Conversion to NinjaRig.
This commit is contained in:
Haifa Bogdan Adnan
parent
84f56f0a4e
commit
2845347881
280 changed files with 18971 additions and 32469 deletions
132
src/crypto/argon2_hasher/hash/Hasher.cpp
Executable file
132
src/crypto/argon2_hasher/hash/Hasher.cpp
Executable file
|
|
@ -0,0 +1,132 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#include "../common/common.h"
|
||||
#include "../crypt/base64.h"
|
||||
#include "../crypt/hex.h"
|
||||
#include "../crypt/random_generator.h"
|
||||
|
||||
#include "crypto/argon2_hasher/common/DLLExport.h"
|
||||
#include "crypto/argon2_hasher/hash/argon2/Argon2.h"
|
||||
#include "Hasher.h"
|
||||
|
||||
vector<Hasher *> *Hasher::m_registeredHashers = NULL;
|
||||
string Hasher::m_appFolder = "";
|
||||
|
||||
typedef void (*hasherLoader)();
|
||||
|
||||
Hasher::Hasher() {
|
||||
m_intensity = 0;
|
||||
m_type = "";
|
||||
m_subType = "";
|
||||
m_shortSubType = "";
|
||||
m_description = "";
|
||||
|
||||
m_computingThreads = 1;
|
||||
|
||||
if(m_registeredHashers == NULL) {
|
||||
m_registeredHashers = new vector<Hasher*>();
|
||||
}
|
||||
|
||||
m_registeredHashers->push_back(this);
|
||||
}
|
||||
|
||||
Hasher::~Hasher() {};
|
||||
|
||||
string Hasher::type() {
|
||||
return m_type;
|
||||
}
|
||||
|
||||
string Hasher::subType(bool shortName) {
|
||||
if(shortName && !(m_shortSubType.empty())) {
|
||||
string shortVersion = m_shortSubType;
|
||||
shortVersion.erase(3);
|
||||
return shortVersion;
|
||||
}
|
||||
else
|
||||
return m_subType;
|
||||
}
|
||||
|
||||
string Hasher::info() {
|
||||
return m_description;
|
||||
}
|
||||
|
||||
int Hasher::computingThreads() {
|
||||
return m_computingThreads;
|
||||
}
|
||||
|
||||
void Hasher::loadHashers(const string &appPath) {
|
||||
m_registeredHashers = new vector<Hasher*>();
|
||||
|
||||
string modulePath = ".";
|
||||
|
||||
size_t lastSlash = appPath.find_last_of("/\\");
|
||||
if (lastSlash != string::npos) {
|
||||
modulePath = appPath.substr(0, lastSlash);
|
||||
if(modulePath.empty()) {
|
||||
modulePath = ".";
|
||||
}
|
||||
}
|
||||
|
||||
m_appFolder = modulePath;
|
||||
|
||||
modulePath += "/modules/";
|
||||
|
||||
vector<string> files = getFiles(modulePath);
|
||||
for(string file : files) {
|
||||
if(file.find(".hsh") != string::npos) {
|
||||
void *dllHandle = dlopen((modulePath + file).c_str(), RTLD_LAZY);
|
||||
if(dllHandle != NULL) {
|
||||
hasherLoader hasherLoaderPtr = (hasherLoader) dlsym(dllHandle, "hasherLoader");
|
||||
(*hasherLoaderPtr)();
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
vector<Hasher *> Hasher::getHashers() {
|
||||
return *m_registeredHashers;
|
||||
}
|
||||
|
||||
vector<Hasher *> Hasher::getActiveHashers() {
|
||||
vector<Hasher *> filtered;
|
||||
for(Hasher *hasher : *m_registeredHashers) {
|
||||
if(hasher->m_intensity != 0)
|
||||
filtered.push_back(hasher);
|
||||
}
|
||||
return filtered;
|
||||
}
|
||||
|
||||
vector<Hasher *> Hasher::getHashers(const string &type) {
|
||||
vector<Hasher *> filtered;
|
||||
for(Hasher *hasher : *m_registeredHashers) {
|
||||
if(hasher->m_type == type)
|
||||
filtered.push_back(hasher);
|
||||
}
|
||||
return filtered;
|
||||
}
|
||||
|
||||
map<int, DeviceInfo> &Hasher::devices() {
|
||||
return m_deviceInfos;
|
||||
}
|
||||
|
||||
void Hasher::storeDeviceInfo(int deviceId, DeviceInfo device) {
|
||||
m_deviceInfosMutex.lock();
|
||||
m_deviceInfos[deviceId] = device;
|
||||
m_deviceInfosMutex.unlock();
|
||||
}
|
||||
|
||||
Argon2Profile *Hasher::getArgon2Profile(xmrig::Algo algorithm, xmrig::Variant variant) {
|
||||
if(algorithm == xmrig::ARGON2) {
|
||||
switch(variant) {
|
||||
case xmrig::VARIANT_CHUKWA:
|
||||
return &argon2profile_3_1_512;
|
||||
case xmrig::VARIANT_CHUKWA_LITE:
|
||||
return &argon2profile_4_1_256;
|
||||
default:
|
||||
return nullptr;
|
||||
}
|
||||
}
|
||||
return nullptr;
|
||||
}
|
||||
63
src/crypto/argon2_hasher/hash/Hasher.h
Executable file
63
src/crypto/argon2_hasher/hash/Hasher.h
Executable file
|
|
@ -0,0 +1,63 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#ifndef ARGON2_HASHER_H
|
||||
#define ARGON2_HASHER_H
|
||||
|
||||
#include "crypto/argon2_hasher/hash/argon2/Defs.h"
|
||||
#include "../../../core/HasherConfig.h"
|
||||
#include "../../../common/xmrig.h"
|
||||
|
||||
struct DeviceInfo {
|
||||
string name;
|
||||
string bus_id;
|
||||
double intensity;
|
||||
};
|
||||
|
||||
#define REGISTER_HASHER(x) extern "C" { DLLEXPORT void hasherLoader() { x *instance = new x(); } }
|
||||
|
||||
class DLLEXPORT Hasher {
|
||||
public:
|
||||
Hasher();
|
||||
virtual ~Hasher();
|
||||
|
||||
virtual bool initialize(xmrig::Algo algorithm, xmrig::Variant variant) = 0;
|
||||
virtual bool configure(xmrig::HasherConfig &config) = 0;
|
||||
virtual void cleanup() = 0;
|
||||
virtual int compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output) = 0;
|
||||
virtual size_t parallelism(int workerIdx) = 0;
|
||||
virtual size_t deviceCount() = 0;
|
||||
|
||||
string type();
|
||||
string subType(bool shortName = false);
|
||||
|
||||
string info();
|
||||
int computingThreads();
|
||||
|
||||
map<int, DeviceInfo> &devices();
|
||||
|
||||
static vector<Hasher*> getHashers(const string &type);
|
||||
static vector<Hasher*> getHashers();
|
||||
static vector<Hasher*> getActiveHashers();
|
||||
static void loadHashers(const string &appPath);
|
||||
|
||||
protected:
|
||||
double m_intensity;
|
||||
string m_type;
|
||||
string m_subType;
|
||||
string m_shortSubType; //max 3 characters
|
||||
string m_description;
|
||||
int m_computingThreads;
|
||||
static string m_appFolder;
|
||||
|
||||
void storeDeviceInfo(int deviceId, DeviceInfo device);
|
||||
Argon2Profile *getArgon2Profile(xmrig::Algo algorithm, xmrig::Variant variant);
|
||||
|
||||
private:
|
||||
static vector<Hasher*> *m_registeredHashers;
|
||||
map<int, DeviceInfo> m_deviceInfos;
|
||||
mutex m_deviceInfosMutex;
|
||||
};
|
||||
|
||||
#endif //ARGON2_HASHER_H
|
||||
143
src/crypto/argon2_hasher/hash/argon2/Argon2.cpp
Executable file
143
src/crypto/argon2_hasher/hash/argon2/Argon2.cpp
Executable file
|
|
@ -0,0 +1,143 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 05/08/2018.
|
||||
//
|
||||
|
||||
#include "../../common/common.h"
|
||||
#include "../../crypt/base64.h"
|
||||
#include "../../crypt/hex.h"
|
||||
#include "../../crypt/random_generator.h"
|
||||
|
||||
#include "blake2/blake2.h"
|
||||
#include "../../common/DLLExport.h"
|
||||
#include "../../../Argon2_constants.h"
|
||||
#include "Argon2.h"
|
||||
#include "Defs.h"
|
||||
|
||||
Argon2::Argon2(argon2BlocksPrehash prehash, argon2BlocksFillerPtr filler, argon2BlocksPosthash posthash, void *memory, void *userData) {
|
||||
m_prehash = prehash;
|
||||
m_filler = filler;
|
||||
m_posthash = posthash;
|
||||
m_outputMemory = m_seedMemory = (uint8_t*)memory;
|
||||
m_userData = userData;
|
||||
m_threads = 1;
|
||||
}
|
||||
|
||||
int Argon2::generateHashes(const Argon2Profile &profile, HashData &hashData) {
|
||||
if(initializeSeeds(profile, hashData)) {
|
||||
if(fillBlocks(profile)) {
|
||||
return encodeHashes(profile, hashData);
|
||||
}
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
bool Argon2::initializeSeeds(const Argon2Profile &profile, HashData &hashData) {
|
||||
if(m_prehash != NULL) {
|
||||
return (*m_prehash)(hashData.input, m_threads, (Argon2Profile*)&profile, m_userData);
|
||||
}
|
||||
else {
|
||||
uint8_t blockhash[ARGON2_PREHASH_SEED_LENGTH];
|
||||
|
||||
for (int i = 0; i < m_threads; i++, (*(nonce(hashData)))++) {
|
||||
initialHash(profile, blockhash, (char *) hashData.input, hashData.inSize, xmrig::ARGON2_HASHLEN);
|
||||
|
||||
memset(blockhash + ARGON2_PREHASH_DIGEST_LENGTH, 0,
|
||||
ARGON2_PREHASH_SEED_LENGTH -
|
||||
ARGON2_PREHASH_DIGEST_LENGTH);
|
||||
|
||||
fillFirstBlocks(profile, blockhash, i);
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
}
|
||||
|
||||
bool Argon2::fillBlocks(const Argon2Profile &profile) {
|
||||
m_outputMemory = (uint8_t *)(*m_filler) (m_threads, (Argon2Profile*)&profile, m_userData);
|
||||
return m_outputMemory != NULL;
|
||||
}
|
||||
|
||||
int Argon2::encodeHashes(const Argon2Profile &profile, HashData &hashData) {
|
||||
if(m_posthash != NULL) {
|
||||
if((*m_posthash)(hashData.output, m_threads, (Argon2Profile*)&profile, m_userData)) {
|
||||
return m_threads;
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
else {
|
||||
if (m_outputMemory != NULL) {
|
||||
uint32_t nonceInfo = *(nonce(hashData)) - m_threads;
|
||||
|
||||
for (int i = 0; i < m_threads; i++, nonceInfo++) {
|
||||
blake2b_long((void *) (hashData.output + i * hashData.outSize), xmrig::ARGON2_HASHLEN,
|
||||
(void *) (m_outputMemory + i * profile.memSize), ARGON2_BLOCK_SIZE);
|
||||
memcpy(hashData.output + i * hashData.outSize + xmrig::ARGON2_HASHLEN, &nonceInfo, 4);
|
||||
}
|
||||
return m_threads;
|
||||
}
|
||||
else
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
|
||||
void Argon2::initialHash(const Argon2Profile &profile, uint8_t *blockhash, const char *data, size_t dataSz,size_t outSz) {
|
||||
blake2b_state BlakeHash;
|
||||
uint32_t value;
|
||||
|
||||
blake2b_init(&BlakeHash, ARGON2_PREHASH_DIGEST_LENGTH);
|
||||
|
||||
value = profile.thrCost;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
value = outSz;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
value = profile.memCost;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
value = profile.tmCost;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
value = ARGON2_VERSION;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
value = ARGON2_TYPE_VALUE;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
value = (uint32_t)dataSz;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)data, dataSz);
|
||||
|
||||
value = xmrig::ARGON2_SALTLEN;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)data, xmrig::ARGON2_SALTLEN);
|
||||
|
||||
value = 0;
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
blake2b_update(&BlakeHash, (const uint8_t *)&value, sizeof(value));
|
||||
|
||||
blake2b_final(&BlakeHash, blockhash, ARGON2_PREHASH_DIGEST_LENGTH);
|
||||
}
|
||||
|
||||
void Argon2::fillFirstBlocks(const Argon2Profile &profile, uint8_t *blockhash, int thread) {
|
||||
block *blocks = (block *)(m_seedMemory + thread * profile.memSize);
|
||||
size_t lane_length = profile.memCost / profile.thrCost;
|
||||
|
||||
for (uint32_t l = 0; l < profile.thrCost; ++l) {
|
||||
*((uint32_t*)(blockhash + ARGON2_PREHASH_DIGEST_LENGTH)) = 0;
|
||||
*((uint32_t*)(blockhash + ARGON2_PREHASH_DIGEST_LENGTH + 4)) = l;
|
||||
|
||||
blake2b_long((void *)(blocks + l * lane_length), ARGON2_BLOCK_SIZE, blockhash,
|
||||
ARGON2_PREHASH_SEED_LENGTH);
|
||||
|
||||
*((uint32_t*)(blockhash + ARGON2_PREHASH_DIGEST_LENGTH)) = 1;
|
||||
|
||||
blake2b_long((void *)(blocks + l * lane_length + 1), ARGON2_BLOCK_SIZE, blockhash,
|
||||
ARGON2_PREHASH_SEED_LENGTH);
|
||||
}
|
||||
}
|
||||
|
||||
void Argon2::setThreads(int threads) {
|
||||
m_threads = threads;
|
||||
}
|
||||
56
src/crypto/argon2_hasher/hash/argon2/Argon2.h
Normal file
56
src/crypto/argon2_hasher/hash/argon2/Argon2.h
Normal file
|
|
@ -0,0 +1,56 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 05/08/2018.
|
||||
//
|
||||
|
||||
#ifndef ARIOMINER_ARGON2_H
|
||||
#define ARIOMINER_ARGON2_H
|
||||
|
||||
#include "Defs.h"
|
||||
#include "crypto/argon2_hasher/hash/Hasher.h"
|
||||
|
||||
typedef bool (*argon2BlocksPrehash)(void *, int, Argon2Profile *, void *); // data_memory
|
||||
typedef void *(*argon2BlocksFillerPtr)(int, Argon2Profile *, void *);
|
||||
typedef bool (*argon2BlocksPosthash)(void *, int, Argon2Profile *, void *); // raw_hash_mem
|
||||
|
||||
struct HashData {
|
||||
uint8_t *input;
|
||||
uint8_t *output;
|
||||
size_t inSize;
|
||||
size_t outSize;
|
||||
};
|
||||
|
||||
class DLLEXPORT Argon2 {
|
||||
public:
|
||||
Argon2(argon2BlocksPrehash prehash, argon2BlocksFillerPtr filler, argon2BlocksPosthash posthash, void *memory, void *userData);
|
||||
|
||||
int generateHashes(const Argon2Profile &profile, HashData &hashData);
|
||||
|
||||
bool initializeSeeds(const Argon2Profile &profile, HashData &hashData);
|
||||
bool fillBlocks(const Argon2Profile &profile);
|
||||
int encodeHashes(const Argon2Profile &profile, HashData &hashData);
|
||||
|
||||
void setThreads(int threads);
|
||||
|
||||
private:
|
||||
void initialHash(const Argon2Profile &profile, uint8_t *blockhash, const char *data, size_t dataSz, size_t outSz);
|
||||
void fillFirstBlocks(const Argon2Profile &profile, uint8_t *blockhash, int thread);
|
||||
|
||||
inline uint32_t *nonce(HashData &hashData)
|
||||
{
|
||||
return reinterpret_cast<uint32_t*>(hashData.input + 39);
|
||||
}
|
||||
|
||||
argon2BlocksPrehash m_prehash;
|
||||
argon2BlocksFillerPtr m_filler;
|
||||
argon2BlocksPosthash m_posthash;
|
||||
|
||||
int m_threads;
|
||||
|
||||
uint8_t *m_seedMemory;
|
||||
uint8_t *m_outputMemory;
|
||||
|
||||
void *m_userData;
|
||||
};
|
||||
|
||||
|
||||
#endif //ARIOMINER_ARGON2_H
|
||||
50
src/crypto/argon2_hasher/hash/argon2/Defs.h
Executable file
50
src/crypto/argon2_hasher/hash/argon2/Defs.h
Executable file
|
|
@ -0,0 +1,50 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 06/08/2018.
|
||||
//
|
||||
|
||||
#ifndef ARIOMINER_DEFS_H
|
||||
#define ARIOMINER_DEFS_H
|
||||
|
||||
#define ARGON2_RAW_LENGTH 32
|
||||
#define ARGON2_TYPE_VALUE 2
|
||||
#define ARGON2_VERSION 0x13
|
||||
|
||||
#define ARGON2_BLOCK_SIZE 1024
|
||||
#define ARGON2_DWORDS_IN_BLOCK ARGON2_BLOCK_SIZE / 4
|
||||
#define ARGON2_QWORDS_IN_BLOCK ARGON2_BLOCK_SIZE / 8
|
||||
#define ARGON2_OWORDS_IN_BLOCK ARGON2_BLOCK_SIZE / 16
|
||||
#define ARGON2_HWORDS_IN_BLOCK ARGON2_BLOCK_SIZE / 32
|
||||
#define ARGON2_512BIT_WORDS_IN_BLOCK ARGON2_BLOCK_SIZE / 64
|
||||
#define ARGON2_PREHASH_DIGEST_LENGTH 64
|
||||
#define ARGON2_PREHASH_SEED_LENGTH 72
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
typedef struct block_ { uint64_t v[ARGON2_QWORDS_IN_BLOCK]; } block;
|
||||
|
||||
typedef struct Argon2Profile_ {
|
||||
uint32_t memCost;
|
||||
uint32_t thrCost;
|
||||
uint32_t tmCost;
|
||||
size_t memSize;
|
||||
int32_t *blockRefs;
|
||||
size_t blockRefsSize;
|
||||
char profileName[15];
|
||||
int32_t *segments; // { start segment / current block, stop segment (excluding) / previous block, addressing type = 0 -> i, 1 -> d }
|
||||
uint32_t segSize;
|
||||
uint32_t segCount;
|
||||
uint32_t succesiveIdxs; // 0 - idx are precalculated, 1 - idx are successive
|
||||
int pwdLen; // in dwords
|
||||
int saltLen; // in dwords
|
||||
} Argon2Profile;
|
||||
|
||||
extern DLLEXPORT Argon2Profile argon2profile_3_1_512;
|
||||
extern DLLEXPORT Argon2Profile argon2profile_4_1_256;
|
||||
|
||||
#ifdef __cplusplus
|
||||
}
|
||||
#endif
|
||||
|
||||
#endif //ARIOMINER_DEFS_H
|
||||
292
src/crypto/argon2_hasher/hash/argon2/argon2profile_3_1_512.c
Normal file
292
src/crypto/argon2_hasher/hash/argon2/argon2profile_3_1_512.c
Normal file
|
|
@ -0,0 +1,292 @@
|
|||
#include <stdint.h>
|
||||
#include <stddef.h>
|
||||
#include "../../common/DLLExport.h"
|
||||
#include "Defs.h"
|
||||
|
||||
int32_t blocks_refs_3_1_512[] = {
|
||||
2, 0, 1,
|
||||
3, 1, 1,
|
||||
4, 2, 1,
|
||||
5, 3, 1,
|
||||
6, 3, 1,
|
||||
7, 3, 1,
|
||||
8, 2, 1,
|
||||
9, 5, 1,
|
||||
10, 0, 1,
|
||||
11, 9, 1,
|
||||
12, 10, 1,
|
||||
13, 9, 1,
|
||||
14, 12, 1,
|
||||
15, 8, 1,
|
||||
16, 5, 1,
|
||||
17, 15, 1,
|
||||
18, 10, 1,
|
||||
19, 14, 1,
|
||||
20, 7, 1,
|
||||
21, 19, 1,
|
||||
22, 14, 1,
|
||||
23, 7, 1,
|
||||
24, 14, 1,
|
||||
25, 23, 1,
|
||||
26, 24, 1,
|
||||
27, 0, 1,
|
||||
28, 9, 1,
|
||||
29, 11, 1,
|
||||
30, 12, 1,
|
||||
31, 29, 1,
|
||||
32, 12, 1,
|
||||
33, 23, 1,
|
||||
34, 30, 1,
|
||||
35, 1, 1,
|
||||
36, 32, 1,
|
||||
37, 8, 1,
|
||||
38, 30, 1,
|
||||
39, 31, 1,
|
||||
40, 15, 1,
|
||||
41, 38, 1,
|
||||
42, 29, 1,
|
||||
43, 18, 1,
|
||||
44, 33, 1,
|
||||
45, 18, 1,
|
||||
46, 39, 1,
|
||||
47, 43, 1,
|
||||
48, 40, 1,
|
||||
49, 38, 1,
|
||||
50, 5, 1,
|
||||
51, 47, 1,
|
||||
52, 14, 1,
|
||||
53, 45, 1,
|
||||
54, 30, 1,
|
||||
55, 13, 1,
|
||||
56, 47, 1,
|
||||
57, 30, 1,
|
||||
58, 21, 1,
|
||||
59, 18, 1,
|
||||
60, 36, 1,
|
||||
61, 58, 1,
|
||||
62, 58, 1,
|
||||
63, 19, 1,
|
||||
64, 59, 1,
|
||||
65, 29, 1,
|
||||
66, 10, 1,
|
||||
67, 48, 1,
|
||||
68, 39, 1,
|
||||
69, 25, 1,
|
||||
70, 63, 1,
|
||||
71, 57, 1,
|
||||
72, 70, 1,
|
||||
73, 16, 1,
|
||||
74, 20, 1,
|
||||
75, 72, 1,
|
||||
76, 67, 1,
|
||||
77, 61, 1,
|
||||
78, 49, 1,
|
||||
79, 63, 1,
|
||||
80, 9, 1,
|
||||
81, 19, 1,
|
||||
82, 80, 1,
|
||||
83, 36, 1,
|
||||
84, 20, 1,
|
||||
85, 23, 1,
|
||||
86, 52, 1,
|
||||
87, 85, 1,
|
||||
88, 75, 1,
|
||||
89, 18, 1,
|
||||
90, 85, 1,
|
||||
91, 2, 1,
|
||||
92, 81, 1,
|
||||
93, 91, 1,
|
||||
94, 91, 1,
|
||||
95, 3, 1,
|
||||
96, 45, 1,
|
||||
97, 16, 1,
|
||||
98, 11, 1,
|
||||
99, 60, 1,
|
||||
100, 89, 1,
|
||||
101, 65, 1,
|
||||
102, 39, 1,
|
||||
103, 63, 1,
|
||||
104, 66, 1,
|
||||
105, 74, 1,
|
||||
106, 54, 1,
|
||||
107, 88, 1,
|
||||
108, 106, 1,
|
||||
109, 107, 1,
|
||||
110, 47, 1,
|
||||
111, 8, 1,
|
||||
112, 95, 1,
|
||||
113, 66, 1,
|
||||
114, 1, 1,
|
||||
115, 2, 1,
|
||||
116, 20, 1,
|
||||
117, 110, 1,
|
||||
118, 47, 1,
|
||||
119, 117, 1,
|
||||
120, 114, 1,
|
||||
121, 37, 1,
|
||||
122, 71, 1,
|
||||
123, 51, 1,
|
||||
124, 122, 1,
|
||||
125, 44, 1,
|
||||
126, 92, 1,
|
||||
127, 120, 1,
|
||||
128, 123, 1,
|
||||
129, 127, 1,
|
||||
130, 11, 1,
|
||||
131, 110, 1,
|
||||
132, 93, 1,
|
||||
133, 20, 1,
|
||||
134, 58, 1,
|
||||
135, 13, 1,
|
||||
136, 73, 1,
|
||||
137, 27, 1,
|
||||
138, 94, 1,
|
||||
139, 110, 1,
|
||||
140, 96, 1,
|
||||
141, 57, 1,
|
||||
142, 137, 1,
|
||||
143, 116, 1,
|
||||
144, 119, 1,
|
||||
145, 141, 1,
|
||||
146, 73, 1,
|
||||
147, 26, 1,
|
||||
148, 103, 1,
|
||||
149, 125, 1,
|
||||
150, 146, 1,
|
||||
151, 149, 1,
|
||||
152, 28, 1,
|
||||
153, 149, 1,
|
||||
154, 125, 1,
|
||||
155, 104, 1,
|
||||
156, 61, 1,
|
||||
157, 128, 1,
|
||||
158, 156, 1,
|
||||
159, 122, 1,
|
||||
160, 96, 1,
|
||||
161, 92, 1,
|
||||
162, 160, 1,
|
||||
163, 154, 1,
|
||||
164, 88, 1,
|
||||
165, 160, 1,
|
||||
166, 134, 1,
|
||||
167, 116, 1,
|
||||
168, 23, 1,
|
||||
169, 167, 1,
|
||||
170, 100, 1,
|
||||
171, 169, 1,
|
||||
172, 169, 1,
|
||||
173, 127, 1,
|
||||
174, 0, 1,
|
||||
175, 78, 1,
|
||||
176, 155, 1,
|
||||
177, 124, 1,
|
||||
178, 138, 1,
|
||||
179, 41, 1,
|
||||
180, 156, 1,
|
||||
181, 173, 1,
|
||||
182, 122, 1,
|
||||
183, 173, 1,
|
||||
184, 112, 1,
|
||||
185, 15, 1,
|
||||
186, 183, 1,
|
||||
187, 171, 1,
|
||||
188, 163, 1,
|
||||
189, 85, 1,
|
||||
190, 45, 1,
|
||||
191, 171, 1,
|
||||
192, 139, 1,
|
||||
193, 188, 1,
|
||||
194, 192, 1,
|
||||
195, 78, 1,
|
||||
196, 5, 1,
|
||||
197, 187, 1,
|
||||
198, 180, 1,
|
||||
199, 195, 1,
|
||||
200, 102, 1,
|
||||
201, 89, 1,
|
||||
202, 165, 1,
|
||||
203, 144, 1,
|
||||
204, 171, 1,
|
||||
205, 152, 1,
|
||||
206, 53, 1,
|
||||
207, 19, 1,
|
||||
208, 206, 1,
|
||||
209, 165, 1,
|
||||
210, 208, 1,
|
||||
211, 76, 1,
|
||||
212, 177, 1,
|
||||
213, 189, 1,
|
||||
214, 43, 1,
|
||||
215, 120, 1,
|
||||
216, 122, 1,
|
||||
217, 189, 1,
|
||||
218, 45, 1,
|
||||
219, 217, 1,
|
||||
220, 207, 1,
|
||||
221, 202, 1,
|
||||
222, 169, 1,
|
||||
223, 194, 1,
|
||||
224, 213, 1,
|
||||
225, 178, 1,
|
||||
226, 175, 1,
|
||||
227, 221, 1,
|
||||
228, 212, 1,
|
||||
229, 220, 1,
|
||||
230, 227, 1,
|
||||
231, 30, 1,
|
||||
232, 34, 1,
|
||||
233, 91, 1,
|
||||
234, 231, 1,
|
||||
235, 154, 1,
|
||||
236, 100, 1,
|
||||
237, 166, 1,
|
||||
238, 216, 1,
|
||||
239, 229, 1,
|
||||
240, 177, 1,
|
||||
241, 123, 1,
|
||||
242, 172, 1,
|
||||
243, 71, 1,
|
||||
244, 241, 1,
|
||||
245, 236, 1,
|
||||
246, 109, 1,
|
||||
247, 4, 1,
|
||||
248, 246, 1,
|
||||
249, 166, 1,
|
||||
250, 248, 1,
|
||||
251, 243, 1,
|
||||
252, 248, 1,
|
||||
253, 39, 1,
|
||||
254, 98, 1,
|
||||
255, 253, 1
|
||||
};
|
||||
|
||||
int32_t segments_3_1_512[] = { // current_idx, previous_idx, seg_type 0=i 1=d
|
||||
2, 1, 0,
|
||||
128, 127, 0,
|
||||
256, 255, 1,
|
||||
384, 383, 1,
|
||||
0, 511, 1,
|
||||
128, 127, 1,
|
||||
256, 255, 1,
|
||||
384, 383, 1,
|
||||
0, 511, 1,
|
||||
128, 127, 1,
|
||||
256, 255, 1,
|
||||
384, 383, 1
|
||||
};
|
||||
|
||||
DLLEXPORT Argon2Profile argon2profile_3_1_512 = {
|
||||
512,
|
||||
1,
|
||||
3,
|
||||
524288, //256 blocks of 1024 bytes
|
||||
blocks_refs_3_1_512,
|
||||
sizeof(blocks_refs_3_1_512) / (3 * sizeof(int32_t)),
|
||||
"3_1_512",
|
||||
segments_3_1_512,
|
||||
128,
|
||||
12,
|
||||
1,
|
||||
32,
|
||||
4
|
||||
};
|
||||
168
src/crypto/argon2_hasher/hash/argon2/argon2profile_4_1_256.c
Normal file
168
src/crypto/argon2_hasher/hash/argon2/argon2profile_4_1_256.c
Normal file
|
|
@ -0,0 +1,168 @@
|
|||
#include <stdint.h>
|
||||
#include <stddef.h>
|
||||
#include "../../common/DLLExport.h"
|
||||
#include "Defs.h"
|
||||
|
||||
int32_t blocks_refs_4_1_256[] = {
|
||||
2, 0, 1,
|
||||
3, 1, 1,
|
||||
4, 2, 1,
|
||||
5, 3, 1,
|
||||
6, 0, 1,
|
||||
7, 4, 1,
|
||||
8, 5, 1,
|
||||
9, 7, 1,
|
||||
10, 7, 1,
|
||||
11, 9, 1,
|
||||
12, 5, 1,
|
||||
13, 11, 1,
|
||||
14, 3, 1,
|
||||
15, 2, 1,
|
||||
16, 12, 1,
|
||||
17, 15, 1,
|
||||
18, 15, 1,
|
||||
19, 10, 1,
|
||||
20, 4, 1,
|
||||
21, 18, 1,
|
||||
22, 17, 1,
|
||||
23, 19, 1,
|
||||
24, 2, 1,
|
||||
25, 23, 1,
|
||||
26, 22, 1,
|
||||
27, 12, 1,
|
||||
28, 23, 1,
|
||||
29, 27, 1,
|
||||
30, 26, 1,
|
||||
31, 19, 1,
|
||||
32, 27, 1,
|
||||
33, 29, 1,
|
||||
34, 32, 1,
|
||||
35, 18, 1,
|
||||
36, 32, 1,
|
||||
37, 16, 1,
|
||||
38, 35, 1,
|
||||
39, 22, 1,
|
||||
40, 30, 1,
|
||||
41, 31, 1,
|
||||
42, 39, 1,
|
||||
43, 36, 1,
|
||||
44, 18, 1,
|
||||
45, 0, 1,
|
||||
46, 36, 1,
|
||||
47, 12, 1,
|
||||
48, 28, 1,
|
||||
49, 39, 1,
|
||||
50, 4, 1,
|
||||
51, 48, 1,
|
||||
52, 48, 1,
|
||||
53, 51, 1,
|
||||
54, 50, 1,
|
||||
55, 3, 1,
|
||||
56, 54, 1,
|
||||
57, 53, 1,
|
||||
58, 48, 1,
|
||||
59, 47, 1,
|
||||
60, 25, 1,
|
||||
61, 53, 1,
|
||||
62, 31, 1,
|
||||
63, 59, 1,
|
||||
64, 45, 1,
|
||||
65, 63, 1,
|
||||
66, 48, 1,
|
||||
67, 58, 1,
|
||||
68, 40, 1,
|
||||
69, 17, 1,
|
||||
70, 62, 1,
|
||||
71, 24, 1,
|
||||
72, 60, 1,
|
||||
73, 71, 1,
|
||||
74, 72, 1,
|
||||
75, 57, 1,
|
||||
76, 69, 1,
|
||||
77, 58, 1,
|
||||
78, 74, 1,
|
||||
79, 69, 1,
|
||||
80, 75, 1,
|
||||
81, 74, 1,
|
||||
82, 56, 1,
|
||||
83, 67, 1,
|
||||
84, 15, 1,
|
||||
85, 83, 1,
|
||||
86, 69, 1,
|
||||
87, 83, 1,
|
||||
88, 85, 1,
|
||||
89, 24, 1,
|
||||
90, 52, 1,
|
||||
91, 70, 1,
|
||||
92, 88, 1,
|
||||
93, 42, 1,
|
||||
94, 61, 1,
|
||||
95, 93, 1,
|
||||
96, 22, 1,
|
||||
97, 37, 1,
|
||||
98, 15, 1,
|
||||
99, 91, 1,
|
||||
100, 14, 1,
|
||||
101, 98, 1,
|
||||
102, 24, 1,
|
||||
103, 84, 1,
|
||||
104, 44, 1,
|
||||
105, 103, 1,
|
||||
106, 12, 1,
|
||||
107, 15, 1,
|
||||
108, 79, 1,
|
||||
109, 35, 1,
|
||||
110, 4, 1,
|
||||
111, 109, 1,
|
||||
112, 90, 1,
|
||||
113, 109, 1,
|
||||
114, 43, 1,
|
||||
115, 73, 1,
|
||||
116, 113, 1,
|
||||
117, 107, 1,
|
||||
118, 51, 1,
|
||||
119, 117, 1,
|
||||
120, 118, 1,
|
||||
121, 115, 1,
|
||||
122, 74, 1,
|
||||
123, 67, 1,
|
||||
124, 102, 1,
|
||||
125, 17, 1,
|
||||
126, 113, 1,
|
||||
127, 110, 1
|
||||
};
|
||||
|
||||
int32_t segments_4_1_256[] = { // current_idx, previous_idx, seg_type 0=i 1=d
|
||||
2, 1, 0,
|
||||
64, 63, 0,
|
||||
128, 127, 1,
|
||||
192, 191, 1,
|
||||
0, 255, 1,
|
||||
64, 63, 1,
|
||||
128, 127, 1,
|
||||
192, 191, 1,
|
||||
0, 255, 1,
|
||||
64, 63, 1,
|
||||
128, 127, 1,
|
||||
192, 191, 1,
|
||||
0, 255, 1,
|
||||
64, 63, 1,
|
||||
128, 127, 1,
|
||||
192, 191, 1
|
||||
};
|
||||
|
||||
DLLEXPORT Argon2Profile argon2profile_4_1_256 = {
|
||||
256,
|
||||
1,
|
||||
4,
|
||||
262144, //256 blocks of 1024 bytes
|
||||
blocks_refs_4_1_256,
|
||||
sizeof(blocks_refs_4_1_256) / (3 * sizeof(int32_t)),
|
||||
"4_1_256",
|
||||
segments_4_1_256,
|
||||
64,
|
||||
16,
|
||||
1,
|
||||
32,
|
||||
4
|
||||
};
|
||||
76
src/crypto/argon2_hasher/hash/argon2/blake2/blake2-config.h
Normal file
76
src/crypto/argon2_hasher/hash/argon2/blake2/blake2-config.h
Normal file
|
|
@ -0,0 +1,76 @@
|
|||
/*
|
||||
BLAKE2 reference source code package - optimized C implementations
|
||||
|
||||
Copyright 2012, Samuel Neves <sneves@dei.uc.pt>. You may use this under the
|
||||
terms of the CC0, the OpenSSL Licence, or the Apache Public License 2.0, at
|
||||
your option. The terms of these licenses can be found at:
|
||||
|
||||
- CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
- OpenSSL license : https://www.openssl.org/source/license.html
|
||||
- Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
More information about the BLAKE2 hash function can be found at
|
||||
https://blake2.net.
|
||||
*/
|
||||
#ifndef BLAKE2_CONFIG_H
|
||||
#define BLAKE2_CONFIG_H
|
||||
|
||||
/* These don't work everywhere */
|
||||
#if defined(__SSE2__) || defined(__x86_64__) || defined(__amd64__) || defined(_M_X64)
|
||||
#define HAVE_SSE2
|
||||
#endif
|
||||
|
||||
#if defined(__SSSE3__)
|
||||
#define HAVE_SSSE3
|
||||
#endif
|
||||
|
||||
#if defined(__SSE4_1__)
|
||||
#define HAVE_SSE41
|
||||
#endif
|
||||
|
||||
#if defined(__AVX__)
|
||||
#define HAVE_AVX
|
||||
#endif
|
||||
|
||||
#if defined(__AVX2__)
|
||||
#define HAVE_AVX2
|
||||
#endif
|
||||
|
||||
#if defined(__XOP__)
|
||||
#define HAVE_XOP
|
||||
#endif
|
||||
|
||||
|
||||
#ifdef HAVE_AVX2
|
||||
#ifndef HAVE_AVX
|
||||
#define HAVE_AVX
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_XOP
|
||||
#ifndef HAVE_AVX
|
||||
#define HAVE_AVX
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_AVX
|
||||
#ifndef HAVE_SSE41
|
||||
#define HAVE_SSE41
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_SSE41
|
||||
#ifndef HAVE_SSSE3
|
||||
#define HAVE_SSSE3
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_SSSE3
|
||||
#define HAVE_SSE2
|
||||
#endif
|
||||
|
||||
#if !defined(HAVE_SSE2)
|
||||
#error "This code requires at least SSE2."
|
||||
#endif
|
||||
|
||||
#endif
|
||||
154
src/crypto/argon2_hasher/hash/argon2/blake2/blake2-impl.h
Normal file
154
src/crypto/argon2_hasher/hash/argon2/blake2/blake2-impl.h
Normal file
|
|
@ -0,0 +1,154 @@
|
|||
/*
|
||||
* Argon2 reference source code package - reference C implementations
|
||||
*
|
||||
* Copyright 2015
|
||||
* Daniel Dinu, Dmitry Khovratovich, Jean-Philippe Aumasson, and Samuel Neves
|
||||
*
|
||||
* You may use this work under the terms of a Creative Commons CC0 1.0
|
||||
* License/Waiver or the Apache Public License 2.0, at your option. The terms of
|
||||
* these licenses can be found at:
|
||||
*
|
||||
* - CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
* - Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* You should have received a copy of both of these licenses along with this
|
||||
* software. If not, they may be obtained at the above URLs.
|
||||
*/
|
||||
|
||||
#ifndef PORTABLE_BLAKE2_IMPL_H
|
||||
#define PORTABLE_BLAKE2_IMPL_H
|
||||
|
||||
#include <stdint.h>
|
||||
#include <string.h>
|
||||
|
||||
#if defined(_MSC_VER)
|
||||
#define BLAKE2_INLINE __inline
|
||||
#elif defined(__GNUC__) || defined(__clang__)
|
||||
#define BLAKE2_INLINE __inline__
|
||||
#else
|
||||
#define BLAKE2_INLINE
|
||||
#endif
|
||||
|
||||
/* Argon2 Team - Begin Code */
|
||||
/*
|
||||
Not an exhaustive list, but should cover the majority of modern platforms
|
||||
Additionally, the code will always be correct---this is only a performance
|
||||
tweak.
|
||||
*/
|
||||
#if (defined(__BYTE_ORDER__) && \
|
||||
(__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)) || \
|
||||
defined(__LITTLE_ENDIAN__) || defined(__ARMEL__) || defined(__MIPSEL__) || \
|
||||
defined(__AARCH64EL__) || defined(__amd64__) || defined(__i386__) || \
|
||||
defined(_M_IX86) || defined(_M_X64) || defined(_M_AMD64) || \
|
||||
defined(_M_ARM)
|
||||
#define NATIVE_LITTLE_ENDIAN
|
||||
#endif
|
||||
/* Argon2 Team - End Code */
|
||||
|
||||
static BLAKE2_INLINE uint32_t load32(const void *src) {
|
||||
#if defined(NATIVE_LITTLE_ENDIAN)
|
||||
uint32_t w;
|
||||
memcpy(&w, src, sizeof w);
|
||||
return w;
|
||||
#else
|
||||
const uint8_t *p = (const uint8_t *)src;
|
||||
uint32_t w = *p++;
|
||||
w |= (uint32_t)(*p++) << 8;
|
||||
w |= (uint32_t)(*p++) << 16;
|
||||
w |= (uint32_t)(*p++) << 24;
|
||||
return w;
|
||||
#endif
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE uint64_t load64(const void *src) {
|
||||
#if defined(NATIVE_LITTLE_ENDIAN)
|
||||
uint64_t w;
|
||||
memcpy(&w, src, sizeof w);
|
||||
return w;
|
||||
#else
|
||||
const uint8_t *p = (const uint8_t *)src;
|
||||
uint64_t w = *p++;
|
||||
w |= (uint64_t)(*p++) << 8;
|
||||
w |= (uint64_t)(*p++) << 16;
|
||||
w |= (uint64_t)(*p++) << 24;
|
||||
w |= (uint64_t)(*p++) << 32;
|
||||
w |= (uint64_t)(*p++) << 40;
|
||||
w |= (uint64_t)(*p++) << 48;
|
||||
w |= (uint64_t)(*p++) << 56;
|
||||
return w;
|
||||
#endif
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void store32(void *dst, uint32_t w) {
|
||||
#if defined(NATIVE_LITTLE_ENDIAN)
|
||||
memcpy(dst, &w, sizeof w);
|
||||
#else
|
||||
uint8_t *p = (uint8_t *)dst;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
#endif
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void store64(void *dst, uint64_t w) {
|
||||
#if defined(NATIVE_LITTLE_ENDIAN)
|
||||
memcpy(dst, &w, sizeof w);
|
||||
#else
|
||||
uint8_t *p = (uint8_t *)dst;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
#endif
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE uint64_t load48(const void *src) {
|
||||
const uint8_t *p = (const uint8_t *)src;
|
||||
uint64_t w = *p++;
|
||||
w |= (uint64_t)(*p++) << 8;
|
||||
w |= (uint64_t)(*p++) << 16;
|
||||
w |= (uint64_t)(*p++) << 24;
|
||||
w |= (uint64_t)(*p++) << 32;
|
||||
w |= (uint64_t)(*p++) << 40;
|
||||
return w;
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void store48(void *dst, uint64_t w) {
|
||||
uint8_t *p = (uint8_t *)dst;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
w >>= 8;
|
||||
*p++ = (uint8_t)w;
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE uint32_t rotr32(const uint32_t w, const unsigned c) {
|
||||
return (w >> c) | (w << (32 - c));
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE uint64_t rotr64(const uint64_t w, const unsigned c) {
|
||||
return (w >> c) | (w << (64 - c));
|
||||
}
|
||||
|
||||
#endif
|
||||
90
src/crypto/argon2_hasher/hash/argon2/blake2/blake2.h
Normal file
90
src/crypto/argon2_hasher/hash/argon2/blake2/blake2.h
Normal file
|
|
@ -0,0 +1,90 @@
|
|||
/*
|
||||
* Argon2 reference source code package - reference C implementations
|
||||
*
|
||||
* Copyright 2015
|
||||
* Daniel Dinu, Dmitry Khovratovich, Jean-Philippe Aumasson, and Samuel Neves
|
||||
*
|
||||
* You may use this work under the terms of a Creative Commons CC0 1.0
|
||||
* License/Waiver or the Apache Public License 2.0, at your option. The terms of
|
||||
* these licenses can be found at:
|
||||
*
|
||||
* - CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
* - Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* You should have received a copy of both of these licenses along with this
|
||||
* software. If not, they may be obtained at the above URLs.
|
||||
*/
|
||||
|
||||
#ifndef PORTABLE_BLAKE2_H
|
||||
#define PORTABLE_BLAKE2_H
|
||||
|
||||
#include <limits.h>
|
||||
|
||||
#if defined(__cplusplus)
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
enum blake2b_constant {
|
||||
BLAKE2B_BLOCKBYTES = 128,
|
||||
BLAKE2B_OUTBYTES = 64,
|
||||
BLAKE2B_KEYBYTES = 64,
|
||||
BLAKE2B_SALTBYTES = 16,
|
||||
BLAKE2B_PERSONALBYTES = 16
|
||||
};
|
||||
|
||||
#pragma pack(push, 1)
|
||||
typedef struct __blake2b_param {
|
||||
uint8_t digest_length; /* 1 */
|
||||
uint8_t key_length; /* 2 */
|
||||
uint8_t fanout; /* 3 */
|
||||
uint8_t depth; /* 4 */
|
||||
uint32_t leaf_length; /* 8 */
|
||||
uint64_t node_offset; /* 16 */
|
||||
uint8_t node_depth; /* 17 */
|
||||
uint8_t inner_length; /* 18 */
|
||||
uint8_t reserved[14]; /* 32 */
|
||||
uint8_t salt[BLAKE2B_SALTBYTES]; /* 48 */
|
||||
uint8_t personal[BLAKE2B_PERSONALBYTES]; /* 64 */
|
||||
} blake2b_param;
|
||||
#pragma pack(pop)
|
||||
|
||||
typedef struct __blake2b_state {
|
||||
uint64_t h[8];
|
||||
uint64_t t[2];
|
||||
uint64_t f[2];
|
||||
uint8_t buf[BLAKE2B_BLOCKBYTES];
|
||||
unsigned buflen;
|
||||
unsigned outlen;
|
||||
uint8_t last_node;
|
||||
} blake2b_state;
|
||||
|
||||
/* Ensure param structs have not been wrongly padded */
|
||||
/* Poor man's static_assert */
|
||||
enum {
|
||||
blake2_size_check_0 = 1 / !!(CHAR_BIT == 8),
|
||||
blake2_size_check_2 =
|
||||
1 / !!(sizeof(blake2b_param) == sizeof(uint64_t) * CHAR_BIT)
|
||||
};
|
||||
|
||||
/* Streaming API */
|
||||
int blake2b_init(blake2b_state *S, size_t outlen);
|
||||
int blake2b_init_key(blake2b_state *S, size_t outlen, const void *key,
|
||||
size_t keylen);
|
||||
int blake2b_init_param(blake2b_state *S, const blake2b_param *P);
|
||||
int blake2b_update(blake2b_state *S, const void *in, size_t inlen);
|
||||
int blake2b_update_static(blake2b_state *S, const char in, size_t inlen);
|
||||
int blake2b_final(blake2b_state *S, void *out, size_t outlen);
|
||||
|
||||
/* Simple API */
|
||||
int blake2b(void *out, size_t outlen, const void *in, size_t inlen,
|
||||
const void *key, size_t keylen);
|
||||
|
||||
/* Argon2 Team - Begin Code */
|
||||
int blake2b_long(void *out, size_t outlen, const void *in, size_t inlen);
|
||||
/* Argon2 Team - End Code */
|
||||
|
||||
#if defined(__cplusplus)
|
||||
}
|
||||
#endif
|
||||
|
||||
#endif
|
||||
|
|
@ -0,0 +1,68 @@
|
|||
/*
|
||||
BLAKE2 reference source code package - optimized C implementations
|
||||
|
||||
Copyright 2012, Samuel Neves <sneves@dei.uc.pt>. You may use this under the
|
||||
terms of the CC0, the OpenSSL Licence, or the Apache Public License 2.0, at
|
||||
your option. The terms of these licenses can be found at:
|
||||
|
||||
- CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
- OpenSSL license : https://www.openssl.org/source/license.html
|
||||
- Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
More information about the BLAKE2 hash function can be found at
|
||||
https://blake2.net.
