Ecosystem/xmrig
Ecosystem/xmrig
1
0
Fork 0
mirror of https://github.com/xmrig/xmrig.git synced 2025-01-11 05:14:40 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge remote-tracking branch 'upstream/evo' into evo

Browse source
This commit is contained in:
SChernykh 2019-09-13 15:27:14 +02:00
commit 6eba700ecf
335 changed files with 38529 additions and 4035 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
.gitignore vendored
View file

@ -1,2 +1,4 @@
/build
/CMakeLists.txt.user
/.idea
/src/backend/opencl/cl/cn/cryptonight_gen.cl

13
CHANGELOG.md
View file

@ -1,3 +1,16 @@
# v3.1.1
- [#1133](https://github.com/xmrig/xmrig/issues/1133) Fixed syslog regression.
- [#1138](https://github.com/xmrig/xmrig/issues/1138) Fixed multiple network bugs.
- [#1141](https://github.com/xmrig/xmrig/issues/1141) Fixed log in background mode.
- [#1142](https://github.com/xmrig/xmrig/pull/1142) RandomX hashrate improved by 0.5-1.5% depending on variant and CPU.
- [#1146](https://github.com/xmrig/xmrig/pull/1146) Fixed race condition in RandomX thread init.
- [#1148](https://github.com/xmrig/xmrig/pull/1148) Fixed, on Linux linker marking entire executable as having an executable stack.
- Fixed, for Argon2 algorithms command line options like `--threads` was ignored.
- Fixed command line options for single pool, free order allowed again.
# v3.1.0
- [#1107](https://github.com/xmrig/xmrig/issues/1107#issuecomment-522235892) Added Argon2 algorithm family: `argon2/chukwa` and `argon2/wrkz`.
# v3.0.0
- **[#1111](https://github.com/xmrig/xmrig/pull/1111) Added RandomX (`rx/test`) algorithm for testing and benchmarking.**
- **[#1036](https://github.com/xmrig/xmrig/pull/1036) Added RandomWOW (`rx/wow`) algorithm for [Wownero](http://wownero.org/).**

83
CMakeLists.txt
View file

@ -1,18 +1,22 @@
cmake_minimum_required(VERSION 2.8)
project(xmrig)
option(WITH_LIBCPUID "Use Libcpuid" ON)
option(WITH_HWLOC "Use hwloc" ON)
option(WITH_CN_LITE "CryptoNight-Lite support" ON)
option(WITH_CN_HEAVY "CryptoNight-Heavy support" ON)
option(WITH_CN_PICO "CryptoNight-Pico support" ON)
option(WITH_CN_GPU "CryptoNight-GPU support" ON)
option(WITH_RANDOMX "RandomX support" ON)
option(WITH_HTTP "HTTP protocol support (client/server)" ON)
option(WITH_LIBCPUID "Enable libcpuid support" ON)
option(WITH_HWLOC "Enable hwloc support" ON)
option(WITH_CN_LITE "Enable CryptoNight-Lite algorithms family" ON)
option(WITH_CN_HEAVY "Enable CryptoNight-Heavy algorithms family" ON)
option(WITH_CN_PICO "Enable CryptoNight-Pico algorithm" ON)
option(WITH_CN_GPU "Enable CryptoNight-GPU algorithm" ON)
option(WITH_RANDOMX "Enable RandomX algorithms family" ON)
option(WITH_ARGON2 "Enable Argon2 algorithms family" ON)
option(WITH_HTTP "Enable HTTP protocol support (client/server)" ON)
option(WITH_DEBUG_LOG "Enable debug log output" OFF)
option(WITH_TLS "Enable OpenSSL support" ON)
option(WITH_ASM "Enable ASM PoW implementations" ON)
option(WITH_EMBEDDED_CONFIG "Enable internal embedded JSON config" OFF)
option(WITH_OPENCL "Enable OpenCL backend" ON)
option(WITH_STRICT_CACHE "Enable strict checks for OpenCL cache" ON)
option(WITH_INTERLEAVE_DEBUG_LOG "Enable debug log for threads interleave" OFF)
option(BUILD_STATIC "Build static binary" OFF)
option(ARM_TARGET "Force use specific ARM target 8 or 7" 0)
@ -151,66 +155,9 @@ add_definitions(/DUNICODE)
find_package(UV REQUIRED)
if (WITH_RANDOMX)
include_directories(src/crypto/randomx)
add_definitions(/DXMRIG_ALGO_RANDOMX)
set(SOURCES_CRYPTO
"${SOURCES_CRYPTO}"
src/crypto/randomx/aes_hash.cpp
src/crypto/randomx/allocator.cpp
src/crypto/randomx/argon2_core.c
src/crypto/randomx/argon2_ref.c
src/crypto/randomx/blake2_generator.cpp
src/crypto/randomx/blake2/blake2b.c
src/crypto/randomx/bytecode_machine.cpp
src/crypto/randomx/dataset.cpp
src/crypto/randomx/instructions_portable.cpp
src/crypto/randomx/randomx.cpp
src/crypto/randomx/reciprocal.c
src/crypto/randomx/soft_aes.cpp
src/crypto/randomx/superscalar.cpp
src/crypto/randomx/virtual_machine.cpp
src/crypto/randomx/virtual_memory.cpp
src/crypto/randomx/vm_compiled_light.cpp
src/crypto/randomx/vm_compiled.cpp
src/crypto/randomx/vm_interpreted_light.cpp
src/crypto/randomx/vm_interpreted.cpp
src/crypto/rx/Rx.cpp
src/crypto/rx/Rx.h
src/crypto/rx/RxAlgo.cpp
src/crypto/rx/RxAlgo.h
src/crypto/rx/RxCache.cpp
src/crypto/rx/RxCache.h
src/crypto/rx/RxConfig.cpp
src/crypto/rx/RxConfig.h
src/crypto/rx/RxDataset.cpp
src/crypto/rx/RxDataset.h
src/crypto/rx/RxVm.cpp
src/crypto/rx/RxVm.h
)
if (NOT ARCH_ID)
set(ARCH_ID ${CMAKE_HOST_SYSTEM_PROCESSOR})
endif()
if (CMAKE_C_COMPILER_ID MATCHES MSVC)
enable_language(ASM_MASM)
list(APPEND SOURCES_CRYPTO
src/crypto/randomx/jit_compiler_x86_static.asm
src/crypto/randomx/jit_compiler_x86.cpp
)
elseif (NOT XMRIG_ARM AND CMAKE_SIZEOF_VOID_P EQUAL 8)
list(APPEND SOURCES_CRYPTO
src/crypto/randomx/jit_compiler_x86_static.S
src/crypto/randomx/jit_compiler_x86.cpp
)
# cheat because cmake and ccache hate each other
set_property(SOURCE src/crypto/randomx/jit_compiler_x86_static.S PROPERTY LANGUAGE C)
endif()
else()
remove_definitions(/DXMRIG_ALGO_RANDOMX)
endif()
include(cmake/flags.cmake)
include(cmake/randomx.cmake)
include(cmake/argon2.cmake)
include(cmake/OpenSSL.cmake)
include(cmake/asm.cmake)
include(cmake/cn-gpu.cmake)
@ -244,4 +191,4 @@ if (WITH_DEBUG_LOG)
endif()
add_executable(${CMAKE_PROJECT_NAME} ${HEADERS} ${SOURCES} ${SOURCES_OS} ${SOURCES_CPUID} ${HEADERS_CRYPTO} ${SOURCES_CRYPTO} ${SOURCES_SYSLOG} ${TLS_SOURCES} ${XMRIG_ASM_SOURCES} ${CN_GPU_SOURCES})
target_link_libraries(${CMAKE_PROJECT_NAME} ${XMRIG_ASM_LIBRARY} ${OPENSSL_LIBRARIES} ${UV_LIBRARIES} ${EXTRA_LIBS} ${CPUID_LIB})
target_link_libraries(${CMAKE_PROJECT_NAME} ${XMRIG_ASM_LIBRARY} ${OPENSSL_LIBRARIES} ${UV_LIBRARIES} ${EXTRA_LIBS} ${CPUID_LIB} ${ARGON2_LIBRARY})

18
cmake/argon2.cmake Normal file
View file

@ -0,0 +1,18 @@
if (WITH_ARGON2)
add_definitions(/DXMRIG_ALGO_ARGON2)
list(APPEND HEADERS_CRYPTO
src/crypto/argon2/Hash.h
src/crypto/argon2/Impl.h
)
list(APPEND SOURCES_CRYPTO
src/crypto/argon2/Impl.cpp
)
add_subdirectory(src/3rdparty/argon2)
set(ARGON2_LIBRARY argon2)
else()
remove_definitions(/DXMRIG_ALGO_ARGON2)
set(ARGON2_LIBRARY "")
endif()

61
cmake/randomx.cmake Normal file
View file

@ -0,0 +1,61 @@
if (WITH_RANDOMX)
add_definitions(/DXMRIG_ALGO_RANDOMX)
list(APPEND HEADERS_CRYPTO
src/crypto/rx/Rx.h
src/crypto/rx/RxAlgo.h
src/crypto/rx/RxCache.h
src/crypto/rx/RxConfig.h
src/crypto/rx/RxDataset.h
src/crypto/rx/RxVm.h
)
list(APPEND SOURCES_CRYPTO
src/crypto/randomx/aes_hash.cpp
src/crypto/randomx/allocator.cpp
src/crypto/randomx/argon2_core.c
src/crypto/randomx/argon2_ref.c
src/crypto/randomx/blake2_generator.cpp
src/crypto/randomx/blake2/blake2b.c
src/crypto/randomx/bytecode_machine.cpp
src/crypto/randomx/dataset.cpp
src/crypto/randomx/instructions_portable.cpp
src/crypto/randomx/randomx.cpp
src/crypto/randomx/reciprocal.c
src/crypto/randomx/soft_aes.cpp
src/crypto/randomx/superscalar.cpp
src/crypto/randomx/virtual_machine.cpp
src/crypto/randomx/virtual_memory.cpp
src/crypto/randomx/vm_compiled_light.cpp
src/crypto/randomx/vm_compiled.cpp
src/crypto/randomx/vm_interpreted_light.cpp
src/crypto/randomx/vm_interpreted.cpp
src/crypto/rx/Rx.cpp
src/crypto/rx/RxAlgo.cpp
src/crypto/rx/RxCache.cpp
src/crypto/rx/RxConfig.cpp
src/crypto/rx/RxDataset.cpp
src/crypto/rx/RxVm.cpp
)
if (CMAKE_C_COMPILER_ID MATCHES MSVC)
enable_language(ASM_MASM)
list(APPEND SOURCES_CRYPTO
src/crypto/randomx/jit_compiler_x86_static.asm
src/crypto/randomx/jit_compiler_x86.cpp
)
elseif (NOT XMRIG_ARM AND CMAKE_SIZEOF_VOID_P EQUAL 8)
list(APPEND SOURCES_CRYPTO
src/crypto/randomx/jit_compiler_x86_static.S
src/crypto/randomx/jit_compiler_x86.cpp
)
# cheat because cmake and ccache hate each other
set_property(SOURCE src/crypto/randomx/jit_compiler_x86_static.S PROPERTY LANGUAGE C)
endif()
if (CMAKE_CXX_COMPILER_ID MATCHES Clang)
set_source_files_properties(src/crypto/randomx/jit_compiler_x86.cpp PROPERTIES COMPILE_FLAGS -Wno-unused-const-variable)
endif()
else()
remove_definitions(/DXMRIG_ALGO_RANDOMX)
endif()

28
doc/ALGORITHMS.md
View file

@ -23,4 +23,30 @@ Since version 3 mining [algorithm](#algorithm-names) should specified for each p
## Algorithm names
* https://xmrig.com/docs/algorithms
| Name | Memory | Version | Notes |
|------|--------|---------|-------|
| `argon2/chukwa` | 512 KB | 3.1.0+ | Argon2id (Chukwa). |
| `argon2/wrkz` | 256 KB | 3.1.0+ | Argon2id (WRKZ) |
| `rx/test` | 2 MB | 3.0.0+ | RandomX (reference configuration). |
| `rx/0` | 2 MB | 3.0.0+ | RandomX (reference configuration), reserved for future use. |
| `rx/wow` | 1 MB | 3.0.0+ | RandomWOW. |
| `rx/loki` | 2 MB | 3.0.0+ | RandomXL. |
| `cn/fast` | 2 MB | 3.0.0+ | CryptoNight variant 1 with half iterations. |
| `cn/rwz` | 2 MB | 2.14.0+ | CryptoNight variant 2 with 3/4 iterations and reversed shuffle operation. |
| `cn/zls` | 2 MB | 2.14.0+ | CryptoNight variant 2 with 3/4 iterations. |
| `cn/double` | 2 MB | 2.14.0+ | CryptoNight variant 2 with double iterations. |
| `cn/r` | 2 MB | 2.13.0+ | CryptoNightR (Monero's variant 4). |
| `cn/wow` | 2 MB | 2.12.0+ | CryptoNightR (Wownero). |
| `cn/gpu` | 2 MB | 2.11.0+ | CryptoNight-GPU. |
| `cn-pico` | 256 KB | 2.10.0+ | CryptoNight-Pico. |
| `cn/half` | 2 MB | 2.9.0+ | CryptoNight variant 2 with half iterations. |
| `cn/2` | 2 MB | 2.8.0+ | CryptoNight variant 2. |
| `cn/xao` | 2 MB | 2.6.4+ | CryptoNight variant 0 (modified). |
| `cn/rto` | 2 MB | 2.6.4+ | CryptoNight variant 1 (modified). |
| `cn-heavy/tube` | 4 MB | 2.6.4+ | CryptoNight-Heavy (modified). |
| `cn-heavy/xhv` | 4 MB | 2.6.3+ | CryptoNight-Heavy (modified). |
| `cn-heavy/0` | 4 MB | 2.6.0+ | CryptoNight-Heavy. |
| `cn/1` | 2 MB | 2.5.0+ | CryptoNight variant 1. |
| `cn-lite/1` | 1 MB | 2.5.0+ | CryptoNight-Lite variant 1. |
| `cn-lite/0` | 1 MB | 0.8.0+ | CryptoNight-Lite variant 0. |
| `cn/0` | 2 MB | 0.5.0+ | CryptoNight (original). |

6
doc/CHANGELOG_OLD.md
View file

@ -1,3 +1,9 @@
# v2.99.6-beta
- Added commands `pause` and `resume` via JSON RPC 2.0 API (`POST /json_rpc`).
- Fixed autoconfig regression (since 2.99.5), mostly `rx/wow` was affected by this bug.
- Fixed user job recovery after donation round.
- Information about AVX2 CPU feature how hidden in miner summary.
# v2.99.5-beta
- [#1066](https://github.com/xmrig/xmrig/issues/1066#issuecomment-518080529) Fixed crash and added error message if pool not ready for RandomX.
- [#1092](https://github.com/xmrig/xmrig/issues/1092) Fixed crash if wrong CPU affinity used.

5
doc/CPU.md
View file

@ -1,7 +1,5 @@
# CPU backend
**Information in this document actual to version 2.99.5+**
All CPU related settings contains in one `cpu` object in config file, CPU backend allow specify multiple profiles and allow switch between them without restrictions by pool request or config change. Default auto-configuration create reasonable minimum of profiles which cover all supported algorithms.
### Example
@ -93,3 +91,6 @@ Mining threads priority, value from `1` (lowest priority) to `5` (highest possib
#### `asm`
Enable/configure or disable ASM optimizations. Possible values: `true`, `false`, `"intel"`, `"ryzen"`, `"bulldozer"`.
#### `argon2-impl` (since v3.1.0)
Allow override automatically detected Argon2 implementation, this option added mostly for debug purposes, default value `null` means autodetect. Other possible values: `"x86_64"`, `"SSE2"`, `"SSSE3"`, `"XOP"`, `"AVX2"`, `"AVX-512F"`. Manual selection has no safe guards, if you CPU not support required instuctions, miner will crash.

9
doc/PERSISTENT_OPTIONS.md Normal file
View file

@ -0,0 +1,9 @@
# Persistent options
Options in list below can't changed in runtime by watching config file or via API.
* `background`
* `donate-level`
* `cpu/argon2-impl`
* `opencl/loader`
* `opencl/platform`

32
doc/data/algorithms.json
View file

@ -1,32 +0,0 @@
{
"current": [
["rx/test", "2 MB", "2.99.5+", "RandomX (reference configuration)."],
["rx/0", "2 MB", "2.99.0+", "RandomX (reference configuration), reserved for future use."],
["rx/wow", "1 MB", "2.99.0+", "RandomWOW."],
["rx/loki", "2 MB", "2.99.0+", "RandomXL."],
["cn/fast", "2 MB", "2.99.0+", "CryptoNight variant 1 with half iterations."],
["cn/rwz", "2 MB", "2.14.0+", "CryptoNight variant 2 with 3/4 iterations and reversed shuffle operation."],
["cn/zls", "2 MB", "2.14.0+", "CryptoNight variant 2 with 3/4 iterations."],
["cn/double", "2 MB", "2.14.0+", "CryptoNight variant 2 with double iterations."],
["cn/r", "2 MB", "2.13.0+", "CryptoNightR (Monero's variant 4)."],
["cn/wow", "2 MB", "2.12.0+", "CryptoNightR (Wownero)."],
["cn/gpu", "2 MB", "2.11.0+", "CryptoNight-GPU."],
["cn-pico", "256 KB", "2.10.0+", "CryptoNight-Pico."],
["cn/half", "2 MB", "2.9.0+", "CryptoNight variant 2 with half iterations."],
["cn/2", "2 MB", "2.8.0+", "CryptoNight variant 2."],
["cn/xao", "2 MB", "2.6.4+", "CryptoNight variant 0 (modified)."],
["cn/rto", "2 MB", "2.6.4+", "CryptoNight variant 1 (modified)."],
["cn-heavy/tube", "4 MB", "2.6.4+", "CryptoNight-Heavy (modified)."],
["cn-heavy/xhv", "4 MB", "2.6.3+", "CryptoNight-Heavy (modified)."],
["cn-heavy/0", "4 MB", "2.6.0+", "CryptoNight-Heavy."],
["cn/1", "2 MB", "2.5.0+", "CryptoNight variant 1."],
["cn-lite/1", "1 MB", "2.5.0+", "CryptoNight-Lite variant 1."],
["cn-lite/0", "1 MB", "0.8.0+", "CryptoNight-Lite variant 0."],
["cn/0", "2 MB", "0.5.0+", "CryptoNight (original)."]
],
"removed": [
["cn/msr", "2 MB", "2.6.3+", "Renamed to cn/fast, still supported as alias."],
["cn/xtl", "2 MB", "2.6.1-2.16.0", "Coin forked to cn/half."],
["cn-lite/ipbc", "1 MB", "2.6.1-2.6.3", "Coin forked to cn-heavy/tube."]
]
}

23
package.json Normal file
View file

@ -0,0 +1,23 @@
{
"name": "xmrig",
"version": "3.0.0",
"description": "RandomX, CryptoNight and Argon2 miner",
"main": "index.js",
"directories": {
"doc": "doc"
},
"scripts": {
"build": "node scripts/generate_cl.js"
},
"repository": {
"type": "git",
"url": "git+https://github.com/xmrig/xmrig.git"
},
"keywords": [],
"author": "",
"license": "GPLv3",
"bugs": {
"url": "https://github.com/xmrig/xmrig/issues"
},
"homepage": "https://github.com/xmrig/xmrig#readme"
}

86
scripts/generate_cl.js Normal file
View file

@ -0,0 +1,86 @@
#!/usr/bin/env node
'use strict';
const fs = require('fs');
const path = require('path');
const { text2h, text2h_bundle, addIncludes } = require('./js/opencl');
const { opencl_minify } = require('./js/opencl_minify');
const cwd = process.cwd();
function cn()
{
const cn = opencl_minify(addIncludes('cryptonight.cl', [
'algorithm.cl',
'wolf-aes.cl',
'wolf-skein.cl',
'jh.cl',
'blake256.cl',
'groestl256.cl',
'fast_int_math_v2.cl',
'fast_div_heavy.cl',
'keccak.cl'
]));
// fs.writeFileSync('cryptonight_gen.cl', cn);
fs.writeFileSync('cryptonight_cl.h', text2h(cn, 'xmrig', 'cryptonight_cl'));
}
function cn_r()
{
const items = {};
items.cryptonight_r_defines_cl = opencl_minify(addIncludes('cryptonight_r_defines.cl', [ 'wolf-aes.cl' ]));
items.cryptonight_r_cl = opencl_minify(fs.readFileSync('cryptonight_r.cl', 'utf8'));
// for (let key in items) {
// fs.writeFileSync(key + '_gen.cl', items[key]);
// }
fs.writeFileSync('cryptonight_r_cl.h', text2h_bundle('xmrig', items));
}
function cn_gpu()
{
const cn_gpu = opencl_minify(addIncludes('cryptonight_gpu.cl', [ 'wolf-aes.cl', 'keccak.cl' ]));
// fs.writeFileSync('cryptonight_gpu_gen.cl', cn_gpu);
fs.writeFileSync('cryptonight_gpu_cl.h', text2h(cn_gpu, 'xmrig', 'cryptonight_gpu_cl'));
}
function rx()
{
let rx = addIncludes('randomx.cl', [
'../cn/algorithm.cl',
'randomx_constants_monero.h',
'randomx_constants_wow.h',
'randomx_constants_loki.h',
'aes.cl',
'blake2b.cl',
'randomx_vm.cl',
'randomx_jit.cl'
]);
rx = rx.replace(/(\t| )*#include "fillAes1Rx4.cl"/g, fs.readFileSync('fillAes1Rx4.cl', 'utf8'));
rx = rx.replace(/(\t| )*#include "blake2b_double_block.cl"/g, fs.readFileSync('blake2b_double_block.cl', 'utf8'));
rx = opencl_minify(rx);
//fs.writeFileSync('randomx_gen.cl', rx);
fs.writeFileSync('randomx_cl.h', text2h(rx, 'xmrig', 'randomx_cl'));
}
process.chdir(path.resolve('src/backend/opencl/cl/cn'));
cn();
cn_r();
cn_gpu();
process.chdir(cwd);
process.chdir(path.resolve('src/backend/opencl/cl/rx'));
rx();

91
scripts/js/opencl.js Normal file
View file

@ -0,0 +1,91 @@
'use strict';
const fs = require('fs');
function bin2h(buf, namespace, name)
{
const size = buf.byteLength;
let out = `#pragma once\n\nnamespace ${namespace} {\n\nstatic unsigned char ${name}[${size}] = {\n `;
let b = 32;
for (let i = 0; i < size; i++) {
out += `0x${buf.readUInt8(i).toString(16).padStart(2, '0')}${size - i > 1 ? ',' : ''}`;
if (--b === 0) {
b = 32;
out += '\n ';
}
}
out += `\n};\n\n} // namespace ${namespace}\n`;
return out;
}
function text2h_internal(text, name)
{
const buf = Buffer.from(text);
const size = buf.byteLength;
let out = `\nstatic char ${name}[${size + 1}] = {\n `;
let b = 32;
for (let i = 0; i < size; i++) {
out += `0x${buf.readUInt8(i).toString(16).padStart(2, '0')},`;
if (--b === 0) {
b = 32;
out += '\n ';
}
}
out += '0x00';
out += '\n};\n';
return out;
}
function text2h(text, namespace, name)
{
return `#pragma once\n\nnamespace ${namespace} {\n` + text2h_internal(text, name) + `\n} // namespace ${namespace}\n`;
}
function text2h_bundle(namespace, items)
{
let out = `#pragma once\n\nnamespace ${namespace} {\n`;
for (let key in items) {
out += text2h_internal(items[key], key);
}
return out + `\n} // namespace ${namespace}\n`;
}
function addInclude(input, name)
{
return input.replace(`#include "${name}"`, fs.readFileSync(name, 'utf8'));
}
function addIncludes(inputFileName, names)
{
let data = fs.readFileSync(inputFileName, 'utf8');
for (let name of names) {
data = addInclude(data, name);
}
return data;
}
module.exports.bin2h = bin2h;
module.exports.text2h = text2h;
module.exports.text2h_bundle = text2h_bundle;
module.exports.addInclude = addInclude;
module.exports.addIncludes = addIncludes;

50
scripts/js/opencl_minify.js Normal file
View file

@ -0,0 +1,50 @@
'use strict';
function opencl_minify(input)
{
let out = input.replace(/\/\*[\s\S]*?\*\/|\/\/.*$/gm, ''); // comments
out = out.replace(/^#\s+/gm, '#'); // macros with spaces
out = out.replace(/\n{2,}/g, '\n'); // empty lines
out = out.replace(/^\s+/gm, ''); // leading whitespace
out = out.replace(/ {2,}/g, ' '); // extra whitespace
let array = out.split('\n').map(line => {
if (line[0] === '#') {
return line
}
line = line.replace(/, /g, ',');
line = line.replace(/ \? /g, '?');
line = line.replace(/ : /g, ':');
line = line.replace(/ = /g, '=');
line = line.replace(/ != /g, '!=');
line = line.replace(/ >= /g, '>=');
line = line.replace(/ <= /g, '<=');
line = line.replace(/ == /g, '==');
line = line.replace(/ \+= /g, '+=');
line = line.replace(/ -= /g, '-=');
line = line.replace(/ \|= /g, '|=');
line = line.replace(/ \| /g, '|');
line = line.replace(/ \|\| /g, '||');
line = line.replace(/ & /g, '&');
line = line.replace(/ && /g, '&&');
line = line.replace(/ > /g, '>');
line = line.replace(/ < /g, '<');
line = line.replace(/ \+ /g, '+');
line = line.replace(/ - /g, '-');
line = line.replace(/ \* /g, '*');
line = line.replace(/ \^ /g, '^');
line = line.replace(/ & /g, '&');
line = line.replace(/ \/ /g, '/');
line = line.replace(/ << /g, '<<');
line = line.replace(/ >> /g, '>>');
line = line.replace(/if \(/g, 'if(');
return line;
});
return array.join('\n');
}
module.exports.opencl_minify = opencl_minify;

25
src/3rdparty/CL/LICENSE vendored Normal file
View file

@ -0,0 +1,25 @@
Copyright (c) 2008-2015 The Khronos Group Inc.
Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and/or associated documentation files (the
"Materials"), to deal in the Materials without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Materials, and to
permit persons to whom the Materials are furnished to do so, subject to
the following conditions:
The above copyright notice and this permission notice shall be included
in all copies or substantial portions of the Materials.
MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
https://www.khronos.org/registry/
THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.

50
src/3rdparty/CL/README.md vendored Normal file
View file

@ -0,0 +1,50 @@
# OpenCL<sup>TM</sup> API Headers
This repository contains C language headers for the OpenCL API.
The authoritative public repository for these headers is located at:
https://github.com/KhronosGroup/OpenCL-Headers
Issues, proposed fixes for issues, and other suggested changes should be
created using Github.
## Branch Structure
The OpenCL API headers in this repository are Unified headers and are designed
to work with all released OpenCL versions. This differs from previous OpenCL
API headers, where version-specific API headers either existed in separate
branches, or in separate folders in a branch.
## Compiling for a Specific OpenCL Version
By default, the OpenCL API headers in this repository are for the latest
OpenCL version (currently OpenCL 2.2). To use these API headers to target
a different OpenCL version, an application may `#define` the preprocessor
value `CL_TARGET_OPENCL_VERSION` before including the OpenCL API headers.
The `CL_TARGET_OPENCL_VERSION` is a three digit decimal value representing
the OpenCL API version.
For example, to enforce usage of no more than the OpenCL 1.2 APIs, you may
include the OpenCL API headers as follows:
```
#define CL_TARGET_OPENCL_VERSION 120
#include <CL/opencl.h>
```
## Directory Structure
```
README.md This file
LICENSE Source license for the OpenCL API headers
CL/ Unified OpenCL API headers tree
```
## License
See [LICENSE](LICENSE).
---
OpenCL and the OpenCL logo are trademarks of Apple Inc. used by permission by Khronos.

1804
src/3rdparty/CL/cl.h vendored Normal file
View file

File diff suppressed because it is too large Load diff

131
src/3rdparty/CL/cl_d3d10.h vendored Normal file
View file

@ -0,0 +1,131 @@
/**********************************************************************************
* Copyright (c) 2008-2015 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
/* $Revision: 11708 $ on $Date: 2010-06-13 23:36:24 -0700 (Sun, 13 Jun 2010) $ */
#ifndef __OPENCL_CL_D3D10_H
#define __OPENCL_CL_D3D10_H
#include <d3d10.h>
#include <CL/cl.h>
#include <CL/cl_platform.h>
#ifdef __cplusplus
extern "C" {
#endif
/******************************************************************************
* cl_khr_d3d10_sharing */
#define cl_khr_d3d10_sharing 1
typedef cl_uint cl_d3d10_device_source_khr;
typedef cl_uint cl_d3d10_device_set_khr;
/******************************************************************************/
/* Error Codes */
#define CL_INVALID_D3D10_DEVICE_KHR -1002
#define CL_INVALID_D3D10_RESOURCE_KHR -1003
#define CL_D3D10_RESOURCE_ALREADY_ACQUIRED_KHR -1004
#define CL_D3D10_RESOURCE_NOT_ACQUIRED_KHR -1005
/* cl_d3d10_device_source_nv */
#define CL_D3D10_DEVICE_KHR 0x4010
#define CL_D3D10_DXGI_ADAPTER_KHR 0x4011
/* cl_d3d10_device_set_nv */
#define CL_PREFERRED_DEVICES_FOR_D3D10_KHR 0x4012
#define CL_ALL_DEVICES_FOR_D3D10_KHR 0x4013
/* cl_context_info */
#define CL_CONTEXT_D3D10_DEVICE_KHR 0x4014
#define CL_CONTEXT_D3D10_PREFER_SHARED_RESOURCES_KHR 0x402C
/* cl_mem_info */
#define CL_MEM_D3D10_RESOURCE_KHR 0x4015
/* cl_image_info */
#define CL_IMAGE_D3D10_SUBRESOURCE_KHR 0x4016
/* cl_command_type */
#define CL_COMMAND_ACQUIRE_D3D10_OBJECTS_KHR 0x4017
#define CL_COMMAND_RELEASE_D3D10_OBJECTS_KHR 0x4018
/******************************************************************************/
typedef CL_API_ENTRY cl_int (CL_API_CALL *clGetDeviceIDsFromD3D10KHR_fn)(
cl_platform_id platform,
cl_d3d10_device_source_khr d3d_device_source,
void * d3d_object,
cl_d3d10_device_set_khr d3d_device_set,
cl_uint num_entries,
cl_device_id * devices,
cl_uint * num_devices) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromD3D10BufferKHR_fn)(
cl_context context,
cl_mem_flags flags,
ID3D10Buffer * resource,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromD3D10Texture2DKHR_fn)(
cl_context context,
cl_mem_flags flags,
ID3D10Texture2D * resource,
UINT subresource,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromD3D10Texture3DKHR_fn)(
cl_context context,
cl_mem_flags flags,
ID3D10Texture3D * resource,
UINT subresource,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueAcquireD3D10ObjectsKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueReleaseD3D10ObjectsKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_0;
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_D3D10_H */

131
src/3rdparty/CL/cl_d3d11.h vendored Normal file
View file

@ -0,0 +1,131 @@
/**********************************************************************************
* Copyright (c) 2008-2015 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
/* $Revision: 11708 $ on $Date: 2010-06-13 23:36:24 -0700 (Sun, 13 Jun 2010) $ */
#ifndef __OPENCL_CL_D3D11_H
#define __OPENCL_CL_D3D11_H
#include <d3d11.h>
#include <CL/cl.h>
#include <CL/cl_platform.h>
#ifdef __cplusplus
extern "C" {
#endif
/******************************************************************************
* cl_khr_d3d11_sharing */
#define cl_khr_d3d11_sharing 1
typedef cl_uint cl_d3d11_device_source_khr;
typedef cl_uint cl_d3d11_device_set_khr;
/******************************************************************************/
/* Error Codes */
#define CL_INVALID_D3D11_DEVICE_KHR -1006
#define CL_INVALID_D3D11_RESOURCE_KHR -1007
#define CL_D3D11_RESOURCE_ALREADY_ACQUIRED_KHR -1008
#define CL_D3D11_RESOURCE_NOT_ACQUIRED_KHR -1009
/* cl_d3d11_device_source */
#define CL_D3D11_DEVICE_KHR 0x4019
#define CL_D3D11_DXGI_ADAPTER_KHR 0x401A
/* cl_d3d11_device_set */
#define CL_PREFERRED_DEVICES_FOR_D3D11_KHR 0x401B
#define CL_ALL_DEVICES_FOR_D3D11_KHR 0x401C
/* cl_context_info */
#define CL_CONTEXT_D3D11_DEVICE_KHR 0x401D
#define CL_CONTEXT_D3D11_PREFER_SHARED_RESOURCES_KHR 0x402D
/* cl_mem_info */
#define CL_MEM_D3D11_RESOURCE_KHR 0x401E
/* cl_image_info */
#define CL_IMAGE_D3D11_SUBRESOURCE_KHR 0x401F
/* cl_command_type */
#define CL_COMMAND_ACQUIRE_D3D11_OBJECTS_KHR 0x4020
#define CL_COMMAND_RELEASE_D3D11_OBJECTS_KHR 0x4021
/******************************************************************************/
typedef CL_API_ENTRY cl_int (CL_API_CALL *clGetDeviceIDsFromD3D11KHR_fn)(
cl_platform_id platform,
cl_d3d11_device_source_khr d3d_device_source,
void * d3d_object,
cl_d3d11_device_set_khr d3d_device_set,
cl_uint num_entries,
cl_device_id * devices,
cl_uint * num_devices) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromD3D11BufferKHR_fn)(
cl_context context,
cl_mem_flags flags,
ID3D11Buffer * resource,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromD3D11Texture2DKHR_fn)(
cl_context context,
cl_mem_flags flags,
ID3D11Texture2D * resource,
UINT subresource,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromD3D11Texture3DKHR_fn)(
cl_context context,
cl_mem_flags flags,
ID3D11Texture3D * resource,
UINT subresource,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueAcquireD3D11ObjectsKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueReleaseD3D11ObjectsKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_2;
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_D3D11_H */

132
src/3rdparty/CL/cl_dx9_media_sharing.h vendored Normal file
View file

@ -0,0 +1,132 @@
/**********************************************************************************
* Copyright (c) 2008-2015 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
/* $Revision: 11708 $ on $Date: 2010-06-13 23:36:24 -0700 (Sun, 13 Jun 2010) $ */
#ifndef __OPENCL_CL_DX9_MEDIA_SHARING_H
#define __OPENCL_CL_DX9_MEDIA_SHARING_H
#include <CL/cl.h>
#include <CL/cl_platform.h>
#ifdef __cplusplus
extern "C" {
#endif
/******************************************************************************/
/* cl_khr_dx9_media_sharing */
#define cl_khr_dx9_media_sharing 1
typedef cl_uint cl_dx9_media_adapter_type_khr;
typedef cl_uint cl_dx9_media_adapter_set_khr;
#if defined(_WIN32)
#include <d3d9.h>
typedef struct _cl_dx9_surface_info_khr
{
IDirect3DSurface9 *resource;
HANDLE shared_handle;
} cl_dx9_surface_info_khr;
#endif
/******************************************************************************/
/* Error Codes */
#define CL_INVALID_DX9_MEDIA_ADAPTER_KHR -1010
#define CL_INVALID_DX9_MEDIA_SURFACE_KHR -1011
#define CL_DX9_MEDIA_SURFACE_ALREADY_ACQUIRED_KHR -1012
#define CL_DX9_MEDIA_SURFACE_NOT_ACQUIRED_KHR -1013
/* cl_media_adapter_type_khr */
#define CL_ADAPTER_D3D9_KHR 0x2020
#define CL_ADAPTER_D3D9EX_KHR 0x2021
#define CL_ADAPTER_DXVA_KHR 0x2022
/* cl_media_adapter_set_khr */
#define CL_PREFERRED_DEVICES_FOR_DX9_MEDIA_ADAPTER_KHR 0x2023
#define CL_ALL_DEVICES_FOR_DX9_MEDIA_ADAPTER_KHR 0x2024
/* cl_context_info */
#define CL_CONTEXT_ADAPTER_D3D9_KHR 0x2025
#define CL_CONTEXT_ADAPTER_D3D9EX_KHR 0x2026
#define CL_CONTEXT_ADAPTER_DXVA_KHR 0x2027
/* cl_mem_info */
#define CL_MEM_DX9_MEDIA_ADAPTER_TYPE_KHR 0x2028
#define CL_MEM_DX9_MEDIA_SURFACE_INFO_KHR 0x2029
/* cl_image_info */
#define CL_IMAGE_DX9_MEDIA_PLANE_KHR 0x202A
/* cl_command_type */
#define CL_COMMAND_ACQUIRE_DX9_MEDIA_SURFACES_KHR 0x202B
#define CL_COMMAND_RELEASE_DX9_MEDIA_SURFACES_KHR 0x202C
/******************************************************************************/
typedef CL_API_ENTRY cl_int (CL_API_CALL *clGetDeviceIDsFromDX9MediaAdapterKHR_fn)(
cl_platform_id platform,
cl_uint num_media_adapters,
cl_dx9_media_adapter_type_khr * media_adapter_type,
void * media_adapters,
cl_dx9_media_adapter_set_khr media_adapter_set,
cl_uint num_entries,
cl_device_id * devices,
cl_uint * num_devices) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromDX9MediaSurfaceKHR_fn)(
cl_context context,
cl_mem_flags flags,
cl_dx9_media_adapter_type_khr adapter_type,
void * surface_info,
cl_uint plane,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueAcquireDX9MediaSurfacesKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueReleaseDX9MediaSurfacesKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_2;
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_DX9_MEDIA_SHARING_H */

