m256i *)(p), r)
|
||||
|
||||
#define LOADU(p) _mm256_loadu_si256( (__m256i *)(p) )
|
||||
#define STOREU(p,r) _mm256_storeu_si256((__m256i *)(p), r)
|
||||
|
||||
#if !defined(__cplusplus) && (!defined(__STDC_VERSION__) || __STDC_VERSION__ < 199901L)
|
||||
#if defined(_MSC_VER)
|
||||
#define INLINE __inline
|
||||
#elif defined(__GNUC__)
|
||||
#define INLINE __inline__
|
||||
#else
|
||||
#define INLINE
|
||||
#endif
|
||||
#else
|
||||
#define INLINE inline
|
||||
#endif
|
||||
|
||||
static INLINE uint64_t LOADU64(void const * p) {
|
||||
uint64_t v;
|
||||
memcpy(&v, p, sizeof v);
|
||||
return v;
|
||||
}
|
||||
|
||||
#define ROTATE16 _mm256_setr_epi8( 2, 3, 4, 5, 6, 7, 0, 1, 10, 11, 12, 13, 14, 15, 8, 9, \
|
||||
2, 3, 4, 5, 6, 7, 0, 1, 10, 11, 12, 13, 14, 15, 8, 9 )
|
||||
|
||||
#define ROTATE24 _mm256_setr_epi8( 3, 4, 5, 6, 7, 0, 1, 2, 11, 12, 13, 14, 15, 8, 9, 10, \
|
||||
3, 4, 5, 6, 7, 0, 1, 2, 11, 12, 13, 14, 15, 8, 9, 10 )
|
||||
|
||||
#define ADD(a, b) _mm256_add_epi64(a, b)
|
||||
#define SUB(a, b) _mm256_sub_epi64(a, b)
|
||||
|
||||
#define XOR(a, b) _mm256_xor_si256(a, b)
|
||||
#define AND(a, b) _mm256_and_si256(a, b)
|
||||
#define OR(a, b) _mm256_or_si256(a, b)
|
||||
|
||||
#define ROT32(x) _mm256_shuffle_epi32((x), _MM_SHUFFLE(2, 3, 0, 1))
|
||||
#define ROT24(x) _mm256_shuffle_epi8((x), ROTATE24)
|
||||
#define ROT16(x) _mm256_shuffle_epi8((x), ROTATE16)
|
||||
#define ROT63(x) _mm256_or_si256(_mm256_srli_epi64((x), 63), ADD((x), (x)))
|
||||
|
||||
#endif
|
||||
|
|
@ -0,0 +1,54 @@
|
|||
#ifndef BLAKE2_AVX2_BLAKE2B_LOAD_AVX2_SIMPLE_H
|
||||
#define BLAKE2_AVX2_BLAKE2B_LOAD_AVX2_SIMPLE_H
|
||||
|
||||
#define BLAKE2B_LOAD_MSG_0_1(b0) b0 = _mm256_set_epi64x(m6, m4, m2, m0);
|
||||
#define BLAKE2B_LOAD_MSG_0_2(b0) b0 = _mm256_set_epi64x(m7, m5, m3, m1);
|
||||
#define BLAKE2B_LOAD_MSG_0_3(b0) b0 = _mm256_set_epi64x(m14, m12, m10, m8);
|
||||
#define BLAKE2B_LOAD_MSG_0_4(b0) b0 = _mm256_set_epi64x(m15, m13, m11, m9);
|
||||
#define BLAKE2B_LOAD_MSG_1_1(b0) b0 = _mm256_set_epi64x(m13, m9, m4, m14);
|
||||
#define BLAKE2B_LOAD_MSG_1_2(b0) b0 = _mm256_set_epi64x(m6, m15, m8, m10);
|
||||
#define BLAKE2B_LOAD_MSG_1_3(b0) b0 = _mm256_set_epi64x(m5, m11, m0, m1);
|
||||
#define BLAKE2B_LOAD_MSG_1_4(b0) b0 = _mm256_set_epi64x(m3, m7, m2, m12);
|
||||
#define BLAKE2B_LOAD_MSG_2_1(b0) b0 = _mm256_set_epi64x(m15, m5, m12, m11);
|
||||
#define BLAKE2B_LOAD_MSG_2_2(b0) b0 = _mm256_set_epi64x(m13, m2, m0, m8);
|
||||
#define BLAKE2B_LOAD_MSG_2_3(b0) b0 = _mm256_set_epi64x(m9, m7, m3, m10);
|
||||