|
||||
*/
|
||||
#ifndef BLAKE2B_LOAD_SSE2_H
|
||||
#define BLAKE2B_LOAD_SSE2_H
|
||||
|
||||
#define LOAD_MSG_0_1(b0, b1) b0 = _mm_set_epi64x(m2, m0); b1 = _mm_set_epi64x(m6, m4)
|
||||
#define LOAD_MSG_0_2(b0, b1) b0 = _mm_set_epi64x(m3, m1); b1 = _mm_set_epi64x(m7, m5)
|
||||
#define LOAD_MSG_0_3(b0, b1) b0 = _mm_set_epi64x(m10, m8); b1 = _mm_set_epi64x(m14, m12)
|
||||
#define LOAD_MSG_0_4(b0, b1) b0 = _mm_set_epi64x(m11, m9); b1 = _mm_set_epi64x(m15, m13)
|
||||
#define LOAD_MSG_1_1(b0, b1) b0 = _mm_set_epi64x(m4, m14); b1 = _mm_set_epi64x(m13, m9)
|
||||
#define LOAD_MSG_1_2(b0, b1) b0 = _mm_set_epi64x(m8, m10); b1 = _mm_set_epi64x(m6, m15)
|
||||
#define LOAD_MSG_1_3(b0, b1) b0 = _mm_set_epi64x(m0, m1); b1 = _mm_set_epi64x(m5, m11)
|
||||
#define LOAD_MSG_1_4(b0, b1) b0 = _mm_set_epi64x(m2, m12); b1 = _mm_set_epi64x(m3, m7)
|
||||
#define LOAD_MSG_2_1(b0, b1) b0 = _mm_set_epi64x(m12, m11); b1 = _mm_set_epi64x(m15, m5)
|
||||
#define LOAD_MSG_2_2(b0, b1) b0 = _mm_set_epi64x(m0, m8); b1 = _mm_set_epi64x(m13, m2)
|
||||
#define LOAD_MSG_2_3(b0, b1) b0 = _mm_set_epi64x(m3, m10); b1 = _mm_set_epi64x(m9, m7)
|
||||
#define LOAD_MSG_2_4(b0, b1) b0 = _mm_set_epi64x(m6, m14); b1 = _mm_set_epi64x(m4, m1)
|
||||
#define LOAD_MSG_3_1(b0, b1) b0 = _mm_set_epi64x(m3, m7); b1 = _mm_set_epi64x(m11, m13)
|
||||
#define LOAD_MSG_3_2(b0, b1) b0 = _mm_set_epi64x(m1, m9); b1 = _mm_set_epi64x(m14, m12)
|
||||
#define LOAD_MSG_3_3(b0, b1) b0 = _mm_set_epi64x(m5, m2); b1 = _mm_set_epi64x(m15, m4)
|
||||
#define LOAD_MSG_3_4(b0, b1) b0 = _mm_set_epi64x(m10, m6); b1 = _mm_set_epi64x(m8, m0)
|
||||
#define LOAD_MSG_4_1(b0, b1) b0 = _mm_set_epi64x(m5, m9); b1 = _mm_set_epi64x(m10, m2)
|
||||
#define LOAD_MSG_4_2(b0, b1) b0 = _mm_set_epi64x(m7, m0); b1 = _mm_set_epi64x(m15, m4)
|
||||
#define LOAD_MSG_4_3(b0, b1) b0 = _mm_set_epi64x(m11, m14); b1 = _mm_set_epi64x(m3, m6)
|
||||
#define LOAD_MSG_4_4(b0, b1) b0 = _mm_set_epi64x(m12, m1); b1 = _mm_set_epi64x(m13, m8)
|
||||
#define LOAD_MSG_5_1(b0, b1) b0 = _mm_set_epi64x(m6, m2); b1 = _mm_set_epi64x(m8, m0)
|
||||
#define LOAD_MSG_5_2(b0, b1) b0 = _mm_set_epi64x(m10, m12); b1 = _mm_set_epi64x(m3, m11)
|
||||
#define LOAD_MSG_5_3(b0, b1) b0 = _mm_set_epi64x(m7, m4); b1 = _mm_set_epi64x(m1, m15)
|
||||
#define LOAD_MSG_5_4(b0, b1) b0 = _mm_set_epi64x(m5, m13); b1 = _mm_set_epi64x(m9, m14)
|
||||
#define LOAD_MSG_6_1(b0, b1) b0 = _mm_set_epi64x(m1, m12); b1 = _mm_set_epi64x(m4, m14)
|
||||
#define LOAD_MSG_6_2(b0, b1) b0 = _mm_set_epi64x(m15, m5); b1 = _mm_set_epi64x(m10, m13)
|
||||
#define LOAD_MSG_6_3(b0, b1) b0 = _mm_set_epi64x(m6, m0); b1 = _mm_set_epi64x(m8, m9)
|
||||
#define LOAD_MSG_6_4(b0, b1) b0 = _mm_set_epi64x(m3, m7); b1 = _mm_set_epi64x(m11, m2)
|
||||
#define LOAD_MSG_7_1(b0, b1) b0 = _mm_set_epi64x(m7, m13); b1 = _mm_set_epi64x(m3, m12)
|
||||
#define LOAD_MSG_7_2(b0, b1) b0 = _mm_set_epi64x(m14, m11); b1 = _mm_set_epi64x(m9, m1)
|
||||
#define LOAD_MSG_7_3(b0, b1) b0 = _mm_set_epi64x(m15, m5); b1 = _mm_set_epi64x(m2, m8)
|
||||
#define LOAD_MSG_7_4(b0, b1) b0 = _mm_set_epi64x(m4, m0); b1 = _mm_set_epi64x(m10, m6)
|
||||
#define LOAD_MSG_8_1(b0, b1) b0 = _mm_set_epi64x(m14, m6); b1 = _mm_set_epi64x(m0, m11)
|
||||
#define LOAD_MSG_8_2(b0, b1) b0 = _mm_set_epi64x(m9, m15); b1 = _mm_set_epi64x(m8, m3)
|
||||
#define LOAD_MSG_8_3(b0, b1) b0 = _mm_set_epi64x(m13, m12); b1 = _mm_set_epi64x(m10, m1)
|
||||
#define LOAD_MSG_8_4(b0, b1) b0 = _mm_set_epi64x(m7, m2); b1 = _mm_set_epi64x(m5, m4)
|
||||
#define LOAD_MSG_9_1(b0, b1) b0 = _mm_set_epi64x(m8, m10); b1 = _mm_set_epi64x(m1, m7)
|
||||
#define LOAD_MSG_9_2(b0, b1) b0 = _mm_set_epi64x(m4, m2); b1 = _mm_set_epi64x(m5, m6)
|
||||
#define LOAD_MSG_9_3(b0, b1) b0 = _mm_set_epi64x(m9, m15); b1 = _mm_set_epi64x(m13, m3)
|
||||
#define LOAD_MSG_9_4(b0, b1) b0 = _mm_set_epi64x(m14, m11); b1 = _mm_set_epi64x(m0, m12)
|
||||
#define LOAD_MSG_10_1(b0, b1) b0 = _mm_set_epi64x(m2, m0); b1 = _mm_set_epi64x(m6, m4)
|
||||
#define LOAD_MSG_10_2(b0, b1) b0 = _mm_set_epi64x(m3, m1); b1 = _mm_set_epi64x(m7, m5)
|
||||
#define LOAD_MSG_10_3(b0, b1) b0 = _mm_set_epi64x(m10, m8); b1 = _mm_set_epi64x(m14, m12)
|
||||
#define LOAD_MSG_10_4(b0, b1) b0 = _mm_set_epi64x(m11, m9); b1 = _mm_set_epi64x(m15, m13)
|
||||
#define LOAD_MSG_11_1(b0, b1) b0 = _mm_set_epi64x(m4, m14); b1 = _mm_set_epi64x(m13, m9)
|
||||
#define LOAD_MSG_11_2(b0, b1) b0 = _mm_set_epi64x(m8, m10); b1 = _mm_set_epi64x(m6, m15)
|
||||
#define LOAD_MSG_11_3(b0, b1) b0 = _mm_set_epi64x(m0, m1); b1 = _mm_set_epi64x(m5, m11)
|
||||
#define LOAD_MSG_11_4(b0, b1) b0 = _mm_set_epi64x(m2, m12); b1 = _mm_set_epi64x(m3, m7)
|
||||
|
||||
|
||||
#endif
|
||||
402
src/crypto/argon2_hasher/hash/argon2/blake2/blake2b-load-sse41.h
Normal file
402
src/crypto/argon2_hasher/hash/argon2/blake2/blake2b-load-sse41.h
Normal file
|
|
@ -0,0 +1,402 @@
|
|||
/*
|
||||
BLAKE2 reference source code package - optimized C implementations
|
||||
|
||||
Copyright 2012, Samuel Neves <sneves@dei.uc.pt>. You may use this under the
|
||||
terms of the CC0, the OpenSSL Licence, or the Apache Public License 2.0, at
|
||||
your option. The terms of these licenses can be found at:
|
||||
|
||||
- CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
- OpenSSL license : https://www.openssl.org/source/license.html
|
||||
- Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
More information about the BLAKE2 hash function can be found at
|
||||
https://blake2.net.
|
||||
*/
|
||||
#ifndef BLAKE2B_LOAD_SSE41_H
|
||||
#define BLAKE2B_LOAD_SSE41_H
|
||||
|
||||
#define LOAD_MSG_0_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m0, m1); \
|
||||
b1 = _mm_unpacklo_epi64(m2, m3); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_0_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m0, m1); \
|
||||
b1 = _mm_unpackhi_epi64(m2, m3); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_0_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m4, m5); \
|
||||
b1 = _mm_unpacklo_epi64(m6, m7); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_0_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m4, m5); \
|
||||
b1 = _mm_unpackhi_epi64(m6, m7); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_1_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m7, m2); \
|
||||
b1 = _mm_unpackhi_epi64(m4, m6); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_1_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m5, m4); \
|
||||
b1 = _mm_alignr_epi8(m3, m7, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_1_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_shuffle_epi32(m0, _MM_SHUFFLE(1,0,3,2)); \
|
||||
b1 = _mm_unpackhi_epi64(m5, m2); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_1_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m6, m1); \
|
||||
b1 = _mm_unpackhi_epi64(m3, m1); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_2_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_alignr_epi8(m6, m5, 8); \
|
||||
b1 = _mm_unpackhi_epi64(m2, m7); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_2_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m4, m0); \
|
||||
b1 = _mm_blend_epi16(m1, m6, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_2_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m5, m1, 0xF0); \
|
||||
b1 = _mm_unpackhi_epi64(m3, m4); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_2_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m7, m3); \
|
||||
b1 = _mm_alignr_epi8(m2, m0, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_3_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m3, m1); \
|
||||
b1 = _mm_unpackhi_epi64(m6, m5); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_3_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m4, m0); \
|
||||
b1 = _mm_unpacklo_epi64(m6, m7); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_3_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m1, m2, 0xF0); \
|
||||
b1 = _mm_blend_epi16(m2, m7, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_3_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m3, m5); \
|
||||
b1 = _mm_unpacklo_epi64(m0, m4); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_4_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m4, m2); \
|
||||
b1 = _mm_unpacklo_epi64(m1, m5); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_4_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m0, m3, 0xF0); \
|
||||
b1 = _mm_blend_epi16(m2, m7, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_4_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m7, m5, 0xF0); \
|
||||
b1 = _mm_blend_epi16(m3, m1, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_4_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_alignr_epi8(m6, m0, 8); \
|
||||
b1 = _mm_blend_epi16(m4, m6, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_5_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m1, m3); \
|
||||
b1 = _mm_unpacklo_epi64(m0, m4); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_5_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m6, m5); \
|
||||
b1 = _mm_unpackhi_epi64(m5, m1); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_5_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m2, m3, 0xF0); \
|
||||
b1 = _mm_unpackhi_epi64(m7, m0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_5_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m6, m2); \
|
||||
b1 = _mm_blend_epi16(m7, m4, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_6_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m6, m0, 0xF0); \
|
||||
b1 = _mm_unpacklo_epi64(m7, m2); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_6_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m2, m7); \
|
||||
b1 = _mm_alignr_epi8(m5, m6, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_6_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m0, m3); \
|
||||
b1 = _mm_shuffle_epi32(m4, _MM_SHUFFLE(1,0,3,2)); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_6_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m3, m1); \
|
||||
b1 = _mm_blend_epi16(m1, m5, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_7_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m6, m3); \
|
||||
b1 = _mm_blend_epi16(m6, m1, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_7_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_alignr_epi8(m7, m5, 8); \
|
||||
b1 = _mm_unpackhi_epi64(m0, m4); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_7_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m2, m7); \
|
||||
b1 = _mm_unpacklo_epi64(m4, m1); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_7_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m0, m2); \
|
||||
b1 = _mm_unpacklo_epi64(m3, m5); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_8_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m3, m7); \
|
||||
b1 = _mm_alignr_epi8(m0, m5, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_8_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m7, m4); \
|
||||
b1 = _mm_alignr_epi8(m4, m1, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_8_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = m6; \
|
||||
b1 = _mm_alignr_epi8(m5, m0, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_8_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_blend_epi16(m1, m3, 0xF0); \
|
||||
b1 = m2; \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_9_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m5, m4); \
|
||||
b1 = _mm_unpackhi_epi64(m3, m0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_9_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m1, m2); \
|
||||
b1 = _mm_blend_epi16(m3, m2, 0xF0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_9_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m7, m4); \
|
||||
b1 = _mm_unpackhi_epi64(m1, m6); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_9_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_alignr_epi8(m7, m5, 8); \
|
||||
b1 = _mm_unpacklo_epi64(m6, m0); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_10_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m0, m1); \
|
||||
b1 = _mm_unpacklo_epi64(m2, m3); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_10_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m0, m1); \
|
||||
b1 = _mm_unpackhi_epi64(m2, m3); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_10_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m4, m5); \
|
||||
b1 = _mm_unpacklo_epi64(m6, m7); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_10_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpackhi_epi64(m4, m5); \
|
||||
b1 = _mm_unpackhi_epi64(m6, m7); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_11_1(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m7, m2); \
|
||||
b1 = _mm_unpackhi_epi64(m4, m6); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_11_2(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m5, m4); \
|
||||
b1 = _mm_alignr_epi8(m3, m7, 8); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_11_3(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_shuffle_epi32(m0, _MM_SHUFFLE(1,0,3,2)); \
|
||||
b1 = _mm_unpackhi_epi64(m5, m2); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#define LOAD_MSG_11_4(b0, b1) \
|
||||
do \
|
||||
{ \
|
||||
b0 = _mm_unpacklo_epi64(m6, m1); \
|
||||
b1 = _mm_unpackhi_epi64(m3, m1); \
|
||||
} while(0)
|
||||
|
||||
|
||||
#endif
|
||||
154
src/crypto/argon2_hasher/hash/argon2/blake2/blake2b-round.h
Normal file
154
src/crypto/argon2_hasher/hash/argon2/blake2/blake2b-round.h
Normal file
|
|
@ -0,0 +1,154 @@
|
|||
/*
|
||||
BLAKE2 reference source code package - optimized C implementations
|
||||
|
||||
Copyright 2012, Samuel Neves <sneves@dei.uc.pt>. You may use this under the
|
||||
terms of the CC0, the OpenSSL Licence, or the Apache Public License 2.0, at
|
||||
your option. The terms of these licenses can be found at:
|
||||
|
||||
- CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
- OpenSSL license : https://www.openssl.org/source/license.html
|
||||
- Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
More information about the BLAKE2 hash function can be found at
|
||||
https://blake2.net.
|
||||
*/
|
||||
#ifndef BLAKE2B_ROUND_H
|
||||
#define BLAKE2B_ROUND_H
|
||||
|
||||
#define LOADU(p) _mm_loadu_si128( (const __m128i *)(p) )
|
||||
#define STOREU(p,r) _mm_storeu_si128((__m128i *)(p), r)
|
||||
|
||||
#define TOF(reg) _mm_castsi128_ps((reg))
|
||||
#define TOI(reg) _mm_castps_si128((reg))
|
||||
|
||||
#define LIKELY(x) __builtin_expect((x),1)
|
||||
|
||||
/* Microarchitecture-specific macros */
|
||||
#ifndef HAVE_XOP
|
||||
#ifdef HAVE_SSSE3
|
||||
#define _mm_roti_epi64(x, c) \
|
||||
(-(c) == 32) ? _mm_shuffle_epi32((x), _MM_SHUFFLE(2,3,0,1)) \
|
||||
: (-(c) == 24) ? _mm_shuffle_epi8((x), r24) \
|
||||
: (-(c) == 16) ? _mm_shuffle_epi8((x), r16) \
|
||||
: (-(c) == 63) ? _mm_xor_si128(_mm_srli_epi64((x), -(c)), _mm_add_epi64((x), (x))) \
|
||||
: _mm_xor_si128(_mm_srli_epi64((x), -(c)), _mm_slli_epi64((x), 64-(-(c))))
|
||||
#else
|
||||
#define _mm_roti_epi64(r, c) _mm_xor_si128(_mm_srli_epi64( (r), -(c) ),_mm_slli_epi64( (r), 64-(-(c)) ))
|
||||
#endif
|
||||
#else
|
||||
/* ... */
|
||||
#endif
|
||||
|
||||
#define G1(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1) \
|
||||
row1l = _mm_add_epi64(_mm_add_epi64(row1l, b0), row2l); \
|
||||
row1h = _mm_add_epi64(_mm_add_epi64(row1h, b1), row2h); \
|
||||
\
|
||||
row4l = _mm_xor_si128(row4l, row1l); \
|
||||
row4h = _mm_xor_si128(row4h, row1h); \
|
||||
\
|
||||
row4l = _mm_roti_epi64(row4l, -32); \
|
||||
row4h = _mm_roti_epi64(row4h, -32); \
|
||||
\
|
||||
row3l = _mm_add_epi64(row3l, row4l); \
|
||||
row3h = _mm_add_epi64(row3h, row4h); \
|
||||
\
|
||||
row2l = _mm_xor_si128(row2l, row3l); \
|
||||
row2h = _mm_xor_si128(row2h, row3h); \
|
||||
\
|
||||
row2l = _mm_roti_epi64(row2l, -24); \
|
||||
row2h = _mm_roti_epi64(row2h, -24); \
|
||||
|
||||
#define G2(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1) \
|
||||
row1l = _mm_add_epi64(_mm_add_epi64(row1l, b0), row2l); \
|
||||
row1h = _mm_add_epi64(_mm_add_epi64(row1h, b1), row2h); \
|
||||
\
|
||||
row4l = _mm_xor_si128(row4l, row1l); \
|
||||
row4h = _mm_xor_si128(row4h, row1h); \
|
||||
\
|
||||
row4l = _mm_roti_epi64(row4l, -16); \
|
||||
row4h = _mm_roti_epi64(row4h, -16); \
|
||||
\
|
||||
row3l = _mm_add_epi64(row3l, row4l); \
|
||||
row3h = _mm_add_epi64(row3h, row4h); \
|
||||
\
|
||||
row2l = _mm_xor_si128(row2l, row3l); \
|
||||
row2h = _mm_xor_si128(row2h, row3h); \
|
||||
\
|
||||
row2l = _mm_roti_epi64(row2l, -63); \
|
||||
row2h = _mm_roti_epi64(row2h, -63); \
|
||||
|
||||
#if defined(HAVE_SSSE3)
|
||||
#define DIAGONALIZE(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h) \
|
||||
t0 = _mm_alignr_epi8(row2h, row2l, 8); \
|
||||
t1 = _mm_alignr_epi8(row2l, row2h, 8); \
|
||||
row2l = t0; \
|
||||
row2h = t1; \
|
||||
\
|
||||
t0 = row3l; \
|
||||
row3l = row3h; \
|
||||
row3h = t0; \
|
||||
\
|
||||
t0 = _mm_alignr_epi8(row4h, row4l, 8); \
|
||||
t1 = _mm_alignr_epi8(row4l, row4h, 8); \
|
||||
row4l = t1; \
|
||||
row4h = t0;
|
||||
|
||||
#define UNDIAGONALIZE(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h) \
|
||||
t0 = _mm_alignr_epi8(row2l, row2h, 8); \
|
||||
t1 = _mm_alignr_epi8(row2h, row2l, 8); \
|
||||
row2l = t0; \
|
||||
row2h = t1; \
|
||||
\
|
||||
t0 = row3l; \
|
||||
row3l = row3h; \
|
||||
row3h = t0; \
|
||||
\
|
||||
t0 = _mm_alignr_epi8(row4l, row4h, 8); \
|
||||
t1 = _mm_alignr_epi8(row4h, row4l, 8); \
|
||||
row4l = t1; \
|
||||
row4h = t0;
|
||||
#else
|
||||
|
||||
#define DIAGONALIZE(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h) \
|
||||
t0 = row4l;\
|
||||
t1 = row2l;\
|
||||
row4l = row3l;\
|
||||
row3l = row3h;\
|
||||
row3h = row4l;\
|
||||
row4l = _mm_unpackhi_epi64(row4h, _mm_unpacklo_epi64(t0, t0)); \
|
||||
row4h = _mm_unpackhi_epi64(t0, _mm_unpacklo_epi64(row4h, row4h)); \
|
||||
row2l = _mm_unpackhi_epi64(row2l, _mm_unpacklo_epi64(row2h, row2h)); \
|
||||
row2h = _mm_unpackhi_epi64(row2h, _mm_unpacklo_epi64(t1, t1))
|
||||
|
||||
#define UNDIAGONALIZE(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h) \
|
||||
t0 = row3l;\
|
||||
row3l = row3h;\
|
||||
row3h = t0;\
|
||||
t0 = row2l;\
|
||||
t1 = row4l;\
|
||||
row2l = _mm_unpackhi_epi64(row2h, _mm_unpacklo_epi64(row2l, row2l)); \
|
||||
row2h = _mm_unpackhi_epi64(t0, _mm_unpacklo_epi64(row2h, row2h)); \
|
||||
row4l = _mm_unpackhi_epi64(row4l, _mm_unpacklo_epi64(row4h, row4h)); \
|
||||
row4h = _mm_unpackhi_epi64(row4h, _mm_unpacklo_epi64(t1, t1))
|
||||
|
||||
#endif
|
||||
|
||||
#if defined(HAVE_SSE41)
|
||||
#include "blake2b-load-sse41.h"
|
||||
#else
|
||||
#include "blake2b-load-sse2.h"
|
||||
#endif
|
||||
|
||||
#define ROUND(r) \
|
||||
LOAD_MSG_ ##r ##_1(b0, b1); \
|
||||
G1(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1); \
|
||||
LOAD_MSG_ ##r ##_2(b0, b1); \
|
||||
G2(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1); \
|
||||
DIAGONALIZE(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h); \
|
||||
LOAD_MSG_ ##r ##_3(b0, b1); \
|
||||
G1(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1); \
|
||||
LOAD_MSG_ ##r ##_4(b0, b1); \
|
||||
G2(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1); \
|
||||
UNDIAGONALIZE(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h);
|
||||
|
||||
#endif
|
||||
514
src/crypto/argon2_hasher/hash/argon2/blake2/blake2b.c
Normal file
514
src/crypto/argon2_hasher/hash/argon2/blake2/blake2b.c
Normal file
|
|
@ -0,0 +1,514 @@
|
|||
/*
|
||||
* Argon2 reference source code package - reference C implementations
|
||||
*
|
||||
* Copyright 2015
|
||||
* Daniel Dinu, Dmitry Khovratovich, Jean-Philippe Aumasson, and Samuel Neves
|
||||
*
|
||||
* You may use this work under the terms of a Creative Commons CC0 1.0
|
||||
* License/Waiver or the Apache Public License 2.0, at your option. The terms of
|
||||
* these licenses can be found at:
|
||||
*
|
||||
* - CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
* - Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* You should have received a copy of both of these licenses along with this
|
||||
* software. If not, they may be obtained at the above URLs.
|
||||
*/
|
||||
|
||||
#include <stdint.h>
|
||||
#include <string.h>
|
||||
#include <stdio.h>
|
||||
|
||||
#include "blake2.h"
|
||||
#include "blake2-impl.h"
|
||||
|
||||
#if !defined(BUILD_REF) && (defined(__x86_64__) || defined(_WIN64))
|
||||
#include "blake2-config.h"
|
||||
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h> /* for _mm_set_epi64x */
|
||||
#endif
|
||||
#include <emmintrin.h>
|
||||
#if defined(HAVE_SSSE3)
|
||||
#include <tmmintrin.h>
|
||||
#endif
|
||||
#if defined(HAVE_SSE41)
|
||||
#include <smmintrin.h>
|
||||
#endif
|
||||
#if defined(HAVE_AVX)
|
||||
#include <immintrin.h>
|
||||
#endif
|
||||
#if defined(HAVE_XOP)
|
||||
#include <x86intrin.h>
|
||||
#endif
|
||||
|
||||
#include "blake2b-round.h"
|
||||
#endif
|
||||
|
||||
static const uint64_t blake2b_IV[8] = {
|
||||
UINT64_C(0x6a09e667f3bcc908), UINT64_C(0xbb67ae8584caa73b),
|
||||
UINT64_C(0x3c6ef372fe94f82b), UINT64_C(0xa54ff53a5f1d36f1),
|
||||
UINT64_C(0x510e527fade682d1), UINT64_C(0x9b05688c2b3e6c1f),
|
||||
UINT64_C(0x1f83d9abfb41bd6b), UINT64_C(0x5be0cd19137e2179)};
|
||||
|
||||
static const unsigned int blake2b_sigma[12][16] = {
|
||||
{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15},
|
||||
{14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3},
|
||||
{11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4},
|
||||
{7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8},
|
||||
{9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13},
|
||||
{2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9},
|
||||
{12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11},
|
||||
{13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10},
|
||||
{6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5},
|
||||
{10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13, 0},
|
||||
{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15},
|
||||
{14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3},
|
||||
};
|
||||
|
||||
static BLAKE2_INLINE void blake2b_set_lastnode(blake2b_state *S) {
|
||||
S->f[1] = (uint64_t)-1;
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void blake2b_set_lastblock(blake2b_state *S) {
|
||||
if (S->last_node) {
|
||||
blake2b_set_lastnode(S);
|
||||
}
|
||||
S->f[0] = (uint64_t)-1;
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void blake2b_increment_counter(blake2b_state *S,
|
||||
uint64_t inc) {
|
||||
S->t[0] += inc;
|
||||
S->t[1] += (S->t[0] < inc);
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void blake2b_invalidate_state(blake2b_state *S) {
|
||||
blake2b_set_lastblock(S); /* invalidate for further use */
|
||||
}
|
||||
|
||||
static BLAKE2_INLINE void blake2b_init0(blake2b_state *S) {
|
||||
memset(S, 0, sizeof(*S));
|
||||
memcpy(S->h, blake2b_IV, sizeof(S->h));
|
||||
}
|
||||
|
||||
int blake2b_init_param(blake2b_state *S, const blake2b_param *P) {
|
||||
const unsigned char *p = (const unsigned char *)P;
|
||||
unsigned int i;
|
||||
|
||||
if (NULL == P || NULL == S) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
blake2b_init0(S);
|
||||
/* IV XOR Parameter Block */
|
||||
for (i = 0; i < 8; ++i) {
|
||||
S->h[i] ^= load64(&p[i * sizeof(S->h[i])]);
|
||||
}
|
||||
S->outlen = P->digest_length;
|
||||
return 0;
|
||||
}
|
||||
|
||||
/* Sequential blake2b initialization */
|
||||
int blake2b_init(blake2b_state *S, size_t outlen) {
|
||||
blake2b_param P;
|
||||
|
||||
if (S == NULL) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
if ((outlen == 0) || (outlen > BLAKE2B_OUTBYTES)) {
|
||||
blake2b_invalidate_state(S);
|
||||
return -1;
|
||||
}
|
||||
|
||||
/* Setup Parameter Block for unkeyed BLAKE2 */
|
||||
P.digest_length = (uint8_t)outlen;
|
||||
P.key_length = 0;
|
||||
P.fanout = 1;
|
||||
P.depth = 1;
|
||||
P.leaf_length = 0;
|
||||
P.node_offset = 0;
|
||||
P.node_depth = 0;
|
||||
P.inner_length = 0;
|
||||
memset(P.reserved, 0, sizeof(P.reserved));
|
||||
memset(P.salt, 0, sizeof(P.salt));
|
||||
memset(P.personal, 0, sizeof(P.personal));
|
||||
|
||||
return blake2b_init_param(S, &P);
|
||||
}
|
||||
|
||||
int blake2b_init_key(blake2b_state *S, size_t outlen, const void *key,
|
||||
size_t keylen) {
|
||||
blake2b_param P;
|
||||
|
||||
if (S == NULL) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
if ((outlen == 0) || (outlen > BLAKE2B_OUTBYTES)) {
|
||||
blake2b_invalidate_state(S);
|
||||
return -1;
|
||||
}
|
||||
|
||||
if ((key == 0) || (keylen == 0) || (keylen > BLAKE2B_KEYBYTES)) {
|
||||
blake2b_invalidate_state(S);
|
||||
return -1;
|
||||
}
|
||||
|
||||
/* Setup Parameter Block for keyed BLAKE2 */
|
||||
P.digest_length = (uint8_t)outlen;
|
||||
P.key_length = (uint8_t)keylen;
|
||||
P.fanout = 1;
|
||||
P.depth = 1;
|
||||
P.leaf_length = 0;
|
||||
P.node_offset = 0;
|
||||
P.node_depth = 0;
|
||||
P.inner_length = 0;
|
||||
memset(P.reserved, 0, sizeof(P.reserved));
|
||||
memset(P.salt, 0, sizeof(P.salt));
|
||||
memset(P.personal, 0, sizeof(P.personal));
|
||||
|
||||
if (blake2b_init_param(S, &P) < 0) {
|
||||
blake2b_invalidate_state(S);
|
||||
return -1;
|
||||
}
|
||||
|
||||
{
|
||||
uint8_t block[BLAKE2B_BLOCKBYTES];
|
||||
memset(block, 0, BLAKE2B_BLOCKBYTES);
|
||||
memcpy(block, key, keylen);
|
||||
blake2b_update(S, block, BLAKE2B_BLOCKBYTES);
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
||||
#if !defined(BUILD_REF) && (defined(__x86_64__) || defined(_WIN64))
|
||||
static void blake2b_compress( blake2b_state *S, const uint8_t block[BLAKE2B_BLOCKBYTES] )
|
||||
{
|
||||
__m128i row1l, row1h;
|
||||
__m128i row2l, row2h;
|
||||
__m128i row3l, row3h;
|
||||
__m128i row4l, row4h;
|
||||
__m128i b0, b1;
|
||||
__m128i t0, t1;
|
||||
#if defined(HAVE_SSSE3) && !defined(HAVE_XOP)
|
||||
const __m128i r16 = _mm_setr_epi8( 2, 3, 4, 5, 6, 7, 0, 1, 10, 11, 12, 13, 14, 15, 8, 9 );
|
||||
const __m128i r24 = _mm_setr_epi8( 3, 4, 5, 6, 7, 0, 1, 2, 11, 12, 13, 14, 15, 8, 9, 10 );
|
||||
#endif
|
||||
#if defined(HAVE_SSE41)
|
||||
const __m128i m0 = LOADU( block + 00 );
|
||||
const __m128i m1 = LOADU( block + 16 );
|
||||
const __m128i m2 = LOADU( block + 32 );
|
||||
const __m128i m3 = LOADU( block + 48 );
|
||||
const __m128i m4 = LOADU( block + 64 );
|
||||
const __m128i m5 = LOADU( block + 80 );
|
||||
const __m128i m6 = LOADU( block + 96 );
|
||||
const __m128i m7 = LOADU( block + 112 );
|
||||
#else
|
||||
const uint64_t m0 = load64(block + 0 * sizeof(uint64_t));
|
||||
const uint64_t m1 = load64(block + 1 * sizeof(uint64_t));
|
||||
const uint64_t m2 = load64(block + 2 * sizeof(uint64_t));
|
||||
const uint64_t m3 = load64(block + 3 * sizeof(uint64_t));
|
||||
const uint64_t m4 = load64(block + 4 * sizeof(uint64_t));
|
||||
const uint64_t m5 = load64(block + 5 * sizeof(uint64_t));
|
||||
const uint64_t m6 = load64(block + 6 * sizeof(uint64_t));
|
||||
const uint64_t m7 = load64(block + 7 * sizeof(uint64_t));
|
||||
const uint64_t m8 = load64(block + 8 * sizeof(uint64_t));
|
||||
const uint64_t m9 = load64(block + 9 * sizeof(uint64_t));
|
||||
const uint64_t m10 = load64(block + 10 * sizeof(uint64_t));
|
||||
const uint64_t m11 = load64(block + 11 * sizeof(uint64_t));
|
||||
const uint64_t m12 = load64(block + 12 * sizeof(uint64_t));
|
||||
const uint64_t m13 = load64(block + 13 * sizeof(uint64_t));
|
||||
const uint64_t m14 = load64(block + 14 * sizeof(uint64_t));
|
||||
const uint64_t m15 = load64(block + 15 * sizeof(uint64_t));
|
||||
#endif
|
||||
row1l = LOADU( &S->h[0] );
|
||||
row1h = LOADU( &S->h[2] );
|
||||
row2l = LOADU( &S->h[4] );
|
||||
row2h = LOADU( &S->h[6] );
|
||||
row3l = LOADU( &blake2b_IV[0] );
|
||||
row3h = LOADU( &blake2b_IV[2] );
|
||||
row4l = _mm_xor_si128( LOADU( &blake2b_IV[4] ), LOADU( &S->t[0] ) );
|
||||
row4h = _mm_xor_si128( LOADU( &blake2b_IV[6] ), LOADU( &S->f[0] ) );
|
||||
ROUND( 0 );
|
||||
ROUND( 1 );
|
||||
ROUND( 2 );
|
||||
ROUND( 3 );
|
||||
ROUND( 4 );
|
||||
ROUND( 5 );
|
||||
ROUND( 6 );
|
||||
ROUND( 7 );
|
||||
ROUND( 8 );
|
||||
ROUND( 9 );
|
||||
ROUND( 10 );
|
||||
ROUND( 11 );
|
||||
row1l = _mm_xor_si128( row3l, row1l );
|
||||
row1h = _mm_xor_si128( row3h, row1h );
|
||||
STOREU( &S->h[0], _mm_xor_si128( LOADU( &S->h[0] ), row1l ) );
|
||||
STOREU( &S->h[2], _mm_xor_si128( LOADU( &S->h[2] ), row1h ) );
|
||||
row2l = _mm_xor_si128( row4l, row2l );
|
||||
row2h = _mm_xor_si128( row4h, row2h );
|
||||
STOREU( &S->h[4], _mm_xor_si128( LOADU( &S->h[4] ), row2l ) );
|
||||
STOREU( &S->h[6], _mm_xor_si128( LOADU( &S->h[6] ), row2h ) );
|
||||
}
|
||||
#else
|
||||
static void blake2b_compress(blake2b_state *S, const uint8_t *block) {
|
||||
uint64_t m[16];
|
||||
uint64_t v[16];
|
||||
unsigned int i, r;
|
||||
|
||||
for (i = 0; i < 16; ++i) {
|
||||
m[i] = load64(block + i * sizeof(m[i]));
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
v[i] = S->h[i];
|
||||
}
|
||||
|
||||
v[8] = blake2b_IV[0];
|
||||
v[9] = blake2b_IV[1];
|
||||
v[10] = blake2b_IV[2];
|
||||
v[11] = blake2b_IV[3];
|
||||
v[12] = blake2b_IV[4] ^ S->t[0];
|
||||
v[13] = blake2b_IV[5] ^ S->t[1];
|
||||
v[14] = blake2b_IV[6] ^ S->f[0];
|
||||
v[15] = blake2b_IV[7] ^ S->f[1];
|
||||
|
||||
#define G(r, i, a, b, c, d) \
|
||||
do { \
|
||||
a = a + b + m[blake2b_sigma[r][2 * i + 0]]; \
|
||||
d = rotr64(d ^ a, 32); \
|
||||
c = c + d; \
|
||||
b = rotr64(b ^ c, 24); \
|
||||
a = a + b + m[blake2b_sigma[r][2 * i + 1]]; \
|
||||
d = rotr64(d ^ a, 16); \
|
||||
c = c + d; \
|
||||
b = rotr64(b ^ c, 63); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define ROUND(r) \
|
||||
do { \
|
||||
G(r, 0, v[0], v[4], v[8], v[12]); \
|
||||
G(r, 1, v[1], v[5], v[9], v[13]); \
|
||||
G(r, 2, v[2], v[6], v[10], v[14]); \
|
||||
G(r, 3, v[3], v[7], v[11], v[15]); \
|
||||
G(r, 4, v[0], v[5], v[10], v[15]); \
|
||||
G(r, 5, v[1], v[6], v[11], v[12]); \
|
||||
G(r, 6, v[2], v[7], v[8], v[13]); \
|
||||
G(r, 7, v[3], v[4], v[9], v[14]); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
for (r = 0; r < 12; ++r) {
|
||||
ROUND(r);
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
S->h[i] = S->h[i] ^ v[i] ^ v[i + 8];
|
||||
}
|
||||
|
||||
#undef G
|
||||
#undef ROUND
|
||||
}
|
||||
#endif
|
||||
|
||||
int blake2b_update(blake2b_state *S, const void *in, size_t inlen) {
|
||||
const uint8_t *pin = (const uint8_t *)in;
|
||||
|
||||
if (inlen == 0) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
/* Sanity check */
|
||||
if (S == NULL || in == NULL) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
/* Is this a reused state? */
|
||||
if (S->f[0] != 0) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
if (S->buflen + inlen > BLAKE2B_BLOCKBYTES) {
|
||||
/* Complete current block */
|
||||
size_t left = S->buflen;
|
||||
size_t fill = BLAKE2B_BLOCKBYTES - left;
|
||||
memcpy(&S->buf[left], pin, fill);
|
||||
blake2b_increment_counter(S, BLAKE2B_BLOCKBYTES);
|
||||
blake2b_compress(S, S->buf);
|
||||
S->buflen = 0;
|
||||
inlen -= fill;
|
||||
pin += fill;
|
||||
/* Avoid buffer copies when possible */
|
||||
while (inlen > BLAKE2B_BLOCKBYTES) {
|
||||
blake2b_increment_counter(S, BLAKE2B_BLOCKBYTES);
|
||||
blake2b_compress(S, pin);
|
||||
inlen -= BLAKE2B_BLOCKBYTES;
|
||||
pin += BLAKE2B_BLOCKBYTES;
|
||||
}
|
||||
}
|
||||
memcpy(&S->buf[S->buflen], pin, inlen);
|
||||
S->buflen += (unsigned int)inlen;
|
||||
return 0;
|
||||
}
|
||||
|
||||
int blake2b_update_static(blake2b_state *S, const char in, size_t inlen) {
|
||||
if (inlen == 0) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
/* Sanity check */
|
||||
if (S == NULL) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
/* Is this a reused state? */
|
||||
if (S->f[0] != 0) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
if (S->buflen + inlen > BLAKE2B_BLOCKBYTES) {
|
||||
/* Complete current block */
|
||||
size_t left = S->buflen;
|
||||
size_t fill = BLAKE2B_BLOCKBYTES - left;
|
||||
memset(&S->buf[left], in, fill);
|
||||
blake2b_increment_counter(S, BLAKE2B_BLOCKBYTES);
|
||||
blake2b_compress(S, S->buf);
|
||||
S->buflen = 0;
|
||||
inlen -= fill;
|
||||
/* Avoid buffer copies when possible */
|
||||
while (inlen > BLAKE2B_BLOCKBYTES) {
|
||||
memset(S->buf, in, BLAKE2B_BLOCKBYTES);
|
||||
blake2b_increment_counter(S, BLAKE2B_BLOCKBYTES);
|
||||
blake2b_compress(S, S->buf);
|
||||
inlen -= BLAKE2B_BLOCKBYTES;
|
||||
}
|
||||
}
|
||||
memset(&S->buf[S->buflen], in, inlen);
|
||||
S->buflen += (unsigned int)inlen;
|
||||
return 0;
|
||||
}
|
||||
|
||||
|
||||
int blake2b_final(blake2b_state *S, void *out, size_t outlen) {
|
||||
uint8_t buffer[BLAKE2B_OUTBYTES] = {0};
|
||||
unsigned int i;
|
||||
|
||||
/* Sanity checks */
|
||||
if (S == NULL || out == NULL || outlen < S->outlen) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
/* Is this a reused state? */
|
||||
if (S->f[0] != 0) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
blake2b_increment_counter(S, S->buflen);
|
||||
blake2b_set_lastblock(S);
|
||||
memset(&S->buf[S->buflen], 0, BLAKE2B_BLOCKBYTES - S->buflen); /* Padding */
|
||||
blake2b_compress(S, S->buf);
|
||||
|
||||
for (i = 0; i < 8; ++i) { /* Output full hash to temp buffer */
|
||||
store64(buffer + sizeof(S->h[i]) * i, S->h[i]);
|
||||
}
|
||||
|
||||
memcpy(out, buffer, S->outlen);
|
||||
return 0;
|
||||
}
|
||||
|
||||
int blake2b(void *out, size_t outlen, const void *in, size_t inlen,
|
||||
const void *key, size_t keylen) {
|
||||
blake2b_state S;
|
||||
int ret = -1;
|
||||
|
||||
/* Verify parameters */
|
||||
if (NULL == in && inlen > 0) {
|
||||
goto fail;
|
||||
}
|
||||
|
||||
if (NULL == out || outlen == 0 || outlen > BLAKE2B_OUTBYTES) {
|
||||
goto fail;
|
||||
}
|
||||
|
||||
if ((NULL == key && keylen > 0) || keylen > BLAKE2B_KEYBYTES) {
|
||||
goto fail;
|
||||
}
|
||||
|
||||
if (keylen > 0) {
|
||||
if (blake2b_init_key(&S, outlen, key, keylen) < 0) {
|
||||
goto fail;
|
||||
}
|
||||
} else {
|
||||
if (blake2b_init(&S, outlen) < 0) {
|
||||
goto fail;
|
||||
}
|
||||
}
|
||||
|
||||
if (blake2b_update(&S, in, inlen) < 0) {
|
||||
goto fail;
|
||||
}
|
||||
ret = blake2b_final(&S, out, outlen);
|
||||
|
||||
fail:
|
||||
return ret;
|
||||
}
|
||||
|
||||
/* Argon2 Team - Begin Code */
|
||||
int blake2b_long(void *pout, size_t outlen, const void *in, size_t inlen) {
|
||||
uint8_t *out = (uint8_t *)pout;
|
||||
blake2b_state blake_state;
|
||||
uint8_t outlen_bytes[sizeof(uint32_t)] = {0};
|
||||
int ret = -1;
|
||||
|
||||
if (outlen > UINT32_MAX) {
|
||||
goto fail;
|
||||
}
|
||||
|
||||
/* Ensure little-endian byte order! */
|
||||
store32(outlen_bytes, (uint32_t)outlen);
|
||||
|
||||
#define TRY(statement) \
|
||||
do { \
|
||||
ret = statement; \
|
||||
if (ret < 0) { \
|
||||
goto fail; \
|
||||
} \
|
||||
} while ((void)0, 0)
|
||||
|
||||
if (outlen <= BLAKE2B_OUTBYTES) {
|
||||
TRY(blake2b_init(&blake_state, outlen));
|
||||
TRY(blake2b_update(&blake_state, outlen_bytes, sizeof(outlen_bytes)));
|
||||
TRY(blake2b_update(&blake_state, in, inlen));
|
||||
TRY(blake2b_final(&blake_state, out, outlen));
|
||||
} else {
|
||||
uint32_t toproduce;
|
||||
uint8_t out_buffer[BLAKE2B_OUTBYTES];
|
||||
uint8_t in_buffer[BLAKE2B_OUTBYTES];
|
||||
TRY(blake2b_init(&blake_state, BLAKE2B_OUTBYTES));
|
||||
TRY(blake2b_update(&blake_state, outlen_bytes, sizeof(outlen_bytes)));
|
||||
TRY(blake2b_update(&blake_state, in, inlen));
|
||||
TRY(blake2b_final(&blake_state, out_buffer, BLAKE2B_OUTBYTES));
|
||||
memcpy(out, out_buffer, BLAKE2B_OUTBYTES / 2);
|
||||
out += BLAKE2B_OUTBYTES / 2;
|
||||
toproduce = (uint32_t)outlen - BLAKE2B_OUTBYTES / 2;
|
||||
|
||||
while (toproduce > BLAKE2B_OUTBYTES) {
|
||||
memcpy(in_buffer, out_buffer, BLAKE2B_OUTBYTES);
|
||||
TRY(blake2b(out_buffer, BLAKE2B_OUTBYTES, in_buffer,
|
||||
BLAKE2B_OUTBYTES, NULL, 0));
|
||||
memcpy(out, out_buffer, BLAKE2B_OUTBYTES / 2);
|
||||
out += BLAKE2B_OUTBYTES / 2;
|
||||
toproduce -= BLAKE2B_OUTBYTES / 2;
|
||||
}
|
||||
|
||||
memcpy(in_buffer, out_buffer, BLAKE2B_OUTBYTES);
|
||||
TRY(blake2b(out_buffer, toproduce, in_buffer, BLAKE2B_OUTBYTES, NULL,
|
||||
0));
|
||||
memcpy(out, out_buffer, toproduce);
|
||||
}
|
||||
fail:
|
||||
return ret;
|
||||
#undef TRY
|
||||
}
|
||||
/* Argon2 Team - End Code */
|
||||
227
src/crypto/argon2_hasher/hash/cpu/CpuHasher.cpp
Executable file
227
src/crypto/argon2_hasher/hash/cpu/CpuHasher.cpp
Executable file
|
|
@ -0,0 +1,227 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#if defined(__x86_64__) || defined(__i386__) || defined(_WIN64)
|
||||
#include <cpuinfo_x86.h>
|
||||
#endif
|
||||
#if defined(__arm__)
|
||||
#include <cpuinfo_arm.h>
|
||||
#endif
|
||||
|
||||
#include <crypto/Argon2_constants.h>
|
||||
|
||||
#include "../../common/common.h"
|
||||
|
||||
#include "crypto/argon2_hasher/hash/Hasher.h"
|
||||
#include "crypto/argon2_hasher/hash/argon2/Argon2.h"
|
||||
|
||||
#include "CpuHasher.h"
|
||||
#include "crypto/argon2_hasher/common/DLLExport.h"
|
||||
|
||||
CpuHasher::CpuHasher() : Hasher() {
|
||||
m_type = "CPU";
|
||||
m_subType = "CPU";
|
||||
m_shortSubType = "CPU";
|
||||
m_optimization = "REF";
|
||||
m_computingThreads = 0;
|
||||
m_availableProcessingThr = 1;
|
||||
m_availableMemoryThr = 1;
|
||||
m_argon2BlocksFillerPtr = nullptr;
|
||||
m_dllHandle = nullptr;
|
||||
m_profile = nullptr;
|
||||
m_threadData = nullptr;
|
||||
}
|
||||
|
||||
CpuHasher::~CpuHasher() {
|
||||
this->cleanup();
|
||||
}
|
||||
|
||||
bool CpuHasher::initialize(xmrig::Algo algorithm, xmrig::Variant variant) {
|
||||
m_profile = getArgon2Profile(algorithm, variant);
|
||||
m_description = detectFeaturesAndMakeDescription();
|
||||
return true;
|
||||
}
|
||||
|
||||
bool CpuHasher::configure(xmrig::HasherConfig &config) {
|
||||
m_intensity = 100;
|
||||
|
||||
if(config.cpuOptimization() != "") {
|
||||
m_description += "Overiding detected optimization feature with " + config.cpuOptimization() + ".\n";
|
||||
m_optimization = config.cpuOptimization();
|
||||
}
|
||||
|
||||
loadArgon2BlockFiller();
|
||||
|
||||
if(m_argon2BlocksFillerPtr == NULL) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - argon2 hashing module not found.";
|
||||
return false;
|
||||
}
|
||||
|
||||
m_computingThreads = min(m_availableProcessingThr, m_availableMemoryThr);
|
||||
|
||||
if (m_computingThreads == 0) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - not enough resources.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if(config.cpuThreads() > -1) {
|
||||
m_intensity = min(100.0 * config.cpuThreads() / m_computingThreads, 100.0);
|
||||
m_computingThreads = min(config.cpuThreads(), m_computingThreads);
|
||||
}
|
||||
|
||||
if (m_intensity == 0) {
|
||||
m_description += "Status: DISABLED - by user.";
|
||||
return false;
|
||||
}
|
||||
|
||||
m_deviceInfo.intensity = m_intensity;
|
||||
|
||||
storeDeviceInfo(0, m_deviceInfo);
|
||||
|
||||
m_threadData = new CpuHasherThread[m_computingThreads];
|
||||
for(int i=0; i < m_computingThreads; i++) {
|
||||
void *buffer = NULL;
|
||||
void *mem = allocateMemory(buffer);
|
||||
if(mem == NULL) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - error allocating memory.";
|
||||
return false;
|
||||
}
|
||||
m_threadData[i].mem = buffer;
|
||||
m_threadData[i].argon2 = new Argon2(NULL, m_argon2BlocksFillerPtr, NULL, mem, mem);
|
||||
m_threadData[i].hashData.outSize = xmrig::ARGON2_HASHLEN + sizeof(uint32_t);
|
||||
}
|
||||
|
||||
m_description += "Status: ENABLED - with " + to_string(m_computingThreads) + " threads.";
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
string CpuHasher::detectFeaturesAndMakeDescription() {
|
||||
stringstream ss;
|
||||
#if defined(__x86_64__) || defined(__i386__) || defined(_WIN64)
|
||||
char brand_string[49];
|
||||
cpu_features::FillX86BrandString(brand_string);
|
||||
m_deviceInfo.name = brand_string;
|
||||
|
||||
ss << brand_string << endl;
|
||||
|
||||
cpu_features::X86Features features = cpu_features::GetX86Info().features;
|
||||
ss << "Optimization features: ";
|
||||
|
||||
#if defined(__x86_64__) || defined(_WIN64)
|
||||
ss << "SSE2 ";
|
||||
m_optimization = "SSE2";
|
||||
#else
|
||||
ss << "none";
|
||||
m_optimization = "REF";
|
||||
#endif
|
||||
|
||||
if(features.ssse3 || features.avx2 || features.avx512f) {
|
||||
if (features.ssse3) {
|
||||
ss << "SSSE3 ";
|
||||
m_optimization = "SSSE3";
|
||||
}
|
||||
if (features.avx) {
|
||||
ss << "AVX ";
|
||||
m_optimization = "AVX";
|
||||
}
|
||||
if (features.avx2) {
|
||||
ss << "AVX2 ";
|
||||
m_optimization = "AVX2";
|
||||
}
|
||||
if (features.avx512f) {
|
||||
ss << "AVX512F ";
|
||||
m_optimization = "AVX512F";
|
||||
}
|
||||
}
|
||||
ss << endl;
|
||||
#endif
|
||||
#if defined(__arm__)
|
||||
m_deviceInfo.name = "ARM processor";
|
||||
|
||||
cpu_features::ArmFeatures features = cpu_features::GetArmInfo().features;
|
||||
ss << "ARM processor" << endl;
|
||||
ss << "Optimization features: ";
|
||||
|
||||
m_optimization = "REF";
|
||||
|
||||
if(features.neon) {
|
||||
ss << "NEON";
|
||||
m_optimization = "NEON";
|
||||
}
|
||||
else {
|
||||
ss << "none";
|
||||
}
|
||||
ss << endl;
|
||||
#endif
|
||||
ss << "Selecting " << m_optimization << " as candidate for hashing algorithm." << endl;
|
||||
|
||||
m_availableProcessingThr = thread::hardware_concurrency();
|
||||
ss << "Parallelism: " << m_availableProcessingThr << " concurent threads supported." << endl;
|
||||
|
||||
//check available memory
|
||||
vector<void *> memoryTest;
|
||||
for(m_availableMemoryThr = 0;m_availableMemoryThr < m_availableProcessingThr;m_availableMemoryThr++) {
|
||||
void *memory = malloc(m_profile->memSize + 64); //64 bytes for alignament - to work on AVX512F optimisations
|
||||
if(memory == NULL)
|
||||
break;
|
||||
memoryTest.push_back(memory);
|
||||
}
|
||||
for(vector<void*>::iterator it=memoryTest.begin(); it != memoryTest.end(); ++it) {
|
||||
free(*it);
|
||||
}
|
||||
ss << "Memory: there is enough memory for " << m_availableMemoryThr << " concurent threads." << endl;
|
||||
|
||||
return ss.str();
|
||||
}
|
||||
|
||||
void CpuHasher::cleanup() {
|
||||
for(int i=0; i < m_computingThreads; i++) {
|
||||
delete m_threadData[i].argon2;
|
||||
free(m_threadData[i].mem);
|
||||
}
|
||||
delete[] m_threadData;
|
||||
if(m_dllHandle != NULL)
|
||||
dlclose(m_dllHandle);
|
||||
}
|
||||
|
||||
void CpuHasher::loadArgon2BlockFiller() {
|
||||
string module_path = m_appFolder;
|
||||
module_path += "/modules/argon2_fill_blocks_" + m_optimization + ".opt";
|
||||
|
||||
m_dllHandle = dlopen(module_path.c_str(), RTLD_LAZY);
|
||||
if(m_dllHandle != NULL)
|
||||
m_argon2BlocksFillerPtr = (argon2BlocksFillerPtr)dlsym(m_dllHandle, "fill_memory_blocks");
|
||||
}
|
||||
|
||||
int CpuHasher::compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output) {
|
||||
CpuHasherThread &threadData = m_threadData[threadIdx];
|
||||
threadData.hashData.input = input;
|
||||
threadData.hashData.inSize = size;
|
||||
threadData.hashData.output = output;
|
||||
return threadData.argon2->generateHashes(*m_profile, threadData.hashData);
|
||||
}
|
||||
|
||||
void *CpuHasher::allocateMemory(void *&buffer) {
|
||||
size_t mem_size = m_profile->memSize + 64;
|
||||
void *mem = malloc(mem_size);
|
||||
buffer = mem;
|
||||
return align(64, m_profile->memSize, mem, mem_size);
|
||||
}
|
||||
|
||||
size_t CpuHasher::parallelism(int workerIdx) {
|
||||
if(workerIdx < 0 || workerIdx > computingThreads())
|
||||
return 0;
|
||||
|
||||
return 1;
|
||||
}
|
||||
|
||||
size_t CpuHasher::deviceCount() {
|
||||
return computingThreads();
|
||||
}
|
||||
|
||||
REGISTER_HASHER(CpuHasher);
|
||||
41
src/crypto/argon2_hasher/hash/cpu/CpuHasher.h
Normal file
41
src/crypto/argon2_hasher/hash/cpu/CpuHasher.h
Normal file
|
|
@ -0,0 +1,41 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#ifndef ARGON2_CPU_HASHER_H
|
||||
#define ARGON2_CPU_HASHER_H
|
||||
|
||||
struct CpuHasherThread {
|
||||
Argon2 *argon2;
|
||||
HashData hashData;
|
||||
void *mem;
|
||||
};
|
||||
|
||||
class CpuHasher : public Hasher {
|
||||
public:
|
||||
CpuHasher();
|
||||
~CpuHasher();
|
||||
|
||||
virtual bool initialize(xmrig::Algo algorithm, xmrig::Variant variant);
|
||||
virtual bool configure(xmrig::HasherConfig &config);
|
||||
virtual void cleanup();
|
||||
virtual int compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output);
|
||||
virtual size_t parallelism(int workerIdx);
|
||||
virtual size_t deviceCount();
|
||||
|
||||
private:
|
||||
string detectFeaturesAndMakeDescription();
|
||||
void loadArgon2BlockFiller();
|
||||
void *allocateMemory(void *&buffer);
|
||||
|
||||
DeviceInfo m_deviceInfo;
|
||||
string m_optimization;
|
||||
int m_availableProcessingThr;
|
||||
int m_availableMemoryThr;
|
||||
void *m_dllHandle;
|
||||
Argon2Profile *m_profile;
|
||||
argon2BlocksFillerPtr m_argon2BlocksFillerPtr;
|
||||
CpuHasherThread *m_threadData;
|
||||
};
|
||||
|
||||
#endif //ARGON2_CPU_HASHER_H
|
||||
567
src/crypto/argon2_hasher/hash/cpu/argon2_opt/blamka-round-opt.h
Normal file
567
src/crypto/argon2_hasher/hash/cpu/argon2_opt/blamka-round-opt.h
Normal file
|
|
@ -0,0 +1,567 @@
|
|||
/*
|
||||
* Argon2 reference source code package - reference C implementations
|
||||
*
|
||||
* Copyright 2015
|
||||
* Daniel Dinu, Dmitry Khovratovich, Jean-Philippe Aumasson, and Samuel Neves
|
||||
*
|
||||
* You may use this work under the terms of a Creative Commons CC0 1.0
|
||||
* License/Waiver or the Apache Public License 2.0, at your option. The terms of
|
||||
* these licenses can be found at:
|
||||
*
|
||||
* - CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
* - Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* You should have received a copy of both of these licenses along with this
|
||||
* software. If not, they may be obtained at the above URLs.
