Update CHANGELOG.md

# Conflicts:
#	src/version.h

.gitignore

@@ -1 +1,2 @@
 /build
+/CMakeLists.txt.user

CHANGELOG.md

@@ -1,6 +1,62 @@
+# v2.1.0
+- [#40](https://github.com/xmrig/xmrig/issues/40) Improved miner shutdown, fixed crash on exit for Linux and OS X.
+- Fixed, login request was contain malformed JSON if username or password has some special characters for example `\`. 
+- [#220](https://github.com/fireice-uk/xmr-stak-cpu/pull/220) Better support for Round Robin DNS, IP address now always chosen randomly instead of stuck on first one.
+- Changed donation address, new [xmrig-proxy](https://github.com/xmrig/xmrig-proxy) is coming soon.
+
+# v2.0.2
+- Better deal with possible duplicate jobs from pool, show warning and ignore duplicates.
+- For Windows builds libuv updated to version 1.13.1 and gcc to 7.1.0.
+
+# v2.0.1
+ - [#27](https://github.com/xmrig/xmrig/issues/27) Fixed possibility crash on 32bit systems.
+
+# v2.0.0
+ - Option `--backup-url` removed, instead now possibility specify multiple pools for example: `-o example1.com:3333 -u user1 -p password1 -k -o example2.com:5555 -u user2 -o example3.com:4444 -u user3`
+ - [#15](https://github.com/xmrig/xmrig/issues/15) Added option `-l, --log-file=FILE` to write log to file.
+ - [#15](https://github.com/xmrig/xmrig/issues/15) Added option `-S, --syslog` to use syslog for logging, Linux only.
+ - [#18](https://github.com/xmrig/xmrig/issues/18) Added nice messages for accepted/rejected shares with diff and network latency.
+ - [#20](https://github.com/xmrig/xmrig/issues/20) Fixed `--cpu-affinity` for more than 32 threads.
+ - Fixed Windows XP support.
+ - Fixed regression, option `--no-color` was not fully disable colored output.
+ - Show resolved pool IP address in miner output.
+ 
+# v1.0.1
+- Fix broken software AES implementation, app has crashed if CPU not support AES-NI, only version 1.0.0 affected.
+
+# v1.0.0
+- Miner complete rewritten in C++ with libuv.
+- This version should be fully compatible (except config file) with previos versions, many new nice features will come in next versions.
+- This is still beta. If you found regression, stability or perfomance issues or have an idea for new feature please fell free to open new [issue](https://github.com/xmrig/xmrig/issues/new).
+- Added new option `--print-time=N`, print hashrate report every N seconds.
+- New hashrate reports, by default every 60 secons.
+- Added Microsoft Visual C++ 2015 and 2017 support.
+- Removed dependency on libcurl.
+- To compile this version from source please switch to [dev](https://github.com/xmrig/xmrig/tree/dev) branch.
+
+# v0.8.2
+- Fixed L2 cache size detection for AMD CPUs (Bulldozer/Piledriver/Steamroller/Excavator architecture).
+
+# v0.8.2
+- Fixed L2 cache size detection for AMD CPUs (Bulldozer/Piledriver/Steamroller/Excavator architecture).
+- Fixed gcc 7.1 support.
+
+# v0.8.1
+- Added nicehash support, detects automaticaly by pool URL, for example `cryptonight.eu.nicehash.com:3355` or manually via option `--nicehash`.
+
+# v0.8.0
+- Added double hash mode, also known as lower power mode. `--av=2` and `--av=4`.
+- Added smart automatic CPU configuration. Default threads count now depends on size of the L3 cache of CPU.
+- Added CryptoNight-Lite support for AEON `-a cryptonight-lite`.
+- Added `--max-cpu-usage` option for auto CPU configuration mode.
+- Added `--safe` option for adjust threads and algorithm variations to current CPU.
+- No more manual steps to enable huge pages on Windows. XMRig will do it automatically.
+- Removed BMI2 algorithm variation.
+- Removed default pool URL.
+
 # v0.6.0
 - Added automatic cryptonight self test.
-- New software AES algorithm variation `--av=4`. Will be automatically selected if cpu not support AES-NI.
+- New software AES algorithm variation. Will be automatically selected if cpu not support AES-NI.
 - Added 32 bit builds.
 - Documented [algorithm variations](https://github.com/xmrig/xmrig#algorithm-variations).
 

CMakeLists.txt

@@ -1,122 +1,197 @@
 cmake_minimum_required(VERSION 3.0)
-project(xmrig C)
+project(xmrig)
+
+option(WITH_LIBCPUID "Use Libcpuid" ON)
+option(WITH_AEON     "CryptoNight-Lite support" ON)
+
+
+include (CheckIncludeFile)
+
 
 set(HEADERS
-    compat.h
-    algo/cryptonight/cryptonight.h
-    algo/cryptonight/cryptonight_p.h
-    elist.h
-    xmrig.h
-    version.h
-    options.h
-    cpu.h
-    persistent_memory.h
-    stratum.h
-    stats.h
-    util.h
-    donate.h
+    src/3rdparty/align.h
+    src/App.h
+    src/Cpu.h
+    src/interfaces/IClientListener.h
+    src/interfaces/IJobResultListener.h
+    src/interfaces/ILogBackend.h
+    src/interfaces/IStrategy.h
+    src/interfaces/IStrategyListener.h
+    src/interfaces/IWorker.h
+    src/log/ConsoleLog.h
+    src/log/FileLog.h
+    src/log/Log.h
+    src/Mem.h
+    src/net/Client.h
+    src/net/Job.h
+    src/net/JobResult.h
+    src/net/Network.h
+    src/net/SubmitResult.h
+    src/net/Url.h
+    src/net/strategies/DonateStrategy.h
+    src/net/strategies/FailoverStrategy.h
+    src/net/strategies/SinglePoolStrategy.h
+    src/Options.h
+    src/Summary.h
+    src/version.h
+    src/workers/DoubleWorker.h
+    src/workers/Handle.h
+    src/workers/Hashrate.h
+    src/workers/SingleWorker.h
+    src/workers/Worker.h
+    src/workers/Workers.h
    )
 
 set(HEADERS_CRYPTO
-    crypto/c_groestl.h
-    crypto/c_blake256.h
-    crypto/c_jh.h
-    crypto/c_skein.h
-   )
-
-set(HEADERS_COMPAT
-    compat/winansi.h
-   )
-
-set(HEADERS_UTILS
-    utils/applog.h
-    utils/threads.h
-    utils/summary.h
+    src/crypto/c_blake256.h
+    src/crypto/c_groestl.h
+    src/crypto/c_jh.h
+    src/crypto/c_keccak.h
+    src/crypto/c_skein.h
+    src/crypto/CryptoNight.h
+    src/crypto/CryptoNight_p.h
+    src/crypto/CryptoNight_test.h
+    src/crypto/groestl_tables.h
+    src/crypto/hash.h
+    src/crypto/skein_port.h
    )
 
 set(SOURCES
-    xmrig.c
-    algo/cryptonight/cryptonight.c
-    util.c
-    options.c
-    cpu.c
-    stratum.c
-    stats.c
-    memory.c
+    src/App.cpp
+    src/log/ConsoleLog.cpp
+    src/log/FileLog.cpp
+    src/log/Log.cpp
+    src/Mem.cpp
+    src/net/Client.cpp
+    src/net/Job.cpp
+    src/net/Network.cpp
+    src/net/Url.cpp
+    src/net/strategies/DonateStrategy.cpp
+    src/net/strategies/FailoverStrategy.cpp
+    src/net/strategies/SinglePoolStrategy.cpp
+    src/Options.cpp
+    src/Summary.cpp
+    src/workers/DoubleWorker.cpp
+    src/workers/Handle.cpp
+    src/workers/Hashrate.cpp
+    src/workers/SingleWorker.cpp
+    src/workers/Worker.cpp
+    src/workers/Workers.cpp
+    src/xmrig.cpp
    )
 
 set(SOURCES_CRYPTO
-    crypto/c_keccak.c
-    crypto/c_groestl.c
-    crypto/c_blake256.c
-    crypto/c_jh.c
-    crypto/c_skein.c
-    crypto/soft_aes.c
-   )
-
-set(SOURCES_UTILS
-    utils/applog.c
-    utils/summary.c
+    src/crypto/c_keccak.c
+    src/crypto/c_groestl.c
+    src/crypto/c_blake256.c
+    src/crypto/c_jh.c
+    src/crypto/c_skein.c
+    src/crypto/soft_aes.c
+    src/crypto/soft_aes.c
+    src/crypto/CryptoNight.cpp
    )
 
 if (WIN32)
-    set(SOURCES_OS win/cpu_win.c win/memory_win.c win/xmrig_win.c win/app.rc compat/winansi.c)
-    set(EXTRA_LIBS ws2_32)
-    add_definitions(/D_WIN32_WINNT=0x600)
+    set(SOURCES_OS
+        res/app.rc
+        src/3rdparty/winansi.cpp
+        src/3rdparty/winansi.h
+        src/App_win.cpp
+        src/Cpu_win.cpp
+        src/Mem_win.cpp
+        src/net/Network_win.cpp
+        )
+
+    add_definitions(/DWIN32)
+    set(EXTRA_LIBS ws2_32 psapi iphlpapi userenv)
+elseif (APPLE)
+    set(SOURCES_OS
+        src/App_unix.cpp
+        src/Cpu_mac.cpp
+        src/Mem_unix.cpp
+        src/net/Network_mac.cpp
+        )
 else()
-    set(SOURCES_OS unix/cpu_unix.c unix/memory_unix.c unix/xmrig_unix.c)
+    set(SOURCES_OS
+        src/App_unix.cpp
+        src/Cpu_unix.cpp
+        src/Mem_unix.cpp
+        src/net/Network_unix.cpp
+        )
+
     set(EXTRA_LIBS pthread)
 endif()
 
-include_directories(.)
-add_definitions(/DUSE_NATIVE_THREADS)
-add_definitions(/D_GNU_SOURCE)
+add_definitions(/DUNICODE)
+#add_definitions(/DAPP_DEBUG)
+
+set(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} "${CMAKE_SOURCE_DIR}/cmake")
+
+find_package(UV REQUIRED)
 
 if ("${CMAKE_BUILD_TYPE}" STREQUAL "")
     set(CMAKE_BUILD_TYPE Release)
 endif()
 
-set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -maes -Wno-pointer-to-int-cast")
-set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Ofast -funroll-loops -fvariable-expansion-in-unroller -ftree-loop-if-convert-stores -fmerge-all-constants -fbranch-target-load-optimize2")
-#set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -gdwarf-2")
-#set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -fprofile-generate")
-#set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -fprofile-use -fprofile-correction")
 
-if (WIN32)
-    set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -static")
+# https://cmake.org/cmake/help/latest/variable/CMAKE_LANG_COMPILER_ID.html
+if (CMAKE_CXX_COMPILER_ID MATCHES GNU)
+
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -maes -Wall -Wno-strict-aliasing")
+    set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Ofast -funroll-loops -fvariable-expansion-in-unroller -ftree-loop-if-convert-stores -fmerge-all-constants -fbranch-target-load-optimize2")
+
+    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -maes -Wall -std=c++14 -fno-exceptions -fno-rtti")
+    set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -Ofast -s -funroll-loops -fvariable-expansion-in-unroller -ftree-loop-if-convert-stores -fmerge-all-constants -fbranch-target-load-optimize2")
+
+    if (WIN32)
+        set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -static")
+        add_definitions(/D__STDC_FORMAT_MACROS)
+    else()
+        set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -static-libgcc -static-libstdc++")
+    endif()
+
+    add_definitions(/D_GNU_SOURCE)
+
+    #set(CMAKE_C_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -gdwarf-2")
+
+elseif (CMAKE_CXX_COMPILER_ID MATCHES MSVC)
+
+    set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} /Ox /Ot /Oi /MT /GL")
+    set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /Ox /Ot /Oi /MT /GL")
+
+elseif (CMAKE_CXX_COMPILER_ID MATCHES Clang)
+
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -maes -Wall")
+    set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Ofast -funroll-loops -fmerge-all-constants")
+
+    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -maes -Wall -std=c++14 -fno-exceptions -fno-rtti")
+    set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -Ofast -funroll-loops -fmerge-all-constants")
+
 endif()
 
-include_directories(compat/jansson)
-add_subdirectory(compat/jansson)
+if (WITH_LIBCPUID)
+    add_subdirectory(src/3rdparty/libcpuid)
 
-find_package(CURL REQUIRED)
-
-if (CURL_FOUND)
-    include_directories(${CURL_INCLUDE_DIRS})
-    add_definitions(/DCURL_STATICLIB)
-    link_directories(${CURL_LIBRARIES})
-endif()
-
-if (CMAKE_SIZEOF_VOID_P EQUAL 8)
-    add_subdirectory(algo/cryptonight/bmi2)
-
-    set(CRYPTONIGHT64
-        algo/cryptonight/cryptonight_av1_aesni.c
-        algo/cryptonight/cryptonight_av2_aesni_stak.c
-        algo/cryptonight/cryptonight_av4_softaes.c
-        algo/cryptonight/cryptonight_av5_aesni_experimental.c
-    )
-
-    add_executable(xmrig ${HEADERS} ${HEADERS_CRYPTO} ${SOURCES} ${SOURCES_CRYPTO} ${HEADERS_UTILS} ${SOURCES_UTILS} ${HEADERS_COMPAT} ${SOURCES_COMPAT} ${SOURCES_OS} ${CRYPTONIGHT64})
-    target_link_libraries(xmrig jansson curl cryptonight_av3_aesni_bmi2 ${EXTRA_LIBS})
+    include_directories(src/3rdparty/libcpuid)
+    set(CPUID_LIB cpuid)
+    set(SOURCES_CPUID src/Cpu.cpp)
 else()
-    set(CRYPTONIGHT32
-        algo/cryptonight/i686/cryptonight_av1_aesni.c
-        algo/cryptonight/i686/cryptonight_av4_softaes.c
-    )
-
-    add_executable(xmrig32 ${HEADERS} ${HEADERS_CRYPTO} ${SOURCES} ${SOURCES_CRYPTO} ${HEADERS_UTILS} ${SOURCES_UTILS} ${HEADERS_COMPAT} ${SOURCES_COMPAT} ${SOURCES_OS} ${CRYPTONIGHT32})
-    target_link_libraries(xmrig32 jansson curl ${EXTRA_LIBS})
+    add_definitions(/DXMRIG_NO_LIBCPUID)
+    set(SOURCES_CPUID src/Cpu_stub.cpp)
 endif()
 
-source_group("HEADERS" FILES ${HEADERS})
+CHECK_INCLUDE_FILE (syslog.h HAVE_SYSLOG_H)
+if (HAVE_SYSLOG_H)
+    add_definitions(/DHAVE_SYSLOG_H)
+    set(SOURCES_SYSLOG src/log/SysLog.h src/log/SysLog.cpp)
+endif()
+
+include_directories(src)
+include_directories(src/3rdparty)
+include_directories(src/3rdparty/jansson)
+include_directories(${UV_INCLUDE_DIR})
+
+add_subdirectory(src/3rdparty/jansson)
+
+add_executable(xmrig ${HEADERS} ${SOURCES} ${SOURCES_OS} ${SOURCES_CPUID} ${HEADERS_CRYPTO} ${SOURCES_CRYPTO} ${SOURCES_SYSLOG})
+target_link_libraries(xmrig jansson ${UV_LIBRARIES} ${EXTRA_LIBS} ${CPUID_LIB})

README.md

@@ -1,26 +1,30 @@
 # XMRig
 XMRig is high performance Monero (XMR) CPU miner, with the official full Windows support.
-Based on cpuminer-multi with heavy optimizations/rewrites and removing a lot of legacy code.
+Originally based on cpuminer-multi with heavy optimizations/rewrites and removing a lot of legacy code, since version 1.0.0 complete rewritten from scratch on C++.
 
-<img src="http://i.imgur.com/GdRDnAu.png" width="596" >
+<img src="https://i.imgur.com/OXoB10D.png" width="628" >
 
 #### Table of contents
 * [Features](#features)
 * [Download](#download)
 * [Usage](#usage)
 * [Algorithm variations](#algorithm-variations)
-* [Build](#build)
+* [Build](https://github.com/xmrig/xmrig/wiki/Build)
 * [Common Issues](#common-issues)
 * [Other information](#other-information)
-* [Donations](#Donations)
+* [Donations](#donations)
+* [Contacts](#contacts)
 
 ## Features
 * High performance (290+ H/s on i7 6700).
 * Official Windows support.
-* Small Windows executable, only 430 KB without dependencies.
+* Small Windows executable, only 535 KB without dependencies.
 * Support for backup (failover) mining server.
 * keepalived support.
 * Command line options compatible with cpuminer.
+* CryptoNight-Lite support for AEON.
+* Smart automatic [CPU configuration](https://github.com/xmrig/xmrig/wiki/Threads).
+* Nicehash support
 * It's open source software.
 