#define BLAKE2B_LOAD_MSG_2_4(b0) b0 = _mm256_set_epi64x(m4, m1, m6, m14);
|
||||
#define BLAKE2B_LOAD_MSG_3_1(b0) b0 = _mm256_set_epi64x(m11, m13, m3, m7);
|
||||
#define BLAKE2B_LOAD_MSG_3_2(b0) b0 = _mm256_set_epi64x(m14, m12, m1, m9);
|
||||
#define BLAKE2B_LOAD_MSG_3_3(b0) b0 = _mm256_set_epi64x(m15, m4, m5, m2);
|
||||
#define BLAKE2B_LOAD_MSG_3_4(b0) b0 = _mm256_set_epi64x(m8, m0, m10, m6);
|
||||
#define BLAKE2B_LOAD_MSG_4_1(b0) b0 = _mm256_set_epi64x(m10, m2, m5, m9);
|
||||
#define BLAKE2B_LOAD_MSG_4_2(b0) b0 = _mm256_set_epi64x(m15, m4, m7, m0);
|
||||
#define BLAKE2B_LOAD_MSG_4_3(b0) b0 = _mm256_set_epi64x(m3, m6, m11, m14);
|
||||
#define BLAKE2B_LOAD_MSG_4_4(b0) b0 = _mm256_set_epi64x(m13, m8, m12, m1);
|
||||
#define BLAKE2B_LOAD_MSG_5_1(b0) b0 = _mm256_set_epi64x(m8, m0, m6, m2);
|
||||
#define BLAKE2B_LOAD_MSG_5_2(b0) b0 = _mm256_set_epi64x(m3, m11, m10, m12);
|
||||
#define BLAKE2B_LOAD_MSG_5_3(b0) b0 = _mm256_set_epi64x(m1, m15, m7, m4);
|
||||
#define BLAKE2B_LOAD_MSG_5_4(b0) b0 = _mm256_set_epi64x(m9, m14, m5, m13);
|
||||
#define BLAKE2B_LOAD_MSG_6_1(b0) b0 = _mm256_set_epi64x(m4, m14, m1, m12);
|
||||
#define BLAKE2B_LOAD_MSG_6_2(b0) b0 = _mm256_set_epi64x(m10, m13, m15, m5);
|
||||
#define BLAKE2B_LOAD_MSG_6_3(b0) b0 = _mm256_set_epi64x(m8, m9, m6, m0);
|
||||
#define BLAKE2B_LOAD_MSG_6_4(b0) b0 = _mm256_set_epi64x(m11, m2, m3, m7);
|
||||
#define BLAKE2B_LOAD_MSG_7_1(b0) b0 = _mm256_set_epi64x(m3, m12, m7, m13);
|
||||
#define BLAKE2B_LOAD_MSG_7_2(b0) b0 = _mm256_set_epi64x(m9, m1, m14, m11);
|
||||
#define BLAKE2B_LOAD_MSG_7_3(b0) b0 = _mm256_set_epi64x(m2, m8, m15, m5);
|
||||
#define BLAKE2B_LOAD_MSG_7_4(b0) b0 = _mm256_set_epi64x(m10, m6, m4, m0);
|
||||
#define BLAKE2B_LOAD_MSG_8_1(b0) b0 = _mm256_set_epi64x(m0, m11, m14, m6);
|
||||
#define BLAKE2B_LOAD_MSG_8_2(b0) b0 = _mm256_set_epi64x(m8, m3, m9, m15);
|
||||
#define BLAKE2B_LOAD_MSG_8_3(b0) b0 = _mm256_set_epi64x(m10, m1, m13, m12);
|
||||
#define BLAKE2B_LOAD_MSG_8_4(b0) b0 = _mm256_set_epi64x(m5, m4, m7, m2);
|
||||
#define BLAKE2B_LOAD_MSG_9_1(b0) b0 = _mm256_set_epi64x(m1, m7, m8, m10);
|
||||
#define BLAKE2B_LOAD_MSG_9_2(b0) b0 = _mm256_set_epi64x(m5, m6, m4, m2);
|
||||
#define BLAKE2B_LOAD_MSG_9_3(b0) b0 = _mm256_set_epi64x(m13, m3, m9, m15);
|
||||
#define BLAKE2B_LOAD_MSG_9_4(b0) b0 = _mm256_set_epi64x(m0, m12, m14, m11);
|
||||
#define BLAKE2B_LOAD_MSG_10_1(b0) b0 = _mm256_set_epi64x(m6, m4, m2, m0);
|
||||
#define BLAKE2B_LOAD_MSG_10_2(b0) b0 = _mm256_set_epi64x(m7, m5, m3, m1);
|
||||
#define BLAKE2B_LOAD_MSG_10_3(b0) b0 = _mm256_set_epi64x(m14, m12, m10, m8);
|
||||