|
||||
*/
|
||||
|
||||
#ifndef BLAKE_ROUND_MKA_OPT_H
|
||||
#define BLAKE_ROUND_MKA_OPT_H
|
||||
|
||||
#include "../../argon2/blake2/blake2-impl.h"
|
||||
|
||||
#if !defined(__NEON__)
|
||||
#include <emmintrin.h>
|
||||
#if defined(__SSSE3__)
|
||||
#include <tmmintrin.h> /* for _mm_shuffle_epi8 and _mm_alignr_epi8 */
|
||||
#endif
|
||||
|
||||
#if (defined(__XOP__) || defined(__AVX__)) && (defined(__GNUC__) || defined(__clang__))
|
||||
#include <x86intrin.h>
|
||||
#endif
|
||||
#else
|
||||
#include <arm_neon.h>
|
||||
#endif
|
||||
|
||||
#if !defined(__NEON__)
|
||||
#if !defined(__AVX512F__)
|
||||
#if !defined(__AVX2__)
|
||||
#if !defined(__XOP__)
|
||||
#if defined(__SSSE3__)
|
||||
#define r16 \
|
||||
(_mm_setr_epi8(2, 3, 4, 5, 6, 7, 0, 1, 10, 11, 12, 13, 14, 15, 8, 9))
|
||||
#define r24 \
|
||||
(_mm_setr_epi8(3, 4, 5, 6, 7, 0, 1, 2, 11, 12, 13, 14, 15, 8, 9, 10))
|
||||
#define _mm_roti_epi64(x, c) \
|
||||
(-(c) == 32) \
|
||||
? _mm_shuffle_epi32((x), _MM_SHUFFLE(2, 3, 0, 1)) \
|
||||
: (-(c) == 24) \
|
||||
? _mm_shuffle_epi8((x), r24) \
|
||||
: (-(c) == 16) \
|
||||
? _mm_shuffle_epi8((x), r16) \
|
||||
: (-(c) == 63) \
|
||||
? _mm_xor_si128(_mm_srli_epi64((x), -(c)), \
|
||||
_mm_add_epi64((x), (x))) \
|
||||
: _mm_xor_si128(_mm_srli_epi64((x), -(c)), \
|
||||
_mm_slli_epi64((x), 64 - (-(c))))
|
||||
#else /* defined(__SSE2__) */
|
||||
#define _mm_roti_epi64(r, c) \
|
||||
_mm_xor_si128(_mm_srli_epi64((r), -(c)), _mm_slli_epi64((r), 64 - (-(c))))
|
||||
#endif
|
||||
#else
|
||||
#endif
|
||||
|
||||
static BLAKE2_INLINE __m128i fBlaMka(__m128i x, __m128i y) {
|
||||
const __m128i z = _mm_mul_epu32(x, y);
|
||||
return _mm_add_epi64(_mm_add_epi64(x, y), _mm_add_epi64(z, z));
|
||||
}
|
||||
|
||||
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
A0 = fBlaMka(A0, B0); \
|
||||
A1 = fBlaMka(A1, B1); \
|
||||
\
|
||||
D0 = _mm_xor_si128(D0, A0); \
|
||||
D1 = _mm_xor_si128(D1, A1); \
|
||||
\
|
||||
D0 = _mm_roti_epi64(D0, -32); \
|
||||
D1 = _mm_roti_epi64(D1, -32); \
|
||||
\
|
||||
C0 = fBlaMka(C0, D0); \
|
||||
C1 = fBlaMka(C1, D1); \
|
||||
\
|
||||
B0 = _mm_xor_si128(B0, C0); \
|
||||
B1 = _mm_xor_si128(B1, C1); \
|
||||
\
|
||||
B0 = _mm_roti_epi64(B0, -24); \
|
||||
B1 = _mm_roti_epi64(B1, -24); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
A0 = fBlaMka(A0, B0); \
|
||||
A1 = fBlaMka(A1, B1); \
|
||||
\
|
||||
D0 = _mm_xor_si128(D0, A0); \
|
||||
D1 = _mm_xor_si128(D1, A1); \
|
||||
\
|
||||
D0 = _mm_roti_epi64(D0, -16); \
|
||||
D1 = _mm_roti_epi64(D1, -16); \
|
||||
\
|
||||
C0 = fBlaMka(C0, D0); \
|
||||
C1 = fBlaMka(C1, D1); \
|
||||
\
|
||||
B0 = _mm_xor_si128(B0, C0); \
|
||||
B1 = _mm_xor_si128(B1, C1); \
|
||||
\
|
||||
B0 = _mm_roti_epi64(B0, -63); \
|
||||
B1 = _mm_roti_epi64(B1, -63); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#if defined(__SSSE3__)
|
||||
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
__m128i t0 = _mm_alignr_epi8(B1, B0, 8); \
|
||||
__m128i t1 = _mm_alignr_epi8(B0, B1, 8); \
|
||||
B0 = t0; \
|
||||
B1 = t1; \
|
||||
\
|
||||
t0 = C0; \
|
||||
C0 = C1; \
|
||||
C1 = t0; \
|
||||
\
|
||||
t0 = _mm_alignr_epi8(D1, D0, 8); \
|
||||
t1 = _mm_alignr_epi8(D0, D1, 8); \
|
||||
D0 = t1; \
|
||||
D1 = t0; \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
__m128i t0 = _mm_alignr_epi8(B0, B1, 8); \
|
||||
__m128i t1 = _mm_alignr_epi8(B1, B0, 8); \
|
||||
B0 = t0; \
|
||||
B1 = t1; \
|
||||
\
|
||||
t0 = C0; \
|
||||
C0 = C1; \
|
||||
C1 = t0; \
|
||||
\
|
||||
t0 = _mm_alignr_epi8(D0, D1, 8); \
|
||||
t1 = _mm_alignr_epi8(D1, D0, 8); \
|
||||
D0 = t1; \
|
||||
D1 = t0; \
|
||||
} while ((void)0, 0)
|
||||
#else /* SSE2 */
|
||||
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
__m128i t0 = D0; \
|
||||
__m128i t1 = B0; \
|
||||
D0 = C0; \
|
||||
C0 = C1; \
|
||||
C1 = D0; \
|
||||
D0 = _mm_unpackhi_epi64(D1, _mm_unpacklo_epi64(t0, t0)); \
|
||||
D1 = _mm_unpackhi_epi64(t0, _mm_unpacklo_epi64(D1, D1)); \
|
||||
B0 = _mm_unpackhi_epi64(B0, _mm_unpacklo_epi64(B1, B1)); \
|
||||
B1 = _mm_unpackhi_epi64(B1, _mm_unpacklo_epi64(t1, t1)); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
__m128i t0, t1; \
|
||||
t0 = C0; \
|
||||
C0 = C1; \
|
||||
C1 = t0; \
|
||||
t0 = B0; \
|
||||
t1 = D0; \
|
||||
B0 = _mm_unpackhi_epi64(B1, _mm_unpacklo_epi64(B0, B0)); \
|
||||
B1 = _mm_unpackhi_epi64(t0, _mm_unpacklo_epi64(B1, B1)); \
|
||||
D0 = _mm_unpackhi_epi64(D0, _mm_unpacklo_epi64(D1, D1)); \
|
||||
D1 = _mm_unpackhi_epi64(D1, _mm_unpacklo_epi64(t1, t1)); \
|
||||
} while ((void)0, 0)
|
||||
#endif
|
||||
|
||||
#define BLAKE2_ROUND(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
} while ((void)0, 0)
|
||||
#else /* __AVX2__ */
|
||||
|
||||
#include <immintrin.h>
|
||||
|
||||
#define rotr32(x) _mm256_shuffle_epi32(x, _MM_SHUFFLE(2, 3, 0, 1))
|
||||
#define rotr24(x) _mm256_shuffle_epi8(x, _mm256_setr_epi8(3, 4, 5, 6, 7, 0, 1, 2, 11, 12, 13, 14, 15, 8, 9, 10, 3, 4, 5, 6, 7, 0, 1, 2, 11, 12, 13, 14, 15, 8, 9, 10))
|
||||
#define rotr16(x) _mm256_shuffle_epi8(x, _mm256_setr_epi8(2, 3, 4, 5, 6, 7, 0, 1, 10, 11, 12, 13, 14, 15, 8, 9, 2, 3, 4, 5, 6, 7, 0, 1, 10, 11, 12, 13, 14, 15, 8, 9))
|
||||
#define rotr63(x) _mm256_xor_si256(_mm256_srli_epi64((x), 63), _mm256_add_epi64((x), (x)))
|
||||
|
||||
#define G1_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
__m256i ml = _mm256_mul_epu32(A0, B0); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
A0 = _mm256_add_epi64(A0, _mm256_add_epi64(B0, ml)); \
|
||||
D0 = _mm256_xor_si256(D0, A0); \
|
||||
D0 = rotr32(D0); \
|
||||
\
|
||||
ml = _mm256_mul_epu32(C0, D0); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
C0 = _mm256_add_epi64(C0, _mm256_add_epi64(D0, ml)); \
|
||||
\
|
||||
B0 = _mm256_xor_si256(B0, C0); \
|
||||
B0 = rotr24(B0); \
|
||||
\
|
||||
ml = _mm256_mul_epu32(A1, B1); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
A1 = _mm256_add_epi64(A1, _mm256_add_epi64(B1, ml)); \
|
||||
D1 = _mm256_xor_si256(D1, A1); \
|
||||
D1 = rotr32(D1); \
|
||||
\
|
||||
ml = _mm256_mul_epu32(C1, D1); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
C1 = _mm256_add_epi64(C1, _mm256_add_epi64(D1, ml)); \
|
||||
\
|
||||
B1 = _mm256_xor_si256(B1, C1); \
|
||||
B1 = rotr24(B1); \
|
||||
} while((void)0, 0);
|
||||
|
||||
#define G2_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
__m256i ml = _mm256_mul_epu32(A0, B0); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
A0 = _mm256_add_epi64(A0, _mm256_add_epi64(B0, ml)); \
|
||||
D0 = _mm256_xor_si256(D0, A0); \
|
||||
D0 = rotr16(D0); \
|
||||
\
|
||||
ml = _mm256_mul_epu32(C0, D0); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
C0 = _mm256_add_epi64(C0, _mm256_add_epi64(D0, ml)); \
|
||||
B0 = _mm256_xor_si256(B0, C0); \
|
||||
B0 = rotr63(B0); \
|
||||
\
|
||||
ml = _mm256_mul_epu32(A1, B1); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
A1 = _mm256_add_epi64(A1, _mm256_add_epi64(B1, ml)); \
|
||||
D1 = _mm256_xor_si256(D1, A1); \
|
||||
D1 = rotr16(D1); \
|
||||
\
|
||||
ml = _mm256_mul_epu32(C1, D1); \
|
||||
ml = _mm256_add_epi64(ml, ml); \
|
||||
C1 = _mm256_add_epi64(C1, _mm256_add_epi64(D1, ml)); \
|
||||
B1 = _mm256_xor_si256(B1, C1); \
|
||||
B1 = rotr63(B1); \
|
||||
} while((void)0, 0);
|
||||
|
||||
#define DIAGONALIZE_1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
B0 = _mm256_permute4x64_epi64(B0, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
C0 = _mm256_permute4x64_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
D0 = _mm256_permute4x64_epi64(D0, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
\
|
||||
B1 = _mm256_permute4x64_epi64(B1, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
C1 = _mm256_permute4x64_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
D1 = _mm256_permute4x64_epi64(D1, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
} while((void)0, 0);
|
||||
|
||||
#define DIAGONALIZE_2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
__m256i tmp1 = _mm256_blend_epi32(B0, B1, 0xCC); \
|
||||
__m256i tmp2 = _mm256_blend_epi32(B0, B1, 0x33); \
|
||||
B1 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
|
||||
B0 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
|
||||
\
|
||||
tmp1 = C0; \
|
||||
C0 = C1; \
|
||||
C1 = tmp1; \
|
||||
\
|
||||
tmp1 = _mm256_blend_epi32(D0, D1, 0xCC); \
|
||||
tmp2 = _mm256_blend_epi32(D0, D1, 0x33); \
|
||||
D0 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
|
||||
D1 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
|
||||
} while(0);
|
||||
|
||||
#define UNDIAGONALIZE_1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
B0 = _mm256_permute4x64_epi64(B0, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
C0 = _mm256_permute4x64_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
D0 = _mm256_permute4x64_epi64(D0, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
\
|
||||
B1 = _mm256_permute4x64_epi64(B1, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
C1 = _mm256_permute4x64_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
D1 = _mm256_permute4x64_epi64(D1, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
} while((void)0, 0);
|
||||
|
||||
#define UNDIAGONALIZE_2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
__m256i tmp1 = _mm256_blend_epi32(B0, B1, 0xCC); \
|
||||
__m256i tmp2 = _mm256_blend_epi32(B0, B1, 0x33); \
|
||||
B0 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
|
||||
B1 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
|
||||
\
|
||||
tmp1 = C0; \
|
||||
C0 = C1; \
|
||||
C1 = tmp1; \
|
||||
\
|
||||
tmp1 = _mm256_blend_epi32(D0, D1, 0x33); \
|
||||
tmp2 = _mm256_blend_epi32(D0, D1, 0xCC); \
|
||||
D0 = _mm256_permute4x64_epi64(tmp1, _MM_SHUFFLE(2,3,0,1)); \
|
||||
D1 = _mm256_permute4x64_epi64(tmp2, _MM_SHUFFLE(2,3,0,1)); \
|
||||
} while((void)0, 0);
|
||||
|
||||
#define BLAKE2_ROUND_1(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do{ \
|
||||
G1_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
G2_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
\
|
||||
DIAGONALIZE_1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
\
|
||||
G1_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
G2_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
\
|
||||
UNDIAGONALIZE_1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
} while((void)0, 0);
|
||||
|
||||
#define BLAKE2_ROUND_2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do{ \
|
||||
G1_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
G2_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
\
|
||||
DIAGONALIZE_2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
\
|
||||
G1_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
G2_AVX2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
\
|
||||
UNDIAGONALIZE_2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
} while((void)0, 0);
|
||||
|
||||
#endif /* __AVX2__ */
|
||||
|
||||
#else /* __AVX512F__ */
|
||||
|
||||
#include <immintrin.h>
|
||||
|
||||
#define ror64(x, n) _mm512_ror_epi64((x), (n))
|
||||
|
||||
static BLAKE2_INLINE __m512i muladd(__m512i x, __m512i y)
|
||||
{
|
||||
__m512i z = _mm512_mul_epu32(x, y);
|
||||
return _mm512_add_epi64(_mm512_add_epi64(x, y), _mm512_add_epi64(z, z));
|
||||
}
|
||||
|
||||
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
A0 = muladd(A0, B0); \
|
||||
A1 = muladd(A1, B1); \
|
||||
\
|
||||
D0 = _mm512_xor_si512(D0, A0); \
|
||||
D1 = _mm512_xor_si512(D1, A1); \
|
||||
\
|
||||
D0 = ror64(D0, 32); \
|
||||
D1 = ror64(D1, 32); \
|
||||
\
|
||||
C0 = muladd(C0, D0); \
|
||||
C1 = muladd(C1, D1); \
|
||||
\
|
||||
B0 = _mm512_xor_si512(B0, C0); \
|
||||
B1 = _mm512_xor_si512(B1, C1); \
|
||||
\
|
||||
B0 = ror64(B0, 24); \
|
||||
B1 = ror64(B1, 24); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
A0 = muladd(A0, B0); \
|
||||
A1 = muladd(A1, B1); \
|
||||
\
|
||||
D0 = _mm512_xor_si512(D0, A0); \
|
||||
D1 = _mm512_xor_si512(D1, A1); \
|
||||
\
|
||||
D0 = ror64(D0, 16); \
|
||||
D1 = ror64(D1, 16); \
|
||||
\
|
||||
C0 = muladd(C0, D0); \
|
||||
C1 = muladd(C1, D1); \
|
||||
\
|
||||
B0 = _mm512_xor_si512(B0, C0); \
|
||||
B1 = _mm512_xor_si512(B1, C1); \
|
||||
\
|
||||
B0 = ror64(B0, 63); \
|
||||
B1 = ror64(B1, 63); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
B0 = _mm512_permutex_epi64(B0, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
B1 = _mm512_permutex_epi64(B1, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
\
|
||||
C0 = _mm512_permutex_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
C1 = _mm512_permutex_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
\
|
||||
D0 = _mm512_permutex_epi64(D0, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
D1 = _mm512_permutex_epi64(D1, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
B0 = _mm512_permutex_epi64(B0, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
B1 = _mm512_permutex_epi64(B1, _MM_SHUFFLE(2, 1, 0, 3)); \
|
||||
\
|
||||
C0 = _mm512_permutex_epi64(C0, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
C1 = _mm512_permutex_epi64(C1, _MM_SHUFFLE(1, 0, 3, 2)); \
|
||||
\
|
||||
D0 = _mm512_permutex_epi64(D0, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
D1 = _mm512_permutex_epi64(D1, _MM_SHUFFLE(0, 3, 2, 1)); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define BLAKE2_ROUND(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define SWAP_HALVES(A0, A1) \
|
||||
do { \
|
||||
__m512i t0, t1; \
|
||||
t0 = _mm512_shuffle_i64x2(A0, A1, _MM_SHUFFLE(1, 0, 1, 0)); \
|
||||
t1 = _mm512_shuffle_i64x2(A0, A1, _MM_SHUFFLE(3, 2, 3, 2)); \
|
||||
A0 = t0; \
|
||||
A1 = t1; \
|
||||
} while((void)0, 0)
|
||||
|
||||
#define SWAP_QUARTERS(A0, A1) \
|
||||
do { \
|
||||
SWAP_HALVES(A0, A1); \
|
||||
A0 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A0); \
|
||||
A1 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A1); \
|
||||
} while((void)0, 0)
|
||||
|
||||
#define UNSWAP_QUARTERS(A0, A1) \
|
||||
do { \
|
||||
A0 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A0); \
|
||||
A1 = _mm512_permutexvar_epi64(_mm512_setr_epi64(0, 1, 4, 5, 2, 3, 6, 7), A1); \
|
||||
SWAP_HALVES(A0, A1); \
|
||||
} while((void)0, 0)
|
||||
|
||||
#define BLAKE2_ROUND_1(A0, C0, B0, D0, A1, C1, B1, D1) \
|
||||
do { \
|
||||
SWAP_HALVES(A0, B0); \
|
||||
SWAP_HALVES(C0, D0); \
|
||||
SWAP_HALVES(A1, B1); \
|
||||
SWAP_HALVES(C1, D1); \
|
||||
BLAKE2_ROUND(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
SWAP_HALVES(A0, B0); \
|
||||
SWAP_HALVES(C0, D0); \
|
||||
SWAP_HALVES(A1, B1); \
|
||||
SWAP_HALVES(C1, D1); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define BLAKE2_ROUND_2(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
SWAP_QUARTERS(A0, A1); \
|
||||
SWAP_QUARTERS(B0, B1); \
|
||||
SWAP_QUARTERS(C0, C1); \
|
||||
SWAP_QUARTERS(D0, D1); \
|
||||
BLAKE2_ROUND(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
UNSWAP_QUARTERS(A0, A1); \
|
||||
UNSWAP_QUARTERS(B0, B1); \
|
||||
UNSWAP_QUARTERS(C0, C1); \
|
||||
UNSWAP_QUARTERS(D0, D1); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#endif /* __AVX512F__ */
|
||||
|
||||
#else /* __NEON__ */
|
||||
|
||||
static BLAKE2_INLINE uint64x2_t fBlaMka(uint64x2_t x, uint64x2_t y) {
|
||||
const uint64x2_t z = vmull_u32(vmovn_u64(x), vmovn_u64(y));
|
||||
return vaddq_u64(vaddq_u64(x, y), vaddq_u64(z, z));
|
||||
}
|
||||
|
||||
#define vrorq_n_u64_32(x) vreinterpretq_u64_u32(vrev64q_u32(vreinterpretq_u32_u64((x))))
|
||||
|
||||
#define vrorq_n_u64_24(x) vcombine_u64( \
|
||||
vreinterpret_u64_u8(vext_u8(vreinterpret_u8_u64(vget_low_u64(x)), vreinterpret_u8_u64(vget_low_u64(x)), 3)), \
|
||||
vreinterpret_u64_u8(vext_u8(vreinterpret_u8_u64(vget_high_u64(x)), vreinterpret_u8_u64(vget_high_u64(x)), 3)))
|
||||
|
||||
#define vrorq_n_u64_16(x) vcombine_u64( \
|
||||
vreinterpret_u64_u8(vext_u8(vreinterpret_u8_u64(vget_low_u64(x)), vreinterpret_u8_u64(vget_low_u64(x)), 2)), \
|
||||
vreinterpret_u64_u8(vext_u8(vreinterpret_u8_u64(vget_high_u64(x)), vreinterpret_u8_u64(vget_high_u64(x)), 2)))
|
||||
|
||||
#define vrorq_n_u64_63(x) veorq_u64(vaddq_u64(x, x), vshrq_n_u64(x, 63))
|
||||
|
||||
#define G1(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
A0 = fBlaMka(A0, B0); \
|
||||
A1 = fBlaMka(A1, B1); \
|
||||
\
|
||||
D0 = veorq_u64(D0, A0); \
|
||||
D1 = veorq_u64(D1, A1); \
|
||||
\
|
||||
D0 = vrorq_n_u64_32(D0); \
|
||||
D1 = vrorq_n_u64_32(D1); \
|
||||
\
|
||||
C0 = fBlaMka(C0, D0); \
|
||||
C1 = fBlaMka(C1, D1); \
|
||||
\
|
||||
B0 = veorq_u64(B0, C0); \
|
||||
B1 = veorq_u64(B1, C1); \
|
||||
\
|
||||
B0 = vrorq_n_u64_24(B0); \
|
||||
B1 = vrorq_n_u64_24(B1); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define G2(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
do { \
|
||||
A0 = fBlaMka(A0, B0); \
|
||||
A1 = fBlaMka(A1, B1); \
|
||||
\
|
||||
D0 = veorq_u64(D0, A0); \
|
||||
D1 = veorq_u64(D1, A1); \
|
||||
\
|
||||
D0 = vrorq_n_u64_16(D0); \
|
||||
D1 = vrorq_n_u64_16(D1); \
|
||||
\
|
||||
C0 = fBlaMka(C0, D0); \
|
||||
C1 = fBlaMka(C1, D1); \
|
||||
\
|
||||
B0 = veorq_u64(B0, C0); \
|
||||
B1 = veorq_u64(B1, C1); \
|
||||
\
|
||||
B0 = vrorq_n_u64_63(B0); \
|
||||
B1 = vrorq_n_u64_63(B1); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
t0 = vextq_u64(B0, B1, 1); \
|
||||
t1 = vextq_u64(B1, B0, 1); \
|
||||
B0 = t0; B1 = t1; t0 = C0; C0 = C1; C1 = t0; \
|
||||
t0 = vextq_u64(D1, D0, 1); t1 = vextq_u64(D0, D1, 1); \
|
||||
D0 = t0; D1 = t1;
|
||||
|
||||
#define UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1) \
|
||||
t0 = vextq_u64(B1, B0, 1); \
|
||||
t1 = vextq_u64(B0, B1, 1); \
|
||||
B0 = t0; B1 = t1; t0 = C0; C0 = C1; C1 = t0; \
|
||||
t0 = vextq_u64(D0, D1, 1); t1 = vextq_u64(D1, D0, 1); \
|
||||
D0 = t0; D1 = t1;
|
||||
|
||||
#define BLAKE2_ROUND(A0, A1, B0, B1, C0, C1, D0, D1) \
|
||||
do { \
|
||||
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
DIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
G1(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
G2(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
\
|
||||
UNDIAGONALIZE(A0, B0, C0, D0, A1, B1, C1, D1); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#endif /* __NEON__ */
|
||||
|
||||
#endif /* BLAKE_ROUND_MKA_OPT_H */
|
||||
|
|
@ -0,0 +1,55 @@
|
|||
/*
|
||||
* Argon2 reference source code package - reference C implementations
|
||||
*
|
||||
* Copyright 2015
|
||||
* Daniel Dinu, Dmitry Khovratovich, Jean-Philippe Aumasson, and Samuel Neves
|
||||
*
|
||||
* You may use this work under the terms of a Creative Commons CC0 1.0
|
||||
* License/Waiver or the Apache Public License 2.0, at your option. The terms of
|
||||
* these licenses can be found at:
|
||||
*
|
||||
* - CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
|
||||
* - Apache 2.0 : http://www.apache.org/licenses/LICENSE-2.0
|
||||
*
|
||||
* You should have received a copy of both of these licenses along with this
|
||||
* software. If not, they may be obtained at the above URLs.
|
||||
*/
|
||||
|
||||
#ifndef BLAKE_ROUND_MKA_H
|
||||
#define BLAKE_ROUND_MKA_H
|
||||
|
||||
#include "../../argon2/blake2/blake2-impl.h"
|
||||
|
||||
/* designed by the Lyra PHC team */
|
||||
static BLAKE2_INLINE uint64_t fBlaMka(uint64_t x, uint64_t y) {
|
||||
const uint64_t m = UINT64_C(0xFFFFFFFF);
|
||||
const uint64_t xy = (x & m) * (y & m);
|
||||
return x + y + 2 * xy;
|
||||
}
|
||||
|
||||
#define G(a, b, c, d) \
|
||||
do { \
|
||||
a = fBlaMka(a, b); \
|
||||
d = rotr64(d ^ a, 32); \
|
||||
c = fBlaMka(c, d); \
|
||||
b = rotr64(b ^ c, 24); \
|
||||
a = fBlaMka(a, b); \
|
||||
d = rotr64(d ^ a, 16); \
|
||||
c = fBlaMka(c, d); \
|
||||
b = rotr64(b ^ c, 63); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define BLAKE2_ROUND_NOMSG(v0, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, \
|
||||
v12, v13, v14, v15) \
|
||||
do { \
|
||||
G(v0, v4, v8, v12); \
|
||||
G(v1, v5, v9, v13); \
|
||||
G(v2, v6, v10, v14); \
|
||||
G(v3, v7, v11, v15); \
|
||||
G(v0, v5, v10, v15); \
|
||||
G(v1, v6, v11, v12); \
|
||||
G(v2, v7, v8, v13); \
|
||||
G(v3, v4, v9, v14); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#endif
|
||||
448
src/crypto/argon2_hasher/hash/cpu/argon2_opt/implementation.c
Executable file
448
src/crypto/argon2_hasher/hash/cpu/argon2_opt/implementation.c
Executable file
|
|
@ -0,0 +1,448 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 06/08/2018.
|
||||
//
|
||||
|
||||
#include <stdint.h>
|
||||
#include <stdio.h>
|
||||
#include <string.h>
|
||||
#include <stdlib.h>
|
||||
|
||||
#include "../../../common/DLLImport.h"
|
||||
#include "../../argon2/Defs.h"
|
||||
#include "../../../common/DLLExport.h"
|
||||
|
||||
#if !defined(BUILD_REF) && (defined(__x86_64__) || defined(_WIN64) || defined(__NEON__))
|
||||
#include "blamka-round-opt.h"
|
||||
#else
|
||||
#include "blamka-round-ref.h"
|
||||
#endif
|
||||
|
||||
void copy_block(block *dst, const block *src) {
|
||||
memcpy(dst->v, src->v, sizeof(uint64_t) * ARGON2_QWORDS_IN_BLOCK);
|
||||
}
|
||||
|
||||
void xor_block(block *dst, const block *src) {
|
||||
int i;
|
||||
for (i = 0; i < ARGON2_QWORDS_IN_BLOCK; ++i) {
|
||||
dst->v[i] ^= src->v[i];
|
||||
}
|
||||
}
|
||||
|
||||
#ifndef BUILD_REF
|
||||
|
||||
#if defined(__AVX512F__)
|
||||
static void fill_block(__m512i *state, const block *ref_block,
|
||||
block *next_block, int with_xor, int keep) {
|
||||
__m512i block_XY[ARGON2_512BIT_WORDS_IN_BLOCK];
|
||||
unsigned int i;
|
||||
|
||||
if (with_xor) {
|
||||
for (i = 0; i < ARGON2_512BIT_WORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm512_xor_si512(
|
||||
state[i], _mm512_loadu_si512((const __m512i *)ref_block->v + i));
|
||||
block_XY[i] = _mm512_xor_si512(
|
||||
state[i], _mm512_loadu_si512((const __m512i *)next_block->v + i));
|
||||
}
|
||||
} else {
|
||||
for (i = 0; i < ARGON2_512BIT_WORDS_IN_BLOCK; i++) {
|
||||
block_XY[i] = state[i] = _mm512_xor_si512(
|
||||
state[i], _mm512_loadu_si512((const __m512i *)ref_block->v + i));
|
||||
}
|
||||
}
|
||||
|
||||
for (i = 0; i < 2; ++i) {
|
||||
BLAKE2_ROUND_1(
|
||||
state[8 * i + 0], state[8 * i + 1], state[8 * i + 2], state[8 * i + 3],
|
||||
state[8 * i + 4], state[8 * i + 5], state[8 * i + 6], state[8 * i + 7]);
|
||||
}
|
||||
|
||||
for (i = 0; i < 2; ++i) {
|
||||
BLAKE2_ROUND_2(
|
||||
state[2 * 0 + i], state[2 * 1 + i], state[2 * 2 + i], state[2 * 3 + i],
|
||||
state[2 * 4 + i], state[2 * 5 + i], state[2 * 6 + i], state[2 * 7 + i]);
|
||||
}
|
||||
|
||||
if(keep) {
|
||||
for (i = 0; i < ARGON2_512BIT_WORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm512_xor_si512(state[i], block_XY[i]);
|
||||
_mm512_storeu_si512((__m512i *)next_block->v + i, state[i]);
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (i = 0; i < ARGON2_512BIT_WORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm512_xor_si512(state[i], block_XY[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
#elif defined(__AVX2__)
|
||||
static void fill_block(__m256i *state, const block *ref_block,
|
||||
block *next_block, int with_xor, int keep) {
|
||||
__m256i block_XY[ARGON2_HWORDS_IN_BLOCK];
|
||||
unsigned int i;
|
||||
|
||||
if (with_xor) {
|
||||
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm256_xor_si256(
|
||||
state[i], _mm256_loadu_si256((const __m256i *)ref_block->v + i));
|
||||
block_XY[i] = _mm256_xor_si256(
|
||||
state[i], _mm256_loadu_si256((const __m256i *)next_block->v + i));
|
||||
}
|
||||
} else {
|
||||
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
|
||||
block_XY[i] = state[i] = _mm256_xor_si256(
|
||||
state[i], _mm256_loadu_si256((const __m256i *)ref_block->v + i));
|
||||
}
|
||||
}
|
||||
|
||||
for (i = 0; i < 4; ++i) {
|
||||
BLAKE2_ROUND_1(state[8 * i + 0], state[8 * i + 4], state[8 * i + 1], state[8 * i + 5],
|
||||
state[8 * i + 2], state[8 * i + 6], state[8 * i + 3], state[8 * i + 7]);
|
||||
}
|
||||
|
||||
for (i = 0; i < 4; ++i) {
|
||||
BLAKE2_ROUND_2(state[ 0 + i], state[ 4 + i], state[ 8 + i], state[12 + i],
|
||||
state[16 + i], state[20 + i], state[24 + i], state[28 + i]);
|
||||
}
|
||||
|
||||
if(keep) {
|
||||
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm256_xor_si256(state[i], block_XY[i]);
|
||||
_mm256_store_si256((__m256i *)next_block->v + i, state[i]);
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (i = 0; i < ARGON2_HWORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm256_xor_si256(state[i], block_XY[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
#elif defined(__AVX__)
|
||||
|
||||
#define I2D(x) _mm256_castsi256_pd(x)
|
||||
#define D2I(x) _mm256_castpd_si256(x)
|
||||
|
||||
static void fill_block(__m128i *state, const block *ref_block,
|
||||
block *next_block, int with_xor, int keep) {
|
||||
__m128i block_XY[ARGON2_OWORDS_IN_BLOCK];
|
||||
unsigned int i;
|
||||
|
||||
__m256i t;
|
||||
__m256i *s256 = (__m256i *) state, *block256 = (__m256i *) block_XY;
|
||||
|
||||
if (with_xor) {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK / 2; i++) {
|
||||
t = D2I(_mm256_xor_pd(I2D(_mm256_loadu_si256(s256 + i)), \
|
||||
I2D(_mm256_loadu_si256((const __m256i *)ref_block->v + i))));
|
||||
_mm256_storeu_si256(s256 + i, t);
|
||||
t = D2I(_mm256_xor_pd(I2D(t), \
|
||||
I2D(_mm256_loadu_si256((const __m256i *)next_block->v + i))));
|
||||
_mm256_storeu_si256(block256 + i, t);
|
||||
}
|
||||
} else {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK / 2; i++) {
|
||||
t = D2I(_mm256_xor_pd(I2D(_mm256_loadu_si256(s256 + i)), \
|
||||
I2D(_mm256_loadu_si256((const __m256i *)ref_block->v + i))));
|
||||
_mm256_storeu_si256(s256 + i, t);
|
||||
_mm256_storeu_si256(block256 + i, t);
|
||||
}
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND(state[8 * i + 0], state[8 * i + 1], state[8 * i + 2],
|
||||
state[8 * i + 3], state[8 * i + 4], state[8 * i + 5],
|
||||
state[8 * i + 6], state[8 * i + 7]);
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND(state[8 * 0 + i], state[8 * 1 + i], state[8 * 2 + i],
|
||||
state[8 * 3 + i], state[8 * 4 + i], state[8 * 5 + i],
|
||||
state[8 * 6 + i], state[8 * 7 + i]);
|
||||
}
|
||||
|
||||
if(keep) {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK / 2; i++) {
|
||||
t = D2I(_mm256_xor_pd(I2D(_mm256_loadu_si256(s256 + i)), \
|
||||
I2D(_mm256_loadu_si256(block256 + i))));
|
||||
|
||||
_mm256_storeu_si256(s256 + i, t);
|
||||
_mm256_storeu_si256((__m256i *)next_block->v + i, t);
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK / 2; i++) {
|
||||
t = D2I(_mm256_xor_pd(I2D(_mm256_loadu_si256(s256 + i)), \
|
||||
I2D(_mm256_loadu_si256(block256 + i))));
|
||||
|
||||
_mm256_storeu_si256(s256 + i, t);
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
#elif defined(__NEON__)
|
||||
static void fill_block(uint64x2_t *state, const block *ref_block,
|
||||
block *next_block, int with_xor, int keep) {
|
||||
uint64x2_t block_XY[ARGON2_OWORDS_IN_BLOCK];
|
||||
uint64x2_t t0, t1;
|
||||
|
||||
unsigned int i;
|
||||
|
||||
if (with_xor) {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
state[i] = veorq_u64(state[i], vld1q_u64(ref_block->v + i*2));
|
||||
block_XY[i] = veorq_u64(state[i], vld1q_u64(next_block->v + i*2));
|
||||
}
|
||||
} else {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
block_XY[i] = state[i] = veorq_u64(state[i], vld1q_u64(ref_block->v + i*2));
|
||||
}
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND(state[8 * i + 0], state[8 * i + 1], state[8 * i + 2],
|
||||
state[8 * i + 3], state[8 * i + 4], state[8 * i + 5],
|
||||
state[8 * i + 6], state[8 * i + 7]);
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND(state[8 * 0 + i], state[8 * 1 + i], state[8 * 2 + i],
|
||||
state[8 * 3 + i], state[8 * 4 + i], state[8 * 5 + i],
|
||||
state[8 * 6 + i], state[8 * 7 + i]);
|
||||
}
|
||||
|
||||
if(keep) {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
state[i] = veorq_u64(state[i], block_XY[i]);
|
||||
vst1q_u64(next_block->v + i*2, state[i]);
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
state[i] = veorq_u64(state[i], block_XY[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
#else
|
||||
static void fill_block(__m128i *state, const block *ref_block,
|
||||
block *next_block, int with_xor, int keep) {
|
||||
__m128i block_XY[ARGON2_OWORDS_IN_BLOCK];
|
||||
unsigned int i;
|
||||
|
||||
if (with_xor) {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm_xor_si128(
|
||||
state[i], _mm_loadu_si128((const __m128i *)ref_block->v + i));
|
||||
block_XY[i] = _mm_xor_si128(
|
||||
state[i], _mm_loadu_si128((const __m128i *)next_block->v + i));
|
||||
}
|
||||
} else {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
block_XY[i] = state[i] = _mm_xor_si128(
|
||||
state[i], _mm_loadu_si128((const __m128i *)ref_block->v + i));
|
||||
}
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND(state[8 * i + 0], state[8 * i + 1], state[8 * i + 2],
|
||||
state[8 * i + 3], state[8 * i + 4], state[8 * i + 5],
|
||||
state[8 * i + 6], state[8 * i + 7]);
|
||||
}
|
||||
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND(state[8 * 0 + i], state[8 * 1 + i], state[8 * 2 + i],
|
||||
state[8 * 3 + i], state[8 * 4 + i], state[8 * 5 + i],
|
||||
state[8 * 6 + i], state[8 * 7 + i]);
|
||||
}
|
||||
|
||||
if(keep) {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm_xor_si128(state[i], block_XY[i]);
|
||||
_mm_storeu_si128((__m128i *)next_block->v + i, state[i]);
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (i = 0; i < ARGON2_OWORDS_IN_BLOCK; i++) {
|
||||
state[i] = _mm_xor_si128(state[i], block_XY[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
#endif
|
||||
|
||||
#else
|
||||
static void fill_block(block *prev_block, const block *ref_block,
|
||||
block *next_block, int with_xor, int keep) {
|
||||
block block_tmp;
|
||||
unsigned i;
|
||||
|
||||
xor_block(prev_block, ref_block);
|
||||
copy_block(&block_tmp, prev_block);
|
||||
|
||||
if (with_xor && next_block != NULL) {
|
||||
xor_block(&block_tmp, next_block);
|
||||
}
|
||||
|
||||
/* Apply Blake2 on columns of 64-bit words: (0,1,...,15) , then
|
||||
(16,17,..31)... finally (112,113,...127) */
|
||||
for (i = 0; i < 8; ++i) {
|
||||
BLAKE2_ROUND_NOMSG(
|
||||
prev_block->v[16 * i], prev_block->v[16 * i + 1], prev_block->v[16 * i + 2],
|
||||
prev_block->v[16 * i + 3], prev_block->v[16 * i + 4], prev_block->v[16 * i + 5],
|
||||
prev_block->v[16 * i + 6], prev_block->v[16 * i + 7], prev_block->v[16 * i + 8],
|
||||
prev_block->v[16 * i + 9], prev_block->v[16 * i + 10], prev_block->v[16 * i + 11],
|
||||
prev_block->v[16 * i + 12], prev_block->v[16 * i + 13], prev_block->v[16 * i + 14],
|
||||
prev_block->v[16 * i + 15]);
|
||||
}
|
||||
|
||||
/* Apply Blake2 on rows of 64-bit words: (0,1,16,17,...112,113), then
|
||||
(2,3,18,19,...,114,115).. finally (14,15,30,31,...,126,127) */
|
||||
for (i = 0; i < 8; i++) {
|
||||
BLAKE2_ROUND_NOMSG(
|
||||
prev_block->v[2 * i], prev_block->v[2 * i + 1], prev_block->v[2 * i + 16],
|
||||
prev_block->v[2 * i + 17], prev_block->v[2 * i + 32], prev_block->v[2 * i + 33],
|
||||
prev_block->v[2 * i + 48], prev_block->v[2 * i + 49], prev_block->v[2 * i + 64],
|
||||
prev_block->v[2 * i + 65], prev_block->v[2 * i + 80], prev_block->v[2 * i + 81],
|
||||
prev_block->v[2 * i + 96], prev_block->v[2 * i + 97], prev_block->v[2 * i + 112],
|
||||
prev_block->v[2 * i + 113]);
|
||||
}
|
||||
|
||||
xor_block(prev_block, &block_tmp);
|
||||
if(keep)
|
||||
copy_block(next_block, prev_block);
|
||||
}
|
||||
|
||||
#endif
|
||||
|
||||
DLLEXPORT void *fill_memory_blocks(int threads, Argon2Profile *profile, void *user_data) {
|
||||
void *memory = user_data;
|
||||
#ifndef BUILD_REF
|
||||
#if defined(__AVX512F__)
|
||||
__m512i state[ARGON2_512BIT_WORDS_IN_BLOCK];
|
||||
uint64_t buff_512[8];
|
||||
#elif defined(__AVX2__)
|
||||
__m256i state[ARGON2_HWORDS_IN_BLOCK];
|
||||
uint64_t buff_256[4];
|
||||
#elif defined(__x86_64__) || defined(_WIN64)
|
||||
__m128i state[ARGON2_OWORDS_IN_BLOCK];
|
||||
#elif defined(__NEON__)
|
||||
uint64x2_t state[ARGON2_OWORDS_IN_BLOCK];
|
||||
#endif
|
||||
#else
|
||||
block state_;
|
||||
block *state = &state_;
|
||||
#endif
|
||||
int lane_length = profile->segSize * 4;
|
||||
int seg_length = profile->segSize;
|
||||
int suc_idx = profile->succesiveIdxs;
|
||||
|
||||
for(int thr = 0; thr < threads;thr++) {
|
||||
block *ref_block = NULL, *curr_block = NULL;
|
||||
|
||||
int32_t ref_idx = 0;
|
||||
int32_t cur_idx = 0;
|
||||
int32_t prev_idx = 0;
|
||||
int32_t seg_type = 0;
|
||||
int32_t idx = 0;
|
||||
int32_t keep = 1;
|
||||
int32_t with_xor = 0;
|
||||
|
||||
block *blocks = (block *)((uint8_t*)memory + thr * profile->memSize);
|
||||
|
||||
int32_t *address = profile->blockRefs;
|
||||
|
||||
for(uint32_t s = 0; s < profile->segCount; s++) {
|
||||
cur_idx = profile->segments[s * 3];
|
||||
prev_idx = profile->segments[s * 3 + 1];
|
||||
seg_type = profile->segments[s * 3 + 2];
|
||||
keep = 1;
|
||||
with_xor = (s >= profile->thrCost * 4) ? 1 : 0;
|
||||
|
||||
idx = (s < profile->thrCost) ? 2 : 0;
|
||||
|
||||
int32_t lane = s % profile->thrCost;
|
||||
int32_t slice = (s / profile->thrCost) % 4;
|
||||
int32_t pass = (s / profile->thrCost) / 4;
|
||||
|
||||
memcpy(state, (void *) (blocks + prev_idx), ARGON2_BLOCK_SIZE);
|
||||
|
||||
if(seg_type == 0) {
|
||||
if(s < profile->thrCost)
|
||||
address = &profile->blockRefs[(s * (profile->segSize - 2)) * 3];
|
||||
else
|
||||
address = &profile->blockRefs[(profile->thrCost * (profile->segSize - 2) + (s - profile->thrCost) * profile->segSize) * 3];
|
||||
}
|
||||
|
||||
for (int i = idx; i < seg_length; ++i, cur_idx ++) {
|
||||
if (seg_type == 1) { // data dependent addressing
|
||||
#ifndef BUILD_REF
|
||||
#if defined(__AVX512F__)
|
||||
_mm512_storeu_si512(buff_512, state[0]);
|
||||
uint64_t pseudo_rand = buff_512[0];
|
||||
#elif defined(__AVX2__)
|
||||
_mm256_storeu_si256(buff_256, state[0]);
|
||||
uint64_t pseudo_rand = buff_256[0];
|
||||
#elif defined(__x86_64__) || defined(_WIN64)
|
||||
uint64_t pseudo_rand = _mm_cvtsi128_si64(state[0]);
|
||||
#elif defined(__NEON__)
|
||||
uint64_t pseudo_rand = 0;
|
||||
vst1q_lane_u64(&pseudo_rand, state[0], 0);
|
||||
#endif
|
||||
#else
|
||||
uint64_t pseudo_rand = state->v[0];
|
||||
#endif
|
||||
uint64_t ref_lane = ((pseudo_rand >> 32)) % profile->thrCost;
|
||||
uint32_t reference_area_size = 0;
|
||||
if(pass > 0) {
|
||||
if (lane == ref_lane) {
|
||||
reference_area_size = lane_length - seg_length + i - 1;
|
||||
} else {
|
||||
reference_area_size = lane_length - seg_length + ((i == 0) ? (-1) : 0);
|
||||
}
|
||||
}
|
||||
else {
|
||||
if (lane == ref_lane) {
|
||||
reference_area_size = slice * seg_length + i - 1;
|
||||
} else {
|
||||
reference_area_size = slice * seg_length + ((i == 0) ? (-1) : 0);
|
||||
}
|
||||
}
|
||||
uint64_t relative_position = pseudo_rand & 0xFFFFFFFF;
|
||||
relative_position = relative_position * relative_position >> 32;
|
||||
|
||||
relative_position = reference_area_size - 1 -
|
||||
(reference_area_size * relative_position >> 32);
|
||||
|
||||
ref_idx = ref_lane * lane_length + (((pass > 0 && slice < 3) ? ((slice + 1) * seg_length) : 0) + relative_position) % lane_length;
|
||||
}
|
||||
else {
|
||||
ref_idx = address[1];
|
||||
if(suc_idx == 0)
|
||||
cur_idx = address[0];
|
||||
keep = address[2];
|
||||
|
||||
address += 3;
|
||||
}
|
||||
|
||||
ref_block = blocks + ref_idx;
|
||||
curr_block = blocks + cur_idx;
|
||||
|
||||
fill_block(state, ref_block, curr_block, with_xor, keep);
|
||||
}
|
||||
}
|
||||
|
||||
uint32_t dst = -1;
|
||||
for(; address < (profile->blockRefs + profile->blockRefsSize * 3); address += 3) {
|
||||
if (address[2] == -1) {
|
||||
curr_block = blocks + address[0];
|
||||
ref_block = blocks + address[1];
|
||||
dst = address[0];
|
||||
xor_block(curr_block, ref_block);
|
||||
}
|
||||
}
|
||||
if(dst != -1)
|
||||
copy_block(blocks, blocks + dst);
|
||||
else
|
||||
copy_block(blocks, state);
|
||||
}
|
||||
|
||||
return memory;
|
||||
}
|
||||
|
||||
4
src/crypto/argon2_hasher/hash/cpu/cpu_features/.clang-format
Executable file
4
src/crypto/argon2_hasher/hash/cpu/cpu_features/.clang-format
Executable file
|
|
@ -0,0 +1,4 @@
|
|||
---
|
||||
Language: Cpp
|
||||
BasedOnStyle: Google
|
||||
...