182
src/3rdparty/CL/cl_dx9_media_sharing_intel.h vendored Normal file
View file

@ -0,0 +1,182 @@
/**********************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
/*****************************************************************************\
Copyright (c) 2013-2019 Intel Corporation All Rights Reserved.
THESE MATERIALS ARE PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THESE
MATERIALS, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
File Name: cl_dx9_media_sharing_intel.h
Abstract:
Notes:
\*****************************************************************************/
#ifndef __OPENCL_CL_DX9_MEDIA_SHARING_INTEL_H
#define __OPENCL_CL_DX9_MEDIA_SHARING_INTEL_H
#include <CL/cl.h>
#include <CL/cl_platform.h>
#include <d3d9.h>
#include <dxvahd.h>
#include <wtypes.h>
#include <d3d9types.h>
#ifdef __cplusplus
extern "C" {
#endif
/***************************************
* cl_intel_dx9_media_sharing extension *
****************************************/
#define cl_intel_dx9_media_sharing 1
typedef cl_uint cl_dx9_device_source_intel;
typedef cl_uint cl_dx9_device_set_intel;
/* error codes */
#define CL_INVALID_DX9_DEVICE_INTEL -1010
#define CL_INVALID_DX9_RESOURCE_INTEL -1011
#define CL_DX9_RESOURCE_ALREADY_ACQUIRED_INTEL -1012
#define CL_DX9_RESOURCE_NOT_ACQUIRED_INTEL -1013
/* cl_dx9_device_source_intel */
#define CL_D3D9_DEVICE_INTEL 0x4022
#define CL_D3D9EX_DEVICE_INTEL 0x4070
#define CL_DXVA_DEVICE_INTEL 0x4071
/* cl_dx9_device_set_intel */
#define CL_PREFERRED_DEVICES_FOR_DX9_INTEL 0x4024
#define CL_ALL_DEVICES_FOR_DX9_INTEL 0x4025
/* cl_context_info */
#define CL_CONTEXT_D3D9_DEVICE_INTEL 0x4026
#define CL_CONTEXT_D3D9EX_DEVICE_INTEL 0x4072
#define CL_CONTEXT_DXVA_DEVICE_INTEL 0x4073
/* cl_mem_info */
#define CL_MEM_DX9_RESOURCE_INTEL 0x4027
#define CL_MEM_DX9_SHARED_HANDLE_INTEL 0x4074
/* cl_image_info */
#define CL_IMAGE_DX9_PLANE_INTEL 0x4075
/* cl_command_type */
#define CL_COMMAND_ACQUIRE_DX9_OBJECTS_INTEL 0x402A
#define CL_COMMAND_RELEASE_DX9_OBJECTS_INTEL 0x402B
/******************************************************************************/
extern CL_API_ENTRY cl_int CL_API_CALL
clGetDeviceIDsFromDX9INTEL(
cl_platform_id platform,
cl_dx9_device_source_intel dx9_device_source,
void* dx9_object,
cl_dx9_device_set_intel dx9_device_set,
cl_uint num_entries,
cl_device_id* devices,
cl_uint* num_devices) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_int (CL_API_CALL* clGetDeviceIDsFromDX9INTEL_fn)(
cl_platform_id platform,
cl_dx9_device_source_intel dx9_device_source,
void* dx9_object,
cl_dx9_device_set_intel dx9_device_set,
cl_uint num_entries,
cl_device_id* devices,
cl_uint* num_devices) CL_EXT_SUFFIX__VERSION_1_1;
extern CL_API_ENTRY cl_mem CL_API_CALL
clCreateFromDX9MediaSurfaceINTEL(
cl_context context,
cl_mem_flags flags,
IDirect3DSurface9* resource,
HANDLE sharedHandle,
UINT plane,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromDX9MediaSurfaceINTEL_fn)(
cl_context context,
cl_mem_flags flags,
IDirect3DSurface9* resource,
HANDLE sharedHandle,
UINT plane,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_1;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueAcquireDX9ObjectsINTEL(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueAcquireDX9ObjectsINTEL_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_1;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueReleaseDX9ObjectsINTEL(
cl_command_queue command_queue,
cl_uint num_objects,
cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueReleaseDX9ObjectsINTEL_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_1;
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_DX9_MEDIA_SHARING_INTEL_H */

132
src/3rdparty/CL/cl_egl.h vendored Normal file
View file

@ -0,0 +1,132 @@
/*******************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
******************************************************************************/
#ifndef __OPENCL_CL_EGL_H
#define __OPENCL_CL_EGL_H
#include <CL/cl.h>
#ifdef __cplusplus
extern "C" {
#endif
/* Command type for events created with clEnqueueAcquireEGLObjectsKHR */
#define CL_COMMAND_EGL_FENCE_SYNC_OBJECT_KHR 0x202F
#define CL_COMMAND_ACQUIRE_EGL_OBJECTS_KHR 0x202D
#define CL_COMMAND_RELEASE_EGL_OBJECTS_KHR 0x202E
/* Error type for clCreateFromEGLImageKHR */
#define CL_INVALID_EGL_OBJECT_KHR -1093
#define CL_EGL_RESOURCE_NOT_ACQUIRED_KHR -1092
/* CLeglImageKHR is an opaque handle to an EGLImage */
typedef void* CLeglImageKHR;
/* CLeglDisplayKHR is an opaque handle to an EGLDisplay */
typedef void* CLeglDisplayKHR;
/* CLeglSyncKHR is an opaque handle to an EGLSync object */
typedef void* CLeglSyncKHR;
/* properties passed to clCreateFromEGLImageKHR */
typedef intptr_t cl_egl_image_properties_khr;
#define cl_khr_egl_image 1
extern CL_API_ENTRY cl_mem CL_API_CALL
clCreateFromEGLImageKHR(cl_context context,
CLeglDisplayKHR egldisplay,
CLeglImageKHR eglimage,
cl_mem_flags flags,
const cl_egl_image_properties_khr * properties,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_mem (CL_API_CALL *clCreateFromEGLImageKHR_fn)(
cl_context context,
CLeglDisplayKHR egldisplay,
CLeglImageKHR eglimage,
cl_mem_flags flags,
const cl_egl_image_properties_khr * properties,
cl_int * errcode_ret);
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueAcquireEGLObjectsKHR(cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueAcquireEGLObjectsKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event);
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueReleaseEGLObjectsKHR(cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueReleaseEGLObjectsKHR_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event);
#define cl_khr_egl_event 1
extern CL_API_ENTRY cl_event CL_API_CALL
clCreateEventFromEGLSyncKHR(cl_context context,
CLeglSyncKHR sync,
CLeglDisplayKHR display,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_event (CL_API_CALL *clCreateEventFromEGLSyncKHR_fn)(
cl_context context,
CLeglSyncKHR sync,
CLeglDisplayKHR display,
cl_int * errcode_ret);
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_EGL_H */

762
src/3rdparty/CL/cl_ext.h vendored Normal file
View file

@ -0,0 +1,762 @@
/*******************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
******************************************************************************/
/* cl_ext.h contains OpenCL extensions which don't have external */
/* (OpenGL, D3D) dependencies. */
#ifndef __CL_EXT_H
#define __CL_EXT_H
#ifdef __cplusplus
extern "C" {
#endif
#include <CL/cl.h>
/* cl_khr_fp64 extension - no extension #define since it has no functions */
/* CL_DEVICE_DOUBLE_FP_CONFIG is defined in CL.h for OpenCL >= 120 */
#if CL_TARGET_OPENCL_VERSION <= 110
#define CL_DEVICE_DOUBLE_FP_CONFIG 0x1032
#endif
/* cl_khr_fp16 extension - no extension #define since it has no functions */
#define CL_DEVICE_HALF_FP_CONFIG 0x1033
/* Memory object destruction
*
* Apple extension for use to manage externally allocated buffers used with cl_mem objects with CL_MEM_USE_HOST_PTR
*
* Registers a user callback function that will be called when the memory object is deleted and its resources
* freed. Each call to clSetMemObjectCallbackFn registers the specified user callback function on a callback
* stack associated with memobj. The registered user callback functions are called in the reverse order in
* which they were registered. The user callback functions are called and then the memory object is deleted
* and its resources freed. This provides a mechanism for the application (and libraries) using memobj to be
* notified when the memory referenced by host_ptr, specified when the memory object is created and used as
* the storage bits for the memory object, can be reused or freed.
*
* The application may not call CL api's with the cl_mem object passed to the pfn_notify.
*
* Please check for the "cl_APPLE_SetMemObjectDestructor" extension using clGetDeviceInfo(CL_DEVICE_EXTENSIONS)
* before using.
*/
#define cl_APPLE_SetMemObjectDestructor 1
cl_int CL_API_ENTRY clSetMemObjectDestructorAPPLE( cl_mem memobj,
void (* pfn_notify)(cl_mem memobj, void * user_data),
void * user_data) CL_EXT_SUFFIX__VERSION_1_0;
/* Context Logging Functions
*
* The next three convenience functions are intended to be used as the pfn_notify parameter to clCreateContext().
* Please check for the "cl_APPLE_ContextLoggingFunctions" extension using clGetDeviceInfo(CL_DEVICE_EXTENSIONS)
* before using.
*
* clLogMessagesToSystemLog forwards on all log messages to the Apple System Logger
*/
#define cl_APPLE_ContextLoggingFunctions 1
extern void CL_API_ENTRY clLogMessagesToSystemLogAPPLE( const char * errstr,
const void * private_info,
size_t cb,
void * user_data) CL_EXT_SUFFIX__VERSION_1_0;
/* clLogMessagesToStdout sends all log messages to the file descriptor stdout */
extern void CL_API_ENTRY clLogMessagesToStdoutAPPLE( const char * errstr,
const void * private_info,
size_t cb,
void * user_data) CL_EXT_SUFFIX__VERSION_1_0;
/* clLogMessagesToStderr sends all log messages to the file descriptor stderr */
extern void CL_API_ENTRY clLogMessagesToStderrAPPLE( const char * errstr,
const void * private_info,
size_t cb,
void * user_data) CL_EXT_SUFFIX__VERSION_1_0;
/************************
* cl_khr_icd extension *
************************/
#define cl_khr_icd 1
/* cl_platform_info */
#define CL_PLATFORM_ICD_SUFFIX_KHR 0x0920
/* Additional Error Codes */
#define CL_PLATFORM_NOT_FOUND_KHR -1001
extern CL_API_ENTRY cl_int CL_API_CALL
clIcdGetPlatformIDsKHR(cl_uint num_entries,
cl_platform_id * platforms,
cl_uint * num_platforms);
typedef CL_API_ENTRY cl_int
(CL_API_CALL *clIcdGetPlatformIDsKHR_fn)(cl_uint num_entries,
cl_platform_id * platforms,
cl_uint * num_platforms);
/*******************************
* cl_khr_il_program extension *
*******************************/
#define cl_khr_il_program 1
/* New property to clGetDeviceInfo for retrieving supported intermediate
* languages
*/
#define CL_DEVICE_IL_VERSION_KHR 0x105B
/* New property to clGetProgramInfo for retrieving for retrieving the IL of a
* program
*/
#define CL_PROGRAM_IL_KHR 0x1169
extern CL_API_ENTRY cl_program CL_API_CALL
clCreateProgramWithILKHR(cl_context context,
const void * il,
size_t length,
cl_int * errcode_ret);
typedef CL_API_ENTRY cl_program
(CL_API_CALL *clCreateProgramWithILKHR_fn)(cl_context context,
const void * il,
size_t length,
cl_int * errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
/* Extension: cl_khr_image2d_from_buffer
*
* This extension allows a 2D image to be created from a cl_mem buffer without
* a copy. The type associated with a 2D image created from a buffer in an
* OpenCL program is image2d_t. Both the sampler and sampler-less read_image
* built-in functions are supported for 2D images and 2D images created from
* a buffer. Similarly, the write_image built-ins are also supported for 2D
* images created from a buffer.
*
* When the 2D image from buffer is created, the client must specify the
* width, height, image format (i.e. channel order and channel data type)
* and optionally the row pitch.
*
* The pitch specified must be a multiple of
* CL_DEVICE_IMAGE_PITCH_ALIGNMENT_KHR pixels.
* The base address of the buffer must be aligned to
* CL_DEVICE_IMAGE_BASE_ADDRESS_ALIGNMENT_KHR pixels.
*/
#define CL_DEVICE_IMAGE_PITCH_ALIGNMENT_KHR 0x104A
#define CL_DEVICE_IMAGE_BASE_ADDRESS_ALIGNMENT_KHR 0x104B
/**************************************
* cl_khr_initialize_memory extension *
**************************************/
#define CL_CONTEXT_MEMORY_INITIALIZE_KHR 0x2030
/**************************************
* cl_khr_terminate_context extension *
**************************************/
#define CL_DEVICE_TERMINATE_CAPABILITY_KHR 0x2031
#define CL_CONTEXT_TERMINATE_KHR 0x2032
#define cl_khr_terminate_context 1
extern CL_API_ENTRY cl_int CL_API_CALL
clTerminateContextKHR(cl_context context) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int
(CL_API_CALL *clTerminateContextKHR_fn)(cl_context context) CL_EXT_SUFFIX__VERSION_1_2;
/*
* Extension: cl_khr_spir
*
* This extension adds support to create an OpenCL program object from a
* Standard Portable Intermediate Representation (SPIR) instance
*/
#define CL_DEVICE_SPIR_VERSIONS 0x40E0
#define CL_PROGRAM_BINARY_TYPE_INTERMEDIATE 0x40E1
/*****************************************
* cl_khr_create_command_queue extension *
*****************************************/
#define cl_khr_create_command_queue 1
typedef cl_bitfield cl_queue_properties_khr;
extern CL_API_ENTRY cl_command_queue CL_API_CALL
clCreateCommandQueueWithPropertiesKHR(cl_context context,
cl_device_id device,
const cl_queue_properties_khr* properties,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_command_queue
(CL_API_CALL *clCreateCommandQueueWithPropertiesKHR_fn)(cl_context context,
cl_device_id device,
const cl_queue_properties_khr* properties,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
/******************************************
* cl_nv_device_attribute_query extension *
******************************************/
/* cl_nv_device_attribute_query extension - no extension #define since it has no functions */
#define CL_DEVICE_COMPUTE_CAPABILITY_MAJOR_NV 0x4000
#define CL_DEVICE_COMPUTE_CAPABILITY_MINOR_NV 0x4001
#define CL_DEVICE_REGISTERS_PER_BLOCK_NV 0x4002
#define CL_DEVICE_WARP_SIZE_NV 0x4003
#define CL_DEVICE_GPU_OVERLAP_NV 0x4004
#define CL_DEVICE_KERNEL_EXEC_TIMEOUT_NV 0x4005
#define CL_DEVICE_INTEGRATED_MEMORY_NV 0x4006
/*********************************
* cl_amd_device_attribute_query *
*********************************/
#define CL_DEVICE_PROFILING_TIMER_OFFSET_AMD 0x4036
/*********************************
* cl_arm_printf extension
*********************************/
#define CL_PRINTF_CALLBACK_ARM 0x40B0
#define CL_PRINTF_BUFFERSIZE_ARM 0x40B1
/***********************************
* cl_ext_device_fission extension
***********************************/
#define cl_ext_device_fission 1
extern CL_API_ENTRY cl_int CL_API_CALL
clReleaseDeviceEXT(cl_device_id device) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_int
(CL_API_CALL *clReleaseDeviceEXT_fn)(cl_device_id device) CL_EXT_SUFFIX__VERSION_1_1;
extern CL_API_ENTRY cl_int CL_API_CALL
clRetainDeviceEXT(cl_device_id device) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_int
(CL_API_CALL *clRetainDeviceEXT_fn)(cl_device_id device) CL_EXT_SUFFIX__VERSION_1_1;
typedef cl_ulong cl_device_partition_property_ext;
extern CL_API_ENTRY cl_int CL_API_CALL
clCreateSubDevicesEXT(cl_device_id in_device,
const cl_device_partition_property_ext * properties,
cl_uint num_entries,
cl_device_id * out_devices,
cl_uint * num_devices) CL_EXT_SUFFIX__VERSION_1_1;
typedef CL_API_ENTRY cl_int
(CL_API_CALL * clCreateSubDevicesEXT_fn)(cl_device_id in_device,
const cl_device_partition_property_ext * properties,
cl_uint num_entries,
cl_device_id * out_devices,
cl_uint * num_devices) CL_EXT_SUFFIX__VERSION_1_1;
/* cl_device_partition_property_ext */
#define CL_DEVICE_PARTITION_EQUALLY_EXT 0x4050
#define CL_DEVICE_PARTITION_BY_COUNTS_EXT 0x4051
#define CL_DEVICE_PARTITION_BY_NAMES_EXT 0x4052
#define CL_DEVICE_PARTITION_BY_AFFINITY_DOMAIN_EXT 0x4053
/* clDeviceGetInfo selectors */
#define CL_DEVICE_PARENT_DEVICE_EXT 0x4054
#define CL_DEVICE_PARTITION_TYPES_EXT 0x4055
#define CL_DEVICE_AFFINITY_DOMAINS_EXT 0x4056
#define CL_DEVICE_REFERENCE_COUNT_EXT 0x4057
#define CL_DEVICE_PARTITION_STYLE_EXT 0x4058
/* error codes */
#define CL_DEVICE_PARTITION_FAILED_EXT -1057
#define CL_INVALID_PARTITION_COUNT_EXT -1058
#define CL_INVALID_PARTITION_NAME_EXT -1059
/* CL_AFFINITY_DOMAINs */
#define CL_AFFINITY_DOMAIN_L1_CACHE_EXT 0x1
#define CL_AFFINITY_DOMAIN_L2_CACHE_EXT 0x2
#define CL_AFFINITY_DOMAIN_L3_CACHE_EXT 0x3
#define CL_AFFINITY_DOMAIN_L4_CACHE_EXT 0x4
#define CL_AFFINITY_DOMAIN_NUMA_EXT 0x10
#define CL_AFFINITY_DOMAIN_NEXT_FISSIONABLE_EXT 0x100
/* cl_device_partition_property_ext list terminators */
#define CL_PROPERTIES_LIST_END_EXT ((cl_device_partition_property_ext) 0)
#define CL_PARTITION_BY_COUNTS_LIST_END_EXT ((cl_device_partition_property_ext) 0)
#define CL_PARTITION_BY_NAMES_LIST_END_EXT ((cl_device_partition_property_ext) 0 - 1)
/***********************************
* cl_ext_migrate_memobject extension definitions
***********************************/
#define cl_ext_migrate_memobject 1
typedef cl_bitfield cl_mem_migration_flags_ext;
#define CL_MIGRATE_MEM_OBJECT_HOST_EXT 0x1
#define CL_COMMAND_MIGRATE_MEM_OBJECT_EXT 0x4040
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueMigrateMemObjectEXT(cl_command_queue command_queue,
cl_uint num_mem_objects,
const cl_mem * mem_objects,
cl_mem_migration_flags_ext flags,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event);
typedef CL_API_ENTRY cl_int
(CL_API_CALL *clEnqueueMigrateMemObjectEXT_fn)(cl_command_queue command_queue,
cl_uint num_mem_objects,
const cl_mem * mem_objects,
cl_mem_migration_flags_ext flags,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event);
/*********************************
* cl_qcom_ext_host_ptr extension
*********************************/
#define cl_qcom_ext_host_ptr 1
#define CL_MEM_EXT_HOST_PTR_QCOM (1 << 29)
#define CL_DEVICE_EXT_MEM_PADDING_IN_BYTES_QCOM 0x40A0
#define CL_DEVICE_PAGE_SIZE_QCOM 0x40A1
#define CL_IMAGE_ROW_ALIGNMENT_QCOM 0x40A2
#define CL_IMAGE_SLICE_ALIGNMENT_QCOM 0x40A3
#define CL_MEM_HOST_UNCACHED_QCOM 0x40A4
#define CL_MEM_HOST_WRITEBACK_QCOM 0x40A5
#define CL_MEM_HOST_WRITETHROUGH_QCOM 0x40A6
#define CL_MEM_HOST_WRITE_COMBINING_QCOM 0x40A7
typedef cl_uint cl_image_pitch_info_qcom;
extern CL_API_ENTRY cl_int CL_API_CALL
clGetDeviceImageInfoQCOM(cl_device_id device,
size_t image_width,
size_t image_height,
const cl_image_format *image_format,
cl_image_pitch_info_qcom param_name,
size_t param_value_size,
void *param_value,
size_t *param_value_size_ret);
typedef struct _cl_mem_ext_host_ptr
{
/* Type of external memory allocation. */
/* Legal values will be defined in layered extensions. */
cl_uint allocation_type;
/* Host cache policy for this external memory allocation. */
cl_uint host_cache_policy;
} cl_mem_ext_host_ptr;
/*******************************************
* cl_qcom_ext_host_ptr_iocoherent extension
********************************************/
/* Cache policy specifying io-coherence */
#define CL_MEM_HOST_IOCOHERENT_QCOM 0x40A9
/*********************************
* cl_qcom_ion_host_ptr extension
*********************************/
#define CL_MEM_ION_HOST_PTR_QCOM 0x40A8
typedef struct _cl_mem_ion_host_ptr
{
/* Type of external memory allocation. */
/* Must be CL_MEM_ION_HOST_PTR_QCOM for ION allocations. */
cl_mem_ext_host_ptr ext_host_ptr;
/* ION file descriptor */
int ion_filedesc;
/* Host pointer to the ION allocated memory */
void* ion_hostptr;
} cl_mem_ion_host_ptr;
/*********************************
* cl_qcom_android_native_buffer_host_ptr extension
*********************************/
#define CL_MEM_ANDROID_NATIVE_BUFFER_HOST_PTR_QCOM 0x40C6
typedef struct _cl_mem_android_native_buffer_host_ptr
{
/* Type of external memory allocation. */
/* Must be CL_MEM_ANDROID_NATIVE_BUFFER_HOST_PTR_QCOM for Android native buffers. */
cl_mem_ext_host_ptr ext_host_ptr;
/* Virtual pointer to the android native buffer */
void* anb_ptr;
} cl_mem_android_native_buffer_host_ptr;
/******************************************
* cl_img_yuv_image extension *
******************************************/
/* Image formats used in clCreateImage */
#define CL_NV21_IMG 0x40D0
#define CL_YV12_IMG 0x40D1
/******************************************
* cl_img_cached_allocations extension *
******************************************/
/* Flag values used by clCreateBuffer */
#define CL_MEM_USE_UNCACHED_CPU_MEMORY_IMG (1 << 26)
#define CL_MEM_USE_CACHED_CPU_MEMORY_IMG (1 << 27)
/******************************************
* cl_img_use_gralloc_ptr extension *
******************************************/
#define cl_img_use_gralloc_ptr 1
/* Flag values used by clCreateBuffer */
#define CL_MEM_USE_GRALLOC_PTR_IMG (1 << 28)
/* To be used by clGetEventInfo: */
#define CL_COMMAND_ACQUIRE_GRALLOC_OBJECTS_IMG 0x40D2
#define CL_COMMAND_RELEASE_GRALLOC_OBJECTS_IMG 0x40D3
/* Error code from clEnqueueReleaseGrallocObjectsIMG */
#define CL_GRALLOC_RESOURCE_NOT_ACQUIRED_IMG 0x40D4
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueAcquireGrallocObjectsIMG(cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueReleaseGrallocObjectsIMG(cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
/*********************************
* cl_khr_subgroups extension
*********************************/
#define cl_khr_subgroups 1
#if !defined(CL_VERSION_2_1)
/* For OpenCL 2.1 and newer, cl_kernel_sub_group_info is declared in CL.h.
In hindsight, there should have been a khr suffix on this type for
the extension, but keeping it un-suffixed to maintain backwards
compatibility. */
typedef cl_uint cl_kernel_sub_group_info;
#endif
/* cl_kernel_sub_group_info */
#define CL_KERNEL_MAX_SUB_GROUP_SIZE_FOR_NDRANGE_KHR 0x2033
#define CL_KERNEL_SUB_GROUP_COUNT_FOR_NDRANGE_KHR 0x2034
extern CL_API_ENTRY cl_int CL_API_CALL
clGetKernelSubGroupInfoKHR(cl_kernel in_kernel,
cl_device_id in_device,
cl_kernel_sub_group_info param_name,
size_t input_value_size,
const void * input_value,
size_t param_value_size,
void * param_value,
size_t * param_value_size_ret) CL_EXT_SUFFIX__VERSION_2_0_DEPRECATED;
typedef CL_API_ENTRY cl_int
(CL_API_CALL * clGetKernelSubGroupInfoKHR_fn)(cl_kernel in_kernel,
cl_device_id in_device,
cl_kernel_sub_group_info param_name,
size_t input_value_size,
const void * input_value,
size_t param_value_size,
void * param_value,
size_t * param_value_size_ret) CL_EXT_SUFFIX__VERSION_2_0_DEPRECATED;
/*********************************
* cl_khr_mipmap_image extension
*********************************/
/* cl_sampler_properties */
#define CL_SAMPLER_MIP_FILTER_MODE_KHR 0x1155
#define CL_SAMPLER_LOD_MIN_KHR 0x1156
#define CL_SAMPLER_LOD_MAX_KHR 0x1157
/*********************************
* cl_khr_priority_hints extension
*********************************/
/* This extension define is for backwards compatibility.
It shouldn't be required since this extension has no new functions. */
#define cl_khr_priority_hints 1
typedef cl_uint cl_queue_priority_khr;
/* cl_command_queue_properties */
#define CL_QUEUE_PRIORITY_KHR 0x1096
/* cl_queue_priority_khr */
#define CL_QUEUE_PRIORITY_HIGH_KHR (1<<0)
#define CL_QUEUE_PRIORITY_MED_KHR (1<<1)
#define CL_QUEUE_PRIORITY_LOW_KHR (1<<2)
/*********************************
* cl_khr_throttle_hints extension
*********************************/
/* This extension define is for backwards compatibility.
It shouldn't be required since this extension has no new functions. */
#define cl_khr_throttle_hints 1
typedef cl_uint cl_queue_throttle_khr;
/* cl_command_queue_properties */
#define CL_QUEUE_THROTTLE_KHR 0x1097
/* cl_queue_throttle_khr */
#define CL_QUEUE_THROTTLE_HIGH_KHR (1<<0)
#define CL_QUEUE_THROTTLE_MED_KHR (1<<1)
#define CL_QUEUE_THROTTLE_LOW_KHR (1<<2)
/*********************************
* cl_khr_subgroup_named_barrier
*********************************/
/* This extension define is for backwards compatibility.
It shouldn't be required since this extension has no new functions. */
#define cl_khr_subgroup_named_barrier 1
/* cl_device_info */
#define CL_DEVICE_MAX_NAMED_BARRIER_COUNT_KHR 0x2035
/**********************************
* cl_arm_import_memory extension *
**********************************/
#define cl_arm_import_memory 1
typedef intptr_t cl_import_properties_arm;
/* Default and valid proporties name for cl_arm_import_memory */
#define CL_IMPORT_TYPE_ARM 0x40B2
/* Host process memory type default value for CL_IMPORT_TYPE_ARM property */
#define CL_IMPORT_TYPE_HOST_ARM 0x40B3
/* DMA BUF memory type value for CL_IMPORT_TYPE_ARM property */
#define CL_IMPORT_TYPE_DMA_BUF_ARM 0x40B4
/* Protected DMA BUF memory type value for CL_IMPORT_TYPE_ARM property */
#define CL_IMPORT_TYPE_PROTECTED_ARM 0x40B5
/* This extension adds a new function that allows for direct memory import into
* OpenCL via the clImportMemoryARM function.
*
* Memory imported through this interface will be mapped into the device's page
* tables directly, providing zero copy access. It will never fall back to copy
* operations and aliased buffers.
*
* Types of memory supported for import are specified as additional extension
* strings.
*
* This extension produces cl_mem allocations which are compatible with all other
* users of cl_mem in the standard API.
*
* This extension maps pages with the same properties as the normal buffer creation
* function clCreateBuffer.
*/
extern CL_API_ENTRY cl_mem CL_API_CALL
clImportMemoryARM( cl_context context,
cl_mem_flags flags,
const cl_import_properties_arm *properties,
void *memory,
size_t size,
cl_int *errcode_ret) CL_EXT_SUFFIX__VERSION_1_0;
/******************************************
* cl_arm_shared_virtual_memory extension *
******************************************/
#define cl_arm_shared_virtual_memory 1
/* Used by clGetDeviceInfo */
#define CL_DEVICE_SVM_CAPABILITIES_ARM 0x40B6
/* Used by clGetMemObjectInfo */
#define CL_MEM_USES_SVM_POINTER_ARM 0x40B7
/* Used by clSetKernelExecInfoARM: */
#define CL_KERNEL_EXEC_INFO_SVM_PTRS_ARM 0x40B8
#define CL_KERNEL_EXEC_INFO_SVM_FINE_GRAIN_SYSTEM_ARM 0x40B9
/* To be used by clGetEventInfo: */
#define CL_COMMAND_SVM_FREE_ARM 0x40BA
#define CL_COMMAND_SVM_MEMCPY_ARM 0x40BB
#define CL_COMMAND_SVM_MEMFILL_ARM 0x40BC
#define CL_COMMAND_SVM_MAP_ARM 0x40BD
#define CL_COMMAND_SVM_UNMAP_ARM 0x40BE
/* Flag values returned by clGetDeviceInfo with CL_DEVICE_SVM_CAPABILITIES_ARM as the param_name. */
#define CL_DEVICE_SVM_COARSE_GRAIN_BUFFER_ARM (1 << 0)
#define CL_DEVICE_SVM_FINE_GRAIN_BUFFER_ARM (1 << 1)
#define CL_DEVICE_SVM_FINE_GRAIN_SYSTEM_ARM (1 << 2)
#define CL_DEVICE_SVM_ATOMICS_ARM (1 << 3)
/* Flag values used by clSVMAllocARM: */
#define CL_MEM_SVM_FINE_GRAIN_BUFFER_ARM (1 << 10)
#define CL_MEM_SVM_ATOMICS_ARM (1 << 11)
typedef cl_bitfield cl_svm_mem_flags_arm;
typedef cl_uint cl_kernel_exec_info_arm;
typedef cl_bitfield cl_device_svm_capabilities_arm;
extern CL_API_ENTRY void * CL_API_CALL
clSVMAllocARM(cl_context context,
cl_svm_mem_flags_arm flags,
size_t size,
cl_uint alignment) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY void CL_API_CALL
clSVMFreeARM(cl_context context,
void * svm_pointer) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueSVMFreeARM(cl_command_queue command_queue,
cl_uint num_svm_pointers,
void * svm_pointers[],
void (CL_CALLBACK * pfn_free_func)(cl_command_queue queue,
cl_uint num_svm_pointers,
void * svm_pointers[],
void * user_data),
void * user_data,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueSVMMemcpyARM(cl_command_queue command_queue,
cl_bool blocking_copy,
void * dst_ptr,
const void * src_ptr,
size_t size,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueSVMMemFillARM(cl_command_queue command_queue,
void * svm_ptr,
const void * pattern,
size_t pattern_size,
size_t size,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueSVMMapARM(cl_command_queue command_queue,
cl_bool blocking_map,
cl_map_flags flags,
void * svm_ptr,
size_t size,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueSVMUnmapARM(cl_command_queue command_queue,
void * svm_ptr,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clSetKernelArgSVMPointerARM(cl_kernel kernel,
cl_uint arg_index,
const void * arg_value) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clSetKernelExecInfoARM(cl_kernel kernel,
cl_kernel_exec_info_arm param_name,
size_t param_value_size,
const void * param_value) CL_EXT_SUFFIX__VERSION_1_2;
/********************************
* cl_arm_get_core_id extension *
********************************/
#ifdef CL_VERSION_1_2
#define cl_arm_get_core_id 1
/* Device info property for bitfield of cores present */
#define CL_DEVICE_COMPUTE_UNITS_BITFIELD_ARM 0x40BF
#endif /* CL_VERSION_1_2 */
/*********************************
* cl_arm_job_slot_selection
*********************************/
#define cl_arm_job_slot_selection 1
/* cl_device_info */
#define CL_DEVICE_JOB_SLOTS_ARM 0x41E0
/* cl_command_queue_properties */
#define CL_QUEUE_JOB_SLOT_ARM 0x41E1
#ifdef __cplusplus
}
#endif
#endif /* __CL_EXT_H */