 ## Download
@@ -34,10 +38,16 @@ Based on cpuminer-multi with heavy optimizations/rewrites and removing a lot of
 xmrig.exe -o xmr-eu.dwarfpool.com:8005 -u YOUR_WALLET -p x -k
 ```
 
+### Failover
+```
+xmrig.exe -o pool.supportxmr.com:5555 -u YOUR_WALLET1 -k -o xmr-eu.dwarfpool.com:8005 -u YOUR_WALLET2 -p x -k
+```
+For failover you can add multiple pools, maximum count not limited.
+
 ### Options
 ```
+  -a, --algo=ALGO       cryptonight (default) or cryptonight-lite
   -o, --url=URL         URL of mining server
-  -b, --backup-url=URL  URL of backup mining server
   -O, --userpass=U:P    username:password pair for mining server
   -u, --user=USERNAME   username for mining server
   -p, --pass=PASSWORD   password for mining server
@@ -51,58 +61,27 @@ xmrig.exe -o xmr-eu.dwarfpool.com:8005 -u YOUR_WALLET -p x -k
       --donate-level=N  donate level, default 5% (5 minutes in 100 minutes)
   -B, --background      run the miner in the background
   -c, --config=FILE     load a JSON-format configuration file
+      --max-cpu-usage=N maximum cpu usage for automatic threads mode (default 75)
+      --safe            safe adjust threads and av settings for current cpu
+      --nicehash        enable nicehash support
+      --print-time=N    print hashrate report every N seconds
   -h, --help            display this help and exit
   -V, --version         output version information and exit
 ```
 
 ## Algorithm variations
-Since version 0.6.0.
-* `--av=1` Default for CPUs with hardware AES.
-* `--av=2` [XMR-Stak-CPU](https://github.com/fireice-uk/xmr-stak-cpu) algorithm.
-* `--av=3` Same as `1` but uses BMI2 instruction [mulx](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mulx_u64).
-* `--av=4` Software AES implementation.
-
-For 32 bit platform only available `1` and `4`.
-
-## Build
-### Ubuntu (Debian-based distros)
-```
-sudo apt-get install git build-essential cmake libcurl4-openssl-dev
-git clone https://github.com/xmrig/xmrig.git
-cd xmrig
-mkdir build
-cd build
-cmake .. -DCMAKE_BUILD_TYPE=Release
-make
-```
-
-### Windows
-It's complicated, you need [MSYS2](http://www.msys2.org/), custom libcurl build, and of course CMake too.
-
-Necessary MSYS2 packages:
-```
-pacman -Sy
-pacman -S mingw-w64-x86_64-gcc
-pacman -S make
-pacman -S mingw-w64-x86_64-cmake
-pacman -S mingw-w64-x86_64-pkg-config
-```
-Configure options for libcurl:
-```
-./configure --disable-shared --enable-optimize --enable-threaded-resolver --disable-libcurl-option --disable-ares --disable-rt --disable-ftp --disable-file --disable-ldap --disable-ldaps --disable-rtsp --disable-dict --disable-telnet --disable-tftp --disable-pop3 --disable-imap --disable-smb --disable-smtp --disable-gopher --disable-manual --disable-ipv6 --disable-sspi --disable-crypto-auth --disable-ntlm-wb --disable-tls-srp --disable-unix-sockets --without-zlib --without-winssl --without-ssl --without-libssh2 --without-nghttp2 --disable-cookies --without-ca-bundle --without-librtmp
-```
-CMake options:
-```
-cmake .. -G "Unix Makefiles" -DCMAKE_BUILD_TYPE=Release -DCURL_INCLUDE_DIR="c:\<path>\curl-7.53.1\include" -DCURL_LIBRARY="c:\<path>\curl-7.53.1\lib\.libs"
-```
+Since version 0.8.0.
+* `--av=1` For CPUs with hardware AES.
+* `--av=2` Lower power mode (double hash) of `1`.
+* `--av=3` Software AES implementation.
+* `--av=4` Lower power mode (double hash) of `3`.
 
 ## Common Issues
 ### HUGE PAGES unavailable
 * Run XMRig as Administrator.
-* Enable SeLockMemoryPrivilege. For Windows 7 pro, or Windows 8 and above see [this article](https://msdn.microsoft.com/en-gb/library/ms190730.aspx).
+* Since version 0.8.0 XMRig automatically enable SeLockMemoryPrivilege for current user, but reboot or sign out still required. [Manual instruction](https://msdn.microsoft.com/en-gb/library/ms190730.aspx).
 
 ## Other information
-* Now only support 64 bit operating systems (Windows/Linux).
 * No HTTP support, only stratum protocol support.
 * No TLS support.
 * Default donation 5% (5 minutes in 100 minutes) can be reduced to 1% via command line option `--donate-level`.

algo/cryptonight/bmi2/CMakeLists.txt

@@ -1,2 +0,0 @@
-set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mbmi2")
-add_library(cryptonight_av3_aesni_bmi2 STATIC ../cryptonight_av3_aesni_bmi2.c)

algo/cryptonight/cryptonight.c

@@ -1,158 +0,0 @@
-/* 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 2016-2017 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 <stdlib.h>
-#include <string.h>
-
-#ifndef BUILD_TEST
-#   include "xmrig.h"
-#endif
-
-#include "crypto/c_groestl.h"
-#include "crypto/c_blake256.h"
-#include "crypto/c_jh.h"
-#include "crypto/c_skein.h"
-#include "cryptonight.h"
-#include "options.h"
-#include "utils/applog.h"
-
-
-const static char test_input[76] = {
-    0x03, 0x05, 0xA0, 0xDB, 0xD6, 0xBF, 0x05, 0xCF, 0x16, 0xE5, 0x03, 0xF3, 0xA6, 0x6F, 0x78, 0x00,
-    0x7C, 0xBF, 0x34, 0x14, 0x43, 0x32, 0xEC, 0xBF, 0xC2, 0x2E, 0xD9, 0x5C, 0x87, 0x00, 0x38, 0x3B,
-    0x30, 0x9A, 0xCE, 0x19, 0x23, 0xA0, 0x96, 0x4B, 0x00, 0x00, 0x00, 0x08, 0xBA, 0x93, 0x9A, 0x62,
-    0x72, 0x4C, 0x0D, 0x75, 0x81, 0xFC, 0xE5, 0x76, 0x1E, 0x9D, 0x8A, 0x0E, 0x6A, 0x1C, 0x3F, 0x92,
-    0x4F, 0xDD, 0x84, 0x93, 0xD1, 0x11, 0x56, 0x49, 0xC0, 0x5E, 0xB6, 0x01
-};
-
-
-const static char test_output[32] = {
-    0x1A, 0x3F, 0xFB, 0xEE, 0x90, 0x9B, 0x42, 0x0D, 0x91, 0xF7, 0xBE, 0x6E, 0x5F, 0xB5, 0x6D, 0xB7,
-    0x1B, 0x31, 0x10, 0xD8, 0x86, 0x01, 0x1E, 0x87, 0x7E, 0xE5, 0x78, 0x6A, 0xFD, 0x08, 0x01, 0x00
-};
-
-
-void cryptonight_av1_aesni(const void* input, size_t size, void* output, struct cryptonight_ctx* ctx);
-void cryptonight_av4_softaes(const void* input, size_t size, void* output, struct cryptonight_ctx* ctx);
-
-#if defined(__x86_64__)
-  void cryptonight_av2_aesni_stak(const void* input, size_t size, void* output, struct cryptonight_ctx* ctx);
-  void cryptonight_av3_aesni_bmi2(const void* input, size_t size, void* output, struct cryptonight_ctx* ctx);
-  void cryptonight_av5_aesni_experimental(const void* input, size_t size, void* output, struct cryptonight_ctx* ctx);
-#endif
-
-void (*cryptonight_hash_ctx)(const void* input, size_t size, void* output, struct cryptonight_ctx* ctx) = NULL;
-
-
-static bool self_test() {
-    char output[32];
-
-    struct cryptonight_ctx *ctx = (struct cryptonight_ctx*) malloc(sizeof(struct cryptonight_ctx));
-    ctx->memory = (uint8_t *) malloc(MEMORY);
-
-    cryptonight_hash_ctx(test_input, sizeof(test_input), output, ctx);
-
-    free(ctx->memory);
-    free(ctx);
-
-    return memcmp(output, test_output, 32) == 0;
-}
-
-
-bool cryptonight_init(int variant)
-{
-    switch (variant) {
-        case XMR_AV1_AESNI:
-            cryptonight_hash_ctx = cryptonight_av1_aesni;
-            break;
-
-#       if defined(__x86_64__)
-        case XMR_AV2_STAK:
-            cryptonight_hash_ctx = cryptonight_av2_aesni_stak;
-            break;
-
-        case XMR_AV3_AESNI_BMI2:
-            cryptonight_hash_ctx = cryptonight_av3_aesni_bmi2;
-            break;
-
-        case XMR_AV5_EXPERIMENTAL:
-            cryptonight_hash_ctx = cryptonight_av5_aesni_experimental;
-            break;
-#       endif
-
-        case XMR_AV4_SOFT_AES:
-             cryptonight_hash_ctx = cryptonight_av4_softaes;
-             break;
-
-        default:
-            break;
-    }
-
-    return self_test();
-}
-
-
-static inline void do_blake_hash(const void* input, size_t len, char* output) {
-    blake256_hash((uint8_t*)output, input, len);
-}
-
-
-static inline void do_groestl_hash(const void* input, size_t len, char* output) {
-    groestl(input, len * 8, (uint8_t*)output);
-}
-
-
-static inline void do_jh_hash(const void* input, size_t len, char* output) {
-    jh_hash(32 * 8, input, 8 * len, (uint8_t*)output);
-}
-
-
-static inline void do_skein_hash(const void* input, size_t len, char* output) {
-    skein_hash(8 * 32, input, 8 * len, (uint8_t*)output);
-}
-
-
-void (* const extra_hashes[4])(const void *, size_t, char *) = {do_blake_hash, do_groestl_hash, do_jh_hash, do_skein_hash};
-
-
-#ifndef BUILD_TEST
-int scanhash_cryptonight(int thr_id, uint32_t *hash, uint32_t *restrict blob, size_t blob_size, uint32_t target, uint32_t max_nonce, unsigned long *restrict hashes_done, struct cryptonight_ctx *restrict ctx) {
-    uint32_t *nonceptr = (uint32_t*) (((char*) blob) + 39);
-    uint32_t n = *nonceptr - 1;
-    const uint32_t first_nonce = n + 1;
-
-    do {
-        *nonceptr = ++n;
-        cryptonight_hash_ctx(blob, blob_size, hash, ctx);
-
-        if (unlikely(hash[7] < target)) {
-            *hashes_done = n - first_nonce + 1;
-            return true;
-        }
-    } while (likely((n <= max_nonce && !work_restart[thr_id].restart)));
-
-    *hashes_done = n - first_nonce + 1;
-    return 0;
-}
-#endif

algo/cryptonight/cryptonight_av1_aesni.c

@@ -1,77 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "cryptonight.h"
-#include "cryptonight_p.h"
-#include "crypto/c_keccak.h"
-
-
-void cryptonight_av1_aesni(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    keccak((const uint8_t *) input, size, ctx->state, 200);
-
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) ctx->memory);
-
-    const uint8_t* l0 = ctx->memory;
-    uint64_t* h0 = (uint64_t*) ctx->state;
-
-    uint64_t al0 = h0[0] ^ h0[4];
-    uint64_t ah0 = h0[1] ^ h0[5];
-    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
-
-    uint64_t idx0 = h0[0] ^ h0[4];
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i cx;
-        cx = _mm_load_si128((__m128i *) &l0[idx0 & 0x1FFFF0]);
-        cx = _mm_aesenc_si128(cx, _mm_set_epi64x(ah0, al0));
-
-        _mm_store_si128((__m128i *) &l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
-        idx0 = _mm_cvtsi128_si64(cx);
-        bx0 = cx;
-
-        uint64_t hi, lo, cl, ch;
-        cl = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[0];
-        ch = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[1];
-        lo = _umul128(idx0, cl, &hi);
-
-        al0 += hi;
-        ah0 += lo;
-
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
-
-        ah0 ^= ch;
-        al0 ^= cl;
-        idx0 = al0;
-    }
-
-    cn_implode_scratchpad((__m128i*) ctx->memory, (__m128i*) ctx->state);
-
-    keccakf(h0, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

algo/cryptonight/cryptonight_av2_aesni_stak.c

@@ -1,77 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "cryptonight.h"
-#include "cryptonight_p.h"
-#include "crypto/c_keccak.h"
-
-
-void cryptonight_av2_aesni_stak(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    keccak((const uint8_t *) input, size, ctx->state, 200);
-
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) ctx->memory);
-
-    const uint8_t* l0 = ctx->memory;
-    uint64_t* h0 = (uint64_t*) ctx->state;
-
-    uint64_t al0 = h0[0] ^ h0[4];
-    uint64_t ah0 = h0[1] ^ h0[5];
-    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
-
-    uint64_t idx0 = h0[0] ^ h0[4];
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i cx;
-        cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
-        cx = _mm_aesenc_si128(cx, _mm_set_epi64x(ah0, al0));
-
-        _mm_store_si128((__m128i *)&l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
-        idx0 = _mm_cvtsi128_si64(cx);
-        bx0 = cx;
-
-        uint64_t hi, lo, cl, ch;
-        cl = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0];
-        ch = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1];
-        lo = _umul128(idx0, cl, &hi);
-
-        al0 += hi;
-        ah0 += lo;
-
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
-
-        ah0 ^= ch;
-        al0 ^= cl;
-        idx0 = al0;
-    }
-
-    cn_implode_scratchpad((__m128i*) ctx->memory, (__m128i*) ctx->state);
-
-    keccakf(h0, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

algo/cryptonight/cryptonight_av3_aesni_bmi2.c

@@ -1,77 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "cryptonight.h"
-#include "cryptonight_p.h"
-#include "crypto/c_keccak.h"
-
-
-void cryptonight_av3_aesni_bmi2(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    keccak((const uint8_t *) input, size, ctx->state, 200);
-
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) ctx->memory);
-
-    const uint8_t* l0 = ctx->memory;
-    uint64_t* h0 = (uint64_t*) ctx->state;
-
-    uint64_t al0 = h0[0] ^ h0[4];
-    uint64_t ah0 = h0[1] ^ h0[5];
-    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
-
-    uint64_t idx0 = h0[0] ^ h0[4];
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i cx;
-        cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
-        cx = _mm_aesenc_si128(cx, _mm_set_epi64x(ah0, al0));
-
-        _mm_store_si128((__m128i *)&l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
-        idx0 = _mm_cvtsi128_si64(cx);
-        bx0 = cx;
-
-        uint64_t hi, lo, cl, ch;
-        cl = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0];
-        ch = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1];
-        lo = _mulx_u64(idx0, cl, &hi);
-
-        al0 += hi;
-        ah0 += lo;
-
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
-
-        ah0 ^= ch;
-        al0 ^= cl;
-        idx0 = al0;
-    }
-
-    cn_implode_scratchpad((__m128i*) ctx->memory, (__m128i*) ctx->state);
-
-    keccakf(h0, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

algo/cryptonight/cryptonight_av4_softaes.c

@@ -1,250 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "cryptonight.h"
-#include "crypto/c_keccak.h"
-
-__m128i soft_aesenc(__m128i in, __m128i key);
-__m128i soft_aeskeygenassist(__m128i key, uint8_t rcon);
-
-#ifdef __GNUC__
-static inline uint64_t _umul128(uint64_t a, uint64_t b, uint64_t* hi)
-{
-    unsigned __int128 r = (unsigned __int128)a * (unsigned __int128)b;
-    *hi = r >> 64;
-    return (uint64_t)r;
-}
-#endif
-
-
-// This will shift and xor tmp1 into itself as 4 32-bit vals such as
-// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
-static inline __m128i sl_xor(__m128i tmp1)
-{
-    __m128i tmp4;
-    tmp4 = _mm_slli_si128(tmp1, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    return tmp1;
-}
-
-
-static inline void aes_genkey_sub(__m128i* xout0, __m128i* xout2, uint8_t rcon)
-{
-    __m128i xout1 = soft_aeskeygenassist(*xout2, rcon);
-    xout1  = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
-    *xout0 = sl_xor(*xout0);
-    *xout0 = _mm_xor_si128(*xout0, xout1);
-    xout1  = soft_aeskeygenassist(*xout0, 0x00);
-    xout1  = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
-    *xout2 = sl_xor(*xout2);
-    *xout2 = _mm_xor_si128(*xout2, xout1);
-}
-
-
-static inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3, __m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
-{
-    __m128i xout0 = _mm_load_si128(memory);
-    __m128i xout2 = _mm_load_si128(memory + 1);
-    *k0 = xout0;
-    *k1 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x1);
-    *k2 = xout0;
-    *k3 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x2);
-    *k4 = xout0;
-    *k5 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x4);
-    *k6 = xout0;
-    *k7 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x8);
-    *k8 = xout0;
-    *k9 = xout2;
-}
-
-
-static inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
-{
-    *x0 = soft_aesenc(*x0, key);
-    *x1 = soft_aesenc(*x1, key);
-    *x2 = soft_aesenc(*x2, key);
-    *x3 = soft_aesenc(*x3, key);
-    *x4 = soft_aesenc(*x4, key);
-    *x5 = soft_aesenc(*x5, key);
-    *x6 = soft_aesenc(*x6, key);
-    *x7 = soft_aesenc(*x7, key);
-}
-
-
-static inline void cn_explode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xin0 = _mm_load_si128(input + 4);
-    xin1 = _mm_load_si128(input + 5);
-    xin2 = _mm_load_si128(input + 6);
-    xin3 = _mm_load_si128(input + 7);
-    xin4 = _mm_load_si128(input + 8);
-    xin5 = _mm_load_si128(input + 9);
-    xin6 = _mm_load_si128(input + 10);
-    xin7 = _mm_load_si128(input + 11);
-
-    for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8) {
-        aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-
-        _mm_store_si128(output + i + 0, xin0);
-        _mm_store_si128(output + i + 1, xin1);
-        _mm_store_si128(output + i + 2, xin2);
-        _mm_store_si128(output + i + 3, xin3);
-        _mm_prefetch((const char*)output + i + 0, _MM_HINT_T2);
-        _mm_store_si128(output + i + 4, xin4);
-        _mm_store_si128(output + i + 5, xin5);
-        _mm_store_si128(output + i + 6, xin6);
-        _mm_store_si128(output + i + 7, xin7);
-        _mm_prefetch((const char*)output + i + 4, _MM_HINT_T2);
-    }
-}
-
-
-static inline void cn_implode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xout0, xout1, xout2, xout3, xout4, xout5, xout6, xout7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(output + 2, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xout0 = _mm_load_si128(output + 4);
-    xout1 = _mm_load_si128(output + 5);
-    xout2 = _mm_load_si128(output + 6);
-    xout3 = _mm_load_si128(output + 7);
-    xout4 = _mm_load_si128(output + 8);
-    xout5 = _mm_load_si128(output + 9);
-    xout6 = _mm_load_si128(output + 10);
-    xout7 = _mm_load_si128(output + 11);
-
-    for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8)
-    {
-        _mm_prefetch((const char*)input + i + 0, _MM_HINT_NTA);
-        xout0 = _mm_xor_si128(_mm_load_si128(input + i + 0), xout0);
-        xout1 = _mm_xor_si128(_mm_load_si128(input + i + 1), xout1);
-        xout2 = _mm_xor_si128(_mm_load_si128(input + i + 2), xout2);
-        xout3 = _mm_xor_si128(_mm_load_si128(input + i + 3), xout3);
-        _mm_prefetch((const char*)input + i + 4, _MM_HINT_NTA);
-        xout4 = _mm_xor_si128(_mm_load_si128(input + i + 4), xout4);
-        xout5 = _mm_xor_si128(_mm_load_si128(input + i + 5), xout5);
-        xout6 = _mm_xor_si128(_mm_load_si128(input + i + 6), xout6);
-        xout7 = _mm_xor_si128(_mm_load_si128(input + i + 7), xout7);
-
-        aes_round(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-    }
-
-    _mm_store_si128(output + 4, xout0);
-    _mm_store_si128(output + 5, xout1);
-    _mm_store_si128(output + 6, xout2);
-    _mm_store_si128(output + 7, xout3);
-    _mm_store_si128(output + 8, xout4);
-    _mm_store_si128(output + 9, xout5);
-    _mm_store_si128(output + 10, xout6);
-    _mm_store_si128(output + 11, xout7);
-}
-
-
-void cryptonight_av4_softaes(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    keccak((const uint8_t *) input, size, ctx->state, 200);
-
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) ctx->memory);
-
-    const uint8_t* l0 = ctx->memory;
-    uint64_t* h0 = (uint64_t*) ctx->state;
-
-    uint64_t al0 = h0[0] ^ h0[4];
-    uint64_t ah0 = h0[1] ^ h0[5];
-    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
-
-    uint64_t idx0 = h0[0] ^ h0[4];
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i cx;
-        cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
-        cx = soft_aesenc(cx, _mm_set_epi64x(ah0, al0));
-
-        _mm_store_si128((__m128i *)&l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
-        idx0 = _mm_cvtsi128_si64(cx);
-        bx0 = cx;
-
-        uint64_t hi, lo, cl, ch;
-        cl = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0];
-        ch = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1];
-        lo = _umul128(idx0, cl, &hi);
-
-        al0 += hi;
-        ah0 += lo;
-
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
-
-        ah0 ^= ch;
-        al0 ^= cl;
-        idx0 = al0;
-    }
-
-    cn_implode_scratchpad((__m128i*) ctx->memory, (__m128i*) ctx->state);
-
-    keccakf(h0, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

algo/cryptonight/cryptonight_av5_aesni_experimental.c

@@ -1,77 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "cryptonight.h"
-#include "cryptonight_p.h"
-#include "crypto/c_keccak.h"
-
-
-void cryptonight_av5_aesni_experimental(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    const uint8_t* memory = ctx->memory;
-
-    keccak((const uint8_t *) input, size, ctx->state, 200);
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) memory);
-
-    uint64_t* state = (uint64_t*) ctx->state;
-
-    uint64_t a[2] __attribute((aligned(16))) = { state[0] ^ state[4], state[1] ^ state[5] };
-    uint64_t c    __attribute((aligned(16)));
-    uint64_t d[2] __attribute((aligned(16)));
-
-    __m128i a_x = _mm_load_si128((__m128i *) &memory[a[0] & 0x1FFFF0]);
-    __m128i b_x = _mm_set_epi64x(state[3] ^ state[7], state[2] ^ state[6]);
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i c_x = _mm_aesenc_si128(a_x, _mm_load_si128((__m128i *) a));
-        c = _mm_cvtsi128_si64(c_x);
-
-        uint64_t *restrict d_ptr = (uint64_t *) &memory[c & 0x1FFFF0];
-        _mm_store_si128((__m128i *) &memory[a[0] & 0x1FFFF0], _mm_xor_si128(b_x, c_x));
-        b_x = c_x;
-
-        d[0] = d_ptr[0];
-        d[1] = d_ptr[1];
-
-        {
-            unsigned __int128 res = (unsigned __int128) c * d[0];
-
-            d_ptr[0] = a[0] += res >> 64;
-            d_ptr[1] = a[1] += (uint64_t) res;
-        }
-
-        a[0] ^= d[0];
-        a[1] ^= d[1];
-
-        a_x = _mm_load_si128((__m128i *) &memory[a[0] & 0x1FFFF0]);
-    }
-
-    cn_implode_scratchpad((__m128i*) memory, (__m128i*) state);
-
-    keccakf(state, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

algo/cryptonight/cryptonight_p.h

@@ -1,223 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 __CRYPTONIGHT_P_H__
-#define __CRYPTONIGHT_P_H__
-
-#include <x86intrin.h>
-
-
-#define aes_genkey_sub(imm8) \
-    __m128i xout1 = _mm_aeskeygenassist_si128(*xout2, (imm8)); \
-    xout1  = _mm_shuffle_epi32(xout1, 0xFF); \
-    *xout0 = sl_xor(*xout0); \
-    *xout0 = _mm_xor_si128(*xout0, xout1); \
-    xout1  = _mm_aeskeygenassist_si128(*xout0, 0x00);\
-    xout1  = _mm_shuffle_epi32(xout1, 0xAA); \
-    *xout2 = sl_xor(*xout2); \
-    *xout2 = _mm_xor_si128(*xout2, xout1); \
-
-
-// This will shift and xor tmp1 into itself as 4 32-bit vals such as
-// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
-inline __m128i sl_xor(__m128i tmp1)
-{
-    __m128i tmp4;
-    tmp4 = _mm_slli_si128(tmp1, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    return tmp1;
-}
-
-
-inline void aes_genkey_sub1(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x1)
-}
-
-
-inline void aes_genkey_sub2(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x2)
-}
-
-
-inline void aes_genkey_sub4(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x4)
-}
-
-
-inline void aes_genkey_sub8(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x8)
-}
-
-
-inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
-{
-    *x0 = _mm_aesenc_si128(*x0, key);
-    *x1 = _mm_aesenc_si128(*x1, key);
-    *x2 = _mm_aesenc_si128(*x2, key);
-    *x3 = _mm_aesenc_si128(*x3, key);
-    *x4 = _mm_aesenc_si128(*x4, key);
-    *x5 = _mm_aesenc_si128(*x5, key);
-    *x6 = _mm_aesenc_si128(*x6, key);
-    *x7 = _mm_aesenc_si128(*x7, key);
-}
-
-
-inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3, __m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
-{
-    __m128i xout0 = _mm_load_si128(memory);
-    __m128i xout2 = _mm_load_si128(memory + 1);
-    *k0 = xout0;
-    *k1 = xout2;
-
-     aes_genkey_sub1(&xout0, &xout2);
-    *k2 = xout0;
-    *k3 = xout2;
-
-     aes_genkey_sub2(&xout0, &xout2);
-    *k4 = xout0;
-    *k5 = xout2;
-
-     aes_genkey_sub4(&xout0, &xout2);
-    *k6 = xout0;
-    *k7 = xout2;
-
-     aes_genkey_sub8(&xout0, &xout2);
-    *k8 = xout0;
-    *k9 = xout2;
-}
-
-
-inline void cn_explode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xin0 = _mm_load_si128(input + 4);
-    xin1 = _mm_load_si128(input + 5);
-    xin2 = _mm_load_si128(input + 6);
-    xin3 = _mm_load_si128(input + 7);
-    xin4 = _mm_load_si128(input + 8);
-    xin5 = _mm_load_si128(input + 9);
-    xin6 = _mm_load_si128(input + 10);
-    xin7 = _mm_load_si128(input + 11);
-
-    for (size_t i = 0; __builtin_expect(i < MEMORY / sizeof(__m128i), 1); i += 8) {
-        aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-
-        _mm_store_si128(output + i + 0, xin0);
-        _mm_store_si128(output + i + 1, xin1);
-        _mm_store_si128(output + i + 2, xin2);
-        _mm_store_si128(output + i + 3, xin3);
-        _mm_prefetch((const char*)output + i + 0, _MM_HINT_T2);
-        _mm_store_si128(output + i + 4, xin4);
-        _mm_store_si128(output + i + 5, xin5);
-        _mm_store_si128(output + i + 6, xin6);
-        _mm_store_si128(output + i + 7, xin7);
-        _mm_prefetch((const char*)output + i + 4, _MM_HINT_T2);
-    }
-}
-
-
-inline void cn_implode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xout0, xout1, xout2, xout3, xout4, xout5, xout6, xout7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(output + 2, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xout0 = _mm_load_si128(output + 4);
-    xout1 = _mm_load_si128(output + 5);
-    xout2 = _mm_load_si128(output + 6);
-    xout3 = _mm_load_si128(output + 7);
-    xout4 = _mm_load_si128(output + 8);
-    xout5 = _mm_load_si128(output + 9);
-    xout6 = _mm_load_si128(output + 10);
-    xout7 = _mm_load_si128(output + 11);
-
-    for (size_t i = 0; __builtin_expect(i < MEMORY / sizeof(__m128i), 1); i += 8)
-    {
-        _mm_prefetch((const char*)input + i + 0, _MM_HINT_NTA);
-        xout0 = _mm_xor_si128(_mm_load_si128(input + i + 0), xout0);
-        xout1 = _mm_xor_si128(_mm_load_si128(input + i + 1), xout1);
-        xout2 = _mm_xor_si128(_mm_load_si128(input + i + 2), xout2);
-        xout3 = _mm_xor_si128(_mm_load_si128(input + i + 3), xout3);
-        _mm_prefetch((const char*)input + i + 4, _MM_HINT_NTA);
-        xout4 = _mm_xor_si128(_mm_load_si128(input + i + 4), xout4);
-        xout5 = _mm_xor_si128(_mm_load_si128(input + i + 5), xout5);
-        xout6 = _mm_xor_si128(_mm_load_si128(input + i + 6), xout6);
-        xout7 = _mm_xor_si128(_mm_load_si128(input + i + 7), xout7);
-
-        aes_round(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-    }
-
-    _mm_store_si128(output + 4, xout0);
-    _mm_store_si128(output + 5, xout1);
-    _mm_store_si128(output + 6, xout2);
-    _mm_store_si128(output + 7, xout3);
-    _mm_store_si128(output + 8, xout4);
-    _mm_store_si128(output + 9, xout5);
-    _mm_store_si128(output + 10, xout6);
-    _mm_store_si128(output + 11, xout7);
-}
-
-
-inline uint64_t _umul128(uint64_t a, uint64_t b, uint64_t* hi)
-{
-    unsigned __int128 r = (unsigned __int128) a * (unsigned __int128) b;
-    *hi = r >> 64;
-    return (uint64_t) r;
-}
-
-
-#endif /* __CRYPTONIGHT_P_H__ */

algo/cryptonight/i686/cryptonight_av1_aesni.c

@@ -1,297 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "algo/cryptonight/cryptonight.h"
-#include "crypto/c_keccak.h"
-
-
-#ifdef __GNUC__
-static inline uint64_t _umul128(uint64_t multiplier, uint64_t multiplicand, uint64_t *product_hi) {
-    // multiplier   = ab = a * 2^32 + b
-    // multiplicand = cd = c * 2^32 + d
-    // ab * cd = a * c * 2^64 + (a * d + b * c) * 2^32 + b * d
-    uint64_t a = multiplier >> 32;
-    uint64_t b = multiplier & 0xFFFFFFFF;
-    uint64_t c = multiplicand >> 32;
-    uint64_t d = multiplicand & 0xFFFFFFFF;
-
-    //uint64_t ac = a * c;
-    uint64_t ad = a * d;
-    //uint64_t bc = b * c;
-    uint64_t bd = b * d;
-
-    uint64_t adbc = ad + (b * c);
-    uint64_t adbc_carry = adbc < ad ? 1 : 0;
-
-    // multiplier * multiplicand = product_hi * 2^64 + product_lo
-    uint64_t product_lo = bd + (adbc << 32);
-    uint64_t product_lo_carry = product_lo < bd ? 1 : 0;
-    *product_hi = (a * c) + (adbc >> 32) + (adbc_carry << 32) + product_lo_carry;
-
-    return product_lo;
-}
-#endif
-
-
-#define HI32(X) \
-    _mm_srli_si128((X), 4)
-
-
-#define EXTRACT64(X) \
-    ((uint64_t)(uint32_t)_mm_cvtsi128_si32(X) | \
-    ((uint64_t)(uint32_t)_mm_cvtsi128_si32(HI32(X)) << 32))
-
-
-#define aes_genkey_sub(imm8) \
-    __m128i xout1 = _mm_aeskeygenassist_si128(*xout2, (imm8)); \
-    xout1 = _mm_shuffle_epi32(xout1, 0xFF); \
-    *xout0 = sl_xor(*xout0); \
-    *xout0 = _mm_xor_si128(*xout0, xout1); \
-    xout1 = _mm_aeskeygenassist_si128(*xout0, 0x00);\
-    xout1 = _mm_shuffle_epi32(xout1, 0xAA); \
-    *xout2 = sl_xor(*xout2); \
-    *xout2 = _mm_xor_si128(*xout2, xout1); \
-
-
-// This will shift and xor tmp1 into itself as 4 32-bit vals such as
-// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
-static inline __m128i sl_xor(__m128i tmp1)
-{
-    __m128i tmp4;
-    tmp4 = _mm_slli_si128(tmp1, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    return tmp1;
-}
-
-
-static inline void aes_genkey_sub1(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x1)
-}
-
-
-static inline void aes_genkey_sub2(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x2)
-}
-
-
-static inline void aes_genkey_sub4(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x4)
-}
-
-
-static inline void aes_genkey_sub8(__m128i* xout0, __m128i* xout2)
-{
-   aes_genkey_sub(0x8)
-}
-
-
-static inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3, __m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
-{
-    __m128i xout0 = _mm_load_si128(memory);
-    __m128i xout2 = _mm_load_si128(memory + 1);
-    *k0 = xout0;
-    *k1 = xout2;
-
-     aes_genkey_sub1(&xout0, &xout2);
-    *k2 = xout0;
-    *k3 = xout2;
-
-     aes_genkey_sub2(&xout0, &xout2);
-    *k4 = xout0;
-    *k5 = xout2;
-
-     aes_genkey_sub4(&xout0, &xout2);
-    *k6 = xout0;
-    *k7 = xout2;
-
-     aes_genkey_sub8(&xout0, &xout2);
-    *k8 = xout0;
-    *k9 = xout2;
-}
-
-
-static inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
-{
-    *x0 = _mm_aesenc_si128(*x0, key);
-    *x1 = _mm_aesenc_si128(*x1, key);
-    *x2 = _mm_aesenc_si128(*x2, key);
-    *x3 = _mm_aesenc_si128(*x3, key);
-    *x4 = _mm_aesenc_si128(*x4, key);
-    *x5 = _mm_aesenc_si128(*x5, key);
-    *x6 = _mm_aesenc_si128(*x6, key);
-    *x7 = _mm_aesenc_si128(*x7, key);
-}
-
-
-static inline void cn_explode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xin0 = _mm_load_si128(input + 4);
-    xin1 = _mm_load_si128(input + 5);
-    xin2 = _mm_load_si128(input + 6);
-    xin3 = _mm_load_si128(input + 7);
-    xin4 = _mm_load_si128(input + 8);
-    xin5 = _mm_load_si128(input + 9);
-    xin6 = _mm_load_si128(input + 10);
-    xin7 = _mm_load_si128(input + 11);
-
-    for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8) {
-        aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-
-        _mm_store_si128(output + i + 0, xin0);
-        _mm_store_si128(output + i + 1, xin1);
-        _mm_store_si128(output + i + 2, xin2);
-        _mm_store_si128(output + i + 3, xin3);
-        _mm_prefetch((const char*)output + i + 0, _MM_HINT_T2);
-        _mm_store_si128(output + i + 4, xin4);
-        _mm_store_si128(output + i + 5, xin5);
-        _mm_store_si128(output + i + 6, xin6);
-        _mm_store_si128(output + i + 7, xin7);
-        _mm_prefetch((const char*)output + i + 4, _MM_HINT_T2);
-    }
-}
-
-
-static inline void cn_implode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xout0, xout1, xout2, xout3, xout4, xout5, xout6, xout7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(output + 2, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xout0 = _mm_load_si128(output + 4);
-    xout1 = _mm_load_si128(output + 5);
-    xout2 = _mm_load_si128(output + 6);
-    xout3 = _mm_load_si128(output + 7);
-    xout4 = _mm_load_si128(output + 8);
-    xout5 = _mm_load_si128(output + 9);
-    xout6 = _mm_load_si128(output + 10);
-    xout7 = _mm_load_si128(output + 11);
-
-    for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8)
-    {
-        _mm_prefetch((const char*)input + i + 0, _MM_HINT_NTA);
-        xout0 = _mm_xor_si128(_mm_load_si128(input + i + 0), xout0);
-        xout1 = _mm_xor_si128(_mm_load_si128(input + i + 1), xout1);
-        xout2 = _mm_xor_si128(_mm_load_si128(input + i + 2), xout2);
-        xout3 = _mm_xor_si128(_mm_load_si128(input + i + 3), xout3);
-        _mm_prefetch((const char*)input + i + 4, _MM_HINT_NTA);
-        xout4 = _mm_xor_si128(_mm_load_si128(input + i + 4), xout4);
-        xout5 = _mm_xor_si128(_mm_load_si128(input + i + 5), xout5);
-        xout6 = _mm_xor_si128(_mm_load_si128(input + i + 6), xout6);
-        xout7 = _mm_xor_si128(_mm_load_si128(input + i + 7), xout7);
-
-        aes_round(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-    }
-
-    _mm_store_si128(output + 4, xout0);
-    _mm_store_si128(output + 5, xout1);
-    _mm_store_si128(output + 6, xout2);
-    _mm_store_si128(output + 7, xout3);
-    _mm_store_si128(output + 8, xout4);
-    _mm_store_si128(output + 9, xout5);
-    _mm_store_si128(output + 10, xout6);
-    _mm_store_si128(output + 11, xout7);
-}
-
-
-void cryptonight_av1_aesni(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    keccak((const uint8_t *) input, size, ctx->state, 200);
-
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) ctx->memory);
-
-    const uint8_t* l0 = ctx->memory;
-    uint64_t* h0 = (uint64_t*) ctx->state;
-
-    uint64_t al0 = h0[0] ^ h0[4];
-    uint64_t ah0 = h0[1] ^ h0[5];
-    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
-
-    uint64_t idx0 = h0[0] ^ h0[4];
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i cx;
-        cx = _mm_load_si128((__m128i *) &l0[idx0 & 0x1FFFF0]);
-        cx = _mm_aesenc_si128(cx, _mm_set_epi64x(ah0, al0));
-
-        _mm_store_si128((__m128i *) &l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
-        idx0 = EXTRACT64(cx);
-        bx0 = cx;
-
-        uint64_t hi, lo, cl, ch;
-        cl = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[0];
-        ch = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[1];
-        lo = _umul128(idx0, cl, &hi);
-
-        al0 += hi;
-        ah0 += lo;
-
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
-
-        ah0 ^= ch;
-        al0 ^= cl;
-        idx0 = al0;
-    }
-
-    cn_implode_scratchpad((__m128i*) ctx->memory, (__m128i*) ctx->state);
-
-    keccakf(h0, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

algo/cryptonight/i686/cryptonight_av4_softaes.c

@@ -1,277 +0,0 @@
-/* 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      fireice-uk  <https://github.com/fireice-uk>
- * Copyright 2016-2017 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 <x86intrin.h>
-#include <string.h>
-
-#include "algo/cryptonight/cryptonight.h"
-#include "crypto/c_keccak.h"
-
-__m128i soft_aesenc(__m128i in, __m128i key);
-__m128i soft_aeskeygenassist(__m128i key, uint8_t rcon);
-
-#ifdef __GNUC__
-static inline uint64_t _umul128(uint64_t multiplier, uint64_t multiplicand, uint64_t *product_hi) {
-    // multiplier   = ab = a * 2^32 + b
-    // multiplicand = cd = c * 2^32 + d
-    // ab * cd = a * c * 2^64 + (a * d + b * c) * 2^32 + b * d
-    uint64_t a = multiplier >> 32;
-    uint64_t b = multiplier & 0xFFFFFFFF;
-    uint64_t c = multiplicand >> 32;
-    uint64_t d = multiplicand & 0xFFFFFFFF;
-
-    //uint64_t ac = a * c;
-    uint64_t ad = a * d;
-    //uint64_t bc = b * c;
-    uint64_t bd = b * d;
-
-    uint64_t adbc = ad + (b * c);
-    uint64_t adbc_carry = adbc < ad ? 1 : 0;
-
-    // multiplier * multiplicand = product_hi * 2^64 + product_lo
-    uint64_t product_lo = bd + (adbc << 32);
-    uint64_t product_lo_carry = product_lo < bd ? 1 : 0;
-    *product_hi = (a * c) + (adbc >> 32) + (adbc_carry << 32) + product_lo_carry;
-
-    return product_lo;
-}
-#endif
-
-
-#define HI32(X) \
-    _mm_srli_si128((X), 4)
-
-
-#define EXTRACT64(X) \
-    ((uint64_t)(uint32_t)_mm_cvtsi128_si32(X) | \
-    ((uint64_t)(uint32_t)_mm_cvtsi128_si32(HI32(X)) << 32))
-
-
-// This will shift and xor tmp1 into itself as 4 32-bit vals such as
-// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
-static inline __m128i sl_xor(__m128i tmp1)
-{
-    __m128i tmp4;
-    tmp4 = _mm_slli_si128(tmp1, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    tmp4 = _mm_slli_si128(tmp4, 0x04);
-    tmp1 = _mm_xor_si128(tmp1, tmp4);
-    return tmp1;
-}
-
-
-static inline void aes_genkey_sub(__m128i* xout0, __m128i* xout2, uint8_t rcon)
-{
-    __m128i xout1 = soft_aeskeygenassist(*xout2, rcon);
-    xout1  = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
-    *xout0 = sl_xor(*xout0);
-    *xout0 = _mm_xor_si128(*xout0, xout1);
-    xout1  = soft_aeskeygenassist(*xout0, 0x00);
-    xout1  = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
-    *xout2 = sl_xor(*xout2);
-    *xout2 = _mm_xor_si128(*xout2, xout1);
-}
-
-
-static inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3, __m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
-{
-    __m128i xout0 = _mm_load_si128(memory);
-    __m128i xout2 = _mm_load_si128(memory + 1);
-    *k0 = xout0;
-    *k1 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x1);
-    *k2 = xout0;
-    *k3 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x2);
-    *k4 = xout0;
-    *k5 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x4);
-    *k6 = xout0;
-    *k7 = xout2;
-
-    aes_genkey_sub(&xout0, &xout2, 0x8);
-    *k8 = xout0;
-    *k9 = xout2;
-}
-
-
-static inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
-{
-    *x0 = soft_aesenc(*x0, key);
-    *x1 = soft_aesenc(*x1, key);
-    *x2 = soft_aesenc(*x2, key);
-    *x3 = soft_aesenc(*x3, key);
-    *x4 = soft_aesenc(*x4, key);
-    *x5 = soft_aesenc(*x5, key);
-    *x6 = soft_aesenc(*x6, key);
-    *x7 = soft_aesenc(*x7, key);
-}
-
-
-static inline void cn_explode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xin0 = _mm_load_si128(input + 4);
-    xin1 = _mm_load_si128(input + 5);
-    xin2 = _mm_load_si128(input + 6);
-    xin3 = _mm_load_si128(input + 7);
-    xin4 = _mm_load_si128(input + 8);
-    xin5 = _mm_load_si128(input + 9);
-    xin6 = _mm_load_si128(input + 10);
-    xin7 = _mm_load_si128(input + 11);
-
-    for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8) {
-        aes_round(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-        aes_round(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
-
-        _mm_store_si128(output + i + 0, xin0);
-        _mm_store_si128(output + i + 1, xin1);
-        _mm_store_si128(output + i + 2, xin2);
-        _mm_store_si128(output + i + 3, xin3);
-        _mm_prefetch((const char*)output + i + 0, _MM_HINT_T2);
-        _mm_store_si128(output + i + 4, xin4);
-        _mm_store_si128(output + i + 5, xin5);
-        _mm_store_si128(output + i + 6, xin6);
-        _mm_store_si128(output + i + 7, xin7);
-        _mm_prefetch((const char*)output + i + 4, _MM_HINT_T2);
-    }
-}
-
-
-static inline void cn_implode_scratchpad(const __m128i* input, __m128i* output)
-{
-    // This is more than we have registers, compiler will assign 2 keys on the stack
-    __m128i xout0, xout1, xout2, xout3, xout4, xout5, xout6, xout7;
-    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
-
