Browse Source

First!

tags/v2.0.0
fireice-uk 5 years ago
parent
commit
6368ab7455
72 changed files with 22860 additions and 54 deletions
  1. +23
    -0
      CMakeLists.txt
  2. +0
    -54
      LICENSE
  3. +73
    -0
      cli-miner.cpp
  4. +87
    -0
      config.jscfg
  5. +186
    -0
      console.cpp
  6. +41
    -0
      console.h
  7. +326
    -0
      crypto/c_blake256.c
  8. +43
    -0
      crypto/c_blake256.h
  9. +360
    -0
      crypto/c_groestl.c
  10. +60
    -0
      crypto/c_groestl.h
  11. +367
    -0
      crypto/c_jh.c
  12. +19
    -0
      crypto/c_jh.h
  13. +128
    -0
      crypto/c_keccak.c
  14. +26
    -0
      crypto/c_keccak.h
  15. +2036
    -0
      crypto/c_skein.c
  16. +47
    -0
      crypto/c_skein.h
  17. +33
    -0
      crypto/cryptonight.h
  18. +344
    -0
      crypto/cryptonight_aesni.c
  19. +169
    -0
      crypto/cryptonight_common.c
  20. +38
    -0
      crypto/groestl_tables.h
  21. +5
    -0
      crypto/hash.h
  22. +153
    -0
      crypto/int-util.h
  23. +179
    -0
      crypto/skein_port.h
  24. +11
    -0
      donate-level.h
  25. +602
    -0
      executor.cpp
  26. +148
    -0
      executor.h
  27. +310
    -0
      jconf.cpp
  28. +52
    -0
      jconf.h
  29. +13
    -0
      jext.h
  30. +677
    -0
      jpsock.cpp
  31. +96
    -0
      jpsock.h
  32. +433
    -0
      minethd.cpp
  33. +123
    -0
      minethd.h
  34. +131
    -0
      msgstruct.h
  35. +271
    -0
      rapidjson/allocators.h
  36. +2602
    -0
      rapidjson/document.h
  37. +299
    -0
      rapidjson/encodedstream.h
  38. +716
    -0
      rapidjson/encodings.h
  39. +74
    -0
      rapidjson/error/en.h
  40. +155
    -0
      rapidjson/error/error.h
  41. +99
    -0
      rapidjson/filereadstream.h
  42. +104
    -0
      rapidjson/filewritestream.h
  43. +151
    -0
      rapidjson/fwd.h
  44. +290
    -0
      rapidjson/internal/biginteger.h
  45. +258
    -0
      rapidjson/internal/diyfp.h
  46. +245
    -0
      rapidjson/internal/dtoa.h
  47. +78
    -0
      rapidjson/internal/ieee754.h
  48. +304
    -0
      rapidjson/internal/itoa.h
  49. +181
    -0
      rapidjson/internal/meta.h
  50. +55
    -0
      rapidjson/internal/pow10.h
  51. +731
    -0
      rapidjson/internal/regex.h
  52. +230
    -0
      rapidjson/internal/stack.h
  53. +58
    -0
      rapidjson/internal/strfunc.h
  54. +269
    -0
      rapidjson/internal/strtod.h
  55. +46
    -0
      rapidjson/internal/swap.h
  56. +115
    -0
      rapidjson/istreamwrapper.h
  57. +70
    -0
      rapidjson/memorybuffer.h
  58. +71
    -0
      rapidjson/memorystream.h
  59. +316
    -0
      rapidjson/msinttypes/inttypes.h
  60. +300
    -0
      rapidjson/msinttypes/stdint.h
  61. +81
    -0
      rapidjson/ostreamwrapper.h
  62. +1358
    -0
      rapidjson/pointer.h
  63. +275
    -0
      rapidjson/prettywriter.h
  64. +615
    -0
      rapidjson/rapidjson.h
  65. +1862
    -0
      rapidjson/reader.h
  66. +2024
    -0
      rapidjson/schema.h
  67. +179
    -0
      rapidjson/stream.h
  68. +121
    -0
      rapidjson/stringbuffer.h
  69. +624
    -0
      rapidjson/writer.h
  70. +88
    -0
      socks.h
  71. +49
    -0
      thdq.hpp
  72. +157
    -0
      xmr-stak-cpu.cbp

+ 23
- 0
CMakeLists.txt View File

@@ -0,0 +1,23 @@
project(xmr-stak-cpu)

cmake_minimum_required(VERSION 2.8.10)

if("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
if (CMAKE_CXX_COMPILER_VERSION VERSION_LESS 5.1)
message(FATAL_ERROR "GCC version must be at least 5.1!")
endif()
endif()

#SET(CMAKE_VERBOSE_MAKEFILE ON)
SET(CMAKE_C_FLAGS "-DNDEBUG -march=westmere -O3 -m64")
SET(CMAKE_CXX_FLAGS "${CMAKE_C_FLAGS} -std=c++11")
SET(CMAKE_SHARED_LIBRARY_LINK_C_FLAGS "-s")
SET(CMAKE_SHARED_LIBRARY_LINK_CXX_FLAGS "-s")
set(EXECUTABLE_OUTPUT_PATH "bin")

file(GLOB SOURCES "crypto/*.c" "*.cpp")

add_executable(xmr-stak-cpu ${SOURCES})
target_link_libraries(xmr-stak-cpu pthread)


+ 0
- 54
LICENSE View File

@@ -618,57 +618,3 @@ an absolute waiver of all civil liability in connection with the
Program, unless a warranty or assumption of liability accompanies a
copy of the Program in return for a fee.

END OF TERMS AND CONDITIONS

How to Apply These Terms to Your New Programs

If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.

To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
state the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.

{one line to give the program's name and a brief idea of what it does.}
Copyright (C) {year} {name of author}

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.

Also add information on how to contact you by electronic and paper mail.

If the program does terminal interaction, make it output a short
notice like this when it starts in an interactive mode:

{project} Copyright (C) {year} {fullname}
This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.

The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, your program's commands
might be different; for a GUI interface, you would use an "about box".

You should also get your employer (if you work as a programmer) or school,
if any, to sign a "copyright disclaimer" for the program, if necessary.
For more information on this, and how to apply and follow the GNU GPL, see
<http://www.gnu.org/licenses/>.

The GNU General Public License does not permit incorporating your program
into proprietary programs. If your program is a subroutine library, you
may consider it more useful to permit linking proprietary applications with
the library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License. But first, please read
<http://www.gnu.org/philosophy/why-not-lgpl.html>.

+ 73
- 0
cli-miner.cpp View File

@@ -0,0 +1,73 @@
/*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/

#include "executor.h"
#include "minethd.h"
#include "jconf.h"
#include "console.h"
#include "donate-level.h"

#include <stdlib.h>
#include <stdio.h>

int main(int argc, char *argv[])
{
if(argc != 2 || strcmp(argv[1], "-h") == 0)
{
printer::inst()->print_msg(L0, "Usage %s [CONFIG FILE]", argv[0]);
return 0;
}

if(!jconf::inst()->parse_config(argv[1]))
return 0;


printer::inst()->print_str("-------------------------------------------------------------------\n");
printer::inst()->print_str("XMR-Stak-CPU mining software, CPU Version.\n");
printer::inst()->print_str("Based on CPU mining code by wolf9466 (heavily optimized by myself).\n");
printer::inst()->print_str("Brought to you by fireice_uk under GPLv3.\n");
char buffer[64];
snprintf(buffer, sizeof(buffer), "Configurable dev donation level is set to %.1f %%\n", fDevDonationLevel * 100.0);
printer::inst()->print_str(buffer);
printer::inst()->print_str("-------------------------------------------------------------------\n");

if (!minethd::self_test())
return 0;

executor::inst()->ex_start();

int key;
while(true)
{
key = get_key();

switch(key)
{
case 'h':
executor::inst()->push_event(ex_event(EV_USR_HASHRATE));
break;
case 'r':
executor::inst()->push_event(ex_event(EV_USR_RESULTS));
break;
case 'c':
executor::inst()->push_event(ex_event(EV_USR_CONNSTAT));
break;
default:
break;
}
}

return 0;
}

+ 87
- 0
config.jscfg View File

@@ -0,0 +1,87 @@
/*
* Number of threads. You can configure them below. Cryptonight uses 2MB of memory, so the optimal setting
* here is the size of your L3 cache divided by 2. Intel mid-to-high end desktop processors have 2MB of L3
* cache per physical core. Low end cpus can have 1.5 or 1 MB while Xeons can have 2, 2.5 or 3MB per core.
*/
"cpu_thread_num" : 2,

/*
* Thread configuration for each thread. Make sure it matches the number above.
* low_power_mode will double the cache usage, and double the single thread performance. It will consume much
* less power (as less cores are working), but will max out at around 80-85% of the maximum performance.
* affine_to_cpu can be either false (no affinity), or the CPU core number. Note that on hyperthreading systems
* it is better to assign threads to physical cores. On Windows this usually means selecting even or odd numbered
* cpu numbers. For Linux it will be usually the lower CPU numbers, so for a 4 physical core CPU you should select
* cpu numbers 0-3.
*/
"cpu_threads_conf" : [
{ "low_power_mode" : false, "affine_to_cpu" : 0 },
{ "low_power_mode" : false, "affine_to_cpu" : 1 },
],

/*
* LARGE PAGE SUPPORT
* By default we will try to allocate large pages. This means you need to "Run As Administrator" on Windows.
* On Linux you will need to configure large page support "sudo sysctl -w vm.nr_hugepages=128" and increase your
* ulimit -l. To do do this you need to add following lines to /etc/security/limits.conf - "* soft memlock 262144"
* and "* hard memlock 262144". You can also do it Windows-style and simply run-as-root, but this is NOT
* recommended for security reasons.
*
* Memory locking means that the kernel can't swap out the page to disk - something that is unlikey to happen on a
* command line system that isn't starved of memory. I haven't observed any difference on a CLI Linux system between
* locked and unlocked memory. If that is your setup see option "no_mlck".
*/

/*
* use_slow_memory defines our behaviour with regards to large pages. There are three possible options here:
* always - Don't even try to use large pages. Always use slow memory.
* warn - We will try to use large pages, but fall back to slow memory if that fails.
* no_mlck - This option is only relevant on Linux, where we can use large pages without locking memory.
* It will never use slow memory, but it won't attempt to mlock
* never - If we fail to allocate large pages we will print an error and exit.
*/
"use_slow_memory" : "warn",

/*
* pool_address - Pool address should be in the form "pool.supportxmr.com:3333". Only stratum pools are supported.
* wallet_address - Your wallet, or pool login.
* pool_password - Can be empty in most cases or "x".
*/
"pool_address" : "pool.supportxmr.com:3333",
"wallet_address" : "",
"pool_password" : "",

/*
* Network timeouts.
* Because of the way this client is written it doesn't need to constantly talk (keep-alive) to the server to make
* sure it is there. We detect a buggy / overloaded server by the call timeout. The default values will be ok for
* nearly all cases. If they aren't the pool has most likely overload issues. Low call timeout values are preferable -
* long timeouts mean that we waste hashes on potentially stale jobs. Connection report will tell you how long the
* server usually takes to process our calls.
*
* call_timeout - How long should we wait for a response from the server before we assume it is dead and drop the connection.
* retry_time - How long should we wait before another connection attempt.
* Both values are in seconds.
*/
"call_timeout" : 10,
"retry_time" : 10,

/*
* Output control.
* Since most people are used to miners printing all the time, that's what we do by default too. This is suboptimal
* really, since you cannot see errors under pages and pages of text and performance stats. Given that we have internal
* performance monitors, there is very little reason to spew out pages of text instead of concise reports.
* Press 'h' (hashrate), 'r' (results) or 'c' (connection) to print reports.
*
* verbose_level - 0 - Don't print anything.
* 1 - Print intro, connection event, disconnect event
* 2 - All of level 1, and new job (block) event if the difficulty is different from the last job
* 3 - All of level 1, and new job (block) event in all cases, result submission event.
*/
"verbose_level" : 3,