423
src/3rdparty/CL/cl_ext_intel.h vendored Normal file
View file

@ -0,0 +1,423 @@
/*******************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
******************************************************************************/
/*****************************************************************************\
Copyright (c) 2013-2019 Intel Corporation All Rights Reserved.
THESE MATERIALS ARE PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THESE
MATERIALS, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
File Name: cl_ext_intel.h
Abstract:
Notes:
\*****************************************************************************/
#ifndef __CL_EXT_INTEL_H
#define __CL_EXT_INTEL_H
#include <CL/cl.h>
#include <CL/cl_platform.h>
#ifdef __cplusplus
extern "C" {
#endif
/***************************************
* cl_intel_thread_local_exec extension *
****************************************/
#define cl_intel_thread_local_exec 1
#define CL_QUEUE_THREAD_LOCAL_EXEC_ENABLE_INTEL (((cl_bitfield)1) << 31)
/***********************************************
* cl_intel_device_partition_by_names extension *
************************************************/
#define cl_intel_device_partition_by_names 1
#define CL_DEVICE_PARTITION_BY_NAMES_INTEL 0x4052
#define CL_PARTITION_BY_NAMES_LIST_END_INTEL -1
/************************************************
* cl_intel_accelerator extension *
* cl_intel_motion_estimation extension *
* cl_intel_advanced_motion_estimation extension *
*************************************************/
#define cl_intel_accelerator 1
#define cl_intel_motion_estimation 1
#define cl_intel_advanced_motion_estimation 1
typedef struct _cl_accelerator_intel* cl_accelerator_intel;
typedef cl_uint cl_accelerator_type_intel;
typedef cl_uint cl_accelerator_info_intel;
typedef struct _cl_motion_estimation_desc_intel {
cl_uint mb_block_type;
cl_uint subpixel_mode;
cl_uint sad_adjust_mode;
cl_uint search_path_type;
} cl_motion_estimation_desc_intel;
/* error codes */
#define CL_INVALID_ACCELERATOR_INTEL -1094
#define CL_INVALID_ACCELERATOR_TYPE_INTEL -1095
#define CL_INVALID_ACCELERATOR_DESCRIPTOR_INTEL -1096
#define CL_ACCELERATOR_TYPE_NOT_SUPPORTED_INTEL -1097
/* cl_accelerator_type_intel */
#define CL_ACCELERATOR_TYPE_MOTION_ESTIMATION_INTEL 0x0
/* cl_accelerator_info_intel */
#define CL_ACCELERATOR_DESCRIPTOR_INTEL 0x4090
#define CL_ACCELERATOR_REFERENCE_COUNT_INTEL 0x4091
#define CL_ACCELERATOR_CONTEXT_INTEL 0x4092
#define CL_ACCELERATOR_TYPE_INTEL 0x4093
/* cl_motion_detect_desc_intel flags */
#define CL_ME_MB_TYPE_16x16_INTEL 0x0
#define CL_ME_MB_TYPE_8x8_INTEL 0x1
#define CL_ME_MB_TYPE_4x4_INTEL 0x2
#define CL_ME_SUBPIXEL_MODE_INTEGER_INTEL 0x0
#define CL_ME_SUBPIXEL_MODE_HPEL_INTEL 0x1
#define CL_ME_SUBPIXEL_MODE_QPEL_INTEL 0x2
#define CL_ME_SAD_ADJUST_MODE_NONE_INTEL 0x0
#define CL_ME_SAD_ADJUST_MODE_HAAR_INTEL 0x1
#define CL_ME_SEARCH_PATH_RADIUS_2_2_INTEL 0x0
#define CL_ME_SEARCH_PATH_RADIUS_4_4_INTEL 0x1
#define CL_ME_SEARCH_PATH_RADIUS_16_12_INTEL 0x5
#define CL_ME_SKIP_BLOCK_TYPE_16x16_INTEL 0x0
#define CL_ME_CHROMA_INTRA_PREDICT_ENABLED_INTEL 0x1
#define CL_ME_LUMA_INTRA_PREDICT_ENABLED_INTEL 0x2
#define CL_ME_SKIP_BLOCK_TYPE_8x8_INTEL 0x4
#define CL_ME_FORWARD_INPUT_MODE_INTEL 0x1
#define CL_ME_BACKWARD_INPUT_MODE_INTEL 0x2
#define CL_ME_BIDIRECTION_INPUT_MODE_INTEL 0x3
#define CL_ME_BIDIR_WEIGHT_QUARTER_INTEL 16
#define CL_ME_BIDIR_WEIGHT_THIRD_INTEL 21
#define CL_ME_BIDIR_WEIGHT_HALF_INTEL 32
#define CL_ME_BIDIR_WEIGHT_TWO_THIRD_INTEL 43
#define CL_ME_BIDIR_WEIGHT_THREE_QUARTER_INTEL 48
#define CL_ME_COST_PENALTY_NONE_INTEL 0x0
#define CL_ME_COST_PENALTY_LOW_INTEL 0x1
#define CL_ME_COST_PENALTY_NORMAL_INTEL 0x2
#define CL_ME_COST_PENALTY_HIGH_INTEL 0x3
#define CL_ME_COST_PRECISION_QPEL_INTEL 0x0
#define CL_ME_COST_PRECISION_HPEL_INTEL 0x1
#define CL_ME_COST_PRECISION_PEL_INTEL 0x2
#define CL_ME_COST_PRECISION_DPEL_INTEL 0x3
#define CL_ME_LUMA_PREDICTOR_MODE_VERTICAL_INTEL 0x0
#define CL_ME_LUMA_PREDICTOR_MODE_HORIZONTAL_INTEL 0x1
#define CL_ME_LUMA_PREDICTOR_MODE_DC_INTEL 0x2
#define CL_ME_LUMA_PREDICTOR_MODE_DIAGONAL_DOWN_LEFT_INTEL 0x3
#define CL_ME_LUMA_PREDICTOR_MODE_DIAGONAL_DOWN_RIGHT_INTEL 0x4
#define CL_ME_LUMA_PREDICTOR_MODE_PLANE_INTEL 0x4
#define CL_ME_LUMA_PREDICTOR_MODE_VERTICAL_RIGHT_INTEL 0x5
#define CL_ME_LUMA_PREDICTOR_MODE_HORIZONTAL_DOWN_INTEL 0x6
#define CL_ME_LUMA_PREDICTOR_MODE_VERTICAL_LEFT_INTEL 0x7
#define CL_ME_LUMA_PREDICTOR_MODE_HORIZONTAL_UP_INTEL 0x8
#define CL_ME_CHROMA_PREDICTOR_MODE_DC_INTEL 0x0
#define CL_ME_CHROMA_PREDICTOR_MODE_HORIZONTAL_INTEL 0x1
#define CL_ME_CHROMA_PREDICTOR_MODE_VERTICAL_INTEL 0x2
#define CL_ME_CHROMA_PREDICTOR_MODE_PLANE_INTEL 0x3
/* cl_device_info */
#define CL_DEVICE_ME_VERSION_INTEL 0x407E
#define CL_ME_VERSION_LEGACY_INTEL 0x0
#define CL_ME_VERSION_ADVANCED_VER_1_INTEL 0x1
#define CL_ME_VERSION_ADVANCED_VER_2_INTEL 0x2
extern CL_API_ENTRY cl_accelerator_intel CL_API_CALL
clCreateAcceleratorINTEL(
cl_context context,
cl_accelerator_type_intel accelerator_type,
size_t descriptor_size,
const void* descriptor,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_accelerator_intel (CL_API_CALL *clCreateAcceleratorINTEL_fn)(
cl_context context,
cl_accelerator_type_intel accelerator_type,
size_t descriptor_size,
const void* descriptor,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clGetAcceleratorInfoINTEL(
cl_accelerator_intel accelerator,
cl_accelerator_info_intel param_name,
size_t param_value_size,
void* param_value,
size_t* param_value_size_ret) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clGetAcceleratorInfoINTEL_fn)(
cl_accelerator_intel accelerator,
cl_accelerator_info_intel param_name,
size_t param_value_size,
void* param_value,
size_t* param_value_size_ret) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clRetainAcceleratorINTEL(
cl_accelerator_intel accelerator) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clRetainAcceleratorINTEL_fn)(
cl_accelerator_intel accelerator) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clReleaseAcceleratorINTEL(
cl_accelerator_intel accelerator) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clReleaseAcceleratorINTEL_fn)(
cl_accelerator_intel accelerator) CL_EXT_SUFFIX__VERSION_1_2;
/******************************************
* cl_intel_simultaneous_sharing extension *
*******************************************/
#define cl_intel_simultaneous_sharing 1
#define CL_DEVICE_SIMULTANEOUS_INTEROPS_INTEL 0x4104
#define CL_DEVICE_NUM_SIMULTANEOUS_INTEROPS_INTEL 0x4105
/***********************************
* cl_intel_egl_image_yuv extension *
************************************/
#define cl_intel_egl_image_yuv 1
#define CL_EGL_YUV_PLANE_INTEL 0x4107
/********************************
* cl_intel_packed_yuv extension *
*********************************/
#define cl_intel_packed_yuv 1
#define CL_YUYV_INTEL 0x4076
#define CL_UYVY_INTEL 0x4077
#define CL_YVYU_INTEL 0x4078
#define CL_VYUY_INTEL 0x4079
/********************************************
* cl_intel_required_subgroup_size extension *
*********************************************/
#define cl_intel_required_subgroup_size 1
#define CL_DEVICE_SUB_GROUP_SIZES_INTEL 0x4108
#define CL_KERNEL_SPILL_MEM_SIZE_INTEL 0x4109
#define CL_KERNEL_COMPILE_SUB_GROUP_SIZE_INTEL 0x410A
/****************************************
* cl_intel_driver_diagnostics extension *
*****************************************/
#define cl_intel_driver_diagnostics 1
typedef cl_uint cl_diagnostics_verbose_level;
#define CL_CONTEXT_SHOW_DIAGNOSTICS_INTEL 0x4106
#define CL_CONTEXT_DIAGNOSTICS_LEVEL_ALL_INTEL ( 0xff )
#define CL_CONTEXT_DIAGNOSTICS_LEVEL_GOOD_INTEL ( 1 )
#define CL_CONTEXT_DIAGNOSTICS_LEVEL_BAD_INTEL ( 1 << 1 )
#define CL_CONTEXT_DIAGNOSTICS_LEVEL_NEUTRAL_INTEL ( 1 << 2 )
/********************************
* cl_intel_planar_yuv extension *
*********************************/
#define CL_NV12_INTEL 0x410E
#define CL_MEM_NO_ACCESS_INTEL ( 1 << 24 )
#define CL_MEM_ACCESS_FLAGS_UNRESTRICTED_INTEL ( 1 << 25 )
#define CL_DEVICE_PLANAR_YUV_MAX_WIDTH_INTEL 0x417E
#define CL_DEVICE_PLANAR_YUV_MAX_HEIGHT_INTEL 0x417F
/*******************************************************
* cl_intel_device_side_avc_motion_estimation extension *
********************************************************/
#define CL_DEVICE_AVC_ME_VERSION_INTEL 0x410B
#define CL_DEVICE_AVC_ME_SUPPORTS_TEXTURE_SAMPLER_USE_INTEL 0x410C
#define CL_DEVICE_AVC_ME_SUPPORTS_PREEMPTION_INTEL 0x410D
#define CL_AVC_ME_VERSION_0_INTEL 0x0; // No support.
#define CL_AVC_ME_VERSION_1_INTEL 0x1; // First supported version.
#define CL_AVC_ME_MAJOR_16x16_INTEL 0x0
#define CL_AVC_ME_MAJOR_16x8_INTEL 0x1
#define CL_AVC_ME_MAJOR_8x16_INTEL 0x2
#define CL_AVC_ME_MAJOR_8x8_INTEL 0x3
#define CL_AVC_ME_MINOR_8x8_INTEL 0x0
#define CL_AVC_ME_MINOR_8x4_INTEL 0x1
#define CL_AVC_ME_MINOR_4x8_INTEL 0x2
#define CL_AVC_ME_MINOR_4x4_INTEL 0x3
#define CL_AVC_ME_MAJOR_FORWARD_INTEL 0x0
#define CL_AVC_ME_MAJOR_BACKWARD_INTEL 0x1
#define CL_AVC_ME_MAJOR_BIDIRECTIONAL_INTEL 0x2
#define CL_AVC_ME_PARTITION_MASK_ALL_INTEL 0x0
#define CL_AVC_ME_PARTITION_MASK_16x16_INTEL 0x7E
#define CL_AVC_ME_PARTITION_MASK_16x8_INTEL 0x7D
#define CL_AVC_ME_PARTITION_MASK_8x16_INTEL 0x7B
#define CL_AVC_ME_PARTITION_MASK_8x8_INTEL 0x77
#define CL_AVC_ME_PARTITION_MASK_8x4_INTEL 0x6F
#define CL_AVC_ME_PARTITION_MASK_4x8_INTEL 0x5F
#define CL_AVC_ME_PARTITION_MASK_4x4_INTEL 0x3F
#define CL_AVC_ME_SEARCH_WINDOW_EXHAUSTIVE_INTEL 0x0
#define CL_AVC_ME_SEARCH_WINDOW_SMALL_INTEL 0x1
#define CL_AVC_ME_SEARCH_WINDOW_TINY_INTEL 0x2
#define CL_AVC_ME_SEARCH_WINDOW_EXTRA_TINY_INTEL 0x3
#define CL_AVC_ME_SEARCH_WINDOW_DIAMOND_INTEL 0x4
#define CL_AVC_ME_SEARCH_WINDOW_LARGE_DIAMOND_INTEL 0x5
#define CL_AVC_ME_SEARCH_WINDOW_RESERVED0_INTEL 0x6
#define CL_AVC_ME_SEARCH_WINDOW_RESERVED1_INTEL 0x7
#define CL_AVC_ME_SEARCH_WINDOW_CUSTOM_INTEL 0x8
#define CL_AVC_ME_SEARCH_WINDOW_16x12_RADIUS_INTEL 0x9
#define CL_AVC_ME_SEARCH_WINDOW_4x4_RADIUS_INTEL 0x2
#define CL_AVC_ME_SEARCH_WINDOW_2x2_RADIUS_INTEL 0xa
#define CL_AVC_ME_SAD_ADJUST_MODE_NONE_INTEL 0x0
#define CL_AVC_ME_SAD_ADJUST_MODE_HAAR_INTEL 0x2
#define CL_AVC_ME_SUBPIXEL_MODE_INTEGER_INTEL 0x0
#define CL_AVC_ME_SUBPIXEL_MODE_HPEL_INTEL 0x1
#define CL_AVC_ME_SUBPIXEL_MODE_QPEL_INTEL 0x3
#define CL_AVC_ME_COST_PRECISION_QPEL_INTEL 0x0
#define CL_AVC_ME_COST_PRECISION_HPEL_INTEL 0x1
#define CL_AVC_ME_COST_PRECISION_PEL_INTEL 0x2
#define CL_AVC_ME_COST_PRECISION_DPEL_INTEL 0x3
#define CL_AVC_ME_BIDIR_WEIGHT_QUARTER_INTEL 0x10
#define CL_AVC_ME_BIDIR_WEIGHT_THIRD_INTEL 0x15
#define CL_AVC_ME_BIDIR_WEIGHT_HALF_INTEL 0x20
#define CL_AVC_ME_BIDIR_WEIGHT_TWO_THIRD_INTEL 0x2B
#define CL_AVC_ME_BIDIR_WEIGHT_THREE_QUARTER_INTEL 0x30
#define CL_AVC_ME_BORDER_REACHED_LEFT_INTEL 0x0
#define CL_AVC_ME_BORDER_REACHED_RIGHT_INTEL 0x2
#define CL_AVC_ME_BORDER_REACHED_TOP_INTEL 0x4
#define CL_AVC_ME_BORDER_REACHED_BOTTOM_INTEL 0x8
#define CL_AVC_ME_SKIP_BLOCK_PARTITION_16x16_INTEL 0x0
#define CL_AVC_ME_SKIP_BLOCK_PARTITION_8x8_INTEL 0x4000
#define CL_AVC_ME_SKIP_BLOCK_16x16_FORWARD_ENABLE_INTEL ( 0x1 << 24 )
#define CL_AVC_ME_SKIP_BLOCK_16x16_BACKWARD_ENABLE_INTEL ( 0x2 << 24 )
#define CL_AVC_ME_SKIP_BLOCK_16x16_DUAL_ENABLE_INTEL ( 0x3 << 24 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_FORWARD_ENABLE_INTEL ( 0x55 << 24 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_BACKWARD_ENABLE_INTEL ( 0xAA << 24 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_DUAL_ENABLE_INTEL ( 0xFF << 24 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_0_FORWARD_ENABLE_INTEL ( 0x1 << 24 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_0_BACKWARD_ENABLE_INTEL ( 0x2 << 24 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_1_FORWARD_ENABLE_INTEL ( 0x1 << 26 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_1_BACKWARD_ENABLE_INTEL ( 0x2 << 26 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_2_FORWARD_ENABLE_INTEL ( 0x1 << 28 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_2_BACKWARD_ENABLE_INTEL ( 0x2 << 28 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_3_FORWARD_ENABLE_INTEL ( 0x1 << 30 )
#define CL_AVC_ME_SKIP_BLOCK_8x8_3_BACKWARD_ENABLE_INTEL ( 0x2 << 30 )
#define CL_AVC_ME_BLOCK_BASED_SKIP_4x4_INTEL 0x00
#define CL_AVC_ME_BLOCK_BASED_SKIP_8x8_INTEL 0x80
#define CL_AVC_ME_INTRA_16x16_INTEL 0x0
#define CL_AVC_ME_INTRA_8x8_INTEL 0x1
#define CL_AVC_ME_INTRA_4x4_INTEL 0x2
#define CL_AVC_ME_INTRA_LUMA_PARTITION_MASK_16x16_INTEL 0x6
#define CL_AVC_ME_INTRA_LUMA_PARTITION_MASK_8x8_INTEL 0x5
#define CL_AVC_ME_INTRA_LUMA_PARTITION_MASK_4x4_INTEL 0x3
#define CL_AVC_ME_INTRA_NEIGHBOR_LEFT_MASK_ENABLE_INTEL 0x60
#define CL_AVC_ME_INTRA_NEIGHBOR_UPPER_MASK_ENABLE_INTEL 0x10
#define CL_AVC_ME_INTRA_NEIGHBOR_UPPER_RIGHT_MASK_ENABLE_INTEL 0x8
#define CL_AVC_ME_INTRA_NEIGHBOR_UPPER_LEFT_MASK_ENABLE_INTEL 0x4
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_VERTICAL_INTEL 0x0
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_HORIZONTAL_INTEL 0x1
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_DC_INTEL 0x2
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_DIAGONAL_DOWN_LEFT_INTEL 0x3
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_DIAGONAL_DOWN_RIGHT_INTEL 0x4
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_PLANE_INTEL 0x4
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_VERTICAL_RIGHT_INTEL 0x5
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_HORIZONTAL_DOWN_INTEL 0x6
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_VERTICAL_LEFT_INTEL 0x7
#define CL_AVC_ME_LUMA_PREDICTOR_MODE_HORIZONTAL_UP_INTEL 0x8
#define CL_AVC_ME_CHROMA_PREDICTOR_MODE_DC_INTEL 0x0
#define CL_AVC_ME_CHROMA_PREDICTOR_MODE_HORIZONTAL_INTEL 0x1
#define CL_AVC_ME_CHROMA_PREDICTOR_MODE_VERTICAL_INTEL 0x2
#define CL_AVC_ME_CHROMA_PREDICTOR_MODE_PLANE_INTEL 0x3
#define CL_AVC_ME_FRAME_FORWARD_INTEL 0x1
#define CL_AVC_ME_FRAME_BACKWARD_INTEL 0x2
#define CL_AVC_ME_FRAME_DUAL_INTEL 0x3
#define CL_AVC_ME_SLICE_TYPE_PRED_INTEL 0x0
#define CL_AVC_ME_SLICE_TYPE_BPRED_INTEL 0x1
#define CL_AVC_ME_SLICE_TYPE_INTRA_INTEL 0x2
#define CL_AVC_ME_INTERLACED_SCAN_TOP_FIELD_INTEL 0x0
#define CL_AVC_ME_INTERLACED_SCAN_BOTTOM_FIELD_INTEL 0x1
#ifdef __cplusplus
}
#endif
#endif /* __CL_EXT_INTEL_H */

171
src/3rdparty/CL/cl_gl.h vendored Normal file
View file

@ -0,0 +1,171 @@
/**********************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
#ifndef __OPENCL_CL_GL_H
#define __OPENCL_CL_GL_H
#include <CL/cl.h>
#ifdef __cplusplus
extern "C" {
#endif
typedef cl_uint cl_gl_object_type;
typedef cl_uint cl_gl_texture_info;
typedef cl_uint cl_gl_platform_info;
typedef struct __GLsync *cl_GLsync;
/* cl_gl_object_type = 0x2000 - 0x200F enum values are currently taken */
#define CL_GL_OBJECT_BUFFER 0x2000
#define CL_GL_OBJECT_TEXTURE2D 0x2001
#define CL_GL_OBJECT_TEXTURE3D 0x2002
#define CL_GL_OBJECT_RENDERBUFFER 0x2003
#ifdef CL_VERSION_1_2
#define CL_GL_OBJECT_TEXTURE2D_ARRAY 0x200E
#define CL_GL_OBJECT_TEXTURE1D 0x200F
#define CL_GL_OBJECT_TEXTURE1D_ARRAY 0x2010
#define CL_GL_OBJECT_TEXTURE_BUFFER 0x2011
#endif
/* cl_gl_texture_info */
#define CL_GL_TEXTURE_TARGET 0x2004
#define CL_GL_MIPMAP_LEVEL 0x2005
#ifdef CL_VERSION_1_2
#define CL_GL_NUM_SAMPLES 0x2012
#endif
extern CL_API_ENTRY cl_mem CL_API_CALL
clCreateFromGLBuffer(cl_context context,
cl_mem_flags flags,
cl_GLuint bufobj,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
#ifdef CL_VERSION_1_2
extern CL_API_ENTRY cl_mem CL_API_CALL
clCreateFromGLTexture(cl_context context,
cl_mem_flags flags,
cl_GLenum target,
cl_GLint miplevel,
cl_GLuint texture,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_2;
#endif
extern CL_API_ENTRY cl_mem CL_API_CALL
clCreateFromGLRenderbuffer(cl_context context,
cl_mem_flags flags,
cl_GLuint renderbuffer,
cl_int * errcode_ret) CL_API_SUFFIX__VERSION_1_0;
extern CL_API_ENTRY cl_int CL_API_CALL
clGetGLObjectInfo(cl_mem memobj,
cl_gl_object_type * gl_object_type,
cl_GLuint * gl_object_name) CL_API_SUFFIX__VERSION_1_0;
extern CL_API_ENTRY cl_int CL_API_CALL
clGetGLTextureInfo(cl_mem memobj,
cl_gl_texture_info param_name,
size_t param_value_size,
void * param_value,
size_t * param_value_size_ret) CL_API_SUFFIX__VERSION_1_0;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueAcquireGLObjects(cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_0;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueReleaseGLObjects(cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem * mem_objects,
cl_uint num_events_in_wait_list,
const cl_event * event_wait_list,
cl_event * event) CL_API_SUFFIX__VERSION_1_0;
/* Deprecated OpenCL 1.1 APIs */
extern CL_API_ENTRY CL_EXT_PREFIX__VERSION_1_1_DEPRECATED cl_mem CL_API_CALL
clCreateFromGLTexture2D(cl_context context,
cl_mem_flags flags,
cl_GLenum target,
cl_GLint miplevel,
cl_GLuint texture,
cl_int * errcode_ret) CL_EXT_SUFFIX__VERSION_1_1_DEPRECATED;
extern CL_API_ENTRY CL_EXT_PREFIX__VERSION_1_1_DEPRECATED cl_mem CL_API_CALL
clCreateFromGLTexture3D(cl_context context,
cl_mem_flags flags,
cl_GLenum target,
cl_GLint miplevel,
cl_GLuint texture,
cl_int * errcode_ret) CL_EXT_SUFFIX__VERSION_1_1_DEPRECATED;
/* cl_khr_gl_sharing extension */
#define cl_khr_gl_sharing 1
typedef cl_uint cl_gl_context_info;
/* Additional Error Codes */
#define CL_INVALID_GL_SHAREGROUP_REFERENCE_KHR -1000
/* cl_gl_context_info */
#define CL_CURRENT_DEVICE_FOR_GL_CONTEXT_KHR 0x2006
#define CL_DEVICES_FOR_GL_CONTEXT_KHR 0x2007
/* Additional cl_context_properties */
#define CL_GL_CONTEXT_KHR 0x2008
#define CL_EGL_DISPLAY_KHR 0x2009
#define CL_GLX_DISPLAY_KHR 0x200A
#define CL_WGL_HDC_KHR 0x200B
#define CL_CGL_SHAREGROUP_KHR 0x200C
extern CL_API_ENTRY cl_int CL_API_CALL
clGetGLContextInfoKHR(const cl_context_properties * properties,
cl_gl_context_info param_name,
size_t param_value_size,
void * param_value,
size_t * param_value_size_ret) CL_API_SUFFIX__VERSION_1_0;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clGetGLContextInfoKHR_fn)(
const cl_context_properties * properties,
cl_gl_context_info param_name,
size_t param_value_size,
void * param_value,
size_t * param_value_size_ret);
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_GL_H */

52
src/3rdparty/CL/cl_gl_ext.h vendored Normal file
View file

@ -0,0 +1,52 @@
/**********************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
#ifndef __OPENCL_CL_GL_EXT_H
#define __OPENCL_CL_GL_EXT_H
#ifdef __cplusplus
extern "C" {
#endif
#include <CL/cl_gl.h>
/*
* cl_khr_gl_event extension
*/
#define CL_COMMAND_GL_FENCE_SYNC_OBJECT_KHR 0x200D
extern CL_API_ENTRY cl_event CL_API_CALL
clCreateEventFromGLsyncKHR(cl_context context,
cl_GLsync cl_GLsync,
cl_int * errcode_ret) CL_EXT_SUFFIX__VERSION_1_1;
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_GL_EXT_H */

1384
src/3rdparty/CL/cl_platform.h vendored Normal file
View file

File diff suppressed because it is too large Load diff

172
src/3rdparty/CL/cl_va_api_media_sharing_intel.h vendored Normal file
View file

@ -0,0 +1,172 @@
/**********************************************************************************
* Copyright (c) 2008-2019 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
**********************************************************************************/
/*****************************************************************************\
Copyright (c) 2013-2019 Intel Corporation All Rights Reserved.
THESE MATERIALS ARE PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THESE
MATERIALS, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
File Name: cl_va_api_media_sharing_intel.h
Abstract:
Notes:
\*****************************************************************************/
#ifndef __OPENCL_CL_VA_API_MEDIA_SHARING_INTEL_H
#define __OPENCL_CL_VA_API_MEDIA_SHARING_INTEL_H
#include <CL/cl.h>
#include <CL/cl_platform.h>
#include <va/va.h>
#ifdef __cplusplus
extern "C" {
#endif
/******************************************
* cl_intel_va_api_media_sharing extension *
*******************************************/
#define cl_intel_va_api_media_sharing 1
/* error codes */
#define CL_INVALID_VA_API_MEDIA_ADAPTER_INTEL -1098
#define CL_INVALID_VA_API_MEDIA_SURFACE_INTEL -1099
#define CL_VA_API_MEDIA_SURFACE_ALREADY_ACQUIRED_INTEL -1100
#define CL_VA_API_MEDIA_SURFACE_NOT_ACQUIRED_INTEL -1101
/* cl_va_api_device_source_intel */
#define CL_VA_API_DISPLAY_INTEL 0x4094
/* cl_va_api_device_set_intel */
#define CL_PREFERRED_DEVICES_FOR_VA_API_INTEL 0x4095
#define CL_ALL_DEVICES_FOR_VA_API_INTEL 0x4096
/* cl_context_info */
#define CL_CONTEXT_VA_API_DISPLAY_INTEL 0x4097
/* cl_mem_info */
#define CL_MEM_VA_API_MEDIA_SURFACE_INTEL 0x4098
/* cl_image_info */
#define CL_IMAGE_VA_API_PLANE_INTEL 0x4099
/* cl_command_type */
#define CL_COMMAND_ACQUIRE_VA_API_MEDIA_SURFACES_INTEL 0x409A
#define CL_COMMAND_RELEASE_VA_API_MEDIA_SURFACES_INTEL 0x409B
typedef cl_uint cl_va_api_device_source_intel;
typedef cl_uint cl_va_api_device_set_intel;
extern CL_API_ENTRY cl_int CL_API_CALL
clGetDeviceIDsFromVA_APIMediaAdapterINTEL(
cl_platform_id platform,
cl_va_api_device_source_intel media_adapter_type,
void* media_adapter,
cl_va_api_device_set_intel media_adapter_set,
cl_uint num_entries,
cl_device_id* devices,
cl_uint* num_devices) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL * clGetDeviceIDsFromVA_APIMediaAdapterINTEL_fn)(
cl_platform_id platform,
cl_va_api_device_source_intel media_adapter_type,
void* media_adapter,
cl_va_api_device_set_intel media_adapter_set,
cl_uint num_entries,
cl_device_id* devices,
cl_uint* num_devices) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_mem CL_API_CALL
clCreateFromVA_APIMediaSurfaceINTEL(
cl_context context,
cl_mem_flags flags,
VASurfaceID* surface,
cl_uint plane,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_mem (CL_API_CALL * clCreateFromVA_APIMediaSurfaceINTEL_fn)(
cl_context context,
cl_mem_flags flags,
VASurfaceID* surface,
cl_uint plane,
cl_int* errcode_ret) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueAcquireVA_APIMediaSurfacesINTEL(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueAcquireVA_APIMediaSurfacesINTEL_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_2;
extern CL_API_ENTRY cl_int CL_API_CALL
clEnqueueReleaseVA_APIMediaSurfacesINTEL(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_2;
typedef CL_API_ENTRY cl_int (CL_API_CALL *clEnqueueReleaseVA_APIMediaSurfacesINTEL_fn)(
cl_command_queue command_queue,
cl_uint num_objects,
const cl_mem* mem_objects,
cl_uint num_events_in_wait_list,
const cl_event* event_wait_list,
cl_event* event) CL_EXT_SUFFIX__VERSION_1_2;
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_CL_VA_API_MEDIA_SHARING_INTEL_H */

86
src/3rdparty/CL/cl_version.h vendored Normal file
View file

@ -0,0 +1,86 @@
/*******************************************************************************
* Copyright (c) 2018 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
******************************************************************************/
#ifndef __CL_VERSION_H
#define __CL_VERSION_H
/* Detect which version to target */
#if !defined(CL_TARGET_OPENCL_VERSION)
#pragma message("cl_version.h: CL_TARGET_OPENCL_VERSION is not defined. Defaulting to 220 (OpenCL 2.2)")
#define CL_TARGET_OPENCL_VERSION 220
#endif
#if CL_TARGET_OPENCL_VERSION != 100 && \
CL_TARGET_OPENCL_VERSION != 110 && \
CL_TARGET_OPENCL_VERSION != 120 && \
CL_TARGET_OPENCL_VERSION != 200 && \
CL_TARGET_OPENCL_VERSION != 210 && \
CL_TARGET_OPENCL_VERSION != 220
#pragma message("cl_version: CL_TARGET_OPENCL_VERSION is not a valid value (100, 110, 120, 200, 210, 220). Defaulting to 220 (OpenCL 2.2)")
#undef CL_TARGET_OPENCL_VERSION
#define CL_TARGET_OPENCL_VERSION 220
#endif
/* OpenCL Version */
#if CL_TARGET_OPENCL_VERSION >= 220 && !defined(CL_VERSION_2_2)
#define CL_VERSION_2_2 1
#endif
#if CL_TARGET_OPENCL_VERSION >= 210 && !defined(CL_VERSION_2_1)
#define CL_VERSION_2_1 1
#endif
#if CL_TARGET_OPENCL_VERSION >= 200 && !defined(CL_VERSION_2_0)
#define CL_VERSION_2_0 1
#endif
#if CL_TARGET_OPENCL_VERSION >= 120 && !defined(CL_VERSION_1_2)
#define CL_VERSION_1_2 1
#endif
#if CL_TARGET_OPENCL_VERSION >= 110 && !defined(CL_VERSION_1_1)
#define CL_VERSION_1_1 1
#endif
#if CL_TARGET_OPENCL_VERSION >= 100 && !defined(CL_VERSION_1_0)
#define CL_VERSION_1_0 1
#endif
/* Allow deprecated APIs for older OpenCL versions. */
#if CL_TARGET_OPENCL_VERSION <= 210 && !defined(CL_USE_DEPRECATED_OPENCL_2_1_APIS)
#define CL_USE_DEPRECATED_OPENCL_2_1_APIS
#endif
#if CL_TARGET_OPENCL_VERSION <= 200 && !defined(CL_USE_DEPRECATED_OPENCL_2_0_APIS)
#define CL_USE_DEPRECATED_OPENCL_2_0_APIS
#endif
#if CL_TARGET_OPENCL_VERSION <= 120 && !defined(CL_USE_DEPRECATED_OPENCL_1_2_APIS)
#define CL_USE_DEPRECATED_OPENCL_1_2_APIS
#endif
#if CL_TARGET_OPENCL_VERSION <= 110 && !defined(CL_USE_DEPRECATED_OPENCL_1_1_APIS)
#define CL_USE_DEPRECATED_OPENCL_1_1_APIS
#endif
#if CL_TARGET_OPENCL_VERSION <= 100 && !defined(CL_USE_DEPRECATED_OPENCL_1_0_APIS)
#define CL_USE_DEPRECATED_OPENCL_1_0_APIS
#endif
#endif /* __CL_VERSION_H */

47
src/3rdparty/CL/opencl.h vendored Normal file
View file

@ -0,0 +1,47 @@
/*******************************************************************************
* Copyright (c) 2008-2015 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and/or associated documentation files (the
* "Materials"), to deal in the Materials without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Materials, and to
* permit persons to whom the Materials are furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
* KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
* SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
* https://www.khronos.org/registry/
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.
******************************************************************************/
/* $Revision: 11708 $ on $Date: 2010-06-13 23:36:24 -0700 (Sun, 13 Jun 2010) $ */
#ifndef __OPENCL_H
#define __OPENCL_H
#ifdef __cplusplus
extern "C" {
#endif
#include <CL/cl.h>
#include <CL/cl_gl.h>
#include <CL/cl_gl_ext.h>
#include <CL/cl_ext.h>
#ifdef __cplusplus
}
#endif
#endif /* __OPENCL_H */

33
src/3rdparty/argon2.h vendored Normal file
View file

@ -0,0 +1,33 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* 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/>.
*/
#ifndef XMRIG_3RDPARTY_ARGON2_H
#define XMRIG_3RDPARTY_ARGON2_H
#include "3rdparty/argon2/include/argon2.h"
#endif /* XMRIG_3RDPARTY_ARGON2_H */

83
src/3rdparty/argon2/CMakeLists.txt vendored Normal file
View file

@ -0,0 +1,83 @@
cmake_minimum_required(VERSION 2.6)
project(Argon2 C)
set(ARGON2_VERSION 1.0)
set(CMAKE_C_STANDARD 99)
set(CMAKE_C_STANDARD_REQUIRED ON)
include(CheckCSourceCompiles)
add_library(argon2 STATIC
lib/argon2.c
lib/core.c
lib/encoding.c
lib/genkat.c
lib/impl-select.c
lib/blake2/blake2.c
)
target_include_directories(argon2 PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}/include)
target_include_directories(argon2 PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/lib)
if (CMAKE_C_COMPILER_ID MATCHES MSVC)
function(add_feature_impl FEATURE MSVC_FLAG DEF)
add_library(argon2-${FEATURE} STATIC arch/x86_64/lib/argon2-${FEATURE}.c)
target_include_directories(argon2-${FEATURE} PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/include)
target_include_directories(argon2-${FEATURE} PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/lib)
set_target_properties(argon2-${FEATURE} PROPERTIES POSITION_INDEPENDENT_CODE True)
target_compile_options(argon2-${FEATURE} PRIVATE ${MSVC_FLAG})
target_compile_definitions(argon2-${FEATURE} PRIVATE ${DEF})
target_link_libraries(argon2 PUBLIC argon2-${FEATURE})
endfunction()
add_feature_impl(sse2 "" HAVE_SSE2)
add_feature_impl(ssse3 "/arch:SSSE3" HAVE_SSSE3)
add_feature_impl(xop "" HAVE_XOP)
add_feature_impl(avx2 "/arch:AVX2" HAVE_AVX2)
add_feature_impl(avx512f "/arch:AVX512F" HAVE_AVX512F)
target_sources(argon2 PRIVATE arch/x86_64/lib/argon2-arch.c arch/x86_64/lib/cpu-flags.c)
elseif (NOT XMRIG_ARM AND CMAKE_SIZEOF_VOID_P EQUAL 8)
function(add_feature_impl FEATURE GCC_FLAG DEF)
add_library(argon2-${FEATURE} STATIC arch/x86_64/lib/argon2-${FEATURE}.c)
target_include_directories(argon2-${FEATURE} PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/include)
target_include_directories(argon2-${FEATURE} PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/lib)
set_target_properties(argon2-${FEATURE} PROPERTIES POSITION_INDEPENDENT_CODE True)
message("-- argon2: detecting feature '${FEATURE}'...")
file(READ arch/x86_64/src/test-feature-${FEATURE}.c SOURCE_CODE)
# try without flag:
check_c_source_compiles("${SOURCE_CODE}" FEATURE_${FEATURE}_NOFLAG)
set(HAS_FEATURE ${FEATURE_${FEATURE}_NOFLAG})
if (NOT "${HAS_FEATURE}")
# try with -m<feature> flag:
set(CMAKE_REQUIRED_FLAGS ${GCC_FLAG})
check_c_source_compiles("${SOURCE_CODE}" FEATURE_${FEATURE}_FLAG)
set(CMAKE_REQUIRED_FLAGS "")
set(HAS_FEATURE ${FEATURE_${FEATURE}_FLAG})
if (${HAS_FEATURE})
target_compile_options(argon2-${FEATURE} PRIVATE ${GCC_FLAG})
endif()
endif()
if (${HAS_FEATURE})
message("-- argon2: feature '${FEATURE}' detected!")
target_compile_definitions(argon2-${FEATURE} PRIVATE ${DEF})
endif()
target_link_libraries(argon2 PUBLIC argon2-${FEATURE})
endfunction()
add_feature_impl(sse2 -msse2 HAVE_SSE2)
add_feature_impl(ssse3 -mssse3 HAVE_SSSE3)
add_feature_impl(xop -mxop HAVE_XOP)
add_feature_impl(avx2 -mavx2 HAVE_AVX2)
add_feature_impl(avx512f -mavx512f HAVE_AVX512F)
target_sources(argon2 PRIVATE arch/x86_64/lib/argon2-arch.c arch/x86_64/lib/cpu-flags.c)
else()
target_sources(argon2 PRIVATE arch/generic/lib/argon2-arch.c)
endif()

21
src/3rdparty/argon2/LICENSE vendored Normal file
View file

@ -0,0 +1,21 @@
MIT License
Copyright (c) 2016 Ondrej Mosnáček
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

58
src/3rdparty/argon2/README.md vendored Normal file
View file

@ -0,0 +1,58 @@
# Argon2 [![Build Status](https://travis-ci.org/WOnder93/argon2.svg?branch=master)](https://travis-ci.org/WOnder93/argon2)
A multi-arch library implementing the Argon2 password hashing algorithm.
This project is based on the [original source code](https://github.com/P-H-C/phc-winner-argon2) by the Argon2 authors. The goal of this project is to provide efficient Argon2 implementations for various HW architectures (x86, SSE, ARM, PowerPC, ...).
For the x86_64 architecture, the library implements a simple CPU dispatch which automatically selects the best implementation based on CPU flags and quick benchmarks.
# Building
## Using GNU autotools
To prepare the build environment, run:
```bash
autoreconf -i
./configure
```
After that, just run `make` to build the library.
### Running tests
After configuring the build environment, run `make check` to run the tests.
### Architecture options
You can specify the target architecture by passing the `--host=...` flag to `./configure`.
Supported architectures:
* `x86_64` &ndash; 64-bit x86 architecture
* `generic` &ndash; use generic C impementation
## Using CMake
To prepare the build environment, run:
```bash
cmake -DCMAKE_BUILD_TYPE=Release .
```
Then you can run `make` to build the library.
## Using QMake/Qt Creator
A [QMake](http://doc.qt.io/qt-4.8/qmake-manual.html) project is also available in the `qmake` directory. You can open it in the [Qt Creator IDE](http://wiki.qt.io/Category:Tools::QtCreator) or build it from terminal:
```bash
cd qmake
# see table below for the list of possible ARCH and CONFIG values
qmake ARCH=... CONFIG+=...
make
```
### Architecture options
For QMake builds you can configure support for different architectures. Use the `ARCH` variable to choose the architecture and the `CONFIG` variable to set additional options.
Supported architectures:
* `x86_64` &ndash; 64-bit x86 architecture
* QMake config flags:
* `USE_SSE2` &ndash; use SSE2 instructions
* `USE_SSSE3` &ndash; use SSSE3 instructions
* `USE_XOP` &ndash; use XOP instructions
* `USE_AVX2` &ndash; use AVX2 instructions
* `USE_AVX512F` &ndash; use AVX-512F instructions
* `generic` &ndash; use generic C impementation

20
src/3rdparty/argon2/arch/generic/lib/argon2-arch.c vendored Normal file
View file

@ -0,0 +1,20 @@
#include <stdint.h>
#include <string.h>
#include <stdlib.h>
#include "impl-select.h"
#define rotr64(x, n) (((x) >> (n)) | ((x) << (64 - (n))))
#include "argon2-template-64.h"
void fill_segment_default(const argon2_instance_t *instance,
argon2_position_t position)
{
fill_segment_64(instance, position);
}
void argon2_get_impl_list(argon2_impl_list *list)
{
list->count = 0;
}

41
src/3rdparty/argon2/arch/x86_64/lib/argon2-arch.c vendored Normal file
View file

@ -0,0 +1,41 @@
#include <stdint.h>
#include <string.h>
#include <stdlib.h>
#include "impl-select.h"
#include "cpu-flags.h"
#include "argon2-sse2.h"
#include "argon2-ssse3.h"
#include "argon2-xop.h"
#include "argon2-avx2.h"
#include "argon2-avx512f.h"
/* NOTE: there is no portable intrinsic for 64-bit rotate, but any
* sane compiler should be able to compile this into a ROR instruction: */
#define rotr64(x, n) ((x) >> (n)) | ((x) << (64 - (n)))
#include "argon2-template-64.h"
void fill_segment_default(const argon2_instance_t *instance,
argon2_position_t position)
{
fill_segment_64(instance, position);
}
void argon2_get_impl_list(argon2_impl_list *list)
{
static const argon2_impl IMPLS[] = {
{ "x86_64", NULL, fill_segment_default },
{ "SSE2", check_sse2, fill_segment_sse2 },
{ "SSSE3", check_ssse3, fill_segment_ssse3 },
{ "XOP", check_xop, fill_segment_xop },
{ "AVX2", check_avx2, fill_segment_avx2 },
{ "AVX-512F", check_avx512f, fill_segment_avx512f },
};
cpu_flags_get();
list->count = sizeof(IMPLS) / sizeof(IMPLS[0]);
list->entries = IMPLS;
}

347
src/3rdparty/argon2/arch/x86_64/lib/argon2-avx2.c vendored Normal file
View file