-    aes_genkey(output + 2, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
-
-    xout0 = _mm_load_si128(output + 4);
-    xout1 = _mm_load_si128(output + 5);
-    xout2 = _mm_load_si128(output + 6);
-    xout3 = _mm_load_si128(output + 7);
-    xout4 = _mm_load_si128(output + 8);
-    xout5 = _mm_load_si128(output + 9);
-    xout6 = _mm_load_si128(output + 10);
-    xout7 = _mm_load_si128(output + 11);
-
-    for (size_t i = 0; i < MEMORY / sizeof(__m128i); i += 8)
-    {
-        _mm_prefetch((const char*)input + i + 0, _MM_HINT_NTA);
-        xout0 = _mm_xor_si128(_mm_load_si128(input + i + 0), xout0);
-        xout1 = _mm_xor_si128(_mm_load_si128(input + i + 1), xout1);
-        xout2 = _mm_xor_si128(_mm_load_si128(input + i + 2), xout2);
-        xout3 = _mm_xor_si128(_mm_load_si128(input + i + 3), xout3);
-        _mm_prefetch((const char*)input + i + 4, _MM_HINT_NTA);
-        xout4 = _mm_xor_si128(_mm_load_si128(input + i + 4), xout4);
-        xout5 = _mm_xor_si128(_mm_load_si128(input + i + 5), xout5);
-        xout6 = _mm_xor_si128(_mm_load_si128(input + i + 6), xout6);
-        xout7 = _mm_xor_si128(_mm_load_si128(input + i + 7), xout7);
-
-        aes_round(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-        aes_round(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
-    }
-
-    _mm_store_si128(output + 4, xout0);
-    _mm_store_si128(output + 5, xout1);
-    _mm_store_si128(output + 6, xout2);
-    _mm_store_si128(output + 7, xout3);
-    _mm_store_si128(output + 8, xout4);
-    _mm_store_si128(output + 9, xout5);
-    _mm_store_si128(output + 10, xout6);
-    _mm_store_si128(output + 11, xout7);
-}
-
-
-void cryptonight_av4_softaes(const void *restrict input, size_t size, void *restrict output, struct cryptonight_ctx *restrict ctx)
-{
-    keccak((const uint8_t *) input, 76, ctx->state, 200);
-
-    cn_explode_scratchpad((__m128i*) ctx->state, (__m128i*) ctx->memory);
-
-    const uint8_t* l0 = ctx->memory;
-    uint64_t* h0 = (uint64_t*) ctx->state;
-
-    uint64_t al0 = h0[0] ^ h0[4];
-    uint64_t ah0 = h0[1] ^ h0[5];
-    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
-
-    uint64_t idx0 = h0[0] ^ h0[4];
-
-    for (size_t i = 0; __builtin_expect(i < 0x80000, 1); i++) {
-        __m128i cx;
-        cx = _mm_load_si128((__m128i *)&l0[idx0 & 0x1FFFF0]);
-        cx = soft_aesenc(cx, _mm_set_epi64x(ah0, al0));
-
-        _mm_store_si128((__m128i *)&l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
-        idx0 = EXTRACT64(cx);
-        bx0 = cx;
-
-        uint64_t hi, lo, cl, ch;
-        cl = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0];
-        ch = ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1];
-        lo = _umul128(idx0, cl, &hi);
-
-        al0 += hi;
-        ah0 += lo;
-
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[0] = al0;
-        ((uint64_t*)&l0[idx0 & 0x1FFFF0])[1] = ah0;
-
-        ah0 ^= ch;
-        al0 ^= cl;
-        idx0 = al0;
-    }
-
-    cn_implode_scratchpad((__m128i*) ctx->memory, (__m128i*) ctx->state);
-
-    keccakf(h0, 24);
-    extra_hashes[ctx->state[0] & 3](ctx->state, 200, output);
-}

cmake/FindUV.cmake

@@ -0,0 +1,8 @@
+find_path(UV_INCLUDE_DIR NAMES uv.h)
+find_library(UV_LIBRARY NAMES uv libuv)
+
+set(UV_LIBRARIES ${UV_LIBRARY})
+set(UV_INCLUDE_DIRS ${UV_INCLUDE_DIR})
+
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(UV DEFAULT_MSG UV_LIBRARY UV_INCLUDE_DIR)

crypto/c_keccak.c

@@ -1,123 +0,0 @@
-// keccak.c
-// 19-Nov-11  Markku-Juhani O. Saarinen <mjos@iki.fi>
-// A baseline Keccak (3rd round) implementation.
-
-#include <stdint.h>
-#include <string.h>
-
-#define HASH_DATA_AREA 136
-#define KECCAK_ROUNDS 24
-
-#ifndef ROTL64
-#define ROTL64(x, y) (((x) << (y)) | ((x) >> (64 - (y))))
-#endif
-
-const uint64_t keccakf_rndc[24] = 
-{
-    0x0000000000000001, 0x0000000000008082, 0x800000000000808a,
-    0x8000000080008000, 0x000000000000808b, 0x0000000080000001,
-    0x8000000080008081, 0x8000000000008009, 0x000000000000008a,
-    0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
-    0x000000008000808b, 0x800000000000008b, 0x8000000000008089,
-    0x8000000000008003, 0x8000000000008002, 0x8000000000000080, 
-    0x000000000000800a, 0x800000008000000a, 0x8000000080008081,
-    0x8000000000008080, 0x0000000080000001, 0x8000000080008008
-};
-
-const int keccakf_rotc[24] = 
-{
-    1,  3,  6,  10, 15, 21, 28, 36, 45, 55, 2,  14, 
-    27, 41, 56, 8,  25, 43, 62, 18, 39, 61, 20, 44
-};
-
-const int keccakf_piln[24] = 
-{
-    10, 7,  11, 17, 18, 3, 5,  16, 8,  21, 24, 4, 
-    15, 23, 19, 13, 12, 2, 20, 14, 22, 9,  6,  1 
-};
-
-// update the state with given number of rounds
-
-void keccakf(uint64_t st[25], int rounds)
-{
-    int i, j, round;
-    uint64_t t, bc[5];
-
-    for (round = 0; round < rounds; ++round) {
-
-        // Theta
-        bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20];
-        bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21];
-        bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22];
-        bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23];
-        bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24];
-
-        for (i = 0; i < 5; ++i) {
-            t = bc[(i + 4) % 5] ^ ROTL64(bc[(i + 1) % 5], 1);
-            st[i     ] ^= t;
-            st[i +  5] ^= t;
-            st[i + 10] ^= t;
-            st[i + 15] ^= t;
-            st[i + 20] ^= t;
-        }
-
-        // Rho Pi
-        t = st[1];
-        for (i = 0; i < 24; ++i) {
-            bc[0] = st[keccakf_piln[i]];
-            st[keccakf_piln[i]] = ROTL64(t, keccakf_rotc[i]);
-            t = bc[0];
-        }
-
-        //  Chi
-        for (j = 0; j < 25; j += 5) {
-            bc[0] = st[j    ];
-            bc[1] = st[j + 1];
-            bc[2] = st[j + 2];
-            bc[3] = st[j + 3];
-            bc[4] = st[j + 4];
-            st[j    ] ^= (~bc[1]) & bc[2];
-            st[j + 1] ^= (~bc[2]) & bc[3];
-            st[j + 2] ^= (~bc[3]) & bc[4];
-            st[j + 3] ^= (~bc[4]) & bc[0];
-            st[j + 4] ^= (~bc[0]) & bc[1];
-        }
-
-        //  Iota
-        st[0] ^= keccakf_rndc[round];
-    }
-}
-
-// compute a keccak hash (md) of given byte length from "in"
-typedef uint64_t state_t[25];
-
-void keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen)
-{
-    state_t st;
-    uint8_t temp[144];
-    int i, rsiz, rsizw;
-
-    rsiz = sizeof(state_t) == mdlen ? HASH_DATA_AREA : 200 - 2 * mdlen;
-    rsizw = rsiz / 8;
-    
-    memset(st, 0, sizeof(st));
-
-    for ( ; inlen >= rsiz; inlen -= rsiz, in += rsiz) {
-        for (i = 0; i < rsizw; i++)
-            st[i] ^= ((uint64_t *) in)[i];
-        keccakf(st, KECCAK_ROUNDS);
-    }
-    
-    // last block and padding
-    memcpy(temp, in, inlen);
-    temp[inlen++] = 1;
-    memset(temp + inlen, 0, rsiz - inlen);
-    temp[rsiz - 1] |= 0x80;
-
-    for (i = 0; i < rsizw; i++)
-        st[i] ^= ((uint64_t *) temp)[i];
-
-    keccakf(st, KECCAK_ROUNDS);
-
-    memcpy(md, st, mdlen);
-}

crypto/hash.c

@@ -1,24 +0,0 @@
-// Copyright (c) 2012-2013 The Cryptonote developers
-// Distributed under the MIT/X11 software license, see the accompanying
-// file COPYING or http://www.opensource.org/licenses/mit-license.php.
-
-#include <stddef.h>
-#include <stdint.h>
-#include <string.h>
-
-#include "hash-ops.h"
-#include "c_keccak.h"
-
-void hash_permutation(union hash_state *state) {
-  keccakf((uint64_t*)state, 24);
-}
-
-void hash_process(union hash_state *state, const uint8_t *buf, size_t count) {
-  keccak1600(buf, count, (uint8_t*)state);
-}
-
-void cn_fast_hash(const void *data, size_t length, char *hash) {
-  union hash_state state;
-  hash_process(&state, data, length);
-  memcpy(hash, &state, HASH_SIZE);
-}

elist.h

@@ -1,274 +0,0 @@
-/* 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 2016-2017 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 _LINUX_LIST_H
-#define _LINUX_LIST_H
-
-/*
- * Simple doubly linked list implementation.
- *
- * Some of the internal functions ("__xxx") are useful when
- * manipulating whole lists rather than single entries, as
- * sometimes we already know the next/prev entries and we can
- * generate better code by using them directly rather than
- * using the generic single-entry routines.
- */
-
-struct list_head {
-	struct list_head *next, *prev;
-};
-
-#define LIST_HEAD_INIT(name) { &(name), &(name) }
-
-#define LIST_HEAD(name) \
-	struct list_head name = LIST_HEAD_INIT(name)
-
-#define INIT_LIST_HEAD(ptr) do { \
-	(ptr)->next = (ptr); (ptr)->prev = (ptr); \
-} while (0)
-
-/*
- * Insert a new entry between two known consecutive entries.
- *
- * This is only for internal list manipulation where we know
- * the prev/next entries already!
- */
-static inline void __list_add(struct list_head *new,
-			      struct list_head *prev,
-			      struct list_head *next)
-{
-	next->prev = new;
-	new->next = next;
-	new->prev = prev;
-	prev->next = new;
-}
-
-/**
- * list_add - add a new entry
- * @new: new entry to be added
- * @head: list head to add it after
- *
- * Insert a new entry after the specified head.
- * This is good for implementing stacks.
- */
-static inline void list_add(struct list_head *new, struct list_head *head)
-{
-	__list_add(new, head, head->next);
-}
-
-/**
- * list_add_tail - add a new entry
- * @new: new entry to be added
- * @head: list head to add it before
- *
- * Insert a new entry before the specified head.
- * This is useful for implementing queues.
- */
-static inline void list_add_tail(struct list_head *new, struct list_head *head)
-{
-	__list_add(new, head->prev, head);
-}
-
-/*
- * Delete a list entry by making the prev/next entries
- * point to each other.
- *
- * This is only for internal list manipulation where we know
- * the prev/next entries already!
- */
-static inline void __list_del(struct list_head *prev, struct list_head *next)
-{
-	next->prev = prev;
-	prev->next = next;
-}
-
-/**
- * list_del - deletes entry from list.
- * @entry: the element to delete from the list.
- * Note: list_empty on entry does not return true after this, the entry is in an undefined state.
- */
-static inline void list_del(struct list_head *entry)
-{
-	__list_del(entry->prev, entry->next);
-	entry->next = (void *) 0;
-	entry->prev = (void *) 0;
-}
-
-/**
- * list_del_init - deletes entry from list and reinitialize it.
- * @entry: the element to delete from the list.
- */
-static inline void list_del_init(struct list_head *entry)
-{
-	__list_del(entry->prev, entry->next);
-	INIT_LIST_HEAD(entry);
-}
-
-/**
- * list_move - delete from one list and add as another's head
- * @list: the entry to move
- * @head: the head that will precede our entry
- */
-static inline void list_move(struct list_head *list, struct list_head *head)
-{
-        __list_del(list->prev, list->next);
-        list_add(list, head);
-}
-
-/**
- * list_move_tail - delete from one list and add as another's tail
- * @list: the entry to move
- * @head: the head that will follow our entry
- */
-static inline void list_move_tail(struct list_head *list,
-				  struct list_head *head)
-{
-        __list_del(list->prev, list->next);
-        list_add_tail(list, head);
-}
-
-/**
- * list_empty - tests whether a list is empty
- * @head: the list to test.
- */
-static inline int list_empty(struct list_head *head)
-{
-	return head->next == head;
-}
-
-static inline void __list_splice(struct list_head *list,
-				 struct list_head *head)
-{
-	struct list_head *first = list->next;
-	struct list_head *last = list->prev;
-	struct list_head *at = head->next;
-
-	first->prev = head;
-	head->next = first;
-
-	last->next = at;
-	at->prev = last;
-}
-
-/**
- * list_splice - join two lists
- * @list: the new list to add.
- * @head: the place to add it in the first list.
- */
-static inline void list_splice(struct list_head *list, struct list_head *head)
-{
-	if (!list_empty(list))
-		__list_splice(list, head);
-}
-
-/**
- * list_splice_init - join two lists and reinitialise the emptied list.
- * @list: the new list to add.
- * @head: the place to add it in the first list.
- *
- * The list at @list is reinitialised
- */
-static inline void list_splice_init(struct list_head *list,
-				    struct list_head *head)
-{
-	if (!list_empty(list)) {
-		__list_splice(list, head);
-		INIT_LIST_HEAD(list);
-	}
-}
-
-/**
- * list_entry - get the struct for this entry
- * @ptr:	the &struct list_head pointer.
- * @type:	the type of the struct this is embedded in.
- * @member:	the name of the list_struct within the struct.
- */
-#define list_entry(ptr, type, member) \
-	((type *)((char *)(ptr)-(unsigned long)(&((type *)0)->member)))
-
-/**
- * list_for_each	-	iterate over a list
- * @pos:	the &struct list_head to use as a loop counter.
- * @head:	the head for your list.
- */
-#define list_for_each(pos, head) \
-	for (pos = (head)->next; pos != (head); \
-        	pos = pos->next)
-/**
- * list_for_each_prev	-	iterate over a list backwards
- * @pos:	the &struct list_head to use as a loop counter.
- * @head:	the head for your list.
- */
-#define list_for_each_prev(pos, head) \
-	for (pos = (head)->prev; pos != (head); \
-        	pos = pos->prev)
-
-/**
- * list_for_each_safe	-	iterate over a list safe against removal of list entry
- * @pos:	the &struct list_head to use as a loop counter.
- * @n:		another &struct list_head to use as temporary storage
- * @head:	the head for your list.
- */
-#define list_for_each_safe(pos, n, head) \
-	for (pos = (head)->next, n = pos->next; pos != (head); \
-		pos = n, n = pos->next)
-
-/**
- * list_for_each_entry	-	iterate over list of given type
- * @pos:	the type * to use as a loop counter.
- * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
- */
-#define list_for_each_entry(pos, head, member)				\
-	for (pos = list_entry((head)->next, typeof(*pos), member);	\
-	     &pos->member != (head); 					\
-	     pos = list_entry(pos->member.next, typeof(*pos), member))
-
-/**
- * list_for_each_entry_safe - iterate over list of given type safe against removal of list entry
- * @pos:	the type * to use as a loop counter.
- * @n:		another type * to use as temporary storage
- * @head:	the head for your list.
- * @member:	the name of the list_struct within the struct.
- */
-#define list_for_each_entry_safe(pos, n, head, member)			\
-	for (pos = list_entry((head)->next, typeof(*pos), member),	\
-		n = list_entry(pos->member.next, typeof(*pos), member);	\
-	     &pos->member != (head); 					\
-	     pos = n, n = list_entry(n->member.next, typeof(*n), member))
-
-/**
- * list_for_each_entry_continue -       iterate over list of given type
- *                      continuing after existing point
- * @pos:        the type * to use as a loop counter.
- * @head:       the head for your list.
- * @member:     the name of the list_struct within the struct.
- */
-#define list_for_each_entry_continue(pos, head, member)			\
-	for (pos = list_entry(pos->member.next, typeof(*pos), member),	\
-		     prefetch(pos->member.next);			\
-	     &pos->member != (head);					\
-	     pos = list_entry(pos->member.next, typeof(*pos), member),	\
-		     prefetch(pos->member.next))
-
-#endif

options.c

@@ -1,440 +0,0 @@
-/* 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 2016-2017 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 <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include <jansson.h>
-#include <curl/curl.h>
-#include <getopt.h>
-
-#include "version.h"
-#include "utils/applog.h"
-#include "options.h"
-#include "cpu.h"
-#include "donate.h"
-#include "algo/cryptonight/cryptonight.h"
-
-
-int64_t opt_affinity     = -1L;
-int     opt_n_threads    = 0;
-int     opt_algo_variant = 0;
-int     opt_retries      = 5;
-int     opt_retry_pause  = 5;
-int     opt_donate_level = DONATE_LEVEL;
-bool    opt_colors       = true;
-bool    opt_keepalive    = false;
-bool    opt_background   = false;
-char    *opt_url         = NULL;
-char    *opt_backup_url  = NULL;
-char    *opt_userpass    = NULL;
-char    *opt_user        = NULL;
-char    *opt_pass        = NULL;
-
-
-static char const usage[] = "\
-Usage: " APP_ID " [OPTIONS]\n\
-Options:\n\
-  -o, --url=URL         URL of mining server\n\
-  -b, --backup-url=URL  URL of backup mining server\n\
-  -O, --userpass=U:P    username:password pair for mining server\n\
-  -u, --user=USERNAME   username for mining server\n\
-  -p, --pass=PASSWORD   password for mining server\n\
-  -t, --threads=N       number of miner threads\n\
-  -v, --av=N            algorithm variation, 0 auto select\n\
-  -k, --keepalive       send keepalived for prevent timeout (need pool support)\n\
-  -r, --retries=N       number of times to retry before switch to backup server (default: 5)\n\
-  -R, --retry-pause=N   time to pause between retries (default: 5)\n\
-      --cpu-affinity    set process affinity to cpu core(s), mask 0x3 for cores 0 and 1\n\
-      --no-color        disable colored output\n\
-      --donate-level=N  donate level, default 5%% (5 minutes in 100 minutes)\n\
-  -B, --background      run the miner in the background\n\
-  -c, --config=FILE     load a JSON-format configuration file\n\
-  -h, --help            display this help and exit\n\
-  -V, --version         output version information and exit\n\
-";
-
-
-static char const short_options[] = "a:c:khBp:Px:r:R:s:t:T:o:u:O:v:Vb:";
-
-
-static struct option const options[] = {
-    { "algo",         1, NULL, 'a'  },
-    { "av",           1, NULL, 'v'  },
-    { "background",   0, NULL, 'B'  },
-    { "backup-url",   1, NULL, 'b'  },
-    { "config",       1, NULL, 'c'  },
-    { "cpu-affinity", 1, NULL, 1020 },
-    { "donate-level", 1, NULL, 1003 },
-    { "help",         0, NULL, 'h'  },
-    { "keepalive",    0, NULL ,'k'  },
-    { "no-color",     0, NULL, 1002 },
-    { "pass",         1, NULL, 'p'  },
-    { "retries",      1, NULL, 'r'  },
-    { "retry-pause",  1, NULL, 'R'  },
-    { "threads",      1, NULL, 't'  },
-    { "url",          1, NULL, 'o'  },
-    { "user",         1, NULL, 'u'  },
-    { "userpass",     1, NULL, 'O'  },
-    { "version",      0, NULL, 'V'  },
-    { 0, 0, 0, 0 }
-};
-
-
-static int get_algo_variant(int variant) {
-   if (variant > XMR_AV0_AUTO && variant < XMR_AV_MAX) {
-       return variant;
-   }
-
-   if (cpu_info.flags & CPU_FLAG_AES) {
-       return XMR_AV1_AESNI;
-   }
-
-   return XMR_AV4_SOFT_AES;
-}
-
-
-static void parse_config(json_t *config, char *ref);
-static char *parse_url(const char *arg);
-
-
-static void parse_arg(int key, char *arg) {
-    char *p;
-    int v;
-    uint64_t ul;
-
-    switch (key)
-    {
-    case 'a':
-        break;
-
-    case 'O': /* --userpass */
-        p = strchr(arg, ':');
-        if (!p) {
-            show_usage_and_exit(1);
-        }
-
-        free(opt_userpass);
-        opt_userpass = strdup(arg);
-        free(opt_user);
-        opt_user = calloc(p - arg + 1, 1);
-        strncpy(opt_user, arg, p - arg);
-        free(opt_pass);
-        opt_pass = strdup(p + 1);
-        break;
-
-    case 'o': /* --url */
-        p = parse_url(arg);
-        if (p) {
-            free(opt_url);
-            opt_url = p;
-        }
-        break;
-
-    case 'b': /* --backup-url */
-        p = parse_url(arg);
-        if (p) {
-            free(opt_backup_url);
-            opt_backup_url = p;
-        }
-        break;
-
-    case 'u': /* --user */
-        free(opt_user);
-        opt_user = strdup(arg);
-        break;
-
-    case 'p': /* --pass */
-        free(opt_pass);
-        opt_pass = strdup(arg);
-        break;
-
-    case 'r': /* --retries */
-        v = atoi(arg);
-        if (v < 1 || v > 1000) {
-            show_usage_and_exit(1);
-        }
-
-        opt_retries = v;
-        break;
-
-    case 'R': /* --retry-pause */
-        v = atoi(arg);
-        if (v < 1 || v > 3600) {
-            show_usage_and_exit(1);
-        }
-
-        opt_retry_pause = v;
-        break;
-
-    case 't': /* --threads */
-        v = atoi(arg);
-        if (v < 1 || v > 1024) {
-            show_usage_and_exit(1);
-        }
-
-        opt_n_threads = v;
-        break;
-
-    case 'k':
-        opt_keepalive = true;
-        break;
-
-    case 'V': /* --version */
-        show_version_and_exit();
-        break;
-
-    case 'h': /* --help */
-        show_usage_and_exit(0);
-        break;
-
-    case 'c': { /* --config */
-        json_error_t err;
-        json_t *config = json_load_file(arg, 0, &err);
-
-        if (!json_is_object(config)) {
-            if (err.line < 0) {
-                applog(LOG_ERR, "%s\n", err.text);
-            }
-            else {
-                applog(LOG_ERR, "%s:%d: %s\n", arg, err.line, err.text);
-            }
-        } else {
-            parse_config(config, arg);
-            json_decref(config);
-        }
-        break;
-    }
-
-    case 'B':
-        opt_background = true;
-        opt_colors = false;
-        break;
-
-    case 'v': /* --av */
-        v = atoi(arg);
-        if (v < 0 || v > XMR_AV_MAX) {
-            show_usage_and_exit(1);
-        }
-
-        opt_algo_variant = v;
-        break;
-
-    case 1020: /* --cpu-affinity */
-        p  = strstr(arg, "0x");
-        ul = p ? strtoul(p, NULL, 16) : atol(arg);
-        if (ul > (1UL << cpu_info.count) -1) {
-            ul = -1;
-        }
-
-        opt_affinity = ul;
-        break;
-
-    case 1002: /* --no-color */
-        opt_colors = false;
-        break;
-
-    case 1003:
-        v = atoi(arg);
-        if (v < 1 || v > 99) {
-            show_usage_and_exit(1);
-        }
-
-        opt_donate_level = v;
-        break;
-
-    default:
-        show_usage_and_exit(1);
-    }
-}
-
-
-static void parse_config(json_t *config, char *ref)
-{
-    int i;
-    char buf[16];
-    json_t *val;
-
-    applog(LOG_ERR, ref);
-
-    for (i = 0; i < ARRAY_SIZE(options); i++) {
-        if (!options[i].name) {
-            break;
-        }
-
-        val = json_object_get(config, options[i].name);
-        if (!val) {
-            continue;
-        }
-
-        if (options[i].has_arg && json_is_string(val)) {
-            char *s = strdup(json_string_value(val));
-            if (!s) {
-                break;
-            }
-
-            parse_arg(options[i].val, s);
-            free(s);
-        }
-        else if (options[i].has_arg && json_is_integer(val)) {
-            sprintf(buf, "%d", (int) json_integer_value(val));
-            parse_arg(options[i].val, buf);
-        }
-        else if (options[i].has_arg && json_is_real(val)) {
-            sprintf(buf, "%f", json_real_value(val));
-            parse_arg(options[i].val, buf);
-        }
-        else if (!options[i].has_arg) {
-            if (json_is_true(val)) {
-               parse_arg(options[i].val, "");
-            }
-        }
-        else {
-            applog(LOG_ERR, "JSON option %s invalid", options[i].name);
-        }
-    }
-}
-
-
-static char *parse_url(const char *arg)
-{
-    char *p = strstr(arg, "://");
-    if (p) {
-        if (strncasecmp(arg, "stratum+tcp://", 14)) {
-            show_usage_and_exit(1);
-        }
-
-        return strdup(arg);
-    }
-
-
-    if (!strlen(arg) || *arg == '/') {
-        show_usage_and_exit(1);
-    }
-
-    char *dest = malloc(strlen(arg) + 16);
-    sprintf(dest, "stratum+tcp://%s", arg);
-
-    return dest;
-}
-
-
-/**
- * Parse application command line via getopt.
- */
-void parse_cmdline(int argc, char *argv[]) {
-    opt_user = strdup("x");
-    opt_pass = strdup("x");
-
-    int key;
-
-    while (1) {
-        key = getopt_long(argc, argv, short_options, options, NULL);
-        if (key < 0) {
-            break;
-        }
-
-        parse_arg(key, optarg);
-    }
-
-    if (optind < argc) {
-        fprintf(stderr, "%s: unsupported non-option argument '%s'\n", argv[0], argv[optind]);
-        show_usage_and_exit(1);
-    }
-
-    if (!opt_url) {
-        opt_url = strdup("stratum+tcp://proxy.xmrig.com:443");
-        opt_keepalive = true;
-
-        if (!opt_backup_url) {
-            opt_backup_url = strdup("stratum+tcp://failover.xmrig.com:80");
-        }
-    }
-
-    if (!opt_userpass) {
-        opt_userpass = malloc(strlen(opt_user) + strlen(opt_pass) + 2);
-        if (!opt_userpass) {
-            proper_exit(1);
-        }
-
-        sprintf(opt_userpass, "%s:%s", opt_user, opt_pass);
-    }
-
-    if (!opt_n_threads) {
-        opt_n_threads = get_optimal_threads_count();
-    }
-
-    opt_algo_variant = get_algo_variant(opt_algo_variant);
-    if (!opt_algo_variant) {
-        opt_algo_variant = get_algo_variant(0);
-    }
-
-    if (opt_donate_level < 1 || opt_donate_level > 99) {
-        opt_donate_level = 1;
-    }
-
-    if (!cryptonight_init(opt_algo_variant)) {
-        applog(LOG_ERR, "Cryptonight hash self-test failed. This might be caused by bad compiler optimizations.");
-        proper_exit(1);
-    }
-}
-
-
-void show_usage_and_exit(int status) {
-    if (status) {
-        fprintf(stderr, "Try \"" APP_ID "\" --help' for more information.\n");
-    }
-    else {
-        printf(usage);
-    }
-
-    proper_exit(status);
-}
-
-
-void show_version_and_exit(void) {
-    printf(APP_NAME " " APP_VERSION "\n built on " __DATE__
-
-    #ifdef __GNUC__
-    " with GCC");
-    printf(" %d.%d.%d", __GNUC__, __GNUC_MINOR__, __GNUC_PATCHLEVEL__);
-    #endif
-
-    printf("\n features:"
-    #ifdef __i386__
-    " i386"
-    #endif
-    #ifdef __x86_64__
-    " x86_64"
-    #endif
-    #ifdef __AES__
-    " AES-NI"
-    #endif
-    "\n");
-
-    printf("\n%s\n", curl_version());
-    #ifdef JANSSON_VERSION
-    printf("libjansson/%s\n", JANSSON_VERSION);
-    #endif
-    proper_exit(0);
-}

res/app.rc

@@ -1,5 +1,5 @@
 #include <windows.h>
-#include "../version.h"
+#include "../src/version.h"
 
 IDI_ICON1    ICON    DISCARDABLE    "app.ico"
 

src/3rdparty/align.h

@@ -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 2016-2017 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 __ALIGN_H__
+#define __ALIGN_H__
+
+#ifdef _MSC_VER
+#   define VAR_ALIGN(x, decl) __declspec(align(x)) decl
+#else
+#   define VAR_ALIGN(x, decl) decl __attribute__ ((aligned(x)))
+#endif
+
+#endif /* __ALIGN_H__ */

src/3rdparty/getopt/getopt.h

@@ -0,0 +1,653 @@
+#ifndef __GETOPT_H__
+/**
+ * DISCLAIMER
+ * This file is part of the mingw-w64 runtime package.
+ *
+ * The mingw-w64 runtime package and its code is distributed in the hope that it 
+ * will be useful but WITHOUT ANY WARRANTY.  ALL WARRANTIES, EXPRESSED OR 
+ * IMPLIED ARE HEREBY DISCLAIMED.  This includes but is not limited to 
+ * warranties of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ */
+ /*
+ * Copyright (c) 2002 Todd C. Miller <Todd.Miller@courtesan.com>
+ *
+ * Permission to use, copy, modify, and distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ *
+ * Sponsored in part by the Defense Advanced Research Projects
+ * Agency (DARPA) and Air Force Research Laboratory, Air Force
+ * Materiel Command, USAF, under agreement number F39502-99-1-0512.
+ */
+/*-
+ * Copyright (c) 2000 The NetBSD Foundation, Inc.
+ * All rights reserved.
+ *
+ * This code is derived from software contributed to The NetBSD Foundation
+ * by Dieter Baron and Thomas Klausner.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. 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 THE FOUNDATION OR 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 THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#pragma warning(disable:4996);
+
+#define __GETOPT_H__
+
+/* All the headers include this file. */
+#include <crtdefs.h>
+#include <errno.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdarg.h>
+#include <stdio.h>
+#include <windows.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define	REPLACE_GETOPT		/* use this getopt as the system getopt(3) */
+
+#ifdef REPLACE_GETOPT
+int	opterr = 1;		/* if error message should be printed */
+int	optind = 1;		/* index into parent argv vector */
+int	optopt = '?';		/* character checked for validity */
+#undef	optreset		/* see getopt.h */
+#define	optreset		__mingw_optreset
+int	optreset;		/* reset getopt */
+char    *optarg;		/* argument associated with option */
+#endif
+
+//extern int optind;		/* index of first non-option in argv      */
+//extern int optopt;		/* single option character, as parsed     */
+//extern int opterr;		/* flag to enable built-in diagnostics... */
+//				/* (user may set to zero, to suppress)    */
+//
+//extern char *optarg;		/* pointer to argument of current option  */
+
+#define PRINT_ERROR	((opterr) && (*options != ':'))
+
+#define FLAG_PERMUTE	0x01	/* permute non-options to the end of argv */
+#define FLAG_ALLARGS	0x02	/* treat non-options as args to option "-1" */
+#define FLAG_LONGONLY	0x04	/* operate as getopt_long_only */
+
+/* return values */
+#define	BADCH		(int)'?'
+#define	BADARG		((*options == ':') ? (int)':' : (int)'?')
+#define	INORDER 	(int)1
+
+#ifndef __CYGWIN__
+#define __progname __argv[0]
+#else
+extern char __declspec(dllimport) *__progname;
+#endif
+
+#ifdef __CYGWIN__
+static char EMSG[] = "";
+#else
+#define	EMSG		""
+#endif
+
+static int getopt_internal(int, char * const *, const char *,
+			   const struct option *, int *, int);
+static int parse_long_options(char * const *, const char *,
+			      const struct option *, int *, int);
+static int gcd(int, int);
+static void permute_args(int, int, int, char * const *);
+
+static char *place = EMSG; /* option letter processing */
+
+/* XXX: set optreset to 1 rather than these two */
+static int nonopt_start = -1; /* first non option argument (for permute) */
+static int nonopt_end = -1;   /* first option after non options (for permute) */
+
+/* Error messages */
+static const char recargchar[] = "option requires an argument -- %c";
+static const char recargstring[] = "option requires an argument -- %s";
+static const char ambig[] = "ambiguous option -- %.*s";
+static const char noarg[] = "option doesn't take an argument -- %.*s";
+static const char illoptchar[] = "unknown option -- %c";
+static const char illoptstring[] = "unknown option -- %s";
+
+static void
+_vwarnx(const char *fmt,va_list ap)
+{
+  (void)fprintf(stderr,"%s: ",__progname);
+  if (fmt != NULL)
+    (void)vfprintf(stderr,fmt,ap);
+  (void)fprintf(stderr,"\n");
+}
+
+static void
+warnx(const char *fmt,...)
+{
+  va_list ap;
+  va_start(ap,fmt);
+  _vwarnx(fmt,ap);
+  va_end(ap);
+}
+
+/*
+ * Compute the greatest common divisor of a and b.
+ */
+static int
+gcd(int a, int b)
+{
+	int c;
+
+	c = a % b;
+	while (c != 0) {
+		a = b;
+		b = c;
+		c = a % b;
+	}
+
+	return (b);
+}
+
+/*
+ * Exchange the block from nonopt_start to nonopt_end with the block
+ * from nonopt_end to opt_end (keeping the same order of arguments
+ * in each block).
+ */
+static void
+permute_args(int panonopt_start, int panonopt_end, int opt_end,
+	char * const *nargv)
+{
+	int cstart, cyclelen, i, j, ncycle, nnonopts, nopts, pos;
+	char *swap;
+
+	/*
+	 * compute lengths of blocks and number and size of cycles
+	 */
+	nnonopts = panonopt_end - panonopt_start;
+	nopts = opt_end - panonopt_end;
+	ncycle = gcd(nnonopts, nopts);
+	cyclelen = (opt_end - panonopt_start) / ncycle;
+
+	for (i = 0; i < ncycle; i++) {
+		cstart = panonopt_end+i;
+		pos = cstart;
+		for (j = 0; j < cyclelen; j++) {
+			if (pos >= panonopt_end)
+				pos -= nnonopts;
+			else
+				pos += nopts;
+			swap = nargv[pos];
+			/* LINTED const cast */
+			((char **) nargv)[pos] = nargv[cstart];
+			/* LINTED const cast */
+			((char **)nargv)[cstart] = swap;
+		}
+	}
+}
+
+#ifdef REPLACE_GETOPT
+/*
+ * getopt --
+ *	Parse argc/argv argument vector.
+ *
+ * [eventually this will replace the BSD getopt]
+ */
+int
+getopt(int nargc, char * const *nargv, const char *options)
+{
+
+	/*
+	 * We don't pass FLAG_PERMUTE to getopt_internal() since
+	 * the BSD getopt(3) (unlike GNU) has never done this.
+	 *
+	 * Furthermore, since many privileged programs call getopt()
+	 * before dropping privileges it makes sense to keep things
+	 * as simple (and bug-free) as possible.
+	 */
+	return (getopt_internal(nargc, nargv, options, NULL, NULL, 0));
+}
+#endif /* REPLACE_GETOPT */
+
+//extern int getopt(int nargc, char * const *nargv, const char *options);
+
+#ifdef _BSD_SOURCE
+/*
+ * BSD adds the non-standard `optreset' feature, for reinitialisation
+ * of `getopt' parsing.  We support this feature, for applications which
+ * proclaim their BSD heritage, before including this header; however,
+ * to maintain portability, developers are advised to avoid it.
+ */
+# define optreset  __mingw_optreset
+extern int optreset;
+#endif
+#ifdef __cplusplus
+}
+#endif
+/*
+ * POSIX requires the `getopt' API to be specified in `unistd.h';
+ * thus, `unistd.h' includes this header.  However, we do not want
+ * to expose the `getopt_long' or `getopt_long_only' APIs, when
+ * included in this manner.  Thus, close the standard __GETOPT_H__
+ * declarations block, and open an additional __GETOPT_LONG_H__
+ * specific block, only when *not* __UNISTD_H_SOURCED__, in which
+ * to declare the extended API.
+ */
+#endif /* !defined(__GETOPT_H__) */
+
+#if !defined(__UNISTD_H_SOURCED__) && !defined(__GETOPT_LONG_H__)
+#define __GETOPT_LONG_H__
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+struct option		/* specification for a long form option...	*/
+{
+  const char *name;		/* option name, without leading hyphens */
+  int         has_arg;		/* does it take an argument?		*/
+  int        *flag;		/* where to save its status, or NULL	*/
+  int         val;		/* its associated status value		*/
+};
+
+enum    		/* permitted values for its `has_arg' field...	*/
+{
+  no_argument = 0,      	/* option never takes an argument	*/
+  required_argument,		/* option always requires an argument	*/
+  optional_argument		/* option may take an argument		*/
+};
+
+/*
+ * parse_long_options --
+ *	Parse long options in argc/argv argument vector.
+ * Returns -1 if short_too is set and the option does not match long_options.
+ */
+static int
+parse_long_options(char * const *nargv, const char *options,
+	const struct option *long_options, int *idx, int short_too)
+{
+	char *current_argv, *has_equal;
+	size_t current_argv_len;
+	int i, ambiguous, match;
+
+#define IDENTICAL_INTERPRETATION(_x, _y)                                \
+	(long_options[(_x)].has_arg == long_options[(_y)].has_arg &&    \
+	 long_options[(_x)].flag == long_options[(_y)].flag &&          \
+	 long_options[(_x)].val == long_options[(_y)].val)
+
+	current_argv = place;
+	match = -1;
+	ambiguous = 0;
+
+	optind++;
+
+	if ((has_equal = strchr(current_argv, '=')) != NULL) {
+		/* argument found (--option=arg) */
+		current_argv_len = has_equal - current_argv;
+		has_equal++;
+	} else
+		current_argv_len = strlen(current_argv);
+
+	for (i = 0; long_options[i].name; i++) {
+		/* find matching long option */
+		if (strncmp(current_argv, long_options[i].name,
+		    current_argv_len))
+			continue;
+
+		if (strlen(long_options[i].name) == current_argv_len) {
+			/* exact match */
+			match = i;
+			ambiguous = 0;
+			break;
+		}
+		/*
+		 * If this is a known short option, don't allow
+		 * a partial match of a single character.
+		 */
+		if (short_too && current_argv_len == 1)
+			continue;
+
+		if (match == -1)	/* partial match */
+			match = i;
+		else if (!IDENTICAL_INTERPRETATION(i, match))
+			ambiguous = 1;
+	}
+	if (ambiguous) {
+		/* ambiguous abbreviation */
+		if (PRINT_ERROR)
+			warnx(ambig, (int)current_argv_len,
+			     current_argv);
+		optopt = 0;
+		return (BADCH);
+	}
+	if (match != -1) {		/* option found */
+		if (long_options[match].has_arg == no_argument
+		    && has_equal) {
+			if (PRINT_ERROR)
+				warnx(noarg, (int)current_argv_len,
+				     current_argv);
+			/*
+			 * XXX: GNU sets optopt to val regardless of flag
+			 */
+			if (long_options[match].flag == NULL)
+				optopt = long_options[match].val;
+			else
+				optopt = 0;
+			return (BADARG);
+		}
+		if (long_options[match].has_arg == required_argument ||
+		    long_options[match].has_arg == optional_argument) {
+			if (has_equal)
+				optarg = has_equal;
+			else if (long_options[match].has_arg ==
+			    required_argument) {
+				/*
+				 * optional argument doesn't use next nargv
+				 */
+				optarg = nargv[optind++];
+			}
+		}
+		if ((long_options[match].has_arg == required_argument)
+		    && (optarg == NULL)) {
+			/*
+			 * Missing argument; leading ':' indicates no error
+			 * should be generated.
+			 */
+			if (PRINT_ERROR)
+				warnx(recargstring,
+				    current_argv);
+			/*
+			 * XXX: GNU sets optopt to val regardless of flag
+			 */
+			if (long_options[match].flag == NULL)
+				optopt = long_options[match].val;
+			else
+				optopt = 0;
+			--optind;
+			return (BADARG);
+		}
+	} else {			/* unknown option */
+		if (short_too) {
+			--optind;
+			return (-1);
+		}
+		if (PRINT_ERROR)
+			warnx(illoptstring, current_argv);
+		optopt = 0;
+		return (BADCH);
+	}
+	if (idx)
+		*idx = match;
+	if (long_options[match].flag) {
+		*long_options[match].flag = long_options[match].val;
+		return (0);
+	} else
+		return (long_options[match].val);
+#undef IDENTICAL_INTERPRETATION
+}
+
+/*
+ * getopt_internal --
+ *	Parse argc/argv argument vector.  Called by user level routines.
+ */
+static int
+getopt_internal(int nargc, char * const *nargv, const char *options,
+	const struct option *long_options, int *idx, int flags)
+{
+	char *oli;				/* option letter list index */
+	int optchar, short_too;
+	static int posixly_correct = -1;
+
+	if (options == NULL)
+		return (-1);
+
+	/*
+	 * XXX Some GNU programs (like cvs) set optind to 0 instead of
+	 * XXX using optreset.  Work around this braindamage.
+	 */
+	if (optind == 0)
+		optind = optreset = 1;
+
+	/*
+	 * Disable GNU extensions if POSIXLY_CORRECT is set or options
+	 * string begins with a '+'.
+	 *
+	 * CV, 2009-12-14: Check POSIXLY_CORRECT anew if optind == 0 or
+	 *                 optreset != 0 for GNU compatibility.
+	 */
+	if (posixly_correct == -1 || optreset != 0)
+		posixly_correct = (getenv("POSIXLY_CORRECT") != NULL);
+	if (*options == '-')
+		flags |= FLAG_ALLARGS;
+	else if (posixly_correct || *options == '+')
+		flags &= ~FLAG_PERMUTE;
+	if (*options == '+' || *options == '-')
+		options++;
+
+	optarg = NULL;
+	if (optreset)
+		nonopt_start = nonopt_end = -1;
+start:
+	if (optreset || !*place) {		/* update scanning pointer */
+		optreset = 0;
+		if (optind >= nargc) {          /* end of argument vector */
+			place = EMSG;
+			if (nonopt_end != -1) {
+				/* do permutation, if we have to */
+				permute_args(nonopt_start, nonopt_end,
+				    optind, nargv);
+				optind -= nonopt_end - nonopt_start;
+			}
+			else if (nonopt_start != -1) {
+				/*
+				 * If we skipped non-options, set optind
+				 * to the first of them.
+				 */
+				optind = nonopt_start;
+			}
+			nonopt_start = nonopt_end = -1;
+			return (-1);
+		}
+		if (*(place = nargv[optind]) != '-' ||