#define BLAKE2B_LOAD_MSG_10_4(b0) b0 = _mm256_set_epi64x(m15, m13, m11, m9);
|
||||
#define BLAKE2B_LOAD_MSG_11_1(b0) b0 = _mm256_set_epi64x(m13, m9, m4, m14);
|
||||
#define BLAKE2B_LOAD_MSG_11_2(b0) b0 = _mm256_set_epi64x(m6, m15, m8, m10);
|
||||
#define BLAKE2B_LOAD_MSG_11_3(b0) b0 = _mm256_set_epi64x(m5, m11, m0, m1);
|
||||
#define BLAKE2B_LOAD_MSG_11_4(b0) b0 = _mm256_set_epi64x(m3, m7, m2, m12);
|
||||
|
||||
#endif
|
||||
|
||||
|
|
@ -0,0 +1,344 @@
|
|||
#include <stddef.h>
|
||||
#include <stdint.h>
|
||||
#include <stdlib.h>
|
||||
#include <string.h>
|
||||
|
||||
#include "blake2.h"
|
||||
#include "blake2b-common.h"
|
||||
#include "blake2bip.h"
|
||||
|
||||
ALIGN(64) static const uint64_t blake2b_IV[8] = {
|
||||
UINT64_C(0x6A09E667F3BCC908), UINT64_C(0xBB67AE8584CAA73B),
|
||||
UINT64_C(0x3C6EF372FE94F82B), UINT64_C(0xA54FF53A5F1D36F1),
|
||||
UINT64_C(0x510E527FADE682D1), UINT64_C(0x9B05688C2B3E6C1F),
|
||||
UINT64_C(0x1F83D9ABFB41BD6B), UINT64_C(0x5BE0CD19137E2179),
|
||||
};
|
||||
|
||||
ALIGN(64) static const uint32_t blake2b_sigma[12][16] = {
|
||||
{ 0, 32, 64, 96, 128, 160, 192, 224, 256, 288, 320, 352, 384, 416, 448, 480},
|
||||
{448, 320, 128, 256, 288, 480, 416, 192, 32, 384, 0, 64, 352, 224, 160, 96},
|
||||
{352, 256, 384, 0, 160, 64, 480, 416, 320, 448, 96, 192, 224, 32, 288, 128},
|
||||
{224, 288, 96, 32, 416, 384, 352, 448, 64, 192, 160, 320, 128, 0, 480, 256},
|
||||
{288, 0, 160, 224, 64, 128, 320, 480, 448, 32, 352, 384, 192, 256, 96, 416},
|
||||
{ 64, 384, 192, 320, 0, 352, 256, 96, 128, 416, 224, 160, 480, 448, 32, 288},
|
||||
{384, 160, 32, 480, 448, 416, 128, 320, 0, 224, 192, 96, 288, 64, 256, 352},
|
||||
{416, 352, 224, 448, 384, 32, 96, 288, 160, 0, 480, 128, 256, 192, 64, 320},
|
||||
{192, 480, 448, 288, 352, 96, 0, 256, 384, 64, 416, 224, 32, 128, 320, 160},
|
||||
{320, 64, 256, 128, 224, 192, 32, 160, 480, 352, 288, 448, 96, 384, 416, 0},
|
||||
{ 0, 32, 64, 96, 128, 160, 192, 224, 256, 288, 320, 352, 384, 416, 448, 480},
|
||||
{448, 320, 128, 256, 288, 480, 416, 192, 32, 384, 0, 64, 352, 224, 160, 96},
|
||||
};
|
||||
|
||||
#define BLAKE2B_G1_V1(a, b, c, d, m) do { \
|
||||
a = ADD(a, m); \
|
||||
a = ADD(a, b); d = XOR(d, a); d = ROT32(d); \
|
||||
c = ADD(c, d); b = XOR(b, c); b = ROT24(b); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_G2_V1(a, b, c, d, m) do { \
|
||||
a = ADD(a, m); \
|
||||
a = ADD(a, b); d = XOR(d, a); d = ROT16(d); \
|
||||
c = ADD(c, d); b = XOR(b, c); b = ROT63(b); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_DIAG_V1(a, b, c, d) do { \
|
||||