@ -0,0 +1,347 @@
#include "argon2-avx2.h"
#ifdef HAVE_AVX2
#include <string.h>
#ifdef __GNUC__
# include <x86intrin.h>
#else
# include <intrin.h>
#endif
#include "cpu-flags.h"
#define r16 (_mm256_setr_epi8( \
2, 3, 4, 5, 6, 7, 0, 1, \
10, 11, 12, 13, 14, 15, 8, 9, \
18, 19, 20, 21, 22, 23, 16, 17, \
26, 27, 28, 29, 30, 31, 24, 25))
#define r24 (_mm256_setr_epi8( \
3, 4, 5, 6, 7, 0, 1, 2, \
11, 12, 13, 14, 15, 8, 9, 10, \
19, 20, 21, 22, 23, 16, 17, 18, \
27, 28, 29, 30, 31, 24, 25, 26))
#define ror64_16(x) _mm256_shuffle_epi8((x), r16)
#define ror64_24(x) _mm256_shuffle_epi8((x), r24)
#define ror64_32(x) _mm256_shuffle_epi32((x), _MM_SHUFFLE(2, 3, 0, 1))
#define ror64_63(x) \
_mm256_xor_si256(_mm256_srli_epi64((x), 63), _mm256_add_epi64((x), (x)))
static __m256i f(__m256i x, __m256i y)
{
__m256i z = _mm256_mul_epu32(x, y);
return _mm256_add_epi64(_mm256_add_epi64(x, y), _mm256_add_epi64(z, z));
}
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm256_xor_si256(D0, A0); \
D1 = _mm256_xor_si256(D1, A1); \
\
D0 = ror64_32(D0); \
D1 = ror64_32(D1); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm256_xor_si256(B0, C0); \
B1 = _mm256_xor_si256(B1, C1); \
\
B0 = ror64_24(B0); \
B1 = ror64_24(B1); \
} while ((void)0, 0)
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm256_xor_si256(D0, A0); \
D1 = _mm256_xor_si256(D1, A1); \
\
D0 = ror64_16(D0); \
D1 = ror64_16(D1); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm256_xor_si256(B0, C0); \
B1 = _mm256_xor_si256(B1, C1); \
\
B0 = ror64_63(B0); \
B1 = ror64_63(B1); \
} while ((void)0, 0)
#define DIAGONALIZE1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
B0 = _mm256_permute4x64_epi64(B0, _MM_SHUFFLE(0, 3, 2, 1)); \
B1 = _mm256_permute4x64_epi64(B1, _MM_SHUFFLE(0, 3, 2, 1)); \
\
C0 = _mm256_permute4x64_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
C1 = _mm256_permute4x64_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
\
D0 = _mm256_permute4x64_epi64(D0, _MM_SHUFFLE(2, 1, 0, 3)); \
D1 = _mm256_permute4x64_epi64(D1, _MM_SHUFFLE(2, 1, 0, 3)); \
} while ((void)0, 0)
#define UNDIAGONALIZE1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
B0 = _mm256_permute4x64_epi64(B0, _MM_SHUFFLE(2, 1, 0, 3)); \
B1 = _mm256_permute4x64_epi64(B1, _MM_SHUFFLE(2, 1, 0, 3)); \
\
C0 = _mm256_permute4x64_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
C1 = _mm256_permute4x64_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
\
D0 = _mm256_permute4x64_epi64(D0, _MM_SHUFFLE(0, 3, 2, 1)); \
D1 = _mm256_permute4x64_epi64(D1, _MM_SHUFFLE(0, 3, 2, 1)); \
} while ((void)0, 0)
#define DIAGONALIZE2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m256i tmp1, tmp2; \
tmp1 = _mm256_blend_epi32(B0, B1, 0xCC); \
tmp2 = _mm256_blend_epi32(B0, B1, 0x33); \
B1 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
B0 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
\
tmp1 = C0; \
C0 = C1; \
C1 = tmp1; \
\
tmp1 = _mm256_blend_epi32(D0, D1, 0xCC); \
tmp2 = _mm256_blend_epi32(D0, D1, 0x33); \
D0 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
D1 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
} while ((void)0, 0)
#define UNDIAGONALIZE2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m256i tmp1, tmp2; \
tmp1 = _mm256_blend_epi32(B0, B1, 0xCC); \
tmp2 = _mm256_blend_epi32(B0, B1, 0x33); \
B0 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
B1 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
\
tmp1 = C0; \
C0 = C1; \
C1 = tmp1; \
\
tmp1 = _mm256_blend_epi32(D0, D1, 0xCC); \
tmp2 = _mm256_blend_epi32(D0, D1, 0x33); \
D1 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
D0 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
} while ((void)0, 0)
#define BLAKE2_ROUND1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
DIAGONALIZE1(A0, B0, C0, D0, A1, B1, C1, D1); \
\
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
UNDIAGONALIZE1(A0, B0, C0, D0, A1, B1, C1, D1); \
} while ((void)0, 0)
#define BLAKE2_ROUND2(A0, A1, B0, B1, C0, C1, D0, D1) \
do { \
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
DIAGONALIZE2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
UNDIAGONALIZE2(A0, B0, C0, D0, A1, B1, C1, D1); \
} while ((void)0, 0)
enum {
ARGON2_HWORDS_IN_BLOCK = ARGON2_OWORDS_IN_BLOCK / 2,
};
static void fill_block(__m256i *s, const block *ref_block, block *next_block,
int with_xor)
{
__m256i block_XY[ARGON2_HWORDS_IN_BLOCK];
unsigned int i;
if (with_xor) {
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
s[i] =_mm256_xor_si256(
s[i], _mm256_loadu_si256((const __m256i *)ref_block->v + i));
block_XY[i] = _mm256_xor_si256(
s[i], _mm256_loadu_si256((const __m256i *)next_block->v + i));
}
} else {
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
block_XY[i] = s[i] =_mm256_xor_si256(
s[i], _mm256_loadu_si256((const __m256i *)ref_block->v + i));
}
}
for (i = 0; i < 4; ++i) {
BLAKE2_ROUND1(
s[8 * i + 0], s[8 * i + 1], s[8 * i + 2], s[8 * i + 3],
s[8 * i + 4], s[8 * i + 5], s[8 * i + 6], s[8 * i + 7]);
}
for (i = 0; i < 4; ++i) {
BLAKE2_ROUND2(
s[4 * 0 + i], s[4 * 1 + i], s[4 * 2 + i], s[4 * 3 + i],
s[4 * 4 + i], s[4 * 5 + i], s[4 * 6 + i], s[4 * 7 + i]);
}
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
s[i] = _mm256_xor_si256(s[i], block_XY[i]);
_mm256_storeu_si256((__m256i *)next_block->v + i, s[i]);
}
}
static void next_addresses(block *address_block, block *input_block)
{
/*Temporary zero-initialized blocks*/
__m256i zero_block[ARGON2_HWORDS_IN_BLOCK];
__m256i zero2_block[ARGON2_HWORDS_IN_BLOCK];
memset(zero_block, 0, sizeof(zero_block));
memset(zero2_block, 0, sizeof(zero2_block));
/*Increasing index counter*/
input_block->v[6]++;
/*First iteration of G*/
fill_block(zero_block, input_block, address_block, 0);
/*Second iteration of G*/
fill_block(zero2_block, address_block, address_block, 0);
}
void fill_segment_avx2(const argon2_instance_t *instance,
argon2_position_t position)
{
block *ref_block = NULL, *curr_block = NULL;
block address_block, input_block;
uint64_t pseudo_rand, ref_index, ref_lane;
uint32_t prev_offset, curr_offset;
uint32_t starting_index, i;
__m256i state[ARGON2_HWORDS_IN_BLOCK];
int data_independent_addressing;
if (instance == NULL) {
return;
}
data_independent_addressing = (instance->type == Argon2_i) ||
(instance->type == Argon2_id && (position.pass == 0) &&
(position.slice < ARGON2_SYNC_POINTS / 2));
if (data_independent_addressing) {
init_block_value(&input_block, 0);
input_block.v[0] = position.pass;
input_block.v[1] = position.lane;
input_block.v[2] = position.slice;
input_block.v[3] = instance->memory_blocks;
input_block.v[4] = instance->passes;
input_block.v[5] = instance->type;
}
starting_index = 0;
if ((0 == position.pass) && (0 == position.slice)) {
starting_index = 2; /* we have already generated the first two blocks */
/* Don't forget to generate the first block of addresses: */
if (data_independent_addressing) {
next_addresses(&address_block, &input_block);
}
}
/* Offset of the current block */
curr_offset = position.lane * instance->lane_length +
position.slice * instance->segment_length + starting_index;
if (0 == curr_offset % instance->lane_length) {
/* Last block in this lane */
prev_offset = curr_offset + instance->lane_length - 1;
} else {
/* Previous block */
prev_offset = curr_offset - 1;
}
memcpy(state, ((instance->memory + prev_offset)->v), ARGON2_BLOCK_SIZE);
for (i = starting_index; i < instance->segment_length;
++i, ++curr_offset, ++prev_offset) {
/*1.1 Rotating prev_offset if needed */
if (curr_offset % instance->lane_length == 1) {
prev_offset = curr_offset - 1;
}
/* 1.2 Computing the index of the reference block */
/* 1.2.1 Taking pseudo-random value from the previous block */
if (data_independent_addressing) {
if (i % ARGON2_ADDRESSES_IN_BLOCK == 0) {
next_addresses(&address_block, &input_block);
}
pseudo_rand = address_block.v[i % ARGON2_ADDRESSES_IN_BLOCK];
} else {
pseudo_rand = instance->memory[prev_offset].v[0];
}
/* 1.2.2 Computing the lane of the reference block */
ref_lane = ((pseudo_rand >> 32)) % instance->lanes;
if ((position.pass == 0) && (position.slice == 0)) {
/* Can not reference other lanes yet */
ref_lane = position.lane;
}
/* 1.2.3 Computing the number of possible reference block within the
* lane.
*/
position.index = i;
ref_index = index_alpha(instance, &position, pseudo_rand & 0xFFFFFFFF,
ref_lane == position.lane);
/* 2 Creating a new block */
ref_block =
instance->memory + instance->lane_length * ref_lane + ref_index;
curr_block = instance->memory + curr_offset;
/* version 1.2.1 and earlier: overwrite, not XOR */
if (0 == position.pass || ARGON2_VERSION_10 == instance->version) {
fill_block(state, ref_block, curr_block, 0);
} else {
fill_block(state, ref_block, curr_block, 1);
}
}
}
int check_avx2(void)
{
return cpu_flags_have_avx2();
}
#else
void fill_segment_avx2(const argon2_instance_t *instance,
argon2_position_t position)
{
}
int check_avx2(void)
{
return 0;
}
#endif

11
src/3rdparty/argon2/arch/x86_64/lib/argon2-avx2.h vendored Normal file
View file

@ -0,0 +1,11 @@
#ifndef ARGON2_AVX2_H
#define ARGON2_AVX2_H
#include "core.h"
void fill_segment_avx2(const argon2_instance_t *instance,
argon2_position_t position);
int check_avx2(void);
#endif // ARGON2_AVX2_H

332
src/3rdparty/argon2/arch/x86_64/lib/argon2-avx512f.c vendored Normal file
View file

@ -0,0 +1,332 @@
#include "argon2-avx512f.h"
#ifdef HAVE_AVX512F
#include <stdint.h>
#include <string.h>
#ifdef __GNUC__
# include <x86intrin.h>
#else
# include <intrin.h>
#endif
#include "cpu-flags.h"
#define ror64(x, n) _mm512_ror_epi64((x), (n))
static __m512i f(__m512i x, __m512i y)
{
__m512i z = _mm512_mul_epu32(x, y);
return _mm512_add_epi64(_mm512_add_epi64(x, y), _mm512_add_epi64(z, z));
}
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm512_xor_si512(D0, A0); \
D1 = _mm512_xor_si512(D1, A1); \
\
D0 = ror64(D0, 32); \
D1 = ror64(D1, 32); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm512_xor_si512(B0, C0); \
B1 = _mm512_xor_si512(B1, C1); \
\
B0 = ror64(B0, 24); \
B1 = ror64(B1, 24); \
} while ((void)0, 0)
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm512_xor_si512(D0, A0); \
D1 = _mm512_xor_si512(D1, A1); \
\
D0 = ror64(D0, 16); \
D1 = ror64(D1, 16); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm512_xor_si512(B0, C0); \
B1 = _mm512_xor_si512(B1, C1); \
\
B0 = ror64(B0, 63); \
B1 = ror64(B1, 63); \
} while ((void)0, 0)
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
B0 = _mm512_permutex_epi64(B0, _MM_SHUFFLE(0, 3, 2, 1)); \
B1 = _mm512_permutex_epi64(B1, _MM_SHUFFLE(0, 3, 2, 1)); \
\
C0 = _mm512_permutex_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
C1 = _mm512_permutex_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
\
D0 = _mm512_permutex_epi64(D0, _MM_SHUFFLE(2, 1, 0, 3)); \
D1 = _mm512_permutex_epi64(D1, _MM_SHUFFLE(2, 1, 0, 3)); \
} while ((void)0, 0)
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
B0 = _mm512_permutex_epi64(B0, _MM_SHUFFLE(2, 1, 0, 3)); \
B1 = _mm512_permutex_epi64(B1, _MM_SHUFFLE(2, 1, 0, 3)); \
\
C0 = _mm512_permutex_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
C1 = _mm512_permutex_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
\
D0 = _mm512_permutex_epi64(D0, _MM_SHUFFLE(0, 3, 2, 1)); \
D1 = _mm512_permutex_epi64(D1, _MM_SHUFFLE(0, 3, 2, 1)); \
} while ((void)0, 0)
#define BLAKE2_ROUND(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
\
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
} while ((void)0, 0)
#define SWAP_HALVES(A0, A1) \
do { \
__m512i t0, t1; \
t0 = _mm512_shuffle_i64x2(A0, A1, _MM_SHUFFLE(1, 0, 1, 0)); \
t1 = _mm512_shuffle_i64x2(A0, A1, _MM_SHUFFLE(3, 2, 3, 2)); \
A0 = t0; \
A1 = t1; \
} while((void)0, 0)
#define SWAP_QUARTERS(A0, A1) \
do { \
SWAP_HALVES(A0, A1); \
A0 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A0); \
A1 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A1); \
} while((void)0, 0)
#define UNSWAP_QUARTERS(A0, A1) \
do { \
A0 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A0); \
A1 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A1); \
SWAP_HALVES(A0, A1); \
} while((void)0, 0)
#define BLAKE2_ROUND1(A0, C0, B0, D0, A1, C1, B1, D1) \
do { \
SWAP_HALVES(A0, B0); \
SWAP_HALVES(C0, D0); \
SWAP_HALVES(A1, B1); \
SWAP_HALVES(C1, D1); \
BLAKE2_ROUND(A0, B0, C0, D0, A1, B1, C1, D1); \
SWAP_HALVES(A0, B0); \
SWAP_HALVES(C0, D0); \
SWAP_HALVES(A1, B1); \
SWAP_HALVES(C1, D1); \
} while ((void)0, 0)
#define BLAKE2_ROUND2(A0, A1, B0, B1, C0, C1, D0, D1) \
do { \
SWAP_QUARTERS(A0, A1); \
SWAP_QUARTERS(B0, B1); \
SWAP_QUARTERS(C0, C1); \
SWAP_QUARTERS(D0, D1); \
BLAKE2_ROUND(A0, B0, C0, D0, A1, B1, C1, D1); \
UNSWAP_QUARTERS(A0, A1); \
UNSWAP_QUARTERS(B0, B1); \
UNSWAP_QUARTERS(C0, C1); \
UNSWAP_QUARTERS(D0, D1); \
} while ((void)0, 0)
enum {
ARGON2_VECS_IN_BLOCK = ARGON2_OWORDS_IN_BLOCK / 4,
};
static void fill_block(__m512i *s, const block *ref_block, block *next_block,
int with_xor)
{
__m512i block_XY[ARGON2_VECS_IN_BLOCK];
unsigned int i;
if (with_xor) {
for (i = 0; i < ARGON2_VECS_IN_BLOCK; i++) {
s[i] =_mm512_xor_si512(
s[i], _mm512_loadu_si512((const __m512i *)ref_block->v + i));
block_XY[i] = _mm512_xor_si512(
s[i], _mm512_loadu_si512((const __m512i *)next_block->v + i));
}
} else {
for (i = 0; i < ARGON2_VECS_IN_BLOCK; i++) {
block_XY[i] = s[i] =_mm512_xor_si512(
s[i], _mm512_loadu_si512((const __m512i *)ref_block->v + i));
}
}
for (i = 0; i < 2; ++i) {
BLAKE2_ROUND1(
s[8 * i + 0], s[8 * i + 1], s[8 * i + 2], s[8 * i + 3],
s[8 * i + 4], s[8 * i + 5], s[8 * i + 6], s[8 * i + 7]);
}
for (i = 0; i < 2; ++i) {
BLAKE2_ROUND2(
s[2 * 0 + i], s[2 * 1 + i], s[2 * 2 + i], s[2 * 3 + i],
s[2 * 4 + i], s[2 * 5 + i], s[2 * 6 + i], s[2 * 7 + i]);
}
for (i = 0; i < ARGON2_VECS_IN_BLOCK; i++) {
s[i] = _mm512_xor_si512(s[i], block_XY[i]);
_mm512_storeu_si512((__m512i *)next_block->v + i, s[i]);
}
}
static void next_addresses(block *address_block, block *input_block)
{
/*Temporary zero-initialized blocks*/
__m512i zero_block[ARGON2_VECS_IN_BLOCK];
__m512i zero2_block[ARGON2_VECS_IN_BLOCK];
memset(zero_block, 0, sizeof(zero_block));
memset(zero2_block, 0, sizeof(zero2_block));
/*Increasing index counter*/
input_block->v[6]++;
/*First iteration of G*/
fill_block(zero_block, input_block, address_block, 0);
/*Second iteration of G*/
fill_block(zero2_block, address_block, address_block, 0);
}
void fill_segment_avx512f(const argon2_instance_t *instance,
argon2_position_t position)
{
block *ref_block = NULL, *curr_block = NULL;
block address_block, input_block;
uint64_t pseudo_rand, ref_index, ref_lane;
uint32_t prev_offset, curr_offset;
uint32_t starting_index, i;
__m512i state[ARGON2_VECS_IN_BLOCK];
int data_independent_addressing;
if (instance == NULL) {
return;
}
data_independent_addressing = (instance->type == Argon2_i) ||
(instance->type == Argon2_id && (position.pass == 0) &&
(position.slice < ARGON2_SYNC_POINTS / 2));
if (data_independent_addressing) {
init_block_value(&input_block, 0);
input_block.v[0] = position.pass;
input_block.v[1] = position.lane;
input_block.v[2] = position.slice;
input_block.v[3] = instance->memory_blocks;
input_block.v[4] = instance->passes;
input_block.v[5] = instance->type;
}
starting_index = 0;
if ((0 == position.pass) && (0 == position.slice)) {
starting_index = 2; /* we have already generated the first two blocks */
/* Don't forget to generate the first block of addresses: */
if (data_independent_addressing) {
next_addresses(&address_block, &input_block);
}
}
/* Offset of the current block */
curr_offset = position.lane * instance->lane_length +
position.slice * instance->segment_length + starting_index;
if (0 == curr_offset % instance->lane_length) {
/* Last block in this lane */
prev_offset = curr_offset + instance->lane_length - 1;
} else {
/* Previous block */
prev_offset = curr_offset - 1;
}
memcpy(state, ((instance->memory + prev_offset)->v), ARGON2_BLOCK_SIZE);
for (i = starting_index; i < instance->segment_length;
++i, ++curr_offset, ++prev_offset) {
/*1.1 Rotating prev_offset if needed */
if (curr_offset % instance->lane_length == 1) {
prev_offset = curr_offset - 1;
}
/* 1.2 Computing the index of the reference block */
/* 1.2.1 Taking pseudo-random value from the previous block */
if (data_independent_addressing) {
if (i % ARGON2_ADDRESSES_IN_BLOCK == 0) {
next_addresses(&address_block, &input_block);
}
pseudo_rand = address_block.v[i % ARGON2_ADDRESSES_IN_BLOCK];
} else {
pseudo_rand = instance->memory[prev_offset].v[0];
}
/* 1.2.2 Computing the lane of the reference block */
ref_lane = ((pseudo_rand >> 32)) % instance->lanes;
if ((position.pass == 0) && (position.slice == 0)) {
/* Can not reference other lanes yet */
ref_lane = position.lane;
}
/* 1.2.3 Computing the number of possible reference block within the
* lane.
*/
position.index = i;
ref_index = index_alpha(instance, &position, pseudo_rand & 0xFFFFFFFF,
ref_lane == position.lane);
/* 2 Creating a new block */
ref_block =
instance->memory + instance->lane_length * ref_lane + ref_index;
curr_block = instance->memory + curr_offset;
/* version 1.2.1 and earlier: overwrite, not XOR */
if (0 == position.pass || ARGON2_VERSION_10 == instance->version) {
fill_block(state, ref_block, curr_block, 0);
} else {
fill_block(state, ref_block, curr_block, 1);
}
}
}
int check_avx512f(void)
{
return cpu_flags_have_avx512f();
}
#else
void fill_segment_avx512f(const argon2_instance_t *instance,
argon2_position_t position)
{
}
int check_avx512f(void)
{
return 0;
}
#endif

11
src/3rdparty/argon2/arch/x86_64/lib/argon2-avx512f.h vendored Normal file
View file

@ -0,0 +1,11 @@
#ifndef ARGON2_AVX512F_H
#define ARGON2_AVX512F_H
#include "core.h"
void fill_segment_avx512f(const argon2_instance_t *instance,
argon2_position_t position);
int check_avx512f(void);
#endif // ARGON2_AVX512F_H

128
src/3rdparty/argon2/arch/x86_64/lib/argon2-sse2.c vendored Normal file
View file

@ -0,0 +1,128 @@
#include "argon2-sse2.h"
#ifdef HAVE_SSE2
#ifdef __GNUC__
# include <x86intrin.h>
#else
# include <intrin.h>
#endif
#include "cpu-flags.h"
#define ror64_16(x) \
_mm_shufflehi_epi16( \
_mm_shufflelo_epi16((x), _MM_SHUFFLE(0, 3, 2, 1)), \
_MM_SHUFFLE(0, 3, 2, 1))
#define ror64_24(x) \
_mm_xor_si128(_mm_srli_epi64((x), 24), _mm_slli_epi64((x), 40))
#define ror64_32(x) _mm_shuffle_epi32((x), _MM_SHUFFLE(2, 3, 0, 1))
#define ror64_63(x) \
_mm_xor_si128(_mm_srli_epi64((x), 63), _mm_add_epi64((x), (x)))
static __m128i f(__m128i x, __m128i y)
{
__m128i z = _mm_mul_epu32(x, y);
return _mm_add_epi64(_mm_add_epi64(x, y), _mm_add_epi64(z, z));
}
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm_xor_si128(D0, A0); \
D1 = _mm_xor_si128(D1, A1); \
\
D0 = ror64_32(D0); \
D1 = ror64_32(D1); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm_xor_si128(B0, C0); \
B1 = _mm_xor_si128(B1, C1); \
\
B0 = ror64_24(B0); \
B1 = ror64_24(B1); \
} while ((void)0, 0)
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm_xor_si128(D0, A0); \
D1 = _mm_xor_si128(D1, A1); \
\
D0 = ror64_16(D0); \
D1 = ror64_16(D1); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm_xor_si128(B0, C0); \
B1 = _mm_xor_si128(B1, C1); \
\
B0 = ror64_63(B0); \
B1 = ror64_63(B1); \
} while ((void)0, 0)
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m128i t0 = D0; \
__m128i t1 = B0; \
D0 = _mm_unpackhi_epi64(D1, _mm_unpacklo_epi64(t0, t0)); \
D1 = _mm_unpackhi_epi64(t0, _mm_unpacklo_epi64(D1, D1)); \
B0 = _mm_unpackhi_epi64(B0, _mm_unpacklo_epi64(B1, B1)); \
B1 = _mm_unpackhi_epi64(B1, _mm_unpacklo_epi64(t1, t1)); \
} while ((void)0, 0)
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m128i t0 = B0; \
__m128i t1 = D0; \
B0 = _mm_unpackhi_epi64(B1, _mm_unpacklo_epi64(B0, B0)); \
B1 = _mm_unpackhi_epi64(t0, _mm_unpacklo_epi64(B1, B1)); \
D0 = _mm_unpackhi_epi64(D0, _mm_unpacklo_epi64(D1, D1)); \
D1 = _mm_unpackhi_epi64(D1, _mm_unpacklo_epi64(t1, t1)); \
} while ((void)0, 0)
#define BLAKE2_ROUND(A0, A1, B0, B1, C0, C1, D0, D1) \
do { \
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
\
G1(A0, B0, C1, D0, A1, B1, C0, D1); \
G2(A0, B0, C1, D0, A1, B1, C0, D1); \
\
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
} while ((void)0, 0)
#include "argon2-template-128.h"
void fill_segment_sse2(const argon2_instance_t *instance,
argon2_position_t position)
{
fill_segment_128(instance, position);
}
int check_sse2(void)
{
return cpu_flags_have_sse2();
}
#else
void fill_segment_sse2(const argon2_instance_t *instance,
argon2_position_t position)
{
}
int check_sse2(void)
{
return 0;
}
#endif

11
src/3rdparty/argon2/arch/x86_64/lib/argon2-sse2.h vendored Normal file
View file

@ -0,0 +1,11 @@
#ifndef ARGON2_SSE2_H
#define ARGON2_SSE2_H
#include "core.h"
void fill_segment_sse2(const argon2_instance_t *instance,
argon2_position_t position);
int check_sse2(void);
#endif // ARGON2_SSE2_H

140
src/3rdparty/argon2/arch/x86_64/lib/argon2-ssse3.c vendored Normal file
View file

@ -0,0 +1,140 @@
#include "argon2-ssse3.h"
#ifdef HAVE_SSSE3
#include <string.h>
#ifdef __GNUC__
# include <x86intrin.h>
#else
# include <intrin.h>
#endif
#include "cpu-flags.h"
#define r16 (_mm_setr_epi8( \
2, 3, 4, 5, 6, 7, 0, 1, \
10, 11, 12, 13, 14, 15, 8, 9))
#define r24 (_mm_setr_epi8( \
3, 4, 5, 6, 7, 0, 1, 2, \
11, 12, 13, 14, 15, 8, 9, 10))
#define ror64_16(x) _mm_shuffle_epi8((x), r16)
#define ror64_24(x) _mm_shuffle_epi8((x), r24)
#define ror64_32(x) _mm_shuffle_epi32((x), _MM_SHUFFLE(2, 3, 0, 1))
#define ror64_63(x) \
_mm_xor_si128(_mm_srli_epi64((x), 63), _mm_add_epi64((x), (x)))
static __m128i f(__m128i x, __m128i y)
{
__m128i z = _mm_mul_epu32(x, y);
return _mm_add_epi64(_mm_add_epi64(x, y), _mm_add_epi64(z, z));
}
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm_xor_si128(D0, A0); \
D1 = _mm_xor_si128(D1, A1); \
\
D0 = ror64_32(D0); \
D1 = ror64_32(D1); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm_xor_si128(B0, C0); \
B1 = _mm_xor_si128(B1, C1); \
\
B0 = ror64_24(B0); \
B1 = ror64_24(B1); \
} while ((void)0, 0)
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm_xor_si128(D0, A0); \
D1 = _mm_xor_si128(D1, A1); \
\
D0 = ror64_16(D0); \
D1 = ror64_16(D1); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm_xor_si128(B0, C0); \
B1 = _mm_xor_si128(B1, C1); \
\
B0 = ror64_63(B0); \
B1 = ror64_63(B1); \
} while ((void)0, 0)
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m128i t0 = _mm_alignr_epi8(B1, B0, 8); \
__m128i t1 = _mm_alignr_epi8(B0, B1, 8); \
B0 = t0; \
B1 = t1; \
\
t0 = _mm_alignr_epi8(D1, D0, 8); \
t1 = _mm_alignr_epi8(D0, D1, 8); \
D0 = t1; \
D1 = t0; \
} while ((void)0, 0)
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m128i t0 = _mm_alignr_epi8(B0, B1, 8); \
__m128i t1 = _mm_alignr_epi8(B1, B0, 8); \
B0 = t0; \
B1 = t1; \
\
t0 = _mm_alignr_epi8(D0, D1, 8); \
t1 = _mm_alignr_epi8(D1, D0, 8); \
D0 = t1; \
D1 = t0; \
} while ((void)0, 0)
#define BLAKE2_ROUND(A0, A1, B0, B1, C0, C1, D0, D1) \
do { \
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
\
G1(A0, B0, C1, D0, A1, B1, C0, D1); \
G2(A0, B0, C1, D0, A1, B1, C0, D1); \
\
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
} while ((void)0, 0)
#include "argon2-template-128.h"
void fill_segment_ssse3(const argon2_instance_t *instance,
argon2_position_t position)
{
fill_segment_128(instance, position);
}
int check_ssse3(void)
{
return cpu_flags_have_ssse3();
}
#else
void fill_segment_ssse3(const argon2_instance_t *instance,
argon2_position_t position)
{
}
int check_ssse3(void)
{
return 0;
}
#endif

11
src/3rdparty/argon2/arch/x86_64/lib/argon2-ssse3.h vendored Normal file
View file

@ -0,0 +1,11 @@
#ifndef ARGON2_SSSE3_H
#define ARGON2_SSSE3_H
#include "core.h"
void fill_segment_ssse3(const argon2_instance_t *instance,
argon2_position_t position);
int check_ssse3(void);
#endif // ARGON2_SSSE3_H

168
src/3rdparty/argon2/arch/x86_64/lib/argon2-template-128.h vendored Normal file
View file

@ -0,0 +1,168 @@
#include <string.h>
#ifdef __GNUC__
# include <x86intrin.h>
#else
# include <intrin.h>
#endif
#include "core.h"
static void fill_block(__m128i *s, const block *ref_block, block *next_block,
int with_xor)
{
__m128i block_XY[ARGON2_OWORDS_IN_BLOCK];
unsigned int i;
if (with_xor) {
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
s[i] = _mm_xor_si128(
s[i], _mm_loadu_si128((const __m128i *)ref_block->v + i));
block_XY[i] = _mm_xor_si128(
s[i], _mm_loadu_si128((const __m128i *)next_block->v + i));
}
} else {
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
block_XY[i] = s[i] = _mm_xor_si128(
s[i], _mm_loadu_si128((const __m128i *)ref_block->v + i));
}
}
for (i = 0; i < 8; ++i) {
BLAKE2_ROUND(
s[8 * i + 0], s[8 * i + 1], s[8 * i + 2], s[8 * i + 3],
s[8 * i + 4], s[8 * i + 5], s[8 * i + 6], s[8 * i + 7]);
}
for (i = 0; i < 8; ++i) {
BLAKE2_ROUND(
s[8 * 0 + i], s[8 * 1 + i], s[8 * 2 + i], s[8 * 3 + i],
s[8 * 4 + i], s[8 * 5 + i], s[8 * 6 + i], s[8 * 7 + i]);
}
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
s[i] = _mm_xor_si128(s[i], block_XY[i]);
_mm_storeu_si128((__m128i *)next_block->v + i, s[i]);
}
}
static void next_addresses(block *address_block, block *input_block)
{
/*Temporary zero-initialized blocks*/
__m128i zero_block[ARGON2_OWORDS_IN_BLOCK];
__m128i zero2_block[ARGON2_OWORDS_IN_BLOCK];
memset(zero_block, 0, sizeof(zero_block));
memset(zero2_block, 0, sizeof(zero2_block));
/*Increasing index counter*/
input_block->v[6]++;
/*First iteration of G*/
fill_block(zero_block, input_block, address_block, 0);
/*Second iteration of G*/
fill_block(zero2_block, address_block, address_block, 0);
}
static void fill_segment_128(const argon2_instance_t *instance,
argon2_position_t position)
{
block *ref_block = NULL, *curr_block = NULL;
block address_block, input_block;
uint64_t pseudo_rand, ref_index, ref_lane;
uint32_t prev_offset, curr_offset;
uint32_t starting_index, i;
__m128i state[ARGON2_OWORDS_IN_BLOCK];
int data_independent_addressing;
if (instance == NULL) {
return;
}
data_independent_addressing = (instance->type == Argon2_i) ||
(instance->type == Argon2_id && (position.pass == 0) &&
(position.slice < ARGON2_SYNC_POINTS / 2));
if (data_independent_addressing) {
init_block_value(&input_block, 0);
input_block.v[0] = position.pass;
input_block.v[1] = position.lane;
input_block.v[2] = position.slice;
input_block.v[3] = instance->memory_blocks;
input_block.v[4] = instance->passes;
input_block.v[5] = instance->type;
}
starting_index = 0;
if ((0 == position.pass) && (0 == position.slice)) {
starting_index = 2; /* we have already generated the first two blocks */
/* Don't forget to generate the first block of addresses: */
if (data_independent_addressing) {
next_addresses(&address_block, &input_block);
}
}
/* Offset of the current block */
curr_offset = position.lane * instance->lane_length +
position.slice * instance->segment_length + starting_index;
if (0 == curr_offset % instance->lane_length) {
/* Last block in this lane */
prev_offset = curr_offset + instance->lane_length - 1;
} else {
/* Previous block */
prev_offset = curr_offset - 1;
}
memcpy(state, ((instance->memory + prev_offset)->v), ARGON2_BLOCK_SIZE);
for (i = starting_index; i < instance->segment_length;
++i, ++curr_offset, ++prev_offset) {
/*1.1 Rotating prev_offset if needed */
if (curr_offset % instance->lane_length == 1) {
prev_offset = curr_offset - 1;
}
/* 1.2 Computing the index of the reference block */
/* 1.2.1 Taking pseudo-random value from the previous block */
if (data_independent_addressing) {
if (i % ARGON2_ADDRESSES_IN_BLOCK == 0) {
next_addresses(&address_block, &input_block);
}
pseudo_rand = address_block.v[i % ARGON2_ADDRESSES_IN_BLOCK];
} else {
pseudo_rand = instance->memory[prev_offset].v[0];
}
/* 1.2.2 Computing the lane of the reference block */
ref_lane = ((pseudo_rand >> 32)) % instance->lanes;
if ((position.pass == 0) && (position.slice == 0)) {
/* Can not reference other lanes yet */
ref_lane = position.lane;
}
/* 1.2.3 Computing the number of possible reference block within the
* lane.
*/
position.index = i;
ref_index = index_alpha(instance, &position, pseudo_rand & 0xFFFFFFFF,
ref_lane == position.lane);
/* 2 Creating a new block */
ref_block =
instance->memory + instance->lane_length * ref_lane + ref_index;
curr_block = instance->memory + curr_offset;
/* version 1.2.1 and earlier: overwrite, not XOR */
if (0 == position.pass || ARGON2_VERSION_10 == instance->version) {
fill_block(state, ref_block, curr_block, 0);
} else {
fill_block(state, ref_block, curr_block, 1);
}
}
}

128
src/3rdparty/argon2/arch/x86_64/lib/argon2-xop.c vendored Normal file
View file

@ -0,0 +1,128 @@
#include "argon2-xop.h"
#ifdef HAVE_XOP
#include <string.h>
#ifdef __GNUC__
# include <x86intrin.h>
#else
# include <intrin.h>
#endif
#include "cpu-flags.h"
#define ror64(x, c) _mm_roti_epi64((x), -(c))
static __m128i f(__m128i x, __m128i y)
{
__m128i z = _mm_mul_epu32(x, y);
return _mm_add_epi64(_mm_add_epi64(x, y), _mm_add_epi64(z, z));
}
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm_xor_si128(D0, A0); \
D1 = _mm_xor_si128(D1, A1); \
\
D0 = ror64(D0, 32); \
D1 = ror64(D1, 32); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm_xor_si128(B0, C0); \
B1 = _mm_xor_si128(B1, C1); \
\
B0 = ror64(B0, 24); \
B1 = ror64(B1, 24); \
} while ((void)0, 0)
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
A0 = f(A0, B0); \
A1 = f(A1, B1); \
\
D0 = _mm_xor_si128(D0, A0); \
D1 = _mm_xor_si128(D1, A1); \
\
D0 = ror64(D0, 16); \
D1 = ror64(D1, 16); \
\
C0 = f(C0, D0); \
C1 = f(C1, D1); \
\
B0 = _mm_xor_si128(B0, C0); \
B1 = _mm_xor_si128(B1, C1); \
\
B0 = ror64(B0, 63); \
B1 = ror64(B1, 63); \
} while ((void)0, 0)
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m128i t0 = _mm_alignr_epi8(B1, B0, 8); \
__m128i t1 = _mm_alignr_epi8(B0, B1, 8); \
B0 = t0; \
B1 = t1; \
\
t0 = _mm_alignr_epi8(D1, D0, 8); \
t1 = _mm_alignr_epi8(D0, D1, 8); \
D0 = t1; \
D1 = t0; \
} while ((void)0, 0)
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
do { \
__m128i t0 = _mm_alignr_epi8(B0, B1, 8); \
__m128i t1 = _mm_alignr_epi8(B1, B0, 8); \
B0 = t0; \
B1 = t1; \
\
t0 = _mm_alignr_epi8(D0, D1, 8); \
t1 = _mm_alignr_epi8(D1, D0, 8); \
D0 = t1; \
D1 = t0; \
} while ((void)0, 0)
#define BLAKE2_ROUND(A0, A1, B0, B1, C0, C1, D0, D1) \
do { \
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
\
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
\
G1(A0, B0, C1, D0, A1, B1, C0, D1); \
G2(A0, B0, C1, D0, A1, B1, C0, D1); \
\
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
} while ((void)0, 0)
#include "argon2-template-128.h"
void fill_segment_xop(const argon2_instance_t *instance,
argon2_position_t position)
{
fill_segment_128(instance, position);
}
int check_xop(void)
{
return cpu_flags_have_xop();
}
#else
void fill_segment_xop(const argon2_instance_t *instance,
argon2_position_t position)
{
}
int check_xop(void)
{
return 0;
}
#endif

11
src/3rdparty/argon2/arch/x86_64/lib/argon2-xop.h vendored Normal file
View file

@ -0,0 +1,11 @@
#ifndef ARGON2_XOP_H
#define ARGON2_XOP_H
#include "core.h"
void fill_segment_xop(const argon2_instance_t *instance,
argon2_position_t position);
int check_xop(void);
#endif // ARGON2_XOP_H

129
src/3rdparty/argon2/arch/x86_64/lib/cpu-flags.c vendored Normal file
View file