+		    (place[1] == '\0' && strchr(options, '-') == NULL)) {
+			place = EMSG;		/* found non-option */
+			if (flags & FLAG_ALLARGS) {
+				/*
+				 * GNU extension:
+				 * return non-option as argument to option 1
+				 */
+				optarg = nargv[optind++];
+				return (INORDER);
+			}
+			if (!(flags & FLAG_PERMUTE)) {
+				/*
+				 * If no permutation wanted, stop parsing
+				 * at first non-option.
+				 */
+				return (-1);
+			}
+			/* do permutation */
+			if (nonopt_start == -1)
+				nonopt_start = optind;
+			else if (nonopt_end != -1) {
+				permute_args(nonopt_start, nonopt_end,
+				    optind, nargv);
+				nonopt_start = optind -
+				    (nonopt_end - nonopt_start);
+				nonopt_end = -1;
+			}
+			optind++;
+			/* process next argument */
+			goto start;
+		}
+		if (nonopt_start != -1 && nonopt_end == -1)
+			nonopt_end = optind;
+
+		/*
+		 * If we have "-" do nothing, if "--" we are done.
+		 */
+		if (place[1] != '\0' && *++place == '-' && place[1] == '\0') {
+			optind++;
+			place = EMSG;
+			/*
+			 * We found an option (--), so if we skipped
+			 * non-options, we have to permute.
+			 */
+			if (nonopt_end != -1) {
+				permute_args(nonopt_start, nonopt_end,
+				    optind, nargv);
+				optind -= nonopt_end - nonopt_start;
+			}
+			nonopt_start = nonopt_end = -1;
+			return (-1);
+		}
+	}
+
+	/*
+	 * Check long options if:
+	 *  1) we were passed some
+	 *  2) the arg is not just "-"
+	 *  3) either the arg starts with -- we are getopt_long_only()
+	 */
+	if (long_options != NULL && place != nargv[optind] &&
+	    (*place == '-' || (flags & FLAG_LONGONLY))) {
+		short_too = 0;
+		if (*place == '-')
+			place++;		/* --foo long option */
+		else if (*place != ':' && strchr(options, *place) != NULL)
+			short_too = 1;		/* could be short option too */
+
+		optchar = parse_long_options(nargv, options, long_options,
+		    idx, short_too);
+		if (optchar != -1) {
+			place = EMSG;
+			return (optchar);
+		}
+	}
+
+	if ((optchar = (int)*place++) == (int)':' ||
+	    (optchar == (int)'-' && *place != '\0') ||
+	    (oli = (char*)strchr(options, optchar)) == NULL) {
+		/*
+		 * If the user specified "-" and  '-' isn't listed in
+		 * options, return -1 (non-option) as per POSIX.
+		 * Otherwise, it is an unknown option character (or ':').
+		 */
+		if (optchar == (int)'-' && *place == '\0')
+			return (-1);
+		if (!*place)
+			++optind;
+		if (PRINT_ERROR)
+			warnx(illoptchar, optchar);
+		optopt = optchar;
+		return (BADCH);
+	}
+	if (long_options != NULL && optchar == 'W' && oli[1] == ';') {
+		/* -W long-option */
+		if (*place)			/* no space */
+			/* NOTHING */;
+		else if (++optind >= nargc) {	/* no arg */
+			place = EMSG;
+			if (PRINT_ERROR)
+				warnx(recargchar, optchar);
+			optopt = optchar;
+			return (BADARG);
+		} else				/* white space */
+			place = nargv[optind];
+		optchar = parse_long_options(nargv, options, long_options,
+		    idx, 0);
+		place = EMSG;
+		return (optchar);
+	}
+	if (*++oli != ':') {			/* doesn't take argument */
+		if (!*place)
+			++optind;
+	} else {				/* takes (optional) argument */
+		optarg = NULL;
+		if (*place)			/* no white space */
+			optarg = place;
+		else if (oli[1] != ':') {	/* arg not optional */
+			if (++optind >= nargc) {	/* no arg */
+				place = EMSG;
+				if (PRINT_ERROR)
+					warnx(recargchar, optchar);
+				optopt = optchar;
+				return (BADARG);
+			} else
+				optarg = nargv[optind];
+		}
+		place = EMSG;
+		++optind;
+	}
+	/* dump back option letter */
+	return (optchar);
+}
+
+/*
+ * getopt_long --
+ *	Parse argc/argv argument vector.
+ */
+int
+getopt_long(int nargc, char * const *nargv, const char *options,
+    const struct option *long_options, int *idx)
+{
+
+	return (getopt_internal(nargc, nargv, options, long_options, idx,
+	    FLAG_PERMUTE));
+}
+
+/*
+ * getopt_long_only --
+ *	Parse argc/argv argument vector.
+ */
+int
+getopt_long_only(int nargc, char * const *nargv, const char *options,
+    const struct option *long_options, int *idx)
+{
+
+	return (getopt_internal(nargc, nargv, options, long_options, idx,
+	    FLAG_PERMUTE|FLAG_LONGONLY));
+}
+
+//extern int getopt_long(int nargc, char * const *nargv, const char *options,
+//    const struct option *long_options, int *idx);
+//extern int getopt_long_only(int nargc, char * const *nargv, const char *options,
+//    const struct option *long_options, int *idx);
+/*
+ * Previous MinGW implementation had...
+ */
+#ifndef HAVE_DECL_GETOPT
+/*
+ * ...for the long form API only; keep this for compatibility.
+ */
+# define HAVE_DECL_GETOPT	1
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* !defined(__UNISTD_H_SOURCED__) && !defined(__GETOPT_LONG_H__) */

src/3rdparty/jansson/CMakeLists.txt

@@ -6,6 +6,8 @@ add_definitions(-DHAVE_CONFIG_H)
 # Add the lib sources.
 file(GLOB JANSSON_SRC *.c)
 
+set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Os")
+
 set(JANSSON_HDR_PRIVATE
    ${CMAKE_CURRENT_SOURCE_DIR}/hashtable.h
    ${CMAKE_CURRENT_SOURCE_DIR}/jansson_private.h

src/3rdparty/jansson/dump.c

@@ -9,13 +9,17 @@
 #define _GNU_SOURCE
 #endif
 
+#include "jansson_private.h"
+
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <assert.h>
+#ifdef HAVE_UNISTD_H
+#include <unistd.h>
+#endif
 
 #include "jansson.h"
-#include "jansson_private.h"
 #include "strbuffer.h"
 #include "utf.h"
 
@@ -25,11 +29,28 @@
 #define FLAGS_TO_INDENT(f)      ((f) & 0x1F)
 #define FLAGS_TO_PRECISION(f)   (((f) >> 11) & 0x1F)
 
+struct buffer {
+    const size_t size;
+    size_t used;
+    char *data;
+};
+
 static int dump_to_strbuffer(const char *buffer, size_t size, void *data)
 {
     return strbuffer_append_bytes((strbuffer_t *)data, buffer, size);
 }
 
+static int dump_to_buffer(const char *buffer, size_t size, void *data)
+{
+    struct buffer *buf = (struct buffer *)data;
+
+    if(buf->used + size <= buf->size)
+        memcpy(&buf->data[buf->used], buffer, size);
+
+    buf->used += size;
+    return 0;
+}
+
 static int dump_to_file(const char *buffer, size_t size, void *data)
 {
     FILE *dest = (FILE *)data;
@@ -38,6 +59,16 @@ static int dump_to_file(const char *buffer, size_t size, void *data)
     return 0;
 }
 
+static int dump_to_fd(const char *buffer, size_t size, void *data)
+{
+    int *dest = (int *)data;
+#ifdef HAVE_UNISTD_H
+    if(write(*dest, buffer, size) == (ssize_t)size)
+        return 0;
+#endif
+    return -1;
+}
+
 /* 32 spaces (the maximum indentation size) */
 static const char whitespace[] = "                                ";
 
@@ -168,6 +199,10 @@ static int compare_keys(const void *key1, const void *key2)
 static int do_dump(const json_t *json, size_t flags, int depth,
                    json_dump_callback_t dump, void *data)
 {
+    int embed = flags & JSON_EMBED;
+
+    flags &= ~JSON_EMBED;
+
     if(!json)
         return -1;
 
@@ -227,11 +262,11 @@ static int do_dump(const json_t *json, size_t flags, int depth,
 
             n = json_array_size(json);
 
-            if(dump("[", 1, data))
+            if(!embed && dump("[", 1, data))
                 goto array_error;
             if(n == 0) {
                 array->visited = 0;
-                return dump("]", 1, data);
+                return embed ? 0 : dump("]", 1, data);
             }
             if(dump_indent(flags, depth + 1, 0, dump, data))
                 goto array_error;
@@ -255,7 +290,7 @@ static int do_dump(const json_t *json, size_t flags, int depth,
             }
 
             array->visited = 0;
-            return dump("]", 1, data);
+            return embed ? 0 : dump("]", 1, data);
 
         array_error:
             array->visited = 0;
@@ -286,11 +321,11 @@ static int do_dump(const json_t *json, size_t flags, int depth,
 
             iter = json_object_iter((json_t *)json);
 
-            if(dump("{", 1, data))
+            if(!embed && dump("{", 1, data))
                 goto object_error;
             if(!iter) {
                 object->visited = 0;
-                return dump("}", 1, data);
+                return embed ? 0 : dump("}", 1, data);
             }
             if(dump_indent(flags, depth + 1, 0, dump, data))
                 goto object_error;
@@ -386,7 +421,7 @@ static int do_dump(const json_t *json, size_t flags, int depth,
             }
 
             object->visited = 0;
-            return dump("}", 1, data);
+            return embed ? 0 : dump("}", 1, data);
 
         object_error:
             object->visited = 0;
@@ -416,11 +451,26 @@ char *json_dumps(const json_t *json, size_t flags)
     return result;
 }
 
+size_t json_dumpb(const json_t *json, char *buffer, size_t size, size_t flags)
+{
+    struct buffer buf = { size, 0, buffer };
+
+    if(json_dump_callback(json, dump_to_buffer, (void *)&buf, flags))
+        return 0;
+
+    return buf.used;
+}
+
 int json_dumpf(const json_t *json, FILE *output, size_t flags)
 {
     return json_dump_callback(json, dump_to_file, (void *)output, flags);
 }
 
+int json_dumpfd(const json_t *json, int output, size_t flags)
+{
+    return json_dump_callback(json, dump_to_fd, (void *)&output, flags);
+}
+
 int json_dump_file(const json_t *json, const char *path, size_t flags)
 {
     int result;

src/3rdparty/jansson/hashtable_seed.c

@@ -168,12 +168,12 @@ static uint32_t generate_seed() {
     int done = 0;
 
 #if !defined(_WIN32) && defined(USE_URANDOM)
-    if (!done && seed_from_urandom(&seed) == 0)
+    if (seed_from_urandom(&seed) == 0)
         done = 1;
 #endif
 
 #if defined(_WIN32) && defined(USE_WINDOWS_CRYPTOAPI)
-    if (!done && seed_from_windows_cryptoapi(&seed) == 0)
+    if (seed_from_windows_cryptoapi(&seed) == 0)
         done = 1;
 #endif
 

src/3rdparty/jansson/jansson.h

@@ -21,11 +21,11 @@ extern "C" {
 /* version */
 
 #define JANSSON_MAJOR_VERSION  2
-#define JANSSON_MINOR_VERSION  9
+#define JANSSON_MINOR_VERSION  10
 #define JANSSON_MICRO_VERSION  0
 
 /* Micro version is omitted if it's 0 */
-#define JANSSON_VERSION  "2.9"
+#define JANSSON_VERSION  "2.10"
 
 /* Version as a 3-byte hex number, e.g. 0x010201 == 1.2.1. Use this
    for numeric comparisons, e.g. #if JANSSON_VERSION_HEX >= ... */
@@ -273,6 +273,7 @@ typedef size_t (*json_load_callback_t)(void *buffer, size_t buflen, void *data);
 json_t *json_loads(const char *input, size_t flags, json_error_t *error);
 json_t *json_loadb(const char *buffer, size_t buflen, size_t flags, json_error_t *error);
 json_t *json_loadf(FILE *input, size_t flags, json_error_t *error);
+json_t *json_loadfd(int input, size_t flags, json_error_t *error);
 json_t *json_load_file(const char *path, size_t flags, json_error_t *error);
 json_t *json_load_callback(json_load_callback_t callback, void *data, size_t flags, json_error_t *error);
 
@@ -288,11 +289,14 @@ json_t *json_load_callback(json_load_callback_t callback, void *data, size_t fla
 #define JSON_ENCODE_ANY         0x200
 #define JSON_ESCAPE_SLASH       0x400
 #define JSON_REAL_PRECISION(n)  (((n) & 0x1F) << 11)
+#define JSON_EMBED              0x10000
 
 typedef int (*json_dump_callback_t)(const char *buffer, size_t size, void *data);
 
 char *json_dumps(const json_t *json, size_t flags);
+size_t json_dumpb(const json_t *json, char *buffer, size_t size, size_t flags);
 int json_dumpf(const json_t *json, FILE *output, size_t flags);
+int json_dumpfd(const json_t *json, int output, size_t flags);
 int json_dump_file(const json_t *json, const char *path, size_t flags);
 int json_dump_callback(const json_t *json, json_dump_callback_t callback, void *data, size_t flags);
 

src/3rdparty/jansson/jansson_private.h

@@ -8,6 +8,7 @@
 #ifndef JANSSON_PRIVATE_H
 #define JANSSON_PRIVATE_H
 
+#include "jansson_private_config.h"
 #include <stddef.h>
 #include "jansson.h"
 #include "hashtable.h"

src/3rdparty/jansson/jansson_private_config.h

@@ -2,7 +2,9 @@
 /* jansson_private_config.h.in.  Generated from configure.ac by autoheader.  */
 
 /* Define to 1 if gcc's __atomic builtins are available */
-#define HAVE_ATOMIC_BUILTINS 1
+#ifndef _MSC_VER
+#   define HAVE_ATOMIC_BUILTINS 1
+#endif
 
 /* Define to 1 if you have the `close' function. */
 #define HAVE_CLOSE 1
@@ -20,7 +22,9 @@
 #define HAVE_GETPID 1
 
 /* Define to 1 if you have the `gettimeofday' function. */
-#define HAVE_GETTIMEOFDAY 1
+#ifndef _MSC_VER
+#   define HAVE_GETTIMEOFDAY 1
+#endif
 
 /* Define to 1 if you have the <inttypes.h> header file. */
 #define HAVE_INTTYPES_H 1
@@ -44,10 +48,14 @@
 #define HAVE_READ 1
 
 /* Define to 1 if you have the <sched.h> header file. */
-#define HAVE_SCHED_H 1
+#ifndef _MSC_VER
+#   define HAVE_SCHED_H 1
+#endif
 
 /* Define to 1 if you have the `sched_yield' function. */
-#define HAVE_SCHED_YIELD 1
+#ifndef _MSC_VER
+#   define HAVE_SCHED_YIELD 1
+#endif
 
 /* Define to 1 if you have the <stdint.h> header file. */
 #define HAVE_STDINT_H 1
@@ -68,19 +76,25 @@
 #define HAVE_SYNC_BUILTINS 1
 
 /* Define to 1 if you have the <sys/param.h> header file. */
-#define HAVE_SYS_PARAM_H 1
+#ifndef _MSC_VER
+#   define HAVE_SYS_PARAM_H 1
+#endif
 
 /* Define to 1 if you have the <sys/stat.h> header file. */
 #define HAVE_SYS_STAT_H 1
 
 /* Define to 1 if you have the <sys/time.h> header file. */
-#define HAVE_SYS_TIME_H 1
+#ifndef _MSC_VER
+#   define HAVE_SYS_TIME_H 1
+#endif
 
 /* Define to 1 if you have the <sys/types.h> header file. */
 #define HAVE_SYS_TYPES_H 1
 
 /* Define to 1 if you have the <unistd.h> header file. */
-#define HAVE_UNISTD_H 1
+#ifndef _MSC_VER
+#   define HAVE_UNISTD_H 1
+#endif
 
 /* Define to 1 if the system has the type 'unsigned long long int'. */
 #define HAVE_UNSIGNED_LONG_LONG_INT 1

src/3rdparty/jansson/load.c

@@ -9,15 +9,19 @@
 #define _GNU_SOURCE
 #endif
 
+#include "jansson_private.h"
+
 #include <errno.h>
 #include <limits.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <assert.h>
+#ifdef HAVE_UNISTD_H
+#include <unistd.h>
+#endif
 
 #include "jansson.h"
-#include "jansson_private.h"
 #include "strbuffer.h"
 #include "utf.h"
 
@@ -340,7 +344,7 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
             /* control character */
             lex_unget_unsave(lex, c);
             if(c == '\n')
-                error_set(error, lex, "unexpected newline", c);
+                error_set(error, lex, "unexpected newline");
             else
                 error_set(error, lex, "control character 0x%x", c);
             goto out;
@@ -914,7 +918,7 @@ static json_t *parse_json(lex_t *lex, size_t flags, json_error_t *error)
 typedef struct
 {
     const char *data;
-    int pos;
+    size_t pos;
 } string_data_t;
 
 static int string_get(void *data)
@@ -1028,6 +1032,45 @@ json_t *json_loadf(FILE *input, size_t flags, json_error_t *error)
     return result;
 }
 
+static int fd_get_func(int *fd)
+{
+    uint8_t c;
+#ifdef HAVE_UNISTD_H
+    if (read(*fd, &c, 1) == 1)
+        return c;
+#endif
+    return EOF;
+}
+
+json_t *json_loadfd(int input, size_t flags, json_error_t *error)
+{
+    lex_t lex;
+    const char *source;
+    json_t *result;
+
+#ifdef HAVE_UNISTD_H
+    if(input == STDIN_FILENO)
+        source = "<stdin>";
+    else
+#endif
+        source = "<stream>";
+
+    jsonp_error_init(error, source);
+
+    if (input < 0) {
+        error_set(error, NULL, "wrong arguments");
+        return NULL;
+    }
+
+    if(lex_init(&lex, (get_func)fd_get_func, flags, &input))
+        return NULL;
+
+    result = parse_json(&lex, flags, error);
+
+    lex_close(&lex);
+    return result;
+}
+
 json_t *json_load_file(const char *path, size_t flags, json_error_t *error)
 {
     json_t *result;

src/3rdparty/libcpuid/CMakeLists.txt

@@ -0,0 +1,38 @@
+cmake_minimum_required (VERSION 2.8)
+project (cpuid C)
+
+add_definitions(/DVERSION="0.4.0")
+
+set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} -Os")
+
+set(HEADERS
+    libcpuid.h
+    libcpuid_types.h
+    libcpuid_constants.h
+    libcpuid_internal.h
+    amd_code_t.h
+    intel_code_t.h
+    recog_amd.h
+    recog_intel.h
+    asm-bits.h
+    libcpuid_util.h
+    )
+
+set(SOURCES
+    cpuid_main.c
+    asm-bits.c
+    recog_amd.c
+    recog_intel.c
+    libcpuid_util.c
+   )
+
+if (CMAKE_CL_64)
+    enable_language(ASM_MASM)
+    set(SOURCES_ASM masm-x64.asm)
+endif()
+
+add_library(cpuid STATIC
+    ${HEADERS}
+    ${SOURCES}
+    ${SOURCES_ASM}
+    )

src/3rdparty/libcpuid/amd_code_t.h

@@ -0,0 +1,39 @@
+/*
+ * Copyright 2016  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ * This file contains a list of internal codes we use in detection. It is
+ * of no external use and isn't a complete list of AMD products.
+ */
+	CODE2(OPTERON_800, 1000),
+	CODE(PHENOM),
+	CODE(PHENOM2),
+	CODE(FUSION_C),
+	CODE(FUSION_E),
+	CODE(FUSION_EA),
+	CODE(FUSION_Z),
+	CODE(FUSION_A),
+	

src/3rdparty/libcpuid/asm-bits.c

@@ -0,0 +1,825 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "libcpuid.h"
+#include "asm-bits.h"
+
+int cpuid_exists_by_eflags(void)
+{
+#if defined(PLATFORM_X64)
+	return 1; /* CPUID is always present on the x86_64 */
+#elif defined(PLATFORM_X86)
+#  if defined(COMPILER_GCC)
+	int result;
+	__asm __volatile(
+		"	pushfl\n"
+		"	pop	%%eax\n"
+		"	mov	%%eax,	%%ecx\n"
+		"	xor	$0x200000,	%%eax\n"
+		"	push	%%eax\n"
+		"	popfl\n"
+		"	pushfl\n"
+		"	pop	%%eax\n"
+		"	xor	%%ecx,	%%eax\n"
+		"	mov	%%eax,	%0\n"
+		"	push	%%ecx\n"
+		"	popfl\n"
+		: "=m"(result)
+		: :"eax", "ecx", "memory");
+	return (result != 0);
+#  elif defined(COMPILER_MICROSOFT)
+	int result;
+	__asm {
+		pushfd
+		pop	eax
+		mov	ecx,	eax
+		xor	eax,	0x200000
+		push	eax
+		popfd
+		pushfd
+		pop	eax
+		xor	eax,	ecx
+		mov	result,	eax
+		push	ecx
+		popfd
+	};
+	return (result != 0);
+#  else
+	return 0;
+#  endif /* COMPILER_MICROSOFT */
+#else
+	return 0;
+#endif /* PLATFORM_X86 */
+}
+
+#ifdef INLINE_ASM_SUPPORTED
+/* 
+ * with MSVC/AMD64, the exec_cpuid() and cpu_rdtsc() functions
+ * are implemented in separate .asm files. Otherwise, use inline assembly
+ */
+void exec_cpuid(uint32_t *regs)
+{
+#ifdef COMPILER_GCC
+#	ifdef PLATFORM_X64
+	__asm __volatile(
+		"	mov	%0,	%%rdi\n"
+
+		"	push	%%rbx\n"
+		"	push	%%rcx\n"
+		"	push	%%rdx\n"
+		
+		"	mov	(%%rdi),	%%eax\n"
+		"	mov	4(%%rdi),	%%ebx\n"
+		"	mov	8(%%rdi),	%%ecx\n"
+		"	mov	12(%%rdi),	%%edx\n"
+		
+		"	cpuid\n"
+		
+		"	movl	%%eax,	(%%rdi)\n"
+		"	movl	%%ebx,	4(%%rdi)\n"
+		"	movl	%%ecx,	8(%%rdi)\n"
+		"	movl	%%edx,	12(%%rdi)\n"
+		"	pop	%%rdx\n"
+		"	pop	%%rcx\n"
+		"	pop	%%rbx\n"
+		:
+		:"m"(regs)
+		:"memory", "eax", "rdi"
+	);
+#	else
+	__asm __volatile(
+		"	mov	%0,	%%edi\n"
+
+		"	push	%%ebx\n"
+		"	push	%%ecx\n"
+		"	push	%%edx\n"
+		
+		"	mov	(%%edi),	%%eax\n"
+		"	mov	4(%%edi),	%%ebx\n"
+		"	mov	8(%%edi),	%%ecx\n"
+		"	mov	12(%%edi),	%%edx\n"
+		
+		"	cpuid\n"
+		
+		"	mov	%%eax,	(%%edi)\n"
+		"	mov	%%ebx,	4(%%edi)\n"
+		"	mov	%%ecx,	8(%%edi)\n"
+		"	mov	%%edx,	12(%%edi)\n"
+		"	pop	%%edx\n"
+		"	pop	%%ecx\n"
+		"	pop	%%ebx\n"
+		:
+		:"m"(regs)
+		:"memory", "eax", "edi"
+	);
+#	endif /* COMPILER_GCC */
+#else
+#  ifdef COMPILER_MICROSOFT
+	__asm {
+		push	ebx
+		push	ecx
+		push	edx
+		push	edi
+		mov	edi,	regs
+		
+		mov	eax,	[edi]
+		mov	ebx,	[edi+4]
+		mov	ecx,	[edi+8]
+		mov	edx,	[edi+12]
+		
+		cpuid
+		
+		mov	[edi],		eax
+		mov	[edi+4],	ebx
+		mov	[edi+8],	ecx
+		mov	[edi+12],	edx
+		
+		pop	edi
+		pop	edx
+		pop	ecx
+		pop	ebx
+	}
+#  else
+#    error "Unsupported compiler"
+#  endif /* COMPILER_MICROSOFT */
+#endif
+}
+#endif /* INLINE_ASSEMBLY_SUPPORTED */
+
+#ifdef INLINE_ASM_SUPPORTED
+void cpu_rdtsc(uint64_t* result)
+{
+	uint32_t low_part, hi_part;
+#ifdef COMPILER_GCC
+	__asm __volatile (
+		"	rdtsc\n"
+		"	mov	%%eax,	%0\n"
+		"	mov	%%edx,	%1\n"
+		:"=m"(low_part), "=m"(hi_part)::"memory", "eax", "edx"
+	);
+#else
+#  ifdef COMPILER_MICROSOFT
+	__asm {
+		rdtsc
+		mov	low_part,	eax
+		mov	hi_part,	edx
+	};
+#  else
+#    error "Unsupported compiler"
+#  endif /* COMPILER_MICROSOFT */
+#endif /* COMPILER_GCC */
+	*result = (uint64_t)low_part + (((uint64_t) hi_part) << 32);
+}
+#endif /* INLINE_ASM_SUPPORTED */
+
+#ifdef INLINE_ASM_SUPPORTED
+void busy_sse_loop(int cycles)
+{
+#ifdef COMPILER_GCC
+#ifndef __APPLE__
+#	define XALIGN ".balign 16\n"
+#else
+#	define XALIGN ".align 4\n"
+#endif
+	__asm __volatile (
+		"	xorps	%%xmm0,	%%xmm0\n"
+		"	xorps	%%xmm1,	%%xmm1\n"
+		"	xorps	%%xmm2,	%%xmm2\n"
+		"	xorps	%%xmm3,	%%xmm3\n"
+		"	xorps	%%xmm4,	%%xmm4\n"
+		"	xorps	%%xmm5,	%%xmm5\n"
+		"	xorps	%%xmm6,	%%xmm6\n"
+		"	xorps	%%xmm7,	%%xmm7\n"
+		XALIGN
+		/* ".bsLoop:\n" */
+		"1:\n"
+		// 0:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 1:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 2:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 3:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 4:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 5:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 6:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 7:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 8:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		// 9:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//10:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//11:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//12:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//13:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//14:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//15:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//16:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//17:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//18:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//19:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//20:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//21:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//22:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//23:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//24:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//25:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//26:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//27:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//28:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//29:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//30:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		//31:
+		"	addps	%%xmm1, %%xmm0\n"
+		"	addps	%%xmm2, %%xmm1\n"
+		"	addps	%%xmm3, %%xmm2\n"
+		"	addps	%%xmm4, %%xmm3\n"
+		"	addps	%%xmm5, %%xmm4\n"
+		"	addps	%%xmm6, %%xmm5\n"
+		"	addps	%%xmm7, %%xmm6\n"
+		"	addps	%%xmm0, %%xmm7\n"
+		
+		"	dec	%%eax\n"
+		/* "jnz	.bsLoop\n" */
+		"	jnz	1b\n"
+		::"a"(cycles)
+	);
+#else
+#  ifdef COMPILER_MICROSOFT
+	__asm {
+		mov	eax,	cycles
+		xorps	xmm0,	xmm0
+		xorps	xmm1,	xmm1
+		xorps	xmm2,	xmm2
+		xorps	xmm3,	xmm3
+		xorps	xmm4,	xmm4
+		xorps	xmm5,	xmm5
+		xorps	xmm6,	xmm6
+		xorps	xmm7,	xmm7
+		//--
+		align 16
+bsLoop:
+		// 0:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 1:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 2:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 3:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 4:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 5:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 6:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 7:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 8:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 9:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 10:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 11:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 12:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 13:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 14:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 15:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 16:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 17:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 18:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 19:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 20:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 21:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 22:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 23:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 24:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 25:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 26:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 27:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 28:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 29:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 30:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		// 31:
+		addps	xmm0,	xmm1
+		addps	xmm1,	xmm2
+		addps	xmm2,	xmm3
+		addps	xmm3,	xmm4
+		addps	xmm4,	xmm5
+		addps	xmm5,	xmm6
+		addps	xmm6,	xmm7
+		addps	xmm7,	xmm0
+		//----------------------
+		dec		eax
+		jnz		bsLoop
+	}
+#  else
+#    error "Unsupported compiler"
+#  endif /* COMPILER_MICROSOFT */
+#endif /* COMPILER_GCC */
+}
+#endif /* INLINE_ASSEMBLY_SUPPORTED */

src/3rdparty/libcpuid/asm-bits.h

@@ -0,0 +1,53 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __ASM_BITS_H__
+#define __ASM_BITS_H__
+#include "libcpuid.h"
+
+/* Determine Compiler: */
+#if defined(_MSC_VER)
+#	define COMPILER_MICROSOFT
+#elif defined(__GNUC__)
+#	define COMPILER_GCC
+#endif
+
+/* Determine Platform */
+#if defined(__x86_64__) || defined(_M_AMD64)
+#	define PLATFORM_X64
+#elif defined(__i386__) || defined(_M_IX86)
+#	define PLATFORM_X86
+#endif
+
+/* Under Windows/AMD64 with MSVC, inline assembly isn't supported */
+#if (defined(COMPILER_GCC) && defined(PLATFORM_X64)) || defined(PLATFORM_X86)
+#	define INLINE_ASM_SUPPORTED
+#endif
+
+int cpuid_exists_by_eflags(void);
+void exec_cpuid(uint32_t *regs);
+void busy_sse_loop(int cycles);
+
+#endif /* __ASM_BITS_H__ */

src/3rdparty/libcpuid/cpuid_main.c

@@ -0,0 +1,389 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#include "libcpuid.h"
+#include "libcpuid_internal.h"
+#include "recog_intel.h"
+#include "recog_amd.h"
+#include "asm-bits.h"
+#include "libcpuid_util.h"
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+#include <stdio.h>
+#include <string.h>
+#include <stdlib.h>
+
+/* Implementation: */
+
+static int _libcpiud_errno = ERR_OK;
+
+int set_error(cpu_error_t err)
+{
+	_libcpiud_errno = (int) err;
+	return (int) err;
+}
+
+static void cpu_id_t_constructor(struct cpu_id_t* id)
+{
+	memset(id, 0, sizeof(struct cpu_id_t));
+	id->l1_data_cache = id->l1_instruction_cache = id->l2_cache = id->l3_cache = id->l4_cache = -1;
+	id->l1_assoc = id->l2_assoc = id->l3_assoc = id->l4_assoc = -1;
+	id->l1_cacheline = id->l2_cacheline = id->l3_cacheline = id->l4_cacheline = -1;
+	id->sse_size = -1;
+}
+
+/* get_total_cpus() system specific code: uses OS routines to determine total number of CPUs */
+#ifdef __APPLE__
+#include <unistd.h>
+#include <mach/clock_types.h>
+#include <mach/clock.h>
+#include <mach/mach.h>
+static int get_total_cpus(void)
+{
+	kern_return_t kr;
+	host_basic_info_data_t basic_info;
+	host_info_t info = (host_info_t)&basic_info;
+	host_flavor_t flavor = HOST_BASIC_INFO;
+	mach_msg_type_number_t count = HOST_BASIC_INFO_COUNT;
+	kr = host_info(mach_host_self(), flavor, info, &count);
+	if (kr != KERN_SUCCESS) return 1;
+	return basic_info.avail_cpus;
+}
+#define GET_TOTAL_CPUS_DEFINED
+#endif
+
+#ifdef _WIN32
+#include <windows.h>
+static int get_total_cpus(void)
+{
+	SYSTEM_INFO system_info;
+	GetSystemInfo(&system_info);
+	return system_info.dwNumberOfProcessors;
+}
+#define GET_TOTAL_CPUS_DEFINED
+#endif
+
+#if defined linux || defined __linux__ || defined __sun
+#include <sys/sysinfo.h>
+#include <unistd.h>
+ 
+static int get_total_cpus(void)
+{
+	return sysconf(_SC_NPROCESSORS_ONLN);
+}
+#define GET_TOTAL_CPUS_DEFINED
+#endif
+
+#if defined __FreeBSD__ || defined __OpenBSD__ || defined __NetBSD__ || defined __bsdi__ || defined __QNX__
+#include <sys/types.h>
+#include <sys/sysctl.h>
+
+static int get_total_cpus(void)
+{
+	int mib[2] = { CTL_HW, HW_NCPU };
+	int ncpus;
+	size_t len = sizeof(ncpus);
+	if (sysctl(mib, 2, &ncpus, &len, (void *) 0, 0) != 0) return 1;
+	return ncpus;
+}
+#define GET_TOTAL_CPUS_DEFINED
+#endif
+
+#ifndef GET_TOTAL_CPUS_DEFINED
+static int get_total_cpus(void)
+{
+	static int warning_printed = 0;
+	if (!warning_printed) {
+		warning_printed = 1;
+		warnf("Your system is not supported by libcpuid -- don't know how to detect the\n");
+		warnf("total number of CPUs on your system. It will be reported as 1.\n");
+		printf("Please use cpu_id_t.logical_cpus field instead.\n");
+	}
+	return 1;
+}
+#endif /* GET_TOTAL_CPUS_DEFINED */
+
+
+static void load_features_common(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	const struct feature_map_t matchtable_edx1[] = {
+		{  0, CPU_FEATURE_FPU },
+		{  1, CPU_FEATURE_VME },
+		{  2, CPU_FEATURE_DE },
+		{  3, CPU_FEATURE_PSE },
+		{  4, CPU_FEATURE_TSC },
+		{  5, CPU_FEATURE_MSR },
+		{  6, CPU_FEATURE_PAE },
+		{  7, CPU_FEATURE_MCE },
+		{  8, CPU_FEATURE_CX8 },
+		{  9, CPU_FEATURE_APIC },
+		{ 11, CPU_FEATURE_SEP },
+		{ 12, CPU_FEATURE_MTRR },
+		{ 13, CPU_FEATURE_PGE },
+		{ 14, CPU_FEATURE_MCA },
+		{ 15, CPU_FEATURE_CMOV },
+		{ 16, CPU_FEATURE_PAT },
+		{ 17, CPU_FEATURE_PSE36 },
+		{ 19, CPU_FEATURE_CLFLUSH },
+		{ 23, CPU_FEATURE_MMX },
+		{ 24, CPU_FEATURE_FXSR },
+		{ 25, CPU_FEATURE_SSE },
+		{ 26, CPU_FEATURE_SSE2 },
+		{ 28, CPU_FEATURE_HT },
+	};
+	const struct feature_map_t matchtable_ecx1[] = {
+		{  0, CPU_FEATURE_PNI },
+		{  1, CPU_FEATURE_PCLMUL },
+		{  3, CPU_FEATURE_MONITOR },
+		{  9, CPU_FEATURE_SSSE3 },
+		{ 12, CPU_FEATURE_FMA3 },
+		{ 13, CPU_FEATURE_CX16 },
+		{ 19, CPU_FEATURE_SSE4_1 },
+		{ 20, CPU_FEATURE_SSE4_2 },
+		{ 22, CPU_FEATURE_MOVBE },
+		{ 23, CPU_FEATURE_POPCNT },
+		{ 25, CPU_FEATURE_AES },
+		{ 26, CPU_FEATURE_XSAVE },
+		{ 27, CPU_FEATURE_OSXSAVE },
+		{ 28, CPU_FEATURE_AVX },
+		{ 29, CPU_FEATURE_F16C },
+		{ 30, CPU_FEATURE_RDRAND },
+	};
+	const struct feature_map_t matchtable_ebx7[] = {
+		{  3, CPU_FEATURE_BMI1 },
+		{  5, CPU_FEATURE_AVX2 },
+		{  8, CPU_FEATURE_BMI2 },
+	};
+	const struct feature_map_t matchtable_edx81[] = {
+		{ 11, CPU_FEATURE_SYSCALL },
+		{ 27, CPU_FEATURE_RDTSCP },
+		{ 29, CPU_FEATURE_LM },
+	};
+	const struct feature_map_t matchtable_ecx81[] = {
+		{  0, CPU_FEATURE_LAHF_LM },
+	};
+	const struct feature_map_t matchtable_edx87[] = {
+		{  8, CPU_FEATURE_CONSTANT_TSC },
+	};
+	if (raw->basic_cpuid[0][0] >= 1) {
+		match_features(matchtable_edx1, COUNT_OF(matchtable_edx1), raw->basic_cpuid[1][3], data);
+		match_features(matchtable_ecx1, COUNT_OF(matchtable_ecx1), raw->basic_cpuid[1][2], data);
+	}
+	if (raw->basic_cpuid[0][0] >= 7) {
+		match_features(matchtable_ebx7, COUNT_OF(matchtable_ebx7), raw->basic_cpuid[7][1], data);
+	}
+	if (raw->ext_cpuid[0][0] >= 0x80000001) {
+		match_features(matchtable_edx81, COUNT_OF(matchtable_edx81), raw->ext_cpuid[1][3], data);
+		match_features(matchtable_ecx81, COUNT_OF(matchtable_ecx81), raw->ext_cpuid[1][2], data);
+	}
+	if (raw->ext_cpuid[0][0] >= 0x80000007) {
+		match_features(matchtable_edx87, COUNT_OF(matchtable_edx87), raw->ext_cpuid[7][3], data);
+	}
+	if (data->flags[CPU_FEATURE_SSE]) {
+		/* apply guesswork to check if the SSE unit width is 128 bit */
+		switch (data->vendor) {
+			case VENDOR_AMD:
+				data->sse_size = (data->ext_family >= 16 && data->ext_family != 17) ? 128 : 64;
+				break;
+			case VENDOR_INTEL:
+				data->sse_size = (data->family == 6 && data->ext_model >= 15) ? 128 : 64;
+				break;
+			default:
+				break;
+		}
+		/* leave the CPU_FEATURE_128BIT_SSE_AUTH 0; the advanced per-vendor detection routines
+		 * will set it accordingly if they detect the needed bit */
+	}
+}
+
+static cpu_vendor_t cpuid_vendor_identify(const uint32_t *raw_vendor, char *vendor_str)
+{
+    int i;
+    cpu_vendor_t vendor = VENDOR_UNKNOWN;
+    const struct { cpu_vendor_t vendor; char match[16]; }
+    matchtable[NUM_CPU_VENDORS] = {
+        /* source: http://www.sandpile.org/ia32/cpuid.htm */
+        { VENDOR_INTEL		, "GenuineIntel" },
+        { VENDOR_AMD		, "AuthenticAMD" },
+        { VENDOR_CYRIX		, "CyrixInstead" },
+        { VENDOR_NEXGEN		, "NexGenDriven" },
+        { VENDOR_TRANSMETA	, "GenuineTMx86" },
+        { VENDOR_UMC		, "UMC UMC UMC " },
+        { VENDOR_CENTAUR	, "CentaurHauls" },
+        { VENDOR_RISE		, "RiseRiseRise" },
+        { VENDOR_SIS		, "SiS SiS SiS " },
+        { VENDOR_NSC		, "Geode by NSC" },
+    };
+
+    memcpy(vendor_str + 0, &raw_vendor[1], 4);
+    memcpy(vendor_str + 4, &raw_vendor[3], 4);
+    memcpy(vendor_str + 8, &raw_vendor[2], 4);
+    vendor_str[12] = 0;
+
+    /* Determine vendor: */
+    for (i = 0; i < NUM_CPU_VENDORS; i++)
+        if (!strcmp(vendor_str, matchtable[i].match)) {
+            vendor = matchtable[i].vendor;
+            break;
+        }
+    return vendor;
+}
+
+static int cpuid_basic_identify(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	int i, j, basic, xmodel, xfamily, ext;
+	char brandstr[64] = {0};
+    data->vendor = cpuid_vendor_identify(raw->basic_cpuid[0], data->vendor_str);
+
+	if (data->vendor == VENDOR_UNKNOWN)
+		return set_error(ERR_CPU_UNKN);
+	basic = raw->basic_cpuid[0][0];
+	if (basic >= 1) {
+		data->family = (raw->basic_cpuid[1][0] >> 8) & 0xf;
+		data->model = (raw->basic_cpuid[1][0] >> 4) & 0xf;
+		data->stepping = raw->basic_cpuid[1][0] & 0xf;
+		xmodel = (raw->basic_cpuid[1][0] >> 16) & 0xf;
+		xfamily = (raw->basic_cpuid[1][0] >> 20) & 0xff;
+		if (data->vendor == VENDOR_AMD && data->family < 0xf)
+			data->ext_family = data->family;
+		else
+			data->ext_family = data->family + xfamily;
+		data->ext_model = data->model + (xmodel << 4);
+	}
+	ext = raw->ext_cpuid[0][0] - 0x8000000;
+	
+	/* obtain the brand string, if present: */
+	if (ext >= 4) {
+		for (i = 0; i < 3; i++)
+			for (j = 0; j < 4; j++)
+				memcpy(brandstr + i * 16 + j * 4,
+				       &raw->ext_cpuid[2 + i][j], 4);
+		brandstr[48] = 0;
+		i = 0;
+		while (brandstr[i] == ' ') i++;
+		strncpy(data->brand_str, brandstr + i, sizeof(data->brand_str));
+		data->brand_str[48] = 0;
+	}
+	load_features_common(raw, data);
+	data->total_logical_cpus = get_total_cpus();
+	return set_error(ERR_OK);
+}
+
+/* Interface: */
+
+int cpuid_get_total_cpus(void)
+{
+	return get_total_cpus();
+}
+
+int cpuid_present(void)
+{
+	return cpuid_exists_by_eflags();
+}
+
+void cpu_exec_cpuid(uint32_t eax, uint32_t* regs)
+{
+	regs[0] = eax;
+	regs[1] = regs[2] = regs[3] = 0;
+	exec_cpuid(regs);
+}
+
+void cpu_exec_cpuid_ext(uint32_t* regs)
+{
+	exec_cpuid(regs);
+}
+
+int cpuid_get_raw_data(struct cpu_raw_data_t* data)
+{
+	unsigned i;
+	if (!cpuid_present())
+		return set_error(ERR_NO_CPUID);
+	for (i = 0; i < 32; i++)
+		cpu_exec_cpuid(i, data->basic_cpuid[i]);
+	for (i = 0; i < 32; i++)
+		cpu_exec_cpuid(0x80000000 + i, data->ext_cpuid[i]);
+	for (i = 0; i < MAX_INTELFN4_LEVEL; i++) {
+		memset(data->intel_fn4[i], 0, sizeof(data->intel_fn4[i]));
+		data->intel_fn4[i][0] = 4;
+		data->intel_fn4[i][2] = i;
+		cpu_exec_cpuid_ext(data->intel_fn4[i]);
+	}
+	for (i = 0; i < MAX_INTELFN11_LEVEL; i++) {
+		memset(data->intel_fn11[i], 0, sizeof(data->intel_fn11[i]));
+		data->intel_fn11[i][0] = 11;
+		data->intel_fn11[i][2] = i;
+		cpu_exec_cpuid_ext(data->intel_fn11[i]);
+	}
+	for (i = 0; i < MAX_INTELFN12H_LEVEL; i++) {
+		memset(data->intel_fn12h[i], 0, sizeof(data->intel_fn12h[i]));
+		data->intel_fn12h[i][0] = 0x12;
+		data->intel_fn12h[i][2] = i;
+		cpu_exec_cpuid_ext(data->intel_fn12h[i]);
+	}
+	for (i = 0; i < MAX_INTELFN14H_LEVEL; i++) {
+		memset(data->intel_fn14h[i], 0, sizeof(data->intel_fn14h[i]));
+		data->intel_fn14h[i][0] = 0x14;
+		data->intel_fn14h[i][2] = i;
+		cpu_exec_cpuid_ext(data->intel_fn14h[i]);
+	}
+	return set_error(ERR_OK);
+}
+
+int cpu_ident_internal(struct cpu_raw_data_t* raw, struct cpu_id_t* data, struct internal_id_info_t* internal)
+{
+	int r;
+	struct cpu_raw_data_t myraw;
+	if (!raw) {
+		if ((r = cpuid_get_raw_data(&myraw)) < 0)
+			return set_error(r);
+		raw = &myraw;
+	}
+	cpu_id_t_constructor(data);
+	if ((r = cpuid_basic_identify(raw, data)) < 0)
+		return set_error(r);
+	switch (data->vendor) {
+		case VENDOR_INTEL:
+			r = cpuid_identify_intel(raw, data, internal);
+			break;
+		case VENDOR_AMD:
+			r = cpuid_identify_amd(raw, data, internal);
+			break;
+		default:
+			break;
+	}
+	return set_error(r);
+}
+
+int cpu_identify(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	struct internal_id_info_t throwaway;
+	return cpu_ident_internal(raw, data, &throwaway);
+}
+
+const char* cpuid_lib_version(void)
+{
+	return VERSION;
+}

src/3rdparty/libcpuid/intel_code_t.h

@@ -0,0 +1,58 @@
+/*
+ * Copyright 2016  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ * This file contains a list of internal codes we use in detection. It is
+ * of no external use and isn't a complete list of intel products.
+ */
+	CODE2(PENTIUM, 2000),
+	
+	CODE(IRWIN),
+	CODE(POTOMAC),
+	CODE(GAINESTOWN),
+	CODE(WESTMERE),
+	
+	CODE(PENTIUM_M),
+	CODE(NOT_CELERON),	
+	
+	CODE(CORE_SOLO),
+	CODE(MOBILE_CORE_SOLO),
+	CODE(CORE_DUO),
+	CODE(MOBILE_CORE_DUO),
+	
+	CODE(WOLFDALE),
+	CODE(MEROM),
+	CODE(PENRYN),
+	CODE(QUAD_CORE),
+	CODE(DUAL_CORE_HT),
+	CODE(QUAD_CORE_HT),
+	CODE(MORE_THAN_QUADCORE),
+	CODE(PENTIUM_D),
+	
+	CODE(SILVERTHORNE),
+	CODE(DIAMONDVILLE),
+	CODE(PINEVIEW),
+	CODE(CEDARVIEW),

src/3rdparty/libcpuid/libcpuid.h

@@ -0,0 +1,677 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __LIBCPUID_H__
+#define __LIBCPUID_H__
+/**
+ * \file     libcpuid.h
+ * \author   Veselin Georgiev
+ * \date     Oct 2008
+ * \version  0.4.0
+ *
+ * Version history:
+ *
+ * * 0.1.0 (2008-10-15): initial adaptation from wxfractgui sources
+ * * 0.1.1 (2009-07-06): Added intel_fn11 fields to cpu_raw_data_t to handle
+ *                       new processor topology enumeration required on Core i7
+ * * 0.1.2 (2009-09-26): Added support for MSR reading through self-extracting
+ *                       kernel driver on Win32.
+ * * 0.1.3 (2010-04-20): Added support for greater more accurate CPU clock
+ *                       measurements with cpu_clock_by_ic()
+ * * 0.2.0 (2011-10-11): Support for AMD Bulldozer CPUs, 128-bit SSE unit size
+ *                       checking. A backwards-incompatible change, since the
+ *                       sizeof cpu_id_t is now different.
+ * * 0.2.1 (2012-05-26): Support for Ivy Bridge, and detecting the presence of
+ *                       the RdRand instruction.
+ * * 0.2.2 (2015-11-04): Support for newer processors up to Haswell and Vishera.
+ *                       Fix clock detection in cpu_clock_by_ic() for Bulldozer.
+ *                       More entries supported in cpu_msrinfo().
+ *                       *BSD and Solaris support (unofficial).
+ * * 0.3.0 (2016-07-09): Support for Skylake; MSR ops in FreeBSD; INFO_VOLTAGE
+ *                       for AMD CPUs. Level 4 cache support for Crystalwell
+ *                       (a backwards-incompatible change since the sizeof
+ *                        cpu_raw_data_t is now different).