d = _mm256_permute4x64_epi64(d, _MM_SHUFFLE(2,1,0,3)); \
|
||||
c = _mm256_permute4x64_epi64(c, _MM_SHUFFLE(1,0,3,2)); \
|
||||
b = _mm256_permute4x64_epi64(b, _MM_SHUFFLE(0,3,2,1)); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_UNDIAG_V1(a, b, c, d) do { \
|
||||
d = _mm256_permute4x64_epi64(d, _MM_SHUFFLE(0,3,2,1)); \
|
||||
c = _mm256_permute4x64_epi64(c, _MM_SHUFFLE(1,0,3,2)); \
|
||||
b = _mm256_permute4x64_epi64(b, _MM_SHUFFLE(2,1,0,3)); \
|
||||
} while(0)
|
||||
|
||||
#if defined(PERMUTE_WITH_SHUFFLES)
|
||||
#include "blake2b-load-avx2.h"
|
||||
#elif defined(PERMUTE_WITH_GATHER)
|
||||
#else
|
||||
#include "blake2b-load-avx2-simple.h"
|
||||
#endif
|
||||
|
||||
#if defined(PERMUTE_WITH_GATHER)
|
||||
ALIGN(64) static const uint32_t indices[12][16] = {
|
||||
{ 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},
|
||||
{14, 4, 9, 13,10, 8, 15, 6, 1, 0, 11, 5,12, 2, 7, 3},
|
||||
};
|
||||
|
||||
#define BLAKE2B_ROUND_V1(a, b, c, d, r, m) do { \
|
||||
__m256i b0; \
|
||||
b0 = _mm256_i32gather_epi64((void *)(m), LOAD128(&indices[r][ 0]), 8); \
|
||||
BLAKE2B_G1_V1(a, b, c, d, b0); \
|
||||
b0 = _mm256_i32gather_epi64((void *)(m), LOAD128(&indices[r][ 4]), 8); \
|
||||
BLAKE2B_G2_V1(a, b, c, d, b0); \
|
||||
BLAKE2B_DIAG_V1(a, b, c, d); \
|
||||
b0 = _mm256_i32gather_epi64((void *)(m), LOAD128(&indices[r][ 8]), 8); \
|
||||
BLAKE2B_G1_V1(a, b, c, d, b0); \
|
||||
b0 = _mm256_i32gather_epi64((void *)(m), LOAD128(&indices[r][12]), 8); \
|
||||
BLAKE2B_G2_V1(a, b, c, d, b0); \
|
||||
BLAKE2B_UNDIAG_V1(a, b, c, d); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_ROUNDS_V1(a, b, c, d, m) do { \
|
||||
int i; \
|
||||
for(i = 0; i < 12; ++i) { \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, i, m); \
|
||||
} \
|
||||
} while(0)
|
||||
#else /* !PERMUTE_WITH_GATHER */
|
||||
#define BLAKE2B_ROUND_V1(a, b, c, d, r, m) do { \
|
||||
__m256i b0; \
|
||||
BLAKE2B_LOAD_MSG_ ##r ##_1(b0); \
|
||||
BLAKE2B_G1_V1(a, b, c, d, b0); \
|
||||
BLAKE2B_LOAD_MSG_ ##r ##_2(b0); \
|
||||
BLAKE2B_G2_V1(a, b, c, d, b0); \
|
||||
BLAKE2B_DIAG_V1(a, b, c, d); \
|
||||
BLAKE2B_LOAD_MSG_ ##r ##_3(b0); \
|
||||
BLAKE2B_G1_V1(a, b, c, d, b0); \
|
||||
BLAKE2B_LOAD_MSG_ ##r ##_4(b0); \
|
||||
BLAKE2B_G2_V1(a, b, c, d, b0); \
|
||||