@ -0,0 +1,129 @@
#include <stdbool.h>
#include <stdint.h>
#include "cpu-flags.h"
#include <stdio.h>
#ifdef _MSC_VER
# include <intrin.h>
#else
# include <cpuid.h>
#endif
#ifndef bit_OSXSAVE
# define bit_OSXSAVE (1 << 27)
#endif
#ifndef bit_SSE2
# define bit_SSE2 (1 << 26)
#endif
#ifndef bit_SSSE3
# define bit_SSSE3 (1 << 9)
#endif
#ifndef bit_AVX2
# define bit_AVX2 (1 << 5)
#endif
#ifndef bit_AVX512F
# define bit_AVX512F (1 << 16)
#endif
#ifndef bit_XOP
# define bit_XOP (1 << 11)
#endif
#define PROCESSOR_INFO (1)
#define EXTENDED_FEATURES (7)
#define EAX_Reg (0)
#define EBX_Reg (1)
#define ECX_Reg (2)
#define EDX_Reg (3)
enum {
X86_64_FEATURE_SSE2 = (1 << 0),
X86_64_FEATURE_SSSE3 = (1 << 1),
X86_64_FEATURE_XOP = (1 << 2),
X86_64_FEATURE_AVX2 = (1 << 3),
X86_64_FEATURE_AVX512F = (1 << 4),
};
static unsigned int cpu_flags;
static inline void cpuid(uint32_t level, int32_t output[4])
{
# ifdef _MSC_VER
__cpuid(output, (int) level);
# else
__cpuid_count(level, 0, output[0], output[1], output[2], output[3]);
# endif
}
static bool has_feature(uint32_t level, uint32_t reg, int32_t bit)
{
int32_t cpu_info[4] = { 0 };
cpuid(level, cpu_info);
return (cpu_info[reg] & bit) != 0;
}
void cpu_flags_get(void)
{
if (has_feature(PROCESSOR_INFO, EDX_Reg, bit_SSE2)) {
cpu_flags |= X86_64_FEATURE_SSE2;
}
if (has_feature(PROCESSOR_INFO, ECX_Reg, bit_SSSE3)) {
cpu_flags |= X86_64_FEATURE_SSSE3;
}
if (!has_feature(PROCESSOR_INFO, ECX_Reg, bit_OSXSAVE)) {
return;
}
if (has_feature(EXTENDED_FEATURES, EBX_Reg, bit_AVX2)) {
cpu_flags |= X86_64_FEATURE_AVX2;
}
if (has_feature(EXTENDED_FEATURES, EBX_Reg, bit_AVX512F)) {
cpu_flags |= X86_64_FEATURE_AVX512F;
}
if (has_feature(0x80000001, ECX_Reg, bit_XOP)) {
cpu_flags |= X86_64_FEATURE_XOP;
}
}
int cpu_flags_have_sse2(void)
{
return cpu_flags & X86_64_FEATURE_SSE2;
}
int cpu_flags_have_ssse3(void)
{
return cpu_flags & X86_64_FEATURE_SSSE3;
}
int cpu_flags_have_xop(void)
{
return cpu_flags & X86_64_FEATURE_XOP;
}
int cpu_flags_have_avx2(void)
{
return cpu_flags & X86_64_FEATURE_AVX2;
}
int cpu_flags_have_avx512f(void)
{
return cpu_flags & X86_64_FEATURE_AVX512F;
}

12
src/3rdparty/argon2/arch/x86_64/lib/cpu-flags.h vendored Normal file
View file

@ -0,0 +1,12 @@
#ifndef ARGON2_CPU_FLAGS_H
#define ARGON2_CPU_FLAGS_H
void cpu_flags_get(void);
int cpu_flags_have_sse2(void);
int cpu_flags_have_ssse3(void);
int cpu_flags_have_xop(void);
int cpu_flags_have_avx2(void);
int cpu_flags_have_avx512f(void);
#endif // ARGON2_CPU_FLAGS_H

8
src/3rdparty/argon2/arch/x86_64/src/test-feature-avx2.c vendored Normal file
View file

@ -0,0 +1,8 @@
#include <x86intrin.h>
void function_avx2(__m256i *dst, const __m256i *a, const __m256i *b)
{
*dst = _mm256_xor_si256(*a, *b);
}
int main(void) { return 0; }

8
src/3rdparty/argon2/arch/x86_64/src/test-feature-avx512f.c vendored Normal file
View file

@ -0,0 +1,8 @@
#include <x86intrin.h>
void function_avx512f(__m512i *dst, const __m512i *a)
{
*dst = _mm512_ror_epi64(*a, 57);
}
int main(void) { return 0; }

8
src/3rdparty/argon2/arch/x86_64/src/test-feature-sse2.c vendored Normal file
View file

@ -0,0 +1,8 @@
#include <x86intrin.h>
void function_sse2(__m128i *dst, const __m128i *a, const __m128i *b)
{
*dst = _mm_xor_si128(*a, *b);
}
int main(void) { return 0; }

8
src/3rdparty/argon2/arch/x86_64/src/test-feature-ssse3.c vendored Normal file
View file

@ -0,0 +1,8 @@
#include <x86intrin.h>
void function_ssse3(__m128i *dst, const __m128i *a, const __m128i *b)
{
*dst = _mm_shuffle_epi8(*a, *b);
}
int main(void) { return 0; }

8
src/3rdparty/argon2/arch/x86_64/src/test-feature-xop.c vendored Normal file
View file

@ -0,0 +1,8 @@
#include <x86intrin.h>
void function_xop(__m128i *dst, const __m128i *a, int b)
{
*dst = _mm_roti_epi64(*a, b);
}
int main(void) { return 0; }

465
src/3rdparty/argon2/include/argon2.h vendored Normal file
View file

@ -0,0 +1,465 @@
/*
* Argon2 source code package
*
* Written by Daniel Dinu and Dmitry Khovratovich, 2015
*
* This work is licensed under a Creative Commons CC0 1.0 License/Waiver.
*
* 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/>.
*/
#ifndef ARGON2_H
#define ARGON2_H
#include <stdint.h>
#include <stddef.h>
#include <stdio.h>
#include <limits.h>
/* Symbols visibility control */
#define ARGON2_PUBLIC
#if defined(__cplusplus)
extern "C" {
#endif
/*
* Argon2 input parameter restrictions
*/
/* Minimum and maximum number of lanes (degree of parallelism) */
#define ARGON2_MIN_LANES UINT32_C(1)
#define ARGON2_MAX_LANES UINT32_C(0xFFFFFF)
/* Minimum and maximum number of threads */
#define ARGON2_MIN_THREADS UINT32_C(1)
#define ARGON2_MAX_THREADS UINT32_C(0xFFFFFF)
/* Number of synchronization points between lanes per pass */
#define ARGON2_SYNC_POINTS UINT32_C(4)
/* Minimum and maximum digest size in bytes */
#define ARGON2_MIN_OUTLEN UINT32_C(4)
#define ARGON2_MAX_OUTLEN UINT32_C(0xFFFFFFFF)
/* Minimum and maximum number of memory blocks (each of BLOCK_SIZE bytes) */
#define ARGON2_MIN_MEMORY (2 * ARGON2_SYNC_POINTS) /* 2 blocks per slice */
#define ARGON2_MIN(a, b) ((a) < (b) ? (a) : (b))
/* Max memory size is addressing-space/2, topping at 2^32 blocks (4 TB) */
#define ARGON2_MAX_MEMORY_BITS \
ARGON2_MIN(UINT32_C(32), (sizeof(void *) * CHAR_BIT - 10 - 1))
#define ARGON2_MAX_MEMORY \
ARGON2_MIN(UINT32_C(0xFFFFFFFF), UINT64_C(1) << ARGON2_MAX_MEMORY_BITS)
/* Minimum and maximum number of passes */
#define ARGON2_MIN_TIME UINT32_C(1)
#define ARGON2_MAX_TIME UINT32_C(0xFFFFFFFF)
/* Minimum and maximum password length in bytes */
#define ARGON2_MIN_PWD_LENGTH UINT32_C(0)
#define ARGON2_MAX_PWD_LENGTH UINT32_C(0xFFFFFFFF)
/* Minimum and maximum associated data length in bytes */
#define ARGON2_MIN_AD_LENGTH UINT32_C(0)
#define ARGON2_MAX_AD_LENGTH UINT32_C(0xFFFFFFFF)
/* Minimum and maximum salt length in bytes */
#define ARGON2_MIN_SALT_LENGTH UINT32_C(8)
#define ARGON2_MAX_SALT_LENGTH UINT32_C(0xFFFFFFFF)
/* Minimum and maximum key length in bytes */
#define ARGON2_MIN_SECRET UINT32_C(0)
#define ARGON2_MAX_SECRET UINT32_C(0xFFFFFFFF)
/* Flags to determine which fields are securely wiped (default = no wipe). */
#define ARGON2_DEFAULT_FLAGS UINT32_C(0)
#define ARGON2_FLAG_CLEAR_PASSWORD (UINT32_C(1) << 0)
#define ARGON2_FLAG_CLEAR_SECRET (UINT32_C(1) << 1)
#define ARGON2_FLAG_GENKAT (UINT32_C(1) << 3)
/* Global flag to determine if we are wiping internal memory buffers. This flag
* is defined in core.c and deafults to 1 (wipe internal memory). */
extern int FLAG_clear_internal_memory;
/* Error codes */
typedef enum Argon2_ErrorCodes {
ARGON2_OK = 0,
ARGON2_OUTPUT_PTR_NULL = -1,
ARGON2_OUTPUT_TOO_SHORT = -2,
ARGON2_OUTPUT_TOO_LONG = -3,
ARGON2_PWD_TOO_SHORT = -4,
ARGON2_PWD_TOO_LONG = -5,
ARGON2_SALT_TOO_SHORT = -6,
ARGON2_SALT_TOO_LONG = -7,
ARGON2_AD_TOO_SHORT = -8,
ARGON2_AD_TOO_LONG = -9,
ARGON2_SECRET_TOO_SHORT = -10,
ARGON2_SECRET_TOO_LONG = -11,
ARGON2_TIME_TOO_SMALL = -12,
ARGON2_TIME_TOO_LARGE = -13,
ARGON2_MEMORY_TOO_LITTLE = -14,
ARGON2_MEMORY_TOO_MUCH = -15,
ARGON2_LANES_TOO_FEW = -16,
ARGON2_LANES_TOO_MANY = -17,
ARGON2_PWD_PTR_MISMATCH = -18, /* NULL ptr with non-zero length */
ARGON2_SALT_PTR_MISMATCH = -19, /* NULL ptr with non-zero length */
ARGON2_SECRET_PTR_MISMATCH = -20, /* NULL ptr with non-zero length */
ARGON2_AD_PTR_MISMATCH = -21, /* NULL ptr with non-zero length */
ARGON2_MEMORY_ALLOCATION_ERROR = -22,
ARGON2_FREE_MEMORY_CBK_NULL = -23,
ARGON2_ALLOCATE_MEMORY_CBK_NULL = -24,
ARGON2_INCORRECT_PARAMETER = -25,
ARGON2_INCORRECT_TYPE = -26,
ARGON2_OUT_PTR_MISMATCH = -27,
ARGON2_THREADS_TOO_FEW = -28,
ARGON2_THREADS_TOO_MANY = -29,
ARGON2_MISSING_ARGS = -30,
ARGON2_ENCODING_FAIL = -31,
ARGON2_DECODING_FAIL = -32,
ARGON2_THREAD_FAIL = -33,
ARGON2_DECODING_LENGTH_FAIL = -34,
ARGON2_VERIFY_MISMATCH = -35
} argon2_error_codes;
/* Memory allocator types --- for external allocation */
typedef int (*allocate_fptr)(uint8_t **memory, size_t bytes_to_allocate);
typedef void (*deallocate_fptr)(uint8_t *memory, size_t bytes_to_allocate);
/* Argon2 external data structures */
/*
*****
* Context: structure to hold Argon2 inputs:
* output array and its length,
* password and its length,
* salt and its length,
* secret and its length,
* associated data and its length,
* number of passes, amount of used memory (in KBytes, can be rounded up a bit)
* number of parallel threads that will be run.
* All the parameters above affect the output hash value.
* Additionally, two function pointers can be provided to allocate and
* deallocate the memory (if NULL, memory will be allocated internally).
* Also, three flags indicate whether to erase password, secret as soon as they
* are pre-hashed (and thus not needed anymore), and the entire memory
*****
* Simplest situation: you have output array out[8], password is stored in
* pwd[32], salt is stored in salt[16], you do not have keys nor associated
* data. You need to spend 1 GB of RAM and you run 5 passes of Argon2d with
* 4 parallel lanes.
* You want to erase the password, but you're OK with last pass not being
* erased. You want to use the default memory allocator.
* Then you initialize:
Argon2_Context(out,8,pwd,32,salt,16,NULL,0,NULL,0,5,1<<20,4,4,NULL,NULL,true,false,false,false)
*/
typedef struct Argon2_Context {
uint8_t *out; /* output array */
uint32_t outlen; /* digest length */
uint8_t *pwd; /* password array */
uint32_t pwdlen; /* password length */
uint8_t *salt; /* salt array */
uint32_t saltlen; /* salt length */
uint8_t *secret; /* key array */
uint32_t secretlen; /* key length */
uint8_t *ad; /* associated data array */
uint32_t adlen; /* associated data length */
uint32_t t_cost; /* number of passes */
uint32_t m_cost; /* amount of memory requested (KB) */
uint32_t lanes; /* number of lanes */
uint32_t threads; /* maximum number of threads */
uint32_t version; /* version number */
allocate_fptr allocate_cbk; /* pointer to memory allocator */
deallocate_fptr free_cbk; /* pointer to memory deallocator */
uint32_t flags; /* array of bool options */
} argon2_context;
/* Argon2 primitive type */
typedef enum Argon2_type {
Argon2_d = 0,
Argon2_i = 1,
Argon2_id = 2
} argon2_type;
/* Version of the algorithm */
typedef enum Argon2_version {
ARGON2_VERSION_10 = 0x10,
ARGON2_VERSION_13 = 0x13,
ARGON2_VERSION_NUMBER = ARGON2_VERSION_13
} argon2_version;
/*
* Function that gives the string representation of an argon2_type.
* @param type The argon2_type that we want the string for
* @param uppercase Whether the string should have the first letter uppercase
* @return NULL if invalid type, otherwise the string representation.
*/
ARGON2_PUBLIC const char *argon2_type2string(argon2_type type, int uppercase);
/*
* Function that performs memory-hard hashing with certain degree of parallelism
* @param context Pointer to the Argon2 internal structure
* @return Error code if smth is wrong, ARGON2_OK otherwise
*/
ARGON2_PUBLIC int argon2_ctx(argon2_context *context, argon2_type type);
/**
* Hashes a password with Argon2i, producing an encoded hash
* @param t_cost Number of iterations
* @param m_cost Sets memory usage to m_cost kibibytes
* @param parallelism Number of threads and compute lanes
* @param pwd Pointer to password
* @param pwdlen Password size in bytes
* @param salt Pointer to salt
* @param saltlen Salt size in bytes
* @param hashlen Desired length of the hash in bytes
* @param encoded Buffer where to write the encoded hash
* @param encodedlen Size of the buffer (thus max size of the encoded hash)
* @pre Different parallelism levels will give different results
* @pre Returns ARGON2_OK if successful
*/
ARGON2_PUBLIC int argon2i_hash_encoded(const uint32_t t_cost,
const uint32_t m_cost,
const uint32_t parallelism,
const void *pwd, const size_t pwdlen,
const void *salt, const size_t saltlen,
const size_t hashlen, char *encoded,
const size_t encodedlen);
/**
* Hashes a password with Argon2i, producing a raw hash by allocating memory at
* @hash
* @param t_cost Number of iterations
* @param m_cost Sets memory usage to m_cost kibibytes
* @param parallelism Number of threads and compute lanes
* @param pwd Pointer to password
* @param pwdlen Password size in bytes
* @param salt Pointer to salt
* @param saltlen Salt size in bytes
* @param hash Buffer where to write the raw hash - updated by the function
* @param hashlen Desired length of the hash in bytes
* @pre Different parallelism levels will give different results
* @pre Returns ARGON2_OK if successful
*/
ARGON2_PUBLIC int argon2i_hash_raw(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash,
const size_t hashlen);
ARGON2_PUBLIC int argon2d_hash_encoded(const uint32_t t_cost,
const uint32_t m_cost,
const uint32_t parallelism,
const void *pwd, const size_t pwdlen,
const void *salt, const size_t saltlen,
const size_t hashlen, char *encoded,
const size_t encodedlen);
ARGON2_PUBLIC int argon2d_hash_raw(const uint32_t t_cost,
const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash,
const size_t hashlen);
ARGON2_PUBLIC int argon2id_hash_encoded(const uint32_t t_cost,
const uint32_t m_cost,
const uint32_t parallelism,
const void *pwd, const size_t pwdlen,
const void *salt, const size_t saltlen,
const size_t hashlen, char *encoded,
const size_t encodedlen);
ARGON2_PUBLIC int argon2id_hash_raw(const uint32_t t_cost,
const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash,
const size_t hashlen);
ARGON2_PUBLIC int argon2id_hash_raw_ex(const uint32_t t_cost,
const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash,
const size_t hashlen,
void *memory);
/* generic function underlying the above ones */
ARGON2_PUBLIC int argon2_hash(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash,
const size_t hashlen, char *encoded,
const size_t encodedlen, argon2_type type,
const uint32_t version);
/**
* Verifies a password against an encoded string
* Encoded string is restricted as in validate_inputs()
* @param encoded String encoding parameters, salt, hash
* @param pwd Pointer to password
* @pre Returns ARGON2_OK if successful
*/
ARGON2_PUBLIC int argon2i_verify(const char *encoded, const void *pwd,
const size_t pwdlen);
ARGON2_PUBLIC int argon2d_verify(const char *encoded, const void *pwd,
const size_t pwdlen);
ARGON2_PUBLIC int argon2id_verify(const char *encoded, const void *pwd,
const size_t pwdlen);
/* generic function underlying the above ones */
ARGON2_PUBLIC int argon2_verify(const char *encoded, const void *pwd,
const size_t pwdlen, argon2_type type);
/**
* Argon2d: Version of Argon2 that picks memory blocks depending
* on the password and salt. Only for side-channel-free
* environment!!
*****
* @param context Pointer to current Argon2 context
* @return Zero if successful, a non zero error code otherwise
*/
ARGON2_PUBLIC int argon2d_ctx(argon2_context *context);
/**
* Argon2i: Version of Argon2 that picks memory blocks
* independent on the password and salt. Good for side-channels,
* but worse w.r.t. tradeoff attacks if only one pass is used.
*****
* @param context Pointer to current Argon2 context
* @return Zero if successful, a non zero error code otherwise
*/
ARGON2_PUBLIC int argon2i_ctx(argon2_context *context);
/**
* Argon2id: Version of Argon2 where the first half-pass over memory is
* password-independent, the rest are password-dependent (on the password and
* salt). OK against side channels (they reduce to 1/2-pass Argon2i), and
* better with w.r.t. tradeoff attacks (similar to Argon2d).
*****
* @param context Pointer to current Argon2 context
* @return Zero if successful, a non zero error code otherwise
*/
ARGON2_PUBLIC int argon2id_ctx(argon2_context *context);
/**
* Verify if a given password is correct for Argon2d hashing
* @param context Pointer to current Argon2 context
* @param hash The password hash to verify. The length of the hash is
* specified by the context outlen member
* @return Zero if successful, a non zero error code otherwise
*/
ARGON2_PUBLIC int argon2d_verify_ctx(argon2_context *context, const char *hash);
/**
* Verify if a given password is correct for Argon2i hashing
* @param context Pointer to current Argon2 context
* @param hash The password hash to verify. The length of the hash is
* specified by the context outlen member
* @return Zero if successful, a non zero error code otherwise
*/
ARGON2_PUBLIC int argon2i_verify_ctx(argon2_context *context, const char *hash);
/**
* Verify if a given password is correct for Argon2id hashing
* @param context Pointer to current Argon2 context
* @param hash The password hash to verify. The length of the hash is
* specified by the context outlen member
* @return Zero if successful, a non zero error code otherwise
*/
ARGON2_PUBLIC int argon2id_verify_ctx(argon2_context *context,
const char *hash);
/* generic function underlying the above ones */
ARGON2_PUBLIC int argon2_verify_ctx(argon2_context *context, const char *hash,
argon2_type type);
/**
* Get the associated error message for given error code
* @return The error message associated with the given error code
*/
ARGON2_PUBLIC const char *argon2_error_message(int error_code);
/**
* Returns the encoded hash length for the given input parameters
* @param t_cost Number of iterations
* @param m_cost Memory usage in kibibytes
* @param parallelism Number of threads; used to compute lanes
* @param saltlen Salt size in bytes
* @param hashlen Hash size in bytes
* @param type The argon2_type that we want the encoded length for
* @return The encoded hash length in bytes
*/
ARGON2_PUBLIC size_t argon2_encodedlen(uint32_t t_cost, uint32_t m_cost,
uint32_t parallelism, uint32_t saltlen,
uint32_t hashlen, argon2_type type);
/* signals availability of argon2_select_impl: */
#define ARGON2_SELECTABLE_IMPL
/**
* Selects the fastest available optimized implementation.
* @param out The file for debug output (e. g. stderr; pass NULL for no
* debug output)
* @param prefix What to print before each line; NULL is equivalent to empty
* string
*/
ARGON2_PUBLIC void argon2_select_impl();
ARGON2_PUBLIC const char *argon2_get_impl_name();
ARGON2_PUBLIC int argon2_select_impl_by_name(const char *name);
/* signals support for passing preallocated memory: */
#define ARGON2_PREALLOCATED_MEMORY
ARGON2_PUBLIC size_t argon2_memory_size(uint32_t m_cost, uint32_t parallelism);
/**
* Function that performs memory-hard hashing with certain degree of parallelism
* @param context Pointer to the Argon2 internal structure
* @param type The Argon2 type
* @param memory Preallocated memory for blocks (or NULL)
* @param memory_size The size of preallocated memory
* @return Error code if smth is wrong, ARGON2_OK otherwise
*/
ARGON2_PUBLIC int argon2_ctx_mem(argon2_context *context, argon2_type type,
void *memory, size_t memory_size);
#if defined(__cplusplus)
}
#endif
#endif

193
src/3rdparty/argon2/lib/argon2-template-64.h vendored Normal file
View file

@ -0,0 +1,193 @@
#include <string.h>
#include "core.h"
#define MASK_32 UINT64_C(0xFFFFFFFF)
#define F(x, y) ((x) + (y) + 2 * ((x) & MASK_32) * ((y) & MASK_32))
#define G(a, b, c, d) \
do { \
a = F(a, b); \
d = rotr64(d ^ a, 32); \
c = F(c, d); \
b = rotr64(b ^ c, 24); \
a = F(a, b); \
d = rotr64(d ^ a, 16); \
c = F(c, d); \
b = rotr64(b ^ c, 63); \
} while ((void)0, 0)
#define BLAKE2_ROUND_NOMSG(v0, v1, v2, v3, v4, v5, v6, v7, \
v8, v9, v10, v11, v12, v13, v14, v15) \
do { \
G(v0, v4, v8, v12); \
G(v1, v5, v9, v13); \
G(v2, v6, v10, v14); \
G(v3, v7, v11, v15); \
G(v0, v5, v10, v15); \
G(v1, v6, v11, v12); \
G(v2, v7, v8, v13); \
G(v3, v4, v9, v14); \
} while ((void)0, 0)
#define BLAKE2_ROUND_NOMSG1(v) \
BLAKE2_ROUND_NOMSG( \
(v)[ 0], (v)[ 1], (v)[ 2], (v)[ 3], \
(v)[ 4], (v)[ 5], (v)[ 6], (v)[ 7], \
(v)[ 8], (v)[ 9], (v)[10], (v)[11], \
(v)[12], (v)[13], (v)[14], (v)[15])
#define BLAKE2_ROUND_NOMSG2(v) \
BLAKE2_ROUND_NOMSG( \
(v)[ 0], (v)[ 1], (v)[ 16], (v)[ 17], \
(v)[ 32], (v)[ 33], (v)[ 48], (v)[ 49], \
(v)[ 64], (v)[ 65], (v)[ 80], (v)[ 81], \
(v)[ 96], (v)[ 97], (v)[112], (v)[113])
static void fill_block(const block *prev_block, const block *ref_block,
block *next_block, int with_xor)
{
block blockR, block_tmp;
copy_block(&blockR, ref_block);
xor_block(&blockR, prev_block);
copy_block(&block_tmp, &blockR);
if (with_xor) {
xor_block(&block_tmp, next_block);
}
/* Apply Blake2 on columns of 64-bit words: (0,1,...,15) , then
(16,17,..31)... finally (112,113,...127) */
BLAKE2_ROUND_NOMSG1(blockR.v + 0 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 1 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 2 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 3 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 4 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 5 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 6 * 16);
BLAKE2_ROUND_NOMSG1(blockR.v + 7 * 16);
/* Apply Blake2 on rows of 64-bit words: (0,1,16,17,...112,113), then
(2,3,18,19,...,114,115).. finally (14,15,30,31,...,126,127) */
BLAKE2_ROUND_NOMSG2(blockR.v + 0 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 1 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 2 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 3 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 4 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 5 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 6 * 2);
BLAKE2_ROUND_NOMSG2(blockR.v + 7 * 2);
copy_block(next_block, &block_tmp);
xor_block(next_block, &blockR);
}
static void next_addresses(block *address_block, block *input_block,
const block *zero_block)
{
input_block->v[6]++;
fill_block(zero_block, input_block, address_block, 0);
fill_block(zero_block, address_block, address_block, 0);
}
static void fill_segment_64(const argon2_instance_t *instance,
argon2_position_t position)
{
block *ref_block, *curr_block, *prev_block;
block address_block, input_block, zero_block;
uint64_t pseudo_rand, ref_index, ref_lane;
uint32_t prev_offset, curr_offset;
uint32_t starting_index, i;
int data_independent_addressing;
if (instance == NULL) {
return;
}
data_independent_addressing = (instance->type == Argon2_i) ||
(instance->type == Argon2_id && (position.pass == 0) &&
(position.slice < ARGON2_SYNC_POINTS / 2));
if (data_independent_addressing) {
init_block_value(&zero_block, 0);
init_block_value(&input_block, 0);
input_block.v[0] = position.pass;
input_block.v[1] = position.lane;
input_block.v[2] = position.slice;
input_block.v[3] = instance->memory_blocks;
input_block.v[4] = instance->passes;
input_block.v[5] = instance->type;
}
starting_index = 0;
if ((0 == position.pass) && (0 == position.slice)) {
starting_index = 2; /* we have already generated the first two blocks */
/* Don't forget to generate the first block of addresses: */
if (data_independent_addressing) {
next_addresses(&address_block, &input_block, &zero_block);
}
}
/* Offset of the current block */
curr_offset = position.lane * instance->lane_length +
position.slice * instance->segment_length + starting_index;
if (0 == curr_offset % instance->lane_length) {
/* Last block in this lane */
prev_offset = curr_offset + instance->lane_length - 1;
} else {
/* Previous block */
prev_offset = curr_offset - 1;
}
for (i = starting_index; i < instance->segment_length;
++i, ++curr_offset, ++prev_offset) {
/*1.1 Rotating prev_offset if needed */
if (curr_offset % instance->lane_length == 1) {
prev_offset = curr_offset - 1;
}
/* 1.2 Computing the index of the reference block */
/* 1.2.1 Taking pseudo-random value from the previous block */
if (data_independent_addressing) {
if (i % ARGON2_ADDRESSES_IN_BLOCK == 0) {
next_addresses(&address_block, &input_block, &zero_block);
}
pseudo_rand = address_block.v[i % ARGON2_ADDRESSES_IN_BLOCK];
} else {
pseudo_rand = instance->memory[prev_offset].v[0];
}
/* 1.2.2 Computing the lane of the reference block */
ref_lane = ((pseudo_rand >> 32)) % instance->lanes;
if ((position.pass == 0) && (position.slice == 0)) {
/* Can not reference other lanes yet */
ref_lane = position.lane;
}
/* 1.2.3 Computing the number of possible reference block within the
* lane.
*/
position.index = i;
ref_index = index_alpha(instance, &position, pseudo_rand & 0xFFFFFFFF,
ref_lane == position.lane);
/* 2 Creating a new block */
ref_block =
instance->memory + instance->lane_length * ref_lane + ref_index;
curr_block = instance->memory + curr_offset;
prev_block = instance->memory + prev_offset;
/* version 1.2.1 and earlier: overwrite, not XOR */
if (0 == position.pass || ARGON2_VERSION_10 == instance->version) {
fill_block(prev_block, ref_block, curr_block, 0);
} else {
fill_block(prev_block, ref_block, curr_block, 1);
}
}
}

504
src/3rdparty/argon2/lib/argon2.c vendored Normal file
View file

@ -0,0 +1,504 @@
/*
* Argon2 source code package
*
* Written by Daniel Dinu and Dmitry Khovratovich, 2015
*
* This work is licensed under a Creative Commons CC0 1.0 License/Waiver.
*
* 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/>.
*/
#include <string.h>
#include <stdlib.h>
#include <stdio.h>
#include "argon2.h"
#include "encoding.h"
#include "core.h"
const char *argon2_type2string(argon2_type type, int uppercase) {
switch (type) {
case Argon2_d:
return uppercase ? "Argon2d" : "argon2d";
case Argon2_i:
return uppercase ? "Argon2i" : "argon2i";
case Argon2_id:
return uppercase ? "Argon2id" : "argon2id";
}
return NULL;
}
static void argon2_compute_memory_blocks(uint32_t *memory_blocks,
uint32_t *segment_length,
uint32_t m_cost, uint32_t lanes)
{
/* Minimum memory_blocks = 8L blocks, where L is the number of lanes */
*memory_blocks = m_cost;
if (*memory_blocks < 2 * ARGON2_SYNC_POINTS * lanes) {
*memory_blocks = 2 * ARGON2_SYNC_POINTS * lanes;
}
*segment_length = *memory_blocks / (lanes * ARGON2_SYNC_POINTS);
/* Ensure that all segments have equal length */
*memory_blocks = *segment_length * (lanes * ARGON2_SYNC_POINTS);
}
size_t argon2_memory_size(uint32_t m_cost, uint32_t parallelism) {
uint32_t memory_blocks, segment_length;
argon2_compute_memory_blocks(&memory_blocks, &segment_length, m_cost,
parallelism);
return memory_blocks * ARGON2_BLOCK_SIZE;
}
int argon2_ctx_mem(argon2_context *context, argon2_type type, void *memory,
size_t memory_size) {
/* 1. Validate all inputs */
int result = validate_inputs(context);
uint32_t memory_blocks, segment_length;
argon2_instance_t instance;
if (ARGON2_OK != result) {
return result;
}
if (Argon2_d != type && Argon2_i != type && Argon2_id != type) {
return ARGON2_INCORRECT_TYPE;
}
/* 2. Align memory size */
argon2_compute_memory_blocks(&memory_blocks, &segment_length,
context->m_cost, context->lanes);
/* check for sufficient memory size: */
if (memory != NULL && (memory_size % ARGON2_BLOCK_SIZE != 0 ||
memory_size / ARGON2_BLOCK_SIZE < memory_blocks)) {
return ARGON2_MEMORY_ALLOCATION_ERROR;
}
instance.version = context->version;
instance.memory = (block *)memory;
instance.passes = context->t_cost;
instance.memory_blocks = memory_blocks;
instance.segment_length = segment_length;
instance.lane_length = segment_length * ARGON2_SYNC_POINTS;
instance.lanes = context->lanes;
instance.threads = context->threads;
instance.type = type;
instance.print_internals = !!(context->flags & ARGON2_FLAG_GENKAT);
instance.keep_memory = memory != NULL;
if (instance.threads > instance.lanes) {
instance.threads = instance.lanes;
}
/* 3. Initialization: Hashing inputs, allocating memory, filling first
* blocks
*/
result = initialize(&instance, context);
if (ARGON2_OK != result) {
return result;
}
/* 4. Filling memory */
result = fill_memory_blocks(&instance);
if (ARGON2_OK != result) {
return result;
}
/* 5. Finalization */
finalize(context, &instance);
return ARGON2_OK;
}
int argon2_ctx(argon2_context *context, argon2_type type) {
return argon2_ctx_mem(context, type, NULL, 0);
}
int argon2_hash(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt, const size_t saltlen,
void *hash, const size_t hashlen, char *encoded,
const size_t encodedlen, argon2_type type,
const uint32_t version){
argon2_context context;
int result;
uint8_t *out;
if (pwdlen > ARGON2_MAX_PWD_LENGTH) {
return ARGON2_PWD_TOO_LONG;
}
if (saltlen > ARGON2_MAX_SALT_LENGTH) {
return ARGON2_SALT_TOO_LONG;
}
if (hashlen > ARGON2_MAX_OUTLEN) {
return ARGON2_OUTPUT_TOO_LONG;
}
if (hashlen < ARGON2_MIN_OUTLEN) {
return ARGON2_OUTPUT_TOO_SHORT;
}
out = malloc(hashlen);
if (!out) {
return ARGON2_MEMORY_ALLOCATION_ERROR;
}
context.out = (uint8_t *)out;
context.outlen = (uint32_t)hashlen;
context.pwd = CONST_CAST(uint8_t *)pwd;
context.pwdlen = (uint32_t)pwdlen;
context.salt = CONST_CAST(uint8_t *)salt;
context.saltlen = (uint32_t)saltlen;
context.secret = NULL;
context.secretlen = 0;
context.ad = NULL;
context.adlen = 0;
context.t_cost = t_cost;
context.m_cost = m_cost;
context.lanes = parallelism;
context.threads = parallelism;
context.allocate_cbk = NULL;
context.free_cbk = NULL;
context.flags = ARGON2_DEFAULT_FLAGS;
context.version = version;
result = argon2_ctx(&context, type);
if (result != ARGON2_OK) {
clear_internal_memory(out, hashlen);
free(out);
return result;
}
/* if raw hash requested, write it */
if (hash) {
memcpy(hash, out, hashlen);
}
/* if encoding requested, write it */
if (encoded && encodedlen) {
if (encode_string(encoded, encodedlen, &context, type) != ARGON2_OK) {
clear_internal_memory(out, hashlen); /* wipe buffers if error */
clear_internal_memory(encoded, encodedlen);
free(out);
return ARGON2_ENCODING_FAIL;
}
}
clear_internal_memory(out, hashlen);
free(out);
return ARGON2_OK;
}
int argon2i_hash_encoded(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, const size_t hashlen,
char *encoded, const size_t encodedlen) {
return argon2_hash(t_cost, m_cost, parallelism, pwd, pwdlen, salt, saltlen,
NULL, hashlen, encoded, encodedlen, Argon2_i,
ARGON2_VERSION_NUMBER);
}
int argon2i_hash_raw(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash, const size_t hashlen) {
return argon2_hash(t_cost, m_cost, parallelism, pwd, pwdlen, salt, saltlen,
hash, hashlen, NULL, 0, Argon2_i, ARGON2_VERSION_NUMBER);
}
int argon2d_hash_encoded(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, const size_t hashlen,
char *encoded, const size_t encodedlen) {
return argon2_hash(t_cost, m_cost, parallelism, pwd, pwdlen, salt, saltlen,
NULL, hashlen, encoded, encodedlen, Argon2_d,
ARGON2_VERSION_NUMBER);
}
int argon2d_hash_raw(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash, const size_t hashlen) {
return argon2_hash(t_cost, m_cost, parallelism, pwd, pwdlen, salt, saltlen,
hash, hashlen, NULL, 0, Argon2_d, ARGON2_VERSION_NUMBER);
}
int argon2id_hash_encoded(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, const size_t hashlen,
char *encoded, const size_t encodedlen) {
return argon2_hash(t_cost, m_cost, parallelism, pwd, pwdlen, salt, saltlen,
NULL, hashlen, encoded, encodedlen, Argon2_id,
ARGON2_VERSION_NUMBER);
}
int argon2id_hash_raw(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash, const size_t hashlen) {
return argon2_hash(t_cost, m_cost, parallelism, pwd, pwdlen, salt, saltlen,
hash, hashlen, NULL, 0, Argon2_id,
ARGON2_VERSION_NUMBER);
}
int argon2id_hash_raw_ex(const uint32_t t_cost, const uint32_t m_cost,
const uint32_t parallelism, const void *pwd,
const size_t pwdlen, const void *salt,
const size_t saltlen, void *hash, const size_t hashlen, void *memory) {
argon2_context context;
context.out = (uint8_t *)hash;
context.outlen = (uint32_t)hashlen;
context.pwd = CONST_CAST(uint8_t *)pwd;
context.pwdlen = (uint32_t)pwdlen;
context.salt = CONST_CAST(uint8_t *)salt;
context.saltlen = (uint32_t)saltlen;
context.secret = NULL;
context.secretlen = 0;
context.ad = NULL;
context.adlen = 0;
context.t_cost = t_cost;
context.m_cost = m_cost;
context.lanes = parallelism;
context.threads = parallelism;
context.allocate_cbk = NULL;
context.free_cbk = NULL;
context.flags = ARGON2_DEFAULT_FLAGS;
context.version = ARGON2_VERSION_NUMBER;
return argon2_ctx_mem(&context, Argon2_id, memory, m_cost * 1024);
}
static int argon2_compare(const uint8_t *b1, const uint8_t *b2, size_t len) {
size_t i;
uint8_t d = 0U;
for (i = 0U; i < len; i++) {
d |= b1[i] ^ b2[i];
}
return (int)((1 & ((d - 1) >> 8)) - 1);
}
int argon2_verify(const char *encoded, const void *pwd, const size_t pwdlen,
argon2_type type) {
argon2_context ctx;
uint8_t *desired_result = NULL;
int ret = ARGON2_OK;
size_t encoded_len;
uint32_t max_field_len;
if (pwdlen > ARGON2_MAX_PWD_LENGTH) {
return ARGON2_PWD_TOO_LONG;
}
if (encoded == NULL) {
return ARGON2_DECODING_FAIL;
}
encoded_len = strlen(encoded);
if (encoded_len > UINT32_MAX) {
return ARGON2_DECODING_FAIL;
}
/* No field can be longer than the encoded length */
max_field_len = (uint32_t)encoded_len;
ctx.saltlen = max_field_len;
ctx.outlen = max_field_len;
ctx.salt = malloc(ctx.saltlen);
ctx.out = malloc(ctx.outlen);
if (!ctx.salt || !ctx.out) {
ret = ARGON2_MEMORY_ALLOCATION_ERROR;
goto fail;
}
ctx.pwd = (uint8_t *)pwd;
ctx.pwdlen = (uint32_t)pwdlen;
ret = decode_string(&ctx, encoded, type);
if (ret != ARGON2_OK) {
goto fail;
}
/* Set aside the desired result, and get a new buffer. */
desired_result = ctx.out;
ctx.out = malloc(ctx.outlen);
if (!ctx.out) {
ret = ARGON2_MEMORY_ALLOCATION_ERROR;
goto fail;
}
ret = argon2_verify_ctx(&ctx, (char *)desired_result, type);
if (ret != ARGON2_OK) {
goto fail;
}
fail:
free(ctx.salt);
free(ctx.out);
free(desired_result);
return ret;
}
int argon2i_verify(const char *encoded, const void *pwd, const size_t pwdlen) {
return argon2_verify(encoded, pwd, pwdlen, Argon2_i);
}
int argon2d_verify(const char *encoded, const void *pwd, const size_t pwdlen) {
return argon2_verify(encoded, pwd, pwdlen, Argon2_d);
}
int argon2id_verify(const char *encoded, const void *pwd, const size_t pwdlen) {
return argon2_verify(encoded, pwd, pwdlen, Argon2_id);
}
int argon2d_ctx(argon2_context *context) {
return argon2_ctx(context, Argon2_d);
}
int argon2i_ctx(argon2_context *context) {
return argon2_ctx(context, Argon2_i);
}
int argon2id_ctx(argon2_context *context) {
return argon2_ctx(context, Argon2_id);
}
int argon2_verify_ctx(argon2_context *context, const char *hash,
argon2_type type) {
int ret = argon2_ctx(context, type);
if (ret != ARGON2_OK) {
return ret;
}
if (argon2_compare((uint8_t *)hash, context->out, context->outlen)) {
return ARGON2_VERIFY_MISMATCH;
}
return ARGON2_OK;
}
int argon2d_verify_ctx(argon2_context *context, const char *hash) {
return argon2_verify_ctx(context, hash, Argon2_d);
}
int argon2i_verify_ctx(argon2_context *context, const char *hash) {
return argon2_verify_ctx(context, hash, Argon2_i);
}
int argon2id_verify_ctx(argon2_context *context, const char *hash) {
return argon2_verify_ctx(context, hash, Argon2_id);
}
const char *argon2_error_message(int error_code) {
switch (error_code) {
case ARGON2_OK:
return "OK";
case ARGON2_OUTPUT_PTR_NULL:
return "Output pointer is NULL";
case ARGON2_OUTPUT_TOO_SHORT:
return "Output is too short";
case ARGON2_OUTPUT_TOO_LONG:
return "Output is too long";
case ARGON2_PWD_TOO_SHORT:
return "Password is too short";
case ARGON2_PWD_TOO_LONG:
return "Password is too long";
case ARGON2_SALT_TOO_SHORT:
return "Salt is too short";
case ARGON2_SALT_TOO_LONG:
return "Salt is too long";
case ARGON2_AD_TOO_SHORT:
return "Associated data is too short";
case ARGON2_AD_TOO_LONG:
return "Associated data is too long";
case ARGON2_SECRET_TOO_SHORT:
return "Secret is too short";
case ARGON2_SECRET_TOO_LONG:
return "Secret is too long";
case ARGON2_TIME_TOO_SMALL:
return "Time cost is too small";
case ARGON2_TIME_TOO_LARGE:
return "Time cost is too large";
case ARGON2_MEMORY_TOO_LITTLE:
return "Memory cost is too small";
case ARGON2_MEMORY_TOO_MUCH:
return "Memory cost is too large";
case ARGON2_LANES_TOO_FEW:
return "Too few lanes";
case ARGON2_LANES_TOO_MANY:
return "Too many lanes";
case ARGON2_PWD_PTR_MISMATCH:
return "Password pointer is NULL, but password length is not 0";
case ARGON2_SALT_PTR_MISMATCH:
return "Salt pointer is NULL, but salt length is not 0";
case ARGON2_SECRET_PTR_MISMATCH:
return "Secret pointer is NULL, but secret length is not 0";
case ARGON2_AD_PTR_MISMATCH:
return "Associated data pointer is NULL, but ad length is not 0";
case ARGON2_MEMORY_ALLOCATION_ERROR:
return "Memory allocation error";
case ARGON2_FREE_MEMORY_CBK_NULL:
return "The free memory callback is NULL";
case ARGON2_ALLOCATE_MEMORY_CBK_NULL:
return "The allocate memory callback is NULL";
case ARGON2_INCORRECT_PARAMETER:
return "Argon2_Context context is NULL";
case ARGON2_INCORRECT_TYPE:
return "There is no such version of Argon2";
case ARGON2_OUT_PTR_MISMATCH:
return "Output pointer mismatch";
case ARGON2_THREADS_TOO_FEW:
return "Not enough threads";
case ARGON2_THREADS_TOO_MANY:
return "Too many threads";
case ARGON2_MISSING_ARGS:
return "Missing arguments";
case ARGON2_ENCODING_FAIL:
return "Encoding failed";
case ARGON2_DECODING_FAIL:
return "Decoding failed";
case ARGON2_THREAD_FAIL:
return "Threading failure";
case ARGON2_DECODING_LENGTH_FAIL:
return "Some of encoded parameters are too long or too short";
case ARGON2_VERIFY_MISMATCH:
return "The password does not match the supplied hash";
default:
return "Unknown error code";
}
}
size_t argon2_encodedlen(uint32_t t_cost, uint32_t m_cost, uint32_t parallelism,
uint32_t saltlen, uint32_t hashlen, argon2_type type) {
return strlen("$$v=$m=,t=,p=$$") + strlen(argon2_type2string(type, 0)) +
numlen(t_cost) + numlen(m_cost) + numlen(parallelism) +
b64len(saltlen) + b64len(hashlen) + numlen(ARGON2_VERSION_NUMBER) +
1;
}