/*
* prefer_ipv4 - IPv6 preference. If the host is available on both IPv4 and IPv6 net, which one should be choose?
* This setting will only be needed in 2020's. No need to worry about it now.
*/
"prefer_ipv4" : true,

+ 186
- 0
console.cpp View File

@@ -0,0 +1,186 @@
/*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/

#include "console.h"
#include <time.h>
#include <stdio.h>
#include <string.h>
#include <stdarg.h>

#ifdef _WIN32
#include <windows.h>

int get_key()
{
DWORD mode, rd;
HANDLE h;

if ((h = GetStdHandle(STD_INPUT_HANDLE)) == NULL)
return -1;

GetConsoleMode( h, &mode );
SetConsoleMode( h, mode & ~(ENABLE_LINE_INPUT | ENABLE_ECHO_INPUT) );

int c = 0;
ReadConsole( h, &c, 1, &rd, NULL );
SetConsoleMode( h, mode );

return c;
}

void set_colour(out_colours cl)
{
WORD attr = 0;

switch(cl)
{
case K_RED:
attr = FOREGROUND_RED | FOREGROUND_INTENSITY;
break;
case K_GREEN:
attr = FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
case K_BLUE:
attr = FOREGROUND_BLUE | FOREGROUND_INTENSITY;
break;
case K_YELLOW:
attr = FOREGROUND_RED | FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
case K_CYAN:
attr = FOREGROUND_BLUE | FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
case K_MAGENTA:
attr = FOREGROUND_BLUE | FOREGROUND_RED | FOREGROUND_INTENSITY;
break;
case K_WHITE:
attr = FOREGROUND_BLUE | FOREGROUND_RED | FOREGROUND_GREEN | FOREGROUND_INTENSITY;
break;
default:
break;
}

SetConsoleTextAttribute(GetStdHandle(STD_OUTPUT_HANDLE), attr);
}

void reset_colour()
{
SetConsoleTextAttribute(GetStdHandle(STD_OUTPUT_HANDLE), FOREGROUND_RED | FOREGROUND_GREEN | FOREGROUND_BLUE);
}

#else
#include <termios.h>
#include <unistd.h>
#include <stdio.h>

int get_key()
{
struct termios oldattr, newattr;
int ch;
tcgetattr( STDIN_FILENO, &oldattr );
newattr = oldattr;
newattr.c_lflag &= ~( ICANON | ECHO );
tcsetattr( STDIN_FILENO, TCSANOW, &newattr );
ch = getchar();
tcsetattr( STDIN_FILENO, TCSANOW, &oldattr );
return ch;
}

void set_colour(out_colours cl)
{
switch(cl)
{
case K_RED:
fputs("\x1B[1;31m", stdout);
break;
case K_GREEN:
fputs("\x1B[1;32m", stdout);
break;
case K_BLUE:
fputs("\x1B[1;34m", stdout);
break;
case K_YELLOW:
fputs("\x1B[1;33m", stdout);
break;
case K_CYAN:
fputs("\x1B[1;36m", stdout);
break;
case K_MAGENTA:
fputs("\x1B[1;35m", stdout);
break;
case K_WHITE:
fputs("\x1B[1;37m", stdout);
break;
default:
break;
}
}

void reset_colour()
{
fputs("\x1B[0m", stdout);
}
#endif // _WIN32

inline void comp_localtime(const time_t* ctime, tm* stime)
{
#ifdef _WIN32
localtime_s(stime, ctime);
#else
localtime_r(ctime, stime);
#endif // __WIN32
}

printer* printer::oInst = nullptr;

printer::printer()
{
verbose_level = LINF;
}

void printer::print_msg(verbosity verbose, const char* fmt, ...)
{
if(verbose > verbose_level)
return;

char buf[1024];
size_t bpos;
tm stime;

time_t now = time(nullptr);
comp_localtime(&now, &stime);
strftime(buf, sizeof(buf), "[%F %T] : ", &stime);
bpos = strlen(buf);

va_list args;
va_start(args, fmt);
vsnprintf(buf+bpos, sizeof(buf)-bpos, fmt, args);
va_end(args);
bpos = strlen(buf);

if(bpos+2 >= sizeof(buf))
return;

buf[bpos] = '\n';
buf[bpos+1] = '\0';

std::unique_lock<std::mutex> lck(print_mutex);
fputs(buf, stdout);
}

void printer::print_str(const char* str)
{
std::unique_lock<std::mutex> lck(print_mutex);
fputs(str, stdout);
}

+ 41
- 0
console.h View File

@@ -0,0 +1,41 @@
#pragma once
#include <mutex>

enum out_colours { K_RED, K_GREEN, K_BLUE, K_YELLOW, K_CYAN, K_MAGENTA, K_WHITE, K_NONE };

// Warning - on Linux get_key will detect control keys, but not on Windows.
// We will only use it for alphanum keys anyway.
int get_key();

void set_colour(out_colours cl);
void reset_colour();

// on MSVC sizeof(long int) = 4, gcc sizeof(long int) = 8, this is the workaround
// now we can use %llu on both compilers
inline long long unsigned int int_port(size_t i)
{
return i;
}

enum verbosity : size_t { L0 = 0, L1 = 1, L2 = 2, L3 = 3, LINF = 100};

class printer
{
public:
static inline printer* inst()
{
if (oInst == nullptr) oInst = new printer;
return oInst;
};

inline void set_verbose_level(size_t level) { verbose_level = (verbosity)level; }
void print_msg(verbosity verbose, const char* fmt, ...);
void print_str(const char* str);

private:
printer();
static printer* oInst;

std::mutex print_mutex;
verbosity verbose_level;
};

+ 326
- 0
crypto/c_blake256.c View File

@@ -0,0 +1,326 @@
/*
* The blake256_* and blake224_* functions are largely copied from
* blake256_light.c and blake224_light.c from the BLAKE website:
*
* http://131002.net/blake/
*
* The hmac_* functions implement HMAC-BLAKE-256 and HMAC-BLAKE-224.
* HMAC is specified by RFC 2104.
*/

#include <string.h>
#include <stdio.h>
#include <stdint.h>
#include "c_blake256.h"

#define U8TO32(p) \
(((uint32_t)((p)[0]) << 24) | ((uint32_t)((p)[1]) << 16) | \
((uint32_t)((p)[2]) << 8) | ((uint32_t)((p)[3]) ))
#define U32TO8(p, v) \
(p)[0] = (uint8_t)((v) >> 24); (p)[1] = (uint8_t)((v) >> 16); \
(p)[2] = (uint8_t)((v) >> 8); (p)[3] = (uint8_t)((v) );

const uint8_t sigma[][16] = {
{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14,15},
{14,10, 4, 8, 9,15,13, 6, 1,12, 0, 2,11, 7, 5, 3},
{11, 8,12, 0, 5, 2,15,13,10,14, 3, 6, 7, 1, 9, 4},
{ 7, 9, 3, 1,13,12,11,14, 2, 6, 5,10, 4, 0,15, 8},
{ 9, 0, 5, 7, 2, 4,10,15,14, 1,11,12, 6, 8, 3,13},
{ 2,12, 6,10, 0,11, 8, 3, 4,13, 7, 5,15,14, 1, 9},
{12, 5, 1,15,14,13, 4,10, 0, 7, 6, 3, 9, 2, 8,11},
{13,11, 7,14,12, 1, 3, 9, 5, 0,15, 4, 8, 6, 2,10},
{ 6,15,14, 9,11, 3, 0, 8,12, 2,13, 7, 1, 4,10, 5},
{10, 2, 8, 4, 7, 6, 1, 5,15,11, 9,14, 3,12,13, 0},
{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14,15},
{14,10, 4, 8, 9,15,13, 6, 1,12, 0, 2,11, 7, 5, 3},
{11, 8,12, 0, 5, 2,15,13,10,14, 3, 6, 7, 1, 9, 4},
{ 7, 9, 3, 1,13,12,11,14, 2, 6, 5,10, 4, 0,15, 8}
};

const uint32_t cst[16] = {
0x243F6A88, 0x85A308D3, 0x13198A2E, 0x03707344,
0xA4093822, 0x299F31D0, 0x082EFA98, 0xEC4E6C89,
0x452821E6, 0x38D01377, 0xBE5466CF, 0x34E90C6C,
0xC0AC29B7, 0xC97C50DD, 0x3F84D5B5, 0xB5470917
};

static const uint8_t padding[] = {
0x80,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
};


void blake256_compress(state *S, const uint8_t *block) {
uint32_t v[16], m[16], i;

#define ROT(x,n) (((x)<<(32-n))|((x)>>(n)))
#define G(a,b,c,d,e) \
v[a] += (m[sigma[i][e]] ^ cst[sigma[i][e+1]]) + v[b]; \
v[d] = ROT(v[d] ^ v[a],16); \
v[c] += v[d]; \
v[b] = ROT(v[b] ^ v[c],12); \
v[a] += (m[sigma[i][e+1]] ^ cst[sigma[i][e]])+v[b]; \
v[d] = ROT(v[d] ^ v[a], 8); \
v[c] += v[d]; \
v[b] = ROT(v[b] ^ v[c], 7);

for (i = 0; i < 16; ++i) m[i] = U8TO32(block + i * 4);
for (i = 0; i < 8; ++i) v[i] = S->h[i];
v[ 8] = S->s[0] ^ 0x243F6A88;
v[ 9] = S->s[1] ^ 0x85A308D3;
v[10] = S->s[2] ^ 0x13198A2E;
v[11] = S->s[3] ^ 0x03707344;
v[12] = 0xA4093822;
v[13] = 0x299F31D0;
v[14] = 0x082EFA98;
v[15] = 0xEC4E6C89;

if (S->nullt == 0) {
v[12] ^= S->t[0];
v[13] ^= S->t[0];
v[14] ^= S->t[1];
v[15] ^= S->t[1];
}

for (i = 0; i < 14; ++i) {
G(0, 4, 8, 12, 0);
G(1, 5, 9, 13, 2);
G(2, 6, 10, 14, 4);
G(3, 7, 11, 15, 6);
G(3, 4, 9, 14, 14);
G(2, 7, 8, 13, 12);
G(0, 5, 10, 15, 8);
G(1, 6, 11, 12, 10);
}

for (i = 0; i < 16; ++i) S->h[i % 8] ^= v[i];
for (i = 0; i < 8; ++i) S->h[i] ^= S->s[i % 4];
}

void blake256_init(state *S) {
S->h[0] = 0x6A09E667;
S->h[1] = 0xBB67AE85;
S->h[2] = 0x3C6EF372;
S->h[3] = 0xA54FF53A;
S->h[4] = 0x510E527F;
S->h[5] = 0x9B05688C;
S->h[6] = 0x1F83D9AB;
S->h[7] = 0x5BE0CD19;
S->t[0] = S->t[1] = S->buflen = S->nullt = 0;
S->s[0] = S->s[1] = S->s[2] = S->s[3] = 0;
}

void blake224_init(state *S) {
S->h[0] = 0xC1059ED8;
S->h[1] = 0x367CD507;
S->h[2] = 0x3070DD17;
S->h[3] = 0xF70E5939;
S->h[4] = 0xFFC00B31;
S->h[5] = 0x68581511;
S->h[6] = 0x64F98FA7;
S->h[7] = 0xBEFA4FA4;
S->t[0] = S->t[1] = S->buflen = S->nullt = 0;
S->s[0] = S->s[1] = S->s[2] = S->s[3] = 0;
}

// datalen = number of bits
void blake256_update(state *S, const uint8_t *data, uint64_t datalen) {
int left = S->buflen >> 3;
int fill = 64 - left;

if (left && (((datalen >> 3) & 0x3F) >= (unsigned) fill)) {
memcpy((void *) (S->buf + left), (void *) data, fill);
S->t[0] += 512;
if (S->t[0] == 0) S->t[1]++;
blake256_compress(S, S->buf);
data += fill;
datalen -= (fill << 3);
left = 0;
}

while (datalen >= 512) {
S->t[0] += 512;
if (S->t[0] == 0) S->t[1]++;
blake256_compress(S, data);
data += 64;
datalen -= 512;
}

if (datalen > 0) {
memcpy((void *) (S->buf + left), (void *) data, datalen >> 3);
S->buflen = (left << 3) + datalen;
} else {
S->buflen = 0;
}
}