BLAKE2B_UNDIAG_V1(a, b, c, d); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_ROUNDS_V1(a, b, c, d, m) do { \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 0, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 1, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 2, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 3, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 4, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 5, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 6, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 7, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 8, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 9, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 10, (m)); \
|
||||
BLAKE2B_ROUND_V1(a, b, c, d, 11, (m)); \
|
||||
} while(0)
|
||||
#endif
|
||||
|
||||
#if defined(PERMUTE_WITH_GATHER)
|
||||
#define DECLARE_MESSAGE_WORDS(m)
|
||||
#elif defined(PERMUTE_WITH_SHUFFLES)
|
||||
#define DECLARE_MESSAGE_WORDS(m) \
|
||||
const __m256i m0 = _mm256_broadcastsi128_si256(LOADU128((m) + 0)); \
|
||||
const __m256i m1 = _mm256_broadcastsi128_si256(LOADU128((m) + 16)); \
|
||||
const __m256i m2 = _mm256_broadcastsi128_si256(LOADU128((m) + 32)); \
|
||||
const __m256i m3 = _mm256_broadcastsi128_si256(LOADU128((m) + 48)); \
|
||||
const __m256i m4 = _mm256_broadcastsi128_si256(LOADU128((m) + 64)); \
|
||||
const __m256i m5 = _mm256_broadcastsi128_si256(LOADU128((m) + 80)); \
|
||||
const __m256i m6 = _mm256_broadcastsi128_si256(LOADU128((m) + 96)); \
|
||||
const __m256i m7 = _mm256_broadcastsi128_si256(LOADU128((m) + 112)); \
|
||||
__m256i t0, t1;
|
||||
#else
|
||||
#define DECLARE_MESSAGE_WORDS(m) \
|
||||
const uint64_t m0 = LOADU64((m) + 0); \
|
||||
const uint64_t m1 = LOADU64((m) + 8); \
|
||||
const uint64_t m2 = LOADU64((m) + 16); \
|
||||
const uint64_t m3 = LOADU64((m) + 24); \
|
||||
const uint64_t m4 = LOADU64((m) + 32); \
|
||||
const uint64_t m5 = LOADU64((m) + 40); \
|
||||
const uint64_t m6 = LOADU64((m) + 48); \
|
||||
const uint64_t m7 = LOADU64((m) + 56); \
|
||||
const uint64_t m8 = LOADU64((m) + 64); \
|
||||
const uint64_t m9 = LOADU64((m) + 72); \
|
||||
const uint64_t m10 = LOADU64((m) + 80); \
|
||||
const uint64_t m11 = LOADU64((m) + 88); \
|
||||
const uint64_t m12 = LOADU64((m) + 96); \
|
||||
const uint64_t m13 = LOADU64((m) + 104); \
|
||||
const uint64_t m14 = LOADU64((m) + 112); \
|
||||
const uint64_t m15 = LOADU64((m) + 120);
|
||||
#endif
|
||||
|
||||
#define BLAKE2B_COMPRESS_V1(a, b, m, t0, t1, f0, f1) do { \
|
||||
DECLARE_MESSAGE_WORDS(m) \
|
||||
const __m256i iv0 = a; \
|
||||
const __m256i iv1 = b; \
|
||||
__m256i c = LOAD(&blake2b_IV[0]); \
|
||||
__m256i d = XOR( \
|
||||
LOAD(&blake2b_IV[4]), \
|
||||
_mm256_set_epi64x(f1, f0, t1, t0) \
|
||||
); \
|
||||
BLAKE2B_ROUNDS_V1(a, b, c, d, m); \
|
||||
a = XOR(a, c); \
|
||||
b = XOR(b, d); \
|
||||