+ * * 0.4.0 (2016-09-30): Better detection of AMD clock multiplier with msrinfo.
+ *                       Support for Intel SGX detection
+ *                       (a backwards-incompatible change since the sizeof
+ *                        cpu_raw_data_t and cpu_id_t is now different).
+ */
+
+/** @mainpage A simple libcpuid introduction
+ *
+ * LibCPUID provides CPU identification and access to the CPUID and RDTSC
+ * instructions on the x86.
+ * <p>
+ * To execute CPUID, use \ref cpu_exec_cpuid <br>
+ * To execute RDTSC, use \ref cpu_rdtsc <br>
+ * To fetch the CPUID info needed for CPU identification, use
+ *   \ref cpuid_get_raw_data <br>
+ * To make sense of that data (decode, extract features), use \ref cpu_identify <br>
+ * To detect the CPU speed, use either \ref cpu_clock, \ref cpu_clock_by_os,
+ * \ref cpu_tsc_mark + \ref cpu_tsc_unmark + \ref cpu_clock_by_mark,
+ * \ref cpu_clock_measure or \ref cpu_clock_by_ic.
+ * Read carefully for pros/cons of each method. <br>
+ * 
+ * To read MSRs, use \ref cpu_msr_driver_open to get a handle, and then
+ * \ref cpu_rdmsr for querying abilities. Some MSR decoding is available on recent
+ * CPUs, and can be queried through \ref cpu_msrinfo; the various types of queries
+ * are described in \ref cpu_msrinfo_request_t.
+ * </p>
+ */
+
+/** @defgroup libcpuid LibCPUID
+ @{ */
+
+/* Include some integer type specifications: */
+#include "libcpuid_types.h"
+
+/* Some limits and other constants */
+#include "libcpuid_constants.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+ * @brief CPU vendor, as guessed from the Vendor String.
+ */
+typedef enum {
+	VENDOR_INTEL = 0,  /*!< Intel CPU */
+	VENDOR_AMD,        /*!< AMD CPU */
+	VENDOR_CYRIX,      /*!< Cyrix CPU */
+	VENDOR_NEXGEN,     /*!< NexGen CPU */
+	VENDOR_TRANSMETA,  /*!< Transmeta CPU */
+	VENDOR_UMC,        /*!< x86 CPU by UMC */
+	VENDOR_CENTAUR,    /*!< x86 CPU by IDT */
+	VENDOR_RISE,       /*!< x86 CPU by Rise Technology */
+	VENDOR_SIS,        /*!< x86 CPU by SiS */
+	VENDOR_NSC,        /*!< x86 CPU by National Semiconductor */
+	
+	NUM_CPU_VENDORS,   /*!< Valid CPU vendor ids: 0..NUM_CPU_VENDORS - 1 */
+	VENDOR_UNKNOWN = -1,
+} cpu_vendor_t;
+#define NUM_CPU_VENDORS NUM_CPU_VENDORS
+
+/**
+ * @brief Contains just the raw CPUID data.
+ *
+ * This contains only the most basic CPU data, required to do identification
+ * and feature recognition. Every processor should be identifiable using this
+ * data only.
+ */
+struct cpu_raw_data_t {
+	/** contains results of CPUID for eax = 0, 1, ...*/
+	uint32_t basic_cpuid[MAX_CPUID_LEVEL][4];
+
+	/** contains results of CPUID for eax = 0x80000000, 0x80000001, ...*/
+	uint32_t ext_cpuid[MAX_EXT_CPUID_LEVEL][4];
+	
+	/** when the CPU is intel and it supports deterministic cache
+	    information: this contains the results of CPUID for eax = 4
+	    and ecx = 0, 1, ... */
+	uint32_t intel_fn4[MAX_INTELFN4_LEVEL][4];
+	
+	/** when the CPU is intel and it supports leaf 0Bh (Extended Topology
+	    enumeration leaf), this stores the result of CPUID with 
+	    eax = 11 and ecx = 0, 1, 2... */
+	uint32_t intel_fn11[MAX_INTELFN11_LEVEL][4];
+	
+	/** when the CPU is intel and supports leaf 12h (SGX enumeration leaf),
+	 *  this stores the result of CPUID with eax = 0x12 and
+	 *  ecx = 0, 1, 2... */
+	uint32_t intel_fn12h[MAX_INTELFN12H_LEVEL][4];
+
+	/** when the CPU is intel and supports leaf 14h (Intel Processor Trace
+	 *  capabilities leaf).
+	 *  this stores the result of CPUID with eax = 0x12 and
+	 *  ecx = 0, 1, 2... */
+	uint32_t intel_fn14h[MAX_INTELFN14H_LEVEL][4];
+};
+
+/**
+ * @brief This contains information about SGX features of the processor
+ * Example usage:
+ * @code
+ * ...
+ * struct cpu_raw_data_t raw;
+ * struct cpu_id_t id;
+ * 
+ * if (cpuid_get_raw_data(&raw) == 0 && cpu_identify(&raw, &id) == 0 && id.sgx.present) {
+ *   printf("SGX is present.\n");
+ *   printf("SGX1 instructions: %s.\n", id.sgx.flags[INTEL_SGX1] ? "present" : "absent");
+ *   printf("SGX2 instructions: %s.\n", id.sgx.flags[INTEL_SGX2] ? "present" : "absent");
+ *   printf("Max 32-bit enclave size: 2^%d bytes.\n", id.sgx.max_enclave_32bit);
+ *   printf("Max 64-bit enclave size: 2^%d bytes.\n", id.sgx.max_enclave_64bit);
+ *   for (int i = 0; i < id.sgx.num_epc_sections; i++) {
+ *     struct cpu_epc_t epc = cpuid_get_epc(i, NULL);
+ *     printf("EPC section #%d: address = %x, size = %d bytes.\n", epc.address, epc.size);
+ *   }
+ * } else {
+ *   printf("SGX is not present.\n");
+ * }
+ * @endcode
+ */ 
+struct cpu_sgx_t {
+	/** Whether SGX is present (boolean) */
+	uint32_t present;
+	
+	/** Max enclave size in 32-bit mode. This is a power-of-two value:
+	 *  if it is "31", then the max enclave size is 2^31 bytes (2 GiB).
+	 */
+	uint8_t max_enclave_32bit;
+	
+	/** Max enclave size in 64-bit mode. This is a power-of-two value:
+	 *  if it is "36", then the max enclave size is 2^36 bytes (64 GiB).
+	 */
+	uint8_t max_enclave_64bit;
+	
+	/**
+	 * contains SGX feature flags. See the \ref cpu_sgx_feature_t
+	 * "INTEL_SGX*" macros below.
+	 */
+	uint8_t flags[SGX_FLAGS_MAX];
+	
+	/** number of Enclave Page Cache (EPC) sections. Info for each
+	 *  section is available through the \ref cpuid_get_epc() function
+	 */
+	int num_epc_sections;
+	
+	/** bit vector of the supported extended  features that can be written
+	 *  to the MISC region of the SSA (Save State Area)
+	 */ 
+	uint32_t misc_select;
+	
+	/** a bit vector of the attributes that can be set to SECS.ATTRIBUTES
+	 *  via ECREATE. Corresponds to bits 0-63 (incl.) of SECS.ATTRIBUTES.
+	 */ 
+	uint64_t secs_attributes;
+	
+	/** a bit vector of the bits that can be set in the XSAVE feature
+	 *  request mask; Corresponds to bits 64-127 of SECS.ATTRIBUTES.
+	 */
+	uint64_t secs_xfrm;
+};
+
+/**
+ * @brief This contains the recognized CPU features/info
+ */
+struct cpu_id_t {
+	/** contains the CPU vendor string, e.g. "GenuineIntel" */
+	char vendor_str[VENDOR_STR_MAX];
+	
+	/** contains the brand string, e.g. "Intel(R) Xeon(TM) CPU 2.40GHz" */
+	char brand_str[BRAND_STR_MAX];
+	
+	/** contains the recognized CPU vendor */
+	cpu_vendor_t vendor;
+	
+	/**
+	 * contain CPU flags. Used to test for features. See
+	 * the \ref cpu_feature_t "CPU_FEATURE_*" macros below.
+	 * @see Features
+	 */
+	uint8_t flags[CPU_FLAGS_MAX];
+	
+	/** CPU family */
+	int32_t family;
+	
+	/** CPU model */
+	int32_t model;
+	
+	/** CPU stepping */
+	int32_t stepping;
+	
+	/** CPU extended family */
+	int32_t ext_family;
+	
+	/** CPU extended model */
+	int32_t ext_model;
+	
+	/** Number of CPU cores on the current processor */
+	int32_t num_cores;
+	
+	/**
+	 * Number of logical processors on the current processor.
+	 * Could be more than the number of physical cores,
+	 * e.g. when the processor has HyperThreading.
+	 */
+	int32_t num_logical_cpus;
+	
+	/**
+	 * The total number of logical processors.
+	 * The same value is availabe through \ref cpuid_get_total_cpus.
+	 *
+	 * This is num_logical_cpus * {total physical processors in the system}
+	 * (but only on a real system, under a VM this number may be lower).
+	 *
+	 * If you're writing a multithreaded program and you want to run it on
+	 * all CPUs, this is the number of threads you need.
+	 *
+	 * @note in a VM, this will exactly match the number of CPUs set in
+	 *       the VM's configuration.
+	 *
+	 */
+	int32_t total_logical_cpus;
+	
+	/**
+	 * L1 data cache size in KB. Could be zero, if the CPU lacks cache.
+	 * If the size cannot be determined, it will be -1.
+	 */
+	int32_t l1_data_cache;
+	
+	/**
+	 * L1 instruction cache size in KB. Could be zero, if the CPU lacks
+	 * cache. If the size cannot be determined, it will be -1.
+	 * @note On some Intel CPUs, whose instruction cache is in fact
+	 * a trace cache, the size will be expressed in K uOps.
+	 */
+	int32_t l1_instruction_cache;
+	
+	/**
+	 * L2 cache size in KB. Could be zero, if the CPU lacks L2 cache.
+	 * If the size of the cache could not be determined, it will be -1
+	 */
+	int32_t l2_cache;
+	
+	/** L3 cache size in KB. Zero on most systems */
+	int32_t l3_cache;
+
+	/** L4 cache size in KB. Zero on most systems */
+	int32_t l4_cache;
+	
+	/** Cache associativity for the L1 data cache. -1 if undetermined */
+	int32_t l1_assoc;
+	
+	/** Cache associativity for the L2 cache. -1 if undetermined */
+	int32_t l2_assoc;
+	
+	/** Cache associativity for the L3 cache. -1 if undetermined */
+	int32_t l3_assoc;
+
+	/** Cache associativity for the L4 cache. -1 if undetermined */
+	int32_t l4_assoc;
+	
+	/** Cache-line size for L1 data cache. -1 if undetermined */
+	int32_t l1_cacheline;
+	
+	/** Cache-line size for L2 cache. -1 if undetermined */
+	int32_t l2_cacheline;
+	
+	/** Cache-line size for L3 cache. -1 if undetermined */
+	int32_t l3_cacheline;
+	
+	/** Cache-line size for L4 cache. -1 if undetermined */
+	int32_t l4_cacheline;
+
+	/**
+	 * The brief and human-friendly CPU codename, which was recognized.<br>
+	 * Examples:
+	 * @code
+	 * +--------+--------+-------+-------+-------+---------------------------------------+-----------------------+
+	 * | Vendor | Family | Model | Step. | Cache |       Brand String                    | cpu_id_t.cpu_codename |
+	 * +--------+--------+-------+-------+-------+---------------------------------------+-----------------------+
+	 * | AMD    |      6 |     8 |     0 |   256 | (not available - will be ignored)     | "K6-2"                |
+	 * | Intel  |     15 |     2 |     5 |   512 | "Intel(R) Xeon(TM) CPU 2.40GHz"       | "Xeon (Prestonia)"    |
+	 * | Intel  |      6 |    15 |    11 |  4096 | "Intel(R) Core(TM)2 Duo CPU E6550..." | "Conroe (Core 2 Duo)" |
+	 * | AMD    |     15 |    35 |     2 |  1024 | "Dual Core AMD Opteron(tm) Proces..." | "Opteron (Dual Core)" |
+	 * +--------+--------+-------+-------+-------+---------------------------------------+-----------------------+
+	 * @endcode
+	 */
+	char cpu_codename[64];
+	
+	/** SSE execution unit size (64 or 128; -1 if N/A) */
+	int32_t sse_size;
+	
+	/**
+	 * contain miscellaneous detection information. Used to test about specifics of
+	 * certain detected features. See \ref cpu_hint_t "CPU_HINT_*" macros below.
+	 * @see Hints
+	 */
+	uint8_t detection_hints[CPU_HINTS_MAX];
+	
+	/** contains information about SGX features if the processor, if present */
+	struct cpu_sgx_t sgx;
+};
+
+/**
+ * @brief CPU feature identifiers
+ *
+ * Usage:
+ * @code
+ * ...
+ * struct cpu_raw_data_t raw;
+ * struct cpu_id_t id;
+ * if (cpuid_get_raw_data(&raw) == 0 && cpu_identify(&raw, &id) == 0) {
+ *     if (id.flags[CPU_FEATURE_SSE2]) {
+ *         // The CPU has SSE2...
+ *         ...
+ *     } else {
+ *         // no SSE2
+ *     }
+ * } else {
+ *   // processor cannot be determined.
+ * }
+ * @endcode
+ */
+typedef enum {
+	CPU_FEATURE_FPU = 0,	/*!< Floating point unit */
+	CPU_FEATURE_VME,	/*!< Virtual mode extension */
+	CPU_FEATURE_DE,		/*!< Debugging extension */
+	CPU_FEATURE_PSE,	/*!< Page size extension */
+	CPU_FEATURE_TSC,	/*!< Time-stamp counter */
+	CPU_FEATURE_MSR,	/*!< Model-specific regsisters, RDMSR/WRMSR supported */
+	CPU_FEATURE_PAE,	/*!< Physical address extension */
+	CPU_FEATURE_MCE,	/*!< Machine check exception */
+	CPU_FEATURE_CX8,	/*!< CMPXCHG8B instruction supported */
+	CPU_FEATURE_APIC,	/*!< APIC support */
+	CPU_FEATURE_MTRR,	/*!< Memory type range registers */
+	CPU_FEATURE_SEP,	/*!< SYSENTER / SYSEXIT instructions supported */
+	CPU_FEATURE_PGE,	/*!< Page global enable */
+	CPU_FEATURE_MCA,	/*!< Machine check architecture */
+	CPU_FEATURE_CMOV,	/*!< CMOVxx instructions supported */
+	CPU_FEATURE_PAT,	/*!< Page attribute table */
+	CPU_FEATURE_PSE36,	/*!< 36-bit page address extension */
+	CPU_FEATURE_PN,		/*!< Processor serial # implemented (Intel P3 only) */
+	CPU_FEATURE_CLFLUSH,	/*!< CLFLUSH instruction supported */
+	CPU_FEATURE_DTS,	/*!< Debug store supported */
+	CPU_FEATURE_ACPI,	/*!< ACPI support (power states) */
+	CPU_FEATURE_MMX,	/*!< MMX instruction set supported */
+	CPU_FEATURE_FXSR,	/*!< FXSAVE / FXRSTOR supported */
+	CPU_FEATURE_SSE,	/*!< Streaming-SIMD Extensions (SSE) supported */
+	CPU_FEATURE_SSE2,	/*!< SSE2 instructions supported */
+	CPU_FEATURE_SS,		/*!< Self-snoop */
+	CPU_FEATURE_HT,		/*!< Hyper-threading supported (but might be disabled) */
+	CPU_FEATURE_TM,		/*!< Thermal monitor */
+	CPU_FEATURE_IA64,	/*!< IA64 supported (Itanium only) */
+	CPU_FEATURE_PBE,	/*!< Pending-break enable */
+	CPU_FEATURE_PNI,	/*!< PNI (SSE3) instructions supported */
+	CPU_FEATURE_PCLMUL,	/*!< PCLMULQDQ instruction supported */
+	CPU_FEATURE_DTS64,	/*!< 64-bit Debug store supported */
+	CPU_FEATURE_MONITOR,	/*!< MONITOR / MWAIT supported */
+	CPU_FEATURE_DS_CPL,	/*!< CPL Qualified Debug Store */
+	CPU_FEATURE_VMX,	/*!< Virtualization technology supported */
+	CPU_FEATURE_SMX,	/*!< Safer mode exceptions */
+	CPU_FEATURE_EST,	/*!< Enhanced SpeedStep */
+	CPU_FEATURE_TM2,	/*!< Thermal monitor 2 */
+	CPU_FEATURE_SSSE3,	/*!< SSSE3 instructionss supported (this is different from SSE3!) */
+	CPU_FEATURE_CID,	/*!< Context ID supported */
+	CPU_FEATURE_CX16,	/*!< CMPXCHG16B instruction supported */
+	CPU_FEATURE_XTPR,	/*!< Send Task Priority Messages disable */
+	CPU_FEATURE_PDCM,	/*!< Performance capabilities MSR supported */
+	CPU_FEATURE_DCA,	/*!< Direct cache access supported */
+	CPU_FEATURE_SSE4_1,	/*!< SSE 4.1 instructions supported */
+	CPU_FEATURE_SSE4_2,	/*!< SSE 4.2 instructions supported */
+	CPU_FEATURE_SYSCALL,	/*!< SYSCALL / SYSRET instructions supported */
+	CPU_FEATURE_XD,		/*!< Execute disable bit supported */
+	CPU_FEATURE_MOVBE,	/*!< MOVBE instruction supported */
+	CPU_FEATURE_POPCNT,	/*!< POPCNT instruction supported */
+	CPU_FEATURE_AES,	/*!< AES* instructions supported */
+	CPU_FEATURE_XSAVE,	/*!< XSAVE/XRSTOR/etc instructions supported */
+	CPU_FEATURE_OSXSAVE,	/*!< non-privileged copy of OSXSAVE supported */
+	CPU_FEATURE_AVX,	/*!< Advanced vector extensions supported */
+	CPU_FEATURE_MMXEXT,	/*!< AMD MMX-extended instructions supported */
+	CPU_FEATURE_3DNOW,	/*!< AMD 3DNow! instructions supported */
+	CPU_FEATURE_3DNOWEXT,	/*!< AMD 3DNow! extended instructions supported */
+	CPU_FEATURE_NX,		/*!< No-execute bit supported */
+	CPU_FEATURE_FXSR_OPT,	/*!< FFXSR: FXSAVE and FXRSTOR optimizations */
+	CPU_FEATURE_RDTSCP,	/*!< RDTSCP instruction supported (AMD-only) */
+	CPU_FEATURE_LM,		/*!< Long mode (x86_64/EM64T) supported */
+	CPU_FEATURE_LAHF_LM,	/*!< LAHF/SAHF supported in 64-bit mode */
+	CPU_FEATURE_CMP_LEGACY,	/*!< core multi-processing legacy mode */
+	CPU_FEATURE_SVM,	/*!< AMD Secure virtual machine */
+	CPU_FEATURE_ABM,	/*!< LZCNT instruction support */
+	CPU_FEATURE_MISALIGNSSE,/*!< Misaligned SSE supported */
+	CPU_FEATURE_SSE4A,	/*!< SSE 4a from AMD */
+	CPU_FEATURE_3DNOWPREFETCH,	/*!< PREFETCH/PREFETCHW support */
+	CPU_FEATURE_OSVW,	/*!< OS Visible Workaround (AMD) */
+	CPU_FEATURE_IBS,	/*!< Instruction-based sampling */
+	CPU_FEATURE_SSE5,	/*!< SSE 5 instructions supported (deprecated, will never be 1) */
+	CPU_FEATURE_SKINIT,	/*!< SKINIT / STGI supported */
+	CPU_FEATURE_WDT,	/*!< Watchdog timer support */
+	CPU_FEATURE_TS,		/*!< Temperature sensor */
+	CPU_FEATURE_FID,	/*!< Frequency ID control */
+	CPU_FEATURE_VID,	/*!< Voltage ID control */
+	CPU_FEATURE_TTP,	/*!< THERMTRIP */
+	CPU_FEATURE_TM_AMD,	/*!< AMD-specified hardware thermal control */
+	CPU_FEATURE_STC,	/*!< Software thermal control */
+	CPU_FEATURE_100MHZSTEPS,/*!< 100 MHz multiplier control */
+	CPU_FEATURE_HWPSTATE,	/*!< Hardware P-state control */
+	CPU_FEATURE_CONSTANT_TSC,	/*!< TSC ticks at constant rate */
+	CPU_FEATURE_XOP,	/*!< The XOP instruction set (same as the old CPU_FEATURE_SSE5) */
+	CPU_FEATURE_FMA3,	/*!< The FMA3 instruction set */
+	CPU_FEATURE_FMA4,	/*!< The FMA4 instruction set */
+	CPU_FEATURE_TBM,	/*!< Trailing bit manipulation instruction support */
+	CPU_FEATURE_F16C,	/*!< 16-bit FP convert instruction support */
+	CPU_FEATURE_RDRAND,     /*!< RdRand instruction */
+	CPU_FEATURE_X2APIC,     /*!< x2APIC, APIC_BASE.EXTD, MSRs 0000_0800h...0000_0BFFh 64-bit ICR (+030h but not +031h), no DFR (+00Eh), SELF_IPI (+040h) also see standard level 0000_000Bh */
+	CPU_FEATURE_CPB,	/*!< Core performance boost */
+	CPU_FEATURE_APERFMPERF,	/*!< MPERF/APERF MSRs support */
+	CPU_FEATURE_PFI,	/*!< Processor Feedback Interface support */
+	CPU_FEATURE_PA,		/*!< Processor accumulator */
+	CPU_FEATURE_AVX2,	/*!< AVX2 instructions */
+	CPU_FEATURE_BMI1,	/*!< BMI1 instructions */
+	CPU_FEATURE_BMI2,	/*!< BMI2 instructions */
+	CPU_FEATURE_HLE,	/*!< Hardware Lock Elision prefixes */
+	CPU_FEATURE_RTM,	/*!< Restricted Transactional Memory instructions */
+	CPU_FEATURE_AVX512F,	/*!< AVX-512 Foundation */
+	CPU_FEATURE_AVX512DQ,	/*!< AVX-512 Double/Quad granular insns */
+	CPU_FEATURE_AVX512PF,	/*!< AVX-512 Prefetch */
+	CPU_FEATURE_AVX512ER,	/*!< AVX-512 Exponential/Reciprocal */
+	CPU_FEATURE_AVX512CD,	/*!< AVX-512 Conflict detection */
+	CPU_FEATURE_SHA_NI,	/*!< SHA-1/SHA-256 instructions */
+	CPU_FEATURE_AVX512BW,	/*!< AVX-512 Byte/Word granular insns */
+	CPU_FEATURE_AVX512VL,	/*!< AVX-512 128/256 vector length extensions */
+	CPU_FEATURE_SGX,	/*!< SGX extensions. Non-autoritative, check cpu_id_t::sgx::present to verify presence */
+	CPU_FEATURE_RDSEED,	/*!< RDSEED instruction */
+	CPU_FEATURE_ADX,	/*!< ADX extensions (arbitrary precision) */
+	/* termination: */
+	NUM_CPU_FEATURES,
+} cpu_feature_t;
+
+/**
+ * @brief CPU detection hints identifiers
+ *
+ * Usage: similar to the flags usage
+ */
+typedef enum {
+	CPU_HINT_SSE_SIZE_AUTH = 0,	/*!< SSE unit size is authoritative (not only a Family/Model guesswork, but based on an actual CPUID bit) */
+	/* termination */
+	NUM_CPU_HINTS,
+} cpu_hint_t;
+
+/**
+ * @brief SGX features flags
+ * \see cpu_sgx_t
+ *
+ * Usage:
+ * @code
+ * ...
+ * struct cpu_raw_data_t raw;
+ * struct cpu_id_t id;
+ * if (cpuid_get_raw_data(&raw) == 0 && cpu_identify(&raw, &id) == 0 && id.sgx.present) {
+ *     if (id.sgx.flags[INTEL_SGX1])
+ *         // The CPU has SGX1 instructions support...
+ *         ...
+ *     } else {
+ *         // no SGX
+ *     }
+ * } else {
+ *   // processor cannot be determined.
+ * }
+ * @endcode
+ */
+ 
+typedef enum {
+	INTEL_SGX1,		/*!< SGX1 instructions support */
+	INTEL_SGX2,		/*!< SGX2 instructions support */
+	
+	/* termination: */
+	NUM_SGX_FEATURES,
+} cpu_sgx_feature_t;
+
+/**
+ * @brief Describes common library error codes
+ */
+typedef enum {
+	ERR_OK       =  0,	/*!< "No error" */
+	ERR_NO_CPUID = -1,	/*!< "CPUID instruction is not supported" */
+	ERR_NO_RDTSC = -2,	/*!< "RDTSC instruction is not supported" */
+	ERR_NO_MEM   = -3,	/*!< "Memory allocation failed" */
+	ERR_OPEN     = -4,	/*!< "File open operation failed" */
+	ERR_BADFMT   = -5,	/*!< "Bad file format" */
+	ERR_NOT_IMP  = -6,	/*!< "Not implemented" */
+	ERR_CPU_UNKN = -7,	/*!< "Unsupported processor" */
+	ERR_NO_RDMSR = -8,	/*!< "RDMSR instruction is not supported" */
+	ERR_NO_DRIVER= -9,	/*!< "RDMSR driver error (generic)" */
+	ERR_NO_PERMS = -10,	/*!< "No permissions to install RDMSR driver" */
+	ERR_EXTRACT  = -11,	/*!< "Cannot extract RDMSR driver (read only media?)" */
+	ERR_HANDLE   = -12,	/*!< "Bad handle" */
+	ERR_INVMSR   = -13,	/*!< "Invalid MSR" */
+	ERR_INVCNB   = -14,	/*!< "Invalid core number" */
+	ERR_HANDLE_R = -15,	/*!< "Error on handle read" */
+	ERR_INVRANGE = -16,	/*!< "Invalid given range" */
+} cpu_error_t;
+
+/**
+ * @brief Internal structure, used in cpu_tsc_mark, cpu_tsc_unmark and
+ *        cpu_clock_by_mark
+ */
+struct cpu_mark_t {
+	uint64_t tsc;		/*!< Time-stamp from RDTSC */
+	uint64_t sys_clock;	/*!< In microsecond resolution */
+};
+
+/**
+ * @brief Returns the total number of logical CPU threads (even if CPUID is not present).
+ *
+ * Under VM, this number (and total_logical_cpus, since they are fetched with the same code)
+ * may be nonsensical, i.e. might not equal NumPhysicalCPUs*NumCoresPerCPU*HyperThreading.
+ * This is because no matter how many logical threads the host machine has, you may limit them
+ * in the VM to any number you like. **This** is the number returned by cpuid_get_total_cpus().
+ *
+ * @returns Number of logical CPU threads available. Equals the \ref cpu_id_t::total_logical_cpus.
+ */
+int cpuid_get_total_cpus(void);
+
+/**
+ * @brief Checks if the CPUID instruction is supported
+ * @retval 1 if CPUID is present
+ * @retval 0 the CPU doesn't have CPUID.
+ */
+int cpuid_present(void);
+
+/**
+ * @brief Executes the CPUID instruction
+ * @param eax - the value of the EAX register when executing CPUID
+ * @param regs - the results will be stored here. regs[0] = EAX, regs[1] = EBX, ...
+ * @note CPUID will be executed with EAX set to the given value and EBX, ECX,
+ *       EDX set to zero.
+ */
+void cpu_exec_cpuid(uint32_t eax, uint32_t* regs);
+
+/**
+ * @brief Executes the CPUID instruction with the given input registers
+ * @note This is just a bit more generic version of cpu_exec_cpuid - it allows
+ *       you to control all the registers.
+ * @param regs - Input/output. Prior to executing CPUID, EAX, EBX, ECX and
+ *               EDX will be set to regs[0], regs[1], regs[2] and regs[3].
+ *               After CPUID, this array will contain the results.
+ */
+void cpu_exec_cpuid_ext(uint32_t* regs);
+
+/**
+ * @brief Obtains the raw CPUID data from the current CPU
+ * @param data - a pointer to cpu_raw_data_t structure
+ * @returns zero if successful, and some negative number on error.
+ *          The error message can be obtained by calling \ref cpuid_error.
+ *          @see cpu_error_t
+ */
+int cpuid_get_raw_data(struct cpu_raw_data_t* data);
+
+/**
+ * @brief Identifies the CPU
+ * @param raw - Input - a pointer to the raw CPUID data, which is obtained
+ *              either by cpuid_get_raw_data or cpuid_deserialize_raw_data.
+ *              Can also be NULL, in which case the functions calls
+ *              cpuid_get_raw_data itself.
+ * @param data - Output - the decoded CPU features/info is written here.
+ * @note The function will not fail, even if some of the information
+ *       cannot be obtained. Even when the CPU is new and thus unknown to
+ *       libcpuid, some generic info, such as "AMD K9 family CPU" will be
+ *       written to data.cpu_codename, and most other things, such as the
+ *       CPU flags, cache sizes, etc. should be detected correctly anyway.
+ *       However, the function CAN fail, if the CPU is completely alien to
+ *       libcpuid.
+ * @note While cpu_identify() and cpuid_get_raw_data() are fast for most
+ *       purposes, running them several thousand times per second can hamper
+ *       performance significantly. Specifically, avoid writing "cpu feature
+ *       checker" wrapping function, which calls cpu_identify and returns the
+ *       value of some flag, if that function is going to be called frequently.
+ * @returns zero if successful, and some negative number on error.
+ *          The error message can be obtained by calling \ref cpuid_error.
+ *          @see cpu_error_t
+ */
+int cpu_identify(struct cpu_raw_data_t* raw, struct cpu_id_t* data);
+
+/**
+ * @brief The return value of cpuid_get_epc().
+ * @details
+ * Describes an EPC (Enclave Page Cache) layout (physical address and size).
+ * A CPU may have one or more EPC areas, and information about each is
+ * fetched via \ref cpuid_get_epc.
+ */ 
+struct cpu_epc_t {
+	uint64_t start_addr;
+	uint64_t length;
+};
+
+/**
+ * @brief Fetches information about an EPC (Enclave Page Cache) area.
+ * @param index - zero-based index, valid range [0..cpu_id_t.egx.num_epc_sections)
+ * @param raw   - a pointer to fetched raw CPUID data. Needed only for testing,
+ *                you can safely pass NULL here (if you pass a real structure,
+ *                it will be used for fetching the leaf 12h data if index < 2;
+ *                otherwise the real CPUID instruction will be used).
+ * @returns the requested data. If the CPU doesn't support SGX, or if
+ *          index >= cpu_id_t.egx.num_epc_sections, both fields of the returned
+ *          structure will be zeros.
+ */
+struct cpu_epc_t cpuid_get_epc(int index, const struct cpu_raw_data_t* raw);
+
+/**
+ * @brief Returns the libcpuid version
+ *
+ * @returns the string representation of the libcpuid version, like "0.1.1"
+ */
+const char* cpuid_lib_version(void);
+
+#ifdef __cplusplus
+}; /* extern "C" */
+#endif
+
+
+/** @} */
+
+#endif /* __LIBCPUID_H__ */

src/3rdparty/libcpuid/libcpuid_constants.h

@@ -0,0 +1,47 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/**
+ * @File     libcpuid_constants.h
+ * @Author   Veselin Georgiev
+ * @Brief    Some limits and constants for libcpuid
+ */
+
+#ifndef __LIBCPUID_CONSTANTS_H__
+#define __LIBCPUID_CONSTANTS_H__
+
+#define VENDOR_STR_MAX		16
+#define BRAND_STR_MAX		64
+#define CPU_FLAGS_MAX		128
+#define MAX_CPUID_LEVEL		32
+#define MAX_EXT_CPUID_LEVEL	32
+#define MAX_INTELFN4_LEVEL	8
+#define MAX_INTELFN11_LEVEL	4
+#define MAX_INTELFN12H_LEVEL	4
+#define MAX_INTELFN14H_LEVEL	4
+#define CPU_HINTS_MAX		16
+#define SGX_FLAGS_MAX		14
+
+#endif /* __LIBCPUID_CONSTANTS_H__ */

src/3rdparty/libcpuid/libcpuid_internal.h

@@ -0,0 +1,106 @@
+/*
+ * Copyright 2016  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __LIBCPUID_INTERNAL_H__
+#define __LIBCPUID_INTERNAL_H__
+/*
+ * This file contains internal undocumented declarations and function prototypes
+ * for the workings of the internal library infrastructure.
+ */
+
+enum _common_codes_t {
+	NA = 0,
+	NC, /* No code */
+};
+
+#define CODE(x) x
+#define CODE2(x, y) x = y
+enum _amd_code_t {
+	#include "amd_code_t.h"
+};
+typedef enum _amd_code_t amd_code_t;
+
+enum _intel_code_t {
+	#include "intel_code_t.h"
+};
+typedef enum _intel_code_t intel_code_t;
+#undef CODE
+#undef CODE2
+
+struct internal_id_info_t {
+	union {
+		amd_code_t   amd;
+		intel_code_t intel;
+	} code;
+	uint64_t bits;
+	int score; // detection (matchtable) score
+};
+
+#define LBIT(x) (((long long) 1) << x)
+
+enum _common_bits_t {
+	_M_                     = LBIT(  0 ),
+	MOBILE_                 = LBIT(  1 ),
+	_MP_                    = LBIT(  2 ),
+};
+
+// additional detection bits for Intel CPUs:
+enum _intel_bits_t {
+	PENTIUM_                = LBIT( 10 ),
+	CELERON_                = LBIT( 11 ),
+	CORE_                   = LBIT( 12 ),
+	_I_                     = LBIT( 13 ),
+	_3                      = LBIT( 14 ),
+	_5                      = LBIT( 15 ),
+	_7                      = LBIT( 16 ),
+	XEON_                   = LBIT( 17 ),
+	ATOM_                   = LBIT( 18 ),
+};
+typedef enum _intel_bits_t intel_bits_t;
+
+enum _amd_bits_t {
+	ATHLON_      = LBIT( 10 ),
+	_XP_         = LBIT( 11 ),
+	DURON_       = LBIT( 12 ),
+	SEMPRON_     = LBIT( 13 ),
+	OPTERON_     = LBIT( 14 ),
+	TURION_      = LBIT( 15 ),
+	_LV_         = LBIT( 16 ),
+	_64_         = LBIT( 17 ),
+	_X2          = LBIT( 18 ),
+	_X3          = LBIT( 19 ),
+	_X4          = LBIT( 20 ),
+	_X6          = LBIT( 21 ),
+	_FX          = LBIT( 22 ),
+	_APU_        = LBIT( 23 ),
+};
+typedef enum _amd_bits_t amd_bits_t;
+
+
+
+int cpu_ident_internal(struct cpu_raw_data_t* raw, struct cpu_id_t* data, 
+		       struct internal_id_info_t* internal);
+
+#endif /* __LIBCPUID_INTERNAL_H__ */

src/3rdparty/libcpuid/libcpuid_types.h

@@ -0,0 +1,37 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/**
+ * @File     libcpuid_types.h
+ * @Author   Veselin Georgiev
+ * @Brief    Type specifications for libcpuid.
+ */
+
+#ifndef __LIBCPUID_TYPES_H__
+#define __LIBCPUID_TYPES_H__
+
+#include <stdint.h>
+
+#endif /* __LIBCPUID_TYPES_H__ */

src/3rdparty/libcpuid/libcpuid_util.c

@@ -0,0 +1,93 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdarg.h>
+#include <string.h>
+#include <ctype.h>
+#include "libcpuid.h"
+#include "libcpuid_util.h"
+
+void match_features(const struct feature_map_t* matchtable, int count, uint32_t reg, struct cpu_id_t* data)
+{
+	int i;
+	for (i = 0; i < count; i++)
+		if (reg & (1u << matchtable[i].bit))
+			data->flags[matchtable[i].feature] = 1;
+}
+
+static int xmatch_entry(char c, const char* p)
+{
+	int i, j;
+	if (c == 0) return -1;
+	if (c == p[0]) return 1;
+	if (p[0] == '.') return 1;
+	if (p[0] == '#' && isdigit(c)) return 1;
+	if (p[0] == '[') {
+		j = 1;
+		while (p[j] && p[j] != ']') j++;
+		if (!p[j]) return -1;
+		for (i = 1; i < j; i++)
+			if (p[i] == c) return j + 1;
+	}
+	return -1;
+}
+
+int match_pattern(const char* s, const char* p)
+{
+	int i, j, dj, k, n, m;
+	n = (int) strlen(s);
+	m = (int) strlen(p);
+	for (i = 0; i < n; i++) {
+		if (xmatch_entry(s[i], p) != -1) {
+			j = 0;
+			k = 0;
+			while (j < m && ((dj = xmatch_entry(s[i + k], p + j)) != -1)) {
+				k++;
+				j += dj;
+			}
+			if (j == m) return i + 1;
+		}
+	}
+	return 0;
+}
+
+struct cpu_id_t* get_cached_cpuid(void)
+{
+	static int initialized = 0;
+	static struct cpu_id_t id;
+	if (initialized) return &id;
+	if (cpu_identify(NULL, &id))
+		memset(&id, 0, sizeof(id));
+	initialized = 1;
+	return &id;
+}
+
+int match_all(uint64_t bits, uint64_t mask)
+{
+	return (bits & mask) == mask;
+}

src/3rdparty/libcpuid/libcpuid_util.h

@@ -0,0 +1,78 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __LIBCPUID_UTIL_H__
+#define __LIBCPUID_UTIL_H__
+
+#define COUNT_OF(array) (sizeof(array) / sizeof(array[0]))
+
+struct feature_map_t {
+	unsigned bit;
+	cpu_feature_t feature;
+};
+ 
+void match_features(const struct feature_map_t* matchtable, int count,
+                    uint32_t reg, struct cpu_id_t* data);
+
+struct match_entry_t {
+	int family, model, stepping, ext_family, ext_model;
+	int ncores, l2cache, l3cache, brand_code;
+	uint64_t model_bits;
+	int model_code;
+	char name[32];
+};
+
+// returns the match score:
+int match_cpu_codename(const struct match_entry_t* matchtable, int count,
+                       struct cpu_id_t* data, int brand_code, uint64_t bits,
+                       int model_code);
+/*
+ * Seek for a pattern in `haystack'.
+ * Pattern may be an fixed string, or contain the special metacharacters
+ * '.' - match any single character
+ * '#' - match any digit
+ * '[<chars>] - match any of the given chars (regex-like ranges are not
+ *              supported)
+ * Return val: 0 if the pattern is not found. Nonzero if it is found (actually,
+ *             x + 1 where x is the index where the match is found).
+ */
+int match_pattern(const char* haystack, const char* pattern);
+
+/*
+ * Gets an initialized cpu_id_t. It is cached, so that internal libcpuid
+ * machinery doesn't need to issue cpu_identify more than once.
+ */
+struct cpu_id_t* get_cached_cpuid(void);
+
+
+/* returns true if all bits of mask are present in `bits'. */
+int match_all(uint64_t bits, uint64_t mask);
+
+/*
+ * Sets the current errno
+ */
+int set_error(cpu_error_t err);
+
+#endif /* __LIBCPUID_UTIL_H__ */

src/3rdparty/libcpuid/masm-x64.asm

@@ -0,0 +1,359 @@
+
+.code
+; procedure exec_cpuid
+; Signature: void exec_cpiud(uint32_t *regs)
+exec_cpuid Proc
+	push	rbx
+	push	rcx
+	push	rdx
+	push	rdi
+	
+	mov	rdi,	rcx
+	
+	mov	eax,	[rdi]
+	mov	ebx,	[rdi+4]
+	mov	ecx,	[rdi+8]
+	mov	edx,	[rdi+12]
+	
+	cpuid
+	
+	mov	[rdi],	eax
+	mov	[rdi+4],	ebx
+	mov	[rdi+8],	ecx
+	mov	[rdi+12],	edx
+	pop	rdi
+	pop	rdx
+	pop	rcx
+	pop	rbx
+	ret
+exec_cpuid endp
+
+; procedure cpu_rdtsc
+; Signature: void cpu_rdtsc(uint64_t *result)
+cpu_rdtsc Proc
+	push	rdx
+	rdtsc
+	mov	[rcx],	eax
+	mov	[rcx+4],	edx
+	pop	rdx
+	ret
+cpu_rdtsc endp
+
+; procedure busy_sse_loop
+; Signature: void busy_sse_loop(int cycles)
+busy_sse_loop Proc
+	; save xmm6 & xmm7 into the shadow area, as Visual C++ 2008
+	; expects that we don't touch them:
+	movups	[rsp + 8],	xmm6
+	movups	[rsp + 24],	xmm7
+
+	xorps	xmm0,	xmm0
+	xorps	xmm1,	xmm1
+	xorps	xmm2,	xmm2
+	xorps	xmm3,	xmm3
+	xorps	xmm4,	xmm4
+	xorps	xmm5,	xmm5
+	xorps	xmm6,	xmm6
+	xorps	xmm7,	xmm7
+	; --
+	align 16
+bsLoop:
+	; 0:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 1:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 2:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 3:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 4:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 5:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 6:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 7:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 8:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 9:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 10:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 11:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 12:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 13:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 14:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 15:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 16:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 17:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 18:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 19:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 20:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 21:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 22:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 23:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 24:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 25:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 26:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 27:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 28:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 29:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 30:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; 31:
+	addps	xmm0,	xmm1
+	addps	xmm1,	xmm2
+	addps	xmm2,	xmm3
+	addps	xmm3,	xmm4
+	addps	xmm4,	xmm5
+	addps	xmm5,	xmm6
+	addps	xmm6,	xmm7
+	addps	xmm7,	xmm0
+	; ----------------------
+	dec		ecx
+	jnz		bsLoop
+
+	; restore xmm6 & xmm7:
+	movups	xmm6,	[rsp + 8]
+	movups	xmm7,	[rsp + 24]
+	ret
+busy_sse_loop endp
+
+END

src/3rdparty/libcpuid/recog_amd.c

@@ -0,0 +1,169 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <ctype.h>
+#include "libcpuid.h"
+#include "libcpuid_util.h"
+#include "libcpuid_internal.h"
+#include "recog_amd.h"
+
+const struct amd_code_str { amd_code_t code; char *str; } amd_code_str[] = {
+	#define CODE(x) { x, #x }
+	#define CODE2(x, y) CODE(x)
+	#include "amd_code_t.h"
+	#undef CODE
+};
+
+struct amd_code_and_bits_t {
+	int code;
+	uint64_t bits;
+};
+
+enum _amd_model_codes_t {
+	// Only for Ryzen CPUs:
+	_1400,
+	_1500,
+	_1600,
+};
+
+static void load_amd_features(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	const struct feature_map_t matchtable_edx81[] = {
+		{ 20, CPU_FEATURE_NX },
+		{ 22, CPU_FEATURE_MMXEXT },
+		{ 25, CPU_FEATURE_FXSR_OPT },
+		{ 30, CPU_FEATURE_3DNOWEXT },
+		{ 31, CPU_FEATURE_3DNOW },
+	};
+	const struct feature_map_t matchtable_ecx81[] = {
+		{  1, CPU_FEATURE_CMP_LEGACY },
+		{  2, CPU_FEATURE_SVM },
+		{  5, CPU_FEATURE_ABM },
+		{  6, CPU_FEATURE_SSE4A },
+		{  7, CPU_FEATURE_MISALIGNSSE },
+		{  8, CPU_FEATURE_3DNOWPREFETCH },
+		{  9, CPU_FEATURE_OSVW },
+		{ 10, CPU_FEATURE_IBS },
+		{ 11, CPU_FEATURE_XOP },
+		{ 12, CPU_FEATURE_SKINIT },
+		{ 13, CPU_FEATURE_WDT },
+		{ 16, CPU_FEATURE_FMA4 },
+		{ 21, CPU_FEATURE_TBM },
+	};
+	const struct feature_map_t matchtable_edx87[] = {
+		{  0, CPU_FEATURE_TS },
+		{  1, CPU_FEATURE_FID },
+		{  2, CPU_FEATURE_VID },
+		{  3, CPU_FEATURE_TTP },
+		{  4, CPU_FEATURE_TM_AMD },
+		{  5, CPU_FEATURE_STC },
+		{  6, CPU_FEATURE_100MHZSTEPS },
+		{  7, CPU_FEATURE_HWPSTATE },
+		/* id 8 is handled in common */
+		{  9, CPU_FEATURE_CPB },
+		{ 10, CPU_FEATURE_APERFMPERF },
+		{ 11, CPU_FEATURE_PFI },
+		{ 12, CPU_FEATURE_PA },
+	};
+	if (raw->ext_cpuid[0][0] >= 0x80000001) {
+		match_features(matchtable_edx81, COUNT_OF(matchtable_edx81), raw->ext_cpuid[1][3], data);
+		match_features(matchtable_ecx81, COUNT_OF(matchtable_ecx81), raw->ext_cpuid[1][2], data);
+	}
+	if (raw->ext_cpuid[0][0] >= 0x80000007)
+		match_features(matchtable_edx87, COUNT_OF(matchtable_edx87), raw->ext_cpuid[7][3], data);
+	if (raw->ext_cpuid[0][0] >= 0x8000001a) {
+		/* We have the extended info about SSE unit size */
+		data->detection_hints[CPU_HINT_SSE_SIZE_AUTH] = 1;
+		data->sse_size = (raw->ext_cpuid[0x1a][0] & 1) ? 128 : 64;
+	}
+}
+
+static void decode_amd_cache_info(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	int l3_result;
+	const int assoc_table[16] = {
+		0, 1, 2, 0, 4, 0, 8, 0, 16, 0, 32, 48, 64, 96, 128, 255
+	};
+	unsigned n = raw->ext_cpuid[0][0];
+	
+	if (n >= 0x80000005) {
+		data->l1_data_cache = (raw->ext_cpuid[5][2] >> 24) & 0xff;
+		data->l1_assoc = (raw->ext_cpuid[5][2] >> 16) & 0xff;
+		data->l1_cacheline = (raw->ext_cpuid[5][2]) & 0xff;
+		data->l1_instruction_cache = (raw->ext_cpuid[5][3] >> 24) & 0xff;
+	}
+	if (n >= 0x80000006) {
+		data->l2_cache = (raw->ext_cpuid[6][2] >> 16) & 0xffff;
+		data->l2_assoc = assoc_table[(raw->ext_cpuid[6][2] >> 12) & 0xf];
+		data->l2_cacheline = (raw->ext_cpuid[6][2]) & 0xff;
+		
+		l3_result = (raw->ext_cpuid[6][3] >> 18);
+		if (l3_result > 0) {
+			l3_result = 512 * l3_result; /* AMD spec says it's a range,
+			                                but we take the lower bound */
+			data->l3_cache = l3_result;
+			data->l3_assoc = assoc_table[(raw->ext_cpuid[6][3] >> 12) & 0xf];
+			data->l3_cacheline = (raw->ext_cpuid[6][3]) & 0xff;
+		} else {
+			data->l3_cache = 0;
+		}
+	}
+}
+
+static void decode_amd_number_of_cores(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	int logical_cpus = -1, num_cores = -1;
+	
+	if (raw->basic_cpuid[0][0] >= 1) {
+		logical_cpus = (raw->basic_cpuid[1][1] >> 16) & 0xff;
+		if (raw->ext_cpuid[0][0] >= 8) {
+			num_cores = 1 + (raw->ext_cpuid[8][2] & 0xff);
+		}
+	}
+	if (data->flags[CPU_FEATURE_HT]) {
+		if (num_cores > 1) {
+			if (data->ext_family >= 23)
+				num_cores /= 2; // e.g., Ryzen 7 reports 16 "real" cores, but they are really just 8.
+			data->num_cores = num_cores;
+			data->num_logical_cpus = logical_cpus;
+		} else {
+			data->num_cores = 1;
+			data->num_logical_cpus = (logical_cpus >= 2 ? logical_cpus : 2);
+		}
+	} else {
+		data->num_cores = data->num_logical_cpus = 1;
+	}
+}
+
+int cpuid_identify_amd(struct cpu_raw_data_t* raw, struct cpu_id_t* data, struct internal_id_info_t* internal)
+{
+	load_amd_features(raw, data);
+	decode_amd_cache_info(raw, data);
+	decode_amd_number_of_cores(raw, data);
+	return 0;
+}

src/3rdparty/libcpuid/recog_amd.h

@@ -0,0 +1,31 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __RECOG_AMD_H__
+#define __RECOG_AMD_H__
+
+int cpuid_identify_amd(struct cpu_raw_data_t* raw, struct cpu_id_t* data, struct internal_id_info_t* internal);
+
+#endif /* __RECOG_AMD_H__ */

src/3rdparty/libcpuid/recog_intel.c

@@ -0,0 +1,542 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#include <string.h>
+#include <ctype.h>
+#include "libcpuid.h"
+#include "libcpuid_util.h"