// datalen = number of bits
void blake224_update(state *S, const uint8_t *data, uint64_t datalen) {
blake256_update(S, data, datalen);
}

void blake256_final_h(state *S, uint8_t *digest, uint8_t pa, uint8_t pb) {
uint8_t msglen[8];
uint32_t lo = S->t[0] + S->buflen, hi = S->t[1];
if (lo < (unsigned) S->buflen) hi++;
U32TO8(msglen + 0, hi);
U32TO8(msglen + 4, lo);

if (S->buflen == 440) { /* one padding byte */
S->t[0] -= 8;
blake256_update(S, &pa, 8);
} else {
if (S->buflen < 440) { /* enough space to fill the block */
if (S->buflen == 0) S->nullt = 1;
S->t[0] -= 440 - S->buflen;
blake256_update(S, padding, 440 - S->buflen);
} else { /* need 2 compressions */
S->t[0] -= 512 - S->buflen;
blake256_update(S, padding, 512 - S->buflen);
S->t[0] -= 440;
blake256_update(S, padding + 1, 440);
S->nullt = 1;
}
blake256_update(S, &pb, 8);
S->t[0] -= 8;
}
S->t[0] -= 64;
blake256_update(S, msglen, 64);

U32TO8(digest + 0, S->h[0]);
U32TO8(digest + 4, S->h[1]);
U32TO8(digest + 8, S->h[2]);
U32TO8(digest + 12, S->h[3]);
U32TO8(digest + 16, S->h[4]);
U32TO8(digest + 20, S->h[5]);
U32TO8(digest + 24, S->h[6]);
U32TO8(digest + 28, S->h[7]);
}

void blake256_final(state *S, uint8_t *digest) {
blake256_final_h(S, digest, 0x81, 0x01);
}

void blake224_final(state *S, uint8_t *digest) {
blake256_final_h(S, digest, 0x80, 0x00);
}

// inlen = number of bytes
void blake256_hash(uint8_t *out, const uint8_t *in, uint64_t inlen) {
state S;
blake256_init(&S);
blake256_update(&S, in, inlen * 8);
blake256_final(&S, out);
}

// inlen = number of bytes
void blake224_hash(uint8_t *out, const uint8_t *in, uint64_t inlen) {
state S;
blake224_init(&S);
blake224_update(&S, in, inlen * 8);
blake224_final(&S, out);
}

// keylen = number of bytes
void hmac_blake256_init(hmac_state *S, const uint8_t *_key, uint64_t keylen) {
const uint8_t *key = _key;
uint8_t keyhash[32];
uint8_t pad[64];
uint64_t i;

if (keylen > 64) {
blake256_hash(keyhash, key, keylen);
key = keyhash;
keylen = 32;
}

blake256_init(&S->inner);
memset(pad, 0x36, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake256_update(&S->inner, pad, 512);

blake256_init(&S->outer);
memset(pad, 0x5c, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake256_update(&S->outer, pad, 512);

memset(keyhash, 0, 32);
}

// keylen = number of bytes
void hmac_blake224_init(hmac_state *S, const uint8_t *_key, uint64_t keylen) {
const uint8_t *key = _key;
uint8_t keyhash[32];
uint8_t pad[64];
uint64_t i;

if (keylen > 64) {
blake256_hash(keyhash, key, keylen);
key = keyhash;
keylen = 28;
}

blake224_init(&S->inner);
memset(pad, 0x36, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake224_update(&S->inner, pad, 512);

blake224_init(&S->outer);
memset(pad, 0x5c, 64);
for (i = 0; i < keylen; ++i) {
pad[i] ^= key[i];
}
blake224_update(&S->outer, pad, 512);

memset(keyhash, 0, 32);
}

// datalen = number of bits
void hmac_blake256_update(hmac_state *S, const uint8_t *data, uint64_t datalen) {
// update the inner state
blake256_update(&S->inner, data, datalen);
}

// datalen = number of bits
void hmac_blake224_update(hmac_state *S, const uint8_t *data, uint64_t datalen) {
// update the inner state
blake224_update(&S->inner, data, datalen);
}

void hmac_blake256_final(hmac_state *S, uint8_t *digest) {
uint8_t ihash[32];
blake256_final(&S->inner, ihash);
blake256_update(&S->outer, ihash, 256);
blake256_final(&S->outer, digest);
memset(ihash, 0, 32);
}

void hmac_blake224_final(hmac_state *S, uint8_t *digest) {
uint8_t ihash[32];
blake224_final(&S->inner, ihash);
blake224_update(&S->outer, ihash, 224);
blake224_final(&S->outer, digest);
memset(ihash, 0, 32);
}

// keylen = number of bytes; inlen = number of bytes
void hmac_blake256_hash(uint8_t *out, const uint8_t *key, uint64_t keylen, const uint8_t *in, uint64_t inlen) {
hmac_state S;
hmac_blake256_init(&S, key, keylen);
hmac_blake256_update(&S, in, inlen * 8);
hmac_blake256_final(&S, out);
}

// keylen = number of bytes; inlen = number of bytes
void hmac_blake224_hash(uint8_t *out, const uint8_t *key, uint64_t keylen, const uint8_t *in, uint64_t inlen) {
hmac_state S;
hmac_blake224_init(&S, key, keylen);
hmac_blake224_update(&S, in, inlen * 8);
hmac_blake224_final(&S, out);
}

+ 43
- 0
crypto/c_blake256.h View File

@@ -0,0 +1,43 @@
#ifndef _BLAKE256_H_
#define _BLAKE256_H_

#include <stdint.h>

typedef struct {
uint32_t h[8], s[4], t[2];
int buflen, nullt;
uint8_t buf[64];
} state;

typedef struct {
state inner;
state outer;
} hmac_state;

void blake256_init(state *);
void blake224_init(state *);

void blake256_update(state *, const uint8_t *, uint64_t);
void blake224_update(state *, const uint8_t *, uint64_t);

void blake256_final(state *, uint8_t *);
void blake224_final(state *, uint8_t *);

void blake256_hash(uint8_t *, const uint8_t *, uint64_t);
void blake224_hash(uint8_t *, const uint8_t *, uint64_t);

/* HMAC functions: */

void hmac_blake256_init(hmac_state *, const uint8_t *, uint64_t);
void hmac_blake224_init(hmac_state *, const uint8_t *, uint64_t);

void hmac_blake256_update(hmac_state *, const uint8_t *, uint64_t);
void hmac_blake224_update(hmac_state *, const uint8_t *, uint64_t);

void hmac_blake256_final(hmac_state *, uint8_t *);
void hmac_blake224_final(hmac_state *, uint8_t *);

void hmac_blake256_hash(uint8_t *, const uint8_t *, uint64_t, const uint8_t *, uint64_t);
void hmac_blake224_hash(uint8_t *, const uint8_t *, uint64_t, const uint8_t *, uint64_t);

#endif /* _BLAKE256_H_ */

+ 360
- 0
crypto/c_groestl.c View File

@@ -0,0 +1,360 @@
/* hash.c April 2012
* Groestl ANSI C code optimised for 32-bit machines
* Author: Thomas Krinninger
*
* This work is based on the implementation of
* Soeren S. Thomsen and Krystian Matusiewicz
*
*
*/

#include "c_groestl.h"
#include "groestl_tables.h"

#define P_TYPE 0
#define Q_TYPE 1

const uint8_t shift_Values[2][8] = {{0,1,2,3,4,5,6,7},{1,3,5,7,0,2,4,6}};

const uint8_t indices_cyclic[15] = {0,1,2,3,4,5,6,7,0,1,2,3,4,5,6};


#define ROTATE_COLUMN_DOWN(v1, v2, amount_bytes, temp_var) {temp_var = (v1<<(8*amount_bytes))|(v2>>(8*(4-amount_bytes))); \
v2 = (v2<<(8*amount_bytes))|(v1>>(8*(4-amount_bytes))); \
v1 = temp_var;}

#define COLUMN(x,y,i,c0,c1,c2,c3,c4,c5,c6,c7,tv1,tv2,tu,tl,t) \
tu = T[2*(uint32_t)x[4*c0+0]]; \
tl = T[2*(uint32_t)x[4*c0+0]+1]; \
tv1 = T[2*(uint32_t)x[4*c1+1]]; \
tv2 = T[2*(uint32_t)x[4*c1+1]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,1,t) \
tu ^= tv1; \
tl ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c2+2]]; \
tv2 = T[2*(uint32_t)x[4*c2+2]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,2,t) \
tu ^= tv1; \
tl ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c3+3]]; \
tv2 = T[2*(uint32_t)x[4*c3+3]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,3,t) \
tu ^= tv1; \
tl ^= tv2; \
tl ^= T[2*(uint32_t)x[4*c4+0]]; \
tu ^= T[2*(uint32_t)x[4*c4+0]+1]; \
tv1 = T[2*(uint32_t)x[4*c5+1]]; \
tv2 = T[2*(uint32_t)x[4*c5+1]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,1,t) \
tl ^= tv1; \
tu ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c6+2]]; \
tv2 = T[2*(uint32_t)x[4*c6+2]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,2,t) \
tl ^= tv1; \
tu ^= tv2; \
tv1 = T[2*(uint32_t)x[4*c7+3]]; \
tv2 = T[2*(uint32_t)x[4*c7+3]+1]; \
ROTATE_COLUMN_DOWN(tv1,tv2,3,t) \
tl ^= tv1; \
tu ^= tv2; \
y[i] = tu; \
y[i+1] = tl;


/* compute one round of P (short variants) */
static void RND512P(uint8_t *x, uint32_t *y, uint32_t r) {
uint32_t temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp;
uint32_t* x32 = (uint32_t*)x;
x32[ 0] ^= 0x00000000^r;
x32[ 2] ^= 0x00000010^r;
x32[ 4] ^= 0x00000020^r;
x32[ 6] ^= 0x00000030^r;
x32[ 8] ^= 0x00000040^r;
x32[10] ^= 0x00000050^r;
x32[12] ^= 0x00000060^r;
x32[14] ^= 0x00000070^r;
COLUMN(x,y, 0, 0, 2, 4, 6, 9, 11, 13, 15, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 2, 2, 4, 6, 8, 11, 13, 15, 1, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 4, 4, 6, 8, 10, 13, 15, 1, 3, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 6, 6, 8, 10, 12, 15, 1, 3, 5, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 8, 8, 10, 12, 14, 1, 3, 5, 7, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,10, 10, 12, 14, 0, 3, 5, 7, 9, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,12, 12, 14, 0, 2, 5, 7, 9, 11, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,14, 14, 0, 2, 4, 7, 9, 11, 13, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
}

/* compute one round of Q (short variants) */
static void RND512Q(uint8_t *x, uint32_t *y, uint32_t r) {
uint32_t temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp;
uint32_t* x32 = (uint32_t*)x;
x32[ 0] = ~x32[ 0];
x32[ 1] ^= 0xffffffff^r;
x32[ 2] = ~x32[ 2];
x32[ 3] ^= 0xefffffff^r;
x32[ 4] = ~x32[ 4];
x32[ 5] ^= 0xdfffffff^r;
x32[ 6] = ~x32[ 6];
x32[ 7] ^= 0xcfffffff^r;
x32[ 8] = ~x32[ 8];
x32[ 9] ^= 0xbfffffff^r;
x32[10] = ~x32[10];
x32[11] ^= 0xafffffff^r;
x32[12] = ~x32[12];
x32[13] ^= 0x9fffffff^r;
x32[14] = ~x32[14];
x32[15] ^= 0x8fffffff^r;
COLUMN(x,y, 0, 2, 6, 10, 14, 1, 5, 9, 13, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 2, 4, 8, 12, 0, 3, 7, 11, 15, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 4, 6, 10, 14, 2, 5, 9, 13, 1, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 6, 8, 12, 0, 4, 7, 11, 15, 3, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y, 8, 10, 14, 2, 6, 9, 13, 1, 5, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,10, 12, 0, 4, 8, 11, 15, 3, 7, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,12, 14, 2, 6, 10, 13, 1, 5, 9, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
COLUMN(x,y,14, 0, 4, 8, 12, 15, 3, 7, 11, temp_v1, temp_v2, temp_upper_value, temp_lower_value, temp);
}