|
||||
1
src/crypto/argon2_hasher/hash/cpu/cpu_features/.gitignore
vendored
Executable file
1
src/crypto/argon2_hasher/hash/cpu/cpu_features/.gitignore
vendored
Executable file
|
|
@ -0,0 +1 @@
|
|||
cmake_build/
|
||||
91
src/crypto/argon2_hasher/hash/cpu/cpu_features/.travis.yml
Executable file
91
src/crypto/argon2_hasher/hash/cpu/cpu_features/.travis.yml
Executable file
|
|
@ -0,0 +1,91 @@
|
|||
language: c
|
||||
|
||||
sudo: false
|
||||
|
||||
cache:
|
||||
directories:
|
||||
- $HOME/cpu_features_archives
|
||||
|
||||
matrix:
|
||||
include:
|
||||
- os: linux
|
||||
compiler: gcc
|
||||
env:
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
- os: linux
|
||||
compiler: clang
|
||||
env:
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
- os: osx
|
||||
compiler: gcc
|
||||
env:
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
- os: osx
|
||||
compiler: clang
|
||||
env:
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
- os: linux-ppc64le
|
||||
compiler: gcc
|
||||
env:
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
- os: linux-ppc64le
|
||||
compiler: clang
|
||||
env:
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
# Toolchains for little-endian, 64-bit ARMv8 for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=aarch64-linux-gnu
|
||||
QEMU_ARCH=aarch64
|
||||
# Toolchains for little-endian, hard-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=arm-linux-gnueabihf
|
||||
QEMU_ARCH=arm
|
||||
# Toolchains for little-endian, 32-bit ARMv8 for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=armv8l-linux-gnueabihf
|
||||
QEMU_ARCH=arm
|
||||
# Toolchains for little-endian, soft-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=arm-linux-gnueabi
|
||||
QEMU_ARCH=arm
|
||||
# Toolchains for big-endian, 64-bit ARMv8 for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=aarch64_be-linux-gnu
|
||||
QEMU_ARCH=DISABLED
|
||||
# Toolchains for big-endian, hard-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=armeb-linux-gnueabihf
|
||||
QEMU_ARCH=DISABLED
|
||||
# Toolchains for big-endian, soft-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=armeb-linux-gnueabi
|
||||
QEMU_ARCH=DISABLED
|
||||
- os: linux
|
||||
env:
|
||||
TOOLCHAIN=CODESCAPE
|
||||
TARGET=mips-mti-linux-gnu
|
||||
QEMU_ARCH=DISABLED
|
||||
|
||||
script:
|
||||
- cmake --version
|
||||
- bash -e -x ./scripts/run_integration.sh
|
||||
165
src/crypto/argon2_hasher/hash/cpu/cpu_features/CMakeLists.txt
Executable file
165
src/crypto/argon2_hasher/hash/cpu/cpu_features/CMakeLists.txt
Executable file
|
|
@ -0,0 +1,165 @@
|
|||
cmake_minimum_required(VERSION 3.0)
|
||||
|
||||
project(CpuFeatures VERSION 0.1.0)
|
||||
|
||||
# Default Build Type to be Release
|
||||
if(NOT CMAKE_BUILD_TYPE)
|
||||
set(CMAKE_BUILD_TYPE "Release" CACHE STRING
|
||||
"Choose the type of build, options are: Debug Release RelWithDebInfo MinSizeRel."
|
||||
FORCE)
|
||||
endif(NOT CMAKE_BUILD_TYPE)
|
||||
|
||||
# BUILD_TESTING is a standard CMake variable, but we declare it here to make it
|
||||
# prominent in the GUI.
|
||||
option(BUILD_TESTING "Enable test (depends on googletest)." OFF)
|
||||
# BUILD_SHARED_LIBS is a standard CMake variable, but we declare it here to make
|
||||
# it prominent in the GUI.
|
||||
option(BUILD_SHARED_LIBS "Build library as shared." OFF)
|
||||
|
||||
#
|
||||
# library : cpu_features
|
||||
#
|
||||
|
||||
set(_HDRS
|
||||
include/cpuinfo_aarch64.h
|
||||
include/cpuinfo_arm.h
|
||||
include/cpuinfo_mips.h
|
||||
include/cpuinfo_ppc.h
|
||||
include/cpuinfo_x86.h
|
||||
include/cpu_features_macros.h
|
||||
)
|
||||
|
||||
add_library(cpu_features
|
||||
${_HDRS}
|
||||
include/internal/bit_utils.h
|
||||
include/internal/linux_features_aggregator.h
|
||||
include/internal/cpuid_x86.h
|
||||
include/internal/filesystem.h
|
||||
include/internal/hwcaps.h
|
||||
include/internal/stack_line_reader.h
|
||||
include/internal/string_view.h
|
||||
include/cpu_features_macros.h
|
||||
src/linux_features_aggregator.c
|
||||
src/cpuid_x86_clang_gcc.c
|
||||
src/cpuid_x86_msvc.c
|
||||
src/cpuinfo_aarch64.c
|
||||
src/cpuinfo_arm.c
|
||||
src/cpuinfo_mips.c
|
||||
src/cpuinfo_ppc.c
|
||||
src/cpuinfo_x86.c
|
||||
src/filesystem.c
|
||||
src/hwcaps.c
|
||||
src/stack_line_reader.c
|
||||
src/string_view.c
|
||||
)
|
||||
|
||||
target_include_directories(cpu_features
|
||||
PUBLIC
|
||||
$<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/include>
|
||||
$<INSTALL_INTERFACE:include/cpu_features>
|
||||
PRIVATE
|
||||
include/internal
|
||||
)
|
||||
set_target_properties(cpu_features PROPERTIES PUBLIC_HEADER "${_HDRS}")
|
||||
target_compile_definitions(cpu_features
|
||||
PUBLIC STACK_LINE_READER_BUFFER_SIZE=1024)
|
||||
target_link_libraries(cpu_features PUBLIC ${CMAKE_DL_LIBS})
|
||||
|
||||
# The use of shared libraries is discouraged.
|
||||
# For API / ABI compatibility reasons, it is recommended to build and use
|
||||
# cpu_features in a subdirectory of your project or as an embedded dependency.
|
||||
if(BUILD_SHARED_LIBS)
|
||||
set_property(TARGET cpu_features PROPERTY POSITION_INDEPENDENT_CODE ON)
|
||||
endif()
|
||||
add_library(CpuFeature::cpu_features ALIAS cpu_features)
|
||||
|
||||
#
|
||||
# program : list_cpu_features
|
||||
#
|
||||
|
||||
add_executable(list_cpu_features src/utils/list_cpu_features.c)
|
||||
target_link_libraries(list_cpu_features PRIVATE cpu_features)
|
||||
add_executable(CpuFeature::list_cpu_features ALIAS list_cpu_features)
|
||||
|
||||
#
|
||||
# tests
|
||||
#
|
||||
|
||||
include(CTest)
|
||||
if(BUILD_TESTING)
|
||||
# Automatically incorporate googletest into the CMake Project if target not
|
||||
# found.
|
||||
if(NOT TARGET gtest OR NOT TARGET gmock_main)
|
||||
# Download and unpack googletest at configure time.
|
||||
configure_file(
|
||||
cmake/googletest.CMakeLists.txt.in
|
||||
googletest-download/CMakeLists.txt
|
||||
)
|
||||
|
||||
execute_process(
|
||||
COMMAND ${CMAKE_COMMAND} -G "${CMAKE_GENERATOR}" .
|
||||
RESULT_VARIABLE result
|
||||
WORKING_DIRECTORY ${CMAKE_BINARY_DIR}/googletest-download)
|
||||
|
||||
if(result)
|
||||
message(FATAL_ERROR "CMake step for googletest failed: ${result}")
|
||||
endif()
|
||||
|
||||
execute_process(
|
||||
COMMAND ${CMAKE_COMMAND} --build .
|
||||
RESULT_VARIABLE result
|
||||
WORKING_DIRECTORY ${CMAKE_BINARY_DIR}/googletest-download)
|
||||
|
||||
if(result)
|
||||
message(FATAL_ERROR "Build step for googletest failed: ${result}")
|
||||
endif()
|
||||
|
||||
# Prevent overriding the parent project's compiler/linker settings on
|
||||
# Windows.
|
||||
set(gtest_force_shared_crt ON CACHE BOOL "" FORCE)
|
||||
|
||||
# Add googletest directly to our build. This defines the gtest and
|
||||
# gtest_main targets.
|
||||
add_subdirectory(${CMAKE_BINARY_DIR}/googletest-src
|
||||
${CMAKE_BINARY_DIR}/googletest-build
|
||||
EXCLUDE_FROM_ALL)
|
||||
endif()
|
||||
|
||||
add_subdirectory(test)
|
||||
endif()
|
||||
|
||||
#
|
||||
# Install
|
||||
#
|
||||
|
||||
include(GNUInstallDirs)
|
||||
install(TARGETS cpu_features list_cpu_features
|
||||
EXPORT CpuFeaturesTargets
|
||||
PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/cpu_features
|
||||
ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR}
|
||||
LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
|
||||
RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR}
|
||||
)
|
||||
install(EXPORT CpuFeaturesTargets
|
||||
NAMESPACE CpuFeatures::
|
||||
DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/CpuFeatures
|
||||
COMPONENT Devel
|
||||
)
|
||||
include(CMakePackageConfigHelpers)
|
||||
configure_package_config_file(cmake/CpuFeaturesConfig.cmake.in
|
||||
"${PROJECT_BINARY_DIR}/CpuFeaturesConfig.cmake"
|
||||
INSTALL_DESTINATION "${CMAKE_INSTALL_LIBDIR}/cmake/CpuFeatures"
|
||||
NO_SET_AND_CHECK_MACRO
|
||||
NO_CHECK_REQUIRED_COMPONENTS_MACRO
|
||||
)
|
||||
write_basic_package_version_file(
|
||||
"${PROJECT_BINARY_DIR}/CpuFeaturesConfigVersion.cmake"
|
||||
COMPATIBILITY SameMajorVersion
|
||||
)
|
||||
install(
|
||||
FILES
|
||||
"${PROJECT_BINARY_DIR}/CpuFeaturesConfig.cmake"
|
||||
"${PROJECT_BINARY_DIR}/CpuFeaturesConfigVersion.cmake"
|
||||
DESTINATION "${CMAKE_INSTALL_LIBDIR}/cmake/CpuFeatures"
|
||||
COMPONENT Devel
|
||||
)
|
||||
23
src/crypto/argon2_hasher/hash/cpu/cpu_features/CONTRIBUTING.md
Executable file
23
src/crypto/argon2_hasher/hash/cpu/cpu_features/CONTRIBUTING.md
Executable file
|
|
@ -0,0 +1,23 @@
|
|||
# How to Contribute
|
||||
|
||||
We'd love to accept your patches and contributions to this project. There are
|
||||
just a few small guidelines you need to follow.
|
||||
|
||||
## Contributor License Agreement
|
||||
|
||||
Contributions to this project must be accompanied by a Contributor License
|
||||
Agreement. You (or your employer) retain the copyright to your contribution;
|
||||
this simply gives us permission to use and redistribute your contributions as
|
||||
part of the project. Head over to <https://cla.developers.google.com/> to see
|
||||
your current agreements on file or to sign a new one.
|
||||
|
||||
You generally only need to submit a CLA once, so if you've already submitted one
|
||||
(even if it was for a different project), you probably don't need to do it
|
||||
again.
|
||||
|
||||
## Code reviews
|
||||
|
||||
All submissions, including submissions by project members, require review. We
|
||||
use GitHub pull requests for this purpose. Consult
|
||||
[GitHub Help](https://help.github.com/articles/about-pull-requests/) for more
|
||||
information on using pull requests.
|
||||
202
src/crypto/argon2_hasher/hash/cpu/cpu_features/LICENSE
Executable file
202
src/crypto/argon2_hasher/hash/cpu/cpu_features/LICENSE
Executable file
|
|
@ -0,0 +1,202 @@
|
|||
|
||||
Apache License
|
||||
Version 2.0, January 2004
|
||||
http://www.apache.org/licenses/
|
||||
|
||||
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
||||
|
||||
1. Definitions.
|
||||
|
||||
"License" shall mean the terms and conditions for use, reproduction,
|
||||
and distribution as defined by Sections 1 through 9 of this document.
|
||||
|
||||
"Licensor" shall mean the copyright owner or entity authorized by
|
||||
the copyright owner that is granting the License.
|
||||
|
||||
"Legal Entity" shall mean the union of the acting entity and all
|
||||
other entities that control, are controlled by, or are under common
|
||||
control with that entity. For the purposes of this definition,
|
||||
"control" means (i) the power, direct or indirect, to cause the
|
||||
direction or management of such entity, whether by contract or
|
||||
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
||||
outstanding shares, or (iii) beneficial ownership of such entity.
|
||||
|
||||
"You" (or "Your") shall mean an individual or Legal Entity
|
||||
exercising permissions granted by this License.
|
||||
|
||||
"Source" form shall mean the preferred form for making modifications,
|
||||
including but not limited to software source code, documentation
|
||||
source, and configuration files.
|
||||
|
||||
"Object" form shall mean any form resulting from mechanical
|
||||
transformation or translation of a Source form, including but
|
||||
not limited to compiled object code, generated documentation,
|
||||
and conversions to other media types.
|
||||
|
||||
"Work" shall mean the work of authorship, whether in Source or
|
||||
Object form, made available under the License, as indicated by a
|
||||
copyright notice that is included in or attached to the work
|
||||
(an example is provided in the Appendix below).
|
||||
|
||||
"Derivative Works" shall mean any work, whether in Source or Object
|
||||
form, that is based on (or derived from) the Work and for which the
|
||||
editorial revisions, annotations, elaborations, or other modifications
|
||||
represent, as a whole, an original work of authorship. For the purposes
|
||||
of this License, Derivative Works shall not include works that remain
|
||||
separable from, or merely link (or bind by name) to the interfaces of,
|
||||
the Work and Derivative Works thereof.
|
||||
|
||||
"Contribution" shall mean any work of authorship, including
|
||||
the original version of the Work and any modifications or additions
|
||||
to that Work or Derivative Works thereof, that is intentionally
|
||||
submitted to Licensor for inclusion in the Work by the copyright owner
|
||||
or by an individual or Legal Entity authorized to submit on behalf of
|
||||
the copyright owner. For the purposes of this definition, "submitted"
|
||||
means any form of electronic, verbal, or written communication sent
|
||||
to the Licensor or its representatives, including but not limited to
|
||||
communication on electronic mailing lists, source code control systems,
|
||||
and issue tracking systems that are managed by, or on behalf of, the
|
||||
Licensor for the purpose of discussing and improving the Work, but
|
||||
excluding communication that is conspicuously marked or otherwise
|
||||
designated in writing by the copyright owner as "Not a Contribution."
|
||||
|
||||
"Contributor" shall mean Licensor and any individual or Legal Entity
|
||||
on behalf of whom a Contribution has been received by Licensor and
|
||||
subsequently incorporated within the Work.
|
||||
|
||||
2. Grant of Copyright License. Subject to the terms and conditions of
|
||||
this License, each Contributor hereby grants to You a perpetual,
|
||||
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
||||
copyright license to reproduce, prepare Derivative Works of,
|
||||
publicly display, publicly perform, sublicense, and distribute the
|
||||
Work and such Derivative Works in Source or Object form.
|
||||
|
||||
3. Grant of Patent License. Subject to the terms and conditions of
|
||||
this License, each Contributor hereby grants to You a perpetual,
|
||||
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
||||
(except as stated in this section) patent license to make, have made,
|
||||
use, offer to sell, sell, import, and otherwise transfer the Work,
|
||||
where such license applies only to those patent claims licensable
|
||||
by such Contributor that are necessarily infringed by their
|
||||
Contribution(s) alone or by combination of their Contribution(s)
|
||||
with the Work to which such Contribution(s) was submitted. If You
|
||||
institute patent litigation against any entity (including a
|
||||
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
||||
or a Contribution incorporated within the Work constitutes direct
|
||||
or contributory patent infringement, then any patent licenses
|
||||
granted to You under this License for that Work shall terminate
|
||||
as of the date such litigation is filed.
|
||||
|
||||
4. Redistribution. You may reproduce and distribute copies of the
|
||||
Work or Derivative Works thereof in any medium, with or without
|
||||
modifications, and in Source or Object form, provided that You
|
||||
meet the following conditions:
|
||||
|
||||
(a) You must give any other recipients of the Work or
|
||||
Derivative Works a copy of this License; and
|
||||
|
||||
(b) You must cause any modified files to carry prominent notices
|
||||
stating that You changed the files; and
|
||||
|
||||
(c) You must retain, in the Source form of any Derivative Works
|
||||
that You distribute, all copyright, patent, trademark, and
|
||||
attribution notices from the Source form of the Work,
|
||||
excluding those notices that do not pertain to any part of
|
||||
the Derivative Works; and
|
||||
|
||||
(d) If the Work includes a "NOTICE" text file as part of its
|
||||
distribution, then any Derivative Works that You distribute must
|
||||
include a readable copy of the attribution notices contained
|
||||
within such NOTICE file, excluding those notices that do not
|
||||
pertain to any part of the Derivative Works, in at least one
|
||||
of the following places: within a NOTICE text file distributed
|
||||
as part of the Derivative Works; within the Source form or
|
||||
documentation, if provided along with the Derivative Works; or,
|
||||
within a display generated by the Derivative Works, if and
|
||||
wherever such third-party notices normally appear. The contents
|
||||
of the NOTICE file are for informational purposes only and
|
||||
do not modify the License. You may add Your own attribution
|
||||
notices within Derivative Works that You distribute, alongside
|
||||
or as an addendum to the NOTICE text from the Work, provided
|
||||
that such additional attribution notices cannot be construed
|
||||
as modifying the License.
|
||||
|
||||
You may add Your own copyright statement to Your modifications and
|
||||
may provide additional or different license terms and conditions
|
||||
for use, reproduction, or distribution of Your modifications, or
|
||||
for any such Derivative Works as a whole, provided Your use,
|
||||
reproduction, and distribution of the Work otherwise complies with
|
||||
the conditions stated in this License.
|
||||
|
||||
5. Submission of Contributions. Unless You explicitly state otherwise,
|
||||
any Contribution intentionally submitted for inclusion in the Work
|
||||
by You to the Licensor shall be under the terms and conditions of
|
||||
this License, without any additional terms or conditions.
|
||||
Notwithstanding the above, nothing herein shall supersede or modify
|
||||
the terms of any separate license agreement you may have executed
|
||||
with Licensor regarding such Contributions.
|
||||
|
||||
6. Trademarks. This License does not grant permission to use the trade
|
||||
names, trademarks, service marks, or product names of the Licensor,
|
||||
except as required for reasonable and customary use in describing the
|
||||
origin of the Work and reproducing the content of the NOTICE file.
|
||||
|
||||
7. Disclaimer of Warranty. Unless required by applicable law or
|
||||
agreed to in writing, Licensor provides the Work (and each
|
||||
Contributor provides its Contributions) on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
||||
implied, including, without limitation, any warranties or conditions
|
||||
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
||||
PARTICULAR PURPOSE. You are solely responsible for determining the
|
||||
appropriateness of using or redistributing the Work and assume any
|
||||
risks associated with Your exercise of permissions under this License.
|
||||
|
||||
8. Limitation of Liability. In no event and under no legal theory,
|
||||
whether in tort (including negligence), contract, or otherwise,
|
||||
unless required by applicable law (such as deliberate and grossly
|
||||
negligent acts) or agreed to in writing, shall any Contributor be
|
||||
liable to You for damages, including any direct, indirect, special,
|
||||
incidental, or consequential damages of any character arising as a
|
||||
result of this License or out of the use or inability to use the
|
||||
Work (including but not limited to damages for loss of goodwill,
|
||||
work stoppage, computer failure or malfunction, or any and all
|
||||
other commercial damages or losses), even if such Contributor
|
||||
has been advised of the possibility of such damages.
|
||||
|
||||
9. Accepting Warranty or Additional Liability. While redistributing
|
||||
the Work or Derivative Works thereof, You may choose to offer,
|
||||
and charge a fee for, acceptance of support, warranty, indemnity,
|
||||
or other liability obligations and/or rights consistent with this
|
||||
License. However, in accepting such obligations, You may act only
|
||||
on Your own behalf and on Your sole responsibility, not on behalf
|
||||
of any other Contributor, and only if You agree to indemnify,
|
||||
defend, and hold each Contributor harmless for any liability
|
||||
incurred by, or claims asserted against, such Contributor by reason
|
||||
of your accepting any such warranty or additional liability.
|
||||
|
||||
END OF TERMS AND CONDITIONS
|
||||
|
||||
APPENDIX: How to apply the Apache License to your work.
|
||||
|
||||
To apply the Apache License to your work, attach the following
|
||||
boilerplate notice, with the fields enclosed by brackets "[]"
|
||||
replaced with your own identifying information. (Don't include
|
||||
the brackets!) The text should be enclosed in the appropriate
|
||||
comment syntax for the file format. We also recommend that a
|
||||
file or class name and description of purpose be included on the
|
||||
same "printed page" as the copyright notice for easier
|
||||
identification within third-party archives.
|
||||
|
||||
Copyright [yyyy] [name of copyright owner]
|
||||
|
||||
Licensed under the Apache License, Version 2.0 (the "License");
|
||||
you may not use this file except in compliance with the License.
|
||||
You may obtain a copy of the License at
|
||||
|
||||
http://www.apache.org/licenses/LICENSE-2.0
|
||||
|
||||
Unless required by applicable law or agreed to in writing, software
|
||||
distributed under the License is distributed on an "AS IS" BASIS,
|
||||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
165
src/crypto/argon2_hasher/hash/cpu/cpu_features/README.md
Executable file
165
src/crypto/argon2_hasher/hash/cpu/cpu_features/README.md
Executable file
|
|
@ -0,0 +1,165 @@
|
|||
# cpu_features [](https://travis-ci.org/google/cpu_features) [](https://ci.appveyor.com/project/gchatelet/cpu-features/branch/master)
|
||||
|
||||
A cross-platform C library to retrieve CPU features (such as available
|
||||
instructions) at runtime.
|
||||
|
||||
## Table of Contents
|
||||
|
||||
- [Design Rationale](#rationale)
|
||||
- [Code samples](#codesample)
|
||||
- [Running sample code](#usagesample)
|
||||
- [What's supported](#support)
|
||||
- [License](#license)
|
||||
- [Build with cmake](#cmake)
|
||||
|
||||
<a name="rationale"></a>
|
||||
## Design Rationale
|
||||
|
||||
- **Simple to use.** See the snippets below for examples.
|
||||
- **Extensible.** Easy to add missing features or architectures.
|
||||
- **Compatible with old compilers** and available on many architectures so it
|
||||
can be used widely. To ensure that cpu_features works on as many platforms
|
||||
as possible, we implemented it in a highly portable version of C: C99.
|
||||
- **Sandbox-compatible.** The library uses a variety of strategies to cope
|
||||
with sandboxed environments or when `cpuid` is unavailable. This is useful
|
||||
when running integration tests in hermetic environments.
|
||||
- **Thread safe, no memory allocation, and raises no exceptions.**
|
||||
cpu_features is suitable for implementing fundamental libc functions like
|
||||
`malloc`, `memcpy`, and `memcmp`.
|
||||
- **Unit tested.**
|
||||
|
||||
<a name="codesample"></a>
|
||||
### Checking features at runtime
|
||||
|
||||
Here's a simple example that executes a codepath if the CPU supports both the
|
||||
AES and the SSE4.2 instruction sets:
|
||||
|
||||
```c
|
||||
#include "cpuinfo_x86.h"
|
||||
|
||||
static const X86Features features = GetX86Info().features;
|
||||
|
||||
void Compute(void) {
|
||||
if (features.aes && features.sse4_2) {
|
||||
// Run optimized code.
|
||||
} else {
|
||||
// Run standard code.
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### Caching for faster evaluation of complex checks
|
||||
|
||||
If you wish, you can read all the features at once into a global variable, and
|
||||
then query for the specific features you care about. Below, we store all the ARM
|
||||
features and then check whether AES and NEON are supported.
|
||||
|
||||
```c
|
||||
#include <stdbool.h>
|
||||
#include "cpuinfo_arm.h"
|
||||
|
||||
static const ArmFeatures features = GetArmInfo().features;
|
||||
static const bool has_aes_and_neon = features.aes && features.neon;
|
||||
|
||||
// use has_aes_and_neon.
|
||||
```
|
||||
|
||||
This is a good approach to take if you're checking for combinations of features
|
||||
when using a compiler that is slow to extract individual bits from bit-packed
|
||||
structures.
|
||||
|
||||
### Checking compile time flags
|
||||
|
||||
The following code determines whether the compiler was told to use the AVX
|
||||
instruction set (e.g., `g++ -mavx`) and sets `has_avx` accordingly.
|
||||
|
||||
```c
|
||||
#include <stdbool.h>
|
||||
#include "cpuinfo_x86.h"
|
||||
|
||||
static const X86Features features = GetX86Info().features;
|
||||
static const bool has_avx = CPU_FEATURES_COMPILED_X86_AVX || features.avx;
|
||||
|
||||
// use has_avx.
|
||||
```
|
||||
|
||||
`CPU_FEATURES_COMPILED_X86_AVX` is set to 1 if the compiler was instructed to
|
||||
use AVX and 0 otherwise, combining compile time and runtime knowledge.
|
||||
|
||||
### Rejecting poor hardware implementations based on microarchitecture
|
||||
|
||||
On x86, the first incarnation of a feature in a microarchitecture might not be
|
||||
the most efficient (e.g. AVX on Sandy Bridge). We provide a function to retrieve
|
||||
the underlying microarchitecture so you can decide whether to use it.
|
||||
|
||||
Below, `has_fast_avx` is set to 1 if the CPU supports the AVX instruction
|
||||
set—but only if it's not Sandy Bridge.
|
||||
|
||||
```c
|
||||
#include <stdbool.h>
|
||||
#include "cpuinfo_x86.h"
|
||||
|
||||
static const X86Info info = GetX86Info();
|
||||
static const X86Microarchitecture uarch = GetX86Microarchitecture(&info);
|
||||
static const bool has_fast_avx = info.features.avx && uarch != INTEL_SNB;
|
||||
|
||||
// use has_fast_avx.
|
||||
```
|
||||
|
||||
This feature is currently available only for x86 microarchitectures.
|
||||
|
||||
<a name="usagesample"></a>
|
||||
### Running sample code
|
||||
|
||||
Building `cpu_features` brings a small executable to test the library.
|
||||
|
||||
```shell
|
||||
% ./build/list_cpu_features
|
||||
arch : x86
|
||||
brand : Intel(R) Xeon(R) CPU E5-1650 0 @ 3.20GHz
|
||||
family : 6 (0x06)
|
||||
model : 45 (0x2D)
|
||||
stepping : 7 (0x07)
|
||||
uarch : INTEL_SNB
|
||||
flags : aes,avx,cx16,smx,sse4_1,sse4_2,ssse3
|
||||
```
|
||||
|
||||
```shell
|
||||
% ./build/list_cpu_features --json
|
||||
{"arch":"x86","brand":" Intel(R) Xeon(R) CPU E5-1650 0 @ 3.20GHz","family":6,"model":45,"stepping":7,"uarch":"INTEL_SNB","flags":["aes","avx","cx16","smx","sse4_1","sse4_2","ssse3"]}
|
||||
```
|
||||
|
||||
<a name="support"></a>
|
||||
## What's supported
|
||||
|
||||
| | x86³ | ARM | AArch64 | MIPSel | POWER |
|
||||
|---------|:----:|:-------:|:-------:|:------:|:-------:|
|
||||
| Android | yes² | yes¹ | yes¹ | yes¹ | N/A |
|
||||
| iOS | N/A | not yet | not yet | N/A | N/A |
|
||||
| Linux | yes² | yes¹ | yes¹ | yes¹ | yes¹ |
|
||||
| MacOs | yes² | N/A | not yet | N/A | no |
|
||||
| Windows | yes² | not yet | not yet | N/A | N/A |
|
||||
|
||||
1. **Features revealed from Linux.** We gather data from several sources
|
||||
depending on availability:
|
||||
+ from glibc's
|
||||
[getauxval](https://www.gnu.org/software/libc/manual/html_node/Auxiliary-Vector.html)
|
||||
+ by parsing `/proc/self/auxv`
|
||||
+ by parsing `/proc/cpuinfo`
|
||||
2. **Features revealed from CPU.** features are retrieved by using the `cpuid`
|
||||
instruction.
|
||||
3. **Microarchitecture detection.** On x86 some features are not always
|
||||
implemented efficiently in hardware (e.g. AVX on Sandybridge). Exposing the
|
||||
microarchitecture allows the client to reject particular microarchitectures.
|
||||
|
||||
|
||||
<a name="license"></a>
|
||||
## License
|
||||
|
||||
The cpu_features library is licensed under the terms of the Apache license.
|
||||
See [LICENSE](LICENSE) for more information.
|
||||
|
||||
<a name="cmake"></a>
|
||||
## Build with CMake
|
||||
|
||||
Please check the [CMake build instructions](cmake/README.md).
|
||||
7
src/crypto/argon2_hasher/hash/cpu/cpu_features/WORKSPACE
Executable file
7
src/crypto/argon2_hasher/hash/cpu/cpu_features/WORKSPACE
Executable file
|
|
@ -0,0 +1,7 @@
|
|||
# ===== googletest =====
|
||||
|
||||
git_repository(
|
||||
name = "com_google_googletest",
|
||||
remote = "https://github.com/google/googletest.git",
|
||||
commit = "c3f65335b79f47b05629e79a54685d899bc53b93",
|
||||
)
|
||||
24
src/crypto/argon2_hasher/hash/cpu/cpu_features/appveyor.yml
Executable file
24
src/crypto/argon2_hasher/hash/cpu/cpu_features/appveyor.yml
Executable file
|
|
@ -0,0 +1,24 @@
|
|||
version: '{build}'
|
||||
shallow_clone: true
|
||||
|
||||
platform: x64
|
||||
|
||||
environment:
|
||||
matrix:
|
||||
- APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2017
|
||||
CMAKE_GENERATOR: "Visual Studio 15 2017 Win64"
|
||||
- APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2015
|
||||
CMAKE_GENERATOR: "Visual Studio 14 2015 Win64"
|
||||
|
||||
matrix:
|
||||
fast_finish: true
|
||||
|
||||
before_build:
|
||||
- cmake --version
|
||||
- cmake -DCMAKE_BUILD_TYPE=Debug -DBUILD_TESTING=ON -H. -Bcmake_build -G "%CMAKE_GENERATOR%"
|
||||
|
||||
build_script:
|
||||
- cmake --build cmake_build --config Debug --target ALL_BUILD
|
||||
|
||||
test_script:
|
||||
- cmake --build cmake_build --config Debug --target RUN_TESTS
|
||||
|
|
@ -0,0 +1,3 @@
|
|||
# CpuFeatures CMake configuration file
|
||||
|
||||
include("${CMAKE_CURRENT_LIST_DIR}/CpuFeaturesTargets.cmake")
|
||||
28
src/crypto/argon2_hasher/hash/cpu/cpu_features/cmake/README.md
Executable file
28
src/crypto/argon2_hasher/hash/cpu/cpu_features/cmake/README.md
Executable file
|
|
@ -0,0 +1,28 @@
|
|||
# CMake build instructions
|
||||
|
||||
## Recommended usage : Incorporating cpu_features into a CMake project
|
||||
|
||||
For API / ABI compatibility reasons, it is recommended to build and use
|
||||
cpu_features in a subdirectory of your project or as an embedded dependency.
|
||||
|
||||
This is similar to the recommended usage of the googletest framework
|
||||
( https://github.com/google/googletest/blob/master/googletest/README.md )
|
||||
|
||||
Build and use step-by-step
|
||||
|
||||
|
||||
1- Download cpu_features and copy it in a sub-directory in your project.
|
||||
or add cpu_features as a git-submodule in your project
|
||||
|
||||
2- You can then use the cmake command `add_subdirectory()` to include
|
||||
cpu_features directly and use the `cpu_features` target in your project.
|
||||
|
||||
3- Add the `cpu_features` target to the `target_link_libraries()` section of
|
||||
your executable or of your library.
|
||||
|
||||
## Enabling tests
|
||||
|
||||
CMake default options for cpu_features is Release built type with tests
|
||||
disabled. To enable testing set cmake `BUILD_TESTING` variable to `ON`,
|
||||
[.travis.yml](../.travis.yml) and [appveyor.yml](../appveyor.yml) have up to
|
||||
date examples.
|
||||
|
|
@ -0,0 +1,15 @@
|
|||
cmake_minimum_required(VERSION 2.8.2)
|
||||
|
||||
project(googletest-download NONE)
|
||||
|
||||
include(ExternalProject)
|
||||
ExternalProject_Add(googletest
|
||||
GIT_REPOSITORY https://github.com/google/googletest.git
|
||||
GIT_TAG master
|
||||
SOURCE_DIR "${CMAKE_BINARY_DIR}/googletest-src"
|
||||
BINARY_DIR "${CMAKE_BINARY_DIR}/googletest-build"
|
||||
CONFIGURE_COMMAND ""
|
||||
BUILD_COMMAND ""
|
||||
INSTALL_COMMAND ""
|
||||
TEST_COMMAND ""
|
||||
)
|
||||
34
src/crypto/argon2_hasher/hash/cpu/cpu_features/cmake/mips32-linux-gcc.cmake
Executable file
34
src/crypto/argon2_hasher/hash/cpu/cpu_features/cmake/mips32-linux-gcc.cmake
Executable file
|
|
@ -0,0 +1,34 @@
|
|||
set(CMAKE_SYSTEM_NAME "Linux")
|
||||
set(CMAKE_SYSTEM_PROCESSOR "mips32")
|
||||
|
||||
if (ENABLE_DSPR2 AND ENABLE_MSA)
|
||||
message(FATAL_ERROR "ENABLE_DSPR2 and ENABLE_MSA cannot be combined.")
|
||||
endif ()
|
||||
|
||||
if (ENABLE_DSPR2)
|
||||
set(HAVE_DSPR2 1 CACHE BOOL "" FORCE)
|
||||
set(MIPS_CFLAGS "-mdspr2")
|
||||
set(MIPS_CXXFLAGS "-mdspr2")
|
||||
elseif (ENABLE_MSA)
|
||||
set(HAVE_MSA 1 CACHE BOOL "" FORCE)
|
||||
set(MIPS_CFLAGS "-mmsa")
|
||||
set(MIPS_CXXFLAGS "-mmsa")
|
||||
endif ()
|
||||
|
||||
if ("${MIPS_CPU}" STREQUAL "")
|
||||
set(MIPS_CFLAGS "${MIPS_CFLAGS} -mips32r2")
|
||||
set(MIPS_CXXFLAGS "${MIPS_CXXFLAGS} -mips32r2")
|
||||
elseif ("${MIPS_CPU}" STREQUAL "p5600")
|
||||
set(P56_FLAGS "-mips32r5 -mload-store-pairs -msched-weight -mhard-float -mfp64")
|
||||
set(MIPS_CFLAGS "${MIPS_CFLAGS} ${P56_FLAGS}")
|
||||
set(MIPS_CXXFLAGS "${MIPS_CXXFLAGS} ${P56_FLAGS}")
|
||||
set(CMAKE_EXE_LINKER_FLAGS "-mfp64 ${CMAKE_EXE_LINKER_FLAGS}")
|
||||
endif ()
|
||||
|
||||
set(CMAKE_C_COMPILER ${CROSS}gcc)
|
||||
set(CMAKE_CXX_COMPILER ${CROSS}g++)
|
||||
set(AS_EXECUTABLE ${CROSS}as)
|
||||
set(CMAKE_C_COMPILER_ARG1 "-EL ${MIPS_CFLAGS}")
|
||||
set(CMAKE_CXX_COMPILER_ARG1 "-EL ${MIPS_CXXFLAGS}")
|
||||
|
||||
set(THREADS_PTHREAD_ARG "2" CACHE STRING "Forcibly set by CMakeLists.txt." FORCE)
|
||||
125
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpu_features_macros.h
Executable file
125
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpu_features_macros.h
Executable file
|
|
@ -0,0 +1,125 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_CPU_FEATURES_MACROS_H_
|
||||
#define CPU_FEATURES_INCLUDE_CPU_FEATURES_MACROS_H_
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Architectures
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
#if ((defined(_M_IX86) || defined(_M_X64) || defined(__i386__) || \
|
||||
defined(__x86_64__)) && \
|
||||
!defined(__pnacl__) && !defined(__CLR_VER))
|
||||
#define CPU_FEATURES_ARCH_X86
|
||||
#endif
|
||||
|
||||
#if (defined(__arm__) || defined(_M_ARM))
|
||||
#define CPU_FEATURES_ARCH_ARM
|
||||
#endif
|
||||
|
||||
#if defined(__aarch64__)
|
||||
#define CPU_FEATURES_ARCH_AARCH64
|
||||
#endif
|
||||
|
||||
#if (defined(CPU_FEATURES_ARCH_AARCH64) || defined(CPU_FEATURES_ARCH_ARM))
|
||||
#define CPU_FEATURES_ARCH_ANY_ARM
|
||||
#endif
|
||||
|
||||
#if defined(__mips__)
|
||||
#define CPU_FEATURES_ARCH_MIPS
|
||||
#endif
|
||||
|
||||
#if defined(__powerpc__)
|
||||
#define CPU_FEATURES_ARCH_PPC
|
||||
#endif
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Os
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
#if defined(__linux__)
|
||||
#define CPU_FEATURES_OS_LINUX_OR_ANDROID
|
||||
#endif
|
||||
|
||||
#if defined(__ANDROID__)
|
||||
#define CPU_FEATURES_OS_ANDROID
|
||||
#endif
|
||||
|
||||
#if (defined(_WIN64) || defined(_WIN32))
|
||||
#define CPU_FEATURES_OS_WINDOWS
|
||||
#endif
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Compilers
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
#if defined(__clang__)
|
||||
#define CPU_FEATURES_COMPILER_CLANG
|
||||
#endif
|
||||
|
||||
#if defined(__GNUC__) && !defined(__clang__)
|
||||
#define CPU_FEATURES_COMPILER_GCC
|
||||
#endif
|
||||
|
||||
#if defined(_MSC_VER)
|
||||
#define CPU_FEATURES_COMPILER_MSC
|
||||
#endif
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Cpp
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
#if defined(__cplusplus)
|
||||
#define CPU_FEATURES_START_CPP_NAMESPACE \
|
||||
namespace cpu_features { \
|
||||
extern "C" {
|
||||
#define CPU_FEATURES_END_CPP_NAMESPACE \
|
||||
} \
|
||||
}
|
||||
#else
|
||||
#define CPU_FEATURES_START_CPP_NAMESPACE
|
||||
#define CPU_FEATURES_END_CPP_NAMESPACE
|
||||
#endif
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Compiler flags
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
// Use the following to check if a feature is known to be available at compile
|
||||
// time. See README.md for an example.
|
||||
#if defined(CPU_FEATURES_ARCH_X86)
|
||||
#define CPU_FEATURES_COMPILED_X86_AES defined(__AES__)
|
||||
#define CPU_FEATURES_COMPILED_X86_F16C defined(__F16C__)
|
||||
#define CPU_FEATURES_COMPILED_X86_BMI defined(__BMI__)
|
||||
#define CPU_FEATURES_COMPILED_X86_BMI2 defined(__BMI2__)
|
||||
#define CPU_FEATURES_COMPILED_X86_SSE (defined(__SSE__) || (_M_IX86_FP >= 1))
|
||||
#define CPU_FEATURES_COMPILED_X86_SSE2 (defined(__SSE2__) || (_M_IX86_FP >= 2))
|
||||
#define CPU_FEATURES_COMPILED_X86_SSE3 defined(__SSE3__)
|
||||
#define CPU_FEATURES_COMPILED_X86_SSSE3 defined(__SSSE3__)
|
||||
#define CPU_FEATURES_COMPILED_X86_SSE4_1 defined(__SSE4_1__)
|
||||
#define CPU_FEATURES_COMPILED_X86_SSE4_2 defined(__SSE4_2__)
|
||||
#define CPU_FEATURES_COMPILED_X86_AVX defined(__AVX__)
|
||||
#define CPU_FEATURES_COMPILED_x86_AVX2 defined(__AVX2__)
|
||||
#endif
|
||||
|
||||
#if defined(CPU_FEATURES_ARCH_ANY_ARM)
|
||||
#define CPU_FEATURES_COMPILED_ANY_ARM_NEON defined(__ARM_NEON__)
|
||||
#endif
|
||||
|
||||
#if defined(CPU_FEATURES_ARCH_MIPS)
|
||||
#define CPU_FEATURES_COMPILED_MIPS_MSA defined(__mips_msa)
|
||||
#endif
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_CPU_FEATURES_MACROS_H_
|
||||
65
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_aarch64.h
Executable file
65
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_aarch64.h
Executable file
|
|
@ -0,0 +1,65 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_CPUINFO_AARCH64_H_
|
||||
#define CPU_FEATURES_INCLUDE_CPUINFO_AARCH64_H_
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
typedef struct {
|
||||
int fp : 1; // Floating-point.
|
||||
int asimd : 1; // Advanced SIMD.
|
||||
int aes : 1; // Hardware-accelerated Advanced Encryption Standard.
|
||||
int pmull : 1; // Polynomial multiply long.
|
||||
int sha1 : 1; // Hardware-accelerated SHA1.
|
||||
int sha2 : 1; // Hardware-accelerated SHA2-256.
|
||||
int crc32 : 1; // Hardware-accelerated CRC-32.
|
||||
|
||||
// Make sure to update Aarch64FeaturesEnum below if you add a field here.
|
||||
} Aarch64Features;
|
||||
|
||||
typedef struct {
|
||||
Aarch64Features features;
|
||||
int implementer;
|
||||
int variant;
|
||||
int part;
|
||||
int revision;
|
||||
} Aarch64Info;
|
||||
|
||||
Aarch64Info GetAarch64Info(void);
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
typedef enum {
|
||||
AARCH64_FP,
|
||||
AARCH64_ASIMD,
|
||||
AARCH64_AES,
|
||||
AARCH64_PMULL,
|
||||
AARCH64_SHA1,
|
||||
AARCH64_SHA2,
|
||||
AARCH64_CRC32,
|
||||
AARCH64_LAST_,
|
||||
} Aarch64FeaturesEnum;
|
||||
|
||||
int GetAarch64FeaturesEnumValue(const Aarch64Features* features,
|
||||
Aarch64FeaturesEnum value);
|
||||
|
||||
const char* GetAarch64FeaturesEnumName(Aarch64FeaturesEnum);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_CPUINFO_AARCH64_H_
|
||||
80
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_arm.h
Executable file
80
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_arm.h
Executable file
|
|
@ -0,0 +1,80 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_CPUINFO_ARM_H_
|
||||
#define CPU_FEATURES_INCLUDE_CPUINFO_ARM_H_
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
typedef struct {
|
||||
int vfp : 1; // Vector Floating Point.
|
||||
int iwmmxt : 1; // Intel Wireless MMX Technology.
|
||||
int neon : 1; // Advanced SIMD.
|
||||
int vfpv3 : 1; // VFP version 3
|
||||
int vfpv3d16 : 1; // VFP version 3 with 16 D-registers
|
||||
int vfpv4 : 1; // VFP version 4 with fast context switching
|
||||
int idiva : 1; // SDIV and UDIV hardware division in ARM mode.
|
||||
int idivt : 1; // SDIV and UDIV hardware division in Thumb mode.
|
||||
int aes : 1; // Hardware-accelerated Advanced Encryption Standard.
|
||||
int pmull : 1; // Polynomial multiply long.
|
||||
int sha1 : 1; // Hardware-accelerated SHA1.
|
||||
int sha2 : 1; // Hardware-accelerated SHA2-256.
|
||||
int crc32 : 1; // Hardware-accelerated CRC-32.
|
||||
|
||||
// Make sure to update ArmFeaturesEnum below if you add a field here.
|
||||
} ArmFeatures;
|
||||
|
||||
typedef struct {
|
||||
ArmFeatures features;
|
||||
int implementer;
|
||||
int architecture;
|
||||
int variant;
|
||||
int part;
|
||||
int revision;
|
||||
} ArmInfo;
|
||||
|
||||
// TODO(user): Add macros to know which features are present at compile
|
||||
// time.