a = XOR(a, iv0); \
|
||||
b = XOR(b, iv1); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_G_V4(m, r, i, a, b, c, d) do { \
|
||||
a = ADD(a, LOAD((uint8_t const *)(m) + blake2b_sigma[r][2*i+0])); \
|
||||
a = ADD(a, b); d = XOR(d, a); d = ROT32(d); \
|
||||
c = ADD(c, d); b = XOR(b, c); b = ROT24(b); \
|
||||
a = ADD(a, LOAD((uint8_t const *)(m) + blake2b_sigma[r][2*i+1])); \
|
||||
a = ADD(a, b); d = XOR(d, a); d = ROT16(d); \
|
||||
c = ADD(c, d); b = XOR(b, c); b = ROT63(b); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_ROUND_V4(v, m, r) do { \
|
||||
BLAKE2B_G_V4(m, r, 0, v[ 0], v[ 4], v[ 8], v[12]); \
|
||||
BLAKE2B_G_V4(m, r, 1, v[ 1], v[ 5], v[ 9], v[13]); \
|
||||
BLAKE2B_G_V4(m, r, 2, v[ 2], v[ 6], v[10], v[14]); \
|
||||
BLAKE2B_G_V4(m, r, 3, v[ 3], v[ 7], v[11], v[15]); \
|
||||
BLAKE2B_G_V4(m, r, 4, v[ 0], v[ 5], v[10], v[15]); \
|
||||
BLAKE2B_G_V4(m, r, 5, v[ 1], v[ 6], v[11], v[12]); \
|
||||
BLAKE2B_G_V4(m, r, 6, v[ 2], v[ 7], v[ 8], v[13]); \
|
||||
BLAKE2B_G_V4(m, r, 7, v[ 3], v[ 4], v[ 9], v[14]); \
|
||||
} while(0)
|
||||
|
||||
#if defined(PERMUTE_WITH_GATHER)
|
||||
#define BLAKE2B_LOADMSG_V4(w, m) do { \
|
||||
int i; \
|
||||
for(i = 0; i < 16; ++i) { \
|
||||
w[i] = _mm256_i32gather_epi64( \
|
||||
(const void *)((m) + i * sizeof(uint64_t)), \
|
||||
_mm_set_epi32(48, 32, 16, 0), \
|
||||
sizeof(uint64_t) \
|
||||
); \
|
||||
} \
|
||||
} while(0)
|
||||
#else
|
||||
#define BLAKE2B_PACK_MSG_V4(w, m) do { \
|
||||
__m256i t0, t1, t2, t3; \
|
||||
t0 = _mm256_unpacklo_epi64(m[ 0], m[ 4]); \
|
||||
t1 = _mm256_unpackhi_epi64(m[ 0], m[ 4]); \
|
||||
t2 = _mm256_unpacklo_epi64(m[ 8], m[12]); \
|
||||
t3 = _mm256_unpackhi_epi64(m[ 8], m[12]); \
|
||||
w[ 0] = _mm256_permute2x128_si256(t0, t2, 0x20); \
|
||||
w[ 2] = _mm256_permute2x128_si256(t0, t2, 0x31); \
|
||||
w[ 1] = _mm256_permute2x128_si256(t1, t3, 0x20); \
|
||||
w[ 3] = _mm256_permute2x128_si256(t1, t3, 0x31); \
|
||||
t0 = _mm256_unpacklo_epi64(m[ 1], m[ 5]); \
|
||||
t1 = _mm256_unpackhi_epi64(m[ 1], m[ 5]); \
|
||||
t2 = _mm256_unpacklo_epi64(m[ 9], m[13]); \
|
||||
t3 = _mm256_unpackhi_epi64(m[ 9], m[13]); \
|
||||
w[ 4] = _mm256_permute2x128_si256(t0, t2, 0x20); \
|
||||
w[ 6] = _mm256_permute2x128_si256(t0, t2, 0x31); \
|
||||
w[ 5] = _mm256_permute2x128_si256(t1, t3, 0x20); \
|
||||
w[ 7] = _mm256_permute2x128_si256(t1, t3, 0x31); \
|
||||
t0 = _mm256_unpacklo_epi64(m[ 2], m[ 6]); \
|
||||
t1 = _mm256_unpackhi_epi64(m[ 2], m[ 6]); \
|
||||
t2 = _mm256_unpacklo_epi64(m[10], m[14]); \
|
||||
t3 = _mm256_unpackhi_epi64(m[10], m[14]); \
|
||||