/* compute compression function (short variants) */
static void F512(uint32_t *h, const uint32_t *m) {
int i;
uint32_t Ptmp[2*COLS512];
uint32_t Qtmp[2*COLS512];
uint32_t y[2*COLS512];
uint32_t z[2*COLS512];

for (i = 0; i < 2*COLS512; i++) {
z[i] = m[i];
Ptmp[i] = h[i]^m[i];
}

/* compute Q(m) */
RND512Q((uint8_t*)z, y, 0x00000000);
RND512Q((uint8_t*)y, z, 0x01000000);
RND512Q((uint8_t*)z, y, 0x02000000);
RND512Q((uint8_t*)y, z, 0x03000000);
RND512Q((uint8_t*)z, y, 0x04000000);
RND512Q((uint8_t*)y, z, 0x05000000);
RND512Q((uint8_t*)z, y, 0x06000000);
RND512Q((uint8_t*)y, z, 0x07000000);
RND512Q((uint8_t*)z, y, 0x08000000);
RND512Q((uint8_t*)y, Qtmp, 0x09000000);

/* compute P(h+m) */
RND512P((uint8_t*)Ptmp, y, 0x00000000);
RND512P((uint8_t*)y, z, 0x00000001);
RND512P((uint8_t*)z, y, 0x00000002);
RND512P((uint8_t*)y, z, 0x00000003);
RND512P((uint8_t*)z, y, 0x00000004);
RND512P((uint8_t*)y, z, 0x00000005);
RND512P((uint8_t*)z, y, 0x00000006);
RND512P((uint8_t*)y, z, 0x00000007);
RND512P((uint8_t*)z, y, 0x00000008);
RND512P((uint8_t*)y, Ptmp, 0x00000009);

/* compute P(h+m) + Q(m) + h */
for (i = 0; i < 2*COLS512; i++) {
h[i] ^= Ptmp[i]^Qtmp[i];
}
}


/* digest up to msglen bytes of input (full blocks only) */
static void Transform(groestlHashState *ctx,
const uint8_t *input,
int msglen) {

/* digest message, one block at a time */
for (; msglen >= SIZE512;
msglen -= SIZE512, input += SIZE512) {
F512(ctx->chaining,(uint32_t*)input);

/* increment block counter */
ctx->block_counter1++;
if (ctx->block_counter1 == 0) ctx->block_counter2++;
}
}

/* given state h, do h <- P(h)+h */
static void OutputTransformation(groestlHashState *ctx) {
int j;
uint32_t temp[2*COLS512];
uint32_t y[2*COLS512];
uint32_t z[2*COLS512];



for (j = 0; j < 2*COLS512; j++) {
temp[j] = ctx->chaining[j];
}
RND512P((uint8_t*)temp, y, 0x00000000);
RND512P((uint8_t*)y, z, 0x00000001);
RND512P((uint8_t*)z, y, 0x00000002);
RND512P((uint8_t*)y, z, 0x00000003);
RND512P((uint8_t*)z, y, 0x00000004);
RND512P((uint8_t*)y, z, 0x00000005);
RND512P((uint8_t*)z, y, 0x00000006);
RND512P((uint8_t*)y, z, 0x00000007);
RND512P((uint8_t*)z, y, 0x00000008);
RND512P((uint8_t*)y, temp, 0x00000009);
for (j = 0; j < 2*COLS512; j++) {
ctx->chaining[j] ^= temp[j];
}
}

/* initialise context */
static void Init(groestlHashState* ctx) {
int i = 0;
/* allocate memory for state and data buffer */

for(;i<(SIZE512/sizeof(uint32_t));i++)
{
ctx->chaining[i] = 0;
}

/* set initial value */
ctx->chaining[2*COLS512-1] = u32BIG((uint32_t)HASH_BIT_LEN);

/* set other variables */
ctx->buf_ptr = 0;
ctx->block_counter1 = 0;
ctx->block_counter2 = 0;
ctx->bits_in_last_byte = 0;
}

/* update state with databitlen bits of input */
static void Update(groestlHashState* ctx,
const BitSequence* input,
DataLength databitlen) {
int index = 0;
int msglen = (int)(databitlen/8);
int rem = (int)(databitlen%8);

/* if the buffer contains data that has not yet been digested, first
add data to buffer until full */
if (ctx->buf_ptr) {
while (ctx->buf_ptr < SIZE512 && index < msglen) {
ctx->buffer[(int)ctx->buf_ptr++] = input[index++];
}
if (ctx->buf_ptr < SIZE512) {
/* buffer still not full, return */
if (rem) {
ctx->bits_in_last_byte = rem;
ctx->buffer[(int)ctx->buf_ptr++] = input[index];
}
return;
}

/* digest buffer */
ctx->buf_ptr = 0;
Transform(ctx, ctx->buffer, SIZE512);
}

/* digest bulk of message */
Transform(ctx, input+index, msglen-index);
index += ((msglen-index)/SIZE512)*SIZE512;

/* store remaining data in buffer */
while (index < msglen) {
ctx->buffer[(int)ctx->buf_ptr++] = input[index++];
}

/* if non-integral number of bytes have been supplied, store
remaining bits in last byte, together with information about
number of bits */
if (rem) {
ctx->bits_in_last_byte = rem;
ctx->buffer[(int)ctx->buf_ptr++] = input[index];
}
}

#define BILB ctx->bits_in_last_byte

/* finalise: process remaining data (including padding), perform
output transformation, and write hash result to 'output' */
static void Final(groestlHashState* ctx,
BitSequence* output) {
int i, j = 0, hashbytelen = HASH_BIT_LEN/8;
uint8_t *s = (BitSequence*)ctx->chaining;

/* pad with '1'-bit and first few '0'-bits */
if (BILB) {
ctx->buffer[(int)ctx->buf_ptr-1] &= ((1<<BILB)-1)<<(8-BILB);
ctx->buffer[(int)ctx->buf_ptr-1] ^= 0x1<<(7-BILB);
BILB = 0;
}
else ctx->buffer[(int)ctx->buf_ptr++] = 0x80;

/* pad with '0'-bits */
if (ctx->buf_ptr > SIZE512-LENGTHFIELDLEN) {
/* padding requires two blocks */
while (ctx->buf_ptr < SIZE512) {
ctx->buffer[(int)ctx->buf_ptr++] = 0;
}
/* digest first padding block */
Transform(ctx, ctx->buffer, SIZE512);
ctx->buf_ptr = 0;
}
while (ctx->buf_ptr < SIZE512-LENGTHFIELDLEN) {
ctx->buffer[(int)ctx->buf_ptr++] = 0;
}

/* length padding */
ctx->block_counter1++;
if (ctx->block_counter1 == 0) ctx->block_counter2++;
ctx->buf_ptr = SIZE512;

while (ctx->buf_ptr > SIZE512-(int)sizeof(uint32_t)) {
ctx->buffer[(int)--ctx->buf_ptr] = (uint8_t)ctx->block_counter1;
ctx->block_counter1 >>= 8;
}
while (ctx->buf_ptr > SIZE512-LENGTHFIELDLEN) {
ctx->buffer[(int)--ctx->buf_ptr] = (uint8_t)ctx->block_counter2;
ctx->block_counter2 >>= 8;
}
/* digest final padding block */
Transform(ctx, ctx->buffer, SIZE512);
/* perform output transformation */
OutputTransformation(ctx);

/* store hash result in output */
for (i = SIZE512-hashbytelen; i < SIZE512; i++,j++) {
output[j] = s[i];
}

/* zeroise relevant variables and deallocate memory */
for (i = 0; i < COLS512; i++) {
ctx->chaining[i] = 0;
}
for (i = 0; i < SIZE512; i++) {
ctx->buffer[i] = 0;
}
}

/* hash bit sequence */
void groestl(const BitSequence* data,
DataLength databitlen,
BitSequence* hashval) {

groestlHashState context;

/* initialise */
Init(&context);


/* process message */
Update(&context, data, databitlen);

/* finalise */
Final(&context, hashval);
}
/*
static int crypto_hash(unsigned char *out,
const unsigned char *in,
unsigned long long len)
{
groestl(in, 8*len, out);
return 0;
}

*/

+ 60
- 0
crypto/c_groestl.h View File

@@ -0,0 +1,60 @@
#ifndef __hash_h
#define __hash_h
/*
#include "crypto_uint8.h"
#include "crypto_uint32.h"
#include "crypto_uint64.h"
#include "crypto_hash.h"

typedef crypto_uint8 uint8_t;
typedef crypto_uint32 uint32_t;
typedef crypto_uint64 uint64_t;
*/
#include <stdint.h>

#include "hash.h"

/* some sizes (number of bytes) */
#define ROWS 8
#define LENGTHFIELDLEN ROWS
#define COLS512 8

#define SIZE512 (ROWS*COLS512)

#define ROUNDS512 10
#define HASH_BIT_LEN 256

#define ROTL32(v, n) ((((v)<<(n))|((v)>>(32-(n))))&li_32(ffffffff))


#define li_32(h) 0x##h##u
#define EXT_BYTE(var,n) ((uint8_t)((uint32_t)(var) >> (8*n)))
#define u32BIG(a) \
((ROTL32(a,8) & li_32(00FF00FF)) | \
(ROTL32(a,24) & li_32(FF00FF00)))


/* NIST API begin */
typedef struct {
uint32_t chaining[SIZE512/sizeof(uint32_t)]; /* actual state */
uint32_t block_counter1,
block_counter2; /* message block counter(s) */
BitSequence buffer[SIZE512]; /* data buffer */
int buf_ptr; /* data buffer pointer */
int bits_in_last_byte; /* no. of message bits in last byte of
data buffer */
} groestlHashState;

/*void Init(hashState*);
void Update(hashState*, const BitSequence*, DataLength);
void Final(hashState*, BitSequence*); */
void groestl(const BitSequence*, DataLength, BitSequence*);
/* NIST API end */

/*
int crypto_hash(unsigned char *out,
const unsigned char *in,
unsigned long long len);
*/

#endif /* __hash_h */

+ 367
- 0
crypto/c_jh.c View File

@@ -0,0 +1,367 @@
/*This program gives the 64-bit optimized bitslice implementation of JH using ANSI C

--------------------------------
Performance

Microprocessor: Intel CORE 2 processor (Core 2 Duo Mobile T6600 2.2GHz)
Operating System: 64-bit Ubuntu 10.04 (Linux kernel 2.6.32-22-generic)
Speed for long message:
1) 45.8 cycles/byte compiler: Intel C++ Compiler 11.1 compilation option: icc -O2
2) 56.8 cycles/byte compiler: gcc 4.4.3 compilation option: gcc -O3

--------------------------------
Last Modified: January 16, 2011
*/

#include "c_jh.h"

#include <stdint.h>
#include <string.h>

/*typedef unsigned long long uint64;*/
typedef uint64_t uint64;

/*define data alignment for different C compilers*/
#if defined(__GNUC__)
#define DATA_ALIGN16(x) x __attribute__ ((aligned(16)))
#else
#define DATA_ALIGN16(x) __declspec(align(16)) x
#endif


typedef struct {
int hashbitlen; /*the message digest size*/
unsigned long long databitlen; /*the message size in bits*/
unsigned long long datasize_in_buffer; /*the size of the message remained in buffer; assumed to be multiple of 8bits except for the last partial block at the end of the message*/
DATA_ALIGN16(uint64 x[8][2]); /*the 1024-bit state, ( x[i][0] || x[i][1] ) is the ith row of the state in the pseudocode*/
unsigned char buffer[64]; /*the 512-bit message block to be hashed;*/
} hashState;