+#include "libcpuid_internal.h"
+#include "recog_intel.h"
+
+const struct intel_bcode_str { intel_code_t code; char *str; } intel_bcode_str[] = {
+	#define CODE(x) { x, #x }
+	#define CODE2(x, y) CODE(x)
+	#include "intel_code_t.h"
+	#undef CODE
+};
+
+typedef struct {
+	int code;
+	uint64_t bits;
+} intel_code_and_bits_t;
+
+enum _intel_model_t {
+	UNKNOWN = -1,
+	_3000 = 100,
+	_3100,
+	_3200,
+	X3200,
+	_3300,
+	X3300,
+	_5100,
+	_5200,
+	_5300,
+	_5400,
+	_2xxx, /* Core i[357] 2xxx */
+	_3xxx, /* Core i[357] 3xxx */
+};
+typedef enum _intel_model_t intel_model_t;
+
+static void load_intel_features(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	const struct feature_map_t matchtable_edx1[] = {
+		{ 18, CPU_FEATURE_PN },
+		{ 21, CPU_FEATURE_DTS },
+		{ 22, CPU_FEATURE_ACPI },
+		{ 27, CPU_FEATURE_SS },
+		{ 29, CPU_FEATURE_TM },
+		{ 30, CPU_FEATURE_IA64 },
+		{ 31, CPU_FEATURE_PBE },
+	};
+	const struct feature_map_t matchtable_ecx1[] = {
+		{  2, CPU_FEATURE_DTS64 },
+		{  4, CPU_FEATURE_DS_CPL },
+		{  5, CPU_FEATURE_VMX },
+		{  6, CPU_FEATURE_SMX },
+		{  7, CPU_FEATURE_EST },
+		{  8, CPU_FEATURE_TM2 },
+		{ 10, CPU_FEATURE_CID },
+		{ 14, CPU_FEATURE_XTPR },
+		{ 15, CPU_FEATURE_PDCM },
+		{ 18, CPU_FEATURE_DCA },
+		{ 21, CPU_FEATURE_X2APIC },
+	};
+	const struct feature_map_t matchtable_edx81[] = {
+		{ 20, CPU_FEATURE_XD },
+	};
+	const struct feature_map_t matchtable_ebx7[] = {
+		{  2, CPU_FEATURE_SGX },
+		{  4, CPU_FEATURE_HLE },
+		{ 11, CPU_FEATURE_RTM },
+		{ 16, CPU_FEATURE_AVX512F },
+		{ 17, CPU_FEATURE_AVX512DQ },
+		{ 18, CPU_FEATURE_RDSEED },
+		{ 19, CPU_FEATURE_ADX },
+		{ 26, CPU_FEATURE_AVX512PF },
+		{ 27, CPU_FEATURE_AVX512ER },
+		{ 28, CPU_FEATURE_AVX512CD },
+		{ 29, CPU_FEATURE_SHA_NI },
+		{ 30, CPU_FEATURE_AVX512BW },
+		{ 31, CPU_FEATURE_AVX512VL },
+	};
+	if (raw->basic_cpuid[0][0] >= 1) {
+		match_features(matchtable_edx1, COUNT_OF(matchtable_edx1), raw->basic_cpuid[1][3], data);
+		match_features(matchtable_ecx1, COUNT_OF(matchtable_ecx1), raw->basic_cpuid[1][2], data);
+	}
+	if (raw->ext_cpuid[0][0] >= 1) {
+		match_features(matchtable_edx81, COUNT_OF(matchtable_edx81), raw->ext_cpuid[1][3], data);
+	}
+	// detect TSX/AVX512:
+	if (raw->basic_cpuid[0][0] >= 7) {
+		match_features(matchtable_ebx7, COUNT_OF(matchtable_ebx7), raw->basic_cpuid[7][1], data);
+	}
+}
+
+enum _cache_type_t {
+	L1I,
+	L1D,
+	L2,
+	L3,
+	L4
+};
+typedef enum _cache_type_t cache_type_t;
+
+static void check_case(uint8_t on, cache_type_t cache, int size, int assoc, int linesize, struct cpu_id_t* data)
+{
+	if (!on) return;
+	switch (cache) {
+		case L1I:
+			data->l1_instruction_cache = size;
+			break;
+		case L1D:
+			data->l1_data_cache = size;
+			data->l1_assoc = assoc;
+			data->l1_cacheline = linesize;
+			break;
+		case L2:
+			data->l2_cache = size;
+			data->l2_assoc = assoc;
+			data->l2_cacheline = linesize;
+			break;
+		case L3:
+			data->l3_cache = size;
+			data->l3_assoc = assoc;
+			data->l3_cacheline = linesize;
+			break;
+		case L4:
+			data->l4_cache = size;
+			data->l4_assoc = assoc;
+			data->l4_cacheline = linesize;
+			break;
+		default:
+			break;
+	}
+}
+
+static void decode_intel_oldstyle_cache_info(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	uint8_t f[256] = {0};
+	int reg, off;
+	uint32_t x;
+	for (reg = 0; reg < 4; reg++) {
+		x = raw->basic_cpuid[2][reg];
+		if (x & 0x80000000) continue;
+		for (off = 0; off < 4; off++) {
+			f[x & 0xff] = 1;
+			x >>= 8;
+		}
+	}
+	
+	check_case(f[0x06], L1I,      8,  4,  32, data);
+	check_case(f[0x08], L1I,     16,  4,  32, data);
+	check_case(f[0x0A], L1D,      8,  2,  32, data);
+	check_case(f[0x0C], L1D,     16,  4,  32, data);
+	check_case(f[0x22],  L3,    512,  4,  64, data);
+	check_case(f[0x23],  L3,   1024,  8,  64, data);
+	check_case(f[0x25],  L3,   2048,  8,  64, data);
+	check_case(f[0x29],  L3,   4096,  8,  64, data);
+	check_case(f[0x2C], L1D,     32,  8,  64, data);
+	check_case(f[0x30], L1I,     32,  8,  64, data);
+	check_case(f[0x39],  L2,    128,  4,  64, data);
+	check_case(f[0x3A],  L2,    192,  6,  64, data);
+	check_case(f[0x3B],  L2,    128,  2,  64, data);
+	check_case(f[0x3C],  L2,    256,  4,  64, data);
+	check_case(f[0x3D],  L2,    384,  6,  64, data);
+	check_case(f[0x3E],  L2,    512,  4,  64, data);
+	check_case(f[0x41],  L2,    128,  4,  32, data);
+	check_case(f[0x42],  L2,    256,  4,  32, data);
+	check_case(f[0x43],  L2,    512,  4,  32, data);
+	check_case(f[0x44],  L2,   1024,  4,  32, data);
+	check_case(f[0x45],  L2,   2048,  4,  32, data);
+	check_case(f[0x46],  L3,   4096,  4,  64, data);
+	check_case(f[0x47],  L3,   8192,  8,  64, data);
+	check_case(f[0x4A],  L3,   6144, 12,  64, data);
+	check_case(f[0x4B],  L3,   8192, 16,  64, data);
+	check_case(f[0x4C],  L3,  12288, 12,  64, data);
+	check_case(f[0x4D],  L3,  16384, 16,  64, data);
+	check_case(f[0x4E],  L2,   6144, 24,  64, data);
+	check_case(f[0x60], L1D,     16,  8,  64, data);
+	check_case(f[0x66], L1D,      8,  4,  64, data);
+	check_case(f[0x67], L1D,     16,  4,  64, data);
+	check_case(f[0x68], L1D,     32,  4,  64, data);
+	/* The following four entries are trace cache. Intel does not
+	 * specify a cache-line size, so we use -1 instead
+	 */
+	check_case(f[0x70], L1I,     12,  8,  -1, data);
+	check_case(f[0x71], L1I,     16,  8,  -1, data);
+	check_case(f[0x72], L1I,     32,  8,  -1, data);
+	check_case(f[0x73], L1I,     64,  8,  -1, data);
+	
+	check_case(f[0x78],  L2,   1024,  4,  64, data);
+	check_case(f[0x79],  L2,    128,  8,  64, data);
+	check_case(f[0x7A],  L2,    256,  8,  64, data);
+	check_case(f[0x7B],  L2,    512,  8,  64, data);
+	check_case(f[0x7C],  L2,   1024,  8,  64, data);
+	check_case(f[0x7D],  L2,   2048,  8,  64, data);
+	check_case(f[0x7F],  L2,    512,  2,  64, data);
+	check_case(f[0x82],  L2,    256,  8,  32, data);
+	check_case(f[0x83],  L2,    512,  8,  32, data);
+	check_case(f[0x84],  L2,   1024,  8,  32, data);
+	check_case(f[0x85],  L2,   2048,  8,  32, data);
+	check_case(f[0x86],  L2,    512,  4,  64, data);
+	check_case(f[0x87],  L2,   1024,  8,  64, data);
+	
+	if (f[0x49]) {
+		/* This flag is overloaded with two meanings. On Xeon MP
+		 * (family 0xf, model 0x6) this means L3 cache. On all other
+		 * CPUs (notably Conroe et al), this is L2 cache. In both cases
+		 * it means 4MB, 16-way associative, 64-byte line size.
+		 */
+		if (data->family == 0xf && data->model == 0x6) {
+			data->l3_cache = 4096;
+			data->l3_assoc = 16;
+			data->l3_cacheline = 64;
+		} else {
+			data->l2_cache = 4096;
+			data->l2_assoc = 16;
+			data->l2_cacheline = 64;
+		}
+	}
+	if (f[0x40]) {
+		/* Again, a special flag. It means:
+		 * 1) If no L2 is specified, then CPU is w/o L2 (0 KB)
+		 * 2) If L2 is specified by other flags, then, CPU is w/o L3.
+		 */
+		if (data->l2_cache == -1) {
+			data->l2_cache = 0;
+		} else {
+			data->l3_cache = 0;
+		}
+	}
+}
+
+static void decode_intel_deterministic_cache_info(struct cpu_raw_data_t* raw,
+                                                  struct cpu_id_t* data)
+{
+	int ecx;
+	int ways, partitions, linesize, sets, size, level, typenumber;
+	cache_type_t type;
+	for (ecx = 0; ecx < MAX_INTELFN4_LEVEL; ecx++) {
+		typenumber = raw->intel_fn4[ecx][0] & 0x1f;
+		if (typenumber == 0) break;
+		level = (raw->intel_fn4[ecx][0] >> 5) & 0x7;
+		if (level == 1 && typenumber == 1)
+			type = L1D;
+		else if (level == 1 && typenumber == 2)
+			type = L1I;
+		else if (level == 2 && typenumber == 3)
+			type = L2;
+		else if (level == 3 && typenumber == 3)
+			type = L3;
+		else if (level == 4 && typenumber == 3)
+			type = L4;
+		else {
+			continue;
+		}
+		ways = ((raw->intel_fn4[ecx][1] >> 22) & 0x3ff) + 1;
+		partitions = ((raw->intel_fn4[ecx][1] >> 12) & 0x3ff) + 1;
+		linesize = (raw->intel_fn4[ecx][1] & 0xfff) + 1;
+		sets = raw->intel_fn4[ecx][2] + 1;
+		size = ways * partitions * linesize * sets / 1024;
+		check_case(1, type, size, ways, linesize, data);
+	}
+}
+
+static int decode_intel_extended_topology(struct cpu_raw_data_t* raw,
+                                           struct cpu_id_t* data)
+{
+	int i, level_type, num_smt = -1, num_core = -1;
+	for (i = 0; i < MAX_INTELFN11_LEVEL; i++) {
+		level_type = (raw->intel_fn11[i][2] & 0xff00) >> 8;
+		switch (level_type) {
+			case 0x01:
+				num_smt = raw->intel_fn11[i][1] & 0xffff;
+				break;
+			case 0x02:
+				num_core = raw->intel_fn11[i][1] & 0xffff;
+				break;
+			default:
+				break;
+		}
+	}
+	if (num_smt == -1 || num_core == -1) return 0;
+	data->num_logical_cpus = num_core;
+	data->num_cores = num_core / num_smt;
+	// make sure num_cores is at least 1. In VMs, the CPUID instruction
+	// is rigged and may give nonsensical results, but we should at least
+	// avoid outputs like data->num_cores == 0.
+	if (data->num_cores <= 0) data->num_cores = 1;
+	return 1;
+}
+
+static void decode_intel_number_of_cores(struct cpu_raw_data_t* raw,
+                                         struct cpu_id_t* data)
+{
+	int logical_cpus = -1, num_cores = -1;
+	
+	if (raw->basic_cpuid[0][0] >= 11) {
+		if (decode_intel_extended_topology(raw, data)) return;
+	}
+	
+	if (raw->basic_cpuid[0][0] >= 1) {
+		logical_cpus = (raw->basic_cpuid[1][1] >> 16) & 0xff;
+		if (raw->basic_cpuid[0][0] >= 4) {
+			num_cores = 1 + ((raw->basic_cpuid[4][0] >> 26) & 0x3f);
+		}
+	}
+	if (data->flags[CPU_FEATURE_HT]) {
+		if (num_cores > 1) {
+			data->num_cores = num_cores;
+			data->num_logical_cpus = logical_cpus;
+		} else {
+			data->num_cores = 1;
+			data->num_logical_cpus = (logical_cpus >= 1 ? logical_cpus : 1);
+			if (data->num_logical_cpus == 1)
+				data->flags[CPU_FEATURE_HT] = 0;
+		}
+	} else {
+		data->num_cores = data->num_logical_cpus = 1;
+	}
+}
+
+static intel_code_and_bits_t get_brand_code_and_bits(struct cpu_id_t* data)
+{
+	intel_code_t code = (intel_code_t) NC;
+	intel_code_and_bits_t result;
+	uint64_t bits = 0;
+	int i = 0;
+	const char* bs = data->brand_str;
+	const char* s;
+	const struct { intel_code_t c; const char *search; } matchtable[] = {
+		{ PENTIUM_M, "Pentium(R) M" },
+		{ CORE_SOLO, "Pentium(R) Dual  CPU" },
+		{ CORE_SOLO, "Pentium(R) Dual-Core" },
+		{ PENTIUM_D, "Pentium(R) D" },
+		{ CORE_SOLO, "Genuine Intel(R) CPU" },
+		{ CORE_SOLO, "Intel(R) Core(TM)" },
+		{ DIAMONDVILLE, "CPU [N ][23]## " },
+		{ SILVERTHORNE, "CPU Z" },
+		{ PINEVIEW, "CPU [ND][45]## " },
+		{ CEDARVIEW, "CPU [ND]#### " },
+	};
+	
+	const struct { uint64_t bit; const char* search; } bit_matchtable[] = {
+		{ XEON_, "Xeon" },
+		{ _MP_, " MP" },
+		{ ATOM_, "Atom(TM) CPU" },
+		{ MOBILE_, "Mobile" },
+		{ CELERON_, "Celeron" },
+		{ PENTIUM_, "Pentium" },
+	};
+	
+	for (i = 0; i < COUNT_OF(bit_matchtable); i++) {
+		if (match_pattern(bs, bit_matchtable[i].search))
+			bits |= bit_matchtable[i].bit;
+	}
+	
+	if ((i = match_pattern(bs, "Core(TM) [im][357]")) != 0) {
+		bits |= CORE_;
+		i--;
+		switch (bs[i + 9]) {
+			case 'i': bits |= _I_; break;
+			case 'm': bits |= _M_; break;
+		}
+		switch (bs[i + 10]) {
+			case '3': bits |= _3; break;
+			case '5': bits |= _5; break;
+			case '7': bits |= _7; break;
+		}
+	}
+	for (i = 0; i < COUNT_OF(matchtable); i++)
+		if (match_pattern(bs, matchtable[i].search)) {
+			code = matchtable[i].c;
+			break;
+		}
+	if (bits & XEON_) {
+		if (match_pattern(bs, "W35##") || match_pattern(bs, "[ELXW]75##"))
+			bits |= _7;
+		else if (match_pattern(bs, "[ELXW]55##"))
+			code = GAINESTOWN;
+		else if (match_pattern(bs, "[ELXW]56##"))
+			code = WESTMERE;
+		else if (data->l3_cache > 0 && data->family == 16)
+			/* restrict by family, since later Xeons also have L3 ... */
+			code = IRWIN;
+	}
+	if (match_all(bits, XEON_ + _MP_) && data->l3_cache > 0)
+		code = POTOMAC;
+	if (code == CORE_SOLO) {
+		s = strstr(bs, "CPU");
+		if (s) {
+			s += 3;
+			while (*s == ' ') s++;
+			if (*s == 'T')
+				bits |= MOBILE_;
+		}
+	}
+	if (code == CORE_SOLO) {
+		switch (data->num_cores) {
+			case 1: break;
+			case 2:
+			{
+				code = CORE_DUO;
+				if (data->num_logical_cpus > 2)
+					code = DUAL_CORE_HT;
+				break;
+			}
+			case 4:
+			{
+				code = QUAD_CORE;
+				if (data->num_logical_cpus > 4)
+					code = QUAD_CORE_HT;
+				break;
+			}
+			default:
+				code = MORE_THAN_QUADCORE; break;
+		}
+	}
+	
+	if (code == CORE_DUO && (bits & MOBILE_) && data->model != 14) {
+		if (data->ext_model < 23) {
+			code = MEROM;
+		} else {
+			code = PENRYN;
+		}
+	}
+	if (data->ext_model == 23 &&
+		(code == CORE_DUO || code == PENTIUM_D || (bits & CELERON_))) {
+		code = WOLFDALE;
+	}
+
+	result.code = code;
+	result.bits = bits;
+	return result;
+}
+
+static void decode_intel_sgx_features(const struct cpu_raw_data_t* raw, struct cpu_id_t* data)
+{
+	struct cpu_epc_t epc;
+	int i;
+	
+	if (raw->basic_cpuid[0][0] < 0x12) return; // no 12h leaf
+	if (raw->basic_cpuid[0x12][0] == 0) return; // no sub-leafs available, probably it's disabled by BIOS
+	
+	// decode sub-leaf 0:
+	if (raw->basic_cpuid[0x12][0] & 1) data->sgx.flags[INTEL_SGX1] = 1;
+	if (raw->basic_cpuid[0x12][0] & 2) data->sgx.flags[INTEL_SGX2] = 1;
+	if (data->sgx.flags[INTEL_SGX1] || data->sgx.flags[INTEL_SGX2])
+		data->sgx.present = 1;
+	data->sgx.misc_select = raw->basic_cpuid[0x12][1];
+	data->sgx.max_enclave_32bit = (raw->basic_cpuid[0x12][3]     ) & 0xff;
+	data->sgx.max_enclave_64bit = (raw->basic_cpuid[0x12][3] >> 8) & 0xff;
+	
+	// decode sub-leaf 1:
+	data->sgx.secs_attributes = raw->intel_fn12h[1][0] | (((uint64_t) raw->intel_fn12h[1][1]) << 32);
+	data->sgx.secs_xfrm       = raw->intel_fn12h[1][2] | (((uint64_t) raw->intel_fn12h[1][3]) << 32);
+	
+	// decode higher-order subleafs, whenever present:
+	data->sgx.num_epc_sections = -1;
+	for (i = 0; i < 1000000; i++) {
+		epc = cpuid_get_epc(i, raw);
+		if (epc.length == 0) {
+			data->sgx.num_epc_sections = i;
+			break;
+		}
+	}
+	if (data->sgx.num_epc_sections == -1) {
+		data->sgx.num_epc_sections = 1000000;
+	}
+}
+
+struct cpu_epc_t cpuid_get_epc(int index, const struct cpu_raw_data_t* raw)
+{
+	uint32_t regs[4];
+	struct cpu_epc_t retval = {0, 0};
+	if (raw && index < MAX_INTELFN12H_LEVEL - 2) {
+		// this was queried already, use the data:
+		memcpy(regs, raw->intel_fn12h[2 + index], sizeof(regs));
+	} else {
+		// query this ourselves:
+		regs[0] = 0x12;
+		regs[2] = 2 + index;
+		regs[1] = regs[3] = 0;
+		cpu_exec_cpuid_ext(regs);
+	}
+	
+	// decode values:
+	if ((regs[0] & 0xf) == 0x1) {
+		retval.start_addr |= (regs[0] & 0xfffff000); // bits [12, 32) -> bits [12, 32)
+		retval.start_addr |= ((uint64_t) (regs[1] & 0x000fffff)) << 32; // bits [0, 20) -> bits [32, 52)
+		retval.length     |= (regs[2] & 0xfffff000); // bits [12, 32) -> bits [12, 32)
+		retval.length     |= ((uint64_t) (regs[3] & 0x000fffff)) << 32; // bits [0, 20) -> bits [32, 52)
+	}
+	return retval;
+}
+
+int cpuid_identify_intel(struct cpu_raw_data_t* raw, struct cpu_id_t* data, struct internal_id_info_t* internal)
+{
+	intel_code_and_bits_t brand;
+
+	load_intel_features(raw, data);
+	if (raw->basic_cpuid[0][0] >= 4) {
+		/* Deterministic way is preferred, being more generic */
+		decode_intel_deterministic_cache_info(raw, data);
+	} else if (raw->basic_cpuid[0][0] >= 2) {
+		decode_intel_oldstyle_cache_info(raw, data);
+	}
+	decode_intel_number_of_cores(raw, data);
+
+	brand = get_brand_code_and_bits(data);
+	
+	internal->code.intel = brand.code;
+	internal->bits = brand.bits;
+	
+	if (data->flags[CPU_FEATURE_SGX]) {
+		// if SGX is indicated by the CPU, verify its presence:
+		decode_intel_sgx_features(raw, data);
+	}
+
+	return 0;
+}

src/3rdparty/libcpuid/recog_intel.h

@@ -0,0 +1,31 @@
+/*
+ * Copyright 2008  Veselin Georgiev,
+ * anrieffNOSPAM @ mgail_DOT.com (convert to gmail)
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 THE AUTHOR 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
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __RECOG_INTEL_H__
+#define __RECOG_INTEL_H__
+
+int cpuid_identify_intel(struct cpu_raw_data_t* raw, struct cpu_id_t* data, struct internal_id_info_t* internal);
+
+#endif /*__RECOG_INTEL_H__*/

src/3rdparty/winansi.cpp

@@ -11,7 +11,7 @@
 #include <stdio.h>
 #include <io.h>
 
-#include "compat/winansi.h"
+#include "winansi.h"
 /*
 * Copyright 2008 Peter Harris <git@peter.is-a-geek.org>
 */
@@ -344,8 +344,8 @@ int winansi_vfprintf(FILE *stream, const char *format, va_list list)
 #endif
     va_end(cp);
 
-    if (len > sizeof(small_buf) - 1) {
-        buf = malloc(len + 1);
+    if ((unsigned) len > sizeof(small_buf) - 1) {
+        buf = (char*)malloc(len + 1);
         if (!buf)
             goto abort;
 
@@ -389,4 +389,4 @@ int winansi_printf(const char *format, ...)
     va_end(list);
 
     return rv;
-}
+}

src/3rdparty/winansi.h

@@ -1,12 +1,11 @@
 /*
  * ANSI emulation wrappers
  */
-#ifdef WIN32
+
 #include <windows.h>
 #include <stddef.h>
 #include <stdio.h>
 
-#define isatty(fd) _isatty(fd)
 #define fileno(fd) _fileno(fd)
 
 #ifdef __cplusplus
@@ -28,5 +27,3 @@ extern "C" {
 #define printf winansi_printf
 #define fprintf winansi_fprintf
 #define vfprintf winansi_vfprintf
-
-#endif

src/App.cpp

@@ -0,0 +1,155 @@
+/* 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 2016-2017 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 <stdlib.h>
+#include <uv.h>
+
+
+#include "App.h"
+#include "Cpu.h"
+#include "crypto/CryptoNight.h"
+#include "log/ConsoleLog.h"
+#include "log/FileLog.h"
+#include "log/Log.h"
+#include "Mem.h"
+#include "net/Network.h"
+#include "Options.h"
+#include "Summary.h"
+#include "version.h"
+#include "workers/Workers.h"
+
+
+#ifdef HAVE_SYSLOG_H
+#   include "log/SysLog.h"
+#endif
+
+
+App *App::m_self = nullptr;
+
+
+
+App::App(int argc, char **argv) :
+    m_network(nullptr),
+    m_options(nullptr)
+{
+    m_self = this;
+
+    Cpu::init();
+    m_options = Options::parse(argc, argv);
+
+    Log::init();
+
+    if (!m_options->background()) {
+        Log::add(new ConsoleLog(m_options->colors()));
+    }
+
+    if (m_options->logFile()) {
+        Log::add(new FileLog(m_options->logFile()));
+    }
+
+#   ifdef HAVE_SYSLOG_H
+    if (m_options->syslog()) {
+        Log::add(new SysLog());
+    }
+#   endif
+
+    m_network = new Network(m_options);
+
+    uv_signal_init(uv_default_loop(), &m_signal);
+}
+
+
+App::~App()
+{
+}
+
+
+int App::exec()
+{
+    if (!m_options->isReady()) {
+        return 0;
+    }
+
+    uv_signal_start(&m_signal, App::onSignal, SIGHUP);
+    uv_signal_start(&m_signal, App::onSignal, SIGTERM);
+    uv_signal_start(&m_signal, App::onSignal, SIGINT);
+
+    background();
+
+    if (!CryptoNight::init(m_options->algo(), m_options->algoVariant())) {
+        LOG_ERR("\"%s\" hash self-test failed.", m_options->algoName());
+        return 1;
+    }
+
+    Mem::allocate(m_options->algo(), m_options->threads(), m_options->doubleHash());
+    Summary::print();
+
+    Workers::start(m_options->affinity());
+
+    m_network->connect();
+
+    const int r = uv_run(uv_default_loop(), UV_RUN_DEFAULT);
+    uv_loop_close(uv_default_loop());
+
+    free(m_network);
+    free(m_options);
+
+    Mem::release();
+
+    return r;
+}
+
+
+void App::close()
+{
+    m_network->stop();
+    Workers::stop();
+
+    uv_stop(uv_default_loop());
+}
+
+
+void App::onSignal(uv_signal_t *handle, int signum)
+{
+    switch (signum)
+    {
+    case SIGHUP:
+        LOG_WARN("SIGHUP received, exiting");
+        break;
+
+    case SIGTERM:
+        LOG_WARN("SIGTERM received, exiting");
+        break;
+
+    case SIGINT:
+        LOG_WARN("SIGINT received, exiting");
+        break;
+
+    default:
+        break;
+    }
+
+    uv_signal_stop(handle);
+    m_self->close();
+}

src/App.h

@@ -0,0 +1,57 @@
+/* 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 2016-2017 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 __APP_H__
+#define __APP_H__
+
+
+#include <uv.h>
+
+
+class Network;
+class Options;
+
+
+class App
+{
+public:
+  App(int argc, char **argv);
+  ~App();
+
+  int exec();
+
+private:
+  void background();
+  void close();
+
+  static void onSignal(uv_signal_t *handle, int signum);
+
+  static App *m_self;
+
+  Network *m_network;
+  Options *m_options;
+  uv_signal_t m_signal;
+};
+
+
+#endif /* __APP_H__ */

src/App_unix.cpp

@@ -21,27 +21,46 @@
  *   along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <string.h>
 
-#include "persistent_memory.h"
-#include "algo/cryptonight/cryptonight.h"
-
-static size_t offset = 0;
+#include <stdlib.h>
+#include <signal.h>
+#include <errno.h>
+#include <unistd.h>
 
 
-void * persistent_calloc(size_t num, size_t size) {
-    void *mem = &persistent_memory[offset];
-    offset += (num * size);
+#include "App.h"
+#include "Cpu.h"
+#include "log/Log.h"
+#include "Options.h"
 
-    memset(mem, 0, num * size);
 
-    return mem;
-}
-
-
-void * create_persistent_ctx(int thr_id) {
-    struct cryptonight_ctx *ctx = (struct cryptonight_ctx *) &persistent_memory[MEMORY - sizeof(struct cryptonight_ctx) * (thr_id + 1)];
-    ctx->memory = &persistent_memory[MEMORY * (thr_id + 1)];
-
-    return ctx;
+void App::background()
+{
+    if (m_options->affinity() != -1L) {
+        Cpu::setAffinity(-1, m_options->affinity());
+    }
+
+    if (!m_options->background()) {
+        return;
+    }
+
+    int i = fork();
+    if (i < 0) {
+        exit(1);
+    }
+
+    if (i > 0) {
+        exit(0);
+    }
+
+    i = setsid();
+
+    if (i < 0) {
+        LOG_ERR("setsid() failed (errno = %d)", errno);
+    }
+
+    i = chdir("/");
+    if (i < 0) {
+        LOG_ERR("chdir() failed (errno = %d)", errno);
+    }
 }

src/App_win.cpp

@@ -21,23 +21,32 @@
  *   along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#ifndef __UTIL_H__
-#define __UTIL_H__
 
-#include <jansson.h>
+#include <winsock2.h>
+#include <windows.h>
 
 
-json_t *json_decode(const char *s);
-
-char *bin2hex(const unsigned char *p, size_t len);
-bool hex2bin(unsigned char *p, const char *hexstr, size_t len);
-
-struct thread_q *tq_new(void);
-void tq_free(struct thread_q *tq);
-bool tq_push(struct thread_q *tq, void *data);
-void *tq_pop(struct thread_q *tq, const struct timespec *abstime);
-void tq_freeze(struct thread_q *tq);
-void tq_thaw(struct thread_q *tq);
+#include "App.h"
+#include "Options.h"
+#include "Cpu.h"
 
 
-#endif /* __UTIL_H__ */
+void App::background()
+{
+    if (m_options->affinity() != -1L) {
+        Cpu::setAffinity(-1, m_options->affinity());
+    }
+
+    if (!m_options->background()) {
+        return;
+    }
+
+    HWND hcon = GetConsoleWindow();
+    if (hcon) {
+        ShowWindow(hcon, SW_HIDE);
+    } else {
+        HANDLE h = GetStdHandle(STD_OUTPUT_HANDLE);
+        CloseHandle(h);
+        FreeConsole();
+    }
+}

src/Cpu.cpp

@@ -0,0 +1,105 @@
+/* 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 2016-2017 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 <libcpuid.h>
+#include <math.h>
+#include <string.h>
+
+#include "Cpu.h"
+
+
+char Cpu::m_brand[64]   = { 0 };
+int Cpu::m_flags        = 0;
+int Cpu::m_l2_cache     = 0;
+int Cpu::m_l3_cache     = 0;
+int Cpu::m_sockets      = 1;
+int Cpu::m_totalCores   = 0;
+int Cpu::m_totalThreads = 0;
+
+
+int Cpu::optimalThreadsCount(int algo, bool doubleHash, int maxCpuUsage)
+{
+    if (m_totalThreads == 1) {
+        return 1;
+    }
+
+    int cache = m_l3_cache ? m_l3_cache : m_l2_cache;
+    int count = 0;
+    const int size = (algo ? 1024 : 2048) * (doubleHash ? 2 : 1);
+
+    if (cache) {
+        count = cache / size;
+    }
+    else {
+        count = m_totalThreads / 2;
+    }
+
+    if (count > m_totalThreads) {
+        count = m_totalThreads;
+    }
+
+    if (((float) count / m_totalThreads * 100) > maxCpuUsage) {
+        count = ceil((float) m_totalThreads * (maxCpuUsage / 100.0));
+    }
+
+    return count < 1 ? 1 : count;
+}
+
+
+void Cpu::initCommon()
+{
+    struct cpu_raw_data_t raw = { 0 };
+    struct cpu_id_t data = { 0 };
+
+    cpuid_get_raw_data(&raw);
+    cpu_identify(&raw, &data);
+
+    strncpy(m_brand, data.brand_str, sizeof(m_brand) - 1);
+
+    m_totalThreads = data.total_logical_cpus;
+    m_sockets = m_totalThreads / data.num_logical_cpus;
+    m_totalCores = data.num_cores *m_sockets;
+
+    m_l3_cache = data.l3_cache > 0 ? data.l3_cache * m_sockets : 0;
+
+    // Workaround for AMD CPUs https://github.com/anrieff/libcpuid/issues/97
+    if (data.vendor == VENDOR_AMD && data.l3_cache <= 0 && data.l2_assoc == 16 && data.ext_family >= 21) {
+        m_l2_cache = data.l2_cache * (m_totalCores / 2) * m_sockets;
+    }
+    else {
+        m_l2_cache = data.l2_cache > 0 ? data.l2_cache * m_totalCores * m_sockets : 0;
+    }
+
+#   if defined(__x86_64__) || defined(_M_AMD64)
+    m_flags |= X86_64;
+#   endif
+
+    if (data.flags[CPU_FEATURE_AES]) {
+        m_flags |= AES;
+    }
+
+    if (data.flags[CPU_FEATURE_BMI2]) {
+        m_flags |= BMI2;
+    }
+}

src/Cpu.h

@@ -21,48 +21,46 @@
  *   along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#ifndef __OPTIONS_H__
-#define __OPTIONS_H__
+#ifndef __CPU_H__
+#define __CPU_H__
+
 
-#include <stdbool.h>
 #include <stdint.h>
 
-#ifndef ARRAY_SIZE
-#   define ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0]))
-#endif
 
+class Cpu
+{
+public:
+    enum Flags {
+        X86_64 = 1,
+        AES    = 2,
+        BMI2   = 4
+    };
 
-enum xmr_algo_variant {
-    XMR_AV0_AUTO,
-    XMR_AV1_AESNI,
-    XMR_AV2_STAK,
-    XMR_AV3_AESNI_BMI2,
-    XMR_AV4_SOFT_AES,
-    XMR_AV5_EXPERIMENTAL,
-    XMR_AV_MAX
+    static int optimalThreadsCount(int algo, bool doubleHash, int maxCpuUsage);
+    static void init();
+    static void setAffinity(int id, uint64_t mask);
+
+    static inline bool hasAES()       { return m_flags & AES; }
+    static inline bool isX64()        { return m_flags & X86_64; }
+    static inline const char *brand() { return m_brand; }
+    static inline int cores()         { return m_totalCores; }
+    static inline int l2()            { return m_l2_cache; }
+    static inline int l3()            { return m_l3_cache; }
+    static inline int sockets()       { return m_sockets; }
+    static inline int threads()       { return m_totalThreads; }
+
+private:
+    static void initCommon();
+
+    static char m_brand[64];
+    static int m_flags;
+    static int m_l2_cache;
+    static int m_l3_cache;
+    static int m_sockets;
+    static int m_totalCores;
+    static int m_totalThreads;
 };
 
 
-extern bool opt_colors;
-extern bool opt_keepalive;
-extern bool opt_background;
-extern char *opt_url;
-extern char *opt_backup_url;
-extern char *opt_userpass;
-extern char *opt_user;
-extern char *opt_pass;
-extern int opt_n_threads;
-extern int opt_algo_variant;
-extern int opt_retry_pause;
-extern int opt_retries;
-extern int opt_donate_level;
-extern int64_t opt_affinity;
-
-void parse_cmdline(int argc, char *argv[]);
-void show_usage_and_exit(int status);
-void show_version_and_exit(void);
-
-extern void proper_exit(int reason);
-
-
-#endif /* __OPTIONS_H__ */
+#endif /* __CPU_H__ */

src/Cpu_mac.cpp

@@ -21,17 +21,25 @@
  *   along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#ifndef __STATS_H__
-#define __STATS_H__
 
-#include <stdbool.h>
-#include <inttypes.h>
+#include <pthread.h>
+#include <sched.h>
+#include <unistd.h>
 
 
-void stats_init();
-void stats_set_target(uint32_t new_target);
-void stats_share_result(bool success);
-void stats_add_hashes(int thr_id, struct timeval *tv_start, unsigned long hashes_done);
+#include "Cpu.h"
 
 
-#endif /* __STATS_H__ */
+void Cpu::init()
+{
+#   ifdef XMRIG_NO_LIBCPUID
+    m_totalThreads = sysconf(_SC_NPROCESSORS_CONF);
+#   endif
+
+    initCommon();
+}
+
+
+void Cpu::setAffinity(int id, uint64_t mask)
+{
+}

src/Cpu_stub.cpp

@@ -21,10 +21,20 @@
  *   along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <cpuid.h>
+
+#ifdef _MSC_VER
+#   include <intrin.h>
+
+#   define bit_AES  (1 << 25)
+#   define bit_BMI2 (1 << 8)
+#else
+#   include <cpuid.h>
+#endif
+
 #include <string.h>
-#include <stdbool.h>
-#include "cpu.h"
+
+
+#include "Cpu.h"
 
 
 #define VENDOR_ID                  (0)
@@ -41,6 +51,11 @@
 #define EDX_Reg  (3)
 
 
+#ifdef _MSC_VER
+static inline void cpuid(int level, int output[4]) {
+    __cpuid(output, level);
+}
+#else
 static inline void cpuid(int level, int output[4]) {
     int a, b, c, d;
     __cpuid_count(level, 0, a, b, c, d);
@@ -50,9 +65,10 @@ static inline void cpuid(int level, int output[4]) {
     output[2] = c;
     output[3] = d;
 }
+#endif
 
 
-static void cpu_brand_string(char* s) {
+static inline void cpu_brand_string(char* s) {
     int cpu_info[4] = { 0 };
     cpuid(VENDOR_ID, cpu_info);
 
@@ -66,7 +82,7 @@ static void cpu_brand_string(char* s) {
 }
 
 
-static bool has_aes_ni()
+static inline bool has_aes_ni()
 {
     int cpu_info[4] = { 0 };
     cpuid(PROCESSOR_INFO, cpu_info);
@@ -75,7 +91,7 @@ static bool has_aes_ni()
 }
 
 
-static bool has_bmi2() {
+static inline bool has_bmi2() {
     int cpu_info[4] = { 0 };
     cpuid(EXTENDED_FEATURES, cpu_info);
 
@@ -83,18 +99,35 @@ static bool has_bmi2() {
 }
 
 
-void cpu_init_common() {
-    cpu_brand_string(cpu_info.brand);
+char Cpu::m_brand[64]   = { 0 };
+int Cpu::m_flags        = 0;
+int Cpu::m_l2_cache     = 0;
+int Cpu::m_l3_cache     = 0;
+int Cpu::m_sockets      = 1;
+int Cpu::m_totalCores   = 0;
+int Cpu::m_totalThreads = 0;
 
-#   ifdef __x86_64__
-    cpu_info.flags |= CPU_FLAG_X86_64;
+
+int Cpu::optimalThreadsCount(int algo, bool doubleHash, int maxCpuUsage)
+{
+    int count = m_totalThreads / 2;
+    return count < 1 ? 1 : count;
+}
+
+
+void Cpu::initCommon()
+{
+    cpu_brand_string(m_brand);
+
+#   if defined(__x86_64__) || defined(_M_AMD64)
+    m_flags |= X86_64;
 #   endif
 
     if (has_aes_ni()) {
-        cpu_info.flags |= CPU_FLAG_AES;
+        m_flags |= AES;
     }
 
     if (has_bmi2()) {
-        cpu_info.flags |= CPU_FLAG_BMI2;
+        m_flags |= BMI2;
     }
 }

src/Cpu_unix.cpp

@@ -21,36 +21,32 @@
  *   along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <unistd.h>
-#include <sched.h>
+
 #include <pthread.h>
-
-#include "cpu.h"
+#include <sched.h>
+#include <unistd.h>
+#include <string.h>
 
 
-struct cpu_info cpu_info = { 0 };
-void cpu_init_common();
+#include "Cpu.h"
 
 
-void cpu_init() {
-    cpu_info.count = sysconf(_SC_NPROCESSORS_CONF);
+void Cpu::init()
+{
+#   ifdef XMRIG_NO_LIBCPUID
+    m_totalThreads = sysconf(_SC_NPROCESSORS_CONF);
+#   endif
 
-    cpu_init_common();
+    initCommon();
 }
 
 
-int get_optimal_threads_count() {
-    int count = cpu_info.count / 2;
-    return count < 1 ? 1 : count;
-}
-
-
-int affine_to_cpu_mask(int id, unsigned long mask)
+void Cpu::setAffinity(int id, uint64_t mask)
 {
     cpu_set_t set;
     CPU_ZERO(&set);
 
-    for (unsigned i = 0; i < cpu_info.count; i++) {
+    for (int i = 0; i < m_totalThreads; i++) {
         if (mask & (1UL << i)) {
             CPU_SET(i, &set);
         }

src/Cpu_win.cpp

@@ -20,34 +20,28 @@
  *   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 <windows.h>
-#include <stdbool.h>
-
-#include "cpu.h"
 
 
-struct cpu_info cpu_info = { 0 };
-void cpu_init_common();
+#include "Cpu.h"
 
 
-void cpu_init() {
+void Cpu::init()
+{
+#   ifdef XMRIG_NO_LIBCPUID
     SYSTEM_INFO sysinfo;
     GetSystemInfo(&sysinfo);
 
-    cpu_info.count = sysinfo.dwNumberOfProcessors;
+    m_totalThreads = sysinfo.dwNumberOfProcessors;
+#   endif
 
-    cpu_init_common();
+    initCommon();
 }
 
 
-int get_optimal_threads_count(int mining_algo) {
-    int count = cpu_info.count / 2;
-    return count < 1 ? 1 : count;
-}
-
-
-int affine_to_cpu_mask(int id, unsigned long mask)
+void Cpu::setAffinity(int id, uint64_t mask)
 {
     if (id == -1) {
         SetProcessAffinityMask(GetCurrentProcess(), mask);

src/Mem.cpp

@@ -0,0 +1,87 @@
+/* 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 2016-2017 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 <memory.h>
+
+
+#include "crypto/CryptoNight.h"
+#include "Mem.h"
+#include "Options.h"
+
+
+bool Mem::m_doubleHash = false;
+int Mem::m_algo        = 0;
+int Mem::m_flags       = 0;
+int Mem::m_threads     = 0;
+size_t Mem::m_offset   = 0;
+uint8_t *Mem::m_memory = nullptr;
+
+
+cryptonight_ctx *Mem::create(int threadId)
+{
+#   ifndef XMRIG_NO_AEON
+    if (m_algo == Options::ALGO_CRYPTONIGHT_LITE) {
+        return createLite(threadId);
+    }
+#   endif
+
+    cryptonight_ctx *ctx = reinterpret_cast<cryptonight_ctx *>(&m_memory[MEMORY - sizeof(cryptonight_ctx) * (threadId + 1)]);
+
+    const int ratio = m_doubleHash ? 2 : 1;
+    ctx->memory = &m_memory[MEMORY * (threadId * ratio + 1)];
+
+    return ctx;
+}
+
+
+
+void *Mem::calloc(size_t num, size_t size)
+{
+    void *mem = &m_memory[m_offset];
+    m_offset += (num * size);
+
+    memset(mem, 0, num * size);
+
+    return mem;
+}
+
+
+#ifndef XMRIG_NO_AEON
+cryptonight_ctx *Mem::createLite(int threadId) {
+    cryptonight_ctx *ctx;
+
+    if (!m_doubleHash) {
+        const size_t offset = MEMORY * (threadId + 1);
+
+        ctx = reinterpret_cast<cryptonight_ctx *>(&m_memory[offset + MEMORY_LITE]);
+        ctx->memory = &m_memory[offset];
+        return ctx;
+    }
+
+    ctx = reinterpret_cast<cryptonight_ctx *>(&m_memory[MEMORY - sizeof(cryptonight_ctx) * (threadId + 1)]);
+    ctx->memory = &m_memory[MEMORY * (threadId + 1)];
+
+    return ctx;
+}
+#endif

src/Mem.h

@@ -0,0 +1,72 @@
+/* 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 2016-2017 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 __MEM_H__
+#define __MEM_H__
+
+
+#include <stddef.h>
+#include <stdint.h>
+
+
+#include "align.h"
+
+
+struct cryptonight_ctx;
+
+
+class Mem
+{
+public:
+    enum Flags {
+        HugepagesAvailable = 1,
+        HugepagesEnabled   = 2,
+        Lock               = 4
+    };
+
+    static bool allocate(int algo, int threads, bool doubleHash);
+    static cryptonight_ctx *create(int threadId);
+    static void *calloc(size_t num, size_t size);
+    static void release();
+
+    static inline bool isDoubleHash()         { return m_doubleHash; }
+    static inline bool isHugepagesAvailable() { return m_flags & HugepagesAvailable; }
+    static inline bool isHugepagesEnabled()   { return m_flags & HugepagesEnabled; }
+    static inline int flags()                 { return m_flags; }
+    static inline int threads()               { return m_threads; }
+
+private:
+    static bool m_doubleHash;
+    static int m_algo;
+    static int m_flags;
+    static int m_threads;
+    static size_t m_offset;
+    VAR_ALIGN(16, static uint8_t *m_memory);
+
+#   ifndef XMRIG_NO_AEON
+    static cryptonight_ctx *createLite(int threadId);
+#   endif
+};
+
+
+#endif /* __MEM_H__ */

src/Mem_unix.cpp

@@ -0,0 +1,86 @@
+/* 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 2016-2017 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 <stdlib.h>
+#include <mm_malloc.h>
+#include <sys/mman.h>
+
+
+#include "crypto/CryptoNight.h"
+#include "log/Log.h"
+#include "Mem.h"
+#include "Options.h"
+
+
+bool Mem::allocate(int algo, int threads, bool doubleHash)
+{
+    m_algo       = algo;
+    m_threads    = threads;
+    m_doubleHash = doubleHash;
+
+    const int ratio   = (doubleHash && algo != Options::ALGO_CRYPTONIGHT_LITE) ? 2 : 1;
+    const size_t size = MEMORY * (threads * ratio + 1);
+
+    m_flags |= HugepagesAvailable;
+
+#   if defined(__APPLE__)
+    m_memory = static_cast<uint8_t*>(mmap(0, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, VM_FLAGS_SUPERPAGE_SIZE_2MB, 0));
+#   else
+    m_memory = static_cast<uint8_t*>(mmap(0, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB | MAP_POPULATE, 0, 0));
+#   endif
+
+    if (m_memory == MAP_FAILED) {
+        m_memory = static_cast<uint8_t*>(_mm_malloc(size, 16));
+        return true;
+    }
+
+    m_flags |= HugepagesEnabled;
+
+    if (madvise(m_memory, size, MADV_RANDOM | MADV_WILLNEED) != 0) {
+        LOG_ERR("madvise failed");
+    }
+
+    if (mlock(m_memory, size) == 0) {
+        m_flags |= Lock;
+    }
+
+    return true;
+}
+
+
+void Mem::release()
+{
+    const int size = MEMORY * (m_threads + 1);
+
+    if (m_flags & HugepagesEnabled) {
+        if (m_flags & Lock) {
+            munlock(m_memory, size);
+        }
+
+        munmap(m_memory, size);
+    }
+    else {
+        _mm_free(m_memory);
+    }
+}

src/Mem_win.cpp

@@ -0,0 +1,181 @@
+/* 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 2016-2017 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 <winsock2.h>
+#include <windows.h>
+#include <ntsecapi.h>
+#include <tchar.h>
+
+#ifdef __GNUC__
+#   include <mm_malloc.h>
+#else
+#   include <malloc.h>
+#endif
+
+#include "log/Log.h"
+#include "crypto/CryptoNight.h"
+#include "Mem.h"
+#include "Options.h"
+
+
+/*****************************************************************
+SetLockPagesPrivilege: a function to obtain or
+release the privilege of locking physical pages.
+
+Inputs:
+
+HANDLE hProcess: Handle for the process for which the
+privilege is needed
+
+BOOL bEnable: Enable (TRUE) or disable?
+
+Return value: TRUE indicates success, FALSE failure.
+
+*****************************************************************/
+/**
+ * AWE Example: https://msdn.microsoft.com/en-us/library/windows/desktop/aa366531(v=vs.85).aspx
+ * Creating a File Mapping Using Large Pages: https://msdn.microsoft.com/en-us/library/aa366543(VS.85).aspx
+ */
+static BOOL SetLockPagesPrivilege() {
+    HANDLE token;
+
+    if (OpenProcessToken(GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &token) != TRUE) {
+        return FALSE;
+    }
+
+    TOKEN_PRIVILEGES tp;
+    tp.PrivilegeCount = 1;
+    tp.Privileges[0].Attributes = SE_PRIVILEGE_ENABLED;
+
+    if (LookupPrivilegeValue(NULL, SE_LOCK_MEMORY_NAME, &(tp.Privileges[0].Luid)) != TRUE) {
+        return FALSE;
+    }
+
+    BOOL rc = AdjustTokenPrivileges(token, FALSE, (PTOKEN_PRIVILEGES) &tp, 0, NULL, NULL);
+    if (rc != TRUE || GetLastError() != ERROR_SUCCESS) {
+        return FALSE;
+    }
+
+    CloseHandle(token);
+
+    return TRUE;
+}
+
+
+static LSA_UNICODE_STRING StringToLsaUnicodeString(LPCTSTR string) {
+    LSA_UNICODE_STRING lsaString;
+    DWORD dwLen = 0;
+
+    dwLen = wcslen(string);
+    lsaString.Buffer = (LPWSTR) string;
+    lsaString.Length = (USHORT)((dwLen) * sizeof(WCHAR));
+    lsaString.MaximumLength = (USHORT)((dwLen + 1) * sizeof(WCHAR));
+    return lsaString;
+}
+
+
+static BOOL ObtainLockPagesPrivilege() {
+    HANDLE token;
+    PTOKEN_USER user = NULL;
+
+    if (OpenProcessToken(GetCurrentProcess(), TOKEN_QUERY, &token) == TRUE) {
+        DWORD size = 0;
+
+        GetTokenInformation(token, TokenUser, NULL, 0, &size);
+        if (size) {
+            user = (PTOKEN_USER) LocalAlloc(LPTR, size);
+        }
+
+        GetTokenInformation(token, TokenUser, user, size, &size);
+        CloseHandle(token);
+    }
+
+    if (!user) {
+        return FALSE;
+    }
+
+    LSA_HANDLE handle;
+    LSA_OBJECT_ATTRIBUTES attributes;
+    ZeroMemory(&attributes, sizeof(attributes));