w[ 8] = _mm256_permute2x128_si256(t0, t2, 0x20); \
|
||||
w[10] = _mm256_permute2x128_si256(t0, t2, 0x31); \
|
||||
w[ 9] = _mm256_permute2x128_si256(t1, t3, 0x20); \
|
||||
w[11] = _mm256_permute2x128_si256(t1, t3, 0x31); \
|
||||
t0 = _mm256_unpacklo_epi64(m[ 3], m[ 7]); \
|
||||
t1 = _mm256_unpackhi_epi64(m[ 3], m[ 7]); \
|
||||
t2 = _mm256_unpacklo_epi64(m[11], m[15]); \
|
||||
t3 = _mm256_unpackhi_epi64(m[11], m[15]); \
|
||||
w[12] = _mm256_permute2x128_si256(t0, t2, 0x20); \
|
||||
w[14] = _mm256_permute2x128_si256(t0, t2, 0x31); \
|
||||
w[13] = _mm256_permute2x128_si256(t1, t3, 0x20); \
|
||||
w[15] = _mm256_permute2x128_si256(t1, t3, 0x31); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_LOADMSG_V4(w, m) do { \
|
||||
__m256i t[16]; \
|
||||
int i; \
|
||||
for(i = 0; i < 16; ++i) { \
|
||||
t[i] = LOADU((m) + i * 32); \
|
||||
} \
|
||||
BLAKE2B_PACK_MSG_V4(w, t); \
|
||||
} while(0)
|
||||
#endif
|
||||
|
||||
#define BLAKE2B_UNPACK_STATE_V4(u, v) do { \
|
||||
__m256i t0, t1, t2, t3; \
|
||||
t0 = _mm256_unpacklo_epi64(v[0], v[1]); \
|
||||
t1 = _mm256_unpackhi_epi64(v[0], v[1]); \
|
||||
t2 = _mm256_unpacklo_epi64(v[2], v[3]); \
|
||||
t3 = _mm256_unpackhi_epi64(v[2], v[3]); \
|
||||
u[0] = _mm256_permute2x128_si256(t0, t2, 0x20); \
|
||||
u[2] = _mm256_permute2x128_si256(t1, t3, 0x20); \
|
||||
u[4] = _mm256_permute2x128_si256(t0, t2, 0x31); \
|
||||
u[6] = _mm256_permute2x128_si256(t1, t3, 0x31); \
|
||||
t0 = _mm256_unpacklo_epi64(v[4], v[5]); \
|
||||
t1 = _mm256_unpackhi_epi64(v[4], v[5]); \
|
||||
t2 = _mm256_unpacklo_epi64(v[6], v[7]); \
|
||||
t3 = _mm256_unpackhi_epi64(v[6], v[7]); \
|
||||
u[1] = _mm256_permute2x128_si256(t0, t2, 0x20); \
|
||||
u[3] = _mm256_permute2x128_si256(t1, t3, 0x20); \
|
||||
u[5] = _mm256_permute2x128_si256(t0, t2, 0x31); \
|
||||
u[7] = _mm256_permute2x128_si256(t1, t3, 0x31); \
|
||||
} while(0)
|
||||
|
||||
#define BLAKE2B_COMPRESS_V4(v, m, counter, flag) do { \
|
||||
__m256i iv[8], w[16]; \
|
||||
int i, r; \
|
||||
for(i = 0; i < 8; ++i) { \
|
||||
iv[i] = v[i]; \
|
||||
} \
|
||||
v[ 8] = _mm256_set1_epi64x(blake2b_IV[0]); \
|
||||
v[ 9] = _mm256_set1_epi64x(blake2b_IV[1]); \
|
||||
v[10] = _mm256_set1_epi64x(blake2b_IV[2]); \
|
||||
v[11] = _mm256_set1_epi64x(blake2b_IV[3]); \
|
||||
v[12] = XOR(_mm256_set1_epi64x(blake2b_IV[4]), counter); \
|
||||
v[13] = _mm256_set1_epi64x(blake2b_IV[5]); \
|
||||
v[14] = XOR(_mm256_set1_epi64x(blake2b_IV[6]), flag); \
|
||||
v[15] = XOR(_mm256_set1_epi64x(blake2b_IV[7]), flag); \
|
||||
BLAKE2B_LOADMSG_V4(w, m); \
|
||||
for(r = 0; r < 12; ++r) { \