/*The initial hash value H(0)*/
const unsigned char JH224_H0[128]={0x2d,0xfe,0xdd,0x62,0xf9,0x9a,0x98,0xac,0xae,0x7c,0xac,0xd6,0x19,0xd6,0x34,0xe7,0xa4,0x83,0x10,0x5,0xbc,0x30,0x12,0x16,0xb8,0x60,0x38,0xc6,0xc9,0x66,0x14,0x94,0x66,0xd9,0x89,0x9f,0x25,0x80,0x70,0x6f,0xce,0x9e,0xa3,0x1b,0x1d,0x9b,0x1a,0xdc,0x11,0xe8,0x32,0x5f,0x7b,0x36,0x6e,0x10,0xf9,0x94,0x85,0x7f,0x2,0xfa,0x6,0xc1,0x1b,0x4f,0x1b,0x5c,0xd8,0xc8,0x40,0xb3,0x97,0xf6,0xa1,0x7f,0x6e,0x73,0x80,0x99,0xdc,0xdf,0x93,0xa5,0xad,0xea,0xa3,0xd3,0xa4,0x31,0xe8,0xde,0xc9,0x53,0x9a,0x68,0x22,0xb4,0xa9,0x8a,0xec,0x86,0xa1,0xe4,0xd5,0x74,0xac,0x95,0x9c,0xe5,0x6c,0xf0,0x15,0x96,0xd,0xea,0xb5,0xab,0x2b,0xbf,0x96,0x11,0xdc,0xf0,0xdd,0x64,0xea,0x6e};
const unsigned char JH256_H0[128]={0xeb,0x98,0xa3,0x41,0x2c,0x20,0xd3,0xeb,0x92,0xcd,0xbe,0x7b,0x9c,0xb2,0x45,0xc1,0x1c,0x93,0x51,0x91,0x60,0xd4,0xc7,0xfa,0x26,0x0,0x82,0xd6,0x7e,0x50,0x8a,0x3,0xa4,0x23,0x9e,0x26,0x77,0x26,0xb9,0x45,0xe0,0xfb,0x1a,0x48,0xd4,0x1a,0x94,0x77,0xcd,0xb5,0xab,0x26,0x2,0x6b,0x17,0x7a,0x56,0xf0,0x24,0x42,0xf,0xff,0x2f,0xa8,0x71,0xa3,0x96,0x89,0x7f,0x2e,0x4d,0x75,0x1d,0x14,0x49,0x8,0xf7,0x7d,0xe2,0x62,0x27,0x76,0x95,0xf7,0x76,0x24,0x8f,0x94,0x87,0xd5,0xb6,0x57,0x47,0x80,0x29,0x6c,0x5c,0x5e,0x27,0x2d,0xac,0x8e,0xd,0x6c,0x51,0x84,0x50,0xc6,0x57,0x5,0x7a,0xf,0x7b,0xe4,0xd3,0x67,0x70,0x24,0x12,0xea,0x89,0xe3,0xab,0x13,0xd3,0x1c,0xd7,0x69};
const unsigned char JH384_H0[128]={0x48,0x1e,0x3b,0xc6,0xd8,0x13,0x39,0x8a,0x6d,0x3b,0x5e,0x89,0x4a,0xde,0x87,0x9b,0x63,0xfa,0xea,0x68,0xd4,0x80,0xad,0x2e,0x33,0x2c,0xcb,0x21,0x48,0xf,0x82,0x67,0x98,0xae,0xc8,0x4d,0x90,0x82,0xb9,0x28,0xd4,0x55,0xea,0x30,0x41,0x11,0x42,0x49,0x36,0xf5,0x55,0xb2,0x92,0x48,0x47,0xec,0xc7,0x25,0xa,0x93,0xba,0xf4,0x3c,0xe1,0x56,0x9b,0x7f,0x8a,0x27,0xdb,0x45,0x4c,0x9e,0xfc,0xbd,0x49,0x63,0x97,0xaf,0xe,0x58,0x9f,0xc2,0x7d,0x26,0xaa,0x80,0xcd,0x80,0xc0,0x8b,0x8c,0x9d,0xeb,0x2e,0xda,0x8a,0x79,0x81,0xe8,0xf8,0xd5,0x37,0x3a,0xf4,0x39,0x67,0xad,0xdd,0xd1,0x7a,0x71,0xa9,0xb4,0xd3,0xbd,0xa4,0x75,0xd3,0x94,0x97,0x6c,0x3f,0xba,0x98,0x42,0x73,0x7f};
const unsigned char JH512_H0[128]={0x6f,0xd1,0x4b,0x96,0x3e,0x0,0xaa,0x17,0x63,0x6a,0x2e,0x5,0x7a,0x15,0xd5,0x43,0x8a,0x22,0x5e,0x8d,0xc,0x97,0xef,0xb,0xe9,0x34,0x12,0x59,0xf2,0xb3,0xc3,0x61,0x89,0x1d,0xa0,0xc1,0x53,0x6f,0x80,0x1e,0x2a,0xa9,0x5,0x6b,0xea,0x2b,0x6d,0x80,0x58,0x8e,0xcc,0xdb,0x20,0x75,0xba,0xa6,0xa9,0xf,0x3a,0x76,0xba,0xf8,0x3b,0xf7,0x1,0x69,0xe6,0x5,0x41,0xe3,0x4a,0x69,0x46,0xb5,0x8a,0x8e,0x2e,0x6f,0xe6,0x5a,0x10,0x47,0xa7,0xd0,0xc1,0x84,0x3c,0x24,0x3b,0x6e,0x71,0xb1,0x2d,0x5a,0xc1,0x99,0xcf,0x57,0xf6,0xec,0x9d,0xb1,0xf8,0x56,0xa7,0x6,0x88,0x7c,0x57,0x16,0xb1,0x56,0xe3,0xc2,0xfc,0xdf,0xe6,0x85,0x17,0xfb,0x54,0x5a,0x46,0x78,0xcc,0x8c,0xdd,0x4b};