+
+    BOOL result = FALSE;
+    if (LsaOpenPolicy(NULL, &attributes, POLICY_ALL_ACCESS, &handle) == 0) {
+        LSA_UNICODE_STRING str = StringToLsaUnicodeString(_T(SE_LOCK_MEMORY_NAME));
+
+        if (LsaAddAccountRights(handle, user->User.Sid, &str, 1) == 0) {
+            LOG_DEBUG("Huge pages support was successfully enabled, but reboot required to use it");
+            result = TRUE;
+        }
+
+        LsaClose(handle);
+    }
+
+    LocalFree(user);
+    return result;
+}
+
+
+static BOOL TrySetLockPagesPrivilege() {
+    if (SetLockPagesPrivilege()) {
+        return TRUE;
+    }
+
+    return ObtainLockPagesPrivilege() && SetLockPagesPrivilege();
+}
+
+
+bool Mem::allocate(int algo, int threads, bool doubleHash)
+{
+    m_algo       = algo;
+    m_threads    = threads;
+    m_doubleHash = doubleHash;
+
+    const int ratio = (doubleHash && algo != Options::ALGO_CRYPTONIGHT_LITE) ? 2 : 1;
+    const size_t size  = MEMORY * (threads * ratio + 1);
+
+    if (TrySetLockPagesPrivilege()) {
+        m_flags |= HugepagesAvailable;
+    }
+
+    m_memory = static_cast<uint8_t*>(VirtualAlloc(NULL, size, MEM_COMMIT | MEM_RESERVE | MEM_LARGE_PAGES, PAGE_READWRITE));
+    if (!m_memory) {
+        m_memory = static_cast<uint8_t*>(_mm_malloc(size, 16));
+    }
+    else {
+        m_flags |= HugepagesEnabled;
+    }
+
+    return true;
+}
+
+
+void Mem::release()
+{
+    if (m_flags & HugepagesEnabled) {
+        VirtualFree(m_memory, 0, MEM_RELEASE);
+    }
+    else {
+        _mm_free(m_memory);
+    }
+}

src/Options.cpp

@@ -0,0 +1,492 @@
+/* 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 2016-2017 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 <jansson.h>
+#include <string.h>
+#include <uv.h>
+
+
+#ifdef _MSC_VER
+#   include "getopt/getopt.h"
+#else
+#   include <getopt.h>
+#endif
+
+
+#include "Cpu.h"
+#include "donate.h"
+#include "net/Url.h"
+#include "Options.h"
+#include "version.h"
+
+
+#ifndef ARRAY_SIZE
+#   define ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0]))
+#endif
+
+
+Options *Options::m_self = nullptr;
+
+
+static char const usage[] = "\
+Usage: " APP_ID " [OPTIONS]\n\
+Options:\n\
+  -a, --algo=ALGO       cryptonight (default) or cryptonight-lite\n\
+  -o, --url=URL         URL of mining server\n\
+  -O, --userpass=U:P    username:password pair for mining server\n\
+  -u, --user=USERNAME   username for mining server\n\
+  -p, --pass=PASSWORD   password for mining server\n\
+  -t, --threads=N       number of miner threads\n\
+  -v, --av=N            algorithm variation, 0 auto select\n\
+  -k, --keepalive       send keepalived for prevent timeout (need pool support)\n\
+  -r, --retries=N       number of times to retry before switch to backup server (default: 5)\n\
+  -R, --retry-pause=N   time to pause between retries (default: 5)\n\
+      --cpu-affinity    set process affinity to CPU core(s), mask 0x3 for cores 0 and 1\n\
+      --no-color        disable colored output\n\
+      --donate-level=N  donate level, default 5%% (5 minutes in 100 minutes)\n\
+  -B, --background      run the miner in the background\n\
+  -c, --config=FILE     load a JSON-format configuration file\n\
+  -l, --log-file=FILE   log all output to a file\n"
+# ifdef HAVE_SYSLOG_H
+"\
+  -S, --syslog          use system log for output messages\n"
+# endif
+"\
+      --max-cpu-usage=N maximum CPU usage for automatic threads mode (default 75)\n\
+      --safe            safe adjust threads and av settings for current CPU\n\
+      --nicehash        enable nicehash support\n\
+      --print-time=N    print hashrate report every N seconds\n\
+  -h, --help            display this help and exit\n\
+  -V, --version         output version information and exit\n\
+";
+
+
+static char const short_options[] = "a:c:khBp:Px:r:R:s:t:T:o:u:O:v:Vl:S";
+
+
+static struct option const options[] = {
+    { "algo",          1, nullptr, 'a'  },
+    { "av",            1, nullptr, 'v'  },
+    { "background",    0, nullptr, 'B'  },
+    { "config",        1, nullptr, 'c'  },
+    { "cpu-affinity",  1, nullptr, 1020 },
+    { "donate-level",  1, nullptr, 1003 },
+    { "help",          0, nullptr, 'h'  },
+    { "keepalive",     0, nullptr ,'k'  },
+    { "log-file",      1, nullptr, 'l'  },
+    { "max-cpu-usage", 1, nullptr, 1004 },
+    { "nicehash",      0, nullptr, 1006 },
+    { "no-color",      0, nullptr, 1002 },
+    { "pass",          1, nullptr, 'p'  },
+    { "print-time",    1, nullptr, 1007 },
+    { "retries",       1, nullptr, 'r'  },
+    { "retry-pause",   1, nullptr, 'R'  },
+    { "safe",          0, nullptr, 1005 },
+    { "syslog",        0, nullptr, 'S'  },
+    { "threads",       1, nullptr, 't'  },
+    { "url",           1, nullptr, 'o'  },
+    { "user",          1, nullptr, 'u'  },
+    { "userpass",      1, nullptr, 'O'  },
+    { "version",       0, nullptr, 'V'  },
+    { 0, 0, 0, 0 }
+};
+
+
+static const char *algo_names[] = {
+    "cryptonight",
+#   ifndef XMRIG_NO_AEON
+    "cryptonight-lite"
+#   endif
+};
+
+
+Options *Options::parse(int argc, char **argv)
+{
+    if (!m_self) {
+        m_self = new Options(argc, argv);
+    }
+
+    return m_self;
+}
+
+
+const char *Options::algoName() const
+{
+    return algo_names[m_algo];
+}
+
+
+Options::Options(int argc, char **argv) :
+    m_background(false),
+    m_colors(true),
+    m_doubleHash(false),
+    m_ready(false),
+    m_safe(false),
+    m_syslog(false),
+    m_logFile(nullptr),
+    m_algo(0),
+    m_algoVariant(0),
+    m_donateLevel(kDonateLevel),
+    m_maxCpuUsage(75),
+    m_printTime(60),
+    m_retries(5),
+    m_retryPause(5),
+    m_threads(0),
+    m_affinity(-1L)
+{
+    m_pools.push_back(new Url());
+
+    int key;
+
+    while (1) {
+        key = getopt_long(argc, argv, short_options, options, NULL);
+        if (key < 0) {
+            break;
+        }
+
+        if (!parseArg(key, optarg)) {
+            return;
+        }
+    }
+
+    if (optind < argc) {
+        fprintf(stderr, "%s: unsupported non-option argument '%s'\n", argv[0], argv[optind]);
+        return;
+    }
+
+    if (!m_pools[0]->isValid()) {
+        fprintf(stderr, "No pool URL supplied. Exiting.");
+        return;
+    }
+
+    m_algoVariant = getAlgoVariant();
+    if (m_algoVariant == AV2_AESNI_DOUBLE || m_algoVariant == AV4_SOFT_AES_DOUBLE) {
+        m_doubleHash = true;
+    }
+
+    if (!m_threads) {
+        m_threads = Cpu::optimalThreadsCount(m_algo, m_doubleHash, m_maxCpuUsage);
+    }
+    else if (m_safe) {
+        const int count = Cpu::optimalThreadsCount(m_algo, m_doubleHash, m_maxCpuUsage);
+        if (m_threads > count) {
+            m_threads = count;
+        }
+    }
+
+    m_ready = true;
+}
+
+
+Options::~Options()
+{
+}
+
+
+bool Options::parseArg(int key, char *arg)
+{
+    char *p;
+    int v;
+
+    switch (key) {
+    case 'a': /* --algo */
+        if (!setAlgo(arg)) {
+            return false;
+        }
+        break;
+
+    case 'O': /* --userpass */
+        if (!m_pools.back()->setUserpass(arg)) {
+            return false;
+        }
+
+        break;
+
+    case 'o': /* --url */
+        if (m_pools.size() > 1 || m_pools[0]->isValid()) {
+            Url *url = new Url(arg);
+            if (url->isValid()) {
+                m_pools.push_back(url);
+            }
+            else {
+                delete url;
+            }
+        }
+        else {
+            m_pools[0]->parse(arg);
+        }
+
+        if (!m_pools.back()->isValid()) {
+            return false;
+        }
+
+        break;
+
+    case 'u': /* --user */
+        m_pools.back()->setUser(arg);
+        break;
+
+    case 'p': /* --pass */
+        m_pools.back()->setPassword(arg);
+        break;
+
+    case 'l': /* --log-file */
+        free(m_logFile);
+        m_logFile = strdup(arg);
+        m_colors = false;
+        break;
+
+    case 'r': /* --retries */
+        v = strtol(arg, nullptr, 10);
+        if (v < 1 || v > 1000) {
+            showUsage(1);
+            return false;
+        }
+
+        m_retries = v;
+        break;
+
+    case 'R': /* --retry-pause */
+        v = strtol(arg, nullptr, 10);
+        if (v < 1 || v > 3600) {
+            showUsage(1);
+            return false;
+        }
+
+        m_retryPause = v;
+        break;
+
+    case 't': /* --threads */
+        v = strtol(arg, nullptr, 10);
+        if (v < 1 || v > 1024) {
+            showUsage(1);
+            return false;
+        }
+
+        m_threads = v;
+        break;
+
+    case 1004: /* --max-cpu-usage */
+        v = strtol(arg, nullptr, 10);
+        if (v < 1 || v > 100) {
+            showUsage(1);
+            return false;
+        }
+
+        m_maxCpuUsage = v;
+        break;
+
+    case 1005: /* --safe */
+        m_safe = true;
+        break;
+
+    case 'k': /* --keepalive */
+        m_pools.back()->setKeepAlive(true);
+        break;
+
+    case 'V': /* --version */
+        showVersion();
+        return false;
+
+    case 'h': /* --help */
+        showUsage(0);
+        return false;
+
+    case 'B': /* --background */
+        m_background = true;
+        m_colors = false;
+        break;
+
+    case 'S': /* --syslog */
+        m_syslog = true;
+        m_colors = false;
+        break;
+
+    case 'v': /* --av */
+        v = strtol(arg, nullptr, 10);
+        if (v < 0 || v > 1000) {
+            showUsage(1);
+            return false;
+        }
+
+        m_algoVariant = v;
+        break;
+
+    case 1020: /* --cpu-affinity */
+        p  = strstr(arg, "0x");
+        m_affinity = p ? strtoull(p, nullptr, 16) : strtoull(arg, nullptr, 10);
+        break;
+
+    case 1002: /* --no-color */
+        m_colors = false;
+        break;
+
+    case 1003: /* --donate-level */
+        v = strtol(arg, nullptr, 10);
+        if (v < 1 || v > 99) {
+            showUsage(1);
+            return false;
+        }
+
+        m_donateLevel = v;
+        break;
+
+    case 1006: /* --nicehash */
+        m_pools.back()->setNicehash(true);
+        break;
+
+    case 1007: /* --print-time */
+        v = strtol(arg, nullptr, 10);
+        if (v < 0 || v > 1000) {
+            showUsage(1);
+            return false;
+        }
+
+        m_printTime = v;
+        break;
+
+    default:
+        showUsage(1);
+        return false;
+    }
+
+    return true;
+}
+
+
+Url *Options::parseUrl(const char *arg) const
+{
+    auto url = new Url(arg);
+    if (!url->isValid()) {
+        delete url;
+        return nullptr;
+    }
+
+    return url;
+}
+
+
+void Options::showUsage(int status) const
+{
+    if (status) {
+        fprintf(stderr, "Try \"" APP_ID "\" --help' for more information.\n");
+    }
+    else {
+        printf(usage);
+    }
+}
+
+
+void Options::showVersion()
+{
+    printf(APP_NAME " " APP_VERSION "\n built on " __DATE__
+
+#   if defined(__clang__)
+    " with clang " __clang_version__);
+#   elif defined(__GNUC__)
+    " with GCC");
+    printf(" %d.%d.%d", __GNUC__, __GNUC_MINOR__, __GNUC_PATCHLEVEL__);
+#   elif defined(_MSC_VER)
+    " with MSVC");
+    printf(" %d", MSVC_VERSION);
+#   else
+    );
+#   endif
+
+    printf("\n features:"
+#   if defined(__i386__) || defined(_M_IX86)
+    " i386"
+#   elif defined(__x86_64__) || defined(_M_AMD64)
+    " x86_64"
+#   endif
+
+#   if defined(__AES__) || defined(_MSC_VER)
+    " AES-NI"
+#   endif
+    "\n");
+
+    printf("\nlibuv/%s\n", uv_version_string());
+    printf("libjansson/%s\n", JANSSON_VERSION);
+}
+
+
+bool Options::setAlgo(const char *algo)
+{
+    for (size_t i = 0; i < ARRAY_SIZE(algo_names); i++) {
+        if (algo_names[i] && !strcmp(algo, algo_names[i])) {
+            m_algo = i;
+            break;
+        }
+
+#       ifndef XMRIG_NO_AEON
+        if (i == ARRAY_SIZE(algo_names) - 1 && !strcmp(algo, "cryptonight-light")) {
+            m_algo = ALGO_CRYPTONIGHT_LITE;
+            break;
+        }
+#       endif
+
+        if (i == ARRAY_SIZE(algo_names) - 1) {
+            showUsage(1);
+            return false;
+        }
+    }
+
+    return true;
+}
+
+
+int Options::getAlgoVariant() const
+{
+#   ifndef XMRIG_NO_AEON
+    if (m_algo == ALGO_CRYPTONIGHT_LITE) {
+        return getAlgoVariantLite();
+    }
+#   endif
+
+    if (m_algoVariant <= AV0_AUTO || m_algoVariant >= AV_MAX) {
+        return Cpu::hasAES() ? AV1_AESNI : AV3_SOFT_AES;
+    }
+
+    if (m_safe && !Cpu::hasAES() && m_algoVariant <= AV2_AESNI_DOUBLE) {
+        return m_algoVariant + 2;
+    }
+
+    return m_algoVariant;
+}
+
+
+#ifndef XMRIG_NO_AEON
+int Options::getAlgoVariantLite() const
+{
+    if (m_algoVariant <= AV0_AUTO || m_algoVariant >= AV_MAX) {
+        return Cpu::hasAES() ? AV2_AESNI_DOUBLE : AV4_SOFT_AES_DOUBLE;
+    }
+
+    if (m_safe && !Cpu::hasAES() && m_algoVariant <= AV2_AESNI_DOUBLE) {
+        return m_algoVariant + 2;
+    }
+
+    return m_algoVariant;
+}
+#endif

src/Options.h

@@ -0,0 +1,110 @@
+/* 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 2016-2017 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 __OPTIONS_H__
+#define __OPTIONS_H__
+
+
+#include <vector>
+#include <stdint.h>
+
+
+class Url;
+
+
+class Options
+{
+public:
+    enum Algo {
+        ALGO_CRYPTONIGHT,      /* CryptoNight (Monero) */
+        ALGO_CRYPTONIGHT_LITE, /* CryptoNight-Lite (AEON) */
+    };
+
+    enum AlgoVariant {
+        AV0_AUTO,
+        AV1_AESNI,
+        AV2_AESNI_DOUBLE,
+        AV3_SOFT_AES,
+        AV4_SOFT_AES_DOUBLE,
+        AV_MAX
+    };
+
+    static inline Options* i() { return m_self; }
+    static Options *parse(int argc, char **argv);
+
+    inline bool background() const                { return m_background; }
+    inline bool colors() const                    { return m_colors; }
+    inline bool doubleHash() const                { return m_doubleHash; }
+    inline bool isReady() const                   { return m_ready; }
+    inline bool syslog() const                    { return m_syslog; }
+    inline const char *logFile() const            { return m_logFile; }
+    inline const std::vector<Url*> &pools() const { return m_pools; }
+    inline int algo() const                       { return m_algo; }
+    inline int algoVariant() const                { return m_algoVariant; }
+    inline int donateLevel() const                { return m_donateLevel; }
+    inline int printTime() const                  { return m_printTime; }
+    inline int retries() const                    { return m_retries; }
+    inline int retryPause() const                 { return m_retryPause; }
+    inline int threads() const                    { return m_threads; }
+    inline int64_t affinity() const               { return m_affinity; }
+
+    const char *algoName() const;
+
+private:
+    Options(int argc, char **argv);
+    ~Options();
+
+    static Options *m_self;
+
+    bool parseArg(int key, char *arg);
+    Url *parseUrl(const char *arg) const;
+    void showUsage(int status) const;
+    void showVersion(void);
+
+    bool setAlgo(const char *algo);
+
+    int getAlgoVariant() const;
+#   ifndef XMRIG_NO_AEON
+    int getAlgoVariantLite() const;
+#   endif
+
+    bool m_background;
+    bool m_colors;
+    bool m_doubleHash;
+    bool m_ready;
+    bool m_safe;
+    bool m_syslog;
+    char *m_logFile;
+    int m_algo;
+    int m_algoVariant;
+    int m_donateLevel;
+    int m_maxCpuUsage;
+    int m_printTime;
+    int m_retries;
+    int m_retryPause;
+    int m_threads;
+    int64_t m_affinity;
+    std::vector<Url*> m_pools;
+};
+
+#endif /* __OPTIONS_H__ */

src/Summary.cpp

@@ -0,0 +1,143 @@
+/* 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 2016-2017 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 <inttypes.h>
+#include <uv.h>
+
+
+#include "Cpu.h"
+#include "log/Log.h"
+#include "Mem.h"
+#include "net/Url.h"
+#include "Options.h"
+#include "Summary.h"
+#include "version.h"
+
+
+static void print_versions()
+{
+    char buf[16];
+
+#   if defined(__clang__)
+    snprintf(buf, 16, " clang/%d.%d.%d", __clang_major__, __clang_minor__, __clang_patchlevel__);
+#   elif defined(__GNUC__)
+    snprintf(buf, 16, " gcc/%d.%d.%d", __GNUC__, __GNUC_MINOR__, __GNUC_PATCHLEVEL__);
+#   elif defined(_MSC_VER)
+    snprintf(buf, 16, " MSVC/%d", MSVC_VERSION);
+#   else
+    buf[0] = '\0';
+#   endif
+
+
+    if (Options::i()->colors()) {
+        Log::i()->text("\x1B[01;32m * \x1B[01;37mVERSIONS:     \x1B[01;36mXMRig/%s\x1B[01;37m libuv/%s%s", APP_VERSION, uv_version_string(), buf);
+    } else {
+        Log::i()->text(" * VERSIONS:     XMRig/%s libuv/%s%s", APP_VERSION, uv_version_string(), buf);
+    }
+}
+
+
+static void print_memory() {
+    if (Options::i()->colors()) {
+        Log::i()->text("\x1B[01;32m * \x1B[01;37mHUGE PAGES:   %s, %s",
+                       Mem::isHugepagesAvailable() ? "\x1B[01;32mavailable" : "\x1B[01;31munavailable",
+                       Mem::isHugepagesEnabled() ? "\x1B[01;32menabled" : "\x1B[01;31mdisabled");
+    }
+    else {
+        Log::i()->text(" * HUGE PAGES:   %s, %s", Mem::isHugepagesAvailable() ? "available" : "unavailable", Mem::isHugepagesEnabled() ? "enabled" : "disabled");
+    }
+}
+
+
+static void print_cpu()
+{
+    if (Options::i()->colors()) {
+        Log::i()->text("\x1B[01;32m * \x1B[01;37mCPU:          %s (%d) %sx64 %sAES-NI",
+                       Cpu::brand(),
+                       Cpu::sockets(),
+                       Cpu::isX64() ? "\x1B[01;32m" : "\x1B[01;31m-",
+                       Cpu::hasAES() ? "\x1B[01;32m" : "\x1B[01;31m-");
+#       ifndef XMRIG_NO_LIBCPUID
+        Log::i()->text("\x1B[01;32m * \x1B[01;37mCPU L2/L3:    %.1f MB/%.1f MB", Cpu::l2() / 1024.0, Cpu::l3() / 1024.0);
+#       endif
+    }
+    else {
+        Log::i()->text(" * CPU:          %s (%d) %sx64 %sAES-NI", Cpu::brand(), Cpu::sockets(), Cpu::isX64() ? "" : "-", Cpu::hasAES() ? "" : "-");
+#       ifndef XMRIG_NO_LIBCPUID
+        Log::i()->text(" * CPU L2/L3:    %.1f MB/%.1f MB", Cpu::l2() / 1024.0, Cpu::l3() / 1024.0);
+#       endif
+    }
+}
+
+
+static void print_threads()
+{
+    char buf[32];
+    if (Options::i()->affinity() != -1L) {
+        snprintf(buf, 32, ", affinity=0x%" PRIX64, Options::i()->affinity());
+    }
+    else {
+        buf[0] = '\0';
+    }
+
+    Log::i()->text(Options::i()->colors() ? "\x1B[01;32m * \x1B[01;37mTHREADS:      \x1B[01;36m%d\x1B[01;37m, %s, av=%d, %sdonate=%d%%%s" : " * THREADS:      %d, %s, av=%d, %sdonate=%d%%%s",
+                   Options::i()->threads(),
+                   Options::i()->algoName(),
+                   Options::i()->algoVariant(),
+                   Options::i()->colors() && Options::i()->donateLevel() == 0 ? "\x1B[01;31m" : "",
+                   Options::i()->donateLevel(),
+                   buf);
+}
+
+
+static void print_pools()
+{
+    const std::vector<Url*> &pools = Options::i()->pools();
+
+    for (size_t i = 0; i < pools.size(); ++i) {
+        Log::i()->text(Options::i()->colors() ? "\x1B[01;32m * \x1B[01;37mPOOL #%d:      \x1B[01;36m%s:%d" : " * POOL #%d:      %s:%d",
+                       i + 1,
+                       pools[i]->host(),
+                       pools[i]->port());
+    }
+
+#   ifdef APP_DEBUG
+    for (size_t i = 0; i < pools.size(); ++i) {
+        Log::i()->text("%s:%d, user: %s, pass: %s, ka: %d, nicehash: %d", pools[i]->host(), pools[i]->port(), pools[i]->user(), pools[i]->password(), pools[i]->isKeepAlive(), pools[i]->isNicehash());
+    }
+#   endif
+}
+
+
+void Summary::print()
+{
+    print_versions();
+    print_memory();
+    print_cpu();
+    print_threads();
+    print_pools();
+}
+
+
+

src/Summary.h

@@ -24,6 +24,12 @@
 #ifndef __SUMMARY_H__
 #define __SUMMARY_H__
 
-void print_summary();
+
+class Summary
+{
+public:
+    static void print();
+};
+
 
 #endif /* __SUMMARY_H__ */

src/crypto/CryptoNight.cpp

@@ -0,0 +1,144 @@
+/* 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 2016-2017 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 "crypto/CryptoNight.h"
+#include "crypto/CryptoNight_p.h"
+#include "crypto/CryptoNight_test.h"
+#include "net/Job.h"
+#include "net/JobResult.h"
+#include "Options.h"
+
+
+void (*cryptonight_hash_ctx)(const void *input, size_t size, void *output, cryptonight_ctx *ctx) = nullptr;
+
+
+static void cryptonight_av1_aesni(const void *input, size_t size, void *output, struct cryptonight_ctx *ctx) {
+    cryptonight_hash<0x80000, MEMORY, 0x1FFFF0, false>(input, size, output, ctx);
+}
+
+
+static void cryptonight_av2_aesni_double(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_double_hash<0x80000, MEMORY, 0x1FFFF0, false>(input, size, output, ctx);
+}
+
+
+static void cryptonight_av3_softaes(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_hash<0x80000, MEMORY, 0x1FFFF0, true>(input, size, output, ctx);
+}
+
+
+static void cryptonight_av4_softaes_double(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_double_hash<0x80000, MEMORY, 0x1FFFF0, true>(input, size, output, ctx);
+}
+
+
+#ifndef XMRIG_NO_AEON
+static void cryptonight_lite_av1_aesni(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_hash<0x40000, MEMORY_LITE, 0xFFFF0, false>(input, size, output, ctx);
+}
+
+
+static void cryptonight_lite_av2_aesni_double(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_double_hash<0x40000, MEMORY_LITE, 0xFFFF0, false>(input, size, output, ctx);
+}
+
+
+static void cryptonight_lite_av3_softaes(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_hash<0x40000, MEMORY_LITE, 0xFFFF0, true>(input, size, output, ctx);
+}
+
+
+static void cryptonight_lite_av4_softaes_double(const void *input, size_t size, void *output, cryptonight_ctx *ctx) {
+    cryptonight_double_hash<0x40000, MEMORY_LITE, 0xFFFF0, true>(input, size, output, ctx);
+}
+
+void (*cryptonight_variations[8])(const void *input, size_t size, void *output, cryptonight_ctx *ctx) = {
+            cryptonight_av1_aesni,
+            cryptonight_av2_aesni_double,
+            cryptonight_av3_softaes,
+            cryptonight_av4_softaes_double,
+            cryptonight_lite_av1_aesni,
+            cryptonight_lite_av2_aesni_double,
+            cryptonight_lite_av3_softaes,
+            cryptonight_lite_av4_softaes_double
+        };
+#else
+void (*cryptonight_variations[4])(const void *input, size_t size, void *output, cryptonight_ctx *ctx) = {
+            cryptonight_av1_aesni,
+            cryptonight_av2_aesni_double,
+            cryptonight_av3_softaes,
+            cryptonight_av4_softaes_double
+        };
+#endif
+
+
+bool CryptoNight::hash(const Job &job, JobResult &result, cryptonight_ctx *ctx)
+{
+    cryptonight_hash_ctx(job.blob(), job.size(), result.result, ctx);
+
+    return *reinterpret_cast<uint64_t*>(result.result + 24) < job.target();
+}
+
+
+bool CryptoNight::init(int algo, int variant)
+{
+    if (variant < 1 || variant > 4) {
+        return false;
+    }
+
+#   ifndef XMRIG_NO_AEON
+    const int index = algo == Options::ALGO_CRYPTONIGHT_LITE ? (variant + 3) : (variant - 1);
+#   else
+    const int index = variant - 1;
+#   endif
+
+    cryptonight_hash_ctx = cryptonight_variations[index];
+
+    return selfTest(algo);
+}
+
+
+void CryptoNight::hash(const uint8_t *input, size_t size, uint8_t *output, cryptonight_ctx *ctx)
+{
+    cryptonight_hash_ctx(input, size, output, ctx);
+}
+
+
+bool CryptoNight::selfTest(int algo) {
+    if (cryptonight_hash_ctx == nullptr) {
+        return false;
+    }
+
+    char output[64];
+
+    struct cryptonight_ctx *ctx = (struct cryptonight_ctx*) _mm_malloc(sizeof(struct cryptonight_ctx), 16);
+    ctx->memory = (uint8_t *) _mm_malloc(MEMORY * 2, 16);
+
+    cryptonight_hash_ctx(test_input, 76, output, ctx);
+
+    _mm_free(ctx->memory);
+    _mm_free(ctx);
+
+    return memcmp(output, algo == Options::ALGO_CRYPTONIGHT_LITE ? test_output1 : test_output0, (Options::i()->doubleHash() ? 64 : 32)) == 0;
+}

src/crypto/CryptoNight.h

@@ -24,22 +24,38 @@
 #ifndef __CRYPTONIGHT_H__
 #define __CRYPTONIGHT_H__
 
+
 #include <stddef.h>
 #include <stdint.h>
-#include <stdbool.h>
 
-#define MEMORY 2097152 /* 2 MiB */
+
+#include "align.h"
+
+
+#define MEMORY      2097152 /* 2 MiB */
+#define MEMORY_LITE 1048576 /* 1 MiB */
 
 
 struct cryptonight_ctx {
-    uint8_t state[200] __attribute__((aligned(16)));
-    uint8_t* memory    __attribute__((aligned(16)));
+    VAR_ALIGN(16, uint8_t state0[200]);
+    VAR_ALIGN(16, uint8_t state1[200]);
+    VAR_ALIGN(16, uint8_t* memory);
 };
 
 
-extern void (* const extra_hashes[4])(const void *, size_t, char *);
+class Job;
+class JobResult;
 
-bool cryptonight_init(int variant);
-int scanhash_cryptonight(int thr_id, uint32_t *hash, uint32_t *restrict blob, size_t blob_size, uint32_t target, uint32_t max_nonce, unsigned long *restrict hashes_done, struct cryptonight_ctx *restrict ctx);
+
+class CryptoNight
+{
+public:
+    static bool hash(const Job &job, JobResult &result, cryptonight_ctx *ctx);
+    static bool init(int algo, int variant);
+    static void hash(const uint8_t *input, size_t size, uint8_t *output, cryptonight_ctx *ctx);
+
+private:
+    static bool selfTest(int algo);
+};
 
 #endif /* __CRYPTONIGHT_H__ */

src/crypto/CryptoNight_p.h

@@ -0,0 +1,452 @@
+/* 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 2016-2017 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 __CRYPTONIGHT_P_H__
+#define __CRYPTONIGHT_P_H__
+
+
+#ifdef __GNUC__
+#   include <x86intrin.h>
+#else
+#   include <intrin.h>
+#   define __restrict__ __restrict
+#endif
+
+
+#include "crypto/CryptoNight.h"
+
+
+extern "C"
+{
+#include "crypto/c_keccak.h"
+#include "crypto/c_groestl.h"
+#include "crypto/c_blake256.h"
+#include "crypto/c_jh.h"
+#include "crypto/c_skein.h"
+
+__m128i soft_aesenc(__m128i in, __m128i key);
+__m128i soft_aeskeygenassist(__m128i key, uint8_t rcon);
+}
+
+
+static inline void do_blake_hash(const void* input, size_t len, char* output) {
+    blake256_hash(reinterpret_cast<uint8_t*>(output), static_cast<const uint8_t*>(input), len);
+}
+
+
+static inline void do_groestl_hash(const void* input, size_t len, char* output) {
+    groestl(static_cast<const uint8_t*>(input), len * 8, reinterpret_cast<uint8_t*>(output));
+}
+
+
+static inline void do_jh_hash(const void* input, size_t len, char* output) {
+    jh_hash(32 * 8, static_cast<const uint8_t*>(input), 8 * len, reinterpret_cast<uint8_t*>(output));
+}
+
+
+static inline void do_skein_hash(const void* input, size_t len, char* output) {
+    xmr_skein(static_cast<const uint8_t*>(input), reinterpret_cast<uint8_t*>(output));
+}
+
+
+void (* const extra_hashes[4])(const void *, size_t, char *) = {do_blake_hash, do_groestl_hash, do_jh_hash, do_skein_hash};
+
+
+
+#if defined(__x86_64__) || defined(_M_AMD64)
+#   define EXTRACT64(X) _mm_cvtsi128_si64(X)
+
+#   ifdef __GNUC__
+static inline uint64_t __umul128(uint64_t a, uint64_t b, uint64_t* hi)
+{
+    unsigned __int128 r = (unsigned __int128) a * (unsigned __int128) b;
+    *hi = r >> 64;
+    return (uint64_t) r;
+}
+#   else
+    #define __umul128 _umul128
+#   endif
+#elif defined(__i386__) || defined(_M_IX86)
+#   define HI32(X) \
+    _mm_srli_si128((X), 4)
+
+
+#   define EXTRACT64(X) \
+    ((uint64_t)(uint32_t)_mm_cvtsi128_si32(X) | \
+    ((uint64_t)(uint32_t)_mm_cvtsi128_si32(HI32(X)) << 32))
+
+static inline uint64_t __umul128(uint64_t multiplier, uint64_t multiplicand, uint64_t *product_hi) {
+    // multiplier   = ab = a * 2^32 + b
+    // multiplicand = cd = c * 2^32 + d
+    // ab * cd = a * c * 2^64 + (a * d + b * c) * 2^32 + b * d
+    uint64_t a = multiplier >> 32;
+    uint64_t b = multiplier & 0xFFFFFFFF;
+    uint64_t c = multiplicand >> 32;
+    uint64_t d = multiplicand & 0xFFFFFFFF;
+
+    //uint64_t ac = a * c;
+    uint64_t ad = a * d;
+    //uint64_t bc = b * c;
+    uint64_t bd = b * d;
+
+    uint64_t adbc = ad + (b * c);
+    uint64_t adbc_carry = adbc < ad ? 1 : 0;
+
+    // multiplier * multiplicand = product_hi * 2^64 + product_lo
+    uint64_t product_lo = bd + (adbc << 32);
+    uint64_t product_lo_carry = product_lo < bd ? 1 : 0;
+    *product_hi = (a * c) + (adbc >> 32) + (adbc_carry << 32) + product_lo_carry;
+
+    return product_lo;
+}
+#endif
+
+
+// This will shift and xor tmp1 into itself as 4 32-bit vals such as
+// sl_xor(a1 a2 a3 a4) = a1 (a2^a1) (a3^a2^a1) (a4^a3^a2^a1)
+static inline __m128i sl_xor(__m128i tmp1)
+{
+    __m128i tmp4;
+    tmp4 = _mm_slli_si128(tmp1, 0x04);
+    tmp1 = _mm_xor_si128(tmp1, tmp4);
+    tmp4 = _mm_slli_si128(tmp4, 0x04);
+    tmp1 = _mm_xor_si128(tmp1, tmp4);
+    tmp4 = _mm_slli_si128(tmp4, 0x04);
+    tmp1 = _mm_xor_si128(tmp1, tmp4);
+    return tmp1;
+}
+
+
+template<uint8_t rcon>
+static inline void aes_genkey_sub(__m128i* xout0, __m128i* xout2)
+{
+    __m128i xout1 = _mm_aeskeygenassist_si128(*xout2, rcon);
+    xout1  = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
+    *xout0 = sl_xor(*xout0);
+    *xout0 = _mm_xor_si128(*xout0, xout1);
+    xout1  = _mm_aeskeygenassist_si128(*xout0, 0x00);
+    xout1  = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
+    *xout2 = sl_xor(*xout2);
+    *xout2 = _mm_xor_si128(*xout2, xout1);
+}
+
+
+static inline void soft_aes_genkey_sub(__m128i* xout0, __m128i* xout2, uint8_t rcon)
+{
+    __m128i xout1 = soft_aeskeygenassist(*xout2, rcon);
+    xout1  = _mm_shuffle_epi32(xout1, 0xFF); // see PSHUFD, set all elems to 4th elem
+    *xout0 = sl_xor(*xout0);
+    *xout0 = _mm_xor_si128(*xout0, xout1);
+    xout1  = soft_aeskeygenassist(*xout0, 0x00);
+    xout1  = _mm_shuffle_epi32(xout1, 0xAA); // see PSHUFD, set all elems to 3rd elem
+    *xout2 = sl_xor(*xout2);
+    *xout2 = _mm_xor_si128(*xout2, xout1);
+}
+
+
+template<bool SOFT_AES>
+static inline void aes_genkey(const __m128i* memory, __m128i* k0, __m128i* k1, __m128i* k2, __m128i* k3, __m128i* k4, __m128i* k5, __m128i* k6, __m128i* k7, __m128i* k8, __m128i* k9)
+{
+    __m128i xout0 = _mm_load_si128(memory);
+    __m128i xout2 = _mm_load_si128(memory +1 );
+    *k0 = xout0;
+    *k1 = xout2;
+
+    SOFT_AES ? soft_aes_genkey_sub(&xout0, &xout2, 0x01) : aes_genkey_sub<0x01>(&xout0, &xout2);
+    *k2 = xout0;
+    *k3 = xout2;
+
+    SOFT_AES ? soft_aes_genkey_sub(&xout0, &xout2, 0x02) : aes_genkey_sub<0x02>(&xout0, &xout2);
+    *k4 = xout0;
+    *k5 = xout2;
+
+    SOFT_AES ? soft_aes_genkey_sub(&xout0, &xout2, 0x04) : aes_genkey_sub<0x04>(&xout0, &xout2);
+    *k6 = xout0;
+    *k7 = xout2;
+
+    SOFT_AES ? soft_aes_genkey_sub(&xout0, &xout2, 0x08) : aes_genkey_sub<0x08>(&xout0, &xout2);
+    *k8 = xout0;
+    *k9 = xout2;
+}
+
+
+template<bool SOFT_AES>
+static inline void aes_round(__m128i key, __m128i* x0, __m128i* x1, __m128i* x2, __m128i* x3, __m128i* x4, __m128i* x5, __m128i* x6, __m128i* x7)
+{
+    if (SOFT_AES) {
+        *x0 = soft_aesenc(*x0, key);
+        *x1 = soft_aesenc(*x1, key);
+        *x2 = soft_aesenc(*x2, key);
+        *x3 = soft_aesenc(*x3, key);
+        *x4 = soft_aesenc(*x4, key);
+        *x5 = soft_aesenc(*x5, key);
+        *x6 = soft_aesenc(*x6, key);
+        *x7 = soft_aesenc(*x7, key);
+    }
+    else {
+        *x0 = _mm_aesenc_si128(*x0, key);
+        *x1 = _mm_aesenc_si128(*x1, key);
+        *x2 = _mm_aesenc_si128(*x2, key);
+        *x3 = _mm_aesenc_si128(*x3, key);
+        *x4 = _mm_aesenc_si128(*x4, key);
+        *x5 = _mm_aesenc_si128(*x5, key);
+        *x6 = _mm_aesenc_si128(*x6, key);
+        *x7 = _mm_aesenc_si128(*x7, key);
+    }
+}
+
+
+template<size_t MEM, bool SOFT_AES>
+static inline void cn_explode_scratchpad(const __m128i *input, __m128i *output)
+{
+    __m128i xin0, xin1, xin2, xin3, xin4, xin5, xin6, xin7;
+    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
+
+    aes_genkey<SOFT_AES>(input, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
+
+    xin0 = _mm_load_si128(input + 4);
+    xin1 = _mm_load_si128(input + 5);
+    xin2 = _mm_load_si128(input + 6);
+    xin3 = _mm_load_si128(input + 7);
+    xin4 = _mm_load_si128(input + 8);
+    xin5 = _mm_load_si128(input + 9);
+    xin6 = _mm_load_si128(input + 10);
+    xin7 = _mm_load_si128(input + 11);
+
+    for (size_t i = 0; i < MEM / sizeof(__m128i); i += 8) {
+        aes_round<SOFT_AES>(k0, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k1, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k2, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k3, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k4, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k5, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k6, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k7, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k8, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+        aes_round<SOFT_AES>(k9, &xin0, &xin1, &xin2, &xin3, &xin4, &xin5, &xin6, &xin7);
+
+        _mm_store_si128(output + i + 0, xin0);
+        _mm_store_si128(output + i + 1, xin1);
+        _mm_store_si128(output + i + 2, xin2);
+        _mm_store_si128(output + i + 3, xin3);
+        _mm_store_si128(output + i + 4, xin4);
+        _mm_store_si128(output + i + 5, xin5);
+        _mm_store_si128(output + i + 6, xin6);
+        _mm_store_si128(output + i + 7, xin7);
+    }
+}
+
+
+template<size_t MEM, bool SOFT_AES>
+static inline void cn_implode_scratchpad(const __m128i *input, __m128i *output)
+{
+    __m128i xout0, xout1, xout2, xout3, xout4, xout5, xout6, xout7;
+    __m128i k0, k1, k2, k3, k4, k5, k6, k7, k8, k9;
+
+    aes_genkey<SOFT_AES>(output + 2, &k0, &k1, &k2, &k3, &k4, &k5, &k6, &k7, &k8, &k9);
+
+    xout0 = _mm_load_si128(output + 4);
+    xout1 = _mm_load_si128(output + 5);
+    xout2 = _mm_load_si128(output + 6);
+    xout3 = _mm_load_si128(output + 7);
+    xout4 = _mm_load_si128(output + 8);
+    xout5 = _mm_load_si128(output + 9);
+    xout6 = _mm_load_si128(output + 10);
+    xout7 = _mm_load_si128(output + 11);
+
+    for (size_t i = 0; i < MEM / sizeof(__m128i); i += 8)
+    {
+        xout0 = _mm_xor_si128(_mm_load_si128(input + i + 0), xout0);
+        xout1 = _mm_xor_si128(_mm_load_si128(input + i + 1), xout1);
+        xout2 = _mm_xor_si128(_mm_load_si128(input + i + 2), xout2);
+        xout3 = _mm_xor_si128(_mm_load_si128(input + i + 3), xout3);
+        xout4 = _mm_xor_si128(_mm_load_si128(input + i + 4), xout4);
+        xout5 = _mm_xor_si128(_mm_load_si128(input + i + 5), xout5);
+        xout6 = _mm_xor_si128(_mm_load_si128(input + i + 6), xout6);
+        xout7 = _mm_xor_si128(_mm_load_si128(input + i + 7), xout7);
+
+        aes_round<SOFT_AES>(k0, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k1, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k2, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k3, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k4, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k5, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k6, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k7, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k8, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+        aes_round<SOFT_AES>(k9, &xout0, &xout1, &xout2, &xout3, &xout4, &xout5, &xout6, &xout7);
+    }
+
+    _mm_store_si128(output + 4, xout0);
+    _mm_store_si128(output + 5, xout1);
+    _mm_store_si128(output + 6, xout2);
+    _mm_store_si128(output + 7, xout3);
+    _mm_store_si128(output + 8, xout4);
+    _mm_store_si128(output + 9, xout5);
+    _mm_store_si128(output + 10, xout6);
+    _mm_store_si128(output + 11, xout7);
+}
+
+
+template<size_t ITERATIONS, size_t MEM, size_t MASK, bool SOFT_AES>
+inline void cryptonight_hash(const void *__restrict__ input, size_t size, void *__restrict__ output, cryptonight_ctx *__restrict__ ctx)
+{
+    keccak(static_cast<const uint8_t*>(input), size, ctx->state0, 200);
+
+    cn_explode_scratchpad<MEM, SOFT_AES>((__m128i*) ctx->state0, (__m128i*) ctx->memory);
+
+    const uint8_t* l0 = ctx->memory;
+    uint64_t* h0 = reinterpret_cast<uint64_t*>(ctx->state0);
+
+    uint64_t al0 = h0[0] ^ h0[4];
+    uint64_t ah0 = h0[1] ^ h0[5];
+    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
+
+    uint64_t idx0 = h0[0] ^ h0[4];
+
+    for (size_t i = 0; i < ITERATIONS; i++) {
+        __m128i cx;
+        cx = _mm_load_si128((__m128i *) &l0[idx0 & MASK]);
+
+        if (SOFT_AES) {
+            cx = soft_aesenc(cx, _mm_set_epi64x(ah0, al0));
+        }
+        else {
+            cx = _mm_aesenc_si128(cx, _mm_set_epi64x(ah0, al0));
+        }
+
+        _mm_store_si128((__m128i *) &l0[idx0 & MASK], _mm_xor_si128(bx0, cx));
+        idx0 = EXTRACT64(cx);
+        bx0 = cx;
+
+        uint64_t hi, lo, cl, ch;
+        cl = ((uint64_t*) &l0[idx0 & MASK])[0];
+        ch = ((uint64_t*) &l0[idx0 & MASK])[1];
+        lo = __umul128(idx0, cl, &hi);
+
+        al0 += hi;
+        ah0 += lo;
+
+        ((uint64_t*)&l0[idx0 & MASK])[0] = al0;
+        ((uint64_t*)&l0[idx0 & MASK])[1] = ah0;
+
+        ah0 ^= ch;
+        al0 ^= cl;
+        idx0 = al0;
+    }
+
+    cn_implode_scratchpad<MEM, SOFT_AES>((__m128i*) ctx->memory, (__m128i*) ctx->state0);
+
+    keccakf(h0, 24);
+    extra_hashes[ctx->state0[0] & 3](ctx->state0, 200, static_cast<char*>(output));
+}
+
+
+template<size_t ITERATIONS, size_t MEM, size_t MASK, bool SOFT_AES>
+inline void cryptonight_double_hash(const void *__restrict__ input, size_t size, void *__restrict__ output, struct cryptonight_ctx *__restrict__ ctx)
+{
+    keccak((const uint8_t *) input,        size, ctx->state0, 200);
+    keccak((const uint8_t *) input + size, size, ctx->state1, 200);
+
+    const uint8_t* l0 = ctx->memory;
+    const uint8_t* l1 = ctx->memory + MEM;
+    uint64_t* h0 = reinterpret_cast<uint64_t*>(ctx->state0);
+    uint64_t* h1 = reinterpret_cast<uint64_t*>(ctx->state1);
+
+    cn_explode_scratchpad<MEM, SOFT_AES>((__m128i*) h0, (__m128i*) l0);
+    cn_explode_scratchpad<MEM, SOFT_AES>((__m128i*) h1, (__m128i*) l1);
+
+    uint64_t al0 = h0[0] ^ h0[4];
+    uint64_t al1 = h1[0] ^ h1[4];
+    uint64_t ah0 = h0[1] ^ h0[5];
+    uint64_t ah1 = h1[1] ^ h1[5];
+
+    __m128i bx0 = _mm_set_epi64x(h0[3] ^ h0[7], h0[2] ^ h0[6]);
+    __m128i bx1 = _mm_set_epi64x(h1[3] ^ h1[7], h1[2] ^ h1[6]);
+
+    uint64_t idx0 = h0[0] ^ h0[4];
+    uint64_t idx1 = h1[0] ^ h1[4];
+
+    for (size_t i = 0; i < ITERATIONS; i++) {
+        __m128i cx0 = _mm_load_si128((__m128i *) &l0[idx0 & MASK]);
+        __m128i cx1 = _mm_load_si128((__m128i *) &l1[idx1 & MASK]);
+
+        if (SOFT_AES) {
+            cx0 = soft_aesenc(cx0, _mm_set_epi64x(ah0, al0));
+            cx1 = soft_aesenc(cx1, _mm_set_epi64x(ah1, al1));
+        }
+        else {
+            cx0 = _mm_aesenc_si128(cx0, _mm_set_epi64x(ah0, al0));
+            cx1 = _mm_aesenc_si128(cx1, _mm_set_epi64x(ah1, al1));
+        }
+
+        _mm_store_si128((__m128i *) &l0[idx0 & MASK], _mm_xor_si128(bx0, cx0));
+        _mm_store_si128((__m128i *) &l1[idx1 & MASK], _mm_xor_si128(bx1, cx1));
+
+        idx0 = EXTRACT64(cx0);
+        idx1 = EXTRACT64(cx1);
+
+        bx0 = cx0;
+        bx1 = cx1;
+