|
||||
|
||||
ArmInfo GetArmInfo(void);
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
typedef enum {
|
||||
ARM_VFP,
|
||||
ARM_IWMMXT,
|
||||
ARM_NEON,
|
||||
ARM_VFPV3,
|
||||
ARM_VFPV3D16,
|
||||
ARM_VFPV4,
|
||||
ARM_IDIVA,
|
||||
ARM_IDIVT,
|
||||
ARM_AES,
|
||||
ARM_PMULL,
|
||||
ARM_SHA1,
|
||||
ARM_SHA2,
|
||||
ARM_CRC32,
|
||||
ARM_LAST_,
|
||||
} ArmFeaturesEnum;
|
||||
|
||||
int GetArmFeaturesEnumValue(const ArmFeatures* features, ArmFeaturesEnum value);
|
||||
|
||||
const char* GetArmFeaturesEnumName(ArmFeaturesEnum);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_CPUINFO_ARM_H_
|
||||
53
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_mips.h
Executable file
53
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_mips.h
Executable file
|
|
@ -0,0 +1,53 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_CPUINFO_MIPS_H_
|
||||
#define CPU_FEATURES_INCLUDE_CPUINFO_MIPS_H_
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
typedef struct {
|
||||
int msa : 1; // MIPS SIMD Architecture
|
||||
// https://www.mips.com/products/architectures/ase/simd/
|
||||
int eva : 1; // Enhanced Virtual Addressing
|
||||
// https://www.mips.com/products/architectures/mips64/
|
||||
|
||||
// Make sure to update MipsFeaturesEnum below if you add a field here.
|
||||
} MipsFeatures;
|
||||
|
||||
typedef struct {
|
||||
MipsFeatures features;
|
||||
} MipsInfo;
|
||||
|
||||
MipsInfo GetMipsInfo(void);
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
typedef enum {
|
||||
MIPS_MSA,
|
||||
MIPS_EVA,
|
||||
MIPS_LAST_,
|
||||
} MipsFeaturesEnum;
|
||||
|
||||
int GetMipsFeaturesEnumValue(const MipsFeatures* features,
|
||||
MipsFeaturesEnum value);
|
||||
|
||||
const char* GetMipsFeaturesEnumName(MipsFeaturesEnum);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_CPUINFO_MIPS_H_
|
||||
141
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_ppc.h
Executable file
141
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_ppc.h
Executable file
|
|
@ -0,0 +1,141 @@
|
|||
// Copyright 2018 IBM
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_CPUINFO_PPC_H_
|
||||
#define CPU_FEATURES_INCLUDE_CPUINFO_PPC_H_
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
#include "internal/hwcaps.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
typedef struct {
|
||||
int ppc32 : 1;
|
||||
int ppc64 : 1;
|
||||
int ppc601 : 1;
|
||||
int altivec : 1;
|
||||
int fpu : 1;
|
||||
int mmu : 1;
|
||||
int mac_4xx : 1;
|
||||
int unifiedcache : 1;
|
||||
int spe : 1;
|
||||
int efpsingle : 1;
|
||||
int efpdouble : 1;
|
||||
int no_tb : 1;
|
||||
int power4 : 1;
|
||||
int power5 : 1;
|
||||
int power5plus : 1;
|
||||
int cell : 1;
|
||||
int booke : 1;
|
||||
int smt : 1;
|
||||
int icachesnoop : 1;
|
||||
int arch205 : 1;
|
||||
int pa6t : 1;
|
||||
int dfp : 1;
|
||||
int power6ext : 1;
|
||||
int arch206 : 1;
|
||||
int vsx : 1;
|
||||
int pseries_perfmon_compat : 1;
|
||||
int truele : 1;
|
||||
int ppcle : 1;
|
||||
int arch207 : 1;
|
||||
int htm : 1;
|
||||
int dscr : 1;
|
||||
int ebb : 1;
|
||||
int isel : 1;
|
||||
int tar : 1;
|
||||
int vcrypto : 1;
|
||||
int htm_nosc : 1;
|
||||
int arch300 : 1;
|
||||
int ieee128 : 1;
|
||||
int darn : 1;
|
||||
int scv : 1;
|
||||
int htm_no_suspend : 1;
|
||||
|
||||
// Make sure to update PPCFeaturesEnum below if you add a field here.
|
||||
} PPCFeatures;
|
||||
|
||||
typedef struct {
|
||||
PPCFeatures features;
|
||||
} PPCInfo;
|
||||
|
||||
// This function is guaranteed to be malloc, memset and memcpy free.
|
||||
PPCInfo GetPPCInfo(void);
|
||||
|
||||
typedef struct {
|
||||
char platform[64]; // 0 terminated string
|
||||
char model[64]; // 0 terminated string
|
||||
char machine[64]; // 0 terminated string
|
||||
char cpu[64]; // 0 terminated string
|
||||
PlatformType type;
|
||||
} PPCPlatformStrings;
|
||||
|
||||
PPCPlatformStrings GetPPCPlatformStrings(void);
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
typedef enum {
|
||||
PPC_32, /* 32 bit mode execution */
|
||||
PPC_64, /* 64 bit mode execution */
|
||||
PPC_601_INSTR, /* Old POWER ISA */
|
||||
PPC_HAS_ALTIVEC, /* SIMD Unit*/
|
||||
PPC_HAS_FPU, /* Floating Point Unit */
|
||||
PPC_HAS_MMU, /* Memory management unit */
|
||||
PPC_HAS_4xxMAC,
|
||||
PPC_UNIFIED_CACHE, /* Unified instruction and data cache */
|
||||
PPC_HAS_SPE, /* Signal processing extention unit */
|
||||
PPC_HAS_EFP_SINGLE, /* SPE single precision fpu */
|
||||
PPC_HAS_EFP_DOUBLE, /* SPE double precision fpu */
|
||||
PPC_NO_TB, /* No timebase */
|
||||
PPC_POWER4,
|
||||
PPC_POWER5,
|
||||
PPC_POWER5_PLUS,
|
||||
PPC_CELL, /* Cell broadband engine */
|
||||
PPC_BOOKE, /* Embedded ISA */
|
||||
PPC_SMT, /* Simultaneous multi-threading */
|
||||
PPC_ICACHE_SNOOP,
|
||||
PPC_ARCH_2_05, /* ISA 2.05 - POWER6 */
|
||||
PPC_PA6T, /* PA Semi 6T core ISA */
|
||||
PPC_HAS_DFP, /* Decimal floating point unit */
|
||||
PPC_POWER6_EXT,
|
||||
PPC_ARCH_2_06, /* ISA 2.06 - POWER7 */
|
||||
PPC_HAS_VSX, /* Vector-scalar extension */
|
||||
PPC_PSERIES_PERFMON_COMPAT, /* Set of backwards compatibile performance
|
||||
monitoring events */
|
||||
PPC_TRUE_LE,
|
||||
PPC_PPC_LE,
|
||||
PPC_ARCH_2_07, /* ISA 2.07 - POWER8 */
|
||||
PPC_HTM, /* Hardware Transactional Memory */
|
||||
PPC_DSCR, /* Data stream control register */
|
||||
PPC_EBB, /* Event base branching */
|
||||
PPC_ISEL, /* Integer select instructions */
|
||||
PPC_TAR, /* Target address register */
|
||||
PPC_VEC_CRYPTO, /* Vector cryptography instructions */
|
||||
PPC_HTM_NOSC, /* Transactions aborted when syscall made*/
|
||||
PPC_ARCH_3_00, /* ISA 3.00 - POWER9 */
|
||||
PPC_HAS_IEEE128, /* VSX IEEE Binary Float 128-bit */
|
||||
PPC_DARN, /* Deliver a random number instruction */
|
||||
PPC_SCV, /* scv syscall */
|
||||
PPC_HTM_NO_SUSPEND, /* TM w/out suspended state */
|
||||
PPC_LAST_,
|
||||
} PPCFeaturesEnum;
|
||||
|
||||
int GetPPCFeaturesEnumValue(const PPCFeatures* features, PPCFeaturesEnum value);
|
||||
|
||||
const char* GetPPCFeaturesEnumName(PPCFeaturesEnum);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_CPUINFO_PPC_H_
|
||||
154
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_x86.h
Executable file
154
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/cpuinfo_x86.h
Executable file
|
|
@ -0,0 +1,154 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_CPUINFO_X86_H_
|
||||
#define CPU_FEATURES_INCLUDE_CPUINFO_X86_H_
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
// See https://en.wikipedia.org/wiki/CPUID for a list of x86 cpu features.
|
||||
typedef struct {
|
||||
int aes : 1;
|
||||
int erms : 1;
|
||||
int f16c : 1;
|
||||
int fma3 : 1;
|
||||
int vpclmulqdq : 1;
|
||||
int bmi1 : 1;
|
||||
int bmi2 : 1;
|
||||
|
||||
int ssse3 : 1;
|
||||
int sse4_1 : 1;
|
||||
int sse4_2 : 1;
|
||||
|
||||
int avx : 1;
|
||||
int avx2 : 1;
|
||||
|
||||
int avx512f : 1;
|
||||
int avx512cd : 1;
|
||||
int avx512er : 1;
|
||||
int avx512pf : 1;
|
||||
int avx512bw : 1;
|
||||
int avx512dq : 1;
|
||||
int avx512vl : 1;
|
||||
int avx512ifma : 1;
|
||||
int avx512vbmi : 1;
|
||||
int avx512vbmi2 : 1;
|
||||
int avx512vnni : 1;
|
||||
int avx512bitalg : 1;
|
||||
int avx512vpopcntdq : 1;
|
||||
int avx512_4vnniw : 1;
|
||||
int avx512_4vbmi2 : 1;
|
||||
|
||||
int smx : 1;
|
||||
int sgx : 1;
|
||||
int cx16 : 1; // aka. CMPXCHG16B
|
||||
|
||||
// Make sure to update X86FeaturesEnum below if you add a field here.
|
||||
} X86Features;
|
||||
|
||||
typedef struct {
|
||||
X86Features features;
|
||||
int family;
|
||||
int model;
|
||||
int stepping;
|
||||
char vendor[13]; // 0 terminated string
|
||||
} X86Info;
|
||||
|
||||
// Calls cpuid and returns an initialized X86info.
|
||||
// This function is guaranteed to be malloc, memset and memcpy free.
|
||||
X86Info GetX86Info(void);
|
||||
|
||||
typedef enum {
|
||||
X86_UNKNOWN,
|
||||
INTEL_CORE, // CORE
|
||||
INTEL_PNR, // PENRYN
|
||||
INTEL_NHM, // NEHALEM
|
||||
INTEL_ATOM_BNL, // BONNELL
|
||||
INTEL_WSM, // WESTMERE
|
||||
INTEL_SNB, // SANDYBRIDGE
|
||||
INTEL_IVB, // IVYBRIDGE
|
||||
INTEL_ATOM_SMT, // SILVERMONT
|
||||
INTEL_HSW, // HASWELL
|
||||
INTEL_BDW, // BROADWELL
|
||||
INTEL_SKL, // SKYLAKE
|
||||
INTEL_ATOM_GMT, // GOLDMONT
|
||||
INTEL_KBL, // KABY LAKE
|
||||
INTEL_CFL, // COFFEE LAKE
|
||||
INTEL_CNL, // CANNON LAKE
|
||||
AMD_HAMMER, // K8
|
||||
AMD_K10, // K10
|
||||
AMD_BOBCAT, // K14
|
||||
AMD_BULLDOZER, // K15
|
||||
AMD_JAGUAR, // K16
|
||||
AMD_ZEN, // K17
|
||||
} X86Microarchitecture;
|
||||
|
||||
// Returns the underlying microarchitecture by looking at X86Info's vendor,
|
||||
// family and model.
|
||||
X86Microarchitecture GetX86Microarchitecture(const X86Info* info);
|
||||
|
||||
// Calls cpuid and fills the brand_string.
|
||||
// - brand_string *must* be of size 49 (beware of array decaying).
|
||||
// - brand_string will be zero terminated.
|
||||
// - This function calls memcpy.
|
||||
void FillX86BrandString(char brand_string[49]);
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
typedef enum {
|
||||
X86_AES,
|
||||
X86_ERMS,
|
||||
X86_F16C,
|
||||
X86_FMA3,
|
||||
X86_VPCLMULQDQ,
|
||||
X86_BMI1,
|
||||
X86_BMI2,
|
||||
X86_SSSE3,
|
||||
X86_SSE4_1,
|
||||
X86_SSE4_2,
|
||||
X86_AVX,
|
||||
X86_AVX2,
|
||||
X86_AVX512F,
|
||||
X86_AVX512CD,
|
||||
X86_AVX512ER,
|
||||
X86_AVX512PF,
|
||||
X86_AVX512BW,
|
||||
X86_AVX512DQ,
|
||||
X86_AVX512VL,
|
||||
X86_AVX512IFMA,
|
||||
X86_AVX512VBMI,
|
||||
X86_AVX512VBMI2,
|
||||
X86_AVX512VNNI,
|
||||
X86_AVX512BITALG,
|
||||
X86_AVX512VPOPCNTDQ,
|
||||
X86_AVX512_4VNNIW,
|
||||
X86_AVX512_4VBMI2,
|
||||
X86_SMX,
|
||||
X86_SGX,
|
||||
X86_CX16,
|
||||
X86_LAST_,
|
||||
} X86FeaturesEnum;
|
||||
|
||||
int GetX86FeaturesEnumValue(const X86Features* features, X86FeaturesEnum value);
|
||||
|
||||
const char* GetX86FeaturesEnumName(X86FeaturesEnum);
|
||||
|
||||
const char* GetX86MicroarchitectureName(X86Microarchitecture);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_CPUINFO_X86_H_
|
||||
39
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/bit_utils.h
Executable file
39
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/bit_utils.h
Executable file
|
|
@ -0,0 +1,39 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_BIT_UTILS_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_BIT_UTILS_H_
|
||||
|
||||
#include <assert.h>
|
||||
#include <stdbool.h>
|
||||
#include <stdint.h>
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
inline static bool IsBitSet(uint32_t reg, uint32_t bit) {
|
||||
return (reg >> bit) & 0x1;
|
||||
}
|
||||
|
||||
inline static uint32_t ExtractBitRange(uint32_t reg, uint32_t msb,
|
||||
uint32_t lsb) {
|
||||
const uint64_t bits = msb - lsb + 1;
|
||||
const uint64_t mask = (1ULL << bits) - 1ULL;
|
||||
assert(msb >= lsb);
|
||||
return (reg >> lsb) & mask;
|
||||
}
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_BIT_UTILS_H_
|
||||
37
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/cpuid_x86.h
Executable file
37
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/cpuid_x86.h
Executable file
|
|
@ -0,0 +1,37 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_CPUID_X86_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_CPUID_X86_H_
|
||||
|
||||
#include <stdint.h>
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
// A struct to hold the result of a call to cpuid.
|
||||
typedef struct {
|
||||
uint32_t eax, ebx, ecx, edx;
|
||||
} Leaf;
|
||||
|
||||
// Retrieves the leaf for a particular cpuid.
|
||||
Leaf CpuId(uint32_t leaf_id);
|
||||
|
||||
// Returns the eax value of the XCR0 register.
|
||||
uint32_t GetXCR0Eax(void);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_CPUID_X86_H_
|
||||
38
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/filesystem.h
Executable file
38
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/filesystem.h
Executable file
|
|
@ -0,0 +1,38 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
// An interface for the filesystem that allows mocking the filesystem in
|
||||
// unittests.
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_FILESYSTEM_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_FILESYSTEM_H_
|
||||
|
||||
#include <stddef.h>
|
||||
#include <stdint.h>
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
// Same as linux "open(filename, O_RDONLY)", retries automatically on EINTR.
|
||||
int CpuFeatures_OpenFile(const char* filename);
|
||||
|
||||
// Same as linux "read(file_descriptor, buffer, buffer_size)", retries
|
||||
// automatically on EINTR.
|
||||
int CpuFeatures_ReadFile(int file_descriptor, void* buffer, size_t buffer_size);
|
||||
|
||||
// Same as linux "close(file_descriptor)".
|
||||
void CpuFeatures_CloseFile(int file_descriptor);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_FILESYSTEM_H_
|
||||
131
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/hwcaps.h
Executable file
131
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/hwcaps.h
Executable file
|
|
@ -0,0 +1,131 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
// Interface to retrieve hardware capabilities. It relies on Linux's getauxval
|
||||
// or `/proc/self/auxval` under the hood.
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_HWCAPS_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_HWCAPS_H_
|
||||
|
||||
#include <stdint.h>
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
// To avoid depending on the linux kernel we reproduce the architecture specific
|
||||
// constants here.
|
||||
|
||||
// http://elixir.free-electrons.com/linux/latest/source/arch/arm64/include/uapi/asm/hwcap.h
|
||||
#define AARCH64_HWCAP_FP (1UL << 0)
|
||||
#define AARCH64_HWCAP_ASIMD (1UL << 1)
|
||||
#define AARCH64_HWCAP_AES (1UL << 3)
|
||||
#define AARCH64_HWCAP_PMULL (1UL << 4)
|
||||
#define AARCH64_HWCAP_SHA1 (1UL << 5)
|
||||
#define AARCH64_HWCAP_SHA2 (1UL << 6)
|
||||
#define AARCH64_HWCAP_CRC32 (1UL << 7)
|
||||
|
||||
// http://elixir.free-electrons.com/linux/latest/source/arch/arm/include/uapi/asm/hwcap.h
|
||||
#define ARM_HWCAP_VFP (1UL << 6)
|
||||
#define ARM_HWCAP_IWMMXT (1UL << 9)
|
||||
#define ARM_HWCAP_NEON (1UL << 12)
|
||||
#define ARM_HWCAP_VFPV3 (1UL << 13)
|
||||
#define ARM_HWCAP_VFPV3D16 (1UL << 14)
|
||||
#define ARM_HWCAP_VFPV4 (1UL << 16)
|
||||
#define ARM_HWCAP_IDIVA (1UL << 17)
|
||||
#define ARM_HWCAP_IDIVT (1UL << 18)
|
||||
#define ARM_HWCAP2_AES (1UL << 0)
|
||||
#define ARM_HWCAP2_PMULL (1UL << 1)
|
||||
#define ARM_HWCAP2_SHA1 (1UL << 2)
|
||||
#define ARM_HWCAP2_SHA2 (1UL << 3)
|
||||
#define ARM_HWCAP2_CRC32 (1UL << 4)
|
||||
|
||||
// http://elixir.free-electrons.com/linux/latest/source/arch/mips/include/uapi/asm/hwcap.h
|
||||
#define MIPS_HWCAP_VZ (1UL << 0)
|
||||
#define MIPS_HWCAP_EVA (1UL << 1)
|
||||
#define MIPS_HWCAP_HTW (1UL << 2)
|
||||
#define MIPS_HWCAP_FPU (1UL << 3)
|
||||
#define MIPS_HWCAP_MIPS32R2 (1UL << 4)
|
||||
#define MIPS_HWCAP_MIPS32R5 (1UL << 5)
|
||||
#define MIPS_HWCAP_MIPS64R6 (1UL << 6)
|
||||
#define MIPS_HWCAP_DSPR1 (1UL << 7)
|
||||
#define MIPS_HWCAP_DSPR2 (1UL << 8)
|
||||
#define MIPS_HWCAP_MSA (1UL << 9)
|
||||
|
||||
// http://elixir.free-electrons.com/linux/latest/source/arch/powerpc/include/uapi/asm/cputable.h
|
||||
#ifndef _UAPI__ASM_POWERPC_CPUTABLE_H
|
||||
/* in AT_HWCAP */
|
||||
#define PPC_FEATURE_32 0x80000000
|
||||
#define PPC_FEATURE_64 0x40000000
|
||||
#define PPC_FEATURE_601_INSTR 0x20000000
|
||||
#define PPC_FEATURE_HAS_ALTIVEC 0x10000000
|
||||
#define PPC_FEATURE_HAS_FPU 0x08000000
|
||||
#define PPC_FEATURE_HAS_MMU 0x04000000
|
||||
#define PPC_FEATURE_HAS_4xxMAC 0x02000000
|
||||
#define PPC_FEATURE_UNIFIED_CACHE 0x01000000
|
||||
#define PPC_FEATURE_HAS_SPE 0x00800000
|
||||
#define PPC_FEATURE_HAS_EFP_SINGLE 0x00400000
|
||||
#define PPC_FEATURE_HAS_EFP_DOUBLE 0x00200000
|
||||
#define PPC_FEATURE_NO_TB 0x00100000
|
||||
#define PPC_FEATURE_POWER4 0x00080000
|
||||
#define PPC_FEATURE_POWER5 0x00040000
|
||||
#define PPC_FEATURE_POWER5_PLUS 0x00020000
|
||||
#define PPC_FEATURE_CELL 0x00010000
|
||||
#define PPC_FEATURE_BOOKE 0x00008000
|
||||
#define PPC_FEATURE_SMT 0x00004000
|
||||
#define PPC_FEATURE_ICACHE_SNOOP 0x00002000
|
||||
#define PPC_FEATURE_ARCH_2_05 0x00001000
|
||||
#define PPC_FEATURE_PA6T 0x00000800
|
||||
#define PPC_FEATURE_HAS_DFP 0x00000400
|
||||
#define PPC_FEATURE_POWER6_EXT 0x00000200
|
||||
#define PPC_FEATURE_ARCH_2_06 0x00000100
|
||||
#define PPC_FEATURE_HAS_VSX 0x00000080
|
||||
|
||||
#define PPC_FEATURE_PSERIES_PERFMON_COMPAT 0x00000040
|
||||
|
||||
/* Reserved - do not use 0x00000004 */
|
||||
#define PPC_FEATURE_TRUE_LE 0x00000002
|
||||
#define PPC_FEATURE_PPC_LE 0x00000001
|
||||
|
||||
/* in AT_HWCAP2 */
|
||||
#define PPC_FEATURE2_ARCH_2_07 0x80000000
|
||||
#define PPC_FEATURE2_HTM 0x40000000
|
||||
#define PPC_FEATURE2_DSCR 0x20000000
|
||||
#define PPC_FEATURE2_EBB 0x10000000
|
||||
#define PPC_FEATURE2_ISEL 0x08000000
|
||||
#define PPC_FEATURE2_TAR 0x04000000
|
||||
#define PPC_FEATURE2_VEC_CRYPTO 0x02000000
|
||||
#define PPC_FEATURE2_HTM_NOSC 0x01000000
|
||||
#define PPC_FEATURE2_ARCH_3_00 0x00800000
|
||||
#define PPC_FEATURE2_HAS_IEEE128 0x00400000
|
||||
#define PPC_FEATURE2_DARN 0x00200000
|
||||
#define PPC_FEATURE2_SCV 0x00100000
|
||||
#define PPC_FEATURE2_HTM_NO_SUSPEND 0x00080000
|
||||
#endif
|
||||
|
||||
typedef struct {
|
||||
unsigned long hwcaps;
|
||||
unsigned long hwcaps2;
|
||||
} HardwareCapabilities;
|
||||
|
||||
HardwareCapabilities CpuFeatures_GetHardwareCapabilities(void);
|
||||
|
||||
typedef struct {
|
||||
char platform[64]; // 0 terminated string
|
||||
char base_platform[64]; // 0 terminated string
|
||||
} PlatformType;
|
||||
|
||||
PlatformType CpuFeatures_GetPlatformType(void);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_HWCAPS_H_
|
||||
|
|
@ -0,0 +1,60 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
// CapabilityConfig provides a way to map cpu features to hardware caps and
|
||||
// /proc/cpuinfo flags. We then provide functions to update capabilities from
|
||||
// either source.
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_LINUX_FEATURES_AGGREGATOR_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_LINUX_FEATURES_AGGREGATOR_H_
|
||||
|
||||
#include <ctype.h>
|
||||
#include <stdint.h>
|
||||
#include "cpu_features_macros.h"
|
||||
#include "internal/hwcaps.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
// Use the following macro to declare setter functions to be used in
|
||||
// CapabilityConfig.
|
||||
#define DECLARE_SETTER(FeatureType, FeatureName) \
|
||||
static void set_##FeatureName(void* const features, bool value) { \
|
||||
((FeatureType*)features)->FeatureName = value; \
|
||||
}
|
||||
|
||||
// Describes the relationship between hardware caps and /proc/cpuinfo flags.
|
||||
typedef struct {
|
||||
const HardwareCapabilities hwcaps_mask;
|
||||
const char* const proc_cpuinfo_flag;
|
||||
void (*set_bit)(void* const, bool); // setter for the corresponding bit.
|
||||
} CapabilityConfig;
|
||||
|
||||
// For every config, looks into flags_line for the presence of the
|
||||
// corresponding proc_cpuinfo_flag, calls `set_bit` accordingly.
|
||||
// Note: features is a pointer to the underlying Feature struct.
|
||||
void CpuFeatures_SetFromFlags(const size_t configs_size,
|
||||
const CapabilityConfig* configs,
|
||||
const StringView flags_line,
|
||||
void* const features);
|
||||
|
||||
// For every config, looks into hwcaps for the presence of the feature. Calls
|
||||
// `set_bit` with true if the hardware capability is found.
|
||||
// Note: features is a pointer to the underlying Feature struct.
|
||||
void CpuFeatures_OverrideFromHwCaps(const size_t configs_size,
|
||||
const CapabilityConfig* configs,
|
||||
const HardwareCapabilities hwcaps,
|
||||
void* const features);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_LINUX_FEATURES_AGGREGATOR_H_
|
||||
|
|
@ -0,0 +1,49 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
// Reads a file line by line and stores the data on the stack. This allows
|
||||
// parsing files in one go without allocating.
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_STACK_LINE_READER_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_STACK_LINE_READER_H_
|
||||
|
||||
#include <stdbool.h>
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
typedef struct {
|
||||
char buffer[STACK_LINE_READER_BUFFER_SIZE];
|
||||
StringView view;
|
||||
int fd;
|
||||
bool skip_mode;
|
||||
} StackLineReader;
|
||||
|
||||
// Initializes a StackLineReader.
|
||||
void StackLineReader_Initialize(StackLineReader* reader, int fd);
|
||||
|
||||
typedef struct {
|
||||
StringView line; // A view of the line.
|
||||
bool eof; // Nothing more to read, we reached EOF.
|
||||
bool full_line; // If false the line was truncated to
|
||||
// STACK_LINE_READER_BUFFER_SIZE.
|
||||
} LineResult;
|
||||
|
||||
// Reads the file pointed to by fd and tries to read a full line.
|
||||
LineResult StackLineReader_NextLine(StackLineReader* reader);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_STACK_LINE_READER_H_
|
||||
108
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/string_view.h
Executable file
108
src/crypto/argon2_hasher/hash/cpu/cpu_features/include/internal/string_view.h
Executable file
|
|
@ -0,0 +1,108 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
// A view over a piece of string. The view is not 0 terminated.
|
||||
#ifndef CPU_FEATURES_INCLUDE_INTERNAL_STRING_VIEW_H_
|
||||
#define CPU_FEATURES_INCLUDE_INTERNAL_STRING_VIEW_H_
|
||||
|
||||
#include <stdbool.h>
|
||||
#include <stddef.h>
|
||||
#include <string.h>
|
||||
#include "cpu_features_macros.h"
|
||||
|
||||
CPU_FEATURES_START_CPP_NAMESPACE
|
||||
|
||||
typedef struct {
|
||||
const char* ptr;
|
||||
size_t size;
|
||||
} StringView;
|
||||
|
||||
#ifdef __cplusplus
|
||||
static const StringView kEmptyStringView = {NULL, 0};
|
||||
#else
|
||||
static const StringView kEmptyStringView;
|
||||
#endif
|
||||
|
||||
// Returns a StringView from the provided string.
|
||||
// Passing NULL is valid only if size is 0.
|
||||
static inline StringView view(const char* str, const size_t size) {
|
||||
StringView view;
|
||||
view.ptr = str;
|
||||
view.size = size;
|
||||
return view;
|
||||
}
|
||||
|
||||
static inline StringView str(const char* str) { return view(str, strlen(str)); }
|
||||
|
||||
// Returns the index of the first occurrence of c in view or -1 if not found.
|
||||
int CpuFeatures_StringView_IndexOfChar(const StringView view, char c);
|
||||
|
||||
// Returns the index of the first occurrence of sub_view in view or -1 if not
|
||||
// found.
|
||||
int CpuFeatures_StringView_IndexOf(const StringView view,
|
||||
const StringView sub_view);
|
||||
|
||||
// Returns whether a is equal to b (same content).
|
||||
bool CpuFeatures_StringView_IsEquals(const StringView a, const StringView b);
|
||||
|
||||
// Returns whether a starts with b.
|
||||
bool CpuFeatures_StringView_StartsWith(const StringView a, const StringView b);
|
||||
|
||||
// Removes count characters from the beginning of view or kEmptyStringView if
|
||||
// count if greater than view.size.
|
||||
StringView CpuFeatures_StringView_PopFront(const StringView str_view,
|
||||
size_t count);
|
||||
|
||||
// Removes count characters from the end of view or kEmptyStringView if count if
|
||||
// greater than view.size.
|
||||
StringView CpuFeatures_StringView_PopBack(const StringView str_view,
|
||||
size_t count);
|
||||
|
||||
// Keeps the count first characters of view or view if count if greater than
|
||||
// view.size.
|
||||
StringView CpuFeatures_StringView_KeepFront(const StringView str_view,
|
||||
size_t count);
|
||||
|
||||
// Retrieves the first character of view. If view is empty the behavior is
|
||||
// undefined.
|
||||
char CpuFeatures_StringView_Front(const StringView view);
|
||||
|
||||
// Retrieves the last character of view. If view is empty the behavior is
|
||||
// undefined.
|
||||
char CpuFeatures_StringView_Back(const StringView view);
|
||||
|
||||
// Removes leading and tailing space characters.
|
||||
StringView CpuFeatures_StringView_TrimWhitespace(StringView view);
|
||||
|
||||
// Convert StringView to positive integer. e.g. "42", "0x2a".
|
||||
// Returns -1 on error.
|
||||
int CpuFeatures_StringView_ParsePositiveNumber(const StringView view);
|
||||
|
||||
// Copies src StringView to dst buffer.
|
||||
void CpuFeatures_StringView_CopyString(const StringView src, char* dst,
|
||||
size_t dst_size);
|
||||
|
||||
// Checks if line contains the specified whitespace separated word.
|
||||
bool CpuFeatures_StringView_HasWord(const StringView line,
|
||||
const char* const word);
|
||||
|
||||
// Get key/value from line. key and value are separated by ": ".
|
||||
// key and value are cleaned up from leading and trailing whitespaces.
|
||||
bool CpuFeatures_StringView_GetAttributeKeyValue(const StringView line,
|
||||
StringView* key,
|
||||
StringView* value);
|
||||
|
||||
CPU_FEATURES_END_CPP_NAMESPACE
|
||||
|
||||
#endif // CPU_FEATURES_INCLUDE_INTERNAL_STRING_VIEW_H_
|
||||
173
src/crypto/argon2_hasher/hash/cpu/cpu_features/scripts/run_integration.sh
Executable file
173
src/crypto/argon2_hasher/hash/cpu/cpu_features/scripts/run_integration.sh
Executable file
|
|
@ -0,0 +1,173 @@
|
|||
#!/bin/bash
|
||||
|
||||
readonly SCRIPT_FOLDER=$(cd -P -- "$(dirname -- "$0")" && pwd -P)
|
||||
readonly PROJECT_FOLDER="${SCRIPT_FOLDER}/.."
|
||||
readonly ARCHIVE_FOLDER=~/cpu_features_archives
|
||||
readonly QEMU_INSTALL=${ARCHIVE_FOLDER}/qemu
|
||||
readonly DEFAULT_CMAKE_ARGS=" -DCMAKE_BUILD_TYPE=Debug -DBUILD_TESTING=ON"
|
||||
|
||||
function extract() {
|
||||
case $1 in
|
||||
*.tar.bz2) tar xjf "$1" ;;
|
||||
*.tar.xz) tar xJf "$1" ;;
|
||||
*.tar.gz) tar xzf "$1" ;;
|
||||
*)
|
||||
echo "don't know how to extract '$1'..."
|
||||
exit 1
|
||||
esac
|
||||
}
|
||||
|
||||
function unpackifnotexists() {
|
||||
mkdir -p "${ARCHIVE_FOLDER}"
|
||||
cd "${ARCHIVE_FOLDER}" || exit
|
||||
local URL=$1
|
||||
local RELATIVE_FOLDER=$2
|
||||
local DESTINATION="${ARCHIVE_FOLDER}/${RELATIVE_FOLDER}"
|
||||
if [[ ! -d "${DESTINATION}" ]] ; then
|
||||
local ARCHIVE_NAME=$(echo ${URL} | sed 's/.*\///')
|
||||
test -f "${ARCHIVE_NAME}" || wget -q "${URL}"
|
||||
extract "${ARCHIVE_NAME}"
|
||||
fi
|
||||
}
|
||||
|
||||
function installqemuifneeded() {
|
||||
local VERSION=${QEMU_VERSION:=2.11.1}
|
||||
local ARCHES=${QEMU_ARCHES:=arm aarch64 i386 x86_64 mips mipsel}
|
||||
local TARGETS=${QEMU_TARGETS:=$(echo "$ARCHES" | sed 's#$# #;s#\([^ ]*\) #\1-linux-user #g')}
|
||||
|
||||
if echo "${VERSION} ${TARGETS}" | cmp --silent ${QEMU_INSTALL}/.build -; then
|
||||
echo "qemu ${VERSION} up to date!"
|
||||
return 0
|
||||
fi
|
||||
|
||||
echo "VERSION: ${VERSION}"
|
||||
echo "TARGETS: ${TARGETS}"
|
||||
|
||||
rm -rf ${QEMU_INSTALL}
|
||||
|
||||
# Checking for a tarball before downloading makes testing easier :-)
|
||||
local QEMU_URL="http://wiki.qemu-project.org/download/qemu-${VERSION}.tar.xz"
|
||||
local QEMU_FOLDER="qemu-${VERSION}"
|
||||
unpackifnotexists ${QEMU_URL} ${QEMU_FOLDER}
|
||||
cd ${QEMU_FOLDER} || exit
|
||||
|
||||
./configure \
|
||||
--prefix="${QEMU_INSTALL}" \
|
||||
--target-list="${TARGETS}" \
|
||||
--disable-docs \
|
||||
--disable-sdl \
|
||||
--disable-gtk \
|
||||
--disable-gnutls \
|
||||
--disable-gcrypt \
|
||||
--disable-nettle \
|
||||
--disable-curses \
|
||||
--static
|
||||
|
||||
make -j4
|
||||
make install
|
||||
|
||||
echo "$VERSION $TARGETS" > ${QEMU_INSTALL}/.build
|
||||
}
|
||||
|
||||
function assert_defined(){
|
||||
local VALUE=${1}
|
||||
: "${VALUE?"${1} needs to be defined"}"
|
||||
}
|
||||
|
||||
function integrate() {
|
||||
cd "${PROJECT_FOLDER}" || exit
|
||||
cmake -H. -B"${BUILD_DIR}" ${DEFAULT_CMAKE_ARGS} ${CMAKE_ADDITIONAL_ARGS}
|
||||
cmake --build "${BUILD_DIR}" --target all
|
||||
|
||||
if [[ -n "${QEMU_ARCH}" ]]; then
|
||||
if [[ "${QEMU_ARCH}" == "DISABLED" ]]; then
|
||||
QEMU="true || "
|
||||
else
|
||||
installqemuifneeded
|
||||
QEMU="${QEMU_INSTALL}/bin/qemu-${QEMU_ARCH} ${QEMU_ARGS}"
|
||||
fi
|
||||
else
|
||||
QEMU=""
|
||||
fi
|
||||
# Run tests
|
||||
for test_binary in ${BUILD_DIR}/test/*_test; do ${QEMU} ${test_binary}; done
|
||||
# Run demo program
|
||||
${QEMU} "${BUILD_DIR}/list_cpu_features"
|
||||
}
|
||||
|
||||
function expand_linaro_config() {
|
||||
assert_defined TARGET
|
||||
local LINARO_ROOT_URL=https://releases.linaro.org/components/toolchain/binaries/7.2-2017.11
|
||||
|
||||
local GCC_URL=${LINARO_ROOT_URL}/${TARGET}/gcc-linaro-7.2.1-2017.11-x86_64_${TARGET}.tar.xz
|
||||
local GCC_RELATIVE_FOLDER="gcc-linaro-7.2.1-2017.11-x86_64_${TARGET}"
|
||||
unpackifnotexists "${GCC_URL}" "${GCC_RELATIVE_FOLDER}"
|
||||
|
||||
local SYSROOT_URL=${LINARO_ROOT_URL}/${TARGET}/sysroot-glibc-linaro-2.25-2017.11-${TARGET}.tar.xz
|
||||
local SYSROOT_RELATIVE_FOLDER=sysroot-glibc-linaro-2.25-2017.11-${TARGET}
|
||||
unpackifnotexists "${SYSROOT_URL}" "${SYSROOT_RELATIVE_FOLDER}"
|
||||
|
||||
local SYSROOT_FOLDER=${ARCHIVE_FOLDER}/${SYSROOT_RELATIVE_FOLDER}
|
||||
local GCC_FOLDER=${ARCHIVE_FOLDER}/${GCC_RELATIVE_FOLDER}
|
||||
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_SYSROOT=${SYSROOT_FOLDER}"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_C_COMPILER=${GCC_FOLDER}/bin/${TARGET}-gcc"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_CXX_COMPILER=${GCC_FOLDER}/bin/${TARGET}-g++"
|
||||
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_FIND_ROOT_PATH_MODE_PROGRAM=NEVER"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_FIND_ROOT_PATH_MODE_INCLUDE=ONLY"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_FIND_ROOT_PATH_MODE_PACKAGE=ONLY"
|
||||
|
||||
QEMU_ARGS+=" -L ${SYSROOT_FOLDER}"
|
||||
QEMU_ARGS+=" -E LD_LIBRARY_PATH=/lib"
|
||||
}
|
||||
|
||||
function expand_codescape_config() {
|
||||
assert_defined TARGET
|
||||
local FLAVOUR=${QEMU_ARCH}-r2-hard
|
||||
local DATE=2016.05-03
|
||||
local CODESCAPE_URL=http://codescape-mips-sdk.imgtec.com/components/toolchain/${DATE}/Codescape.GNU.Tools.Package.${DATE}.for.MIPS.MTI.Linux.CentOS-5.x86_64.tar.gz
|
||||
local GCC_URL=${CODESCAPE_URL}
|
||||
local GCC_RELATIVE_FOLDER=${TARGET}/${DATE}
|
||||
unpackifnotexists "${GCC_URL}" "${GCC_RELATIVE_FOLDER}"
|
||||
|
||||
local SYSROOT_URL=${CODESCAPE_URL}
|
||||
local SYSROOT_FOLDER=${ARCHIVE_FOLDER}/${GCC_RELATIVE_FOLDER}/sysroot/${FLAVOUR}
|
||||
unpackifnotexists "${SYSROOT_URL}" "${SYSROOT_RELATIVE_FOLDER}"
|
||||
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DENABLE_MSA=1"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DMIPS_CPU=p5600"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_TOOLCHAIN_FILE=cmake/mips32-linux-gcc.cmake"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCROSS=${TARGET}-"
|
||||
CMAKE_ADDITIONAL_ARGS+=" -DCMAKE_FIND_ROOT_PATH=${ARCHIVE_FOLDER}/${GCC_RELATIVE_FOLDER}"
|
||||
|
||||
QEMU_ARGS+=" -L ${SYSROOT_FOLDER}"
|
||||
QEMU_ARGS+=" -E LD_LIBRARY_PATH=/lib"
|
||||
QEMU_ARGS+=" -cpu P5600"
|
||||
}
|
||||
|
||||
function expand_environment_and_integrate() {
|
||||
assert_defined PROJECT_FOLDER
|
||||
assert_defined TARGET
|
||||
|
||||
BUILD_DIR="${PROJECT_FOLDER}/cmake_build/${TARGET}"
|
||||
mkdir -p "${BUILD_DIR}"
|
||||
|
||||
CMAKE_ADDITIONAL_ARGS=""
|
||||
QEMU_ARGS=""
|
||||
|
||||
case ${TOOLCHAIN} in
|
||||
LINARO) expand_linaro_config ;;
|
||||
CODESCAPE) expand_codescape_config ;;
|
||||
NATIVE) QEMU_ARCH="" ;;
|
||||
*)
|
||||
echo "Unknown toolchain '${TOOLCHAIN}'..."
|
||||
exit 1
|
||||
esac
|
||||
integrate
|
||||
}
|
||||
|
||||
if [ "${CONTINUOUS_INTEGRATION}" = "true" ]; then
|
||||
QEMU_ARCHES=${QEMU_ARCH}
|
||||
expand_environment_and_integrate
|
||||
fi
|
||||
80
src/crypto/argon2_hasher/hash/cpu/cpu_features/scripts/test_integration.sh
Executable file
80
src/crypto/argon2_hasher/hash/cpu/cpu_features/scripts/test_integration.sh
Executable file
|
|
@ -0,0 +1,80 @@
|
|||
source "$(dirname -- "$0")"/run_integration.sh
|
||||
|
||||
# Toolchains for little-endian, 64-bit ARMv8 for GNU/Linux systems
|
||||
function set_aarch64-linux-gnu() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=aarch64-linux-gnu
|
||||
QEMU_ARCH=aarch64
|
||||
}
|
||||
|
||||
# Toolchains for little-endian, hard-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
function set_arm-linux-gnueabihf() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=arm-linux-gnueabihf
|
||||
QEMU_ARCH=arm
|
||||
}
|
||||
|
||||
# Toolchains for little-endian, 32-bit ARMv8 for GNU/Linux systems
|
||||
function set_armv8l-linux-gnueabihf() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=armv8l-linux-gnueabihf
|
||||
QEMU_ARCH=arm
|
||||
}
|
||||
|
||||
# Toolchains for little-endian, soft-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
function set_arm-linux-gnueabi() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=arm-linux-gnueabi
|
||||
QEMU_ARCH=arm
|
||||
}
|
||||
|
||||
# Toolchains for big-endian, 64-bit ARMv8 for GNU/Linux systems
|
||||
function set_aarch64_be-linux-gnu() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=aarch64_be-linux-gnu
|
||||
QEMU_ARCH="DISABLED"
|
||||
}
|
||||
|
||||
# Toolchains for big-endian, hard-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
function set_armeb-linux-gnueabihf() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=armeb-linux-gnueabihf
|
||||
QEMU_ARCH="DISABLED"
|
||||
}
|
||||
|
||||
# Toolchains for big-endian, soft-float, 32-bit ARMv7 (and earlier) for GNU/Linux systems
|
||||
function set_armeb-linux-gnueabi() {
|
||||
TOOLCHAIN=LINARO
|
||||
TARGET=armeb-linux-gnueabi
|
||||
QEMU_ARCH="DISABLED"
|
||||
}
|
||||
|
||||
|
||||
function set_mips() {
|
||||
TOOLCHAIN=CODESCAPE
|
||||
TARGET=mips-mti-linux-gnu
|
||||
QEMU_ARCH="DISABLED"
|
||||
}
|
||||
|
||||
function set_native() {
|
||||
TOOLCHAIN=NATIVE
|
||||
TARGET=native
|
||||
QEMU_ARCH=""
|
||||
}
|
||||
|
||||
ENVIRONMENTS="
|
||||
set_aarch64-linux-gnu
|
||||
set_arm-linux-gnueabihf
|
||||
set_armv8l-linux-gnueabihf
|
||||
set_arm-linux-gnueabi
|
||||
set_aarch64_be-linux-gnu
|
||||
set_armeb-linux-gnueabihf
|
||||
set_armeb-linux-gnueabi
|
||||
set_native
|
||||
set_mips
|
||||
"
|
||||
|
||||
for SET_ENVIRONMENT in ${ENVIRONMENTS}; do
|
||||
${SET_ENVIRONMENT}
|
||||
expand_environment_and_integrate
|
||||
done
|
||||
36
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuid_x86_clang_gcc.c
Executable file
36
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuid_x86_clang_gcc.c
Executable file
|
|
@ -0,0 +1,36 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/cpuid_x86.h"
|
||||
|
||||
#if defined(CPU_FEATURES_ARCH_X86)
|
||||
#if defined(CPU_FEATURES_COMPILER_CLANG) || defined(CPU_FEATURES_COMPILER_GCC)
|
||||
|
||||
#include <cpuid.h>
|
||||
|
||||
Leaf CpuId(uint32_t leaf_id) {
|
||||
Leaf leaf;
|
||||
__cpuid_count(leaf_id, 0, leaf.eax, leaf.ebx, leaf.ecx, leaf.edx);
|
||||
return leaf;
|
||||
}
|
||||
|
||||
uint32_t GetXCR0Eax(void) {
|
||||
uint32_t eax, edx;
|
||||
__asm("XGETBV" : "=a"(eax), "=d"(edx) : "c"(0));
|
||||
return eax;
|
||||
}
|
||||
|
||||
#endif // defined(CPU_FEATURES_COMPILER_CLANG) ||
|
||||
// defined(CPU_FEATURES_COMPILER_GCC)
|
||||
#endif // defined(CPU_FEATURES_ARCH_X86)
|
||||
34
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuid_x86_msvc.c
Executable file
34
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuid_x86_msvc.c
Executable file
|
|
@ -0,0 +1,34 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/cpuid_x86.h"
|
||||
|
||||
#if defined(CPU_FEATURES_ARCH_X86) && defined(CPU_FEATURES_COMPILER_MSC)
|
||||
#include <immintrin.h>
|
||||
#include <intrin.h> // For __cpuidex()
|
||||
|
||||
Leaf CpuId(uint32_t leaf_id) {
|
||||
Leaf leaf;
|
||||
int data[4];
|
||||
__cpuid(data, leaf_id);
|
||||
leaf.eax = data[0];
|
||||
leaf.ebx = data[1];
|
||||
leaf.ecx = data[2];
|
||||
leaf.edx = data[3];
|
||||
return leaf;
|
||||
}
|
||||
|
||||
uint32_t GetXCR0Eax(void) { return _xgetbv(0); }
|
||||
|
||||
#endif // defined(CPU_FEATURES_ARCH_X86) && defined(CPU_FEATURES_COMPILER_MSC)
|
||||
141
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_aarch64.c
Executable file
141
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_aarch64.c
Executable file
|
|
@ -0,0 +1,141 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_aarch64.h"
|
||||
|
||||
#include "internal/filesystem.h"
|
||||
#include "internal/hwcaps.h"
|
||||
#include "internal/linux_features_aggregator.h"
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#include <ctype.h>
|
||||
|
||||
DECLARE_SETTER(Aarch64Features, fp)
|
||||
DECLARE_SETTER(Aarch64Features, asimd)
|
||||
DECLARE_SETTER(Aarch64Features, aes)
|
||||
DECLARE_SETTER(Aarch64Features, pmull)
|
||||
DECLARE_SETTER(Aarch64Features, sha1)
|
||||
DECLARE_SETTER(Aarch64Features, sha2)
|
||||
DECLARE_SETTER(Aarch64Features, crc32)
|
||||
|
||||
static const CapabilityConfig kConfigs[] = {
|
||||
{{AARCH64_HWCAP_FP, 0}, "fp", &set_fp}, //
|
||||
{{AARCH64_HWCAP_ASIMD, 0}, "asimd", &set_asimd}, //
|
||||
{{AARCH64_HWCAP_AES, 0}, "aes", &set_aes}, //
|
||||
{{AARCH64_HWCAP_PMULL, 0}, "pmull", &set_pmull}, //
|
||||
{{AARCH64_HWCAP_SHA1, 0}, "sha1", &set_sha1}, //
|
||||
{{AARCH64_HWCAP_SHA2, 0}, "sha2", &set_sha2}, //
|
||||
{{AARCH64_HWCAP_CRC32, 0}, "crc32", &set_crc32}, //
|
||||
};
|
||||
|
||||
static const size_t kConfigsSize = sizeof(kConfigs) / sizeof(CapabilityConfig);
|
||||
|
||||
static bool HandleAarch64Line(const LineResult result,
|
||||
Aarch64Info* const info) {
|
||||
StringView line = result.line;
|
||||
StringView key, value;
|
||||
if (CpuFeatures_StringView_GetAttributeKeyValue(line, &key, &value)) {
|
||||
if (CpuFeatures_StringView_IsEquals(key, str("Features"))) {
|
||||
CpuFeatures_SetFromFlags(kConfigsSize, kConfigs, value, &info->features);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU implementer"))) {
|
||||
info->implementer = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU variant"))) {
|
||||
info->variant = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU part"))) {
|
||||
info->part = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU revision"))) {
|
||||
info->revision = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
}
|
||||
}
|
||||
return !result.eof;
|
||||
}
|
||||
|
||||
static void FillProcCpuInfoData(Aarch64Info* const info) {
|
||||
const int fd = CpuFeatures_OpenFile("/proc/cpuinfo");
|
||||
if (fd >= 0) {
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, fd);
|
||||
for (;;) {
|
||||
if (!HandleAarch64Line(StackLineReader_NextLine(&reader), info)) {
|
||||
break;
|
||||
}
|
||||
}
|
||||
CpuFeatures_CloseFile(fd);
|
||||
}
|
||||
}
|
||||
|
||||
static const Aarch64Info kEmptyAarch64Info;
|
||||
|
||||
Aarch64Info GetAarch64Info(void) {
|
||||
// capabilities are fetched from both getauxval and /proc/cpuinfo so we can
|
||||
// have some information if the executable is sandboxed (aka no access to
|
||||
// /proc/cpuinfo).