/*42 round constants, each round constant is 32-byte (256-bit)*/
const unsigned char E8_bitslice_roundconstant[42][32]={
{0x72,0xd5,0xde,0xa2,0xdf,0x15,0xf8,0x67,0x7b,0x84,0x15,0xa,0xb7,0x23,0x15,0x57,0x81,0xab,0xd6,0x90,0x4d,0x5a,0x87,0xf6,0x4e,0x9f,0x4f,0xc5,0xc3,0xd1,0x2b,0x40},
{0xea,0x98,0x3a,0xe0,0x5c,0x45,0xfa,0x9c,0x3,0xc5,0xd2,0x99,0x66,0xb2,0x99,0x9a,0x66,0x2,0x96,0xb4,0xf2,0xbb,0x53,0x8a,0xb5,0x56,0x14,0x1a,0x88,0xdb,0xa2,0x31},
{0x3,0xa3,0x5a,0x5c,0x9a,0x19,0xe,0xdb,0x40,0x3f,0xb2,0xa,0x87,0xc1,0x44,0x10,0x1c,0x5,0x19,0x80,0x84,0x9e,0x95,0x1d,0x6f,0x33,0xeb,0xad,0x5e,0xe7,0xcd,0xdc},
{0x10,0xba,0x13,0x92,0x2,0xbf,0x6b,0x41,0xdc,0x78,0x65,0x15,0xf7,0xbb,0x27,0xd0,0xa,0x2c,0x81,0x39,0x37,0xaa,0x78,0x50,0x3f,0x1a,0xbf,0xd2,0x41,0x0,0x91,0xd3},
{0x42,0x2d,0x5a,0xd,0xf6,0xcc,0x7e,0x90,0xdd,0x62,0x9f,0x9c,0x92,0xc0,0x97,0xce,0x18,0x5c,0xa7,0xb,0xc7,0x2b,0x44,0xac,0xd1,0xdf,0x65,0xd6,0x63,0xc6,0xfc,0x23},
{0x97,0x6e,0x6c,0x3,0x9e,0xe0,0xb8,0x1a,0x21,0x5,0x45,0x7e,0x44,0x6c,0xec,0xa8,0xee,0xf1,0x3,0xbb,0x5d,0x8e,0x61,0xfa,0xfd,0x96,0x97,0xb2,0x94,0x83,0x81,0x97},
{0x4a,0x8e,0x85,0x37,0xdb,0x3,0x30,0x2f,0x2a,0x67,0x8d,0x2d,0xfb,0x9f,0x6a,0x95,0x8a,0xfe,0x73,0x81,0xf8,0xb8,0x69,0x6c,0x8a,0xc7,0x72,0x46,0xc0,0x7f,0x42,0x14},
{0xc5,0xf4,0x15,0x8f,0xbd,0xc7,0x5e,0xc4,0x75,0x44,0x6f,0xa7,0x8f,0x11,0xbb,0x80,0x52,0xde,0x75,0xb7,0xae,0xe4,0x88,0xbc,0x82,0xb8,0x0,0x1e,0x98,0xa6,0xa3,0xf4},
{0x8e,0xf4,0x8f,0x33,0xa9,0xa3,0x63,0x15,0xaa,0x5f,0x56,0x24,0xd5,0xb7,0xf9,0x89,0xb6,0xf1,0xed,0x20,0x7c,0x5a,0xe0,0xfd,0x36,0xca,0xe9,0x5a,0x6,0x42,0x2c,0x36},
{0xce,0x29,0x35,0x43,0x4e,0xfe,0x98,0x3d,0x53,0x3a,0xf9,0x74,0x73,0x9a,0x4b,0xa7,0xd0,0xf5,0x1f,0x59,0x6f,0x4e,0x81,0x86,0xe,0x9d,0xad,0x81,0xaf,0xd8,0x5a,0x9f},
{0xa7,0x5,0x6,0x67,0xee,0x34,0x62,0x6a,0x8b,0xb,0x28,0xbe,0x6e,0xb9,0x17,0x27,0x47,0x74,0x7,0x26,0xc6,0x80,0x10,0x3f,0xe0,0xa0,0x7e,0x6f,0xc6,0x7e,0x48,0x7b},
{0xd,0x55,0xa,0xa5,0x4a,0xf8,0xa4,0xc0,0x91,0xe3,0xe7,0x9f,0x97,0x8e,0xf1,0x9e,0x86,0x76,0x72,0x81,0x50,0x60,0x8d,0xd4,0x7e,0x9e,0x5a,0x41,0xf3,0xe5,0xb0,0x62},
{0xfc,0x9f,0x1f,0xec,0x40,0x54,0x20,0x7a,0xe3,0xe4,0x1a,0x0,0xce,0xf4,0xc9,0x84,0x4f,0xd7,0x94,0xf5,0x9d,0xfa,0x95,0xd8,0x55,0x2e,0x7e,0x11,0x24,0xc3,0x54,0xa5},
{0x5b,0xdf,0x72,0x28,0xbd,0xfe,0x6e,0x28,0x78,0xf5,0x7f,0xe2,0xf,0xa5,0xc4,0xb2,0x5,0x89,0x7c,0xef,0xee,0x49,0xd3,0x2e,0x44,0x7e,0x93,0x85,0xeb,0x28,0x59,0x7f},
{0x70,0x5f,0x69,0x37,0xb3,0x24,0x31,0x4a,0x5e,0x86,0x28,0xf1,0x1d,0xd6,0xe4,0x65,0xc7,0x1b,0x77,0x4,0x51,0xb9,0x20,0xe7,0x74,0xfe,0x43,0xe8,0x23,0xd4,0x87,0x8a},
{0x7d,0x29,0xe8,0xa3,0x92,0x76,0x94,0xf2,0xdd,0xcb,0x7a,0x9,0x9b,0x30,0xd9,0xc1,0x1d,0x1b,0x30,0xfb,0x5b,0xdc,0x1b,0xe0,0xda,0x24,0x49,0x4f,0xf2,0x9c,0x82,0xbf},
{0xa4,0xe7,0xba,0x31,0xb4,0x70,0xbf,0xff,0xd,0x32,0x44,0x5,0xde,0xf8,0xbc,0x48,0x3b,0xae,0xfc,0x32,0x53,0xbb,0xd3,0x39,0x45,0x9f,0xc3,0xc1,0xe0,0x29,0x8b,0xa0},
{0xe5,0xc9,0x5,0xfd,0xf7,0xae,0x9,0xf,0x94,0x70,0x34,0x12,0x42,0x90,0xf1,0x34,0xa2,0x71,0xb7,0x1,0xe3,0x44,0xed,0x95,0xe9,0x3b,0x8e,0x36,0x4f,0x2f,0x98,0x4a},
{0x88,0x40,0x1d,0x63,0xa0,0x6c,0xf6,0x15,0x47,0xc1,0x44,0x4b,0x87,0x52,0xaf,0xff,0x7e,0xbb,0x4a,0xf1,0xe2,0xa,0xc6,0x30,0x46,0x70,0xb6,0xc5,0xcc,0x6e,0x8c,0xe6},
{0xa4,0xd5,0xa4,0x56,0xbd,0x4f,0xca,0x0,0xda,0x9d,0x84,0x4b,0xc8,0x3e,0x18,0xae,0x73,0x57,0xce,0x45,0x30,0x64,0xd1,0xad,0xe8,0xa6,0xce,0x68,0x14,0x5c,0x25,0x67},
{0xa3,0xda,0x8c,0xf2,0xcb,0xe,0xe1,0x16,0x33,0xe9,0x6,0x58,0x9a,0x94,0x99,0x9a,0x1f,0x60,0xb2,0x20,0xc2,0x6f,0x84,0x7b,0xd1,0xce,0xac,0x7f,0xa0,0xd1,0x85,0x18},
{0x32,0x59,0x5b,0xa1,0x8d,0xdd,0x19,0xd3,0x50,0x9a,0x1c,0xc0,0xaa,0xa5,0xb4,0x46,0x9f,0x3d,0x63,0x67,0xe4,0x4,0x6b,0xba,0xf6,0xca,0x19,0xab,0xb,0x56,0xee,0x7e},
{0x1f,0xb1,0x79,0xea,0xa9,0x28,0x21,0x74,0xe9,0xbd,0xf7,0x35,0x3b,0x36,0x51,0xee,0x1d,0x57,0xac,0x5a,0x75,0x50,0xd3,0x76,0x3a,0x46,0xc2,0xfe,0xa3,0x7d,0x70,0x1},
{0xf7,0x35,0xc1,0xaf,0x98,0xa4,0xd8,0x42,0x78,0xed,0xec,0x20,0x9e,0x6b,0x67,0x79,0x41,0x83,0x63,0x15,0xea,0x3a,0xdb,0xa8,0xfa,0xc3,0x3b,0x4d,0x32,0x83,0x2c,0x83},
{0xa7,0x40,0x3b,0x1f,0x1c,0x27,0x47,0xf3,0x59,0x40,0xf0,0x34,0xb7,0x2d,0x76,0x9a,0xe7,0x3e,0x4e,0x6c,0xd2,0x21,0x4f,0xfd,0xb8,0xfd,0x8d,0x39,0xdc,0x57,0x59,0xef},
{0x8d,0x9b,0xc,0x49,0x2b,0x49,0xeb,0xda,0x5b,0xa2,0xd7,0x49,0x68,0xf3,0x70,0xd,0x7d,0x3b,0xae,0xd0,0x7a,0x8d,0x55,0x84,0xf5,0xa5,0xe9,0xf0,0xe4,0xf8,0x8e,0x65},
{0xa0,0xb8,0xa2,0xf4,0x36,0x10,0x3b,0x53,0xc,0xa8,0x7,0x9e,0x75,0x3e,0xec,0x5a,0x91,0x68,0x94,0x92,0x56,0xe8,0x88,0x4f,0x5b,0xb0,0x5c,0x55,0xf8,0xba,0xbc,0x4c},
{0xe3,0xbb,0x3b,0x99,0xf3,0x87,0x94,0x7b,0x75,0xda,0xf4,0xd6,0x72,0x6b,0x1c,0x5d,0x64,0xae,0xac,0x28,0xdc,0x34,0xb3,0x6d,0x6c,0x34,0xa5,0x50,0xb8,0x28,0xdb,0x71},
{0xf8,0x61,0xe2,0xf2,0x10,0x8d,0x51,0x2a,0xe3,0xdb,0x64,0x33,0x59,0xdd,0x75,0xfc,0x1c,0xac,0xbc,0xf1,0x43,0xce,0x3f,0xa2,0x67,0xbb,0xd1,0x3c,0x2,0xe8,0x43,0xb0},
{0x33,0xa,0x5b,0xca,0x88,0x29,0xa1,0x75,0x7f,0x34,0x19,0x4d,0xb4,0x16,0x53,0x5c,0x92,0x3b,0x94,0xc3,0xe,0x79,0x4d,0x1e,0x79,0x74,0x75,0xd7,0xb6,0xee,0xaf,0x3f},
{0xea,0xa8,0xd4,0xf7,0xbe,0x1a,0x39,0x21,0x5c,0xf4,0x7e,0x9,0x4c,0x23,0x27,0x51,0x26,0xa3,0x24,0x53,0xba,0x32,0x3c,0xd2,0x44,0xa3,0x17,0x4a,0x6d,0xa6,0xd5,0xad},
{0xb5,0x1d,0x3e,0xa6,0xaf,0xf2,0xc9,0x8,0x83,0x59,0x3d,0x98,0x91,0x6b,0x3c,0x56,0x4c,0xf8,0x7c,0xa1,0x72,0x86,0x60,0x4d,0x46,0xe2,0x3e,0xcc,0x8,0x6e,0xc7,0xf6},
{0x2f,0x98,0x33,0xb3,0xb1,0xbc,0x76,0x5e,0x2b,0xd6,0x66,0xa5,0xef,0xc4,0xe6,0x2a,0x6,0xf4,0xb6,0xe8,0xbe,0xc1,0xd4,0x36,0x74,0xee,0x82,0x15,0xbc,0xef,0x21,0x63},
{0xfd,0xc1,0x4e,0xd,0xf4,0x53,0xc9,0x69,0xa7,0x7d,0x5a,0xc4,0x6,0x58,0x58,0x26,0x7e,0xc1,0x14,0x16,0x6,0xe0,0xfa,0x16,0x7e,0x90,0xaf,0x3d,0x28,0x63,0x9d,0x3f},
{0xd2,0xc9,0xf2,0xe3,0x0,0x9b,0xd2,0xc,0x5f,0xaa,0xce,0x30,0xb7,0xd4,0xc,0x30,0x74,0x2a,0x51,0x16,0xf2,0xe0,0x32,0x98,0xd,0xeb,0x30,0xd8,0xe3,0xce,0xf8,0x9a},
{0x4b,0xc5,0x9e,0x7b,0xb5,0xf1,0x79,0x92,0xff,0x51,0xe6,0x6e,0x4,0x86,0x68,0xd3,0x9b,0x23,0x4d,0x57,0xe6,0x96,0x67,0x31,0xcc,0xe6,0xa6,0xf3,0x17,0xa,0x75,0x5},
{0xb1,0x76,0x81,0xd9,0x13,0x32,0x6c,0xce,0x3c,0x17,0x52,0x84,0xf8,0x5,0xa2,0x62,0xf4,0x2b,0xcb,0xb3,0x78,0x47,0x15,0x47,0xff,0x46,0x54,0x82,0x23,0x93,0x6a,0x48},
{0x38,0xdf,0x58,0x7,0x4e,0x5e,0x65,0x65,0xf2,0xfc,0x7c,0x89,0xfc,0x86,0x50,0x8e,0x31,0x70,0x2e,0x44,0xd0,0xb,0xca,0x86,0xf0,0x40,0x9,0xa2,0x30,0x78,0x47,0x4e},
{0x65,0xa0,0xee,0x39,0xd1,0xf7,0x38,0x83,0xf7,0x5e,0xe9,0x37,0xe4,0x2c,0x3a,0xbd,0x21,0x97,0xb2,0x26,0x1,0x13,0xf8,0x6f,0xa3,0x44,0xed,0xd1,0xef,0x9f,0xde,0xe7},
{0x8b,0xa0,0xdf,0x15,0x76,0x25,0x92,0xd9,0x3c,0x85,0xf7,0xf6,0x12,0xdc,0x42,0xbe,0xd8,0xa7,0xec,0x7c,0xab,0x27,0xb0,0x7e,0x53,0x8d,0x7d,0xda,0xaa,0x3e,0xa8,0xde},
{0xaa,0x25,0xce,0x93,0xbd,0x2,0x69,0xd8,0x5a,0xf6,0x43,0xfd,0x1a,0x73,0x8,0xf9,0xc0,0x5f,0xef,0xda,0x17,0x4a,0x19,0xa5,0x97,0x4d,0x66,0x33,0x4c,0xfd,0x21,0x6a},
{0x35,0xb4,0x98,0x31,0xdb,0x41,0x15,0x70,0xea,0x1e,0xf,0xbb,0xed,0xcd,0x54,0x9b,0x9a,0xd0,0x63,0xa1,0x51,0x97,0x40,0x72,0xf6,0x75,0x9d,0xbf,0x91,0x47,0x6f,0xe2}};


static void E8(hashState *state); /*The bijective function E8, in bitslice form*/
static void F8(hashState *state); /*The compression function F8 */

/*The API functions*/
static HashReturn Init(hashState *state, int hashbitlen);
static HashReturn Update(hashState *state, const BitSequence *data, DataLength databitlen);
static HashReturn Final(hashState *state, BitSequence *hashval);
HashReturn jh_hash(int hashbitlen, const BitSequence *data,DataLength databitlen, BitSequence *hashval);

/*swapping bit 2i with bit 2i+1 of 64-bit x*/
#define SWAP1(x) (x) = ((((x) & 0x5555555555555555ULL) << 1) | (((x) & 0xaaaaaaaaaaaaaaaaULL) >> 1));
/*swapping bits 4i||4i+1 with bits 4i+2||4i+3 of 64-bit x*/
#define SWAP2(x) (x) = ((((x) & 0x3333333333333333ULL) << 2) | (((x) & 0xccccccccccccccccULL) >> 2));
/*swapping bits 8i||8i+1||8i+2||8i+3 with bits 8i+4||8i+5||8i+6||8i+7 of 64-bit x*/
#define SWAP4(x) (x) = ((((x) & 0x0f0f0f0f0f0f0f0fULL) << 4) | (((x) & 0xf0f0f0f0f0f0f0f0ULL) >> 4));
/*swapping bits 16i||16i+1||......||16i+7 with bits 16i+8||16i+9||......||16i+15 of 64-bit x*/
#define SWAP8(x) (x) = ((((x) & 0x00ff00ff00ff00ffULL) << 8) | (((x) & 0xff00ff00ff00ff00ULL) >> 8));
/*swapping bits 32i||32i+1||......||32i+15 with bits 32i+16||32i+17||......||32i+31 of 64-bit x*/
#define SWAP16(x) (x) = ((((x) & 0x0000ffff0000ffffULL) << 16) | (((x) & 0xffff0000ffff0000ULL) >> 16));
/*swapping bits 64i||64i+1||......||64i+31 with bits 64i+32||64i+33||......||64i+63 of 64-bit x*/
#define SWAP32(x) (x) = (((x) << 32) | ((x) >> 32));

/*The MDS transform*/
#define L(m0,m1,m2,m3,m4,m5,m6,m7) \
(m4) ^= (m1); \
(m5) ^= (m2); \
(m6) ^= (m0) ^ (m3); \
(m7) ^= (m0); \
(m0) ^= (m5); \
(m1) ^= (m6); \
(m2) ^= (m4) ^ (m7); \
(m3) ^= (m4);