+        uint64_t hi, lo, cl, ch;
+        cl = ((uint64_t*) &l0[idx0 & MASK])[0];
+        ch = ((uint64_t*) &l0[idx0 & MASK])[1];
+        lo = __umul128(idx0, cl, &hi);
+
+        al0 += hi;
+        ah0 += lo;
+
+        ((uint64_t*) &l0[idx0 & MASK])[0] = al0;
+        ((uint64_t*) &l0[idx0 & MASK])[1] = ah0;
+
+        ah0 ^= ch;
+        al0 ^= cl;
+        idx0 = al0;
+
+        cl = ((uint64_t*) &l1[idx1 & MASK])[0];
+        ch = ((uint64_t*) &l1[idx1 & MASK])[1];
+        lo = __umul128(idx1, cl, &hi);
+
+        al1 += hi;
+        ah1 += lo;
+
+        ((uint64_t*) &l1[idx1 & MASK])[0] = al1;
+        ((uint64_t*) &l1[idx1 & MASK])[1] = ah1;
+
+        ah1 ^= ch;
+        al1 ^= cl;
+        idx1 = al1;
+    }
+
+    cn_implode_scratchpad<MEM, SOFT_AES>((__m128i*) l0, (__m128i*) h0);
+    cn_implode_scratchpad<MEM, SOFT_AES>((__m128i*) l1, (__m128i*) h1);
+
+    keccakf(h0, 24);
+    keccakf(h1, 24);
+
+    extra_hashes[ctx->state0[0] & 3](ctx->state0, 200, static_cast<char*>(output));
+    extra_hashes[ctx->state1[0] & 3](ctx->state1, 200, static_cast<char*>(output) + 32);
+}
+
+#endif /* __CRYPTONIGHT_P_H__ */

src/crypto/CryptoNight_test.h

@@ -0,0 +1,60 @@
+/* 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 2016-2017 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 __CRYPTONIGHT_TEST_H__
+#define __CRYPTONIGHT_TEST_H__
+
+
+const static uint8_t test_input[152] = {
+    0x01, 0x00, 0xFB, 0x8E, 0x8A, 0xC8, 0x05, 0x89, 0x93, 0x23, 0x37, 0x1B, 0xB7, 0x90, 0xDB, 0x19,
+    0x21, 0x8A, 0xFD, 0x8D, 0xB8, 0xE3, 0x75, 0x5D, 0x8B, 0x90, 0xF3, 0x9B, 0x3D, 0x55, 0x06, 0xA9,
+    0xAB, 0xCE, 0x4F, 0xA9, 0x12, 0x24, 0x45, 0x00, 0x00, 0x00, 0x00, 0xEE, 0x81, 0x46, 0xD4, 0x9F,
+    0xA9, 0x3E, 0xE7, 0x24, 0xDE, 0xB5, 0x7D, 0x12, 0xCB, 0xC6, 0xC6, 0xF3, 0xB9, 0x24, 0xD9, 0x46,
+    0x12, 0x7C, 0x7A, 0x97, 0x41, 0x8F, 0x93, 0x48, 0x82, 0x8F, 0x0F, 0x02,
+    0x03, 0x05, 0xA0, 0xDB, 0xD6, 0xBF, 0x05, 0xCF, 0x16, 0xE5, 0x03, 0xF3, 0xA6, 0x6F, 0x78, 0x00,
+    0x7C, 0xBF, 0x34, 0x14, 0x43, 0x32, 0xEC, 0xBF, 0xC2, 0x2E, 0xD9, 0x5C, 0x87, 0x00, 0x38, 0x3B,
+    0x30, 0x9A, 0xCE, 0x19, 0x23, 0xA0, 0x96, 0x4B, 0x00, 0x00, 0x00, 0x08, 0xBA, 0x93, 0x9A, 0x62,
+    0x72, 0x4C, 0x0D, 0x75, 0x81, 0xFC, 0xE5, 0x76, 0x1E, 0x9D, 0x8A, 0x0E, 0x6A, 0x1C, 0x3F, 0x92,
+    0x4F, 0xDD, 0x84, 0x93, 0xD1, 0x11, 0x56, 0x49, 0xC0, 0x5E, 0xB6, 0x01
+};
+
+
+const static uint8_t test_output0[64] = {
+    0x1B, 0x60, 0x6A, 0x3F, 0x4A, 0x07, 0xD6, 0x48, 0x9A, 0x1B, 0xCD, 0x07, 0x69, 0x7B, 0xD1, 0x66,
+    0x96, 0xB6, 0x1C, 0x8A, 0xE9, 0x82, 0xF6, 0x1A, 0x90, 0x16, 0x0F, 0x4E, 0x52, 0x82, 0x8A, 0x7F,
+    0x1A, 0x3F, 0xFB, 0xEE, 0x90, 0x9B, 0x42, 0x0D, 0x91, 0xF7, 0xBE, 0x6E, 0x5F, 0xB5, 0x6D, 0xB7,
+    0x1B, 0x31, 0x10, 0xD8, 0x86, 0x01, 0x1E, 0x87, 0x7E, 0xE5, 0x78, 0x6A, 0xFD, 0x08, 0x01, 0x00
+};
+
+
+#ifndef XMRIG_NO_AEON
+const static uint8_t test_output1[64] = {
+    0x28, 0xA2, 0x2B, 0xAD, 0x3F, 0x93, 0xD1, 0x40, 0x8F, 0xCA, 0x47, 0x2E, 0xB5, 0xAD, 0x1C, 0xBE,
+    0x75, 0xF2, 0x1D, 0x05, 0x3C, 0x8C, 0xE5, 0xB3, 0xAF, 0x10, 0x5A, 0x57, 0x71, 0x3E, 0x21, 0xDD,
+    0x36, 0x95, 0xB4, 0xB5, 0x3B, 0xB0, 0x03, 0x58, 0xB0, 0xAD, 0x38, 0xDC, 0x16, 0x0F, 0xEB, 0x9E,
+    0x00, 0x4E, 0xEC, 0xE0, 0x9B, 0x83, 0xA7, 0x2E, 0xF6, 0xBA, 0x98, 0x64, 0xD3, 0x51, 0x0C, 0x88,
+};
+#endif
+
+
+#endif /* __CRYPTONIGHT_TEST_H__ */

src/crypto/c_keccak.c

@@ -0,0 +1,176 @@
+// keccak.c
+// 19-Nov-11  Markku-Juhani O. Saarinen <mjos@iki.fi>
+// A baseline Keccak (3rd round) implementation.
+
+#include <stdint.h>
+#include <memory.h>
+
+#define HASH_DATA_AREA 136
+#define KECCAK_ROUNDS 24
+
+#ifndef ROTL64
+#define ROTL64(x, y) (((x) << (y)) | ((x) >> (64 - (y))))
+#endif
+
+const uint64_t keccakf_rndc[24] = 
+{
+    0x0000000000000001, 0x0000000000008082, 0x800000000000808a,
+    0x8000000080008000, 0x000000000000808b, 0x0000000080000001,
+    0x8000000080008081, 0x8000000000008009, 0x000000000000008a,
+    0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
+    0x000000008000808b, 0x800000000000008b, 0x8000000000008089,
+    0x8000000000008003, 0x8000000000008002, 0x8000000000000080, 
+    0x000000000000800a, 0x800000008000000a, 0x8000000080008081,
+    0x8000000000008080, 0x0000000080000001, 0x8000000080008008
+};
+
+// update the state with given number of rounds
+
+void keccakf(uint64_t st[25], int rounds)
+{
+    int i, j, round;
+    uint64_t t, bc[5];
+
+    for (round = 0; round < rounds; ++round) {
+
+        // Theta
+        bc[0] = st[0] ^ st[5] ^ st[10] ^ st[15] ^ st[20];
+        bc[1] = st[1] ^ st[6] ^ st[11] ^ st[16] ^ st[21];
+        bc[2] = st[2] ^ st[7] ^ st[12] ^ st[17] ^ st[22];
+        bc[3] = st[3] ^ st[8] ^ st[13] ^ st[18] ^ st[23];
+        bc[4] = st[4] ^ st[9] ^ st[14] ^ st[19] ^ st[24];
+
+        for (i = 0; i < 5; ++i) {
+            t = bc[(i + 4) % 5] ^ ROTL64(bc[(i + 1) % 5], 1);
+            st[i     ] ^= t;
+            st[i +  5] ^= t;
+            st[i + 10] ^= t;
+            st[i + 15] ^= t;
+            st[i + 20] ^= t;
+        }
+
+        // Rho Pi
+        t = st[1];
+        st[ 1] = ROTL64(st[ 6], 44);
+        st[ 6] = ROTL64(st[ 9], 20);
+        st[ 9] = ROTL64(st[22], 61);
+        st[22] = ROTL64(st[14], 39);
+        st[14] = ROTL64(st[20], 18);
+        st[20] = ROTL64(st[ 2], 62);
+        st[ 2] = ROTL64(st[12], 43);
+        st[12] = ROTL64(st[13], 25);
+        st[13] = ROTL64(st[19],  8);
+        st[19] = ROTL64(st[23], 56);
+        st[23] = ROTL64(st[15], 41);
+        st[15] = ROTL64(st[ 4], 27);
+        st[ 4] = ROTL64(st[24], 14);
+        st[24] = ROTL64(st[21],  2);
+        st[21] = ROTL64(st[ 8], 55);
+        st[ 8] = ROTL64(st[16], 45);
+        st[16] = ROTL64(st[ 5], 36);
+        st[ 5] = ROTL64(st[ 3], 28);
+        st[ 3] = ROTL64(st[18], 21);
+        st[18] = ROTL64(st[17], 15);
+        st[17] = ROTL64(st[11], 10);
+        st[11] = ROTL64(st[ 7],  6);
+        st[ 7] = ROTL64(st[10],  3);
+        st[10] = ROTL64(t, 1);
+
+        //  Chi
+        // unrolled loop, where only last iteration is different
+        j = 0;
+        bc[0] = st[j    ];
+        bc[1] = st[j + 1];
+
+        st[j    ] ^= (~st[j + 1]) & st[j + 2];
+        st[j + 1] ^= (~st[j + 2]) & st[j + 3];
+        st[j + 2] ^= (~st[j + 3]) & st[j + 4];
+        st[j + 3] ^= (~st[j + 4]) & bc[0];
+        st[j + 4] ^= (~bc[0]) & bc[1];
+
+        j = 5;
+        bc[0] = st[j    ];
+        bc[1] = st[j + 1];
+
+        st[j    ] ^= (~st[j + 1]) & st[j + 2];
+        st[j + 1] ^= (~st[j + 2]) & st[j + 3];
+        st[j + 2] ^= (~st[j + 3]) & st[j + 4];
+        st[j + 3] ^= (~st[j + 4]) & bc[0];
+        st[j + 4] ^= (~bc[0]) & bc[1];
+
+        j = 10;
+        bc[0] = st[j    ];
+        bc[1] = st[j + 1];
+
+        st[j    ] ^= (~st[j + 1]) & st[j + 2];
+        st[j + 1] ^= (~st[j + 2]) & st[j + 3];
+        st[j + 2] ^= (~st[j + 3]) & st[j + 4];
+        st[j + 3] ^= (~st[j + 4]) & bc[0];
+        st[j + 4] ^= (~bc[0]) & bc[1];
+
+        j = 15;
+        bc[0] = st[j    ];
+        bc[1] = st[j + 1];
+
+        st[j    ] ^= (~st[j + 1]) & st[j + 2];
+        st[j + 1] ^= (~st[j + 2]) & st[j + 3];
+        st[j + 2] ^= (~st[j + 3]) & st[j + 4];
+        st[j + 3] ^= (~st[j + 4]) & bc[0];
+        st[j + 4] ^= (~bc[0]) & bc[1];
+
+        j = 20;
+        bc[0] = st[j    ];
+        bc[1] = st[j + 1];
+        bc[2] = st[j + 2];
+        bc[3] = st[j + 3];
+        bc[4] = st[j + 4];
+
+        st[j    ] ^= (~bc[1]) & bc[2];
+        st[j + 1] ^= (~bc[2]) & bc[3];
+        st[j + 2] ^= (~bc[3]) & bc[4];
+        st[j + 3] ^= (~bc[4]) & bc[0];
+        st[j + 4] ^= (~bc[0]) & bc[1];
+        
+        //  Iota
+        st[0] ^= keccakf_rndc[round];
+    }
+}
+
+// compute a keccak hash (md) of given byte length from "in"
+typedef uint64_t state_t[25];
+
+void keccak(const uint8_t *in, int inlen, uint8_t *md, int mdlen)
+{
+    state_t st;
+    uint8_t temp[144];
+    int i, rsiz, rsizw;
+
+    rsiz = sizeof(state_t) == mdlen ? HASH_DATA_AREA : 200 - 2 * mdlen;
+    rsizw = rsiz / 8;
+    
+    memset(st, 0, sizeof(st));
+
+    for ( ; inlen >= rsiz; inlen -= rsiz, in += rsiz) {
+        for (i = 0; i < rsizw; i++)
+            st[i] ^= ((uint64_t *) in)[i];
+        keccakf(st, KECCAK_ROUNDS);
+    }
+    
+    // last block and padding
+    memcpy(temp, in, inlen);
+    temp[inlen++] = 1;
+    memset(temp + inlen, 0, rsiz - inlen);
+    temp[rsiz - 1] |= 0x80;
+
+    for (i = 0; i < rsizw; i++)
+        st[i] ^= ((uint64_t *) temp)[i];
+
+    keccakf(st, KECCAK_ROUNDS);
+
+    memcpy(md, st, mdlen);
+}
+
+void keccak1600(const uint8_t *in, int inlen, uint8_t *md)
+{
+    keccak(in, inlen, md, sizeof(state_t));
+}

src/crypto/c_skein.c

@@ -0,0 +1,701 @@
+/***********************************************************************
+**
+** Implementation of the Skein hash function.
+**
+** Source code author: Doug Whiting, 2008.
+**
+** This algorithm and source code is released to the public domain.
+** 
+************************************************************************/
+
+#define  SKEIN_PORT_CODE /* instantiate any code in skein_port.h */
+
+#include <stddef.h>                          /* get size_t definition */
+#include <string.h>      /* get the memcpy/memset functions */
+#include "c_skein.h"       /* get the Skein API definitions   */
+
+#ifndef SKEIN_512_NIST_MAX_HASHBITS
+#define SKEIN_512_NIST_MAX_HASHBITS (512)
+#endif
+
+#define  SKEIN_MODIFIER_WORDS  ( 2)          /* number of modifier (tweak) words */
+
+#define  SKEIN_512_STATE_WORDS ( 8)
+#define  SKEIN_MAX_STATE_WORDS (16)
+
+#define  SKEIN_512_STATE_BYTES ( 8*SKEIN_512_STATE_WORDS)
+#define  SKEIN_512_STATE_BITS  (64*SKEIN_512_STATE_WORDS)
+#define  SKEIN_512_BLOCK_BYTES ( 8*SKEIN_512_STATE_WORDS)
+
+#define SKEIN_RND_SPECIAL       (1000u)
+#define SKEIN_RND_KEY_INITIAL   (SKEIN_RND_SPECIAL+0u)
+#define SKEIN_RND_KEY_INJECT    (SKEIN_RND_SPECIAL+1u)
+#define SKEIN_RND_FEED_FWD      (SKEIN_RND_SPECIAL+2u)
+
+typedef struct
+{
+  size_t  hashBitLen;                      /* size of hash result, in bits */
+  size_t  bCnt;                            /* current byte count in buffer b[] */
+  u64b_t  T[SKEIN_MODIFIER_WORDS];         /* tweak words: T[0]=byte cnt, T[1]=flags */
+} Skein_Ctxt_Hdr_t;
+
+typedef struct                               /*  512-bit Skein hash context structure */
+{
+  Skein_Ctxt_Hdr_t h;                      /* common header context variables */
+  u64b_t  X[SKEIN_512_STATE_WORDS];        /* chaining variables */
+  u08b_t  b[SKEIN_512_BLOCK_BYTES];        /* partial block buffer (8-byte aligned) */
+} Skein_512_Ctxt_t;
+
+/*   Skein APIs for (incremental) "straight hashing" */
+static int  Skein_512_Init  (Skein_512_Ctxt_t *ctx, size_t hashBitLen);
+static int  Skein_512_Update(Skein_512_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt);
+static int  Skein_512_Final (Skein_512_Ctxt_t *ctx, u08b_t * hashVal);
+
+#ifndef SKEIN_TREE_HASH
+#define SKEIN_TREE_HASH (1)
+#endif
+
+/*****************************************************************
+** "Internal" Skein definitions
+**    -- not needed for sequential hashing API, but will be 
+**           helpful for other uses of Skein (e.g., tree hash mode).
+**    -- included here so that they can be shared between
+**           reference and optimized code.
+******************************************************************/
+
+/* tweak word T[1]: bit field starting positions */
+#define SKEIN_T1_BIT(BIT)       ((BIT) - 64)            /* offset 64 because it's the second word  */
+
+#define SKEIN_T1_POS_TREE_LVL   SKEIN_T1_BIT(112)       /* bits 112..118: level in hash tree       */
+#define SKEIN_T1_POS_BIT_PAD    SKEIN_T1_BIT(119)       /* bit  119     : partial final input byte */
+#define SKEIN_T1_POS_BLK_TYPE   SKEIN_T1_BIT(120)       /* bits 120..125: type field               */
+#define SKEIN_T1_POS_FIRST      SKEIN_T1_BIT(126)       /* bits 126     : first block flag         */
+#define SKEIN_T1_POS_FINAL      SKEIN_T1_BIT(127)       /* bit  127     : final block flag         */
+
+/* tweak word T[1]: flag bit definition(s) */
+#define SKEIN_T1_FLAG_FIRST     (((u64b_t)  1 ) << SKEIN_T1_POS_FIRST)
+#define SKEIN_T1_FLAG_FINAL     (((u64b_t)  1 ) << SKEIN_T1_POS_FINAL)
+#define SKEIN_T1_FLAG_BIT_PAD   (((u64b_t)  1 ) << SKEIN_T1_POS_BIT_PAD)
+
+/* tweak word T[1]: tree level bit field mask */
+#define SKEIN_T1_TREE_LVL_MASK  (((u64b_t)0x7F) << SKEIN_T1_POS_TREE_LVL)
+#define SKEIN_T1_TREE_LEVEL(n)  (((u64b_t) (n)) << SKEIN_T1_POS_TREE_LVL)
+
+/* tweak word T[1]: block type field */
+#define SKEIN_BLK_TYPE_KEY      ( 0)                    /* key, for MAC and KDF */
+#define SKEIN_BLK_TYPE_CFG      ( 4)                    /* configuration block */
+#define SKEIN_BLK_TYPE_PERS     ( 8)                    /* personalization string */
+#define SKEIN_BLK_TYPE_PK       (12)                    /* public key (for digital signature hashing) */
+#define SKEIN_BLK_TYPE_KDF      (16)                    /* key identifier for KDF */
+#define SKEIN_BLK_TYPE_NONCE    (20)                    /* nonce for PRNG */
+#define SKEIN_BLK_TYPE_MSG      (48)                    /* message processing */
+#define SKEIN_BLK_TYPE_OUT      (63)                    /* output stage */
+#define SKEIN_BLK_TYPE_MASK     (63)                    /* bit field mask */
+
+#define SKEIN_T1_BLK_TYPE(T)   (((u64b_t) (SKEIN_BLK_TYPE_##T)) << SKEIN_T1_POS_BLK_TYPE)
+#define SKEIN_T1_BLK_TYPE_KEY   SKEIN_T1_BLK_TYPE(KEY)  /* key, for MAC and KDF */
+#define SKEIN_T1_BLK_TYPE_CFG   SKEIN_T1_BLK_TYPE(CFG)  /* configuration block */
+#define SKEIN_T1_BLK_TYPE_PERS  SKEIN_T1_BLK_TYPE(PERS) /* personalization string */
+#define SKEIN_T1_BLK_TYPE_PK    SKEIN_T1_BLK_TYPE(PK)   /* public key (for digital signature hashing) */
+#define SKEIN_T1_BLK_TYPE_KDF   SKEIN_T1_BLK_TYPE(KDF)  /* key identifier for KDF */
+#define SKEIN_T1_BLK_TYPE_NONCE SKEIN_T1_BLK_TYPE(NONCE)/* nonce for PRNG */
+#define SKEIN_T1_BLK_TYPE_MSG   SKEIN_T1_BLK_TYPE(MSG)  /* message processing */
+#define SKEIN_T1_BLK_TYPE_OUT   SKEIN_T1_BLK_TYPE(OUT)  /* output stage */
+#define SKEIN_T1_BLK_TYPE_MASK  SKEIN_T1_BLK_TYPE(MASK) /* field bit mask */
+
+#define SKEIN_T1_BLK_TYPE_CFG_FINAL       (SKEIN_T1_BLK_TYPE_CFG | SKEIN_T1_FLAG_FINAL)
+#define SKEIN_T1_BLK_TYPE_OUT_FINAL       (SKEIN_T1_BLK_TYPE_OUT | SKEIN_T1_FLAG_FINAL)
+
+#define SKEIN_VERSION           (1)
+
+#ifndef SKEIN_ID_STRING_LE      /* allow compile-time personalization */
+#define SKEIN_ID_STRING_LE      (0x33414853)            /* "SHA3" (little-endian)*/
+#endif
+
+#define SKEIN_MK_64(hi32,lo32)  ((lo32) + (((u64b_t) (hi32)) << 32))
+#define SKEIN_SCHEMA_VER        SKEIN_MK_64(SKEIN_VERSION,SKEIN_ID_STRING_LE)
+#define SKEIN_KS_PARITY         SKEIN_MK_64(0x1BD11BDA,0xA9FC1A22)
+
+#define SKEIN_CFG_STR_LEN       (4*8)
+
+/* bit field definitions in config block treeInfo word */
+#define SKEIN_CFG_TREE_LEAF_SIZE_POS  ( 0)
+#define SKEIN_CFG_TREE_NODE_SIZE_POS  ( 8)
+#define SKEIN_CFG_TREE_MAX_LEVEL_POS  (16)
+
+#define SKEIN_CFG_TREE_LEAF_SIZE_MSK  (((u64b_t) 0xFF) << SKEIN_CFG_TREE_LEAF_SIZE_POS)
+#define SKEIN_CFG_TREE_NODE_SIZE_MSK  (((u64b_t) 0xFF) << SKEIN_CFG_TREE_NODE_SIZE_POS)
+#define SKEIN_CFG_TREE_MAX_LEVEL_MSK  (((u64b_t) 0xFF) << SKEIN_CFG_TREE_MAX_LEVEL_POS)
+
+#define SKEIN_CFG_TREE_INFO(leaf,node,maxLvl)                   \
+  ( (((u64b_t)(leaf  )) << SKEIN_CFG_TREE_LEAF_SIZE_POS) |    \
+  (((u64b_t)(node  )) << SKEIN_CFG_TREE_NODE_SIZE_POS) |    \
+  (((u64b_t)(maxLvl)) << SKEIN_CFG_TREE_MAX_LEVEL_POS) )
+
+#define SKEIN_CFG_TREE_INFO_SEQUENTIAL SKEIN_CFG_TREE_INFO(0,0,0) /* use as treeInfo in InitExt() call for sequential processing */
+
+/*
+**   Skein macros for getting/setting tweak words, etc.
+**   These are useful for partial input bytes, hash tree init/update, etc.
+**/
+#define Skein_Get_Tweak(ctxPtr,TWK_NUM)         ((ctxPtr)->h.T[TWK_NUM])
+#define Skein_Set_Tweak(ctxPtr,TWK_NUM,tVal)    {(ctxPtr)->h.T[TWK_NUM] = (tVal);}
+
+#define Skein_Get_T0(ctxPtr)    Skein_Get_Tweak(ctxPtr,0)
+#define Skein_Get_T1(ctxPtr)    Skein_Get_Tweak(ctxPtr,1)
+#define Skein_Set_T0(ctxPtr,T0) Skein_Set_Tweak(ctxPtr,0,T0)
+#define Skein_Set_T1(ctxPtr,T1) Skein_Set_Tweak(ctxPtr,1,T1)
+
+/* set both tweak words at once */
+#define Skein_Set_T0_T1(ctxPtr,T0,T1)           \
+{                                           \
+  Skein_Set_T0(ctxPtr,(T0));                  \
+  Skein_Set_T1(ctxPtr,(T1));                  \
+}
+
+#define Skein_Set_Type(ctxPtr,BLK_TYPE)         \
+  Skein_Set_T1(ctxPtr,SKEIN_T1_BLK_TYPE_##BLK_TYPE)
+
+/* set up for starting with a new type: h.T[0]=0; h.T[1] = NEW_TYPE; h.bCnt=0; */
+#define Skein_Start_New_Type(ctxPtr,BLK_TYPE)   \
+{ Skein_Set_T0_T1(ctxPtr,0,SKEIN_T1_FLAG_FIRST | SKEIN_T1_BLK_TYPE_##BLK_TYPE); (ctxPtr)->h.bCnt=0; }
+
+#define Skein_Clear_First_Flag(hdr)      { (hdr).T[1] &= ~SKEIN_T1_FLAG_FIRST;       }
+#define Skein_Set_Bit_Pad_Flag(hdr)      { (hdr).T[1] |=  SKEIN_T1_FLAG_BIT_PAD;     }
+
+#define Skein_Set_Tree_Level(hdr,height) { (hdr).T[1] |= SKEIN_T1_TREE_LEVEL(height);}
+
+/*****************************************************************
+** "Internal" Skein definitions for debugging and error checking
+******************************************************************/
+#define Skein_Show_Block(bits,ctx,X,blkPtr,wPtr,ksEvenPtr,ksOddPtr)
+#define Skein_Show_Round(bits,ctx,r,X)
+#define Skein_Show_R_Ptr(bits,ctx,r,X_ptr)
+#define Skein_Show_Final(bits,ctx,cnt,outPtr)
+#define Skein_Show_Key(bits,ctx,key,keyBytes)
+
+
+#ifndef SKEIN_ERR_CHECK        /* run-time checks (e.g., bad params, uninitialized context)? */
+#define Skein_Assert(x,retCode)/* default: ignore all Asserts, for performance */
+#define Skein_assert(x)
+#elif   defined(SKEIN_ASSERT)
+#include <assert.h>     
+#define Skein_Assert(x,retCode) assert(x) 
+#define Skein_assert(x)         assert(x) 
+#else
+#include <assert.h>     
+#define Skein_Assert(x,retCode) { if (!(x)) return retCode; } /*  caller  error */
+#define Skein_assert(x)         assert(x)                     /* internal error */
+#endif
+
+/*****************************************************************
+** Skein block function constants (shared across Ref and Opt code)
+******************************************************************/
+enum    
+{   
+  /* Skein_512 round rotation constants */
+  R_512_0_0=46, R_512_0_1=36, R_512_0_2=19, R_512_0_3=37,
+  R_512_1_0=33, R_512_1_1=27, R_512_1_2=14, R_512_1_3=42,
+  R_512_2_0=17, R_512_2_1=49, R_512_2_2=36, R_512_2_3=39,
+  R_512_3_0=44, R_512_3_1= 9, R_512_3_2=54, R_512_3_3=56,
+  R_512_4_0=39, R_512_4_1=30, R_512_4_2=34, R_512_4_3=24,
+  R_512_5_0=13, R_512_5_1=50, R_512_5_2=10, R_512_5_3=17,
+  R_512_6_0=25, R_512_6_1=29, R_512_6_2=39, R_512_6_3=43,
+  R_512_7_0= 8, R_512_7_1=35, R_512_7_2=56, R_512_7_3=22,
+};
+
+#ifndef SKEIN_ROUNDS
+#define SKEIN_512_ROUNDS_TOTAL (72)
+#else                                        /* allow command-line define in range 8*(5..14)   */
+#define SKEIN_512_ROUNDS_TOTAL (8*((((SKEIN_ROUNDS/ 10) + 5) % 10) + 5))
+#endif
+
+
+/*
+***************** Pre-computed Skein IVs *******************
+**
+** NOTE: these values are not "magic" constants, but
+** are generated using the Threefish block function.
+** They are pre-computed here only for speed; i.e., to
+** avoid the need for a Threefish call during Init().
+**
+** The IV for any fixed hash length may be pre-computed.
+** Only the most common values are included here.
+**
+************************************************************
+**/
+
+#define MK_64 SKEIN_MK_64
+
+/* blkSize =  512 bits. hashSize =  256 bits */
+const u64b_t SKEIN_512_IV_256[] =
+    {
+    MK_64(0xCCD044A1,0x2FDB3E13),
+    MK_64(0xE8359030,0x1A79A9EB),
+    MK_64(0x55AEA061,0x4F816E6F),
+    MK_64(0x2A2767A4,0xAE9B94DB),
+    MK_64(0xEC06025E,0x74DD7683),
+    MK_64(0xE7A436CD,0xC4746251),
+    MK_64(0xC36FBAF9,0x393AD185),
+    MK_64(0x3EEDBA18,0x33EDFC13)
+    };
+
+#ifndef SKEIN_USE_ASM
+#define SKEIN_USE_ASM   (0)                     /* default is all C code (no ASM) */
+#endif
+
+#ifndef SKEIN_LOOP
+#define SKEIN_LOOP 001                          /* default: unroll 256 and 512, but not 1024 */
+#endif
+
+#define BLK_BITS        (WCNT*64)               /* some useful definitions for code here */
+#define KW_TWK_BASE     (0)
+#define KW_KEY_BASE     (3)
+#define ks              (kw + KW_KEY_BASE)                
+#define ts              (kw + KW_TWK_BASE)
+
+#ifdef SKEIN_DEBUG
+#define DebugSaveTweak(ctx) { ctx->h.T[0] = ts[0]; ctx->h.T[1] = ts[1]; }
+#else
+#define DebugSaveTweak(ctx)
+#endif
+
+/*****************************  Skein_512 ******************************/
+#if !(SKEIN_USE_ASM & 512)
+static void Skein_512_Process_Block(Skein_512_Ctxt_t *ctx,const u08b_t *blkPtr,size_t blkCnt,size_t byteCntAdd)
+    { /* do it in C */
+    enum
+        {
+        WCNT = SKEIN_512_STATE_WORDS
+        };
+#undef  RCNT
+#define RCNT  (SKEIN_512_ROUNDS_TOTAL/8)
+
+#ifdef  SKEIN_LOOP                              /* configure how much to unroll the loop */
+#define SKEIN_UNROLL_512 (((SKEIN_LOOP)/10)%10)
+#else
+#define SKEIN_UNROLL_512 (0)
+#endif
+
+#if SKEIN_UNROLL_512
+#if (RCNT % SKEIN_UNROLL_512)
+#error "Invalid SKEIN_UNROLL_512"               /* sanity check on unroll count */
+#endif
+    size_t  r;
+    u64b_t  kw[WCNT+4+RCNT*2];                  /* key schedule words : chaining vars + tweak + "rotation"*/
+#else
+    u64b_t  kw[WCNT+4];                         /* key schedule words : chaining vars + tweak */
+#endif
+    u64b_t  X0,X1,X2,X3,X4,X5,X6,X7;            /* local copy of vars, for speed */
+    u64b_t  w [WCNT];                           /* local copy of input block */
+#ifdef SKEIN_DEBUG
+    const u64b_t *Xptr[8];                      /* use for debugging (help compiler put Xn in registers) */
+    Xptr[0] = &X0;  Xptr[1] = &X1;  Xptr[2] = &X2;  Xptr[3] = &X3;
+    Xptr[4] = &X4;  Xptr[5] = &X5;  Xptr[6] = &X6;  Xptr[7] = &X7;
+#endif
+
+    Skein_assert(blkCnt != 0);                  /* never call with blkCnt == 0! */
+    ts[0] = ctx->h.T[0];
+    ts[1] = ctx->h.T[1];
+    do  {
+        /* this implementation only supports 2**64 input bytes (no carry out here) */
+        ts[0] += byteCntAdd;                    /* update processed length */
+
+        /* precompute the key schedule for this block */
+        ks[0] = ctx->X[0];
+        ks[1] = ctx->X[1];
+        ks[2] = ctx->X[2];
+        ks[3] = ctx->X[3];
+        ks[4] = ctx->X[4];
+        ks[5] = ctx->X[5];
+        ks[6] = ctx->X[6];
+        ks[7] = ctx->X[7];
+        ks[8] = ks[0] ^ ks[1] ^ ks[2] ^ ks[3] ^ 
+                ks[4] ^ ks[5] ^ ks[6] ^ ks[7] ^ SKEIN_KS_PARITY;
+
+        ts[2] = ts[0] ^ ts[1];
+
+        Skein_Get64_LSB_First(w,blkPtr,WCNT); /* get input block in little-endian format */
+        DebugSaveTweak(ctx);
+        Skein_Show_Block(BLK_BITS,&ctx->h,ctx->X,blkPtr,w,ks,ts);
+
+        X0   = w[0] + ks[0];                    /* do the first full key injection */
+        X1   = w[1] + ks[1];
+        X2   = w[2] + ks[2];
+        X3   = w[3] + ks[3];
+        X4   = w[4] + ks[4];
+        X5   = w[5] + ks[5] + ts[0];
+        X6   = w[6] + ks[6] + ts[1];
+        X7   = w[7] + ks[7];
+
+        blkPtr += SKEIN_512_BLOCK_BYTES;
+
+        Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INITIAL,Xptr);
+        /* run the rounds */
+#define Round512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                  \
+    X##p0 += X##p1; X##p1 = RotL_64(X##p1,ROT##_0); X##p1 ^= X##p0; \
+    X##p2 += X##p3; X##p3 = RotL_64(X##p3,ROT##_1); X##p3 ^= X##p2; \
+    X##p4 += X##p5; X##p5 = RotL_64(X##p5,ROT##_2); X##p5 ^= X##p4; \
+    X##p6 += X##p7; X##p7 = RotL_64(X##p7,ROT##_3); X##p7 ^= X##p6; \
+
+#if SKEIN_UNROLL_512 == 0                       
+#define R512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)      /* unrolled */  \
+    Round512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                      \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,rNum,Xptr);
+
+#define I512(R)                                                     \
+    X0   += ks[((R)+1) % 9];   /* inject the key schedule value */  \
+    X1   += ks[((R)+2) % 9];                                        \
+    X2   += ks[((R)+3) % 9];                                        \
+    X3   += ks[((R)+4) % 9];                                        \
+    X4   += ks[((R)+5) % 9];                                        \
+    X5   += ks[((R)+6) % 9] + ts[((R)+1) % 3];                      \
+    X6   += ks[((R)+7) % 9] + ts[((R)+2) % 3];                      \
+    X7   += ks[((R)+8) % 9] +     (R)+1;                            \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+#else                                       /* looping version */
+#define R512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                      \
+    Round512(p0,p1,p2,p3,p4,p5,p6,p7,ROT,rNum)                      \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,4*(r-1)+rNum,Xptr);
+
+#define I512(R)                                                     \
+    X0   += ks[r+(R)+0];        /* inject the key schedule value */ \
+    X1   += ks[r+(R)+1];                                            \
+    X2   += ks[r+(R)+2];                                            \
+    X3   += ks[r+(R)+3];                                            \
+    X4   += ks[r+(R)+4];                                            \
+    X5   += ks[r+(R)+5] + ts[r+(R)+0];                              \
+    X6   += ks[r+(R)+6] + ts[r+(R)+1];                              \
+    X7   += ks[r+(R)+7] +    r+(R)   ;                              \
+    ks[r +       (R)+8] = ks[r+(R)-1];  /* rotate key schedule */   \
+    ts[r +       (R)+2] = ts[r+(R)-1];                              \
+    Skein_Show_R_Ptr(BLK_BITS,&ctx->h,SKEIN_RND_KEY_INJECT,Xptr);
+
+    for (r=1;r < 2*RCNT;r+=2*SKEIN_UNROLL_512)   /* loop thru it */
+#endif                         /* end of looped code definitions */
+        {
+#define R512_8_rounds(R)  /* do 8 full rounds */  \
+        R512(0,1,2,3,4,5,6,7,R_512_0,8*(R)+ 1);   \
+        R512(2,1,4,7,6,5,0,3,R_512_1,8*(R)+ 2);   \
+        R512(4,1,6,3,0,5,2,7,R_512_2,8*(R)+ 3);   \
+        R512(6,1,0,7,2,5,4,3,R_512_3,8*(R)+ 4);   \
+        I512(2*(R));                              \
+        R512(0,1,2,3,4,5,6,7,R_512_4,8*(R)+ 5);   \
+        R512(2,1,4,7,6,5,0,3,R_512_5,8*(R)+ 6);   \
+        R512(4,1,6,3,0,5,2,7,R_512_6,8*(R)+ 7);   \
+        R512(6,1,0,7,2,5,4,3,R_512_7,8*(R)+ 8);   \
+        I512(2*(R)+1);        /* and key injection */
+
+        R512_8_rounds( 0);
+
+#define R512_Unroll_R(NN) ((SKEIN_UNROLL_512 == 0 && SKEIN_512_ROUNDS_TOTAL/8 > (NN)) || (SKEIN_UNROLL_512 > (NN)))
+
+  #if   R512_Unroll_R( 1)
+        R512_8_rounds( 1);
+  #endif
+  #if   R512_Unroll_R( 2)
+        R512_8_rounds( 2);
+  #endif
+  #if   R512_Unroll_R( 3)
+        R512_8_rounds( 3);
+  #endif
+  #if   R512_Unroll_R( 4)
+        R512_8_rounds( 4);
+  #endif
+  #if   R512_Unroll_R( 5)
+        R512_8_rounds( 5);
+  #endif
+  #if   R512_Unroll_R( 6)
+        R512_8_rounds( 6);
+  #endif
+  #if   R512_Unroll_R( 7)
+        R512_8_rounds( 7);
+  #endif
+  #if   R512_Unroll_R( 8)
+        R512_8_rounds( 8);
+  #endif
+  #if   R512_Unroll_R( 9)
+        R512_8_rounds( 9);
+  #endif
+  #if   R512_Unroll_R(10)
+        R512_8_rounds(10);
+  #endif
+  #if   R512_Unroll_R(11)
+        R512_8_rounds(11);
+  #endif
+  #if   R512_Unroll_R(12)
+        R512_8_rounds(12);
+  #endif
+  #if   R512_Unroll_R(13)
+        R512_8_rounds(13);
+  #endif
+  #if   R512_Unroll_R(14)
+        R512_8_rounds(14);
+  #endif
+  #if  (SKEIN_UNROLL_512 > 14)
+#error  "need more unrolling in Skein_512_Process_Block"
+  #endif
+        }
+
+        /* do the final "feedforward" xor, update context chaining vars */
+        ctx->X[0] = X0 ^ w[0];
+        ctx->X[1] = X1 ^ w[1];
+        ctx->X[2] = X2 ^ w[2];
+        ctx->X[3] = X3 ^ w[3];
+        ctx->X[4] = X4 ^ w[4];
+        ctx->X[5] = X5 ^ w[5];
+        ctx->X[6] = X6 ^ w[6];
+        ctx->X[7] = X7 ^ w[7];
+        Skein_Show_Round(BLK_BITS,&ctx->h,SKEIN_RND_FEED_FWD,ctx->X);
+
+        ts[1] &= ~SKEIN_T1_FLAG_FIRST;
+        }
+    while (--blkCnt);
+    ctx->h.T[0] = ts[0];
+    ctx->h.T[1] = ts[1];
+    }
+#endif
+
+/*****************************************************************/
+/*     512-bit Skein                                             */
+/*****************************************************************/
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* init the context for a straight hashing operation  */
+static int Skein_512_Init(Skein_512_Ctxt_t *ctx, size_t hashBitLen)
+    {
+    union
+        {
+        u08b_t  b[SKEIN_512_STATE_BYTES];
+        u64b_t  w[SKEIN_512_STATE_WORDS];
+        } cfg;                              /* config block */
+        
+    Skein_Assert(hashBitLen > 0,SKEIN_BAD_HASHLEN);
+    ctx->h.hashBitLen = hashBitLen;         /* output hash bit count */
+
+    switch (hashBitLen)
+        {             /* use pre-computed values, where available */
+#ifndef SKEIN_NO_PRECOMP
+        case  256: memcpy(ctx->X,SKEIN_512_IV_256,sizeof(ctx->X));  break;
+#endif
+        default:
+            /* here if there is no precomputed IV value available */
+            /* build/process the config block, type == CONFIG (could be precomputed) */
+            Skein_Start_New_Type(ctx,CFG_FINAL);        /* set tweaks: T0=0; T1=CFG | FINAL */
+
+            cfg.w[0] = Skein_Swap64(SKEIN_SCHEMA_VER);  /* set the schema, version */
+            cfg.w[1] = Skein_Swap64(hashBitLen);        /* hash result length in bits */
+            cfg.w[2] = Skein_Swap64(SKEIN_CFG_TREE_INFO_SEQUENTIAL);
+            memset(&cfg.w[3],0,sizeof(cfg) - 3*sizeof(cfg.w[0])); /* zero pad config block */
+
+            /* compute the initial chaining values from config block */
+            memset(ctx->X,0,sizeof(ctx->X));            /* zero the chaining variables */
+            Skein_512_Process_Block(ctx,cfg.b,1,SKEIN_CFG_STR_LEN);
+            break;
+        }
+
+    /* The chaining vars ctx->X are now initialized for the given hashBitLen. */
+    /* Set up to process the data message portion of the hash (default) */
+    Skein_Start_New_Type(ctx,MSG);              /* T0=0, T1= MSG type */
+
+    return SKEIN_SUCCESS;
+    }
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* process the input bytes */
+static int Skein_512_Update(Skein_512_Ctxt_t *ctx, const u08b_t *msg, size_t msgByteCnt)
+    {
+    size_t n;
+
+    Skein_Assert(ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    /* process full blocks, if any */
+    if (msgByteCnt + ctx->h.bCnt > SKEIN_512_BLOCK_BYTES)
+        {
+        if (ctx->h.bCnt)                              /* finish up any buffered message data */
+            {
+            n = SKEIN_512_BLOCK_BYTES - ctx->h.bCnt;  /* # bytes free in buffer b[] */
+            if (n)
+                {
+                Skein_assert(n < msgByteCnt);         /* check on our logic here */
+                memcpy(&ctx->b[ctx->h.bCnt],msg,n);
+                msgByteCnt  -= n;
+                msg         += n;
+                ctx->h.bCnt += n;
+                }
+            Skein_assert(ctx->h.bCnt == SKEIN_512_BLOCK_BYTES);
+            Skein_512_Process_Block(ctx,ctx->b,1,SKEIN_512_BLOCK_BYTES);
+            ctx->h.bCnt = 0;
+            }
+        /* now process any remaining full blocks, directly from input message data */
+        if (msgByteCnt > SKEIN_512_BLOCK_BYTES)
+            {
+            n = (msgByteCnt-1) / SKEIN_512_BLOCK_BYTES;   /* number of full blocks to process */
+            Skein_512_Process_Block(ctx,msg,n,SKEIN_512_BLOCK_BYTES);
+            msgByteCnt -= n * SKEIN_512_BLOCK_BYTES;
+            msg        += n * SKEIN_512_BLOCK_BYTES;
+            }
+        Skein_assert(ctx->h.bCnt == 0);
+        }
+
+    /* copy any remaining source message data bytes into b[] */
+    if (msgByteCnt)
+        {
+        Skein_assert(msgByteCnt + ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES);
+        memcpy(&ctx->b[ctx->h.bCnt],msg,msgByteCnt);
+        ctx->h.bCnt += msgByteCnt;
+        }
+
+    return SKEIN_SUCCESS;
+    }
+   
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize the hash computation and output the result */
+static int Skein_512_Final(Skein_512_Ctxt_t *ctx, u08b_t *hashVal)
+    {
+    size_t i,n,byteCnt;
+    u64b_t X[SKEIN_512_STATE_WORDS];
+    Skein_Assert(ctx->h.bCnt <= SKEIN_512_BLOCK_BYTES,SKEIN_FAIL);    /* catch uninitialized context */
+
+    ctx->h.T[1] |= SKEIN_T1_FLAG_FINAL;                 /* tag as the final block */
+    if (ctx->h.bCnt < SKEIN_512_BLOCK_BYTES)            /* zero pad b[] if necessary */
+        memset(&ctx->b[ctx->h.bCnt],0,SKEIN_512_BLOCK_BYTES - ctx->h.bCnt);
+
+    Skein_512_Process_Block(ctx,ctx->b,1,ctx->h.bCnt);  /* process the final block */
+    
+    /* now output the result */
+    byteCnt = (ctx->h.hashBitLen + 7) >> 3;             /* total number of output bytes */
+
+    /* run Threefish in "counter mode" to generate output */
+    memset(ctx->b,0,sizeof(ctx->b));  /* zero out b[], so it can hold the counter */
+    memcpy(X,ctx->X,sizeof(X));       /* keep a local copy of counter mode "key" */
+    for (i=0;i*SKEIN_512_BLOCK_BYTES < byteCnt;i++)
+        {
+        ((u64b_t *)ctx->b)[0]= Skein_Swap64((u64b_t) i); /* build the counter block */
+        Skein_Start_New_Type(ctx,OUT_FINAL);
+        Skein_512_Process_Block(ctx,ctx->b,1,sizeof(u64b_t)); /* run "counter mode" */
+        n = byteCnt - i*SKEIN_512_BLOCK_BYTES;   /* number of output bytes left to go */
+        if (n >= SKEIN_512_BLOCK_BYTES)
+            n  = SKEIN_512_BLOCK_BYTES;
+        Skein_Put64_LSB_First(hashVal+i*SKEIN_512_BLOCK_BYTES,ctx->X,n);   /* "output" the ctr mode bytes */
+        Skein_Show_Final(512,&ctx->h,n,hashVal+i*SKEIN_512_BLOCK_BYTES);
+        memcpy(ctx->X,X,sizeof(X));   /* restore the counter mode key for next time */
+        }
+    return SKEIN_SUCCESS;
+    }
+
+#if defined(SKEIN_CODE_SIZE) || defined(SKEIN_PERF)
+static size_t Skein_512_API_CodeSize(void)
+    {
+    return ((u08b_t *) Skein_512_API_CodeSize) -
+           ((u08b_t *) Skein_512_Init);
+    }
+#endif
+
+typedef struct
+{
+  uint_t  statebits;                      /* 256, 512, or 1024 */
+  union
+  {
+    Skein_Ctxt_Hdr_t h;                 /* common header "overlay" */
+    Skein_512_Ctxt_t ctx_512;
+  } u;
+}
+hashState;
+
+/* "incremental" hashing API */
+static SkeinHashReturn Init  (hashState *state, int hashbitlen);
+static SkeinHashReturn Update(hashState *state, const SkeinBitSequence *data, SkeinDataLength databitlen);
+static SkeinHashReturn Final (hashState *state,       SkeinBitSequence *hashval);
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* select the context size and init the context */
+static SkeinHashReturn Init(hashState *state, int hashbitlen)
+{
+    state->statebits = 64*SKEIN_512_STATE_WORDS;
+    return Skein_512_Init(&state->u.ctx_512,(size_t) hashbitlen);
+}
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* process data to be hashed */
+static SkeinHashReturn Update(hashState *state, const SkeinBitSequence *data, SkeinDataLength databitlen)
+{
+  /* only the final Update() call is allowed do partial bytes, else assert an error */
+  Skein_Assert((state->u.h.T[1] & SKEIN_T1_FLAG_BIT_PAD) == 0 || databitlen == 0, SKEIN_FAIL);
+
+  Skein_Assert(state->statebits % 256 == 0 && (state->statebits-256) < 1024,SKEIN_FAIL);
+  if ((databitlen & 7) == 0)  /* partial bytes? */
+  {
+    return Skein_512_Update(&state->u.ctx_512,data,databitlen >> 3);
+  }
+  else
+  {   /* handle partial final byte */
+    size_t bCnt = (databitlen >> 3) + 1;                  /* number of bytes to handle (nonzero here!) */
+    u08b_t b,mask;
+
+    mask = (u08b_t) (1u << (7 - (databitlen & 7)));       /* partial byte bit mask */
+    b    = (u08b_t) ((data[bCnt-1] & (0-mask)) | mask);   /* apply bit padding on final byte */
+
+    Skein_512_Update(&state->u.ctx_512,data,bCnt-1); /* process all but the final byte    */
+    Skein_512_Update(&state->u.ctx_512,&b  ,  1   ); /* process the (masked) partial byte */
+    Skein_Set_Bit_Pad_Flag(state->u.h);                    /* set tweak flag for the final call */
+
+    return SKEIN_SUCCESS;
+  }
+}
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* finalize hash computation and output the result (hashbitlen bits) */
+static SkeinHashReturn Final(hashState *state, SkeinBitSequence *hashval)
+{
+  Skein_Assert(state->statebits % 256 == 0 && (state->statebits-256) < 1024,FAIL);
+  return Skein_512_Final(&state->u.ctx_512,hashval);
+}
+
+/*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/
+/* all-in-one hash function */
+SkeinHashReturn skein_hash(int hashbitlen, const SkeinBitSequence *data, /* all-in-one call */
+                SkeinDataLength databitlen,SkeinBitSequence *hashval)
+{
+  hashState  state;
+  SkeinHashReturn r = Init(&state,hashbitlen);
+  if (r == SKEIN_SUCCESS)
+  { /* these calls do not fail when called properly */
+    r = Update(&state,data,databitlen);
+    Final(&state,hashval);
+  }
+  return r;
+}
+
+void xmr_skein(const SkeinBitSequence *data, SkeinBitSequence *hashval){
+  #define XMR_HASHBITLEN 256
+  #define XMR_DATABITLEN 1600
+
+  // Init
+  hashState  state;
+  state.statebits = 64*SKEIN_512_STATE_WORDS;
+
+  // Skein_512_Init(&state.u.ctx_512, (size_t)XMR_HASHBITLEN);
+  state.u.ctx_512.h.hashBitLen = XMR_HASHBITLEN;
+  memcpy(state.u.ctx_512.X,SKEIN_512_IV_256,sizeof(state.u.ctx_512.X));
+  Skein_512_Ctxt_t* ctx = &(state.u.ctx_512);
+  Skein_Start_New_Type(ctx,MSG);
+
+  // Update
+  if ((XMR_DATABITLEN & 7) == 0){  /* partial bytes? */
+    Skein_512_Update(&state.u.ctx_512,data,XMR_DATABITLEN >> 3);
+  }else{   /* handle partial final byte */
+    size_t bCnt = (XMR_DATABITLEN >> 3) + 1;                  /* number of bytes to handle (nonzero here!) */
+    u08b_t b,mask;
+
+    mask = (u08b_t) (1u << (7 - (XMR_DATABITLEN & 7)));       /* partial byte bit mask */
+    b    = (u08b_t) ((data[bCnt-1] & (0-mask)) | mask);   /* apply bit padding on final byte */
+
+    Skein_512_Update(&state.u.ctx_512,data,bCnt-1); /* process all but the final byte    */
+    Skein_512_Update(&state.u.ctx_512,&b  ,  1   ); /* process the (masked) partial byte */
+    Skein_Set_Bit_Pad_Flag(state.u.h);                    /* set tweak flag for the final call */
+  }
+
+  // Finalize
+  Skein_512_Final(&state.u.ctx_512, hashval);
+}

src/crypto/c_skein.h

@@ -44,4 +44,6 @@ typedef u08b_t   SkeinBitSequence;               /* bit stream type */
 SkeinHashReturn skein_hash(int hashbitlen,   const SkeinBitSequence *data,
         SkeinDataLength databitlen, SkeinBitSequence *hashval);
 
+void xmr_skein(const SkeinBitSequence *data, SkeinBitSequence *hashval);
+
 #endif  /* ifndef _SKEIN_H_ */