|
||||
Aarch64Info info = kEmptyAarch64Info;
|
||||
|
||||
FillProcCpuInfoData(&info);
|
||||
CpuFeatures_OverrideFromHwCaps(kConfigsSize, kConfigs,
|
||||
CpuFeatures_GetHardwareCapabilities(),
|
||||
&info.features);
|
||||
|
||||
return info;
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
int GetAarch64FeaturesEnumValue(const Aarch64Features* features,
|
||||
Aarch64FeaturesEnum value) {
|
||||
switch (value) {
|
||||
case AARCH64_FP:
|
||||
return features->fp;
|
||||
case AARCH64_ASIMD:
|
||||
return features->asimd;
|
||||
case AARCH64_AES:
|
||||
return features->aes;
|
||||
case AARCH64_PMULL:
|
||||
return features->pmull;
|
||||
case AARCH64_SHA1:
|
||||
return features->sha1;
|
||||
case AARCH64_SHA2:
|
||||
return features->sha2;
|
||||
case AARCH64_CRC32:
|
||||
return features->crc32;
|
||||
case AARCH64_LAST_:
|
||||
break;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
const char* GetAarch64FeaturesEnumName(Aarch64FeaturesEnum value) {
|
||||
switch (value) {
|
||||
case AARCH64_FP:
|
||||
return "fp";
|
||||
case AARCH64_ASIMD:
|
||||
return "asimd";
|
||||
case AARCH64_AES:
|
||||
return "aes";
|
||||
case AARCH64_PMULL:
|
||||
return "pmull";
|
||||
case AARCH64_SHA1:
|
||||
return "sha1";
|
||||
case AARCH64_SHA2:
|
||||
return "sha2";
|
||||
case AARCH64_CRC32:
|
||||
return "crc32";
|
||||
case AARCH64_LAST_:
|
||||
break;
|
||||
}
|
||||
return "unknown feature";
|
||||
}
|
||||
259
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_arm.c
Executable file
259
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_arm.c
Executable file
|
|
@ -0,0 +1,259 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_arm.h"
|
||||
|
||||
#include "internal/bit_utils.h"
|
||||
#include "internal/filesystem.h"
|
||||
#include "internal/hwcaps.h"
|
||||
#include "internal/linux_features_aggregator.h"
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#include <ctype.h>
|
||||
|
||||
DECLARE_SETTER(ArmFeatures, vfp)
|
||||
DECLARE_SETTER(ArmFeatures, iwmmxt)
|
||||
DECLARE_SETTER(ArmFeatures, neon)
|
||||
DECLARE_SETTER(ArmFeatures, vfpv3)
|
||||
DECLARE_SETTER(ArmFeatures, vfpv3d16)
|
||||
DECLARE_SETTER(ArmFeatures, vfpv4)
|
||||
DECLARE_SETTER(ArmFeatures, idiva)
|
||||
DECLARE_SETTER(ArmFeatures, idivt)
|
||||
DECLARE_SETTER(ArmFeatures, aes)
|
||||
DECLARE_SETTER(ArmFeatures, pmull)
|
||||
DECLARE_SETTER(ArmFeatures, sha1)
|
||||
DECLARE_SETTER(ArmFeatures, sha2)
|
||||
DECLARE_SETTER(ArmFeatures, crc32)
|
||||
|
||||
static const CapabilityConfig kConfigs[] = {
|
||||
{{ARM_HWCAP_VFP, 0}, "vfp", &set_vfp}, //
|
||||
{{ARM_HWCAP_IWMMXT, 0}, "iwmmxt", &set_iwmmxt}, //
|
||||
{{ARM_HWCAP_NEON, 0}, "neon", &set_neon}, //
|
||||
{{ARM_HWCAP_VFPV3, 0}, "vfpv3", &set_vfpv3}, //
|
||||
{{ARM_HWCAP_VFPV3D16, 0}, "vfpv3d16", &set_vfpv3d16}, //
|
||||
{{ARM_HWCAP_VFPV4, 0}, "vfpv4", &set_vfpv4}, //
|
||||
{{ARM_HWCAP_IDIVA, 0}, "idiva", &set_idiva}, //
|
||||
{{ARM_HWCAP_IDIVT, 0}, "idivt", &set_idivt}, //
|
||||
{{0, ARM_HWCAP2_AES}, "aes", &set_aes}, //
|
||||
{{0, ARM_HWCAP2_PMULL}, "pmull", &set_pmull}, //
|
||||
{{0, ARM_HWCAP2_SHA1}, "sha1", &set_sha1}, //
|
||||
{{0, ARM_HWCAP2_SHA2}, "sha2", &set_sha2}, //
|
||||
{{0, ARM_HWCAP2_CRC32}, "crc32", &set_crc32}, //
|
||||
};
|
||||
|
||||
static const size_t kConfigsSize = sizeof(kConfigs) / sizeof(CapabilityConfig);
|
||||
|
||||
typedef struct {
|
||||
bool processor_reports_armv6;
|
||||
bool hardware_reports_goldfish;
|
||||
} ProcCpuInfoData;
|
||||
|
||||
static int IndexOfNonDigit(StringView str) {
|
||||
size_t index = 0;
|
||||
while (str.size && isdigit(CpuFeatures_StringView_Front(str))) {
|
||||
str = CpuFeatures_StringView_PopFront(str, 1);
|
||||
++index;
|
||||
}
|
||||
return index;
|
||||
}
|
||||
|
||||
static bool HandleArmLine(const LineResult result, ArmInfo* const info,
|
||||
ProcCpuInfoData* const proc_info) {
|
||||
StringView line = result.line;
|
||||
StringView key, value;
|
||||
if (CpuFeatures_StringView_GetAttributeKeyValue(line, &key, &value)) {
|
||||
if (CpuFeatures_StringView_IsEquals(key, str("Features"))) {
|
||||
CpuFeatures_SetFromFlags(kConfigsSize, kConfigs, value, &info->features);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU implementer"))) {
|
||||
info->implementer = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU variant"))) {
|
||||
info->variant = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU part"))) {
|
||||
info->part = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU revision"))) {
|
||||
info->revision = CpuFeatures_StringView_ParsePositiveNumber(value);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("CPU architecture"))) {
|
||||
// CPU architecture is a number that may be followed by letters. e.g.
|
||||
// "6TEJ", "7".
|
||||
const StringView digits =
|
||||
CpuFeatures_StringView_KeepFront(value, IndexOfNonDigit(value));
|
||||
info->architecture = CpuFeatures_StringView_ParsePositiveNumber(digits);
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("Processor"))) {
|
||||
proc_info->processor_reports_armv6 =
|
||||
CpuFeatures_StringView_IndexOf(value, str("(v6l)")) >= 0;
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("Hardware"))) {
|
||||
proc_info->hardware_reports_goldfish =
|
||||
CpuFeatures_StringView_IsEquals(value, str("Goldfish"));
|
||||
}
|
||||
}
|
||||
return !result.eof;
|
||||
}
|
||||
|
||||
static uint32_t GetCpuId(const ArmInfo* const info) {
|
||||
return (ExtractBitRange(info->implementer, 7, 0) << 24) |
|
||||
(ExtractBitRange(info->variant, 3, 0) << 20) |
|
||||
(ExtractBitRange(info->part, 11, 0) << 4) |
|
||||
(ExtractBitRange(info->revision, 3, 0) << 0);
|
||||
}
|
||||
|
||||
static void FixErrors(ArmInfo* const info,
|
||||
ProcCpuInfoData* const proc_cpu_info_data) {
|
||||
// Fixing Samsung kernel reporting invalid cpu architecture.
|
||||
// http://code.google.com/p/android/issues/detail?id=10812
|
||||
if (proc_cpu_info_data->processor_reports_armv6 && info->architecture >= 7) {
|
||||
info->architecture = 6;
|
||||
}
|
||||
|
||||
// Handle kernel configuration bugs that prevent the correct reporting of CPU
|
||||
// features.
|
||||
switch (GetCpuId(info)) {
|
||||
case 0x4100C080:
|
||||
// Special case: The emulator-specific Android 4.2 kernel fails to report
|
||||
// support for the 32-bit ARM IDIV instruction. Technically, this is a
|
||||
// feature of the virtual CPU implemented by the emulator. Note that it
|
||||
// could also support Thumb IDIV in the future, and this will have to be
|
||||
// slightly updated.
|
||||
if (info->architecture >= 7 &&
|
||||
proc_cpu_info_data->hardware_reports_goldfish) {
|
||||
info->features.idiva = true;
|
||||
}
|
||||
break;
|
||||
case 0x511004D0:
|
||||
// https://crbug.com/341598.
|
||||
info->features.neon = false;
|
||||
break;
|
||||
case 0x510006F2:
|
||||
case 0x510006F3:
|
||||
// The Nexus 4 (Qualcomm Krait) kernel configuration forgets to report
|
||||
// IDIV support.
|
||||
info->features.idiva = true;
|
||||
info->features.idivt = true;
|
||||
break;
|
||||
}
|
||||
|
||||
// Propagate cpu features.
|
||||
if (info->features.vfpv4) info->features.vfpv3 = true;
|
||||
if (info->features.neon) info->features.vfpv3 = true;
|
||||
if (info->features.vfpv3) info->features.vfp = true;
|
||||
}
|
||||
|
||||
static void FillProcCpuInfoData(ArmInfo* const info,
|
||||
ProcCpuInfoData* proc_cpu_info_data) {
|
||||
const int fd = CpuFeatures_OpenFile("/proc/cpuinfo");
|
||||
if (fd >= 0) {
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, fd);
|
||||
for (;;) {
|
||||
if (!HandleArmLine(StackLineReader_NextLine(&reader), info,
|
||||
proc_cpu_info_data)) {
|
||||
break;
|
||||
}
|
||||
}
|
||||
CpuFeatures_CloseFile(fd);
|
||||
}
|
||||
}
|
||||
|
||||
static const ArmInfo kEmptyArmInfo;
|
||||
|
||||
static const ProcCpuInfoData kEmptyProcCpuInfoData;
|
||||
|
||||
ArmInfo GetArmInfo(void) {
|
||||
// capabilities are fetched from both getauxval and /proc/cpuinfo so we can
|
||||
// have some information if the executable is sandboxed (aka no access to
|
||||
// /proc/cpuinfo).
|
||||
ArmInfo info = kEmptyArmInfo;
|
||||
ProcCpuInfoData proc_cpu_info_data = kEmptyProcCpuInfoData;
|
||||
|
||||
FillProcCpuInfoData(&info, &proc_cpu_info_data);
|
||||
CpuFeatures_OverrideFromHwCaps(kConfigsSize, kConfigs,
|
||||
CpuFeatures_GetHardwareCapabilities(),
|
||||
&info.features);
|
||||
|
||||
FixErrors(&info, &proc_cpu_info_data);
|
||||
|
||||
return info;
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
int GetArmFeaturesEnumValue(const ArmFeatures* features,
|
||||
ArmFeaturesEnum value) {
|
||||
switch (value) {
|
||||
case ARM_VFP:
|
||||
return features->vfp;
|
||||
case ARM_IWMMXT:
|
||||
return features->iwmmxt;
|
||||
case ARM_NEON:
|
||||
return features->neon;
|
||||
case ARM_VFPV3:
|
||||
return features->vfpv3;
|
||||
case ARM_VFPV3D16:
|
||||
return features->vfpv3d16;
|
||||
case ARM_VFPV4:
|
||||
return features->vfpv4;
|
||||
case ARM_IDIVA:
|
||||
return features->idiva;
|
||||
case ARM_IDIVT:
|
||||
return features->idivt;
|
||||
case ARM_AES:
|
||||
return features->aes;
|
||||
case ARM_PMULL:
|
||||
return features->pmull;
|
||||
case ARM_SHA1:
|
||||
return features->sha1;
|
||||
case ARM_SHA2:
|
||||
return features->sha2;
|
||||
case ARM_CRC32:
|
||||
return features->crc32;
|
||||
case ARM_LAST_:
|
||||
break;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
const char* GetArmFeaturesEnumName(ArmFeaturesEnum value) {
|
||||
switch (value) {
|
||||
case ARM_VFP:
|
||||
return "vfp";
|
||||
case ARM_IWMMXT:
|
||||
return "iwmmxt";
|
||||
case ARM_NEON:
|
||||
return "neon";
|
||||
case ARM_VFPV3:
|
||||
return "vfpv3";
|
||||
case ARM_VFPV3D16:
|
||||
return "vfpv3d16";
|
||||
case ARM_VFPV4:
|
||||
return "vfpv4";
|
||||
case ARM_IDIVA:
|
||||
return "idiva";
|
||||
case ARM_IDIVT:
|
||||
return "idivt";
|
||||
case ARM_AES:
|
||||
return "aes";
|
||||
case ARM_PMULL:
|
||||
return "pmull";
|
||||
case ARM_SHA1:
|
||||
return "sha1";
|
||||
case ARM_SHA2:
|
||||
return "sha2";
|
||||
case ARM_CRC32:
|
||||
return "crc32";
|
||||
case ARM_LAST_:
|
||||
break;
|
||||
}
|
||||
return "unknown feature";
|
||||
}
|
||||
98
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_mips.c
Executable file
98
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_mips.c
Executable file
|
|
@ -0,0 +1,98 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_mips.h"
|
||||
|
||||
#include "internal/filesystem.h"
|
||||
#include "internal/linux_features_aggregator.h"
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
DECLARE_SETTER(MipsFeatures, msa)
|
||||
DECLARE_SETTER(MipsFeatures, eva)
|
||||
|
||||
static const CapabilityConfig kConfigs[] = {
|
||||
{{MIPS_HWCAP_MSA, 0}, "msa", &set_msa}, //
|
||||
{{MIPS_HWCAP_EVA, 0}, "eva", &set_eva}, //
|
||||
};
|
||||
static const size_t kConfigsSize = sizeof(kConfigs) / sizeof(CapabilityConfig);
|
||||
|
||||
static bool HandleMipsLine(const LineResult result,
|
||||
MipsFeatures* const features) {
|
||||
StringView key, value;
|
||||
// See tests for an example.
|
||||
if (CpuFeatures_StringView_GetAttributeKeyValue(result.line, &key, &value)) {
|
||||
if (CpuFeatures_StringView_IsEquals(key, str("ASEs implemented"))) {
|
||||
CpuFeatures_SetFromFlags(kConfigsSize, kConfigs, value, features);
|
||||
}
|
||||
}
|
||||
return !result.eof;
|
||||
}
|
||||
|
||||
static void FillProcCpuInfoData(MipsFeatures* const features) {
|
||||
const int fd = CpuFeatures_OpenFile("/proc/cpuinfo");
|
||||
if (fd >= 0) {
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, fd);
|
||||
for (;;) {
|
||||
if (!HandleMipsLine(StackLineReader_NextLine(&reader), features)) {
|
||||
break;
|
||||
}
|
||||
}
|
||||
CpuFeatures_CloseFile(fd);
|
||||
}
|
||||
}
|
||||
|
||||
static const MipsInfo kEmptyMipsInfo;
|
||||
|
||||
MipsInfo GetMipsInfo(void) {
|
||||
// capabilities are fetched from both getauxval and /proc/cpuinfo so we can
|
||||
// have some information if the executable is sandboxed (aka no access to
|
||||
// /proc/cpuinfo).
|
||||
MipsInfo info = kEmptyMipsInfo;
|
||||
|
||||
FillProcCpuInfoData(&info.features);
|
||||
CpuFeatures_OverrideFromHwCaps(kConfigsSize, kConfigs,
|
||||
CpuFeatures_GetHardwareCapabilities(),
|
||||
&info.features);
|
||||
return info;
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
int GetMipsFeaturesEnumValue(const MipsFeatures* features,
|
||||
MipsFeaturesEnum value) {
|
||||
switch (value) {
|
||||
case MIPS_MSA:
|
||||
return features->msa;
|
||||
case MIPS_EVA:
|
||||
return features->eva;
|
||||
case MIPS_LAST_:
|
||||
break;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
const char* GetMipsFeaturesEnumName(MipsFeaturesEnum value) {
|
||||
switch (value) {
|
||||
case MIPS_MSA:
|
||||
return "msa";
|
||||
case MIPS_EVA:
|
||||
return "eva";
|
||||
case MIPS_LAST_:
|
||||
break;
|
||||
}
|
||||
return "unknown feature";
|
||||
}
|
||||
358
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_ppc.c
Executable file
358
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_ppc.c
Executable file
|
|
@ -0,0 +1,358 @@
|
|||
// Copyright 2018 IBM.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include <stdbool.h>
|
||||
#include <string.h>
|
||||
|
||||
#include "cpuinfo_ppc.h"
|
||||
#include "internal/bit_utils.h"
|
||||
#include "internal/filesystem.h"
|
||||
#include "internal/linux_features_aggregator.h"
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
DECLARE_SETTER(PPCFeatures, ppc32)
|
||||
DECLARE_SETTER(PPCFeatures, ppc64)
|
||||
DECLARE_SETTER(PPCFeatures, ppc601)
|
||||
DECLARE_SETTER(PPCFeatures, altivec)
|
||||
DECLARE_SETTER(PPCFeatures, fpu)
|
||||
DECLARE_SETTER(PPCFeatures, mmu)
|
||||
DECLARE_SETTER(PPCFeatures, mac_4xx)
|
||||
DECLARE_SETTER(PPCFeatures, unifiedcache)
|
||||
DECLARE_SETTER(PPCFeatures, spe)
|
||||
DECLARE_SETTER(PPCFeatures, efpsingle)
|
||||
DECLARE_SETTER(PPCFeatures, efpdouble)
|
||||
DECLARE_SETTER(PPCFeatures, no_tb)
|
||||
DECLARE_SETTER(PPCFeatures, power4)
|
||||
DECLARE_SETTER(PPCFeatures, power5)
|
||||
DECLARE_SETTER(PPCFeatures, power5plus)
|
||||
DECLARE_SETTER(PPCFeatures, cell)
|
||||
DECLARE_SETTER(PPCFeatures, booke)
|
||||
DECLARE_SETTER(PPCFeatures, smt)
|
||||
DECLARE_SETTER(PPCFeatures, icachesnoop)
|
||||
DECLARE_SETTER(PPCFeatures, arch205)
|
||||
DECLARE_SETTER(PPCFeatures, pa6t)
|
||||
DECLARE_SETTER(PPCFeatures, dfp)
|
||||
DECLARE_SETTER(PPCFeatures, power6ext)
|
||||
DECLARE_SETTER(PPCFeatures, arch206)
|
||||
DECLARE_SETTER(PPCFeatures, vsx)
|
||||
DECLARE_SETTER(PPCFeatures, pseries_perfmon_compat)
|
||||
DECLARE_SETTER(PPCFeatures, truele)
|
||||
DECLARE_SETTER(PPCFeatures, ppcle)
|
||||
DECLARE_SETTER(PPCFeatures, arch207)
|
||||
DECLARE_SETTER(PPCFeatures, htm)
|
||||
DECLARE_SETTER(PPCFeatures, dscr)
|
||||
DECLARE_SETTER(PPCFeatures, ebb)
|
||||
DECLARE_SETTER(PPCFeatures, isel)
|
||||
DECLARE_SETTER(PPCFeatures, tar)
|
||||
DECLARE_SETTER(PPCFeatures, vcrypto)
|
||||
DECLARE_SETTER(PPCFeatures, htm_nosc)
|
||||
DECLARE_SETTER(PPCFeatures, arch300)
|
||||
DECLARE_SETTER(PPCFeatures, ieee128)
|
||||
DECLARE_SETTER(PPCFeatures, darn)
|
||||
DECLARE_SETTER(PPCFeatures, scv)
|
||||
DECLARE_SETTER(PPCFeatures, htm_no_suspend)
|
||||
|
||||
static const CapabilityConfig kConfigs[] = {
|
||||
{{PPC_FEATURE_32, 0}, "ppc32", &set_ppc32},
|
||||
{{PPC_FEATURE_64, 0}, "ppc64", &set_ppc64},
|
||||
{{PPC_FEATURE_601_INSTR, 0}, "ppc601", &set_ppc601},
|
||||
{{PPC_FEATURE_HAS_ALTIVEC, 0}, "altivec", &set_altivec},
|
||||
{{PPC_FEATURE_HAS_FPU, 0}, "fpu", &set_fpu},
|
||||
{{PPC_FEATURE_HAS_MMU, 0}, "mmu", &set_mmu},
|
||||
{{PPC_FEATURE_HAS_4xxMAC, 0}, "4xxmac", &set_mac_4xx},
|
||||
{{PPC_FEATURE_UNIFIED_CACHE, 0}, "ucache", &set_unifiedcache},
|
||||
{{PPC_FEATURE_HAS_SPE, 0}, "spe", &set_spe},
|
||||
{{PPC_FEATURE_HAS_EFP_SINGLE, 0}, "efpsingle", &set_efpsingle},
|
||||
{{PPC_FEATURE_HAS_EFP_DOUBLE, 0}, "efpdouble", &set_efpdouble},
|
||||
{{PPC_FEATURE_NO_TB, 0}, "notb", &set_no_tb},
|
||||
{{PPC_FEATURE_POWER4, 0}, "power4", &set_power4},
|
||||
{{PPC_FEATURE_POWER5, 0}, "power5", &set_power5},
|
||||
{{PPC_FEATURE_POWER5_PLUS, 0}, "power5+", &set_power5plus},
|
||||
{{PPC_FEATURE_CELL, 0}, "cellbe", &set_cell},
|
||||
{{PPC_FEATURE_BOOKE, 0}, "booke", &set_booke},
|
||||
{{PPC_FEATURE_SMT, 0}, "smt", &set_smt},
|
||||
{{PPC_FEATURE_ICACHE_SNOOP, 0}, "ic_snoop", &set_icachesnoop},
|
||||
{{PPC_FEATURE_ARCH_2_05, 0}, "arch_2_05", &set_arch205},
|
||||
{{PPC_FEATURE_PA6T, 0}, "pa6t", &set_pa6t},
|
||||
{{PPC_FEATURE_HAS_DFP, 0}, "dfp", &set_dfp},
|
||||
{{PPC_FEATURE_POWER6_EXT, 0}, "power6x", &set_power6ext},
|
||||
{{PPC_FEATURE_ARCH_2_06, 0}, "arch_2_06", &set_arch206},
|
||||
{{PPC_FEATURE_HAS_VSX, 0}, "vsx", &set_vsx},
|
||||
{{PPC_FEATURE_PSERIES_PERFMON_COMPAT, 0},
|
||||
"archpmu",
|
||||
&set_pseries_perfmon_compat},
|
||||
{{PPC_FEATURE_TRUE_LE, 0}, "true_le", &set_truele},
|
||||
{{PPC_FEATURE_PPC_LE, 0}, "ppcle", &set_ppcle},
|
||||
{{0, PPC_FEATURE2_ARCH_2_07}, "arch_2_07", &set_arch207},
|
||||
{{0, PPC_FEATURE2_HTM}, "htm", &set_htm},
|
||||
{{0, PPC_FEATURE2_DSCR}, "dscr", &set_dscr},
|
||||
{{0, PPC_FEATURE2_EBB}, "ebb", &set_ebb},
|
||||
{{0, PPC_FEATURE2_ISEL}, "isel", &set_isel},
|
||||
{{0, PPC_FEATURE2_TAR}, "tar", &set_tar},
|
||||
{{0, PPC_FEATURE2_VEC_CRYPTO}, "vcrypto", &set_vcrypto},
|
||||
{{0, PPC_FEATURE2_HTM_NOSC}, "htm-nosc", &set_htm_nosc},
|
||||
{{0, PPC_FEATURE2_ARCH_3_00}, "arch_3_00", &set_arch300},
|
||||
{{0, PPC_FEATURE2_HAS_IEEE128}, "ieee128", &set_ieee128},
|
||||
{{0, PPC_FEATURE2_DARN}, "darn", &set_darn},
|
||||
{{0, PPC_FEATURE2_SCV}, "scv", &set_scv},
|
||||
{{0, PPC_FEATURE2_HTM_NO_SUSPEND}, "htm-no-suspend", &set_htm_no_suspend},
|
||||
};
|
||||
static const size_t kConfigsSize = sizeof(kConfigs) / sizeof(CapabilityConfig);
|
||||
|
||||
static bool HandlePPCLine(const LineResult result,
|
||||
PPCPlatformStrings* const strings) {
|
||||
StringView line = result.line;
|
||||
StringView key, value;
|
||||
if (CpuFeatures_StringView_GetAttributeKeyValue(line, &key, &value)) {
|
||||
if (CpuFeatures_StringView_HasWord(key, "platform")) {
|
||||
CpuFeatures_StringView_CopyString(value, strings->platform,
|
||||
sizeof(strings->platform));
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("model"))) {
|
||||
CpuFeatures_StringView_CopyString(value, strings->model,
|
||||
sizeof(strings->platform));
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("machine"))) {
|
||||
CpuFeatures_StringView_CopyString(value, strings->machine,
|
||||
sizeof(strings->platform));
|
||||
} else if (CpuFeatures_StringView_IsEquals(key, str("cpu"))) {
|
||||
CpuFeatures_StringView_CopyString(value, strings->cpu,
|
||||
sizeof(strings->platform));
|
||||
}
|
||||
}
|
||||
return !result.eof;
|
||||
}
|
||||
|
||||
static void FillProcCpuInfoData(PPCPlatformStrings* const strings) {
|
||||
const int fd = CpuFeatures_OpenFile("/proc/cpuinfo");
|
||||
if (fd >= 0) {
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, fd);
|
||||
for (;;) {
|
||||
if (!HandlePPCLine(StackLineReader_NextLine(&reader), strings)) {
|
||||
break;
|
||||
}
|
||||
}
|
||||
CpuFeatures_CloseFile(fd);
|
||||
}
|
||||
}
|
||||
|
||||
static const PPCInfo kEmptyPPCInfo;
|
||||
|
||||
PPCInfo GetPPCInfo(void) {
|
||||
/*
|
||||
* On Power feature flags aren't currently in cpuinfo so we only look at
|
||||
* the auxilary vector.
|
||||
*/
|
||||
PPCInfo info = kEmptyPPCInfo;
|
||||
|
||||
CpuFeatures_OverrideFromHwCaps(kConfigsSize, kConfigs,
|
||||
CpuFeatures_GetHardwareCapabilities(),
|
||||
&info.features);
|
||||
return info;
|
||||
}
|
||||
|
||||
static const PPCPlatformStrings kEmptyPPCPlatformStrings;
|
||||
|
||||
PPCPlatformStrings GetPPCPlatformStrings(void) {
|
||||
PPCPlatformStrings strings = kEmptyPPCPlatformStrings;
|
||||
|
||||
FillProcCpuInfoData(&strings);
|
||||
strings.type = CpuFeatures_GetPlatformType();
|
||||
return strings;
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
int GetPPCFeaturesEnumValue(const PPCFeatures* features,
|
||||
PPCFeaturesEnum value) {
|
||||
switch (value) {
|
||||
case PPC_32:
|
||||
return features->ppc32;
|
||||
case PPC_64:
|
||||
return features->ppc64;
|
||||
case PPC_601_INSTR:
|
||||
return features->ppc601;
|
||||
case PPC_HAS_ALTIVEC:
|
||||
return features->altivec;
|
||||
case PPC_HAS_FPU:
|
||||
return features->fpu;
|
||||
case PPC_HAS_MMU:
|
||||
return features->mmu;
|
||||
case PPC_HAS_4xxMAC:
|
||||
return features->mac_4xx;
|
||||
case PPC_UNIFIED_CACHE:
|
||||
return features->unifiedcache;
|
||||
case PPC_HAS_SPE:
|
||||
return features->spe;
|
||||
case PPC_HAS_EFP_SINGLE:
|
||||
return features->efpsingle;
|
||||
case PPC_HAS_EFP_DOUBLE:
|
||||
return features->efpdouble;
|
||||
case PPC_NO_TB:
|
||||
return features->no_tb;
|
||||
case PPC_POWER4:
|
||||
return features->power4;
|
||||
case PPC_POWER5:
|
||||
return features->power5;
|
||||
case PPC_POWER5_PLUS:
|
||||
return features->power5plus;
|
||||
case PPC_CELL:
|
||||
return features->cell;
|
||||
case PPC_BOOKE:
|
||||
return features->booke;
|
||||
case PPC_SMT:
|
||||
return features->smt;
|
||||
case PPC_ICACHE_SNOOP:
|
||||
return features->icachesnoop;
|
||||
case PPC_ARCH_2_05:
|
||||
return features->arch205;
|
||||
case PPC_PA6T:
|
||||
return features->pa6t;
|
||||
case PPC_HAS_DFP:
|
||||
return features->dfp;
|
||||
case PPC_POWER6_EXT:
|
||||
return features->power6ext;
|
||||
case PPC_ARCH_2_06:
|
||||
return features->arch206;
|
||||
case PPC_HAS_VSX:
|
||||
return features->vsx;
|
||||
case PPC_PSERIES_PERFMON_COMPAT:
|
||||
return features->pseries_perfmon_compat;
|
||||
case PPC_TRUE_LE:
|
||||
return features->truele;
|
||||
case PPC_PPC_LE:
|
||||
return features->ppcle;
|
||||
case PPC_ARCH_2_07:
|
||||
return features->arch207;
|
||||
case PPC_HTM:
|
||||
return features->htm;
|
||||
case PPC_DSCR:
|
||||
return features->dscr;
|
||||
case PPC_EBB:
|
||||
return features->ebb;
|
||||
case PPC_ISEL:
|
||||
return features->isel;
|
||||
case PPC_TAR:
|
||||
return features->tar;
|
||||
case PPC_VEC_CRYPTO:
|
||||
return features->vcrypto;
|
||||
case PPC_HTM_NOSC:
|
||||
return features->htm_nosc;
|
||||
case PPC_ARCH_3_00:
|
||||
return features->arch300;
|
||||
case PPC_HAS_IEEE128:
|
||||
return features->ieee128;
|
||||
case PPC_DARN:
|
||||
return features->darn;
|
||||
case PPC_SCV:
|
||||
return features->scv;
|
||||
case PPC_HTM_NO_SUSPEND:
|
||||
return features->htm_no_suspend;
|
||||
case PPC_LAST_:
|
||||
break;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
/* Have used the same names as glibc */
|
||||
const char* GetPPCFeaturesEnumName(PPCFeaturesEnum value) {
|
||||
switch (value) {
|
||||
case PPC_32:
|
||||
return "ppc32";
|
||||
case PPC_64:
|
||||
return "ppc64";
|
||||
case PPC_601_INSTR:
|
||||
return "ppc601";
|
||||
case PPC_HAS_ALTIVEC:
|
||||
return "altivec";
|
||||
case PPC_HAS_FPU:
|
||||
return "fpu";
|
||||
case PPC_HAS_MMU:
|
||||
return "mmu";
|
||||
case PPC_HAS_4xxMAC:
|
||||
return "4xxmac";
|
||||
case PPC_UNIFIED_CACHE:
|
||||
return "ucache";
|
||||
case PPC_HAS_SPE:
|
||||
return "spe";
|
||||
case PPC_HAS_EFP_SINGLE:
|
||||
return "efpsingle";
|
||||
case PPC_HAS_EFP_DOUBLE:
|
||||
return "efpdouble";
|
||||
case PPC_NO_TB:
|
||||
return "notb";
|
||||
case PPC_POWER4:
|
||||
return "power4";
|
||||
case PPC_POWER5:
|
||||
return "power5";
|
||||
case PPC_POWER5_PLUS:
|
||||
return "power5+";
|
||||
case PPC_CELL:
|
||||
return "cellbe";
|
||||
case PPC_BOOKE:
|
||||
return "booke";
|
||||
case PPC_SMT:
|
||||
return "smt";
|
||||
case PPC_ICACHE_SNOOP:
|
||||
return "ic_snoop";
|
||||
case PPC_ARCH_2_05:
|
||||
return "arch_2_05";
|
||||
case PPC_PA6T:
|
||||
return "pa6t";
|
||||
case PPC_HAS_DFP:
|
||||
return "dfp";
|
||||
case PPC_POWER6_EXT:
|
||||
return "power6x";
|
||||
case PPC_ARCH_2_06:
|
||||
return "arch_2_06";
|
||||
case PPC_HAS_VSX:
|
||||
return "vsx";
|
||||
case PPC_PSERIES_PERFMON_COMPAT:
|
||||
return "archpmu";
|
||||
case PPC_TRUE_LE:
|
||||
return "true_le";
|
||||
case PPC_PPC_LE:
|
||||
return "ppcle";
|
||||
case PPC_ARCH_2_07:
|
||||
return "arch_2_07";
|
||||
case PPC_HTM:
|
||||
return "htm";
|
||||
case PPC_DSCR:
|
||||
return "dscr";
|
||||
case PPC_EBB:
|
||||
return "ebb";
|
||||
case PPC_ISEL:
|
||||
return "isel";
|
||||
case PPC_TAR:
|
||||
return "tar";
|
||||
case PPC_VEC_CRYPTO:
|
||||
return "vcrypto";
|
||||
case PPC_HTM_NOSC:
|
||||
return "htm-nosc";
|
||||
case PPC_ARCH_3_00:
|
||||
return "arch_3_00";
|
||||
case PPC_HAS_IEEE128:
|
||||
return "ieee128";
|
||||
case PPC_DARN:
|
||||
return "darn";
|
||||
case PPC_SCV:
|
||||
return "scv";
|
||||
case PPC_HTM_NO_SUSPEND:
|
||||
return "htm-no-suspend";
|
||||
case PPC_LAST_:
|
||||
break;
|
||||
}
|
||||
return "unknown_feature";
|
||||
}
|
||||
447
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_x86.c
Executable file
447
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/cpuinfo_x86.c
Executable file
|
|
@ -0,0 +1,447 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_x86.h"
|
||||
#include "internal/bit_utils.h"
|
||||
#include "internal/cpuid_x86.h"
|
||||
|
||||
#include <stdbool.h>
|
||||
#include <string.h>
|
||||
|
||||
static const Leaf kEmptyLeaf;
|
||||
|
||||
static Leaf SafeCpuId(uint32_t max_cpuid_leaf, uint32_t leaf_id) {
|
||||
if (leaf_id <= max_cpuid_leaf) {
|
||||
return CpuId(leaf_id);
|
||||
} else {
|
||||
return kEmptyLeaf;
|
||||
}
|
||||
}
|
||||
|
||||
#define MASK_XMM 0x2
|
||||
#define MASK_YMM 0x4
|
||||
#define MASK_MASKREG 0x20
|
||||
#define MASK_ZMM0_15 0x40
|
||||
#define MASK_ZMM16_31 0x80
|
||||
|
||||
static bool HasMask(uint32_t value, uint32_t mask) {
|
||||
return (value & mask) == mask;
|
||||
}
|
||||
|
||||
// Checks that operating system saves and restores xmm registers during context
|
||||
// switches.
|
||||
static bool HasXmmOsXSave(uint32_t xcr0_eax) {
|
||||
return HasMask(xcr0_eax, MASK_XMM);
|
||||
}
|
||||
|
||||
// Checks that operating system saves and restores ymm registers during context
|
||||
// switches.
|
||||
static bool HasYmmOsXSave(uint32_t xcr0_eax) {
|
||||
return HasMask(xcr0_eax, MASK_XMM | MASK_YMM);
|
||||
}
|
||||
|
||||
// Checks that operating system saves and restores zmm registers during context
|
||||
// switches.
|
||||
static bool HasZmmOsXSave(uint32_t xcr0_eax) {
|
||||
return HasMask(xcr0_eax, MASK_XMM | MASK_YMM | MASK_MASKREG | MASK_ZMM0_15 |
|
||||
MASK_ZMM16_31);
|
||||
}
|
||||
|
||||
static void SetVendor(const Leaf leaf, char* const vendor) {
|
||||
*(uint32_t*)(vendor) = leaf.ebx;
|
||||
*(uint32_t*)(vendor + 4) = leaf.edx;
|
||||
*(uint32_t*)(vendor + 8) = leaf.ecx;
|
||||
vendor[12] = '\0';
|
||||
}
|
||||
|
||||
static int IsVendor(const Leaf leaf, const char* const name) {
|
||||
const uint32_t ebx = *(const uint32_t*)(name);
|
||||
const uint32_t edx = *(const uint32_t*)(name + 4);
|
||||
const uint32_t ecx = *(const uint32_t*)(name + 8);
|
||||
return leaf.ebx == ebx && leaf.ecx == ecx && leaf.edx == edx;
|
||||
}
|
||||
|
||||
// Reference https://en.wikipedia.org/wiki/CPUID.
|
||||
static void ParseCpuId(const uint32_t max_cpuid_leaf, X86Info* info) {
|
||||
const Leaf leaf_1 = SafeCpuId(max_cpuid_leaf, 1);
|
||||
const Leaf leaf_7 = SafeCpuId(max_cpuid_leaf, 7);
|
||||
|
||||
const bool have_xsave = IsBitSet(leaf_1.ecx, 26);
|
||||
const bool have_osxsave = IsBitSet(leaf_1.ecx, 27);
|
||||
const uint32_t xcr0_eax = (have_xsave && have_osxsave) ? GetXCR0Eax() : 0;
|
||||
const bool have_sse_os_support = HasXmmOsXSave(xcr0_eax);
|
||||
const bool have_avx_os_support = HasYmmOsXSave(xcr0_eax);
|
||||
const bool have_avx512_os_support = HasZmmOsXSave(xcr0_eax);
|
||||
|
||||
const uint32_t family = ExtractBitRange(leaf_1.eax, 11, 8);
|
||||
const uint32_t extended_family = ExtractBitRange(leaf_1.eax, 27, 20);
|
||||
const uint32_t model = ExtractBitRange(leaf_1.eax, 7, 4);
|
||||
const uint32_t extended_model = ExtractBitRange(leaf_1.eax, 19, 16);
|
||||
|
||||
X86Features* const features = &info->features;
|
||||
|
||||
info->family = extended_family + family;
|
||||
info->model = (extended_model << 4) + model;
|
||||
info->stepping = ExtractBitRange(leaf_1.eax, 3, 0);
|
||||
|
||||
features->smx = IsBitSet(leaf_1.ecx, 6);
|
||||
features->cx16 = IsBitSet(leaf_1.ecx, 13);
|
||||
features->aes = IsBitSet(leaf_1.ecx, 25);
|
||||
features->f16c = IsBitSet(leaf_1.ecx, 29);
|
||||
features->sgx = IsBitSet(leaf_7.ebx, 2);
|
||||
features->bmi1 = IsBitSet(leaf_7.ebx, 3);
|
||||
features->bmi2 = IsBitSet(leaf_7.ebx, 8);
|
||||
features->erms = IsBitSet(leaf_7.ebx, 9);
|
||||
features->vpclmulqdq = IsBitSet(leaf_7.ecx, 10);
|
||||
|
||||
if (have_sse_os_support) {
|
||||
features->ssse3 = IsBitSet(leaf_1.ecx, 9);
|
||||
features->sse4_1 = IsBitSet(leaf_1.ecx, 19);
|
||||
features->sse4_2 = IsBitSet(leaf_1.ecx, 20);
|
||||
}
|
||||
|
||||
if (have_avx_os_support) {
|
||||
features->fma3 = IsBitSet(leaf_1.ecx, 12);
|
||||
features->avx = IsBitSet(leaf_1.ecx, 28);
|
||||
features->avx2 = IsBitSet(leaf_7.ebx, 5);
|
||||
}
|
||||
|
||||
if (have_avx512_os_support) {
|
||||
features->avx512f = IsBitSet(leaf_7.ebx, 16);
|
||||
features->avx512cd = IsBitSet(leaf_7.ebx, 28);
|
||||
features->avx512er = IsBitSet(leaf_7.ebx, 27);
|
||||
features->avx512pf = IsBitSet(leaf_7.ebx, 26);
|
||||
features->avx512bw = IsBitSet(leaf_7.ebx, 30);
|
||||
features->avx512dq = IsBitSet(leaf_7.ebx, 17);
|
||||
features->avx512vl = IsBitSet(leaf_7.ebx, 31);
|
||||
features->avx512ifma = IsBitSet(leaf_7.ebx, 21);
|
||||
features->avx512vbmi = IsBitSet(leaf_7.ecx, 1);
|
||||
features->avx512vbmi2 = IsBitSet(leaf_7.ecx, 6);
|
||||
features->avx512vnni = IsBitSet(leaf_7.ecx, 11);
|
||||
features->avx512bitalg = IsBitSet(leaf_7.ecx, 12);
|
||||
features->avx512vpopcntdq = IsBitSet(leaf_7.ecx, 14);
|
||||
features->avx512_4vnniw = IsBitSet(leaf_7.edx, 2);
|
||||
features->avx512_4vbmi2 = IsBitSet(leaf_7.edx, 3);
|
||||
}
|
||||
}
|
||||
|
||||
static const X86Info kEmptyX86Info;
|
||||
|
||||
X86Info GetX86Info(void) {
|
||||
X86Info info = kEmptyX86Info;
|
||||
const Leaf leaf_0 = CpuId(0);
|
||||
const uint32_t max_cpuid_leaf = leaf_0.eax;
|
||||
SetVendor(leaf_0, info.vendor);
|
||||
if (IsVendor(leaf_0, "GenuineIntel") || IsVendor(leaf_0, "AuthenticAMD")) {
|
||||
ParseCpuId(max_cpuid_leaf, &info);
|
||||
}
|
||||
return info;
|
||||
}
|
||||
|
||||
#define CPUID(FAMILY, MODEL) (((FAMILY & 0xFF) << 8) | (MODEL & 0xFF))
|
||||
|
||||
X86Microarchitecture GetX86Microarchitecture(const X86Info* info) {
|
||||
if (memcmp(info->vendor, "GenuineIntel", sizeof(info->vendor)) == 0) {
|
||||
switch (CPUID(info->family, info->model)) {
|
||||
case CPUID(0x06, 0x35):
|
||||
case CPUID(0x06, 0x36):
|
||||
// https://en.wikipedia.org/wiki/Bonnell_(microarchitecture)
|
||||
return INTEL_ATOM_BNL;
|
||||
case CPUID(0x06, 0x37):
|
||||
case CPUID(0x06, 0x4C):
|
||||
// https://en.wikipedia.org/wiki/Silvermont
|
||||
return INTEL_ATOM_SMT;
|
||||
case CPUID(0x06, 0x5C):
|
||||
// https://en.wikipedia.org/wiki/Goldmont
|
||||
return INTEL_ATOM_GMT;
|
||||
case CPUID(0x06, 0x0F):
|
||||
case CPUID(0x06, 0x16):
|
||||
// https://en.wikipedia.org/wiki/Intel_Core_(microarchitecture)
|
||||
return INTEL_CORE;
|
||||
case CPUID(0x06, 0x17):
|
||||
case CPUID(0x06, 0x1D):
|
||||
// https://en.wikipedia.org/wiki/Penryn_(microarchitecture)
|
||||
return INTEL_PNR;
|
||||
case CPUID(0x06, 0x1A):
|
||||
case CPUID(0x06, 0x1E):
|
||||
case CPUID(0x06, 0x1F):
|
||||
case CPUID(0x06, 0x2E):
|
||||
// https://en.wikipedia.org/wiki/Nehalem_(microarchitecture)
|
||||
return INTEL_NHM;
|
||||
case CPUID(0x06, 0x25):
|
||||
case CPUID(0x06, 0x2C):
|
||||
case CPUID(0x06, 0x2F):
|
||||
// https://en.wikipedia.org/wiki/Westmere_(microarchitecture)
|
||||
return INTEL_WSM;
|
||||
case CPUID(0x06, 0x2A):
|
||||
case CPUID(0x06, 0x2D):
|
||||
// https://en.wikipedia.org/wiki/Sandy_Bridge#Models_and_steppings
|
||||
return INTEL_SNB;
|
||||
case CPUID(0x06, 0x3A):
|
||||
case CPUID(0x06, 0x3E):
|
||||
// https://en.wikipedia.org/wiki/Ivy_Bridge_(microarchitecture)#Models_and_steppings
|
||||
return INTEL_IVB;
|
||||
case CPUID(0x06, 0x3C):
|
||||
case CPUID(0x06, 0x3F):
|
||||
case CPUID(0x06, 0x45):
|
||||
case CPUID(0x06, 0x46):
|
||||
// https://en.wikipedia.org/wiki/Haswell_(microarchitecture)
|
||||
return INTEL_HSW;
|
||||
case CPUID(0x06, 0x3D):
|
||||
case CPUID(0x06, 0x47):
|
||||
case CPUID(0x06, 0x4F):
|
||||
case CPUID(0x06, 0x56):
|
||||
// https://en.wikipedia.org/wiki/Broadwell_(microarchitecture)
|
||||
return INTEL_BDW;
|
||||
case CPUID(0x06, 0x4E):
|
||||
case CPUID(0x06, 0x55):
|
||||
case CPUID(0x06, 0x5E):
|
||||
// https://en.wikipedia.org/wiki/Skylake_(microarchitecture)
|
||||
return INTEL_SKL;
|
||||
case CPUID(0x06, 0x8E):
|
||||
case CPUID(0x06, 0x9E):
|
||||
// https://en.wikipedia.org/wiki/Kaby_Lake
|
||||
return INTEL_KBL;
|
||||
default:
|
||||
return X86_UNKNOWN;
|
||||
}
|
||||
}
|
||||
if (memcmp(info->vendor, "AuthenticAMD", sizeof(info->vendor)) == 0) {
|
||||
switch (info->family) {
|
||||
// https://en.wikipedia.org/wiki/List_of_AMD_CPU_microarchitectures
|
||||
case 0x0F:
|
||||
return AMD_HAMMER;
|
||||
case 0x10:
|
||||
return AMD_K10;
|
||||
case 0x14:
|
||||
return AMD_BOBCAT;
|
||||
case 0x15:
|
||||
return AMD_BULLDOZER;
|
||||
case 0x16:
|
||||
return AMD_JAGUAR;
|
||||
case 0x17:
|
||||
return AMD_ZEN;
|
||||
default:
|
||||
return X86_UNKNOWN;
|
||||
}
|
||||
}
|
||||
return X86_UNKNOWN;
|
||||
}
|
||||
|
||||
static void SetString(const uint32_t max_cpuid_ext_leaf, const uint32_t leaf_id,
|
||||
char* buffer) {
|
||||
const Leaf leaf = SafeCpuId(max_cpuid_ext_leaf, leaf_id);
|
||||
// We allow calling memcpy from SetString which is only called when requesting
|
||||
// X86BrandString.