90
src/3rdparty/argon2/lib/blake2/blake2-impl.h vendored Normal file
View file

@ -0,0 +1,90 @@
#ifndef ARGON2_BLAKE2_IMPL_H
#define ARGON2_BLAKE2_IMPL_H
#include <stdint.h>
/* Argon2 Team - Begin Code */
/*
Not an exhaustive list, but should cover the majority of modern platforms
Additionally, the code will always be correct---this is only a performance
tweak.
*/
#if (defined(__BYTE_ORDER__) && \
(__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)) || \
defined(__LITTLE_ENDIAN__) || defined(__ARMEL__) || defined(__MIPSEL__) || \
defined(__AARCH64EL__) || defined(__amd64__) || defined(__i386__) || \
defined(_M_IX86) || defined(_M_X64) || defined(_M_AMD64) || \
defined(_M_ARM)
#define NATIVE_LITTLE_ENDIAN
#endif
/* Argon2 Team - End Code */
static inline uint32_t load32(const void *src) {
#if defined(NATIVE_LITTLE_ENDIAN)
return *(const uint32_t *)src;
#else
const uint8_t *p = (const uint8_t *)src;
uint32_t w = *p++;
w |= (uint32_t)(*p++) << 8;
w |= (uint32_t)(*p++) << 16;
w |= (uint32_t)(*p++) << 24;
return w;
#endif
}
static inline uint64_t load64(const void *src) {
#if defined(NATIVE_LITTLE_ENDIAN)
return *(const uint64_t *)src;
#else
const uint8_t *p = (const uint8_t *)src;
uint64_t w = *p++;
w |= (uint64_t)(*p++) << 8;
w |= (uint64_t)(*p++) << 16;
w |= (uint64_t)(*p++) << 24;
w |= (uint64_t)(*p++) << 32;
w |= (uint64_t)(*p++) << 40;
w |= (uint64_t)(*p++) << 48;
w |= (uint64_t)(*p++) << 56;
return w;
#endif
}
static inline void store32(void *dst, uint32_t w) {
#if defined(NATIVE_LITTLE_ENDIAN)
*(uint32_t *)dst = w;
#else
uint8_t *p = (uint8_t *)dst;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
#endif
}
static inline void store64(void *dst, uint64_t w) {
#if defined(NATIVE_LITTLE_ENDIAN)
*(uint64_t *)dst = w;
#else
uint8_t *p = (uint8_t *)dst;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
w >>= 8;
*p++ = (uint8_t)w;
#endif
}
#endif // ARGON2_BLAKE2_IMPL_H

225
src/3rdparty/argon2/lib/blake2/blake2.c vendored Normal file
View file

@ -0,0 +1,225 @@
#include <string.h>
#include "blake2/blake2.h"
#include "blake2/blake2-impl.h"
#include "core.h"
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)
};
#define rotr64(x, n) (((x) >> (n)) | ((x) << (64 - (n))))
static const unsigned int blake2b_sigma[12][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},
};
#define G(m, r, i, a, b, c, d) \
do { \
a = a + b + m[blake2b_sigma[r][2 * i + 0]]; \
d = rotr64(d ^ a, 32); \
c = c + d; \
b = rotr64(b ^ c, 24); \
a = a + b + m[blake2b_sigma[r][2 * i + 1]]; \
d = rotr64(d ^ a, 16); \
c = c + d; \
b = rotr64(b ^ c, 63); \
} while ((void)0, 0)
#define ROUND(m, v, r) \
do { \
G(m, r, 0, v[0], v[4], v[ 8], v[12]); \
G(m, r, 1, v[1], v[5], v[ 9], v[13]); \
G(m, r, 2, v[2], v[6], v[10], v[14]); \
G(m, r, 3, v[3], v[7], v[11], v[15]); \
G(m, r, 4, v[0], v[5], v[10], v[15]); \
G(m, r, 5, v[1], v[6], v[11], v[12]); \
G(m, r, 6, v[2], v[7], v[ 8], v[13]); \
G(m, r, 7, v[3], v[4], v[ 9], v[14]); \
} while ((void)0, 0)
void blake2b_compress(blake2b_state *S, const void *block, uint64_t f0)
{
uint64_t m[16];
uint64_t v[16];
m[ 0] = load64((const uint64_t *)block + 0);
m[ 1] = load64((const uint64_t *)block + 1);
m[ 2] = load64((const uint64_t *)block + 2);
m[ 3] = load64((const uint64_t *)block + 3);
m[ 4] = load64((const uint64_t *)block + 4);
m[ 5] = load64((const uint64_t *)block + 5);
m[ 6] = load64((const uint64_t *)block + 6);
m[ 7] = load64((const uint64_t *)block + 7);
m[ 8] = load64((const uint64_t *)block + 8);
m[ 9] = load64((const uint64_t *)block + 9);
m[10] = load64((const uint64_t *)block + 10);
m[11] = load64((const uint64_t *)block + 11);
m[12] = load64((const uint64_t *)block + 12);
m[13] = load64((const uint64_t *)block + 13);
m[14] = load64((const uint64_t *)block + 14);
m[15] = load64((const uint64_t *)block + 15);
v[ 0] = S->h[0];
v[ 1] = S->h[1];
v[ 2] = S->h[2];
v[ 3] = S->h[3];
v[ 4] = S->h[4];
v[ 5] = S->h[5];
v[ 6] = S->h[6];
v[ 7] = S->h[7];
v[ 8] = blake2b_IV[0];
v[ 9] = blake2b_IV[1];
v[10] = blake2b_IV[2];
v[11] = blake2b_IV[3];
v[12] = blake2b_IV[4] ^ S->t[0];
v[13] = blake2b_IV[5] ^ S->t[1];
v[14] = blake2b_IV[6] ^ f0;
v[15] = blake2b_IV[7];
ROUND(m, v, 0);
ROUND(m, v, 1);
ROUND(m, v, 2);
ROUND(m, v, 3);
ROUND(m, v, 4);
ROUND(m, v, 5);
ROUND(m, v, 6);
ROUND(m, v, 7);
ROUND(m, v, 8);
ROUND(m, v, 9);
ROUND(m, v, 10);
ROUND(m, v, 11);
S->h[0] ^= v[0] ^ v[ 8];
S->h[1] ^= v[1] ^ v[ 9];
S->h[2] ^= v[2] ^ v[10];
S->h[3] ^= v[3] ^ v[11];
S->h[4] ^= v[4] ^ v[12];
S->h[5] ^= v[5] ^ v[13];
S->h[6] ^= v[6] ^ v[14];
S->h[7] ^= v[7] ^ v[15];
}
static void blake2b_increment_counter(blake2b_state *S, uint64_t inc)
{
S->t[0] += inc;
S->t[1] += (S->t[0] < inc);
}
static void blake2b_init_state(blake2b_state *S)
{
memcpy(S->h, blake2b_IV, sizeof(S->h));
S->t[1] = S->t[0] = 0;
S->buflen = 0;
}
void blake2b_init(blake2b_state *S, size_t outlen)
{
blake2b_init_state(S);
/* XOR initial state with param block: */
S->h[0] ^= (uint64_t)outlen | (UINT64_C(1) << 16) | (UINT64_C(1) << 24);
}
void blake2b_update(blake2b_state *S, const void *in, size_t inlen)
{
const uint8_t *pin = (const uint8_t *)in;
if (S->buflen + inlen > BLAKE2B_BLOCKBYTES) {
size_t left = S->buflen;
size_t fill = BLAKE2B_BLOCKBYTES - left;
memcpy(&S->buf[left], pin, fill);
blake2b_increment_counter(S, BLAKE2B_BLOCKBYTES);
blake2b_compress(S, S->buf, 0);
S->buflen = 0;
inlen -= fill;
pin += fill;
/* Avoid buffer copies when possible */
while (inlen > BLAKE2B_BLOCKBYTES) {
blake2b_increment_counter(S, BLAKE2B_BLOCKBYTES);
blake2b_compress(S, pin, 0);
inlen -= BLAKE2B_BLOCKBYTES;
pin += BLAKE2B_BLOCKBYTES;
}
}
memcpy(&S->buf[S->buflen], pin, inlen);
S->buflen += inlen;
}
void blake2b_final(blake2b_state *S, void *out, size_t outlen)
{
uint8_t buffer[BLAKE2B_OUTBYTES] = {0};
unsigned int i;
blake2b_increment_counter(S, S->buflen);
memset(&S->buf[S->buflen], 0, BLAKE2B_BLOCKBYTES - S->buflen); /* Padding */
blake2b_compress(S, S->buf, UINT64_C(0xFFFFFFFFFFFFFFFF));
for (i = 0; i < 8; ++i) { /* Output full hash to temp buffer */
store64(buffer + i * sizeof(uint64_t), S->h[i]);
}
memcpy(out, buffer, outlen);
clear_internal_memory(buffer, sizeof(buffer));
clear_internal_memory(S->buf, sizeof(S->buf));
clear_internal_memory(S->h, sizeof(S->h));
}
void blake2b_long(void *out, size_t outlen, const void *in, size_t inlen)
{
uint8_t *pout = (uint8_t *)out;
blake2b_state blake_state;
uint8_t outlen_bytes[sizeof(uint32_t)] = {0};
store32(outlen_bytes, (uint32_t)outlen);
if (outlen <= BLAKE2B_OUTBYTES) {
blake2b_init(&blake_state, outlen);
blake2b_update(&blake_state, outlen_bytes, sizeof(outlen_bytes));
blake2b_update(&blake_state, in, inlen);
blake2b_final(&blake_state, pout, outlen);
} else {
uint32_t toproduce;
uint8_t out_buffer[BLAKE2B_OUTBYTES];
blake2b_init(&blake_state, BLAKE2B_OUTBYTES);
blake2b_update(&blake_state, outlen_bytes, sizeof(outlen_bytes));
blake2b_update(&blake_state, in, inlen);
blake2b_final(&blake_state, out_buffer, BLAKE2B_OUTBYTES);
memcpy(pout, out_buffer, BLAKE2B_OUTBYTES / 2);
pout += BLAKE2B_OUTBYTES / 2;
toproduce = (uint32_t)outlen - BLAKE2B_OUTBYTES / 2;
while (toproduce > BLAKE2B_OUTBYTES) {
blake2b_init(&blake_state, BLAKE2B_OUTBYTES);
blake2b_update(&blake_state, out_buffer, BLAKE2B_OUTBYTES);
blake2b_final(&blake_state, out_buffer, BLAKE2B_OUTBYTES);
memcpy(pout, out_buffer, BLAKE2B_OUTBYTES / 2);
pout += BLAKE2B_OUTBYTES / 2;
toproduce -= BLAKE2B_OUTBYTES / 2;
}
blake2b_init(&blake_state, toproduce);
blake2b_update(&blake_state, out_buffer, BLAKE2B_OUTBYTES);
blake2b_final(&blake_state, out_buffer, toproduce);
memcpy(pout, out_buffer, toproduce);
clear_internal_memory(out_buffer, sizeof(out_buffer));
}
}

30
src/3rdparty/argon2/lib/blake2/blake2.h vendored Normal file
View file

@ -0,0 +1,30 @@
#ifndef ARGON2_BLAKE2_H
#define ARGON2_BLAKE2_H
#include <stddef.h>
#include <stdint.h>
enum blake2b_constant {
BLAKE2B_BLOCKBYTES = 128,
BLAKE2B_OUTBYTES = 64,
BLAKE2B_KEYBYTES = 64,
BLAKE2B_SALTBYTES = 16,
BLAKE2B_PERSONALBYTES = 16
};
typedef struct __blake2b_state {
uint64_t h[8];
uint64_t t[2];
uint8_t buf[BLAKE2B_BLOCKBYTES];
size_t buflen;
} blake2b_state;
/* Streaming API */
void blake2b_init(blake2b_state *S, size_t outlen);
void blake2b_update(blake2b_state *S, const void *in, size_t inlen);
void blake2b_final(blake2b_state *S, void *out, size_t outlen);
void blake2b_long(void *out, size_t outlen, const void *in, size_t inlen);
#endif // ARGON2_BLAKE2_H

539
src/3rdparty/argon2/lib/core.c vendored Normal file
View file

@ -0,0 +1,539 @@
/*
* Argon2 source code package
*
* Written by Daniel Dinu and Dmitry Khovratovich, 2015
*
* This work is licensed under a Creative Commons CC0 1.0 License/Waiver.
*
* 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/>.
*/
/*For memory wiping*/
#ifdef _MSC_VER
#include <windows.h>
#include <winbase.h> /* For SecureZeroMemory */
#endif
#if defined __STDC_LIB_EXT1__
#define __STDC_WANT_LIB_EXT1__ 1
#endif
#define VC_GE_2005(version) (version >= 1400)
#include <inttypes.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "core.h"
#include "blake2/blake2.h"
#include "blake2/blake2-impl.h"
#include "genkat.h"
#if defined(__clang__)
#if __has_attribute(optnone)
#define NOT_OPTIMIZED __attribute__((optnone))
#endif
#elif defined(__GNUC__)
#define GCC_VERSION \
(__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__)
#if GCC_VERSION >= 40400
#define NOT_OPTIMIZED __attribute__((optimize("O0")))
#endif
#endif
#ifndef NOT_OPTIMIZED
#define NOT_OPTIMIZED
#endif
/***************Instance and Position constructors**********/
void init_block_value(block *b, uint8_t in) { memset(b->v, in, sizeof(b->v)); }
void copy_block(block *dst, const block *src) {
memcpy(dst->v, src->v, sizeof(uint64_t) * ARGON2_QWORDS_IN_BLOCK);
}
void xor_block(block *dst, const block *src) {
int i;
for (i = 0; i < ARGON2_QWORDS_IN_BLOCK; ++i) {
dst->v[i] ^= src->v[i];
}
}
static void load_block(block *dst, const void *input) {
unsigned i;
for (i = 0; i < ARGON2_QWORDS_IN_BLOCK; ++i) {
dst->v[i] = load64((const uint8_t *)input + i * sizeof(dst->v[i]));
}
}
static void store_block(void *output, const block *src) {
unsigned i;
for (i = 0; i < ARGON2_QWORDS_IN_BLOCK; ++i) {
store64((uint8_t *)output + i * sizeof(src->v[i]), src->v[i]);
}
}
/***************Memory functions*****************/
int allocate_memory(const argon2_context *context,
argon2_instance_t *instance) {
size_t blocks = instance->memory_blocks;
size_t memory_size = blocks * ARGON2_BLOCK_SIZE;
/* 0. Check for memory supplied by user: */
/* NOTE: Sufficient memory size is already checked in argon2_ctx_mem() */
if (instance->memory != NULL) {
return ARGON2_OK;
}
/* 1. Check for multiplication overflow */
if (blocks != 0 && memory_size / ARGON2_BLOCK_SIZE != blocks) {
return ARGON2_MEMORY_ALLOCATION_ERROR;
}
/* 2. Try to allocate with appropriate allocator */
if (context->allocate_cbk) {
(context->allocate_cbk)((uint8_t **)&instance->memory, memory_size);
} else {
instance->memory = malloc(memory_size);
}
if (instance->memory == NULL) {
return ARGON2_MEMORY_ALLOCATION_ERROR;
}
return ARGON2_OK;
}
void free_memory(const argon2_context *context,
const argon2_instance_t *instance) {
size_t memory_size = instance->memory_blocks * ARGON2_BLOCK_SIZE;
clear_internal_memory(instance->memory, memory_size);
if (instance->keep_memory) {
/* user-supplied memory -- do not free */
return;
}
if (context->free_cbk) {
(context->free_cbk)((uint8_t *)instance->memory, memory_size);
} else {
free(instance->memory);
}
}
void NOT_OPTIMIZED secure_wipe_memory(void *v, size_t n) {
#if defined(_MSC_VER) && VC_GE_2005(_MSC_VER)
SecureZeroMemory(v, n);
#elif defined memset_s
memset_s(v, n, 0, n);
#elif defined(__OpenBSD__)
explicit_bzero(v, n);
#else
static void *(*const volatile memset_sec)(void *, int, size_t) = &memset;
memset_sec(v, 0, n);
#endif
}
/* Memory clear flag defaults to true. */
int FLAG_clear_internal_memory = 0;
void clear_internal_memory(void *v, size_t n) {
if (FLAG_clear_internal_memory && v) {
secure_wipe_memory(v, n);
}
}
void finalize(const argon2_context *context, argon2_instance_t *instance) {
if (context != NULL && instance != NULL) {
block blockhash;
uint32_t l;
copy_block(&blockhash, instance->memory + instance->lane_length - 1);
/* XOR the last blocks */
for (l = 1; l < instance->lanes; ++l) {
uint32_t last_block_in_lane =
l * instance->lane_length + (instance->lane_length - 1);
xor_block(&blockhash, instance->memory + last_block_in_lane);
}
/* Hash the result */
{
uint8_t blockhash_bytes[ARGON2_BLOCK_SIZE];
store_block(blockhash_bytes, &blockhash);
blake2b_long(context->out, context->outlen, blockhash_bytes,
ARGON2_BLOCK_SIZE);
/* clear blockhash and blockhash_bytes */
clear_internal_memory(blockhash.v, ARGON2_BLOCK_SIZE);
clear_internal_memory(blockhash_bytes, ARGON2_BLOCK_SIZE);
}
if (instance->print_internals) {
print_tag(context->out, context->outlen);
}
free_memory(context, instance);
}
}
uint32_t index_alpha(const argon2_instance_t *instance,
const argon2_position_t *position, uint32_t pseudo_rand,
int same_lane) {
/*
* Pass 0:
* This lane : all already finished segments plus already constructed
* blocks in this segment
* Other lanes : all already finished segments
* Pass 1+:
* This lane : (SYNC_POINTS - 1) last segments plus already constructed
* blocks in this segment
* Other lanes : (SYNC_POINTS - 1) last segments
*/
uint32_t reference_area_size;
uint64_t relative_position;
uint32_t start_position, absolute_position;
if (0 == position->pass) {
/* First pass */
if (0 == position->slice) {
/* First slice */
reference_area_size =
position->index - 1; /* all but the previous */
} else {
if (same_lane) {
/* The same lane => add current segment */
reference_area_size =
position->slice * instance->segment_length +
position->index - 1;
} else {
reference_area_size =
position->slice * instance->segment_length +
((position->index == 0) ? (-1) : 0);
}
}
} else {
/* Second pass */
if (same_lane) {
reference_area_size = instance->lane_length -
instance->segment_length + position->index -
1;
} else {
reference_area_size = instance->lane_length -
instance->segment_length +
((position->index == 0) ? (-1) : 0);
}
}
/* 1.2.4. Mapping pseudo_rand to 0..<reference_area_size-1> and produce
* relative position */
relative_position = pseudo_rand;
relative_position = relative_position * relative_position >> 32;
relative_position = reference_area_size - 1 -
(reference_area_size * relative_position >> 32);
/* 1.2.5 Computing starting position */
start_position = 0;
if (0 != position->pass) {
start_position = (position->slice == ARGON2_SYNC_POINTS - 1)
? 0
: (position->slice + 1) * instance->segment_length;
}
/* 1.2.6. Computing absolute position */
absolute_position = (start_position + relative_position) %
instance->lane_length; /* absolute position */
return absolute_position;
}
/* Single-threaded version for p=1 case */
static int fill_memory_blocks_st(argon2_instance_t *instance) {
uint32_t r, s, l;
for (r = 0; r < instance->passes; ++r) {
for (s = 0; s < ARGON2_SYNC_POINTS; ++s) {
for (l = 0; l < instance->lanes; ++l) {
argon2_position_t position = { r, l, (uint8_t)s, 0 };
fill_segment(instance, position);
}
}
if (instance->print_internals) {
internal_kat(instance, r); /* Print all memory blocks */
}
}
return ARGON2_OK;
}
int fill_memory_blocks(argon2_instance_t *instance) {
if (instance == NULL || instance->lanes == 0) {
return ARGON2_INCORRECT_PARAMETER;
}
return fill_memory_blocks_st(instance);
}
int validate_inputs(const argon2_context *context) {
if (NULL == context) {
return ARGON2_INCORRECT_PARAMETER;
}
if (NULL == context->out) {
return ARGON2_OUTPUT_PTR_NULL;
}
/* Validate output length */
if (ARGON2_MIN_OUTLEN > context->outlen) {
return ARGON2_OUTPUT_TOO_SHORT;
}
if (ARGON2_MAX_OUTLEN < context->outlen) {
return ARGON2_OUTPUT_TOO_LONG;
}
/* Validate password (required param) */
if (NULL == context->pwd) {
if (0 != context->pwdlen) {
return ARGON2_PWD_PTR_MISMATCH;
}
}
if (ARGON2_MIN_PWD_LENGTH > context->pwdlen) {
return ARGON2_PWD_TOO_SHORT;
}
if (ARGON2_MAX_PWD_LENGTH < context->pwdlen) {
return ARGON2_PWD_TOO_LONG;
}
/* Validate salt (required param) */
if (NULL == context->salt) {
if (0 != context->saltlen) {
return ARGON2_SALT_PTR_MISMATCH;
}
}
if (ARGON2_MIN_SALT_LENGTH > context->saltlen) {
return ARGON2_SALT_TOO_SHORT;
}
if (ARGON2_MAX_SALT_LENGTH < context->saltlen) {
return ARGON2_SALT_TOO_LONG;
}
/* Validate secret (optional param) */
if (NULL == context->secret) {
if (0 != context->secretlen) {
return ARGON2_SECRET_PTR_MISMATCH;
}
} else {
if (ARGON2_MIN_SECRET > context->secretlen) {
return ARGON2_SECRET_TOO_SHORT;
}
if (ARGON2_MAX_SECRET < context->secretlen) {
return ARGON2_SECRET_TOO_LONG;
}
}
/* Validate associated data (optional param) */
if (NULL == context->ad) {
if (0 != context->adlen) {
return ARGON2_AD_PTR_MISMATCH;
}
} else {
if (ARGON2_MIN_AD_LENGTH > context->adlen) {
return ARGON2_AD_TOO_SHORT;
}
if (ARGON2_MAX_AD_LENGTH < context->adlen) {
return ARGON2_AD_TOO_LONG;
}
}
/* Validate memory cost */
if (ARGON2_MIN_MEMORY > context->m_cost) {
return ARGON2_MEMORY_TOO_LITTLE;
}
if (ARGON2_MAX_MEMORY < context->m_cost) {
return ARGON2_MEMORY_TOO_MUCH;
}
if (context->m_cost < 8 * context->lanes) {
return ARGON2_MEMORY_TOO_LITTLE;
}
/* Validate time cost */
if (ARGON2_MIN_TIME > context->t_cost) {
return ARGON2_TIME_TOO_SMALL;
}
if (ARGON2_MAX_TIME < context->t_cost) {
return ARGON2_TIME_TOO_LARGE;
}
/* Validate lanes */
if (ARGON2_MIN_LANES > context->lanes) {
return ARGON2_LANES_TOO_FEW;
}
if (ARGON2_MAX_LANES < context->lanes) {
return ARGON2_LANES_TOO_MANY;
}
/* Validate threads */
if (ARGON2_MIN_THREADS > context->threads) {
return ARGON2_THREADS_TOO_FEW;
}
if (ARGON2_MAX_THREADS < context->threads) {
return ARGON2_THREADS_TOO_MANY;
}
if (NULL != context->allocate_cbk && NULL == context->free_cbk) {
return ARGON2_FREE_MEMORY_CBK_NULL;
}
if (NULL == context->allocate_cbk && NULL != context->free_cbk) {
return ARGON2_ALLOCATE_MEMORY_CBK_NULL;
}
return ARGON2_OK;
}
void fill_first_blocks(uint8_t *blockhash, const argon2_instance_t *instance) {
uint32_t l;
/* Make the first and second block in each lane as G(H0||0||i) or
G(H0||1||i) */
uint8_t blockhash_bytes[ARGON2_BLOCK_SIZE];
for (l = 0; l < instance->lanes; ++l) {
store32(blockhash + ARGON2_PREHASH_DIGEST_LENGTH, 0);
store32(blockhash + ARGON2_PREHASH_DIGEST_LENGTH + 4, l);
blake2b_long(blockhash_bytes, ARGON2_BLOCK_SIZE, blockhash,
ARGON2_PREHASH_SEED_LENGTH);
load_block(&instance->memory[l * instance->lane_length + 0],
blockhash_bytes);
store32(blockhash + ARGON2_PREHASH_DIGEST_LENGTH, 1);
blake2b_long(blockhash_bytes, ARGON2_BLOCK_SIZE, blockhash,
ARGON2_PREHASH_SEED_LENGTH);
load_block(&instance->memory[l * instance->lane_length + 1],
blockhash_bytes);
}
clear_internal_memory(blockhash_bytes, ARGON2_BLOCK_SIZE);
}
void initial_hash(uint8_t *blockhash, argon2_context *context,
argon2_type type) {
blake2b_state BlakeHash;
uint8_t value[sizeof(uint32_t)];
if (NULL == context || NULL == blockhash) {
return;
}
blake2b_init(&BlakeHash, ARGON2_PREHASH_DIGEST_LENGTH);
store32(&value, context->lanes);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
store32(&value, context->outlen);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
store32(&value, context->m_cost);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
store32(&value, context->t_cost);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
store32(&value, context->version);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
store32(&value, (uint32_t)type);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
store32(&value, context->pwdlen);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
if (context->pwd != NULL) {
blake2b_update(&BlakeHash, (const uint8_t *)context->pwd,
context->pwdlen);
if (context->flags & ARGON2_FLAG_CLEAR_PASSWORD) {
secure_wipe_memory(context->pwd, context->pwdlen);
context->pwdlen = 0;
}
}
store32(&value, context->saltlen);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
if (context->salt != NULL) {
blake2b_update(&BlakeHash, (const uint8_t *)context->salt,
context->saltlen);
}
store32(&value, context->secretlen);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
if (context->secret != NULL) {
blake2b_update(&BlakeHash, (const uint8_t *)context->secret,
context->secretlen);
if (context->flags & ARGON2_FLAG_CLEAR_SECRET) {
secure_wipe_memory(context->secret, context->secretlen);
context->secretlen = 0;
}
}
store32(&value, context->adlen);
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
if (context->ad != NULL) {
blake2b_update(&BlakeHash, (const uint8_t *)context->ad,
context->adlen);
}
blake2b_final(&BlakeHash, blockhash, ARGON2_PREHASH_DIGEST_LENGTH);
}
int initialize(argon2_instance_t *instance, argon2_context *context) {
uint8_t blockhash[ARGON2_PREHASH_SEED_LENGTH];
int result = ARGON2_OK;
if (instance == NULL || context == NULL)
return ARGON2_INCORRECT_PARAMETER;
instance->context_ptr = context;
/* 1. Memory allocation */
result = allocate_memory(context, instance);
if (result != ARGON2_OK) {
return result;
}
/* 2. Initial hashing */
/* H_0 + 8 extra bytes to produce the first blocks */
/* uint8_t blockhash[ARGON2_PREHASH_SEED_LENGTH]; */
/* Hashing all inputs */
initial_hash(blockhash, context, instance->type);
/* Zeroing 8 extra bytes */
clear_internal_memory(blockhash + ARGON2_PREHASH_DIGEST_LENGTH,
ARGON2_PREHASH_SEED_LENGTH -
ARGON2_PREHASH_DIGEST_LENGTH);
if (instance->print_internals) {
initial_kat(blockhash, context, instance->type);
}
/* 3. Creating first blocks, we always have at least two blocks in a slice
*/
fill_first_blocks(blockhash, instance);
/* Clearing the hash */
clear_internal_memory(blockhash, ARGON2_PREHASH_SEED_LENGTH);
return ARGON2_OK;
}

226
src/3rdparty/argon2/lib/core.h vendored Normal file
View file

@ -0,0 +1,226 @@
/*
* Argon2 source code package
*
* Written by Daniel Dinu and Dmitry Khovratovich, 2015
*
* This work is licensed under a Creative Commons CC0 1.0 License/Waiver.
*
* 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/>.
*/
#ifndef ARGON2_CORE_H
#define ARGON2_CORE_H
#include "argon2.h"
#if defined(_MSC_VER)
#define ALIGN(n) __declspec(align(16))
#elif defined(__GNUC__) || defined(__clang)
#define ALIGN(x) __attribute__((__aligned__(x)))
#else
#define ALIGN(x)
#endif
#define CONST_CAST(x) (x)(uintptr_t)
/**********************Argon2 internal constants*******************************/
enum argon2_core_constants {
/* Memory block size in bytes */
ARGON2_BLOCK_SIZE = 1024,
ARGON2_QWORDS_IN_BLOCK = ARGON2_BLOCK_SIZE / 8,
ARGON2_OWORDS_IN_BLOCK = ARGON2_BLOCK_SIZE / 16,
/* Number of pseudo-random values generated by one call to Blake in Argon2i
to
generate reference block positions */
ARGON2_ADDRESSES_IN_BLOCK = 128,
/* Pre-hashing digest length and its extension*/
ARGON2_PREHASH_DIGEST_LENGTH = 64,
ARGON2_PREHASH_SEED_LENGTH = 72
};
/*************************Argon2 internal data types***********************/
/*
* Structure for the (1KB) memory block implemented as 128 64-bit words.
* Memory blocks can be copied, XORed. Internal words can be accessed by [] (no
* bounds checking).
*/
typedef struct block_ { uint64_t v[ARGON2_QWORDS_IN_BLOCK]; } block;
/*****************Functions that work with the block******************/
/* Initialize each byte of the block with @in */
void init_block_value(block *b, uint8_t in);
/* Copy block @src to block @dst */
void copy_block(block *dst, const block *src);
/* XOR @src onto @dst bytewise */
void xor_block(block *dst, const block *src);
/*
* Argon2 instance: memory pointer, number of passes, amount of memory, type,
* and derived values.
* Used to evaluate the number and location of blocks to construct in each
* thread
*/
typedef struct Argon2_instance_t {
block *memory; /* Memory pointer */
uint32_t version;
uint32_t passes; /* Number of passes */
uint32_t memory_blocks; /* Number of blocks in memory */
uint32_t segment_length;
uint32_t lane_length;
uint32_t lanes;
uint32_t threads;
argon2_type type;
int print_internals; /* whether to print the memory blocks */
int keep_memory;
argon2_context *context_ptr; /* points back to original context */
} argon2_instance_t;
/*
* Argon2 position: where we construct the block right now. Used to distribute
* work between threads.
*/
typedef struct Argon2_position_t {
uint32_t pass;
uint32_t lane;
uint8_t slice;
uint32_t index;
} argon2_position_t;
/*Struct that holds the inputs for thread handling FillSegment*/
typedef struct Argon2_thread_data {
argon2_instance_t *instance_ptr;
argon2_position_t pos;
} argon2_thread_data;
/*************************Argon2 core functions********************************/
/* Allocates memory to the given pointer, uses the appropriate allocator as
* specified in the context. Total allocated memory is num*size.
* @param context argon2_context which specifies the allocator
* @param instance the Argon2 instance
* @return ARGON2_OK if memory is allocated successfully
*/
int allocate_memory(const argon2_context *context,
argon2_instance_t *instance);
/*
* Frees memory at the given pointer, uses the appropriate deallocator as
* specified in the context. Also cleans the memory using clear_internal_memory.
* @param context argon2_context which specifies the deallocator
* @param instance the Argon2 instance
*/
void free_memory(const argon2_context *context,
const argon2_instance_t *instance);
/* Function that securely cleans the memory. This ignores any flags set
* regarding clearing memory. Usually one just calls clear_internal_memory.
* @param mem Pointer to the memory
* @param s Memory size in bytes
*/
void secure_wipe_memory(void *v, size_t n);
/* Function that securely clears the memory if FLAG_clear_internal_memory is
* set. If the flag isn't set, this function does nothing.
* @param mem Pointer to the memory
* @param s Memory size in bytes
*/
ARGON2_PUBLIC void clear_internal_memory(void *v, size_t n);
/*
* Computes absolute position of reference block in the lane following a skewed
* distribution and using a pseudo-random value as input
* @param instance Pointer to the current instance
* @param position Pointer to the current position
* @param pseudo_rand 32-bit pseudo-random value used to determine the position
* @param same_lane Indicates if the block will be taken from the current lane.
* If so we can reference the current segment
* @pre All pointers must be valid
*/
uint32_t index_alpha(const argon2_instance_t *instance,
const argon2_position_t *position, uint32_t pseudo_rand,
int same_lane);
/*
* Function that validates all inputs against predefined restrictions and return
* an error code
* @param context Pointer to current Argon2 context
* @return ARGON2_OK if everything is all right, otherwise one of error codes
* (all defined in <argon2.h>
*/
int validate_inputs(const argon2_context *context);
/*
* Hashes all the inputs into @a blockhash[PREHASH_DIGEST_LENGTH], clears
* password and secret if needed
* @param context Pointer to the Argon2 internal structure containing memory
* pointer, and parameters for time and space requirements.
* @param blockhash Buffer for pre-hashing digest
* @param type Argon2 type
* @pre @a blockhash must have at least @a PREHASH_DIGEST_LENGTH bytes
* allocated
*/
void initial_hash(uint8_t *blockhash, argon2_context *context,
argon2_type type);
/*
* Function creates first 2 blocks per lane
* @param instance Pointer to the current instance
* @param blockhash Pointer to the pre-hashing digest
* @pre blockhash must point to @a PREHASH_SEED_LENGTH allocated values
*/
void fill_first_blocks(uint8_t *blockhash, const argon2_instance_t *instance);
/*
* Function allocates memory, hashes the inputs with Blake, and creates first
* two blocks. Returns the pointer to the main memory with 2 blocks per lane
* initialized
* @param context Pointer to the Argon2 internal structure containing memory
* pointer, and parameters for time and space requirements.
* @param instance Current Argon2 instance
* @return Zero if successful, -1 if memory failed to allocate. @context->state
* will be modified if successful.
*/
int initialize(argon2_instance_t *instance, argon2_context *context);
/*
* XORing the last block of each lane, hashing it, making the tag. Deallocates
* the memory.
* @param context Pointer to current Argon2 context (use only the out parameters
* from it)
* @param instance Pointer to current instance of Argon2
* @pre instance->state must point to necessary amount of memory
* @pre context->out must point to outlen bytes of memory
* @pre if context->free_cbk is not NULL, it should point to a function that
* deallocates memory
*/
void finalize(const argon2_context *context, argon2_instance_t *instance);
/*
* Function that fills the segment using previous segments also from other
* threads
* @param instance Pointer to the current instance
* @param position Current position
* @pre all block pointers must be valid
*/
void fill_segment(const argon2_instance_t *instance,
argon2_position_t position);
/*
* Function that fills the entire memory t_cost times based on the first two
* blocks in each lane
* @param instance Pointer to the current instance
* @return ARGON2_OK if successful, @context->state
*/
int fill_memory_blocks(argon2_instance_t *instance);
#endif