/*Two Sboxes are computed in parallel, each Sbox implements S0 and S1, selected by a constant bit*/
/*The reason to compute two Sboxes in parallel is to try to fully utilize the parallel processing power*/
#define SS(m0,m1,m2,m3,m4,m5,m6,m7,cc0,cc1) \
m3 = ~(m3); \
m7 = ~(m7); \
m0 ^= ((~(m2)) & (cc0)); \
m4 ^= ((~(m6)) & (cc1)); \
temp0 = (cc0) ^ ((m0) & (m1));\
temp1 = (cc1) ^ ((m4) & (m5));\
m0 ^= ((m2) & (m3)); \
m4 ^= ((m6) & (m7)); \
m3 ^= ((~(m1)) & (m2)); \
m7 ^= ((~(m5)) & (m6)); \
m1 ^= ((m0) & (m2)); \
m5 ^= ((m4) & (m6)); \
m2 ^= ((m0) & (~(m3))); \
m6 ^= ((m4) & (~(m7))); \
m0 ^= ((m1) | (m3)); \
m4 ^= ((m5) | (m7)); \
m3 ^= ((m1) & (m2)); \
m7 ^= ((m5) & (m6)); \
m1 ^= (temp0 & (m0)); \
m5 ^= (temp1 & (m4)); \
m2 ^= temp0; \
m6 ^= temp1;

/*The bijective function E8, in bitslice form*/
static void E8(hashState *state)
{
uint64 i,roundnumber,temp0,temp1;

for (roundnumber = 0; roundnumber < 42; roundnumber = roundnumber+7) {
/*round 7*roundnumber+0: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+0])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+0])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP1(state->x[1][i]); SWAP1(state->x[3][i]); SWAP1(state->x[5][i]); SWAP1(state->x[7][i]);
}

/*round 7*roundnumber+1: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+1])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+1])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP2(state->x[1][i]); SWAP2(state->x[3][i]); SWAP2(state->x[5][i]); SWAP2(state->x[7][i]);
}

/*round 7*roundnumber+2: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+2])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+2])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP4(state->x[1][i]); SWAP4(state->x[3][i]); SWAP4(state->x[5][i]); SWAP4(state->x[7][i]);
}

/*round 7*roundnumber+3: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+3])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+3])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP8(state->x[1][i]); SWAP8(state->x[3][i]); SWAP8(state->x[5][i]); SWAP8(state->x[7][i]);
}

/*round 7*roundnumber+4: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+4])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+4])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP16(state->x[1][i]); SWAP16(state->x[3][i]); SWAP16(state->x[5][i]); SWAP16(state->x[7][i]);
}

/*round 7*roundnumber+5: Sbox, MDS and Swapping layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+5])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+5])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
SWAP32(state->x[1][i]); SWAP32(state->x[3][i]); SWAP32(state->x[5][i]); SWAP32(state->x[7][i]);
}

/*round 7*roundnumber+6: Sbox and MDS layers*/
for (i = 0; i < 2; i++) {
SS(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i],((uint64*)E8_bitslice_roundconstant[roundnumber+6])[i],((uint64*)E8_bitslice_roundconstant[roundnumber+6])[i+2] );
L(state->x[0][i],state->x[2][i],state->x[4][i],state->x[6][i],state->x[1][i],state->x[3][i],state->x[5][i],state->x[7][i]);
}
/*round 7*roundnumber+6: swapping layer*/
for (i = 1; i < 8; i = i+2) {
temp0 = state->x[i][0]; state->x[i][0] = state->x[i][1]; state->x[i][1] = temp0;
}
}

}

/*The compression function F8 */
static void F8(hashState *state)
{
uint64 i;

/*xor the 512-bit message with the fist half of the 1024-bit hash state*/
for (i = 0; i < 8; i++) state->x[i >> 1][i & 1] ^= ((uint64*)state->buffer)[i];

/*the bijective function E8 */
E8(state);

/*xor the 512-bit message with the second half of the 1024-bit hash state*/
for (i = 0; i < 8; i++) state->x[(8+i) >> 1][(8+i) & 1] ^= ((uint64*)state->buffer)[i];
}

/*before hashing a message, initialize the hash state as H0 */
static HashReturn Init(hashState *state, int hashbitlen)
{
state->databitlen = 0;
state->datasize_in_buffer = 0;

/*initialize the initial hash value of JH*/
state->hashbitlen = hashbitlen;

/*load the intital hash value into state*/
switch (hashbitlen)
{
case 224: memcpy(state->x,JH224_H0,128); break;
case 256: memcpy(state->x,JH256_H0,128); break;
case 384: memcpy(state->x,JH384_H0,128); break;
case 512: memcpy(state->x,JH512_H0,128); break;
}

return(SUCCESS);
}


/*hash each 512-bit message block, except the last partial block*/
static HashReturn Update(hashState *state, const BitSequence *data, DataLength databitlen)
{
DataLength index; /*the starting address of the data to be compressed*/

state->databitlen += databitlen;
index = 0;

/*if there is remaining data in the buffer, fill it to a full message block first*/
/*we assume that the size of the data in the buffer is the multiple of 8 bits if it is not at the end of a message*/

/*There is data in the buffer, but the incoming data is insufficient for a full block*/
if ( (state->datasize_in_buffer > 0 ) && (( state->datasize_in_buffer + databitlen) < 512) ) {
if ( (databitlen & 7) == 0 ) {
memcpy(state->buffer + (state->datasize_in_buffer >> 3), data, 64-(state->datasize_in_buffer >> 3)) ;
}
else memcpy(state->buffer + (state->datasize_in_buffer >> 3), data, 64-(state->datasize_in_buffer >> 3)+1) ;
state->datasize_in_buffer += databitlen;
databitlen = 0;
}

/*There is data in the buffer, and the incoming data is sufficient for a full block*/
if ( (state->datasize_in_buffer > 0 ) && (( state->datasize_in_buffer + databitlen) >= 512) ) {
memcpy( state->buffer + (state->datasize_in_buffer >> 3), data, 64-(state->datasize_in_buffer >> 3) ) ;
index = 64-(state->datasize_in_buffer >> 3);
databitlen = databitlen - (512 - state->datasize_in_buffer);
F8(state);
state->datasize_in_buffer = 0;
}

/*hash the remaining full message blocks*/
for ( ; databitlen >= 512; index = index+64, databitlen = databitlen - 512) {
memcpy(state->buffer, data+index, 64);
F8(state);
}

/*store the partial block into buffer, assume that -- if part of the last byte is not part of the message, then that part consists of 0 bits*/
if ( databitlen > 0) {
if ((databitlen & 7) == 0)
memcpy(state->buffer, data+index, (databitlen & 0x1ff) >> 3);
else
memcpy(state->buffer, data+index, ((databitlen & 0x1ff) >> 3)+1);
state->datasize_in_buffer = databitlen;
}

return(SUCCESS);
}

/*pad the message, process the padded block(s), truncate the hash value H to obtain the message digest*/
static HashReturn Final(hashState *state, BitSequence *hashval)
{
unsigned int i;

if ( (state->databitlen & 0x1ff) == 0 ) {
/*pad the message when databitlen is multiple of 512 bits, then process the padded block*/
memset(state->buffer, 0, 64);
state->buffer[0] = 0x80;
state->buffer[63] = state->databitlen & 0xff;
state->buffer[62] = (state->databitlen >> 8) & 0xff;
state->buffer[61] = (state->databitlen >> 16) & 0xff;
state->buffer[60] = (state->databitlen >> 24) & 0xff;
state->buffer[59] = (state->databitlen >> 32) & 0xff;
state->buffer[58] = (state->databitlen >> 40) & 0xff;
state->buffer[57] = (state->databitlen >> 48) & 0xff;
state->buffer[56] = (state->databitlen >> 56) & 0xff;
F8(state);
}
else {
/*set the rest of the bytes in the buffer to 0*/
if ( (state->datasize_in_buffer & 7) == 0)
for (i = (state->databitlen & 0x1ff) >> 3; i < 64; i++) state->buffer[i] = 0;
else
for (i = ((state->databitlen & 0x1ff) >> 3)+1; i < 64; i++) state->buffer[i] = 0;

/*pad and process the partial block when databitlen is not multiple of 512 bits, then hash the padded blocks*/
state->buffer[((state->databitlen & 0x1ff) >> 3)] |= 1 << (7- (state->databitlen & 7));

F8(state);
memset(state->buffer, 0, 64);
state->buffer[63] = state->databitlen & 0xff;
state->buffer[62] = (state->databitlen >> 8) & 0xff;
state->buffer[61] = (state->databitlen >> 16) & 0xff;
state->buffer[60] = (state->databitlen >> 24) & 0xff;
state->buffer[59] = (state->databitlen >> 32) & 0xff;
state->buffer[58] = (state->databitlen >> 40) & 0xff;
state->buffer[57] = (state->databitlen >> 48) & 0xff;
state->buffer[56] = (state->databitlen >> 56) & 0xff;
F8(state);
}

/*truncating the final hash value to generate the message digest*/
switch(state->hashbitlen) {
case 224: memcpy(hashval,(unsigned char*)state->x+64+36,28); break;
case 256: memcpy(hashval,(unsigned char*)state->x+64+32,32); break;
case 384: memcpy(hashval,(unsigned char*)state->x+64+16,48); break;
case 512: memcpy(hashval,(unsigned char*)state->x+64,64); break;
}

return(SUCCESS);
}

/* hash a message,
three inputs: message digest size in bits (hashbitlen); message (data); message length in bits (databitlen)
one output: message digest (hashval)
*/
HashReturn jh_hash(int hashbitlen, const BitSequence *data,DataLength databitlen, BitSequence *hashval)
{
hashState state;

if ( hashbitlen == 224 || hashbitlen == 256 || hashbitlen == 384 || hashbitlen == 512 ) {
Init(&state, hashbitlen);
Update(&state, data, databitlen);
Final(&state, hashval);
return SUCCESS;
}
else
return(BAD_HASHLEN);
}

+ 19
- 0
crypto/c_jh.h View File

@@ -0,0 +1,19 @@
/*This program gives the 64-bit optimized bitslice implementation of JH using ANSI C

--------------------------------
Performance

Microprocessor: Intel CORE 2 processor (Core 2 Duo Mobile T6600 2.2GHz)
Operating System: 64-bit Ubuntu 10.04 (Linux kernel 2.6.32-22-generic)
Speed for long message:
1) 45.8 cycles/byte compiler: Intel C++ Compiler 11.1 compilation option: icc -O2
2) 56.8 cycles/byte compiler: gcc 4.4.3 compilation option: gcc -O3

--------------------------------
Last Modified: January 16, 2011
*/
#pragma once

#include "hash.h"

HashReturn jh_hash(int hashbitlen, const BitSequence *data, DataLength databitlen, BitSequence *hashval);

+ 128
- 0
crypto/c_keccak.c View File

@@ -0,0 +1,128 @@
// keccak.c
// 19-Nov-11 Markku-Juhani O. Saarinen <mjos@iki.fi>
// A baseline Keccak (3rd round) implementation.