|
||||
memcpy(buffer, &leaf, sizeof(Leaf));
|
||||
}
|
||||
|
||||
void FillX86BrandString(char brand_string[49]) {
|
||||
const Leaf leaf_ext_0 = CpuId(0x80000000);
|
||||
const uint32_t max_cpuid_leaf_ext = leaf_ext_0.eax;
|
||||
SetString(max_cpuid_leaf_ext, 0x80000002, brand_string);
|
||||
SetString(max_cpuid_leaf_ext, 0x80000003, brand_string + 16);
|
||||
SetString(max_cpuid_leaf_ext, 0x80000004, brand_string + 32);
|
||||
brand_string[48] = '\0';
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Introspection functions
|
||||
|
||||
int GetX86FeaturesEnumValue(const X86Features* features,
|
||||
X86FeaturesEnum value) {
|
||||
switch (value) {
|
||||
case X86_AES:
|
||||
return features->aes;
|
||||
case X86_ERMS:
|
||||
return features->erms;
|
||||
case X86_F16C:
|
||||
return features->f16c;
|
||||
case X86_FMA3:
|
||||
return features->fma3;
|
||||
case X86_VPCLMULQDQ:
|
||||
return features->vpclmulqdq;
|
||||
case X86_BMI1:
|
||||
return features->bmi1;
|
||||
case X86_BMI2:
|
||||
return features->bmi2;
|
||||
case X86_SSSE3:
|
||||
return features->ssse3;
|
||||
case X86_SSE4_1:
|
||||
return features->sse4_1;
|
||||
case X86_SSE4_2:
|
||||
return features->sse4_2;
|
||||
case X86_AVX:
|
||||
return features->avx;
|
||||
case X86_AVX2:
|
||||
return features->avx2;
|
||||
case X86_AVX512F:
|
||||
return features->avx512f;
|
||||
case X86_AVX512CD:
|
||||
return features->avx512cd;
|
||||
case X86_AVX512ER:
|
||||
return features->avx512er;
|
||||
case X86_AVX512PF:
|
||||
return features->avx512pf;
|
||||
case X86_AVX512BW:
|
||||
return features->avx512bw;
|
||||
case X86_AVX512DQ:
|
||||
return features->avx512dq;
|
||||
case X86_AVX512VL:
|
||||
return features->avx512vl;
|
||||
case X86_AVX512IFMA:
|
||||
return features->avx512ifma;
|
||||
case X86_AVX512VBMI:
|
||||
return features->avx512vbmi;
|
||||
case X86_AVX512VBMI2:
|
||||
return features->avx512vbmi2;
|
||||
case X86_AVX512VNNI:
|
||||
return features->avx512vnni;
|
||||
case X86_AVX512BITALG:
|
||||
return features->avx512bitalg;
|
||||
case X86_AVX512VPOPCNTDQ:
|
||||
return features->avx512vpopcntdq;
|
||||
case X86_AVX512_4VNNIW:
|
||||
return features->avx512_4vnniw;
|
||||
case X86_AVX512_4VBMI2:
|
||||
return features->avx512_4vbmi2;
|
||||
case X86_SMX:
|
||||
return features->smx;
|
||||
case X86_SGX:
|
||||
return features->sgx;
|
||||
case X86_CX16:
|
||||
return features->cx16;
|
||||
case X86_LAST_:
|
||||
break;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
const char* GetX86FeaturesEnumName(X86FeaturesEnum value) {
|
||||
switch (value) {
|
||||
case X86_AES:
|
||||
return "aes";
|
||||
case X86_ERMS:
|
||||
return "erms";
|
||||
case X86_F16C:
|
||||
return "f16c";
|
||||
case X86_FMA3:
|
||||
return "fma3";
|
||||
case X86_VPCLMULQDQ:
|
||||
return "vpclmulqdq";
|
||||
case X86_BMI1:
|
||||
return "bmi1";
|
||||
case X86_BMI2:
|
||||
return "bmi2";
|
||||
case X86_SSSE3:
|
||||
return "ssse3";
|
||||
case X86_SSE4_1:
|
||||
return "sse4_1";
|
||||
case X86_SSE4_2:
|
||||
return "sse4_2";
|
||||
case X86_AVX:
|
||||
return "avx";
|
||||
case X86_AVX2:
|
||||
return "avx2";
|
||||
case X86_AVX512F:
|
||||
return "avx512f";
|
||||
case X86_AVX512CD:
|
||||
return "avx512cd";
|
||||
case X86_AVX512ER:
|
||||
return "avx512er";
|
||||
case X86_AVX512PF:
|
||||
return "avx512pf";
|
||||
case X86_AVX512BW:
|
||||
return "avx512bw";
|
||||
case X86_AVX512DQ:
|
||||
return "avx512dq";
|
||||
case X86_AVX512VL:
|
||||
return "avx512vl";
|
||||
case X86_AVX512IFMA:
|
||||
return "avx512ifma";
|
||||
case X86_AVX512VBMI:
|
||||
return "avx512vbmi";
|
||||
case X86_AVX512VBMI2:
|
||||
return "avx512vbmi2";
|
||||
case X86_AVX512VNNI:
|
||||
return "avx512vnni";
|
||||
case X86_AVX512BITALG:
|
||||
return "avx512bitalg";
|
||||
case X86_AVX512VPOPCNTDQ:
|
||||
return "avx512vpopcntdq";
|
||||
case X86_AVX512_4VNNIW:
|
||||
return "avx512_4vnniw";
|
||||
case X86_AVX512_4VBMI2:
|
||||
return "avx512_4vbmi2";
|
||||
case X86_SMX:
|
||||
return "smx";
|
||||
case X86_SGX:
|
||||
return "sgx";
|
||||
case X86_CX16:
|
||||
return "cx16";
|
||||
case X86_LAST_:
|
||||
break;
|
||||
}
|
||||
return "unknown_feature";
|
||||
}
|
||||
|
||||
const char* GetX86MicroarchitectureName(X86Microarchitecture uarch) {
|
||||
switch (uarch) {
|
||||
case X86_UNKNOWN:
|
||||
return "X86_UNKNOWN";
|
||||
case INTEL_CORE:
|
||||
return "INTEL_CORE";
|
||||
case INTEL_PNR:
|
||||
return "INTEL_PNR";
|
||||
case INTEL_NHM:
|
||||
return "INTEL_NHM";
|
||||
case INTEL_ATOM_BNL:
|
||||
return "INTEL_ATOM_BNL";
|
||||
case INTEL_WSM:
|
||||
return "INTEL_WSM";
|
||||
case INTEL_SNB:
|
||||
return "INTEL_SNB";
|
||||
case INTEL_IVB:
|
||||
return "INTEL_IVB";
|
||||
case INTEL_ATOM_SMT:
|
||||
return "INTEL_ATOM_SMT";
|
||||
case INTEL_HSW:
|
||||
return "INTEL_HSW";
|
||||
case INTEL_BDW:
|
||||
return "INTEL_BDW";
|
||||
case INTEL_SKL:
|
||||
return "INTEL_SKL";
|
||||
case INTEL_ATOM_GMT:
|
||||
return "INTEL_ATOM_GMT";
|
||||
case INTEL_KBL:
|
||||
return "INTEL_KBL";
|
||||
case INTEL_CFL:
|
||||
return "INTEL_CFL";
|
||||
case INTEL_CNL:
|
||||
return "INTEL_CNL";
|
||||
case AMD_HAMMER:
|
||||
return "AMD_HAMMER";
|
||||
case AMD_K10:
|
||||
return "AMD_K10";
|
||||
case AMD_BOBCAT:
|
||||
return "AMD_BOBCAT";
|
||||
case AMD_BULLDOZER:
|
||||
return "AMD_BULLDOZER";
|
||||
case AMD_JAGUAR:
|
||||
return "AMD_JAGUAR";
|
||||
case AMD_ZEN:
|
||||
return "AMD_ZEN";
|
||||
}
|
||||
return "unknown microarchitecture";
|
||||
}
|
||||
57
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/filesystem.c
Executable file
57
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/filesystem.c
Executable file
|
|
@ -0,0 +1,57 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/filesystem.h"
|
||||
|
||||
#include <errno.h>
|
||||
#include <fcntl.h>
|
||||
#include <sys/stat.h>
|
||||
#include <sys/types.h>
|
||||
|
||||
#if defined(_MSC_VER)
|
||||
#include <io.h>
|
||||
int CpuFeatures_OpenFile(const char* filename) {
|
||||
return _open(filename, _O_RDONLY);
|
||||
}
|
||||
|
||||
void CpuFeatures_CloseFile(int file_descriptor) { _close(file_descriptor); }
|
||||
|
||||
int CpuFeatures_ReadFile(int file_descriptor, void* buffer,
|
||||
size_t buffer_size) {
|
||||
return _read(file_descriptor, buffer, buffer_size);
|
||||
}
|
||||
|
||||
#else
|
||||
#include <unistd.h>
|
||||
|
||||
int CpuFeatures_OpenFile(const char* filename) {
|
||||
int result;
|
||||
do {
|
||||
result = open(filename, O_RDONLY);
|
||||
} while (result == -1L && errno == EINTR);
|
||||
return result;
|
||||
}
|
||||
|
||||
void CpuFeatures_CloseFile(int file_descriptor) { close(file_descriptor); }
|
||||
|
||||
int CpuFeatures_ReadFile(int file_descriptor, void* buffer,
|
||||
size_t buffer_size) {
|
||||
int result;
|
||||
do {
|
||||
result = read(file_descriptor, buffer, buffer_size);
|
||||
} while (result == -1L && errno == EINTR);
|
||||
return result;
|
||||
}
|
||||
|
||||
#endif
|
||||
194
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/hwcaps.c
Executable file
194
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/hwcaps.c
Executable file
|
|
@ -0,0 +1,194 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include <stdlib.h>
|
||||
#include <string.h>
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
#include "internal/filesystem.h"
|
||||
#include "internal/hwcaps.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#if defined(NDEBUG)
|
||||
#define D(...)
|
||||
#else
|
||||
#include <stdio.h>
|
||||
#define D(...) \
|
||||
do { \
|
||||
printf(__VA_ARGS__); \
|
||||
fflush(stdout); \
|
||||
} while (0)
|
||||
#endif
|
||||
|
||||
#if defined(CPU_FEATURES_ARCH_MIPS) || defined(CPU_FEATURES_ARCH_ANY_ARM)
|
||||
#define HWCAPS_ANDROID_MIPS_OR_ARM
|
||||
#endif
|
||||
|
||||
#if defined(CPU_FEATURES_OS_LINUX_OR_ANDROID) && \
|
||||
!defined(HWCAPS_ANDROID_MIPS_OR_ARM)
|
||||
#define HWCAPS_REGULAR_LINUX
|
||||
#endif
|
||||
|
||||
#if defined(HWCAPS_ANDROID_MIPS_OR_ARM) || defined(HWCAPS_REGULAR_LINUX)
|
||||
#define HWCAPS_SUPPORTED
|
||||
#endif
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Implementation of GetElfHwcapFromGetauxval
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
// On Linux we simply use getauxval.
|
||||
#if defined(HWCAPS_REGULAR_LINUX)
|
||||
#include <dlfcn.h>
|
||||
#include <sys/auxv.h>
|
||||
static unsigned long GetElfHwcapFromGetauxval(uint32_t hwcap_type) {
|
||||
return getauxval(hwcap_type);
|
||||
}
|
||||
#endif // defined(HWCAPS_REGULAR_LINUX)
|
||||
|
||||
// On Android we probe the system's C library for a 'getauxval' function and
|
||||
// call it if it exits, or return 0 for failure. This function is available
|
||||
// since API level 20.
|
||||
//
|
||||
// This code does *NOT* check for '__ANDROID_API__ >= 20' to support the edge
|
||||
// case where some NDK developers use headers for a platform that is newer than
|
||||
// the one really targetted by their application. This is typically done to use
|
||||
// newer native APIs only when running on more recent Android versions, and
|
||||
// requires careful symbol management.
|
||||
//
|
||||
// Note that getauxval() can't really be re-implemented here, because its
|
||||
// implementation does not parse /proc/self/auxv. Instead it depends on values
|
||||
// that are passed by the kernel at process-init time to the C runtime
|
||||
// initialization layer.
|
||||
#if defined(HWCAPS_ANDROID_MIPS_OR_ARM)
|
||||
#include <dlfcn.h>
|
||||
#define AT_HWCAP 16
|
||||
#define AT_HWCAP2 26
|
||||
#define AT_PLATFORM 15
|
||||
#define AT_BASE_PLATFORM 24
|
||||
|
||||
typedef unsigned long getauxval_func_t(unsigned long);
|
||||
|
||||
static uint32_t GetElfHwcapFromGetauxval(uint32_t hwcap_type) {
|
||||
uint32_t ret = 0;
|
||||
void* libc_handle = NULL;
|
||||
getauxval_func_t* func = NULL;
|
||||
|
||||
dlerror(); // Cleaning error state before calling dlopen.
|
||||
libc_handle = dlopen("libc.so", RTLD_NOW);
|
||||
if (!libc_handle) {
|
||||
D("Could not dlopen() C library: %s\n", dlerror());
|
||||
return 0;
|
||||
}
|
||||
func = (getauxval_func_t*)dlsym(libc_handle, "getauxval");
|
||||
if (!func) {
|
||||
D("Could not find getauxval() in C library\n");
|
||||
} else {
|
||||
// Note: getauxval() returns 0 on failure. Doesn't touch errno.
|
||||
ret = (uint32_t)(*func)(hwcap_type);
|
||||
}
|
||||
dlclose(libc_handle);
|
||||
return ret;
|
||||
}
|
||||
#endif // defined(HWCAPS_ANDROID_MIPS_OR_ARM)
|
||||
|
||||
#if defined(HWCAPS_SUPPORTED)
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Implementation of GetHardwareCapabilities for Android and Linux
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
// Fallback when getauxval is not available, retrieves hwcaps from
|
||||
// "/proc/self/auxv".
|
||||
static uint32_t GetElfHwcapFromProcSelfAuxv(uint32_t hwcap_type) {
|
||||
struct {
|
||||
uint32_t tag;
|
||||
uint32_t value;
|
||||
} entry;
|
||||
uint32_t result = 0;
|
||||
const char filepath[] = "/proc/self/auxv";
|
||||
const int fd = CpuFeatures_OpenFile(filepath);
|
||||
if (fd < 0) {
|
||||
D("Could not open %s\n", filepath);
|
||||
return 0;
|
||||
}
|
||||
for (;;) {
|
||||
const int ret = CpuFeatures_ReadFile(fd, (char*)&entry, sizeof entry);
|
||||
if (ret < 0) {
|
||||
D("Error while reading %s\n", filepath);
|
||||
break;
|
||||
}
|
||||
// Detect end of list.
|
||||
if (ret == 0 || (entry.tag == 0 && entry.value == 0)) {
|
||||
break;
|
||||
}
|
||||
if (entry.tag == hwcap_type) {
|
||||
result = entry.value;
|
||||
break;
|
||||
}
|
||||
}
|
||||
CpuFeatures_CloseFile(fd);
|
||||
return result;
|
||||
}
|
||||
|
||||
// Retrieves hardware capabilities by first trying to call getauxval, if not
|
||||
// available falls back to reading "/proc/self/auxv".
|
||||
static unsigned long GetHardwareCapabilitiesFor(uint32_t type) {
|
||||
unsigned long hwcaps = GetElfHwcapFromGetauxval(type);
|
||||
if (!hwcaps) {
|
||||
D("Parsing /proc/self/auxv to extract ELF hwcaps!\n");
|
||||
hwcaps = GetElfHwcapFromProcSelfAuxv(type);
|
||||
}
|
||||
return hwcaps;
|
||||
}
|
||||
|
||||
HardwareCapabilities CpuFeatures_GetHardwareCapabilities(void) {
|
||||
HardwareCapabilities capabilities;
|
||||
capabilities.hwcaps = GetHardwareCapabilitiesFor(AT_HWCAP);
|
||||
capabilities.hwcaps2 = GetHardwareCapabilitiesFor(AT_HWCAP2);
|
||||
return capabilities;
|
||||
}
|
||||
|
||||
PlatformType kEmptyPlatformType;
|
||||
|
||||
PlatformType CpuFeatures_GetPlatformType(void) {
|
||||
PlatformType type = kEmptyPlatformType;
|
||||
char *platform = (char *)GetHardwareCapabilitiesFor(AT_PLATFORM);
|
||||
char *base_platform = (char *)GetHardwareCapabilitiesFor(AT_BASE_PLATFORM);
|
||||
|
||||
if (platform != NULL)
|
||||
CpuFeatures_StringView_CopyString(str(platform), type.platform,
|
||||
sizeof(type.platform));
|
||||
if (base_platform != NULL)
|
||||
CpuFeatures_StringView_CopyString(str(base_platform), type.base_platform,
|
||||
sizeof(type.base_platform));
|
||||
return type;
|
||||
}
|
||||
#else // (defined(HWCAPS_SUPPORTED)
|
||||
|
||||
PlatformType kEmptyPlatformType;
|
||||
|
||||
PlatformType CpuFeatures_GetPlatformType(void) {
|
||||
PlatformType type = kEmptyPlatformType;
|
||||
return type;
|
||||
}
|
||||
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
// Implementation of GetHardwareCapabilities for unsupported platforms.
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
const HardwareCapabilities kEmptyHardwareCapabilities;
|
||||
HardwareCapabilities CpuFeatures_GetHardwareCapabilities(void) {
|
||||
return kEmptyHardwareCapabilities;
|
||||
}
|
||||
#endif
|
||||
|
|
@ -0,0 +1,51 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/linux_features_aggregator.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
void CpuFeatures_SetFromFlags(const size_t configs_size,
|
||||
const CapabilityConfig* configs,
|
||||
const StringView flags_line,
|
||||
void* const features) {
|
||||
size_t i = 0;
|
||||
for (; i < configs_size; ++i) {
|
||||
const CapabilityConfig config = configs[i];
|
||||
config.set_bit(features, CpuFeatures_StringView_HasWord(
|
||||
flags_line, config.proc_cpuinfo_flag));
|
||||
}
|
||||
}
|
||||
|
||||
static bool IsSet(const uint32_t mask, const uint32_t value) {
|
||||
return (value & mask) == mask;
|
||||
}
|
||||
|
||||
static bool IsHwCapsSet(const HardwareCapabilities hwcaps_mask,
|
||||
const HardwareCapabilities hwcaps) {
|
||||
return IsSet(hwcaps_mask.hwcaps, hwcaps.hwcaps) &&
|
||||
IsSet(hwcaps_mask.hwcaps2, hwcaps.hwcaps2);
|
||||
}
|
||||
|
||||
void CpuFeatures_OverrideFromHwCaps(const size_t configs_size,
|
||||
const CapabilityConfig* configs,
|
||||
const HardwareCapabilities hwcaps,
|
||||
void* const features) {
|
||||
size_t i = 0;
|
||||
for (; i < configs_size; ++i) {
|
||||
const CapabilityConfig* config = &configs[i];
|
||||
if (IsHwCapsSet(config->hwcaps_mask, hwcaps)) {
|
||||
config->set_bit(features, true);
|
||||
}
|
||||
}
|
||||
}
|
||||
131
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/stack_line_reader.c
Executable file
131
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/stack_line_reader.c
Executable file
|
|
@ -0,0 +1,131 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "internal/filesystem.h"
|
||||
|
||||
#include <assert.h>
|
||||
#include <errno.h>
|
||||
#include <stdio.h>
|
||||
|
||||
void StackLineReader_Initialize(StackLineReader* reader, int fd) {
|
||||
reader->view.ptr = reader->buffer;
|
||||
reader->view.size = 0;
|
||||
reader->skip_mode = false;
|
||||
reader->fd = fd;
|
||||
}
|
||||
|
||||
// Replaces the content of buffer with bytes from the file.
|
||||
static int LoadFullBuffer(StackLineReader* reader) {
|
||||
const int read = CpuFeatures_ReadFile(reader->fd, reader->buffer,
|
||||
STACK_LINE_READER_BUFFER_SIZE);
|
||||
assert(read >= 0);
|
||||
reader->view.ptr = reader->buffer;
|
||||
reader->view.size = read;
|
||||
return read;
|
||||
}
|
||||
|
||||
// Appends with bytes from the file to buffer, filling the remaining space.
|
||||
static int LoadMore(StackLineReader* reader) {
|
||||
char* const ptr = reader->buffer + reader->view.size;
|
||||
const size_t size_to_read = STACK_LINE_READER_BUFFER_SIZE - reader->view.size;
|
||||
const int read = CpuFeatures_ReadFile(reader->fd, ptr, size_to_read);
|
||||
assert(read >= 0);
|
||||
assert(read <= (int)size_to_read);
|
||||
reader->view.size += read;
|
||||
return read;
|
||||
}
|
||||
|
||||
static int IndexOfEol(StackLineReader* reader) {
|
||||
return CpuFeatures_StringView_IndexOfChar(reader->view, '\n');
|
||||
}
|
||||
|
||||
// Relocate buffer's pending bytes at the beginning of the array and fills the
|
||||
// remaining space with bytes from the file.
|
||||
static int BringToFrontAndLoadMore(StackLineReader* reader) {
|
||||
if (reader->view.size && reader->view.ptr != reader->buffer) {
|
||||
memmove(reader->buffer, reader->view.ptr, reader->view.size);
|
||||
}
|
||||
reader->view.ptr = reader->buffer;
|
||||
return LoadMore(reader);
|
||||
}
|
||||
|
||||
// Loads chunks of buffer size from disks until it contains a newline character
|
||||
// or end of file.
|
||||
static void SkipToNextLine(StackLineReader* reader) {
|
||||
for (;;) {
|
||||
const int read = LoadFullBuffer(reader);
|
||||
if (read == 0) {
|
||||
break;
|
||||
} else {
|
||||
const int eol_index = IndexOfEol(reader);
|
||||
if (eol_index >= 0) {
|
||||
reader->view =
|
||||
CpuFeatures_StringView_PopFront(reader->view, eol_index + 1);
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
static LineResult CreateLineResult(bool eof, bool full_line, StringView view) {
|
||||
LineResult result;
|
||||
result.eof = eof;
|
||||
result.full_line = full_line;
|
||||
result.line = view;
|
||||
return result;
|
||||
}
|
||||
|
||||
// Helper methods to provide clearer semantic in StackLineReader_NextLine.
|
||||
static LineResult CreateEOFLineResult(StringView view) {
|
||||
return CreateLineResult(true, true, view);
|
||||
}
|
||||
|
||||
static LineResult CreateTruncatedLineResult(StringView view) {
|
||||
return CreateLineResult(false, false, view);
|
||||
}
|
||||
|
||||
static LineResult CreateValidLineResult(StringView view) {
|
||||
return CreateLineResult(false, true, view);
|
||||
}
|
||||
|
||||
LineResult StackLineReader_NextLine(StackLineReader* reader) {
|
||||
if (reader->skip_mode) {
|
||||
SkipToNextLine(reader);
|
||||
reader->skip_mode = false;
|
||||
}
|
||||
{
|
||||
const bool can_load_more =
|
||||
reader->view.size < STACK_LINE_READER_BUFFER_SIZE;
|
||||
int eol_index = IndexOfEol(reader);
|
||||
if (eol_index < 0 && can_load_more) {
|
||||
const int read = BringToFrontAndLoadMore(reader);
|
||||
if (read == 0) {
|
||||
return CreateEOFLineResult(reader->view);
|
||||
}
|
||||
eol_index = IndexOfEol(reader);
|
||||
}
|
||||
if (eol_index < 0) {
|
||||
reader->skip_mode = true;
|
||||
return CreateTruncatedLineResult(reader->view);
|
||||
}
|
||||
{
|
||||
StringView line =
|
||||
CpuFeatures_StringView_KeepFront(reader->view, eol_index);
|
||||
reader->view =
|
||||
CpuFeatures_StringView_PopFront(reader->view, eol_index + 1);
|
||||
return CreateValidLineResult(line);
|
||||
}
|
||||
}
|
||||
}
|
||||
182
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/string_view.c
Executable file
182
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/string_view.c
Executable file
|
|
@ -0,0 +1,182 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#include <assert.h>
|
||||
#include <ctype.h>
|
||||
#include <string.h>
|
||||
|
||||
int CpuFeatures_StringView_IndexOfChar(const StringView view, char c) {
|
||||
if (view.ptr && view.size) {
|
||||
const char* const found = (const char*)memchr(view.ptr, c, view.size);
|
||||
if (found) {
|
||||
return found - view.ptr;
|
||||
}
|
||||
}
|
||||
return -1;
|
||||
}
|
||||
|
||||
int CpuFeatures_StringView_IndexOf(const StringView view,
|
||||
const StringView sub_view) {
|
||||
if (sub_view.size) {
|
||||
StringView remainder = view;
|
||||
while (remainder.size >= sub_view.size) {
|
||||
const int found_index =
|
||||
CpuFeatures_StringView_IndexOfChar(remainder, sub_view.ptr[0]);
|
||||
if (found_index < 0) break;
|
||||
remainder = CpuFeatures_StringView_PopFront(remainder, found_index);
|
||||
if (CpuFeatures_StringView_StartsWith(remainder, sub_view)) {
|
||||
return remainder.ptr - view.ptr;
|
||||
}
|
||||
remainder = CpuFeatures_StringView_PopFront(remainder, 1);
|
||||
}
|
||||
}
|
||||
return -1;
|
||||
}
|
||||
|
||||
bool CpuFeatures_StringView_IsEquals(const StringView a, const StringView b) {
|
||||
if (a.size == b.size) {
|
||||
return a.ptr == b.ptr || memcmp(a.ptr, b.ptr, b.size) == 0;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
bool CpuFeatures_StringView_StartsWith(const StringView a, const StringView b) {
|
||||
return a.ptr && b.ptr && b.size && a.size >= b.size
|
||||
? memcmp(a.ptr, b.ptr, b.size) == 0
|
||||
: false;
|
||||
}
|
||||
|
||||
StringView CpuFeatures_StringView_PopFront(const StringView str_view,
|
||||
size_t count) {
|
||||
if (count > str_view.size) {
|
||||
return kEmptyStringView;
|
||||
}
|
||||
return view(str_view.ptr + count, str_view.size - count);
|
||||
}
|
||||
|
||||
StringView CpuFeatures_StringView_PopBack(const StringView str_view,
|
||||
size_t count) {
|
||||
if (count > str_view.size) {
|
||||
return kEmptyStringView;
|
||||
}
|
||||
return view(str_view.ptr, str_view.size - count);
|
||||
}
|
||||
|
||||
StringView CpuFeatures_StringView_KeepFront(const StringView str_view,
|
||||
size_t count) {
|
||||
return count <= str_view.size ? view(str_view.ptr, count) : str_view;
|
||||
}
|
||||
|
||||
char CpuFeatures_StringView_Front(const StringView view) {
|
||||
assert(view.size);
|
||||
assert(view.ptr);
|
||||
return view.ptr[0];
|
||||
}
|
||||
|
||||
char CpuFeatures_StringView_Back(const StringView view) {
|
||||
assert(view.size);
|
||||
return view.ptr[view.size - 1];
|
||||
}
|
||||
|
||||
StringView CpuFeatures_StringView_TrimWhitespace(StringView view) {
|
||||
while (view.size && isspace(CpuFeatures_StringView_Front(view)))
|
||||
view = CpuFeatures_StringView_PopFront(view, 1);
|
||||
while (view.size && isspace(CpuFeatures_StringView_Back(view)))
|
||||
view = CpuFeatures_StringView_PopBack(view, 1);
|
||||
return view;
|
||||
}
|
||||
|
||||
static int HexValue(const char c) {
|
||||
if (c >= '0' && c <= '9') return c - '0';
|
||||
if (c >= 'a' && c <= 'f') return c - 'a' + 10;
|
||||
if (c >= 'A' && c <= 'F') return c - 'A' + 10;
|
||||
return -1;
|
||||
}
|
||||
|
||||
// Returns -1 if view contains non digits.
|
||||
static int ParsePositiveNumberWithBase(const StringView view, int base) {
|
||||
int result = 0;
|
||||
StringView remainder = view;
|
||||
for (; remainder.size;
|
||||
remainder = CpuFeatures_StringView_PopFront(remainder, 1)) {
|
||||
const int value = HexValue(CpuFeatures_StringView_Front(remainder));
|
||||
if (value < 0 || value >= base) return -1;
|
||||
result = (result * base) + value;
|
||||
}
|
||||
return result;
|
||||
}
|
||||
|
||||
int CpuFeatures_StringView_ParsePositiveNumber(const StringView view) {
|
||||
if (view.size) {
|
||||
const StringView hex_prefix = str("0x");
|
||||
if (CpuFeatures_StringView_StartsWith(view, hex_prefix)) {
|
||||
const StringView span_no_prefix =
|
||||
CpuFeatures_StringView_PopFront(view, hex_prefix.size);
|
||||
return ParsePositiveNumberWithBase(span_no_prefix, 16);
|
||||
}
|
||||
return ParsePositiveNumberWithBase(view, 10);
|
||||
}
|
||||
return -1;
|
||||
}
|
||||
|
||||
void CpuFeatures_StringView_CopyString(const StringView src, char* dst,
|
||||
size_t dst_size) {
|
||||
if (dst_size > 0) {
|
||||
const size_t max_copy_size = dst_size - 1;
|
||||
const size_t copy_size =
|
||||
src.size > max_copy_size ? max_copy_size : src.size;
|
||||
memcpy(dst, src.ptr, copy_size);
|
||||
dst[copy_size] = '\0';
|
||||
}
|
||||
}
|
||||
|
||||
bool CpuFeatures_StringView_HasWord(const StringView line,
|
||||
const char* const word_str) {
|
||||
const StringView word = str(word_str);
|
||||
StringView remainder = line;
|
||||
for (;;) {
|
||||
const int index_of_word = CpuFeatures_StringView_IndexOf(remainder, word);
|
||||
if (index_of_word < 0) {
|
||||
return false;
|
||||
} else {
|
||||
const StringView before =
|
||||
CpuFeatures_StringView_KeepFront(line, index_of_word);
|
||||
const StringView after =
|
||||
CpuFeatures_StringView_PopFront(line, index_of_word + word.size);
|
||||
const bool valid_before =
|
||||
before.size == 0 || CpuFeatures_StringView_Back(before) == ' ';
|
||||
const bool valid_after =
|
||||
after.size == 0 || CpuFeatures_StringView_Front(after) == ' ';
|
||||
if (valid_before && valid_after) return true;
|
||||
remainder =
|
||||
CpuFeatures_StringView_PopFront(remainder, index_of_word + word.size);
|
||||
}
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
bool CpuFeatures_StringView_GetAttributeKeyValue(const StringView line,
|
||||
StringView* key,
|
||||
StringView* value) {
|
||||
const StringView sep = str(": ");
|
||||
const int index_of_separator = CpuFeatures_StringView_IndexOf(line, sep);
|
||||
if (index_of_separator < 0) return false;
|
||||
*value = CpuFeatures_StringView_TrimWhitespace(
|
||||
CpuFeatures_StringView_PopFront(line, index_of_separator + sep.size));
|
||||
*key = CpuFeatures_StringView_TrimWhitespace(
|
||||
CpuFeatures_StringView_KeepFront(line, index_of_separator));
|
||||
return true;
|
||||
}
|
||||
237
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/utils/list_cpu_features.c
Executable file
237
src/crypto/argon2_hasher/hash/cpu/cpu_features/src/utils/list_cpu_features.c
Executable file
|
|
@ -0,0 +1,237 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include <stdio.h>
|
||||
#include <stdlib.h>
|
||||
#include <string.h>
|
||||
|
||||
#include "cpu_features_macros.h"
|
||||
#include "cpuinfo_aarch64.h"
|
||||
#include "cpuinfo_arm.h"
|
||||
#include "cpuinfo_mips.h"
|
||||
#include "cpuinfo_ppc.h"
|
||||
#include "cpuinfo_x86.h"
|
||||
|
||||
static void PrintEscapedAscii(const char* str) {
|
||||
putchar('"');
|
||||
for (; str && *str; ++str) {
|
||||
switch (*str) {
|
||||
case '\"':
|
||||
case '\\':
|
||||
case '/':
|
||||
case '\b':
|
||||
case '\f':
|
||||
case '\n':
|
||||
case '\r':
|
||||
case '\t':
|
||||
putchar('\\');
|
||||
}
|
||||
putchar(*str);
|
||||
}
|
||||
putchar('"');
|
||||
}
|
||||
|
||||
static void PrintVoid(void) {}
|
||||
static void PrintComma(void) { putchar(','); }
|
||||
static void PrintLineFeed(void) { putchar('\n'); }
|
||||
static void PrintOpenBrace(void) { putchar('{'); }
|
||||
static void PrintCloseBrace(void) { putchar('}'); }
|
||||
static void PrintOpenBracket(void) { putchar('['); }
|
||||
static void PrintCloseBracket(void) { putchar(']'); }
|
||||
static void PrintString(const char* field) { printf("%s", field); }
|
||||
static void PrintAlignedHeader(const char* field) { printf("%-15s : ", field); }
|
||||
static void PrintIntValue(int value) { printf("%d", value); }
|
||||
static void PrintDecHexValue(int value) {
|
||||
printf("%3d (0x%02X)", value, value);
|
||||
}
|
||||
static void PrintJsonHeader(const char* field) {
|
||||
PrintEscapedAscii(field);
|
||||
putchar(':');
|
||||
}
|
||||
|
||||
typedef struct {
|
||||
void (*Start)(void);
|
||||
void (*ArrayStart)(void);
|
||||
void (*ArraySeparator)(void);
|
||||
void (*ArrayEnd)(void);
|
||||
void (*PrintString)(const char* value);
|
||||
void (*PrintValue)(int value);
|
||||
void (*EndField)(void);
|
||||
void (*StartField)(const char* field);
|
||||
void (*End)(void);
|
||||
} Printer;
|
||||
|
||||
static Printer getJsonPrinter(void) {
|
||||
return (Printer){
|
||||
.Start = &PrintOpenBrace,
|
||||
.ArrayStart = &PrintOpenBracket,
|
||||
.ArraySeparator = &PrintComma,
|
||||
.ArrayEnd = &PrintCloseBracket,
|
||||
.PrintString = &PrintEscapedAscii,
|
||||
.PrintValue = &PrintIntValue,
|
||||
.EndField = &PrintComma,
|
||||
.StartField = &PrintJsonHeader,
|
||||
.End = &PrintCloseBrace,
|
||||
};
|
||||
}
|
||||
|
||||
static Printer getTextPrinter(void) {
|
||||
return (Printer){
|
||||
.Start = &PrintVoid,
|
||||
.ArrayStart = &PrintVoid,
|
||||
.ArraySeparator = &PrintComma,
|
||||
.ArrayEnd = &PrintVoid,
|
||||
.PrintString = &PrintString,
|
||||
.PrintValue = &PrintDecHexValue,
|
||||
.EndField = &PrintLineFeed,
|
||||
.StartField = &PrintAlignedHeader,
|
||||
.End = &PrintVoid,
|
||||
};
|
||||
}
|
||||
|
||||
// Prints a named numeric value in both decimal and hexadecimal.
|
||||
static void PrintN(const Printer p, const char* field, int value) {
|
||||
p.StartField(field);
|
||||
p.PrintValue(value);
|
||||
p.EndField();
|
||||
}
|
||||
|
||||
// Prints a named string.
|
||||
static void PrintS(const Printer p, const char* field, const char* value) {
|
||||
p.StartField(field);
|
||||
p.PrintString(value);
|
||||
p.EndField();
|
||||
}
|
||||
|
||||
static int cmp(const void* p1, const void* p2) {
|
||||
return strcmp(*(const char* const*)p1, *(const char* const*)p2);
|
||||
}
|
||||
|
||||
#define DEFINE_PRINT_FLAGS(HasFeature, FeatureName, FeatureType, LastEnum) \
|
||||
static void PrintFlags(const Printer p, const FeatureType* features) { \
|
||||
size_t i; \
|
||||
const char* ptrs[LastEnum] = {0}; \
|
||||
size_t count = 0; \
|
||||
for (i = 0; i < LastEnum; ++i) { \
|
||||
if (HasFeature(features, i)) { \
|
||||
ptrs[count] = FeatureName(i); \
|
||||
++count; \
|
||||
} \
|
||||
} \
|
||||
qsort(ptrs, count, sizeof(char*), cmp); \
|
||||
p.StartField("flags"); \
|
||||
p.ArrayStart(); \
|
||||
for (i = 0; i < count; ++i) { \
|
||||
if (i > 0) p.ArraySeparator(); \
|
||||
p.PrintString(ptrs[i]); \
|
||||
} \
|
||||
p.ArrayEnd(); \
|
||||
}
|
||||
|
||||
#if defined(CPU_FEATURES_ARCH_X86)
|
||||
DEFINE_PRINT_FLAGS(GetX86FeaturesEnumValue, GetX86FeaturesEnumName, X86Features,
|
||||
X86_LAST_)
|
||||
#elif defined(CPU_FEATURES_ARCH_ARM)
|
||||
DEFINE_PRINT_FLAGS(GetArmFeaturesEnumValue, GetArmFeaturesEnumName, ArmFeatures,
|
||||
ARM_LAST_)
|
||||
#elif defined(CPU_FEATURES_ARCH_AARCH64)
|
||||
DEFINE_PRINT_FLAGS(GetAarch64FeaturesEnumValue, GetAarch64FeaturesEnumName,
|
||||
Aarch64Features, AARCH64_LAST_)
|
||||
#elif defined(CPU_FEATURES_ARCH_MIPS)
|
||||
DEFINE_PRINT_FLAGS(GetMipsFeaturesEnumValue, GetMipsFeaturesEnumName,
|
||||
MipsFeatures, MIPS_LAST_)
|
||||
#elif defined(CPU_FEATURES_ARCH_PPC)
|
||||
DEFINE_PRINT_FLAGS(GetPPCFeaturesEnumValue, GetPPCFeaturesEnumName, PPCFeatures,
|
||||
PPC_LAST_)
|
||||
#endif
|
||||
|
||||
static void PrintFeatures(const Printer printer) {
|
||||
#if defined(CPU_FEATURES_ARCH_X86)
|
||||
char brand_string[49];
|
||||
const X86Info info = GetX86Info();
|
||||
FillX86BrandString(brand_string);
|
||||
PrintS(printer, "arch", "x86");
|
||||
PrintS(printer, "brand", brand_string);
|
||||
PrintN(printer, "family", info.family);
|
||||
PrintN(printer, "model", info.model);
|
||||
PrintN(printer, "stepping", info.stepping);
|
||||
PrintS(printer, "uarch",
|
||||
GetX86MicroarchitectureName(GetX86Microarchitecture(&info)));
|
||||
PrintFlags(printer, &info.features);
|
||||
#elif defined(CPU_FEATURES_ARCH_ARM)
|
||||
const ArmInfo info = GetArmInfo();
|
||||
PrintS(printer, "arch", "ARM");
|
||||
PrintN(printer, "implementer", info.implementer);
|
||||
PrintN(printer, "architecture", info.architecture);
|
||||
PrintN(printer, "variant", info.variant);
|
||||
PrintN(printer, "part", info.part);
|
||||
PrintN(printer, "revision", info.revision);
|
||||
PrintFlags(printer, &info.features);
|
||||
#elif defined(CPU_FEATURES_ARCH_AARCH64)
|
||||
const Aarch64Info info = GetAarch64Info();
|
||||
PrintS(printer, "arch", "aarch64");
|
||||
PrintN(printer, "implementer", info.implementer);
|
||||
PrintN(printer, "variant", info.variant);
|
||||
PrintN(printer, "part", info.part);
|
||||
PrintN(printer, "revision", info.revision);
|
||||
PrintFlags(printer, &info.features);
|
||||
#elif defined(CPU_FEATURES_ARCH_MIPS)
|
||||
const MipsInfo info = GetMipsInfo();
|
||||
PrintS(printer, "arch", "mips");
|
||||
PrintFlags(printer, &info.features);
|
||||
#elif defined(CPU_FEATURES_ARCH_PPC)
|
||||
const PPCInfo info = GetPPCInfo();
|
||||
const PPCPlatformStrings strings = GetPPCPlatformStrings();
|
||||
PrintS(printer, "arch", "ppc");
|
||||
PrintS(printer, "platform", strings.platform);
|
||||
PrintS(printer, "model", strings.model);
|
||||
PrintS(printer, "machine", strings.machine);
|
||||
PrintS(printer, "cpu", strings.cpu);
|
||||
PrintS(printer, "instruction set", strings.type.platform);
|
||||
PrintS(printer, "microarchitecture", strings.type.base_platform);
|
||||
PrintFlags(printer, &info.features);
|
||||
#endif
|
||||
}
|
||||
|
||||
static void showUsage(const char* name) {
|
||||
printf(
|
||||
"\n"
|
||||
"Usage: %s [options]\n"
|
||||
" Options:\n"
|
||||
" -h | --help Show help message.\n"
|
||||
" -j | --json Format output as json instead of plain text.\n"
|
||||
"\n",
|
||||
name);
|
||||
}
|
||||
|
||||
int main(int argc, char** argv) {
|
||||
Printer printer = getTextPrinter();
|
||||
int i = 1;
|
||||
for (; i < argc; ++i) {
|
||||
const char* arg = argv[i];
|
||||
if (strcmp(arg, "-j") == 0 || strcmp(arg, "--json") == 0) {
|
||||
printer = getJsonPrinter();
|
||||
} else {
|
||||
showUsage(argv[0]);
|
||||
if (strcmp(arg, "-h") == 0 || strcmp(arg, "--help") == 0)
|
||||
return EXIT_SUCCESS;
|
||||
return EXIT_FAILURE;
|
||||
}
|
||||
}
|
||||
printer.Start();
|
||||
PrintFeatures(printer);
|
||||
printer.End();
|
||||
PrintLineFeed();
|
||||
return EXIT_SUCCESS;
|
||||
}
|
||||
79
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/CMakeLists.txt
Executable file
79
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/CMakeLists.txt
Executable file
|
|
@ -0,0 +1,79 @@
|
|||
#
|
||||
# libraries for tests
|
||||
#
|
||||
|
||||
set(CMAKE_CXX_STANDARD 11)
|
||||
set(CMAKE_CXX_STANDARD_REQUIRED ON)
|
||||
set(CMAKE_CXX_EXTENSIONS OFF) # prefer use of -std11 instead of -gnustd11
|
||||
|
||||
include_directories(../include)
|
||||
add_definitions(-DCPU_FEATURES_TEST)
|
||||
|
||||
##------------------------------------------------------------------------------
|
||||
add_library(string_view ../src/string_view.c)
|
||||
##------------------------------------------------------------------------------
|
||||
add_library(filesystem_for_testing filesystem_for_testing.cc)
|
||||
##------------------------------------------------------------------------------
|
||||
add_library(hwcaps_for_testing hwcaps_for_testing.cc)
|
||||
target_link_libraries(hwcaps_for_testing filesystem_for_testing)
|
||||
##------------------------------------------------------------------------------
|
||||
add_library(stack_line_reader ../src/stack_line_reader.c)
|
||||
target_compile_definitions(stack_line_reader PUBLIC STACK_LINE_READER_BUFFER_SIZE=1024)
|
||||
target_link_libraries(stack_line_reader string_view)
|
||||
##------------------------------------------------------------------------------
|
||||
add_library(stack_line_reader_for_test ../src/stack_line_reader.c)
|
||||
target_compile_definitions(stack_line_reader_for_test PUBLIC STACK_LINE_READER_BUFFER_SIZE=16)
|
||||
target_link_libraries(stack_line_reader_for_test string_view filesystem_for_testing)
|
||||
##------------------------------------------------------------------------------
|
||||
add_library(all_libraries ../src/stack_line_reader.c ../src/linux_features_aggregator.c)
|
||||
target_link_libraries(all_libraries hwcaps_for_testing stack_line_reader string_view)
|
||||
|
||||
#
|
||||
# tests
|
||||
#
|
||||
link_libraries(gtest gmock_main)
|
||||
|
||||
## bit_utils_test
|
||||
add_executable(bit_utils_test bit_utils_test.cc)
|
||||
target_link_libraries(bit_utils_test)
|
||||
add_test(NAME bit_utils_test COMMAND bit_utils_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## string_view_test
|
||||
add_executable(string_view_test string_view_test.cc ../src/string_view.c)
|
||||
target_link_libraries(string_view_test string_view)
|
||||
add_test(NAME string_view_test COMMAND string_view_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## stack_line_reader_test
|
||||
add_executable(stack_line_reader_test stack_line_reader_test.cc)
|
||||
target_link_libraries(stack_line_reader_test stack_line_reader_for_test)
|
||||
add_test(NAME stack_line_reader_test COMMAND stack_line_reader_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## linux_features_aggregator_test
|
||||
add_executable(linux_features_aggregator_test linux_features_aggregator_test.cc)
|
||||
target_link_libraries(linux_features_aggregator_test all_libraries)
|
||||
add_test(NAME linux_features_aggregator_test COMMAND linux_features_aggregator_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## cpuinfo_x86_test
|
||||
add_executable(cpuinfo_x86_test cpuinfo_x86_test.cc ../src/cpuinfo_x86.c)
|
||||
target_link_libraries(cpuinfo_x86_test all_libraries)
|
||||
add_test(NAME cpuinfo_x86_test COMMAND cpuinfo_x86_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## cpuinfo_arm_test
|
||||
add_executable(cpuinfo_arm_test cpuinfo_arm_test.cc ../src/cpuinfo_arm.c)
|
||||
target_link_libraries(cpuinfo_arm_test all_libraries)
|
||||
add_test(NAME cpuinfo_arm_test COMMAND cpuinfo_arm_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## cpuinfo_aarch64_test
|
||||
add_executable(cpuinfo_aarch64_test cpuinfo_aarch64_test.cc ../src/cpuinfo_aarch64.c)
|
||||
target_link_libraries(cpuinfo_aarch64_test all_libraries)
|
||||
add_test(NAME cpuinfo_aarch64_test COMMAND cpuinfo_aarch64_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## cpuinfo_mips_test
|
||||
add_executable(cpuinfo_mips_test cpuinfo_mips_test.cc ../src/cpuinfo_mips.c)
|
||||
target_link_libraries(cpuinfo_mips_test all_libraries)
|
||||
add_test(NAME cpuinfo_mips_test COMMAND cpuinfo_mips_test)
|
||||
##------------------------------------------------------------------------------
|
||||
## cpuinfo_ppc_test
|
||||
add_executable(cpuinfo_ppc_test cpuinfo_ppc_test.cc ../src/cpuinfo_ppc.c)
|
||||
target_link_libraries(cpuinfo_ppc_test all_libraries)
|
||||
add_test(NAME cpuinfo_ppc_test COMMAND cpuinfo_ppc_test)
|
||||
53
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/bit_utils_test.cc
Executable file
53
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/bit_utils_test.cc
Executable file
|
|
@ -0,0 +1,53 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/bit_utils.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
namespace {
|
||||
|
||||
TEST(UtilsTest, IsBitSet) {
|
||||
for (size_t bit_set = 0; bit_set < 32; ++bit_set) {
|
||||
const uint32_t value = 1UL << bit_set;
|
||||
for (size_t i = 0; i < 32; ++i) {
|
||||
EXPECT_EQ(IsBitSet(value, i), i == bit_set);
|
||||
}
|
||||
}
|
||||
|
||||
// testing 0, all bits should be 0.
|
||||
for (size_t i = 0; i < 32; ++i) {
|
||||
EXPECT_FALSE(IsBitSet(0, i));
|
||||
}
|
||||
|
||||
// testing ~0, all bits should be 1.
|
||||
for (size_t i = 0; i < 32; ++i) {
|
||||
EXPECT_TRUE(IsBitSet(-1, i));
|
||||
}
|
||||
}
|
||||
|
||||
TEST(UtilsTest, ExtractBitRange) {
|
||||
// Extracting all bits gives the same number.
|
||||
EXPECT_EQ(ExtractBitRange(123, 31, 0), 123);
|
||||
// Extracting 1 bit gives parity.