432
src/3rdparty/argon2/lib/encoding.c vendored Normal file
View file

@ -0,0 +1,432 @@
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <limits.h>
#include "encoding.h"
#include "core.h"
/*
* Example code for a decoder and encoder of "hash strings", with Argon2
* parameters.
*
* This code comprises three sections:
*
* -- The first section contains generic Base64 encoding and decoding
* functions. It is conceptually applicable to any hash function
* implementation that uses Base64 to encode and decode parameters,
* salts and outputs. It could be made into a library, provided that
* the relevant functions are made public (non-static) and be given
* reasonable names to avoid collisions with other functions.
*
* -- The second section is specific to Argon2. It encodes and decodes
* the parameters, salts and outputs. It does not compute the hash
* itself.
*
* The code was originally written by Thomas Pornin <pornin@bolet.org>,
* to whom comments and remarks may be sent. It is released under what
* should amount to Public Domain or its closest equivalent; the
* following mantra is supposed to incarnate that fact with all the
* proper legal rituals:
*
* ---------------------------------------------------------------------
* This file is provided under the terms of Creative Commons CC0 1.0
* Public Domain Dedication. To the extent possible under law, the
* author (Thomas Pornin) has waived all copyright and related or
* neighboring rights to this file. This work is published from: Canada.
* ---------------------------------------------------------------------
*
* Copyright (c) 2015 Thomas Pornin
*/
/* ==================================================================== */
/*
* Common code; could be shared between different hash functions.
*
* Note: the Base64 functions below assume that uppercase letters (resp.
* lowercase letters) have consecutive numerical codes, that fit on 8
* bits. All modern systems use ASCII-compatible charsets, where these
* properties are true. If you are stuck with a dinosaur of a system
* that still defaults to EBCDIC then you already have much bigger
* interoperability issues to deal with.
*/
/*
* Some macros for constant-time comparisons. These work over values in
* the 0..255 range. Returned value is 0x00 on "false", 0xFF on "true".
*/
#define EQ(x, y) ((((0U - ((unsigned)(x) ^ (unsigned)(y))) >> 8) & 0xFF) ^ 0xFF)
#define GT(x, y) ((((unsigned)(y) - (unsigned)(x)) >> 8) & 0xFF)
#define GE(x, y) (GT(y, x) ^ 0xFF)
#define LT(x, y) GT(y, x)
#define LE(x, y) GE(y, x)
/*
* Convert value x (0..63) to corresponding Base64 character.
*/
static int b64_byte_to_char(unsigned x) {
return (LT(x, 26) & (x + 'A')) |
(GE(x, 26) & LT(x, 52) & (x + ('a' - 26))) |
(GE(x, 52) & LT(x, 62) & (x + ('0' - 52))) | (EQ(x, 62) & '+') |
(EQ(x, 63) & '/');
}
/*
* Convert character c to the corresponding 6-bit value. If character c
* is not a Base64 character, then 0xFF (255) is returned.
*/
static unsigned b64_char_to_byte(int c) {
unsigned x;
x = (GE(c, 'A') & LE(c, 'Z') & (c - 'A')) |
(GE(c, 'a') & LE(c, 'z') & (c - ('a' - 26))) |
(GE(c, '0') & LE(c, '9') & (c - ('0' - 52))) | (EQ(c, '+') & 62) |
(EQ(c, '/') & 63);
return x | (EQ(x, 0) & (EQ(c, 'A') ^ 0xFF));
}
/*
* Convert some bytes to Base64. 'dst_len' is the length (in characters)
* of the output buffer 'dst'; if that buffer is not large enough to
* receive the result (including the terminating 0), then (size_t)-1
* is returned. Otherwise, the zero-terminated Base64 string is written
* in the buffer, and the output length (counted WITHOUT the terminating
* zero) is returned.
*/
static size_t to_base64(char *dst, size_t dst_len, const void *src,
size_t src_len) {
size_t olen;
const unsigned char *buf;
unsigned acc, acc_len;
olen = (src_len / 3) << 2;
switch (src_len % 3) {
case 2:
olen++;
/* fall through */
case 1:
olen += 2;
break;
}
if (dst_len <= olen) {
return (size_t)-1;
}
acc = 0;
acc_len = 0;
buf = (const unsigned char *)src;
while (src_len-- > 0) {
acc = (acc << 8) + (*buf++);
acc_len += 8;
while (acc_len >= 6) {
acc_len -= 6;
*dst++ = (char)b64_byte_to_char((acc >> acc_len) & 0x3F);
}
}
if (acc_len > 0) {
*dst++ = (char)b64_byte_to_char((acc << (6 - acc_len)) & 0x3F);
}
*dst++ = 0;
return olen;
}
/*
* Decode Base64 chars into bytes. The '*dst_len' value must initially
* contain the length of the output buffer '*dst'; when the decoding
* ends, the actual number of decoded bytes is written back in
* '*dst_len'.
*
* Decoding stops when a non-Base64 character is encountered, or when
* the output buffer capacity is exceeded. If an error occurred (output
* buffer is too small, invalid last characters leading to unprocessed
* buffered bits), then NULL is returned; otherwise, the returned value
* points to the first non-Base64 character in the source stream, which
* may be the terminating zero.
*/
static const char *from_base64(void *dst, size_t *dst_len, const char *src) {
size_t len;
unsigned char *buf;
unsigned acc, acc_len;
buf = (unsigned char *)dst;
len = 0;
acc = 0;
acc_len = 0;
for (;;) {
unsigned d;
d = b64_char_to_byte(*src);
if (d == 0xFF) {
break;
}
src++;
acc = (acc << 6) + d;
acc_len += 6;
if (acc_len >= 8) {
acc_len -= 8;
if ((len++) >= *dst_len) {
return NULL;
}
*buf++ = (acc >> acc_len) & 0xFF;
}
}
/*
* If the input length is equal to 1 modulo 4 (which is
* invalid), then there will remain 6 unprocessed bits;
* otherwise, only 0, 2 or 4 bits are buffered. The buffered
* bits must also all be zero.
*/
if (acc_len > 4 || (acc & (((unsigned)1 << acc_len) - 1)) != 0) {
return NULL;
}
*dst_len = len;
return src;
}
/*
* Decode decimal integer from 'str'; the value is written in '*v'.
* Returned value is a pointer to the next non-decimal character in the
* string. If there is no digit at all, or the value encoding is not
* minimal (extra leading zeros), or the value does not fit in an
* 'unsigned long', then NULL is returned.
*/
static const char *decode_decimal(const char *str, unsigned long *v) {
const char *orig;
unsigned long acc;
acc = 0;
for (orig = str;; str++) {
int c;
c = *str;
if (c < '0' || c > '9') {
break;
}
c -= '0';
if (acc > (ULONG_MAX / 10)) {
return NULL;
}
acc *= 10;
if ((unsigned long)c > (ULONG_MAX - acc)) {
return NULL;
}
acc += (unsigned long)c;
}
if (str == orig || (*orig == '0' && str != (orig + 1))) {
return NULL;
}
*v = acc;
return str;
}
/* ==================================================================== */
/*
* Code specific to Argon2.
*
* The code below applies the following format:
*
* $argon2<T>[$v=<num>]$m=<num>,t=<num>,p=<num>$<bin>$<bin>
*
* where <T> is either 'd', 'id', or 'i', <num> is a decimal integer (positive,
* fits in an 'unsigned long'), and <bin> is Base64-encoded data (no '=' padding
* characters, no newline or whitespace).
*
* The last two binary chunks (encoded in Base64) are, in that order,
* the salt and the output. Both are required. The binary salt length and the
* output length must be in the allowed ranges defined in argon2.h.
*
* The ctx struct must contain buffers large enough to hold the salt and pwd
* when it is fed into decode_string.
*/
int decode_string(argon2_context *ctx, const char *str, argon2_type type) {
/* check for prefix */
#define CC(prefix) \
do { \
size_t cc_len = strlen(prefix); \
if (strncmp(str, prefix, cc_len) != 0) { \
return ARGON2_DECODING_FAIL; \
} \
str += cc_len; \
} while ((void)0, 0)
/* optional prefix checking with supplied code */
#define CC_opt(prefix, code) \
do { \
size_t cc_len = strlen(prefix); \
if (strncmp(str, prefix, cc_len) == 0) { \
str += cc_len; \
{ code; } \
} \
} while ((void)0, 0)
/* Decoding prefix into uint32_t decimal */
#define DECIMAL_U32(x) \
do { \
unsigned long dec_x; \
str = decode_decimal(str, &dec_x); \
if (str == NULL || dec_x > UINT32_MAX) { \
return ARGON2_DECODING_FAIL; \
} \
(x) = (uint32_t)dec_x; \
} while ((void)0, 0)
/* Decoding base64 into a binary buffer */
#define BIN(buf, max_len, len) \
do { \
size_t bin_len = (max_len); \
str = from_base64(buf, &bin_len, str); \
if (str == NULL || bin_len > UINT32_MAX) { \
return ARGON2_DECODING_FAIL; \
} \
(len) = (uint32_t)bin_len; \
} while ((void)0, 0)
size_t maxsaltlen = ctx->saltlen;
size_t maxoutlen = ctx->outlen;
int validation_result;
const char* type_string;
/* We should start with the argon2_type we are using */
type_string = argon2_type2string(type, 0);
if (!type_string) {
return ARGON2_INCORRECT_TYPE;
}
CC("$");
CC(type_string);
/* Reading the version number if the default is suppressed */
ctx->version = ARGON2_VERSION_10;
CC_opt("$v=", DECIMAL_U32(ctx->version));
CC("$m=");
DECIMAL_U32(ctx->m_cost);
CC(",t=");
DECIMAL_U32(ctx->t_cost);
CC(",p=");
DECIMAL_U32(ctx->lanes);
ctx->threads = ctx->lanes;
CC("$");
BIN(ctx->salt, maxsaltlen, ctx->saltlen);
CC("$");
BIN(ctx->out, maxoutlen, ctx->outlen);
/* The rest of the fields get the default values */
ctx->secret = NULL;
ctx->secretlen = 0;
ctx->ad = NULL;
ctx->adlen = 0;
ctx->allocate_cbk = NULL;
ctx->free_cbk = NULL;
ctx->flags = ARGON2_DEFAULT_FLAGS;
/* On return, must have valid context */
validation_result = validate_inputs(ctx);
if (validation_result != ARGON2_OK) {
return validation_result;
}
/* Can't have any additional characters */
if (*str == 0) {
return ARGON2_OK;
} else {
return ARGON2_DECODING_FAIL;
}
#undef CC
#undef CC_opt
#undef DECIMAL_U32
#undef BIN
}
int encode_string(char *dst, size_t dst_len, argon2_context *ctx,
argon2_type type) {
#define SS(str) \
do { \
size_t pp_len = strlen(str); \
if (pp_len >= dst_len) { \
return ARGON2_ENCODING_FAIL; \
} \
memcpy(dst, str, pp_len + 1); \
dst += pp_len; \
dst_len -= pp_len; \
} while ((void)0, 0)
#define SX(x) \
do { \
char tmp[30]; \
sprintf(tmp, "%lu", (unsigned long)(x)); \
SS(tmp); \
} while ((void)0, 0)
#define SB(buf, len) \
do { \
size_t sb_len = to_base64(dst, dst_len, buf, len); \
if (sb_len == (size_t)-1) { \
return ARGON2_ENCODING_FAIL; \
} \
dst += sb_len; \
dst_len -= sb_len; \
} while ((void)0, 0)
const char* type_string = argon2_type2string(type, 0);
int validation_result = validate_inputs(ctx);
if (!type_string) {
return ARGON2_ENCODING_FAIL;
}
if (validation_result != ARGON2_OK) {
return validation_result;
}
SS("$");
SS(type_string);
SS("$v=");
SX(ctx->version);
SS("$m=");
SX(ctx->m_cost);
SS(",t=");
SX(ctx->t_cost);
SS(",p=");
SX(ctx->lanes);
SS("$");
SB(ctx->salt, ctx->saltlen);
SS("$");
SB(ctx->out, ctx->outlen);
return ARGON2_OK;
#undef SS
#undef SX
#undef SB
}
size_t b64len(uint32_t len) {
size_t olen = ((size_t)len / 3) << 2;
switch (len % 3) {
case 2:
olen++;
/* fall through */
case 1:
olen += 2;
break;
}
return olen;
}
size_t numlen(uint32_t num) {
size_t len = 1;
while (num >= 10) {
++len;
num = num / 10;
}
return len;
}

40
src/3rdparty/argon2/lib/encoding.h vendored Normal file
View file

@ -0,0 +1,40 @@
#ifndef ENCODING_H
#define ENCODING_H
#include "argon2.h"
#define ARGON2_MAX_DECODED_LANES UINT32_C(255)
#define ARGON2_MIN_DECODED_SALT_LEN UINT32_C(8)
#define ARGON2_MIN_DECODED_OUT_LEN UINT32_C(12)
/*
* encode an Argon2 hash string into the provided buffer. 'dst_len'
* contains the size, in characters, of the 'dst' buffer; if 'dst_len'
* is less than the number of required characters (including the
* terminating 0), then this function returns ARGON2_ENCODING_ERROR.
*
* on success, ARGON2_OK is returned.
*/
int encode_string(char *dst, size_t dst_len, argon2_context *ctx,
argon2_type type);
/*
* Decodes an Argon2 hash string into the provided structure 'ctx'.
* The only fields that must be set prior to this call are ctx.saltlen and
* ctx.outlen (which must be the maximal salt and out length values that are
* allowed), ctx.salt and ctx.out (which must be buffers of the specified
* length), and ctx.pwd and ctx.pwdlen which must hold a valid password.
*
* Invalid input string causes an error. On success, the ctx is valid and all
* fields have been initialized.
*
* Returned value is ARGON2_OK on success, other ARGON2_ codes on error.
*/
int decode_string(argon2_context *ctx, const char *str, argon2_type type);
/* Returns the length of the encoded byte stream with length len */
size_t b64len(uint32_t len);
/* Returns the length of the encoded number num */
size_t numlen(uint32_t num);
#endif

117
src/3rdparty/argon2/lib/genkat.c vendored Normal file
View file

@ -0,0 +1,117 @@
/*
* Argon2 source code package
*
* Written by Daniel Dinu and Dmitry Khovratovich, 2015
*
* This work is licensed under a Creative Commons CC0 1.0 License/Waiver.
*
* 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/>.
*/
#include <inttypes.h>
#include <stdio.h>
#include "genkat.h"
void initial_kat(const uint8_t *blockhash, const argon2_context *context,
argon2_type type) {
unsigned i;
if (blockhash != NULL && context != NULL) {
printf("=======================================\n");
printf("%s version number %d\n", argon2_type2string(type, 1),
context->version);
printf("=======================================\n");
printf("Memory: %u KiB, Iterations: %u, Parallelism: %u lanes, Tag "
"length: %u bytes\n",
context->m_cost, context->t_cost, context->lanes,
context->outlen);
printf("Password[%u]: ", context->pwdlen);
if (context->flags & ARGON2_FLAG_CLEAR_PASSWORD) {
printf("CLEARED\n");
} else {
for (i = 0; i < context->pwdlen; ++i) {
printf("%2.2x ", ((unsigned char *)context->pwd)[i]);
}
printf("\n");
}
printf("Salt[%u]: ", context->saltlen);
for (i = 0; i < context->saltlen; ++i) {
printf("%2.2x ", ((unsigned char *)context->salt)[i]);
}
printf("\n");
printf("Secret[%u]: ", context->secretlen);
if (context->flags & ARGON2_FLAG_CLEAR_SECRET) {
printf("CLEARED\n");
} else {
for (i = 0; i < context->secretlen; ++i) {
printf("%2.2x ", ((unsigned char *)context->secret)[i]);
}
printf("\n");
}
printf("Associated data[%u]: ", context->adlen);
for (i = 0; i < context->adlen; ++i) {
printf("%2.2x ", ((unsigned char *)context->ad)[i]);
}
printf("\n");
printf("Pre-hashing digest: ");
for (i = 0; i < ARGON2_PREHASH_DIGEST_LENGTH; ++i) {
printf("%2.2x ", ((unsigned char *)blockhash)[i]);
}
printf("\n");
}
}
void print_tag(const void *out, uint32_t outlen) {
unsigned i;
if (out != NULL) {
printf("Tag: ");
for (i = 0; i < outlen; ++i) {
printf("%2.2x ", ((uint8_t *)out)[i]);
}
printf("\n");
}
}
void internal_kat(const argon2_instance_t *instance, uint32_t pass) {
if (instance != NULL) {
uint32_t i, j;
printf("\n After pass %u:\n", pass);
for (i = 0; i < instance->memory_blocks; ++i) {
uint32_t how_many_words =
(instance->memory_blocks > ARGON2_QWORDS_IN_BLOCK)
? 1
: ARGON2_QWORDS_IN_BLOCK;
for (j = 0; j < how_many_words; ++j)
printf("Block %.4u [%3u]: %016" PRIx64 "\n", i, j,
instance->memory[i].v[j]);
}
}
}

47
src/3rdparty/argon2/lib/genkat.h vendored Normal file
View file

@ -0,0 +1,47 @@
/*
* Argon2 source code package
*
* Written by Daniel Dinu and Dmitry Khovratovich, 2015
*
* This work is licensed under a Creative Commons CC0 1.0 License/Waiver.
*
* 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/>.
*/
#ifndef ARGON2_KAT_H
#define ARGON2_KAT_H
#include "core.h"
/*
* Initial KAT function that prints the inputs to the file
* @param blockhash Array that contains pre-hashing digest
* @param context Holds inputs
* @param type Argon2 type
* @pre blockhash must point to INPUT_INITIAL_HASH_LENGTH bytes
* @pre context member pointers must point to allocated memory of size according
* to the length values
*/
void initial_kat(const uint8_t *blockhash, const argon2_context *context,
argon2_type type);
/*
* Function that prints the output tag
* @param out output array pointer
* @param outlen digest length
* @pre out must point to @a outlen bytes
**/
void print_tag(const void *out, uint32_t outlen);
/*
* Function that prints the internal state at given moment
* @param instance pointer to the current instance
* @param pass current pass number
* @pre instance must have necessary memory allocated
**/
void internal_kat(const argon2_instance_t *instance, uint32_t pass);
#endif

115
src/3rdparty/argon2/lib/impl-select.c vendored Normal file
View file

@ -0,0 +1,115 @@
#include <time.h>
#include <string.h>
#include "impl-select.h"
#include "argon2.h"
#define BENCH_SAMPLES 1024
#define BENCH_MEM_BLOCKS 512
static argon2_impl selected_argon_impl = {
"default", NULL, fill_segment_default
};
/* the benchmark routine is not thread-safe, so we can use a global var here: */
static block memory[BENCH_MEM_BLOCKS];
static uint64_t benchmark_impl(const argon2_impl *impl) {
clock_t time;
unsigned int i;
uint64_t bench;
argon2_instance_t instance;
argon2_position_t pos;
memset(memory, 0, sizeof(memory));
instance.version = ARGON2_VERSION_NUMBER;
instance.memory = memory;
instance.passes = 1;
instance.memory_blocks = BENCH_MEM_BLOCKS;
instance.segment_length = BENCH_MEM_BLOCKS / ARGON2_SYNC_POINTS;
instance.lane_length = instance.segment_length * ARGON2_SYNC_POINTS;
instance.lanes = 1;
instance.threads = 1;
instance.type = Argon2_i;
pos.lane = 0;
pos.pass = 0;
pos.slice = 0;
pos.index = 0;
/* warm-up cache: */
impl->fill_segment(&instance, pos);
/* OK, now measure: */
bench = 0;
time = clock();
for (i = 0; i < BENCH_SAMPLES; i++) {
impl->fill_segment(&instance, pos);
}
time = clock() - time;
bench = (uint64_t)time;
return bench;
}
void argon2_select_impl()
{
argon2_impl_list impls;
unsigned int i;
const argon2_impl *best_impl = NULL;
uint64_t best_bench = UINT_MAX;
argon2_get_impl_list(&impls);
for (i = 0; i < impls.count; i++) {
const argon2_impl *impl = &impls.entries[i];
uint64_t bench;
if (impl->check != NULL && !impl->check()) {
continue;
}
bench = benchmark_impl(impl);
if (bench < best_bench) {
best_bench = bench;
best_impl = impl;
}
}
if (best_impl != NULL) {
selected_argon_impl = *best_impl;
}
}
void fill_segment(const argon2_instance_t *instance, argon2_position_t position)
{
selected_argon_impl.fill_segment(instance, position);
}
const char *argon2_get_impl_name()
{
return selected_argon_impl.name;
}
int argon2_select_impl_by_name(const char *name)
{
argon2_impl_list impls;
unsigned int i;
argon2_get_impl_list(&impls);
for (i = 0; i < impls.count; i++) {
const argon2_impl *impl = &impls.entries[i];
if (strcmp(impl->name, name) == 0) {
selected_argon_impl = *impl;
return 1;
}
}
return 0;
}

23
src/3rdparty/argon2/lib/impl-select.h vendored Normal file
View file

@ -0,0 +1,23 @@
#ifndef ARGON2_IMPL_SELECT_H
#define ARGON2_IMPL_SELECT_H
#include "core.h"
typedef struct Argon2_impl {
const char *name;
int (*check)(void);
void (*fill_segment)(const argon2_instance_t *instance,
argon2_position_t position);
} argon2_impl;
typedef struct Argon2_impl_list {
const argon2_impl *entries;
size_t count;
} argon2_impl_list;
void argon2_get_impl_list(argon2_impl_list *list);
void fill_segment_default(const argon2_instance_t *instance,
argon2_position_t position);
#endif // ARGON2_IMPL_SELECT_H

68
src/3rdparty/base32/base32.h vendored Normal file
View file

@ -0,0 +1,68 @@
// Base32 implementation
//
// Copyright 2010 Google Inc.
// Author: Markus Gutschke
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
// Encode and decode from base32 encoding using the following alphabet:
// ABCDEFGHIJKLMNOPQRSTUVWXYZ234567
// This alphabet is documented in RFC 4648/3548
//
// We allow white-space and hyphens, but all other characters are considered
// invalid.
//
// All functions return the number of output bytes or -1 on error. If the
// output buffer is too small, the result will silently be truncated.
#ifndef XMRIG_BASE32_H
#define XMRIG_BASE32_H
#include <stdint.h>
int base32_encode(const uint8_t *data, int length, uint8_t *result, int bufSize) {
if (length < 0 || length > (1 << 28)) {
return -1;
}
int count = 0;
if (length > 0) {
int buffer = data[0];
int next = 1;
int bitsLeft = 8;
while (count < bufSize && (bitsLeft > 0 || next < length)) {
if (bitsLeft < 5) {
if (next < length) {
buffer <<= 8;
buffer |= data[next++] & 0xFF;
bitsLeft += 8;
} else {
int pad = 5 - bitsLeft;
buffer <<= pad;
bitsLeft += pad;
}
}
int index = 0x1F & (buffer >> (bitsLeft - 5));
bitsLeft -= 5;
result[count++] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567"[index];
}
}
if (count < bufSize) {
result[count] = '\000';
}
return count;
}
#endif /* XMRIG_BASE32_H */

36
src/3rdparty/cl.h vendored Normal file
View file

@ -0,0 +1,36 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* 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/>.
*/
#ifndef XMRIG_CL_H
#define XMRIG_CL_H
#if defined(__APPLE__)
# include <OpenCL/cl.h>
#else
# include "3rdparty/CL/cl.h"
#endif
#endif /* XMRIG_CL_H */

5
src/Summary.cpp
View file

@ -71,13 +71,12 @@ static void print_cpu(Config *)
{
const ICpuInfo *info = Cpu::info();
Log::print(GREEN_BOLD(" * ") WHITE_BOLD("%-13s%s (%zu)") " %sx64 %sAES %sAVX2",
Log::print(GREEN_BOLD(" * ") WHITE_BOLD("%-13s%s (%zu)") " %sx64 %sAES",
"CPU",
info->brand(),
info->packages(),
info->isX64() ? GREEN_BOLD_S : RED_BOLD_S "-",
info->hasAES() ? GREEN_BOLD_S : RED_BOLD_S "-",
info->hasAVX2() ? GREEN_BOLD_S : RED_BOLD_S "-"
info->hasAES() ? GREEN_BOLD_S : RED_BOLD_S "-"
);
# if defined(XMRIG_FEATURE_LIBCPUID) || defined (XMRIG_FEATURE_HWLOC)
Log::print(WHITE_BOLD(" %-13s") BLACK_BOLD("L2:") WHITE_BOLD("%.1f MB") BLACK_BOLD(" L3:") WHITE_BOLD("%.1f MB")

3
src/backend/backend.cmake
View file

@ -1,13 +1,16 @@
include (src/backend/cpu/cpu.cmake)
include (src/backend/opencl/opencl.cmake)
include (src/backend/common/common.cmake)
set(HEADERS_BACKEND
"${HEADERS_BACKEND_COMMON}"
"${HEADERS_BACKEND_CPU}"
"${HEADERS_BACKEND_OPENCL}"
)
set(SOURCES_BACKEND
"${SOURCES_BACKEND_COMMON}"
"${SOURCES_BACKEND_CPU}"
"${SOURCES_BACKEND_OPENCL}"
)

44
src/backend/common/Tags.h Normal file
View file

@ -0,0 +1,44 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018 Lee Clagett <https://github.com/vtnerd>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* 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/>.
*/
#ifndef XMRIG_TAGS_H
#define XMRIG_TAGS_H
namespace xmrig {
const char *cpu_tag();
#ifdef XMRIG_FEATURE_OPENCL
const char *ocl_tag();
#endif
} // namespace xmrig
#endif /* XMRIG_TAGS_H */

6
src/backend/common/Thread.h
View file

@ -42,18 +42,18 @@ template<class T>
class Thread
{
public:
inline Thread(IBackend *backend, size_t index, const T &config) : m_index(index), m_config(config), m_backend(backend) {}
inline Thread(IBackend *backend, size_t id, const T &config) : m_id(id), m_config(config), m_backend(backend) {}
inline ~Thread() { m_thread.join(); delete m_worker; }
inline const T &config() const { return m_config; }
inline IBackend *backend() const { return m_backend; }
inline IWorker *worker() const { return m_worker; }
inline size_t index() const { return m_index; }
inline size_t id() const { return m_id; }
inline void setWorker(IWorker *worker) { m_worker = worker; }
inline void start(void (*callback) (void *)) { m_thread = std::thread(callback, this); }
private:
const size_t m_index = 0;
const size_t m_id = 0;
const T m_config;
IBackend *m_backend;
IWorker *m_worker = nullptr;

19
src/backend/common/Threads.cpp
View file

@ -25,13 +25,20 @@
#include "backend/common/Threads.h"
#include "backend/cpu/CpuThreads.h"
#include "crypto/cn/CnAlgo.h"
#include "rapidjson/document.h"
#ifdef XMRIG_FEATURE_OPENCL
# include "backend/opencl/OclThreads.h"
#endif
namespace xmrig {
static const char *kAsterisk = "*";
static const char *kCn2 = "cn/2";
} // namespace xmrig
@ -61,7 +68,11 @@ size_t xmrig::Threads<T>::read(const rapidjson::Value &value)
if (!threads.isEmpty()) {
move(member.name.GetString(), std::move(threads));
}
}
}
for (auto &member : value.GetObject()) {
if (member.value.IsArray() || member.value.IsObject()) {
continue;
}
@ -109,6 +120,10 @@ xmrig::String xmrig::Threads<T>::profileName(const Algorithm &algorithm, bool st
return String();
}
if (algorithm.family() == Algorithm::CN && CnAlgo<>::base(algorithm) == Algorithm::CN_2 && has(kCn2)) {
return kCn2;
}
if (name.contains("/")) {
const String base = name.split('/').at(0);
if (has(base)) {
@ -148,4 +163,8 @@ namespace xmrig {
template class Threads<CpuThreads>;
#ifdef XMRIG_FEATURE_OPENCL
template class Threads<OclThreads>;
#endif
} // namespace xmrig

8
src/backend/common/Threads.h
View file

@ -47,7 +47,13 @@ public:
inline bool isExist(const Algorithm &algo) const { return isDisabled(algo) || m_aliases.count(algo) > 0 || has(algo.shortName()); }
inline const T &get(const Algorithm &algo, bool strict = false) const { return get(profileName(algo, strict)); }
inline void disable(const Algorithm &algo) { m_disabled.insert(algo); }
inline void move(const char *profile, T &&threads) { m_profiles.insert({ profile, threads }); }
inline void move(const char *profile, T &&threads)
{
if (!threads.isEmpty()) {
m_profiles.insert({ profile, std::move(threads) });
}
}
const T &get(const String &profileName) const;
size_t read(const rapidjson::Value &value);

2
src/backend/common/Worker.h
View file

@ -28,7 +28,7 @@
#include <atomic>
#include <stdint.h>
#include <cstdint>
#include "backend/common/interfaces/IWorker.h"

17
src/backend/common/WorkerJob.h
View file

@ -56,6 +56,7 @@ public:
}
if (index() == 1 && job.index() == 0 && job == m_jobs[0]) {
m_index = 0;
return;
}
@ -63,18 +64,18 @@ public:
}
inline void nextRound(uint32_t reserveCount)
inline void nextRound(uint32_t rounds, uint32_t roundSize)
{
m_rounds[index()]++;
if ((m_rounds[index()] % reserveCount) == 0) {
if ((m_rounds[index()] % rounds) == 0) {
for (size_t i = 0; i < N; ++i) {
*nonce(i) = Nonce::next(index(), *nonce(i), reserveCount, currentJob().isNicehash());
*nonce(i) = Nonce::next(index(), *nonce(i), rounds * roundSize, currentJob().isNicehash());
}
}
else {
for (size_t i = 0; i < N; ++i) {
*nonce(i) += 1;
*nonce(i) += roundSize;
}
}
}
@ -111,15 +112,15 @@ inline uint32_t *xmrig::WorkerJob<1>::nonce(size_t)
template<>
inline void xmrig::WorkerJob<1>::nextRound(uint32_t reserveCount)
inline void xmrig::WorkerJob<1>::nextRound(uint32_t rounds, uint32_t roundSize)
{
m_rounds[index()]++;
if ((m_rounds[index()] % reserveCount) == 0) {
*nonce() = Nonce::next(index(), *nonce(), reserveCount, currentJob().isNicehash());
if ((m_rounds[index()] % rounds) == 0) {
*nonce() = Nonce::next(index(), *nonce(), rounds * roundSize, currentJob().isNicehash());
}
else {
*nonce() += 1;
*nonce() += roundSize;
}
}

49
src/backend/common/Workers.cpp
View file

@ -31,6 +31,11 @@
#include "base/io/log/Log.h"
#ifdef XMRIG_FEATURE_OPENCL
# include "backend/opencl/OclWorker.h"
#endif
namespace xmrig {
@ -129,7 +134,7 @@ void xmrig::Workers<T>::tick(uint64_t)
return;
}
d_ptr->hashrate->add(handle->index(), handle->worker()->hashCount(), handle->worker()->timestamp());
d_ptr->hashrate->add(handle->id(), handle->worker()->hashCount(), handle->worker()->timestamp());
}
d_ptr->hashrate->updateHighest();
@ -137,7 +142,7 @@ void xmrig::Workers<T>::tick(uint64_t)
template<class T>
xmrig::IWorker *xmrig::Workers<T>::create(Thread<CpuLaunchData> *)
xmrig::IWorker *xmrig::Workers<T>::create(Thread<T> *)
{
return nullptr;
}
@ -146,15 +151,18 @@ xmrig::IWorker *xmrig::Workers<T>::create(Thread<CpuLaunchData> *)
template<class T>
void xmrig::Workers<T>::onReady(void *arg)
{
Thread<T> *handle = static_cast<Thread<T>* >(arg);
auto handle = static_cast<Thread<T>* >(arg);
IWorker *worker = create(handle);
if (!worker || !worker->selfTest()) {
LOG_ERR("thread %zu error: \"hash self-test failed\".", worker->id());
delete worker;
return;
}
assert(handle->backend() != nullptr);
handle->setWorker(worker);
handle->backend()->start(worker);
}
@ -166,32 +174,21 @@ namespace xmrig {
template<>
xmrig::IWorker *xmrig::Workers<CpuLaunchData>::create(Thread<CpuLaunchData> *handle)
{
const int intensity = handle->config().intensity;
# if defined(XMRIG_ALGO_RANDOMX) || defined(XMRIG_ALGO_CN_GPU)
if (intensity > handle->config().algorithm.maxIntensity()) {
LOG_WARN("CPU thread %zu warning: \"intensity %d not supported for %s algorithm\".", handle->index(), handle->config().intensity, handle->config().algorithm.shortName());
return new CpuWorker<1>(handle->index(), handle->config());
}
# endif
switch (intensity) {
switch (handle->config().intensity) {
case 1:
return new CpuWorker<1>(handle->index(), handle->config());
return new CpuWorker<1>(handle->id(), handle->config());
case 2:
return new CpuWorker<2>(handle->index(), handle->config());
return new CpuWorker<2>(handle->id(), handle->config());
case 3:
return new CpuWorker<3>(handle->index(), handle->config());
return new CpuWorker<3>(handle->id(), handle->config());
case 4:
return new CpuWorker<4>(handle->index(), handle->config());
return new CpuWorker<4>(handle->id(), handle->config());
case 5:
return new CpuWorker<5>(handle->index(), handle->config());
return new CpuWorker<5>(handle->id(), handle->config());
}
return nullptr;
@ -201,4 +198,16 @@ xmrig::IWorker *xmrig::Workers<CpuLaunchData>::create(Thread<CpuLaunchData> *han
template class Workers<CpuLaunchData>;
#ifdef XMRIG_FEATURE_OPENCL
template<>
xmrig::IWorker *xmrig::Workers<OclLaunchData>::create(Thread<OclLaunchData> *handle)
{
return new OclWorker(handle->id(), handle->config());
}
template class Workers<OclLaunchData>;
#endif
} // namespace xmrig

16
src/backend/common/Workers.h
View file

@ -31,6 +31,11 @@
#include "backend/cpu/CpuLaunchData.h"
#ifdef XMRIG_FEATURE_OPENCL
# include "backend/opencl/OclLaunchData.h"
#endif
namespace xmrig {
@ -52,7 +57,7 @@ public:
void tick(uint64_t ticks);
private:
static IWorker *create(Thread<CpuLaunchData> *handle);
static IWorker *create(Thread<T> *handle);
static void onReady(void *arg);
std::vector<Thread<T> *> m_workers;
@ -62,11 +67,16 @@ private:
template<>
IWorker *Workers<CpuLaunchData>::create(Thread<CpuLaunchData> *handle);
extern template class Workers<CpuLaunchData>;
#ifdef XMRIG_FEATURE_OPENCL
template<>
IWorker *Workers<OclLaunchData>::create(Thread<OclLaunchData> *handle);
extern template class Workers<OclLaunchData>;
#endif
} // namespace xmrig

7
src/backend/common/common.cmake
View file

@ -1,13 +1,16 @@
set(HEADERS_BACKEND_COMMON
src/backend/common/Hashrate.h
src/backend/common/Tags.h
src/backend/common/interfaces/IBackend.h
src/backend/common/interfaces/IRxListener.h
src/backend/common/interfaces/IThread.h
src/backend/common/interfaces/IWorker.h
src/backend/common/Hashrate.h
src/backend/common/misc/PciTopology.h
src/backend/common/Thread.h
src/backend/common/Threads.h
src/backend/common/Worker.h
src/backend/common/Workers.h
src/backend/common/WorkerJob.h
src/backend/common/Workers.h
)
set(SOURCES_BACKEND_COMMON

2
src/backend/common/interfaces/IBackend.h
View file

@ -37,6 +37,7 @@ namespace xmrig {
class Algorithm;
class Hashrate;
class IApiRequest;
class IWorker;
class Job;
class String;
@ -61,6 +62,7 @@ public:
# ifdef XMRIG_FEATURE_API
virtual rapidjson::Value toJSON(rapidjson::Document &doc) const = 0;
virtual void handleRequest(IApiRequest &request) = 0;
# endif
};

44
src/backend/common/interfaces/IRxListener.h Normal file
View file

@ -0,0 +1,44 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2016-2018 XMRig <support@xmrig.com>
*
* 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/>.
*/
#ifndef XMRIG_IRXLISTENER_H
#define XMRIG_IRXLISTENER_H
namespace xmrig {
class IRxListener
{
public:
virtual ~IRxListener() = default;
# ifdef XMRIG_ALGO_RANDOMX
virtual void onDatasetReady() = 0;
# endif
};
} /* namespace xmrig */
#endif // XMRIG_IRXLISTENER_H