|
||||
BLAKE2B_ROUND_V4(v, w, r); \
|
||||
} \
|
||||
for(i = 0; i < 8; ++i) { \
|
||||
v[i] = XOR(XOR(v[i], v[i+8]), iv[i]); \
|
||||
} \
|
||||
} while(0)
|
||||
|
||||
void blake2bip_final(const blake2b_state *S, uchar *out, u32 blockidx) {
|
||||
ALIGN(64) uint8_t buffer[4 * BLAKE2B_BLOCKBYTES]; // 4 * 128
|
||||
memset(buffer, 0, 4 * BLAKE2B_BLOCKBYTES); // zero whole buffer
|
||||
memcpy(buffer, S->buf, S->buflen);
|
||||
uint32_t b = htole32(4 * blockidx);
|
||||
memcpy(buffer + S->buflen, &b, 4);
|
||||
size_t inlen = S->buflen + 4;
|
||||
buffer[0*128+inlen-4] += 0;
|
||||
memcpy(buffer + 128, buffer, inlen);
|
||||
buffer[1*128+inlen-4] += 1;
|
||||
memcpy(buffer + 256, buffer, inlen);
|
||||
buffer[2*128+inlen-4] += 2;
|
||||
memcpy(buffer + 384, buffer, inlen);
|
||||
buffer[3*128+inlen-4] += 3;
|
||||
|
||||
__m256i v[16], s[8]; // 16 * 32, 8 * 32
|
||||
union {
|
||||
__m256i v;
|
||||
uint64_t w[4];
|
||||
} counter;
|
||||
int i;
|
||||
|
||||
for(i = 0; i < 8; ++i) {
|
||||
v[i] = _mm256_set1_epi64x(S->h[i]); // initialize all 256/64 = 4 lanes
|
||||
}
|
||||
|
||||
#if 0 // these look blake2bp specific
|
||||
v[0] = XOR(v[0], _mm256_set1_epi64x(0x02040040UL));
|
||||
v[1] = XOR(v[1], _mm256_set_epi64x(3, 2, 1, 0));
|
||||
v[2] = XOR(v[2], _mm256_set1_epi64x(0x00004000UL));
|
||||
#endif
|
||||
|
||||
__m256i x4 = _mm256_set_epi64x( 0, 0, 0,inlen);
|
||||
__m256i f0 = _mm256_set_epi64x(~0, ~0, ~0, ~0);
|
||||
|
||||
counter.v = _mm256_add_epi64(counter.v, x4);
|
||||
|
||||
BLAKE2B_COMPRESS_V4(v, buffer, counter.v, f0);
|
||||
|
||||
BLAKE2B_UNPACK_STATE_V4(s, v);
|
||||
|
||||
// big loop obsoleted as it would if(!inlen) break
|
||||
|
||||
memcpy(out, s, 256); // instead of return blake2b_root(out, (void *)s);
|
||||
}
|
||||
|
|
@ -0,0 +1,11 @@
|
|||
#ifndef BLAKE2_AVX2_BLAKE2BIP_H
|
||||
#define BLAKE2_AVX2_BLAKE2BIP_H
|
||||
|
||||
#include <stddef.h>
|
||||
|
||||
typedef uint32_t u32;
|
||||
typedef unsigned char uchar;
|
||||
|
||||
void blake2bip_final(const blake2b_state *midstate, uchar *hashout, u32 blockidx);
|
||||
|
||||
#endif
|
||||
|
|
@ -523,9 +523,6 @@ struct equi {
|
|||
#endif
|
||||
uchar hashes[BLAKESINPARALLEL * 64];
|
||||
blake2b_state state = blake_ctx;
|
||||
state->counter += state->buflen;
|
||||
state->lastblock = 1;
|
||||
memset(state->buf+state->buflen, 0, BLAKE2B_BLOCKBYTES-state->buflen); // Padding
|
||||
for (u32 block = id; block < NBLOCKS; block += nthreads) {
|
||||
blake2bip_final(&state, hashes, block);
|
||||
#ifndef HASHONLY
|
||||
|
|
|
|||
Loading…
Reference in New Issue