#include <stdint.h>
#include <memory.h>

#define HASH_DATA_AREA 136
#define KECCAK_ROUNDS 24

#ifndef ROTL64
#define ROTL64(x, y) (((x) << (y)) | ((x) >> (64 - (y))))
#endif

const uint64_t keccakf_rndc[24] =
{
0x0000000000000001, 0x0000000000008082, 0x800000000000808a,
0x8000000080008000, 0x000000000000808b, 0x0000000080000001,
0x8000000080008081, 0x8000000000008009, 0x000000000000008a,
0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
0x000000008000808b, 0x800000000000008b, 0x8000000000008089,
0x8000000000008003, 0x8000000000008002, 0x8000000000000080,
0x000000000000800a, 0x800000008000000a, 0x8000000080008081,
0x8000000000008080, 0x0000000080000001, 0x8000000080008008
};

const int keccakf_rotc[24] =
{
1, 3, 6, 10, 15, 21, 28, 36, 45, 55, 2, 14,
27, 41, 56, 8, 25, 43, 62, 18, 39, 61, 20, 44
};

const int keccakf_piln[24] =
{
10, 7, 11, 17, 18, 3, 5, 16, 8, 21, 24, 4,
15, 23, 19, 13, 12, 2, 20, 14, 22, 9, 6, 1
};

// update the state with given number of rounds

void keccakf(uint64_t st[25], int rounds)
{
int i, j, round;
uint64_t t, bc[5];

for (round = 0; round < rounds; ++round) {

// Theta
bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20];
bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21];
bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22];
bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23];
bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24];

for (i = 0; i < 5; ++i) {
t = bc[(i + 4) % 5] ^ ROTL64(bc[(i + 1) % 5], 1);
st[i ] ^= t;
st[i + 5] ^= t;
st[i + 10] ^= t;
st[i + 15] ^= t;
st[i + 20] ^= t;
}

// Rho Pi
t = st[1];
for (i = 0; i < 24; ++i) {
bc[0] = st[keccakf_piln[i]];
st[keccakf_piln[i]] = ROTL64(t, keccakf_rotc[i]);
t = bc[0];
}

// Chi
for (j = 0; j < 25; j += 5) {
bc[0] = st[j ];
bc[1] = st[j + 1];
bc[2] = st[j + 2];
bc[3] = st[j + 3];
bc[4] = st[j + 4];
st[j ] ^= (~bc[1]) & bc[2];
st[j + 1] ^= (~bc[2]) & bc[3];
st[j + 2] ^= (~bc[3]) & bc[4];
st[j + 3] ^= (~bc[4]) & bc[0];
st[j + 4] ^= (~bc[0]) & bc[1];
}

// Iota
st[0] ^= keccakf_rndc[round];
}
}

// compute a keccak hash (md) of given byte length from "in"
typedef uint64_t state_t[25];

void keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen)
{
state_t st;
uint8_t temp[144];
int i, rsiz, rsizw;

rsiz = sizeof(state_t) == mdlen ? HASH_DATA_AREA : 200 - 2 * mdlen;
rsizw = rsiz / 8;
memset(st, 0, sizeof(st));

for ( ; inlen >= rsiz; inlen -= rsiz, in += rsiz) {
for (i = 0; i < rsizw; i++)
st[i] ^= ((uint64_t *) in)[i];
keccakf(st, KECCAK_ROUNDS);
}
// last block and padding
memcpy(temp, in, inlen);
temp[inlen++] = 1;
memset(temp + inlen, 0, rsiz - inlen);
temp[rsiz - 1] |= 0x80;

for (i = 0; i < rsizw; i++)
st[i] ^= ((uint64_t *) temp)[i];

keccakf(st, KECCAK_ROUNDS);

memcpy(md, st, mdlen);
}

void keccak1600(const uint8_t *in, int inlen, uint8_t *md)
{
keccak(in, inlen, md, sizeof(state_t));
}

+ 26
- 0
crypto/c_keccak.h View File

@@ -0,0 +1,26 @@
// keccak.h
// 19-Nov-11 Markku-Juhani O. Saarinen <mjos@iki.fi>

#ifndef KECCAK_H
#define KECCAK_H

#include <stdint.h>
#include <string.h>

#ifndef KECCAK_ROUNDS
#define KECCAK_ROUNDS 24
#endif

#ifndef ROTL64
#define ROTL64(x, y) (((x) << (y)) | ((x) >> (64 - (y))))
#endif

// compute a keccak hash (md) of given byte length from "in"
int keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen);

// update the state
void keccakf(uint64_t st[25], int norounds);

void keccak1600(const uint8_t *in, int inlen, uint8_t *md);

#endif

+ 2036
- 0
crypto/c_skein.c
File diff suppressed because it is too large
View File


+ 47
- 0
crypto/c_skein.h View File

@@ -0,0 +1,47 @@
#ifndef _SKEIN_H_
#define _SKEIN_H_ 1
/**************************************************************************
**
** Interface declarations and internal definitions for Skein hashing.
**
** Source code author: Doug Whiting, 2008.
**
** This algorithm and source code is released to the public domain.
**
***************************************************************************
**
** The following compile-time switches may be defined to control some
** tradeoffs between speed, code size, error checking, and security.
**
** The "default" note explains what happens when the switch is not defined.
**
** SKEIN_DEBUG -- make callouts from inside Skein code
** to examine/display intermediate values.
** [default: no callouts (no overhead)]
**
** SKEIN_ERR_CHECK -- how error checking is handled inside Skein
** code. If not defined, most error checking
** is disabled (for performance). Otherwise,
** the switch value is interpreted as:
** 0: use assert() to flag errors
** 1: return SKEIN_FAIL to flag errors
**
***************************************************************************/
#include "skein_port.h" /* get platform-specific definitions */

typedef enum
{
SKEIN_SUCCESS = 0, /* return codes from Skein calls */
SKEIN_FAIL = 1,
SKEIN_BAD_HASHLEN = 2
}
SkeinHashReturn;

typedef size_t SkeinDataLength; /* bit count type */
typedef u08b_t SkeinBitSequence; /* bit stream type */

/* "all-in-one" call */
SkeinHashReturn skein_hash(int hashbitlen, const SkeinBitSequence *data,
SkeinDataLength databitlen, SkeinBitSequence *hashval);

#endif /* ifndef _SKEIN_H_ */

+ 33
- 0
crypto/cryptonight.h View File

@@ -0,0 +1,33 @@
#ifndef __CRYPTONIGHT_H_INCLUDED
#define __CRYPTONIGHT_H_INCLUDED

#ifdef __cplusplus
extern "C" {
#endif

#include <stddef.h>
#include <inttypes.h>

#define MEMORY 2097152

typedef struct {
uint8_t hash_state[224]; // Need only 200, explicit align
uint8_t* long_state;
uint8_t ctx_info[24]; //Use some of the extra memory for flags
} cryptonight_ctx;

typedef struct {
const char* warning;
} alloc_msg;

size_t cryptonight_init(size_t use_fast_mem, size_t use_mlock, alloc_msg* msg);
cryptonight_ctx* cryptonight_alloc_ctx(size_t use_fast_mem, size_t use_mlock, alloc_msg* msg);
void cryptonight_free_ctx(cryptonight_ctx* ctx);
void cryptonight_hash_ctx(const void* input, size_t len, void* output, cryptonight_ctx* ctx);
void cryptonight_double_hash_ctx(const void* input, size_t len, void* output, cryptonight_ctx* __restrict ctx0, cryptonight_ctx* __restrict ctx1);

#ifdef __cplusplus
}
#endif

#endif

+ 344
- 0
crypto/cryptonight_aesni.c View File

@@ -0,0 +1,344 @@
/*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/

#include "cryptonight.h"
#include <memory.h>
#include <stdio.h>

#ifdef __GNUC__
#include <x86intrin.h>
static inline uint64_t _umul128(uint64_t a, uint64_t b, uint64_t* hi)
{
unsigned __int128 r = (unsigned __int128)a * (unsigned __int128)b;
*hi = r >> 64;
return (uint64_t)r;
}

#define _mm256_set_m128i(v0, v1) _mm256_insertf128_si256(_mm256_castsi128_si256(v1), (v0), 1)
#else
#include <intrin.h>
#endif // __GNUC__

#if !defined(_LP64) && !defined(_WIN64)
#error You are trying to do a 32-bit build. This will all end in tears. I know it.
#endif

void keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen);
void keccakf(uint64_t st[25], int rounds);
extern void(*const extra_hashes[4])(const void *, size_t, char *);

// This will shift and xor tmp1 into itself as 4 32-bit vals such as
// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
static inline __m128i sl_xor(__m128i tmp1)
{
__m128i tmp4;
tmp4 = _mm_slli_si128(tmp1, 0x04);
tmp1 = _mm_xor_si128(tmp1, tmp4);
tmp4 = _mm_slli_si128(tmp4, 0x04);
tmp1 = _mm_xor_si128(tmp1, tmp4);
tmp4 = _mm_slli_si128(tmp4, 0x04);
tmp1 = _mm_xor_si128(tmp1, tmp4);
return tmp1;
}

static inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3,
__m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
{
__m128i xout0, xout1, xout2;

xout0 = _mm_load_si128(memory);
xout2 = _mm_load_si128(memory+1);
*k0 = xout0;
*k1 = xout2;

xout1 = _mm_aeskeygenassist_si128(xout2, 0x01);
xout1 = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
xout0 = sl_xor(xout0);
xout0 = _mm_xor_si128(xout0, xout1);

xout1 = _mm_aeskeygenassist_si128(xout0, 0x00);
xout1 = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
xout2 = sl_xor(xout2);
xout2 = _mm_xor_si128(xout2, xout1);
*k2 = xout0;
*k3 = xout2;

xout1 = _mm_aeskeygenassist_si128(xout2, 0x02);
xout1 = _mm_shuffle_epi32(xout1, 0xFF);
xout0 = sl_xor(xout0);
xout0 = _mm_xor_si128(xout0, xout1);

xout1 = _mm_aeskeygenassist_si128(xout0, 0x00);
xout1 = _mm_shuffle_epi32(xout1, 0xAA);
xout2 = sl_xor(xout2);
xout2 = _mm_xor_si128(xout2, xout1);
*k4 = xout0;
*k5 = xout2;

xout1 = _mm_aeskeygenassist_si128(xout2, 0x04);
xout1 = _mm_shuffle_epi32(xout1, 0xFF);
xout0 = sl_xor(xout0);
xout0 = _mm_xor_si128(xout0, xout1);

xout1 = _mm_aeskeygenassist_si128(xout0, 0x00);
xout1 = _mm_shuffle_epi32(xout1, 0xAA);
xout2 = sl_xor(xout2);
xout2 = _mm_xor_si128(xout2, xout1);
*k6 = xout0;
*k7 = xout2;

xout1 = _mm_aeskeygenassist_si128(xout2, 0x08);
xout1 = _mm_shuffle_epi32(xout1, 0xFF);
xout0 = sl_xor(xout0);
xout0 = _mm_xor_si128(xout0, xout1);

xout1 = _mm_aeskeygenassist_si128(xout0, 0x00);
xout1 = _mm_shuffle_epi32(xout1, 0xAA);
xout2 = sl_xor(xout2);
xout2 = _mm_xor_si128(xout2, xout1);
*k8 = xout0;
*k9 = xout2;
}

static inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
{
*x0 = _mm_aesenc_si128(*x0, key);
*x1 = _mm_aesenc_si128(*x1, key);
*x2 = _mm_aesenc_si128(*x2, key);
*x3 = _mm_aesenc_si128(*x3, key);
*x4 = _mm_aesenc_si128(*x4, key);
*x5 = _mm_aesenc_si128(*x5, key);
*x6 = _mm_aesenc_si128(*x6, key);
*x7 = _mm_aesenc_si128(*x7, key);
}

void cn_explode_scratchpad(const __m128i* input, __m128i* output)
{
// This is more than we have registers, compiler will assign 2 keys on the stack
__m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
__m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;

aes_genkey(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);

xin0 = _mm_load_si128(input + 4);
xin1 = _mm_load_si128(input + 5);
xin2 = _mm_load_si128(input + 6);
xin3 = _mm_load_si128(input + 7);
xin4 = _mm_load_si128(input + 8);
xin5 = _mm_load_si128(input + 9);
xin6 = _mm_load_si128(input + 10);
xin7 = _mm_load_si128(input + 11);

for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8)
{
aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);

_mm_store_si128(output + i + 0, xin0);
_mm_store_si128(output + i + 1, xin1);
_mm_store_si128(output + i + 2, xin2);
_mm_store_si128(output + i + 3, xin3);
_mm_prefetch(output + i + 0, _MM_HINT_T2);
_mm_store_si128(output + i + 4, xin4);
_mm_store_si128(output + i + 5, xin5);
_mm_store_si128(output + i + 6, xin6);
_mm_store_si128(output + i + 7, xin7);
_mm_prefetch(output + i + 4, _MM_HINT_T2);
}
}