62
src/backend/common/misc/PciTopology.h Normal file
View file

@ -0,0 +1,62 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018 Lee Clagett <https://github.com/vtnerd>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* 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/>.
*/
#ifndef XMRIG_PCITOPOLOGY_H
#define XMRIG_PCITOPOLOGY_H
#include <stdio.h>
#include "base/tools/String.h"
namespace xmrig {
class PciTopology
{
public:
PciTopology() = default;
PciTopology(uint32_t bus, uint32_t device, uint32_t function) : bus(bus), device(device), function(function) {}
uint32_t bus = 0;
uint32_t device = 0;
uint32_t function = 0;
String toString() const
{
char *buf = new char[8]();
snprintf(buf, 8, "%02x:%02x.%01x", bus, device, function);
return buf;
}
};
} // namespace xmrig
#endif /* XMRIG_PCITOPOLOGY_H */

228
src/backend/cpu/CpuBackend.cpp
View file

@ -28,6 +28,7 @@
#include "backend/common/Hashrate.h"
#include "backend/common/interfaces/IWorker.h"
#include "backend/common/Tags.h"
#include "backend/common/Workers.h"
#include "backend/cpu/Cpu.h"
#include "backend/cpu/CpuBackend.h"
@ -43,6 +44,16 @@
#include "rapidjson/document.h"
#ifdef XMRIG_FEATURE_API
# include "base/api/interfaces/IApiRequest.h"
#endif
#ifdef XMRIG_ALGO_ARGON2
# include "crypto/argon2/Impl.h"
#endif
namespace xmrig {
@ -51,29 +62,63 @@ extern template class Threads<CpuThreads>;
static const char *tag = CYAN_BG_BOLD(" cpu ");
static const String kType = "cpu";
static std::mutex mutex;
struct LaunchStatus
struct CpuLaunchStatus
{
public:
inline void reset()
inline size_t hugePages() const { return m_hugePages; }
inline size_t memory() const { return m_ways * m_memory; }
inline size_t pages() const { return m_pages; }
inline size_t threads() const { return m_threads; }
inline size_t ways() const { return m_ways; }
inline void start(const std::vector<CpuLaunchData> &threads, size_t memory)
{
hugePages = 0;
memory = 0;
pages = 0;
started = 0;
threads = 0;
ways = 0;
ts = Chrono::steadyMSecs();
m_hugePages = 0;
m_memory = memory;
m_pages = 0;
m_started = 0;
m_threads = threads.size();
m_ways = 0;
m_ts = Chrono::steadyMSecs();
for (const CpuLaunchData &data : threads) {
m_ways += data.intensity;
}
}
size_t hugePages = 0;
size_t memory = 0;
size_t pages = 0;
size_t started = 0;
size_t threads = 0;
size_t ways = 0;
uint64_t ts = 0;
inline bool started(const std::pair<size_t, size_t> &hugePages)
{
m_started++;
m_hugePages += hugePages.first;
m_pages += hugePages.second;
return m_started == m_threads;
}
inline void print() const
{
LOG_INFO("%s" GREEN_BOLD(" READY") " threads " CYAN_BOLD("%zu(%zu)") " huge pages %s%zu/%zu %1.0f%%\x1B[0m memory " CYAN_BOLD("%zu KB") BLACK_BOLD(" (%" PRIu64 " ms)"),
tag,
m_threads, m_ways,
(m_hugePages == m_pages ? GREEN_BOLD_S : (m_hugePages == 0 ? RED_BOLD_S : YELLOW_BOLD_S)),
m_hugePages, m_pages,
m_hugePages == 0 ? 0.0 : static_cast<double>(m_hugePages) / m_pages * 100.0,
m_ways * m_memory / 1024,
Chrono::steadyMSecs() - m_ts
);
}
private:
size_t m_hugePages = 0;
size_t m_memory = 0;
size_t m_pages = 0;
size_t m_started = 0;
size_t m_threads = 0;
size_t m_ways = 0;
uint64_t m_ts = 0;
};
@ -95,16 +140,7 @@ public:
algo.l3() / 1024
);
workers.stop();
status.reset();
status.memory = algo.l3();
status.threads = threads.size();
for (const CpuLaunchData &data : threads) {
status.ways += static_cast<size_t>(data.intensity);
}
status.start(threads, algo.l3());
workers.start(threads);
}
@ -113,14 +149,45 @@ public:
{
std::lock_guard<std::mutex> lock(mutex);
return status.ways;
return status.ways();
}
rapidjson::Value hugePages(int version, rapidjson::Document &doc)
{
std::pair<unsigned, unsigned> pages(0, 0);
# ifdef XMRIG_ALGO_RANDOMX
if (algo.family() == Algorithm::RANDOM_X) {
pages = Rx::hugePages();
}
# endif
mutex.lock();
pages.first += status.hugePages();
pages.second += status.pages();
mutex.unlock();
rapidjson::Value hugepages;
if (version > 1) {
hugepages.SetArray();
hugepages.PushBack(pages.first, doc.GetAllocator());
hugepages.PushBack(pages.second, doc.GetAllocator());
}
else {
hugepages = pages.first == pages.second;
}
return hugepages;
}
Algorithm algo;
Controller *controller;
LaunchStatus status;
std::mutex mutex;
CpuLaunchStatus status;
std::vector<CpuLaunchData> threads;
String profileName;
Workers<CpuLaunchData> workers;
@ -130,6 +197,12 @@ public:
} // namespace xmrig
const char *xmrig::cpu_tag()
{
return tag;
}
xmrig::CpuBackend::CpuBackend(Controller *controller) :
d_ptr(new CpuBackendPrivate(controller))
{
@ -143,25 +216,6 @@ xmrig::CpuBackend::~CpuBackend()
}
std::pair<unsigned, unsigned> xmrig::CpuBackend::hugePages() const
{
std::pair<unsigned, unsigned> pages(0, 0);
# ifdef XMRIG_ALGO_RANDOMX
if (d_ptr->algo.family() == Algorithm::RANDOM_X) {
pages = Rx::hugePages();
}
# endif
std::lock_guard<std::mutex> lock(d_ptr->mutex);
pages.first += d_ptr->status.hugePages;
pages.second += d_ptr->status.pages;
return pages;
}
bool xmrig::CpuBackend::isEnabled() const
{
return d_ptr->controller->config()->cpu().isEnabled();
@ -192,8 +246,17 @@ const xmrig::String &xmrig::CpuBackend::type() const
}
void xmrig::CpuBackend::prepare(const Job &)
void xmrig::CpuBackend::prepare(const Job &nextJob)
{
# ifdef XMRIG_ALGO_ARGON2
if (nextJob.algorithm().family() == Algorithm::ARGON2 && argon2::Impl::select(d_ptr->controller->config()->cpu().argon2Impl())) {
LOG_INFO("%s use " WHITE_BOLD("argon2") " implementation " CSI "1;%dm" "%s",
tag,
argon2::Impl::name() == "default" ? 33 : 32,
argon2::Impl::name().data()
);
}
# endif
}
@ -205,11 +268,11 @@ void xmrig::CpuBackend::printHashrate(bool details)
char num[8 * 3] = { 0 };
Log::print(WHITE_BOLD_S "| CPU THREAD | AFFINITY | 10s H/s | 60s H/s | 15m H/s |");
Log::print(WHITE_BOLD_S "| CPU # | AFFINITY | 10s H/s | 60s H/s | 15m H/s |");
size_t i = 0;
for (const CpuLaunchData &data : d_ptr->threads) {
Log::print("| %13zu | %8" PRId64 " | %7s | %7s | %7s |",
Log::print("| %8zu | %8" PRId64 " | %7s | %7s | %7s |",
i,
data.affinity,
Hashrate::format(hashrate()->calc(i, Hashrate::ShortInterval), num, sizeof num / 3),
@ -219,6 +282,14 @@ void xmrig::CpuBackend::printHashrate(bool details)
i++;
}
# ifdef XMRIG_FEATURE_OPENCL
Log::print(WHITE_BOLD_S "| - | - | %7s | %7s | %7s |",
Hashrate::format(hashrate()->calc(Hashrate::ShortInterval), num, sizeof num / 3),
Hashrate::format(hashrate()->calc(Hashrate::MediumInterval), num + 8, sizeof num / 3),
Hashrate::format(hashrate()->calc(Hashrate::LargeInterval), num + 8 * 2, sizeof num / 3)
);
# endif
}
@ -231,7 +302,7 @@ void xmrig::CpuBackend::setJob(const Job &job)
const CpuConfig &cpu = d_ptr->controller->config()->cpu();
std::vector<CpuLaunchData> threads = cpu.get(d_ptr->controller->miner(), job.algorithm());
if (d_ptr->threads.size() == threads.size() && std::equal(d_ptr->threads.begin(), d_ptr->threads.end(), threads.begin())) {
if (!d_ptr->threads.empty() && d_ptr->threads.size() == threads.size() && std::equal(d_ptr->threads.begin(), d_ptr->threads.end(), threads.begin())) {
return;
}
@ -239,13 +310,13 @@ void xmrig::CpuBackend::setJob(const Job &job)
d_ptr->profileName = cpu.threads().profileName(job.algorithm());
if (d_ptr->profileName.isNull() || threads.empty()) {
d_ptr->workers.stop();
LOG_WARN("%s " RED_BOLD("disabled") YELLOW(" (no suitable configuration found)"), tag);
LOG_WARN(YELLOW_BOLD_S "CPU disabled, no suitable configuration for algo %s", job.algorithm().shortName());
return;
return stop();
}
stop();
d_ptr->threads = std::move(threads);
d_ptr->start();
}
@ -253,28 +324,13 @@ void xmrig::CpuBackend::setJob(const Job &job)
void xmrig::CpuBackend::start(IWorker *worker)
{
d_ptr->mutex.lock();
mutex.lock();
const auto pages = worker->memory()->hugePages();
d_ptr->status.started++;
d_ptr->status.hugePages += pages.first;
d_ptr->status.pages += pages.second;
if (d_ptr->status.started == d_ptr->status.threads) {
const double percent = d_ptr->status.hugePages == 0 ? 0.0 : static_cast<double>(d_ptr->status.hugePages) / d_ptr->status.pages * 100.0;
const size_t memory = d_ptr->status.ways * d_ptr->status.memory / 1024;
LOG_INFO("%s" GREEN_BOLD(" READY") " threads " CYAN_BOLD("%zu(%zu)") " huge pages %s%zu/%zu %1.0f%%\x1B[0m memory " CYAN_BOLD("%zu KB") BLACK_BOLD(" (%" PRIu64 " ms)"),
tag,
d_ptr->status.threads, d_ptr->status.ways,
(d_ptr->status.hugePages == d_ptr->status.pages ? GREEN_BOLD_S : (d_ptr->status.hugePages == 0 ? RED_BOLD_S : YELLOW_BOLD_S)),
d_ptr->status.hugePages, d_ptr->status.pages, percent, memory,
Chrono::steadyMSecs() - d_ptr->status.ts
);
if (d_ptr->status.started(worker->memory()->hugePages())) {
d_ptr->status.print();
}
d_ptr->mutex.unlock();
mutex.unlock();
worker->start();
}
@ -282,6 +338,10 @@ void xmrig::CpuBackend::start(IWorker *worker)
void xmrig::CpuBackend::stop()
{
if (d_ptr->threads.empty()) {
return;
}
const uint64_t ts = Chrono::steadyMSecs();
d_ptr->workers.stop();
@ -319,13 +379,11 @@ rapidjson::Value xmrig::CpuBackend::toJSON(rapidjson::Document &doc) const
out.AddMember("asm", false, allocator);
# endif
const auto pages = hugePages();
# ifdef XMRIG_ALGO_ARGON2
out.AddMember("argon2-impl", argon2::Impl::name().toJSON(), allocator);
# endif
rapidjson::Value hugepages(rapidjson::kArrayType);
hugepages.PushBack(pages.first, allocator);
hugepages.PushBack(pages.second, allocator);
out.AddMember("hugepages", hugepages, allocator);
out.AddMember("hugepages", d_ptr->hugePages(2, doc), allocator);
out.AddMember("memory", static_cast<uint64_t>(d_ptr->algo.isValid() ? (d_ptr->ways() * d_ptr->algo.l3()) : 0), allocator);
if (d_ptr->threads.empty() || !hashrate()) {
@ -357,4 +415,12 @@ rapidjson::Value xmrig::CpuBackend::toJSON(rapidjson::Document &doc) const
return out;
}
void xmrig::CpuBackend::handleRequest(IApiRequest &request)
{
if (request.type() == IApiRequest::REQ_SUMMARY) {
request.reply().AddMember("hugepages", d_ptr->hugePages(request.version(), request.doc()), request.doc().GetAllocator());
}
}
#endif

3
src/backend/cpu/CpuBackend.h
View file

@ -46,8 +46,6 @@ public:
CpuBackend(Controller *controller);
~CpuBackend() override;
std::pair<unsigned, unsigned> hugePages() const;
protected:
bool isEnabled() const override;
bool isEnabled(const Algorithm &algorithm) const override;
@ -63,6 +61,7 @@ protected:
# ifdef XMRIG_FEATURE_API
rapidjson::Value toJSON(rapidjson::Document &doc) const override;
void handleRequest(IApiRequest &request) override;
# endif
private:

31
src/backend/cpu/CpuConfig.cpp
View file

@ -62,6 +62,11 @@ static const char *kRx = "rx";
static const char *kRxWOW = "rx/wow";
#endif
#ifdef XMRIG_ALGO_ARGON2
static const char *kArgon2 = "argon2";
static const char *kArgon2Impl = "argon2-impl";
#endif
extern template class Threads<CpuThreads>;
}
@ -94,6 +99,10 @@ rapidjson::Value xmrig::CpuConfig::toJSON(rapidjson::Document &doc) const
obj.AddMember(StringRef(kAsm), m_assembly.toJSON(), allocator);
# endif
# ifdef XMRIG_ALGO_ARGON2
obj.AddMember(StringRef(kArgon2Impl), m_argon2Impl.toJSON(), allocator);
# endif
m_threads.toJSON(obj, doc);
return obj;
@ -112,14 +121,14 @@ std::vector<xmrig::CpuLaunchData> xmrig::CpuConfig::get(const Miner *miner, cons
out.reserve(threads.count());
for (const CpuThread &thread : threads.data()) {
out.push_back(CpuLaunchData(miner, algorithm, *this, thread));
out.emplace_back(miner, algorithm, *this, thread);
}
return out;
}
void xmrig::CpuConfig::read(const rapidjson::Value &value)
void xmrig::CpuConfig::read(const rapidjson::Value &value, uint32_t version)
{
if (value.IsObject()) {
m_enabled = Json::getBool(value, kEnabled, m_enabled);
@ -132,9 +141,17 @@ void xmrig::CpuConfig::read(const rapidjson::Value &value)
m_assembly = Json::getValue(value, kAsm);
# endif
# ifdef XMRIG_ALGO_ARGON2
m_argon2Impl = Json::getString(value, kArgon2Impl);
# endif
if (!m_threads.read(value)) {
generate();
}
if (version == 0) {
generateArgon2();
}
}
else if (value.IsBool() && value.IsFalse()) {
m_enabled = false;
@ -174,6 +191,16 @@ void xmrig::CpuConfig::generate()
m_threads.move(kRx, cpu->threads(Algorithm::RX_0));
m_threads.move(kRxWOW, cpu->threads(Algorithm::RX_WOW));
# endif
generateArgon2();
}
void xmrig::CpuConfig::generateArgon2()
{
# ifdef XMRIG_ALGO_ARGON2
m_threads.move(kArgon2, Cpu::info()->threads(Algorithm::AR2_CHUKWA));
# endif
}

5
src/backend/cpu/CpuConfig.h
View file

@ -49,17 +49,19 @@ public:
bool isHwAES() const;
rapidjson::Value toJSON(rapidjson::Document &doc) const;
std::vector<CpuLaunchData> get(const Miner *miner, const Algorithm &algorithm) const;
void read(const rapidjson::Value &value);
void read(const rapidjson::Value &value, uint32_t version);
inline bool isEnabled() const { return m_enabled; }
inline bool isHugePages() const { return m_hugePages; }
inline bool isShouldSave() const { return m_shouldSave; }
inline const Assembly &assembly() const { return m_assembly; }
inline const String &argon2Impl() const { return m_argon2Impl; }
inline const Threads<CpuThreads> &threads() const { return m_threads; }
inline int priority() const { return m_priority; }
private:
void generate();
void generateArgon2();
void setAesMode(const rapidjson::Value &aesMode);
inline void setPriority(int priority) { m_priority = (priority >= -1 && priority <= 5) ? priority : -1; }
@ -70,6 +72,7 @@ private:
bool m_hugePages = true;
bool m_shouldSave = false;
int m_priority = -1;
String m_argon2Impl;
Threads<CpuThreads> m_threads;
};

7
src/backend/cpu/CpuLaunchData.cpp
View file

@ -24,6 +24,9 @@
*/
#include <algorithm>
#include "backend/cpu/CpuLaunchData.h"
#include "backend/cpu/CpuConfig.h"
@ -33,10 +36,10 @@ xmrig::CpuLaunchData::CpuLaunchData(const Miner *miner, const Algorithm &algorit
assembly(config.assembly()),
hugePages(config.isHugePages()),
hwAES(config.isHwAES()),
intensity(thread.intensity()),
priority(config.priority()),
affinity(thread.affinity()),
miner(miner)
miner(miner),
intensity(std::min<uint32_t>(thread.intensity(), algorithm.maxIntensity()))
{
}

2
src/backend/cpu/CpuLaunchData.h
View file

@ -58,10 +58,10 @@ public:
const Assembly assembly;
const bool hugePages;
const bool hwAES;
const int intensity;
const int priority;
const int64_t affinity;
const Miner *miner;
const uint32_t intensity;
};

6
src/backend/cpu/CpuThread.cpp
View file

@ -31,11 +31,11 @@
xmrig::CpuThread::CpuThread(const rapidjson::Value &value)
{
if (value.IsArray() && value.Size() >= 2) {
m_intensity = value[0].GetInt();
m_intensity = value[0].GetUint();
m_affinity = value[1].GetInt();
}
else if (value.IsInt()) {
m_intensity = -1;
m_intensity = 0;
m_affinity = value.GetInt();
}
}
@ -44,7 +44,7 @@ xmrig::CpuThread::CpuThread(const rapidjson::Value &value)
rapidjson::Value xmrig::CpuThread::toJSON(rapidjson::Document &doc) const
{
using namespace rapidjson;
if (m_intensity == -1) {
if (m_intensity == 0) {
return Value(m_affinity);
}

12
src/backend/cpu/CpuThread.h
View file

@ -35,15 +35,15 @@ namespace xmrig {
class CpuThread
{
public:
inline constexpr CpuThread() {}
inline constexpr CpuThread(int64_t affinity, int intensity) : m_intensity(intensity), m_affinity(affinity) {}
inline constexpr CpuThread() = default;
inline constexpr CpuThread(int64_t affinity, uint32_t intensity) : m_affinity(affinity), m_intensity(intensity) {}
CpuThread(const rapidjson::Value &value);
inline bool isEqual(const CpuThread &other) const { return other.m_affinity == m_affinity && other.m_intensity == m_intensity; }
inline bool isValid() const { return m_intensity == -1 || (m_intensity >= 1 && m_intensity <= 5); }
inline int intensity() const { return m_intensity == -1 ? 1 : m_intensity; }
inline bool isValid() const { return m_intensity <= 5; }
inline int64_t affinity() const { return m_affinity; }
inline uint32_t intensity() const { return m_intensity == 0 ? 1 : m_intensity; }
inline bool operator!=(const CpuThread &other) const { return !isEqual(other); }
inline bool operator==(const CpuThread &other) const { return isEqual(other); }
@ -51,8 +51,8 @@ public:
rapidjson::Value toJSON(rapidjson::Document &doc) const;
private:
int m_intensity = -1;
int64_t m_affinity = -1;
int64_t m_affinity = -1;
uint32_t m_intensity = 0;
};

4
src/backend/cpu/CpuThreads.cpp
View file

@ -94,7 +94,7 @@ xmrig::CpuThreads::CpuThreads(const rapidjson::Value &value)
}
}
else if (value.IsObject()) {
int intensity = Json::getInt(value, kIntensity, 1);
uint32_t intensity = Json::getUint(value, kIntensity, 1);
const size_t threads = std::min<unsigned>(Json::getUint(value, kThreads), 1024);
m_affinity = getAffinityMask(Json::getValue(value, kAffinity));
m_format = ObjectFormat;
@ -110,7 +110,7 @@ xmrig::CpuThreads::CpuThreads(const rapidjson::Value &value)
}
xmrig::CpuThreads::CpuThreads(size_t count, int intensity)
xmrig::CpuThreads::CpuThreads(size_t count, uint32_t intensity)
{
m_data.reserve(count);

4
src/backend/cpu/CpuThreads.h
View file

@ -42,13 +42,13 @@ public:
inline CpuThreads(size_t count) : m_data(count) {}
CpuThreads(const rapidjson::Value &value);
CpuThreads(size_t count, int intensity);
CpuThreads(size_t count, uint32_t intensity);
inline bool isEmpty() const { return m_data.empty(); }
inline const std::vector<CpuThread> &data() const { return m_data; }
inline size_t count() const { return m_data.size(); }
inline void add(CpuThread &&thread) { m_data.push_back(thread); }
inline void add(int64_t affinity, int intensity) { add(CpuThread(affinity, intensity)); }
inline void add(int64_t affinity, uint32_t intensity) { add(CpuThread(affinity, intensity)); }
inline void reserve(size_t capacity) { m_data.reserve(capacity); }
rapidjson::Value toJSON(rapidjson::Document &doc) const;

26
src/backend/cpu/CpuWorker.cpp
View file

@ -24,7 +24,7 @@
*/
#include <assert.h>
#include <cassert>
#include <thread>
@ -46,15 +46,15 @@
namespace xmrig {
static constexpr uint32_t kReserveCount = 4096;
static constexpr uint32_t kReserveCount = 32768;
} // namespace xmrig
template<size_t N>
xmrig::CpuWorker<N>::CpuWorker(size_t index, const CpuLaunchData &data) :
Worker(index, data.affinity, data.priority),
xmrig::CpuWorker<N>::CpuWorker(size_t id, const CpuLaunchData &data) :
Worker(id, data.affinity, data.priority),
m_algorithm(data.algorithm),
m_assembly(data.assembly),
m_hwAES(data.hwAES),
@ -90,6 +90,8 @@ void xmrig::CpuWorker<N>::allocateRandomX_VM()
if (Nonce::sequence(Nonce::CPU) == 0) {
return;
}
dataset = Rx::dataset(m_job.currentJob(), m_node);
}
if (!m_vm) {
@ -118,7 +120,6 @@ bool xmrig::CpuWorker<N>::selfTest()
verify(Algorithm::CN_XAO, test_output_xao) &&
verify(Algorithm::CN_RTO, test_output_rto) &&
verify(Algorithm::CN_HALF, test_output_half) &&
verify2(Algorithm::CN_WOW, test_output_wow) &&
verify2(Algorithm::CN_R, test_output_r) &&
verify(Algorithm::CN_RWZ, test_output_rwz) &&
verify(Algorithm::CN_ZLS, test_output_zls) &&
@ -156,6 +157,13 @@ bool xmrig::CpuWorker<N>::selfTest()
}
# endif
# ifdef XMRIG_ALGO_ARGON2
if (m_algorithm.family() == Algorithm::ARGON2) {
return verify(Algorithm::AR2_CHUKWA, argon2_chukwa_test_out) &&
verify(Algorithm::AR2_WRKZ, argon2_wrkz_test_out);
}
# endif
return false;
}
@ -200,11 +208,11 @@ void xmrig::CpuWorker<N>::start()
for (size_t i = 0; i < N; ++i) {
if (*reinterpret_cast<uint64_t*>(m_hash + (i * 32) + 24) < job.target()) {
JobResults::submit(JobResult(job, *m_job.nonce(i), m_hash + (i * 32)));
JobResults::submit(job, *m_job.nonce(i), m_hash + (i * 32));
}
}
m_job.nextRound(kReserveCount);
m_job.nextRound(kReserveCount, 1);
m_count += N;
std::this_thread::yield();
@ -291,6 +299,10 @@ void xmrig::CpuWorker<N>::allocateCnCtx()
template<size_t N>
void xmrig::CpuWorker<N>::consumeJob()
{
if (Nonce::sequence(Nonce::CPU) == 0) {
return;
}
m_job.add(m_miner->job(), Nonce::sequence(Nonce::CPU), kReserveCount);
# ifdef XMRIG_ALGO_RANDOMX

8
src/backend/cpu/CpuWorker.h
View file

@ -30,7 +30,7 @@
#include "backend/common/Worker.h"
#include "backend/common/WorkerJob.h"
#include "backend/cpu/CpuLaunchData.h"
#include "base/net/stratum/Job.h"
#include "base/tools/Object.h"
#include "net/JobResult.h"
@ -44,7 +44,9 @@ template<size_t N>
class CpuWorker : public Worker
{
public:
CpuWorker(size_t index, const CpuLaunchData &data);
XMRIG_DISABLE_COPY_MOVE_DEFAULT(CpuWorker)
CpuWorker(size_t id, const CpuLaunchData &data);
~CpuWorker() override;
protected:
@ -71,7 +73,7 @@ private:
const CnHash::AlgoVariant m_av;
const Miner *m_miner;
cryptonight_ctx *m_ctx[N];
uint8_t m_hash[N * 32];
uint8_t m_hash[N * 32]{ 0 };
VirtualMemory *m_memory = nullptr;
WorkerJob<N> m_job;

2
src/backend/cpu/platform/AdvancedCpuInfo.cpp
View file

@ -145,7 +145,7 @@ xmrig::CpuThreads xmrig::AdvancedCpuInfo::threads(const Algorithm &algorithm) co
count = threads() / 2;
}
int intensity = algorithm.maxIntensity() == 1 ? -1 : 1;
uint32_t intensity = algorithm.maxIntensity() == 1 ? 0 : 1;
# ifdef XMRIG_ALGO_CN_PICO
if (algorithm == Algorithm::CN_PICO_0 && (count / cores()) >= 2) {

7
src/backend/cpu/platform/BasicCpuInfo.cpp
View file

@ -52,7 +52,6 @@
#define VENDOR_ID (0)
#define PROCESSOR_INFO (1)
#define CACHE_TLB_DESCRIPTOR (2)
#define EXTENDED_FEATURES (7)
#define PROCESSOR_BRAND_STRING_1 (0x80000002)
#define PROCESSOR_BRAND_STRING_2 (0x80000003)
@ -222,5 +221,11 @@ xmrig::CpuThreads xmrig::BasicCpuInfo::threads(const Algorithm &algorithm) const
}
# endif
# ifdef XMRIG_ALGO_ARGON2
if (algorithm.family() == Algorithm::ARGON2) {
return count;
}
# endif
return CpuThreads(std::max<size_t>(count / 2, 1), 1);
}

2
src/backend/cpu/platform/HwlocCpuInfo.cpp
View file

@ -251,7 +251,7 @@ void xmrig::HwlocCpuInfo::processTopLevelCache(hwloc_obj_t cache, const Algorith
int L2_associativity = 0;
size_t extra = 0;
const size_t scratchpad = algorithm.l3();
int intensity = algorithm.maxIntensity() == 1 ? -1 : 1;
uint32_t intensity = algorithm.maxIntensity() == 1 ? 0 : 1;
if (cache->attr->cache.depth == 3) {
for (size_t i = 0; i < cache->arity; ++i) {

403
src/backend/opencl/OclBackend.cpp Normal file
View file

@ -0,0 +1,403 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* 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/>.
*/
#include <mutex>
#include "backend/common/Hashrate.h"
#include "backend/common/interfaces/IWorker.h"
#include "backend/common/Tags.h"
#include "backend/common/Workers.h"
#include "backend/opencl/OclBackend.h"
#include "backend/opencl/OclConfig.h"
#include "backend/opencl/OclLaunchData.h"
#include "backend/opencl/OclWorker.h"
#include "backend/opencl/wrappers/OclContext.h"
#include "backend/opencl/wrappers/OclLib.h"
#include "base/io/log/Log.h"
#include "base/net/stratum/Job.h"
#include "base/tools/Chrono.h"
#include "base/tools/String.h"
#include "core/config/Config.h"
#include "core/Controller.h"
#include "rapidjson/document.h"
#ifdef XMRIG_FEATURE_API
# include "base/api/interfaces/IApiRequest.h"
#endif
namespace xmrig {
extern template class Threads<OclThreads>;
constexpr const size_t oneMiB = 1024u * 1024u;
static const char *tag = MAGENTA_BG_BOLD(WHITE_BOLD_S " ocl ");
static const String kType = "opencl";
static std::mutex mutex;
static void printDisabled(const char *reason)
{
Log::print(GREEN_BOLD(" * ") WHITE_BOLD("%-13s") RED_BOLD("disabled") "%s", "OPENCL", reason);
}
struct OclLaunchStatus
{
public:
inline bool started() { m_started++; return m_started == m_threads; }
inline size_t threads() const { return m_threads; }
inline void start(size_t threads)
{
m_started = 0;
m_threads = threads;
m_ts = Chrono::steadyMSecs();
OclWorker::ready = false;
}
inline void print() const
{
LOG_INFO("%s" GREEN_BOLD(" READY") " threads " CYAN_BOLD("%zu") BLACK_BOLD(" (%" PRIu64 " ms)"),
tag,
m_threads,
Chrono::steadyMSecs() - m_ts
);
}
private:
size_t m_started = 0;
size_t m_threads = 0;
uint64_t m_ts = 0;
};
class OclBackendPrivate
{
public:
inline OclBackendPrivate(Controller *controller) :
controller(controller)
{
init(controller->config()->cl());
}
void init(const OclConfig &cl)
{
if (!cl.isEnabled()) {
return printDisabled("");
}
if (!OclLib::init(cl.loader())) {
return printDisabled(RED_S " (failed to load OpenCL runtime)");
}
platform = cl.platform();
if (!platform.isValid()) {
return printDisabled(RED_S " (selected OpenCL platform NOT found)");
}
devices = platform.devices();
if (devices.empty()) {
return printDisabled(RED_S " (no devices)");
}
Log::print(GREEN_BOLD(" * ") WHITE_BOLD("%-13s") CYAN_BOLD("#%zu ") WHITE_BOLD("%s") "/" WHITE_BOLD("%s"), "OPENCL", platform.index(), platform.name().data(), platform.version().data());
for (const OclDevice &device : devices) {
Log::print(GREEN_BOLD(" * ") WHITE_BOLD("%-13s") CYAN_BOLD("#%zu") YELLOW(" %s") " %s " WHITE_BOLD("%uMHz") " cu:" WHITE_BOLD("%u") " mem:" CYAN("%zu/%zu") " MB", "OPENCL GPU",
device.index(),
device.hasTopology() ? device.topology().toString().data() : "n/a",
device.printableName().data(),
device.clock(),
device.computeUnits(),
device.freeMemSize() / oneMiB,
device.globalMemSize() / oneMiB);
}
}
inline void start()
{
LOG_INFO("%s use profile " BLUE_BG(WHITE_BOLD_S " %s ") WHITE_BOLD_S " (" CYAN_BOLD("%zu") WHITE_BOLD(" threads)") " scratchpad " CYAN_BOLD("%zu KB"),
tag,
profileName.data(),
threads.size(),
algo.l3() / 1024
);
Log::print(WHITE_BOLD("| # | GPU | BUS ID | I | W | SI | MC | U | MEM | NAME"));
size_t i = 0;
for (const auto &data : threads) {
Log::print("|" CYAN_BOLD("%3zu") " |" CYAN_BOLD("%4u") " |" YELLOW(" %7s") " |" CYAN_BOLD("%5u") " |" CYAN_BOLD("%3u") " |"
CYAN_BOLD("%3u") " |" CYAN_BOLD("%3s") " |" CYAN_BOLD("%3u") " |" CYAN("%5zu") " | %s",
i,
data.thread.index(),
data.device.hasTopology() ? data.device.topology().toString().data() : "n/a",
data.thread.intensity(),
data.thread.worksize(),
data.thread.stridedIndex(),
data.thread.stridedIndex() == 2 ? std::to_string(data.thread.memChunk()).c_str() : "-",
data.thread.unrollFactor(),
data.thread.intensity() * algo.l3() / oneMiB,
data.device.printableName().data()
);
i++;
}
status.start(threads.size());
workers.start(threads);
}
Algorithm algo;
Controller *controller;
OclContext context;
OclLaunchStatus status;
OclPlatform platform;
std::vector<OclDevice> devices;
std::vector<OclLaunchData> threads;
String profileName;
Workers<OclLaunchData> workers;
};
} // namespace xmrig
const char *xmrig::ocl_tag()
{
return tag;
}
xmrig::OclBackend::OclBackend(Controller *controller) :
d_ptr(new OclBackendPrivate(controller))
{
d_ptr->workers.setBackend(this);
}
xmrig::OclBackend::~OclBackend()
{
delete d_ptr;
OclLib::close();
}
bool xmrig::OclBackend::isEnabled() const
{
return d_ptr->controller->config()->cl().isEnabled() && OclLib::isInitialized() && d_ptr->platform.isValid() && !d_ptr->devices.empty();
}
bool xmrig::OclBackend::isEnabled(const Algorithm &algorithm) const
{
return !d_ptr->controller->config()->cl().threads().get(algorithm).isEmpty();
}
const xmrig::Hashrate *xmrig::OclBackend::hashrate() const
{
return d_ptr->workers.hashrate();
}
const xmrig::String &xmrig::OclBackend::profileName() const
{
return d_ptr->profileName;
}
const xmrig::String &xmrig::OclBackend::type() const
{
return kType;
}
void xmrig::OclBackend::prepare(const Job &)
{
}
void xmrig::OclBackend::printHashrate(bool details)
{
if (!details || !hashrate()) {
return;
}
char num[8 * 3] = { 0 };
Log::print(WHITE_BOLD_S "| OPENCL # | AFFINITY | 10s H/s | 60s H/s | 15m H/s |");
size_t i = 0;
for (const OclLaunchData &data : d_ptr->threads) {
Log::print("| %8zu | %8" PRId64 " | %7s | %7s | %7s |" CYAN_BOLD(" #%u") YELLOW(" %s") " %s",
i,
data.affinity,
Hashrate::format(hashrate()->calc(i, Hashrate::ShortInterval), num, sizeof num / 3),
Hashrate::format(hashrate()->calc(i, Hashrate::MediumInterval), num + 8, sizeof num / 3),
Hashrate::format(hashrate()->calc(i, Hashrate::LargeInterval), num + 8 * 2, sizeof num / 3),
data.device.index(),
data.device.hasTopology() ? data.device.topology().toString().data() : "n/a",
data.device.printableName().data()
);
i++;
}
Log::print(WHITE_BOLD_S "| - | - | %7s | %7s | %7s |",
Hashrate::format(hashrate()->calc(Hashrate::ShortInterval), num, sizeof num / 3),
Hashrate::format(hashrate()->calc(Hashrate::MediumInterval), num + 8, sizeof num / 3),
Hashrate::format(hashrate()->calc(Hashrate::LargeInterval), num + 8 * 2, sizeof num / 3)
);
}
void xmrig::OclBackend::setJob(const Job &job)
{
if (!isEnabled()) {
return stop();
}
const OclConfig &cl = d_ptr->controller->config()->cl();
std::vector<OclLaunchData> threads = cl.get(d_ptr->controller->miner(), job.algorithm(), d_ptr->platform, d_ptr->devices, tag);
if (!d_ptr->threads.empty() && d_ptr->threads.size() == threads.size() && std::equal(d_ptr->threads.begin(), d_ptr->threads.end(), threads.begin())) {
return;
}
d_ptr->algo = job.algorithm();
d_ptr->profileName = cl.threads().profileName(job.algorithm());
if (d_ptr->profileName.isNull() || threads.empty()) {
LOG_WARN("%s " RED_BOLD("disabled") YELLOW(" (no suitable configuration found)"), tag);
return stop();
}
if (!d_ptr->context.init(d_ptr->devices, threads, job)) {
LOG_WARN("%s " RED_BOLD("disabled") YELLOW(" (OpenCL context unavailable)"), tag);
return stop();
}
stop();
d_ptr->threads = std::move(threads);
d_ptr->start();
}
void xmrig::OclBackend::start(IWorker *worker)
{
mutex.lock();
if (d_ptr->status.started()) {
d_ptr->status.print();
OclWorker::ready = true;
}
mutex.unlock();
worker->start();
}
void xmrig::OclBackend::stop()
{
if (d_ptr->threads.empty()) {
return;
}
const uint64_t ts = Chrono::steadyMSecs();
d_ptr->workers.stop();
d_ptr->threads.clear();
LOG_INFO("%s" YELLOW(" stopped") BLACK_BOLD(" (%" PRIu64 " ms)"), tag, Chrono::steadyMSecs() - ts);
}
void xmrig::OclBackend::tick(uint64_t ticks)
{
d_ptr->workers.tick(ticks);
}
#ifdef XMRIG_FEATURE_API
rapidjson::Value xmrig::OclBackend::toJSON(rapidjson::Document &doc) const
{
using namespace rapidjson;
auto &allocator = doc.GetAllocator();
Value out(kObjectType);
out.AddMember("type", type().toJSON(), allocator);
out.AddMember("enabled", isEnabled(), allocator);
out.AddMember("algo", d_ptr->algo.toJSON(), allocator);
out.AddMember("profile", profileName().toJSON(), allocator);
out.AddMember("platform", d_ptr->platform.toJSON(doc), allocator);
if (d_ptr->threads.empty() || !hashrate()) {
return out;
}
Value threads(kArrayType);
const Hashrate *hr = hashrate();
size_t i = 0;
for (const OclLaunchData &data : d_ptr->threads) {
Value thread = data.thread.toJSON(doc);
thread.AddMember("affinity", data.affinity, allocator);
Value hashrate(kArrayType);
hashrate.PushBack(Hashrate::normalize(hr->calc(i, Hashrate::ShortInterval)), allocator);
hashrate.PushBack(Hashrate::normalize(hr->calc(i, Hashrate::MediumInterval)), allocator);
hashrate.PushBack(Hashrate::normalize(hr->calc(i, Hashrate::LargeInterval)), allocator);
i++;
thread.AddMember("hashrate", hashrate, allocator);
threads.PushBack(thread, allocator);
}
out.AddMember("threads", threads, allocator);
return out;
}
void xmrig::OclBackend::handleRequest(IApiRequest &)
{
}
#endif

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