|
||||
EXPECT_EQ(ExtractBitRange(123, 0, 0), 1);
|
||||
EXPECT_EQ(ExtractBitRange(122, 0, 0), 0);
|
||||
|
||||
EXPECT_EQ(ExtractBitRange(0xF0, 7, 4), 0xF);
|
||||
EXPECT_EQ(ExtractBitRange(0x42 << 2, 10, 2), 0x42);
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
74
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_aarch64_test.cc
Executable file
74
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_aarch64_test.cc
Executable file
|
|
@ -0,0 +1,74 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_aarch64.h"
|
||||
#include "filesystem_for_testing.h"
|
||||
#include "hwcaps_for_testing.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
namespace {
|
||||
|
||||
void DisableHardwareCapabilities() { SetHardwareCapabilities(0, 0); }
|
||||
|
||||
TEST(CpuinfoAarch64Test, FromHardwareCap) {
|
||||
SetHardwareCapabilities(AARCH64_HWCAP_FP | AARCH64_HWCAP_AES, 0);
|
||||
GetEmptyFilesystem(); // disabling /proc/cpuinfo
|
||||
const auto info = GetAarch64Info();
|
||||
EXPECT_TRUE(info.features.fp);
|
||||
EXPECT_FALSE(info.features.asimd);
|
||||
EXPECT_TRUE(info.features.aes);
|
||||
EXPECT_FALSE(info.features.pmull);
|
||||
EXPECT_FALSE(info.features.sha1);
|
||||
EXPECT_FALSE(info.features.sha2);
|
||||
EXPECT_FALSE(info.features.crc32);
|
||||
}
|
||||
|
||||
TEST(CpuinfoAarch64Test, ARMCortexA53) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(Processor : AArch64 Processor rev 3 (aarch64)
|
||||
processor : 0
|
||||
processor : 1
|
||||
processor : 2
|
||||
processor : 3
|
||||
processor : 4
|
||||
processor : 5
|
||||
processor : 6
|
||||
processor : 7
|
||||
Features : fp asimd evtstrm aes pmull sha1 sha2 crc32
|
||||
CPU implementer : 0x41
|
||||
CPU architecture: AArch64
|
||||
CPU variant : 0x0
|
||||
CPU part : 0xd03
|
||||
CPU revision : 3)");
|
||||
const auto info = GetAarch64Info();
|
||||
EXPECT_EQ(info.implementer, 0x41);
|
||||
EXPECT_EQ(info.variant, 0x0);
|
||||
EXPECT_EQ(info.part, 0xd03);
|
||||
EXPECT_EQ(info.revision, 3);
|
||||
|
||||
EXPECT_TRUE(info.features.fp);
|
||||
EXPECT_TRUE(info.features.asimd);
|
||||
EXPECT_TRUE(info.features.aes);
|
||||
EXPECT_TRUE(info.features.pmull);
|
||||
EXPECT_TRUE(info.features.sha1);
|
||||
EXPECT_TRUE(info.features.sha2);
|
||||
EXPECT_TRUE(info.features.crc32);
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
182
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_arm_test.cc
Executable file
182
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_arm_test.cc
Executable file
|
|
@ -0,0 +1,182 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_arm.h"
|
||||
#include "filesystem_for_testing.h"
|
||||
#include "hwcaps_for_testing.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
namespace {
|
||||
|
||||
void DisableHardwareCapabilities() { SetHardwareCapabilities(0, 0); }
|
||||
|
||||
TEST(CpuinfoArmTest, FromHardwareCap) {
|
||||
SetHardwareCapabilities(ARM_HWCAP_NEON, ARM_HWCAP2_AES | ARM_HWCAP2_CRC32);
|
||||
GetEmptyFilesystem(); // disabling /proc/cpuinfo
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_TRUE(info.features.vfp); // triggered by vfpv3
|
||||
EXPECT_TRUE(info.features.vfpv3); // triggered by neon
|
||||
EXPECT_TRUE(info.features.neon);
|
||||
EXPECT_TRUE(info.features.aes);
|
||||
EXPECT_TRUE(info.features.crc32);
|
||||
|
||||
EXPECT_FALSE(info.features.vfpv4);
|
||||
EXPECT_FALSE(info.features.iwmmxt);
|
||||
EXPECT_FALSE(info.features.vfpv3d16);
|
||||
EXPECT_FALSE(info.features.idiva);
|
||||
EXPECT_FALSE(info.features.idivt);
|
||||
EXPECT_FALSE(info.features.pmull);
|
||||
EXPECT_FALSE(info.features.sha1);
|
||||
EXPECT_FALSE(info.features.sha2);
|
||||
}
|
||||
|
||||
TEST(CpuinfoArmTest, ODroidFromCpuInfo) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo", R"(processor : 0
|
||||
model name : ARMv7 Processor rev 3 (v71)
|
||||
BogoMIPS : 120.00
|
||||
Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae
|
||||
CPU implementer : 0x41
|
||||
CPU architecture: 7
|
||||
CPU variant : 0x2
|
||||
CPU part : 0xc0f
|
||||
CPU revision : 3)");
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_EQ(info.implementer, 0x41);
|
||||
EXPECT_EQ(info.variant, 0x2);
|
||||
EXPECT_EQ(info.part, 0xc0f);
|
||||
EXPECT_EQ(info.revision, 3);
|
||||
EXPECT_EQ(info.architecture, 7);
|
||||
|
||||
EXPECT_TRUE(info.features.vfp);
|
||||
EXPECT_FALSE(info.features.iwmmxt);
|
||||
EXPECT_TRUE(info.features.neon);
|
||||
EXPECT_TRUE(info.features.vfpv3);
|
||||
EXPECT_FALSE(info.features.vfpv3d16);
|
||||
EXPECT_TRUE(info.features.vfpv4);
|
||||
EXPECT_TRUE(info.features.idiva);
|
||||
EXPECT_TRUE(info.features.idivt);
|
||||
EXPECT_FALSE(info.features.aes);
|
||||
EXPECT_FALSE(info.features.pmull);
|
||||
EXPECT_FALSE(info.features.sha1);
|
||||
EXPECT_FALSE(info.features.sha2);
|
||||
EXPECT_FALSE(info.features.crc32);
|
||||
}
|
||||
|
||||
// http://code.google.com/p/android/issues/detail?id=10812
|
||||
TEST(CpuinfoArmTest, InvalidArmv7) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(Processor : ARMv6-compatible processor rev 6 (v6l)
|
||||
BogoMIPS : 199.47
|
||||
Features : swp half thumb fastmult vfp edsp java
|
||||
CPU implementer : 0x41
|
||||
CPU architecture: 7
|
||||
CPU variant : 0x0
|
||||
CPU part : 0xb76
|
||||
CPU revision : 6
|
||||
|
||||
Hardware : SPICA
|
||||
Revision : 0020
|
||||
Serial : 33323613546d00ec )");
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_EQ(info.architecture, 6);
|
||||
}
|
||||
|
||||
// https://crbug.com/341598.
|
||||
TEST(CpuinfoArmTest, InvalidNeon) {
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(Processor: ARMv7 Processory rev 0 (v71)
|
||||
processor: 0
|
||||
BogoMIPS: 13.50
|
||||
|
||||
Processor: 1
|
||||
BogoMIPS: 13.50
|
||||
|
||||
Features: swp half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt
|
||||
CPU implementer : 0x51
|
||||
CPU architecture: 7
|
||||
CPU variant: 0x1
|
||||
CPU part: 0x04d
|
||||
CPU revision: 0
|
||||
|
||||
Hardware: SAMSUNG M2
|
||||
Revision: 0010
|
||||
Serial: 00001e030000354e)");
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_FALSE(info.features.neon);
|
||||
}
|
||||
|
||||
// The Nexus 4 (Qualcomm Krait) kernel configuration forgets to report IDIV
|
||||
// support.
|
||||
TEST(CpuinfoArmTest, Nexus4_0x510006f2) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(CPU implementer : 0x51
|
||||
CPU architecture: 7
|
||||
CPU variant : 0x0
|
||||
CPU part : 0x6f
|
||||
CPU revision : 2)");
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_TRUE(info.features.idiva);
|
||||
EXPECT_TRUE(info.features.idivt);
|
||||
}
|
||||
|
||||
// The Nexus 4 (Qualcomm Krait) kernel configuration forgets to report IDIV
|
||||
// support.
|
||||
TEST(CpuinfoArmTest, Nexus4_0x510006f3) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(CPU implementer : 0x51
|
||||
CPU architecture: 7
|
||||
CPU variant : 0x0
|
||||
CPU part : 0x6f
|
||||
CPU revision : 3)");
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_TRUE(info.features.idiva);
|
||||
EXPECT_TRUE(info.features.idivt);
|
||||
}
|
||||
|
||||
// The emulator-specific Android 4.2 kernel fails to report support for the
|
||||
// 32-bit ARM IDIV instruction. Technically, this is a feature of the virtual
|
||||
// CPU implemented by the emulator.
|
||||
TEST(CpuinfoArmTest, EmulatorSpecificIdiv) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(Processor : ARMv7 Processor rev 0 (v7l)
|
||||
BogoMIPS : 629.14
|
||||
Features : swp half thumb fastmult vfp edsp neon vfpv3
|
||||
CPU implementer : 0x41
|
||||
CPU architecture: 7
|
||||
CPU variant : 0x0
|
||||
CPU part : 0xc08
|
||||
CPU revision : 0
|
||||
|
||||
Hardware : Goldfish
|
||||
Revision : 0000
|
||||
Serial : 0000000000000000)");
|
||||
const auto info = GetArmInfo();
|
||||
EXPECT_TRUE(info.features.idiva);
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
125
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_mips_test.cc
Executable file
125
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_mips_test.cc
Executable file
|
|
@ -0,0 +1,125 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_mips.h"
|
||||
#include "filesystem_for_testing.h"
|
||||
#include "hwcaps_for_testing.h"
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
namespace {
|
||||
|
||||
void DisableHardwareCapabilities() { SetHardwareCapabilities(0, 0); }
|
||||
|
||||
TEST(CpuinfoMipsTest, FromHardwareCapBoth) {
|
||||
SetHardwareCapabilities(MIPS_HWCAP_EVA | MIPS_HWCAP_MSA, 0);
|
||||
GetEmptyFilesystem(); // disabling /proc/cpuinfo
|
||||
const auto info = GetMipsInfo();
|
||||
EXPECT_TRUE(info.features.msa);
|
||||
EXPECT_TRUE(info.features.eva);
|
||||
}
|
||||
|
||||
TEST(CpuinfoMipsTest, FromHardwareCapOnlyOne) {
|
||||
SetHardwareCapabilities(MIPS_HWCAP_MSA, 0);
|
||||
GetEmptyFilesystem(); // disabling /proc/cpuinfo
|
||||
const auto info = GetMipsInfo();
|
||||
EXPECT_TRUE(info.features.msa);
|
||||
EXPECT_FALSE(info.features.eva);
|
||||
}
|
||||
|
||||
TEST(CpuinfoMipsTest, Ci40) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo", R"(system type : IMG Pistachio SoC (B0)
|
||||
machine : IMG Marduk – Ci40 with cc2520
|
||||
processor : 0
|
||||
cpu model : MIPS interAptiv (multi) V2.0 FPU V0.0
|
||||
BogoMIPS : 363.72
|
||||
wait instruction : yes
|
||||
microsecond timers : yes
|
||||
tlb_entries : 64
|
||||
extra interrupt vector : yes
|
||||
hardware watchpoint : yes, count: 4, address/irw mask: [0x0ffc, 0x0ffc, 0x0ffb, 0x0ffb]
|
||||
isa : mips1 mips2 mips32r1 mips32r2
|
||||
ASEs implemented : mips16 dsp mt eva
|
||||
shadow register sets : 1
|
||||
kscratch registers : 0
|
||||
package : 0
|
||||
core : 0
|
||||
VCED exceptions : not available
|
||||
VCEI exceptions : not available
|
||||
VPE : 0
|
||||
)");
|
||||
const auto info = GetMipsInfo();
|
||||
EXPECT_FALSE(info.features.msa);
|
||||
EXPECT_TRUE(info.features.eva);
|
||||
}
|
||||
|
||||
TEST(CpuinfoMipsTest, AR7161) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(system type : Atheros AR7161 rev 2
|
||||
machine : NETGEAR WNDR3700/WNDR3800/WNDRMAC
|
||||
processor : 0
|
||||
cpu model : MIPS 24Kc V7.4
|
||||
BogoMIPS : 452.19
|
||||
wait instruction : yes
|
||||
microsecond timers : yes
|
||||
tlb_entries : 16
|
||||
extra interrupt vector : yes
|
||||
hardware watchpoint : yes, count: 4, address/irw mask: [0x0000, 0x0f98, 0x0f78, 0x0df8]
|
||||
ASEs implemented : mips16
|
||||
shadow register sets : 1
|
||||
kscratch registers : 0
|
||||
core : 0
|
||||
VCED exceptions : not available
|
||||
VCEI exceptions : not available
|
||||
)");
|
||||
const auto info = GetMipsInfo();
|
||||
EXPECT_FALSE(info.features.msa);
|
||||
EXPECT_FALSE(info.features.eva);
|
||||
}
|
||||
|
||||
TEST(CpuinfoMipsTest, Goldfish) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo", R"(system type : MIPS-Goldfish
|
||||
Hardware : goldfish
|
||||
Revison : 1
|
||||
processor : 0
|
||||
cpu model : MIPS 24Kc V0.0 FPU V0.0
|
||||
BogoMIPS : 1042.02
|
||||
wait instruction : yes
|
||||
microsecond timers : yes
|
||||
tlb_entries : 16
|
||||
extra interrupt vector : yes
|
||||
hardware watchpoint : yes, count: 1, address/irw mask: [0x0ff8]
|
||||
ASEs implemented :
|
||||
shadow register sets : 1
|
||||
core : 0
|
||||
VCED exceptions : not available
|
||||
VCEI exceptions : not available
|
||||
)");
|
||||
const auto info = GetMipsInfo();
|
||||
EXPECT_FALSE(info.features.msa);
|
||||
EXPECT_FALSE(info.features.eva);
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
119
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_ppc_test.cc
Executable file
119
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_ppc_test.cc
Executable file
|
|
@ -0,0 +1,119 @@
|
|||
// Copyright 2018 IBM.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "cpuinfo_ppc.h"
|
||||
#include "filesystem_for_testing.h"
|
||||
#include "hwcaps_for_testing.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
namespace {
|
||||
|
||||
void DisableHardwareCapabilities() { SetHardwareCapabilities(0, 0); }
|
||||
|
||||
TEST(CpustringsPPCTest, FromHardwareCap) {
|
||||
SetHardwareCapabilities(PPC_FEATURE_HAS_FPU | PPC_FEATURE_HAS_VSX,
|
||||
PPC_FEATURE2_ARCH_3_00);
|
||||
GetEmptyFilesystem(); // disabling /proc/cpuinfo
|
||||
const auto info = GetPPCInfo();
|
||||
EXPECT_TRUE(info.features.fpu);
|
||||
EXPECT_FALSE(info.features.mmu);
|
||||
EXPECT_TRUE(info.features.vsx);
|
||||
EXPECT_TRUE(info.features.arch300);
|
||||
EXPECT_FALSE(info.features.power4);
|
||||
EXPECT_FALSE(info.features.altivec);
|
||||
EXPECT_FALSE(info.features.vcrypto);
|
||||
EXPECT_FALSE(info.features.htm);
|
||||
}
|
||||
|
||||
TEST(CpustringsPPCTest, Blade) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(processor : 14
|
||||
cpu : POWER7 (architected), altivec supported
|
||||
clock : 3000.000000MHz
|
||||
revision : 2.1 (pvr 003f 0201)
|
||||
|
||||
processor : 15
|
||||
cpu : POWER7 (architected), altivec supported
|
||||
clock : 3000.000000MHz
|
||||
revision : 2.1 (pvr 003f 0201)
|
||||
|
||||
timebase : 512000000
|
||||
platform : pSeries
|
||||
model : IBM,8406-70Y
|
||||
machine : CHRP IBM,8406-70Y)");
|
||||
SetPlatformTypes("power7", "power8");
|
||||
const auto strings = GetPPCPlatformStrings();
|
||||
ASSERT_STREQ(strings.platform, "pSeries");
|
||||
ASSERT_STREQ(strings.model, "IBM,8406-70Y");
|
||||
ASSERT_STREQ(strings.machine, "CHRP IBM,8406-70Y");
|
||||
ASSERT_STREQ(strings.cpu, "POWER7 (architected), altivec supported");
|
||||
ASSERT_STREQ(strings.type.platform, "power7");
|
||||
ASSERT_STREQ(strings.type.base_platform, "power8");
|
||||
}
|
||||
|
||||
TEST(CpustringsPPCTest, Firestone) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(processor : 126
|
||||
cpu : POWER8 (raw), altivec supported
|
||||
clock : 2061.000000MHz
|
||||
revision : 2.0 (pvr 004d 0200)
|
||||
|
||||
processor : 127
|
||||
cpu : POWER8 (raw), altivec supported
|
||||
clock : 2061.000000MHz
|
||||
revision : 2.0 (pvr 004d 0200)
|
||||
|
||||
timebase : 512000000
|
||||
platform : PowerNV
|
||||
model : 8335-GTA
|
||||
machine : PowerNV 8335-GTA
|
||||
firmware : OPAL v3)");
|
||||
const auto strings = GetPPCPlatformStrings();
|
||||
ASSERT_STREQ(strings.platform, "PowerNV");
|
||||
ASSERT_STREQ(strings.model, "8335-GTA");
|
||||
ASSERT_STREQ(strings.machine, "PowerNV 8335-GTA");
|
||||
ASSERT_STREQ(strings.cpu, "POWER8 (raw), altivec supported");
|
||||
}
|
||||
|
||||
TEST(CpustringsPPCTest, w8) {
|
||||
DisableHardwareCapabilities();
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
fs.CreateFile("/proc/cpuinfo",
|
||||
R"(processor : 143
|
||||
cpu : POWER9, altivec supported
|
||||
clock : 2300.000000MHz
|
||||
revision : 2.2 (pvr 004e 1202)
|
||||
|
||||
timebase : 512000000
|
||||
platform : PowerNV
|
||||
model : 0000000000000000
|
||||
machine : PowerNV 0000000000000000
|
||||
firmware : OPAL
|
||||
MMU : Radix)");
|
||||
const auto strings = GetPPCPlatformStrings();
|
||||
ASSERT_STREQ(strings.platform, "PowerNV");
|
||||
ASSERT_STREQ(strings.model, "0000000000000000");
|
||||
ASSERT_STREQ(strings.machine, "PowerNV 0000000000000000");
|
||||
ASSERT_STREQ(strings.cpu, "POWER9, altivec supported");
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
172
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_x86_test.cc
Executable file
172
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/cpuinfo_x86_test.cc
Executable file
|
|
@ -0,0 +1,172 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include <cassert>
|
||||
#include <cstdio>
|
||||
#include <map>
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
#include "cpuinfo_x86.h"
|
||||
#include "internal/cpuid_x86.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
class FakeCpu {
|
||||
public:
|
||||
Leaf CpuId(uint32_t leaf_id) const {
|
||||
const auto itr = cpuid_leaves_.find(leaf_id);
|
||||
EXPECT_TRUE(itr != cpuid_leaves_.end()) << "Missing leaf " << leaf_id;
|
||||
return itr->second;
|
||||
}
|
||||
|
||||
uint32_t GetXCR0Eax() const { return xcr0_eax_; }
|
||||
|
||||
void SetLeaves(std::map<uint32_t, Leaf> configuration) {
|
||||
cpuid_leaves_ = std::move(configuration);
|
||||
}
|
||||
|
||||
void SetOsBackupsExtendedRegisters(bool os_backups_extended_registers) {
|
||||
xcr0_eax_ = os_backups_extended_registers ? -1 : 0;
|
||||
}
|
||||
|
||||
private:
|
||||
std::map<uint32_t, Leaf> cpuid_leaves_;
|
||||
uint32_t xcr0_eax_;
|
||||
};
|
||||
|
||||
auto* g_fake_cpu = new FakeCpu();
|
||||
|
||||
extern "C" Leaf CpuId(uint32_t leaf_id) { return g_fake_cpu->CpuId(leaf_id); }
|
||||
extern "C" uint32_t GetXCR0Eax(void) { return g_fake_cpu->GetXCR0Eax(); }
|
||||
|
||||
namespace {
|
||||
|
||||
TEST(CpuidX86Test, SandyBridge) {
|
||||
g_fake_cpu->SetOsBackupsExtendedRegisters(true);
|
||||
g_fake_cpu->SetLeaves({
|
||||
{0x00000000, Leaf{0x0000000D, 0x756E6547, 0x6C65746E, 0x49656E69}},
|
||||
{0x00000001, Leaf{0x000206A6, 0x00100800, 0x1F9AE3BF, 0xBFEBFBFF}},
|
||||
{0x00000007, Leaf{0x00000000, 0x00000000, 0x00000000, 0x00000000}},
|
||||
});
|
||||
const auto info = GetX86Info();
|
||||
EXPECT_STREQ(info.vendor, "GenuineIntel");
|
||||
EXPECT_EQ(info.family, 0x06);
|
||||
EXPECT_EQ(info.model, 0x02A);
|
||||
EXPECT_EQ(info.stepping, 0x06);
|
||||
// Leaf 7 is zeroed out so none of the Leaf 7 flags are set.
|
||||
const auto features = info.features;
|
||||
EXPECT_FALSE(features.erms);
|
||||
EXPECT_FALSE(features.avx2);
|
||||
EXPECT_FALSE(features.avx512f);
|
||||
EXPECT_FALSE(features.avx512cd);
|
||||
EXPECT_FALSE(features.avx512er);
|
||||
EXPECT_FALSE(features.avx512pf);
|
||||
EXPECT_FALSE(features.avx512bw);
|
||||
EXPECT_FALSE(features.avx512dq);
|
||||
EXPECT_FALSE(features.avx512vl);
|
||||
EXPECT_FALSE(features.avx512ifma);
|
||||
EXPECT_FALSE(features.avx512vbmi);
|
||||
EXPECT_FALSE(features.avx512vbmi2);
|
||||
EXPECT_FALSE(features.avx512vnni);
|
||||
EXPECT_FALSE(features.avx512bitalg);
|
||||
EXPECT_FALSE(features.avx512vpopcntdq);
|
||||
EXPECT_FALSE(features.avx512_4vnniw);
|
||||
EXPECT_FALSE(features.avx512_4vbmi2);
|
||||
// All old cpu features should be set.
|
||||
EXPECT_TRUE(features.aes);
|
||||
EXPECT_TRUE(features.ssse3);
|
||||
EXPECT_TRUE(features.sse4_1);
|
||||
EXPECT_TRUE(features.sse4_2);
|
||||
EXPECT_TRUE(features.avx);
|
||||
}
|
||||
|
||||
TEST(CpuidX86Test, SandyBridgeTestOsSupport) {
|
||||
g_fake_cpu->SetLeaves({
|
||||
{0x00000000, Leaf{0x0000000D, 0x756E6547, 0x6C65746E, 0x49656E69}},
|
||||
{0x00000001, Leaf{0x000206A6, 0x00100800, 0x1F9AE3BF, 0xBFEBFBFF}},
|
||||
{0x00000007, Leaf{0x00000000, 0x00000000, 0x00000000, 0x00000000}},
|
||||
});
|
||||
// avx is disabled if os does not support backing up ymm registers.
|
||||
g_fake_cpu->SetOsBackupsExtendedRegisters(false);
|
||||
EXPECT_FALSE(GetX86Info().features.avx);
|
||||
// avx is disabled if os does not support backing up ymm registers.
|
||||
g_fake_cpu->SetOsBackupsExtendedRegisters(true);
|
||||
EXPECT_TRUE(GetX86Info().features.avx);
|
||||
}
|
||||
|
||||
TEST(CpuidX86Test, SkyLake) {
|
||||
g_fake_cpu->SetOsBackupsExtendedRegisters(true);
|
||||
g_fake_cpu->SetLeaves({
|
||||
{0x00000000, Leaf{0x00000016, 0x756E6547, 0x6C65746E, 0x49656E69}},
|
||||
{0x00000001, Leaf{0x000406E3, 0x00100800, 0x7FFAFBBF, 0xBFEBFBFF}},
|
||||
{0x00000007, Leaf{0x00000000, 0x029C67AF, 0x00000000, 0x00000000}},
|
||||
});
|
||||
const auto info = GetX86Info();
|
||||
EXPECT_STREQ(info.vendor, "GenuineIntel");
|
||||
EXPECT_EQ(info.family, 0x06);
|
||||
EXPECT_EQ(info.model, 0x04E);
|
||||
EXPECT_EQ(info.stepping, 0x03);
|
||||
EXPECT_EQ(GetX86Microarchitecture(&info), X86Microarchitecture::INTEL_SKL);
|
||||
}
|
||||
|
||||
TEST(CpuidX86Test, Branding) {
|
||||
g_fake_cpu->SetLeaves({
|
||||
{0x00000000, Leaf{0x00000016, 0x756E6547, 0x6C65746E, 0x49656E69}},
|
||||
{0x00000001, Leaf{0x000406E3, 0x00100800, 0x7FFAFBBF, 0xBFEBFBFF}},
|
||||
{0x00000007, Leaf{0x00000000, 0x029C67AF, 0x00000000, 0x00000000}},
|
||||
{0x80000000, Leaf{0x80000008, 0x00000000, 0x00000000, 0x00000000}},
|
||||
{0x80000001, Leaf{0x00000000, 0x00000000, 0x00000121, 0x2C100000}},
|
||||
{0x80000002, Leaf{0x65746E49, 0x2952286C, 0x726F4320, 0x4D542865}},
|
||||
{0x80000003, Leaf{0x37692029, 0x3035362D, 0x43205530, 0x40205550}},
|
||||
{0x80000004, Leaf{0x352E3220, 0x7A484730, 0x00000000, 0x00000000}},
|
||||
});
|
||||
char brand_string[49];
|
||||
FillX86BrandString(brand_string);
|
||||
EXPECT_STREQ(brand_string, "Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz");
|
||||
}
|
||||
|
||||
// http://users.atw.hu/instlatx64/AuthenticAMD0630F81_K15_Godavari_CPUID.txt
|
||||
TEST(CpuidX86Test, AMD_K15) {
|
||||
g_fake_cpu->SetLeaves({
|
||||
{0x00000000, Leaf{0x0000000D, 0x68747541, 0x444D4163, 0x69746E65}},
|
||||
{0x00000001, Leaf{0x00630F81, 0x00040800, 0x3E98320B, 0x178BFBFF}},
|
||||
{0x00000007, Leaf{0x00000000, 0x00000000, 0x00000000, 0x00000000}},
|
||||
{0x80000000, Leaf{0x8000001E, 0x68747541, 0x444D4163, 0x69746E65}},
|
||||
{0x80000001, Leaf{0x00630F81, 0x10000000, 0x0FEBBFFF, 0x2FD3FBFF}},
|
||||
{0x80000002, Leaf{0x20444D41, 0x372D3841, 0x4B303736, 0x64615220}},
|
||||
{0x80000003, Leaf{0x206E6F65, 0x202C3752, 0x43203031, 0x75706D6F}},
|
||||
{0x80000004, Leaf{0x43206574, 0x7365726F, 0x2B433420, 0x00204736}},
|
||||
{0x80000005, Leaf{0xFF40FF18, 0xFF40FF30, 0x10040140, 0x60030140}},
|
||||
});
|
||||
const auto info = GetX86Info();
|
||||
|
||||
EXPECT_STREQ(info.vendor, "AuthenticAMD");
|
||||
EXPECT_EQ(info.family, 0x15);
|
||||
EXPECT_EQ(info.model, 0x38);
|
||||
EXPECT_EQ(info.stepping, 0x01);
|
||||
EXPECT_EQ(GetX86Microarchitecture(&info),
|
||||
X86Microarchitecture::AMD_BULLDOZER);
|
||||
|
||||
char brand_string[49];
|
||||
FillX86BrandString(brand_string);
|
||||
EXPECT_STREQ(brand_string, "AMD A8-7670K Radeon R7, 10 Compute Cores 4C+6G ");
|
||||
}
|
||||
|
||||
// TODO(user): test what happens when xsave/osxsave are not present.
|
||||
// TODO(user): test what happens when xmm/ymm/zmm os support are not
|
||||
// present.
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
103
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/filesystem_for_testing.cc
Executable file
103
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/filesystem_for_testing.cc
Executable file
|
|
@ -0,0 +1,103 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "filesystem_for_testing.h"
|
||||
|
||||
#include <cassert>
|
||||
#include <climits>
|
||||
#include <cstdio>
|
||||
#include <cstring>
|
||||
#include <utility>
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
FakeFile::FakeFile(int file_descriptor, const char* content)
|
||||
: file_descriptor_(file_descriptor), content_(content) {}
|
||||
|
||||
FakeFile::~FakeFile() { assert(!opened_); }
|
||||
|
||||
void FakeFile::Open() {
|
||||
assert(!opened_);
|
||||
opened_ = true;
|
||||
}
|
||||
|
||||
void FakeFile::Close() {
|
||||
assert(opened_);
|
||||
opened_ = false;
|
||||
}
|
||||
|
||||
int FakeFile::Read(int fd, void* buf, size_t count) {
|
||||
assert(count < INT_MAX);
|
||||
assert(fd == file_descriptor_);
|
||||
const size_t remainder = content_.size() - head_index_;
|
||||
const size_t read = count > remainder ? remainder : count;
|
||||
memcpy(buf, content_.data() + head_index_, read);
|
||||
head_index_ += read;
|
||||
assert(read < INT_MAX);
|
||||
return read;
|
||||
}
|
||||
|
||||
void FakeFilesystem::Reset() { files_.clear(); }
|
||||
|
||||
FakeFile* FakeFilesystem::CreateFile(const std::string& filename,
|
||||
const char* content) {
|
||||
auto& file = files_[filename];
|
||||
file =
|
||||
std::unique_ptr<FakeFile>(new FakeFile(next_file_descriptor_++, content));
|
||||
return file.get();
|
||||
}
|
||||
|
||||
FakeFile* FakeFilesystem::FindFileOrNull(const std::string& filename) const {
|
||||
const auto itr = files_.find(filename);
|
||||
return itr == files_.end() ? nullptr : itr->second.get();
|
||||
}
|
||||
|
||||
FakeFile* FakeFilesystem::FindFileOrDie(const int file_descriptor) const {
|
||||
for (const auto& filename_file_pair : files_) {
|
||||
FakeFile* const file_ptr = filename_file_pair.second.get();
|
||||
if (file_ptr->GetFileDescriptor() == file_descriptor) {
|
||||
return file_ptr;
|
||||
}
|
||||
}
|
||||
assert(false);
|
||||
return nullptr;
|
||||
}
|
||||
|
||||
static FakeFilesystem* kFilesystem = new FakeFilesystem();
|
||||
|
||||
FakeFilesystem& GetEmptyFilesystem() {
|
||||
kFilesystem->Reset();
|
||||
return *kFilesystem;
|
||||
}
|
||||
|
||||
extern "C" int CpuFeatures_OpenFile(const char* filename) {
|
||||
auto* const file = kFilesystem->FindFileOrNull(filename);
|
||||
if (file) {
|
||||
file->Open();
|
||||
return file->GetFileDescriptor();
|
||||
}
|
||||
return -1;
|
||||
}
|
||||
|
||||
extern "C" void CpuFeatures_CloseFile(int file_descriptor) {
|
||||
kFilesystem->FindFileOrDie(file_descriptor)->Close();
|
||||
}
|
||||
|
||||
extern "C" int CpuFeatures_ReadFile(int file_descriptor, void* buffer,
|
||||
size_t buffer_size) {
|
||||
return kFilesystem->FindFileOrDie(file_descriptor)
|
||||
->Read(file_descriptor, buffer, buffer_size);
|
||||
}
|
||||
|
||||
} // namespace cpu_features
|
||||
61
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/filesystem_for_testing.h
Executable file
61
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/filesystem_for_testing.h
Executable file
|
|
@ -0,0 +1,61 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
// Implements a fake filesystem, useful for tests.
|
||||
#ifndef CPU_FEATURES_TEST_FILESYSTEM_FOR_TESTING_H_
|
||||
#define CPU_FEATURES_TEST_FILESYSTEM_FOR_TESTING_H_
|
||||
|
||||
#include <memory>
|
||||
#include <string>
|
||||
#include <unordered_map>
|
||||
|
||||
#include "internal/filesystem.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
class FakeFile {
|
||||
public:
|
||||
explicit FakeFile(int file_descriptor, const char* content);
|
||||
~FakeFile();
|
||||
|
||||
void Open();
|
||||
void Close();
|
||||
int Read(int fd, void* buf, size_t count);
|
||||
|
||||
int GetFileDescriptor() const { return file_descriptor_; }
|
||||
|
||||
private:
|
||||
const int file_descriptor_;
|
||||
const std::string content_;
|
||||
bool opened_ = false;
|
||||
size_t head_index_ = 0;
|
||||
};
|
||||
|
||||
class FakeFilesystem {
|
||||
public:
|
||||
void Reset();
|
||||
FakeFile* CreateFile(const std::string& filename, const char* content);
|
||||
FakeFile* FindFileOrDie(const int file_descriptor) const;
|
||||
FakeFile* FindFileOrNull(const std::string& filename) const;
|
||||
|
||||
private:
|
||||
size_t next_file_descriptor_ = 0;
|
||||
std::unordered_map<std::string, std::unique_ptr<FakeFile>> files_;
|
||||
};
|
||||
|
||||
FakeFilesystem& GetEmptyFilesystem();
|
||||
|
||||
} // namespace cpu_features
|
||||
|
||||
#endif // CPU_FEATURES_TEST_FILESYSTEM_FOR_TESTING_H_
|
||||
45
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/hwcaps_for_testing.cc
Executable file
45
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/hwcaps_for_testing.cc
Executable file
|
|
@ -0,0 +1,45 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include <string.h>
|
||||
|
||||
#include "hwcaps_for_testing.h"
|
||||
#include "internal/string_view.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
namespace {
|
||||
static auto* const g_hardware_capabilities = new HardwareCapabilities();
|
||||
static auto* const g_platform_types = new PlatformType();
|
||||
} // namespace
|
||||
|
||||
void SetHardwareCapabilities(uint32_t hwcaps, uint32_t hwcaps2) {
|
||||
g_hardware_capabilities->hwcaps = hwcaps;
|
||||
g_hardware_capabilities->hwcaps2 = hwcaps2;
|
||||
}
|
||||
|
||||
HardwareCapabilities CpuFeatures_GetHardwareCapabilities(void) {
|
||||
return *g_hardware_capabilities;
|
||||
}
|
||||
|
||||
void SetPlatformTypes(const char* platform, const char* base_platform) {
|
||||
CpuFeatures_StringView_CopyString(str(platform), g_platform_types->platform,
|
||||
sizeof(g_platform_types->platform));
|
||||
CpuFeatures_StringView_CopyString(str(base_platform),
|
||||
g_platform_types->base_platform,
|
||||
sizeof(g_platform_types->base_platform));
|
||||
}
|
||||
|
||||
PlatformType CpuFeatures_GetPlatformType(void) { return *g_platform_types; }
|
||||
} // namespace cpu_features
|
||||
27
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/hwcaps_for_testing.h
Executable file
27
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/hwcaps_for_testing.h
Executable file
|
|
@ -0,0 +1,27 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#ifndef CPU_FEATURES_TEST_HWCAPS_FOR_TESTING_H_
|
||||
#define CPU_FEATURES_TEST_HWCAPS_FOR_TESTING_H_
|
||||
|
||||
#include "internal/hwcaps.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
void SetHardwareCapabilities(uint32_t hwcaps, uint32_t hwcaps2);
|
||||
void SetPlatformTypes(const char *platform, const char *base_platform);
|
||||
|
||||
} // namespace cpu_features
|
||||
|
||||
#endif // CPU_FEATURES_TEST_HWCAPS_FOR_TESTING_H_
|
||||
|
|
@ -0,0 +1,95 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include <array>
|
||||
|
||||
#include "internal/linux_features_aggregator.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
namespace {
|
||||
|
||||
struct Features {
|
||||
bool a = false;
|
||||
bool b = false;
|
||||
bool c = false;
|
||||
};
|
||||
|
||||
DECLARE_SETTER(Features, a)
|
||||
DECLARE_SETTER(Features, b)
|
||||
DECLARE_SETTER(Features, c)
|
||||
|
||||
class LinuxFeatureAggregatorTest : public testing::Test {
|
||||
public:
|
||||
const std::array<CapabilityConfig, 3> kConfigs = {
|
||||
{{{0b0001, 0b0000}, "a", &set_a},
|
||||
{{0b0010, 0b0000}, "b", &set_b},
|
||||
{{0b0000, 0b1100}, "c", &set_c}}};
|
||||
};
|
||||
|
||||
TEST_F(LinuxFeatureAggregatorTest, FromFlagsEmpty) {
|
||||
Features features;
|
||||
CpuFeatures_SetFromFlags(kConfigs.size(), kConfigs.data(), str(""),
|
||||
&features);
|
||||
EXPECT_FALSE(features.a);
|
||||
EXPECT_FALSE(features.b);
|
||||
EXPECT_FALSE(features.c);
|
||||
}
|
||||
|
||||
TEST_F(LinuxFeatureAggregatorTest, FromFlagsAllSet) {
|
||||
Features features;
|
||||
CpuFeatures_SetFromFlags(kConfigs.size(), kConfigs.data(), str("a c b"),
|
||||
&features);
|
||||
EXPECT_TRUE(features.a);
|
||||
EXPECT_TRUE(features.b);
|
||||
EXPECT_TRUE(features.c);
|
||||
}
|
||||
|
||||
TEST_F(LinuxFeatureAggregatorTest, FromFlagsOnlyA) {
|
||||
Features features;
|
||||
CpuFeatures_SetFromFlags(kConfigs.size(), kConfigs.data(), str("a"),
|
||||
&features);
|
||||
EXPECT_TRUE(features.a);
|
||||
EXPECT_FALSE(features.b);
|
||||
EXPECT_FALSE(features.c);
|
||||
}
|
||||
|
||||
TEST_F(LinuxFeatureAggregatorTest, FromHwcapsNone) {
|
||||
HardwareCapabilities capability;
|
||||
capability.hwcaps = 0; // matches none
|
||||
capability.hwcaps2 = 0; // matches none
|
||||
Features features;
|
||||
CpuFeatures_OverrideFromHwCaps(kConfigs.size(), kConfigs.data(), capability,
|
||||
&features);
|
||||
EXPECT_FALSE(features.a);
|
||||
EXPECT_FALSE(features.b);
|
||||
EXPECT_FALSE(features.c);
|
||||
}
|
||||
|
||||
TEST_F(LinuxFeatureAggregatorTest, FromHwcapsSet) {
|
||||
HardwareCapabilities capability;
|
||||
capability.hwcaps = 0b0010; // matches b but not a
|
||||
capability.hwcaps2 = 0b1111; // matches c
|
||||
Features features;
|
||||
CpuFeatures_OverrideFromHwCaps(kConfigs.size(), kConfigs.data(), capability,
|
||||
&features);
|
||||
EXPECT_FALSE(features.a);
|
||||
EXPECT_TRUE(features.b);
|
||||
EXPECT_TRUE(features.c);
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
132
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/stack_line_reader_test.cc
Executable file
132
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/stack_line_reader_test.cc
Executable file
|
|
@ -0,0 +1,132 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/stack_line_reader.h"
|
||||
#include "filesystem_for_testing.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
bool operator==(const StringView& a, const StringView& b) {
|
||||
return CpuFeatures_StringView_IsEquals(a, b);
|
||||
}
|
||||
|
||||
namespace {
|
||||
|
||||
std::string ToString(StringView view) { return {view.ptr, view.size}; }
|
||||
|
||||
TEST(StackLineReaderTest, Empty) {
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
auto* file = fs.CreateFile("/proc/cpuinfo", "");
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, file->GetFileDescriptor());
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_TRUE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str(""));
|
||||
}
|
||||
}
|
||||
|
||||
TEST(StackLineReaderTest, ManySmallLines) {
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
auto* file = fs.CreateFile("/proc/cpuinfo", "a\nb\nc");
|
||||
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, file->GetFileDescriptor());
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("a"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("b"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_TRUE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("c"));
|
||||
}
|
||||
}
|
||||
|
||||
TEST(StackLineReaderTest, TruncatedLine) {
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
auto* file = fs.CreateFile("/proc/cpuinfo", R"(First
|
||||
Second
|
||||
More than 16 characters, this will be truncated.
|
||||
last)");
|
||||
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, file->GetFileDescriptor());
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("First"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("Second"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_FALSE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("More than 16 cha"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_TRUE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("last"));
|
||||
}
|
||||
}
|
||||
|
||||
TEST(StackLineReaderTest, TruncatedLines) {
|
||||
auto& fs = GetEmptyFilesystem();
|
||||
auto* file = fs.CreateFile("/proc/cpuinfo", R"(More than 16 characters
|
||||
Another line that is too long)");
|
||||
|
||||
StackLineReader reader;
|
||||
StackLineReader_Initialize(&reader, file->GetFileDescriptor());
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_FALSE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("More than 16 cha"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_FALSE(result.eof);
|
||||
EXPECT_FALSE(result.full_line);
|
||||
EXPECT_EQ(result.line, str("Another line tha"));
|
||||
}
|
||||
{
|
||||
const auto result = StackLineReader_NextLine(&reader);
|
||||
EXPECT_TRUE(result.eof);
|
||||
EXPECT_TRUE(result.full_line);
|
||||
EXPECT_EQ(result.line, str(""));
|
||||
}
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
144
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/string_view_test.cc
Executable file
144
src/crypto/argon2_hasher/hash/cpu/cpu_features/test/string_view_test.cc
Executable file
|
|
@ -0,0 +1,144 @@
|
|||
// Copyright 2017 Google Inc.
|
||||
//
|
||||
// Licensed under the Apache License, Version 2.0 (the "License");
|
||||
// you may not use this file except in compliance with the License.
|
||||
// You may obtain a copy of the License at
|
||||
//
|
||||
// http://www.apache.org/licenses/LICENSE-2.0
|
||||
//
|
||||
// Unless required by applicable law or agreed to in writing, software
|
||||
// distributed under the License is distributed on an "AS IS" BASIS,
|
||||
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
// See the License for the specific language governing permissions and
|
||||
// limitations under the License.
|
||||
|
||||
#include "internal/string_view.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
namespace cpu_features {
|
||||
|
||||
bool operator==(const StringView& a, const StringView& b) {
|
||||
return CpuFeatures_StringView_IsEquals(a, b);
|
||||
}
|
||||
|
||||
namespace {
|
||||
|
||||
TEST(StringViewTest, Empty) {
|
||||
EXPECT_EQ(kEmptyStringView.ptr, nullptr);
|
||||
EXPECT_EQ(kEmptyStringView.size, 0);
|
||||
}
|
||||
|
||||
TEST(StringViewTest, Build) {
|
||||
const auto view = str("test");
|
||||
EXPECT_EQ(view.ptr[0], 't');
|
||||
EXPECT_EQ(view.size, 4);
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_IndexOfChar) {
|
||||
// Found.
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOfChar(str("test"), 'e'), 1);
|
||||
// Not found.
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOfChar(str("test"), 'z'), -1);
|
||||
// Empty.
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOfChar(kEmptyStringView, 'z'), -1);
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_IndexOf) {
|
||||
// Found.
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOf(str("test"), str("es")), 1);
|
||||
// Not found.
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOf(str("test"), str("aa")), -1);
|
||||
// Empty.
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOf(kEmptyStringView, str("aa")), -1);
|
||||
EXPECT_EQ(CpuFeatures_StringView_IndexOf(str("aa"), kEmptyStringView), -1);
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_StartsWith) {
|
||||
EXPECT_TRUE(CpuFeatures_StringView_StartsWith(str("test"), str("te")));
|
||||
EXPECT_FALSE(CpuFeatures_StringView_StartsWith(str("test"), str("")));
|
||||
EXPECT_FALSE(
|
||||
CpuFeatures_StringView_StartsWith(str("test"), kEmptyStringView));
|
||||
EXPECT_FALSE(
|
||||
CpuFeatures_StringView_StartsWith(kEmptyStringView, str("test")));
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_IsEquals) {
|
||||
EXPECT_TRUE(
|
||||
CpuFeatures_StringView_IsEquals(kEmptyStringView, kEmptyStringView));
|
||||
EXPECT_TRUE(CpuFeatures_StringView_IsEquals(kEmptyStringView, str("")));
|
||||
EXPECT_TRUE(CpuFeatures_StringView_IsEquals(str(""), kEmptyStringView));
|
||||
EXPECT_TRUE(CpuFeatures_StringView_IsEquals(str("a"), str("a")));
|
||||
EXPECT_FALSE(CpuFeatures_StringView_IsEquals(str("a"), str("b")));
|
||||
EXPECT_FALSE(CpuFeatures_StringView_IsEquals(str("a"), kEmptyStringView));
|
||||
EXPECT_FALSE(CpuFeatures_StringView_IsEquals(kEmptyStringView, str("a")));
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_PopFront) {
|
||||
EXPECT_EQ(CpuFeatures_StringView_PopFront(str("test"), 2), str("st"));
|
||||
EXPECT_EQ(CpuFeatures_StringView_PopFront(str("test"), 0), str("test"));
|
||||
EXPECT_EQ(CpuFeatures_StringView_PopFront(str("test"), 4), str(""));
|
||||
EXPECT_EQ(CpuFeatures_StringView_PopFront(str("test"), 100), str(""));
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_ParsePositiveNumber) {
|
||||
EXPECT_EQ(CpuFeatures_StringView_ParsePositiveNumber(str("42")), 42);
|
||||
EXPECT_EQ(CpuFeatures_StringView_ParsePositiveNumber(str("0x2a")), 42);
|
||||
EXPECT_EQ(CpuFeatures_StringView_ParsePositiveNumber(str("0x2A")), 42);
|
||||
|
||||
EXPECT_EQ(CpuFeatures_StringView_ParsePositiveNumber(str("-0x2A")), -1);
|
||||
EXPECT_EQ(CpuFeatures_StringView_ParsePositiveNumber(str("abc")), -1);
|
||||
EXPECT_EQ(CpuFeatures_StringView_ParsePositiveNumber(str("")), -1);
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_CopyString) {
|
||||
char buf[4];
|
||||
buf[0] = 'X';
|
||||
|
||||
// Empty
|
||||
CpuFeatures_StringView_CopyString(str(""), buf, sizeof(buf));
|
||||
EXPECT_STREQ(buf, "");
|
||||
|
||||
// Less
|
||||
CpuFeatures_StringView_CopyString(str("a"), buf, sizeof(buf));
|
||||
EXPECT_STREQ(buf, "a");
|
||||
|
||||
// exact
|
||||
CpuFeatures_StringView_CopyString(str("abc"), buf, sizeof(buf));
|
||||
EXPECT_STREQ(buf, "abc");
|
||||
|
||||
// More
|
||||
CpuFeatures_StringView_CopyString(str("abcd"), buf, sizeof(buf));
|
||||
EXPECT_STREQ(buf, "abc");
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_HasWord) {
|
||||
// Find flags at beginning, middle and end.
|
||||
EXPECT_TRUE(
|
||||
CpuFeatures_StringView_HasWord(str("first middle last"), "first"));
|
||||
EXPECT_TRUE(
|
||||
CpuFeatures_StringView_HasWord(str("first middle last"), "middle"));
|
||||
EXPECT_TRUE(CpuFeatures_StringView_HasWord(str("first middle last"), "last"));
|
||||
// Do not match partial flags
|
||||
EXPECT_FALSE(
|
||||
CpuFeatures_StringView_HasWord(str("first middle last"), "irst"));
|
||||
EXPECT_FALSE(CpuFeatures_StringView_HasWord(str("first middle last"), "mid"));
|
||||
EXPECT_FALSE(CpuFeatures_StringView_HasWord(str("first middle last"), "las"));
|
||||
}
|
||||
|
||||
TEST(StringViewTest, CpuFeatures_StringView_GetAttributeKeyValue) {
|
||||
const StringView line = str(" key : first middle last ");
|
||||
StringView key, value;
|
||||
EXPECT_TRUE(CpuFeatures_StringView_GetAttributeKeyValue(line, &key, &value));
|
||||
EXPECT_EQ(key, str("key"));
|
||||
EXPECT_EQ(value, str("first middle last"));
|
||||
}
|
||||
|
||||
TEST(StringViewTest, FailingGetAttributeKeyValue) {
|
||||
const StringView line = str("key first middle last");
|
||||
StringView key, value;
|
||||
EXPECT_FALSE(CpuFeatures_StringView_GetAttributeKeyValue(line, &key, &value));
|
||||
}
|
||||
|
||||
} // namespace
|
||||
} // namespace cpu_features
|
||||
353
src/crypto/argon2_hasher/hash/gpu/cuda/blake2b.cu
Normal file
353
src/crypto/argon2_hasher/hash/gpu/cuda/blake2b.cu
Normal file
|
|
@ -0,0 +1,353 @@
|
|||
#define BLOCK_BYTES 32
|
||||
#define OUT_BYTES 16
|
||||
#define BLAKE_SHARED_MEM 480
|
||||
#define BLAKE_SHARED_MEM_UINT 120
|
||||
|
||||
#define G(m, r, i, a, b, c, d) \
|
||||
do { \
|
||||
a = a + b + m[blake2b_sigma[r][2 * i + 0]]; \
|
||||
d = rotr64(d ^ a, 32); \
|
||||
c = c + d; \
|
||||
b = rotr64(b ^ c, 24); \
|
||||
a = a + b + m[blake2b_sigma[r][2 * i + 1]]; \
|
||||
d = rotr64(d ^ a, 16); \
|
||||
c = c + d; \
|
||||
b = rotr64(b ^ c, 63); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define G_S(m, a, b, c, d) \
|
||||
do { \
|
||||
a = a + b + m; \
|
||||
d = rotr64(d ^ a, 32); \
|
||||
c = c + d; \
|
||||
b = rotr64(b ^ c, 24); \
|
||||
a = a + b + m; \
|
||||
d = rotr64(d ^ a, 16); \
|
||||
c = c + d; \
|
||||
b = rotr64(b ^ c, 63); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define ROUND(m, t, r) \
|
||||
do { \
|
||||
G(m, r, t, v0, v1, v2, v3); \
|
||||
v1 = __shfl_sync(0xFFFFFFFF, v1, t + 1, 4); \
|
||||
v2 = __shfl_sync(0xFFFFFFFF, v2, t + 2, 4); \
|
||||
v3 = __shfl_sync(0xFFFFFFFF, v3, t + 3, 4); \
|
||||
G(m, r, (t + 4), v0, v1, v2, v3); \
|
||||
v1 = __shfl_sync(0xFFFFFFFF, v1, t + 3, 4); \
|
||||
v2 = __shfl_sync(0xFFFFFFFF, v2, t + 2, 4); \
|
||||
v3 = __shfl_sync(0xFFFFFFFF, v3, t + 1, 4); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
#define ROUND_S(m, t) \
|
||||
do { \
|
||||
G_S(m, v0, v1, v2, v3); \
|
||||
v1 = __shfl_sync(0xFFFFFFFF, v1, t + 1, 4); \
|
||||
v2 = __shfl_sync(0xFFFFFFFF, v2, t + 2, 4); \
|
||||
v3 = __shfl_sync(0xFFFFFFFF, v3, t + 3, 4); \
|
||||
G_S(m, v0, v1, v2, v3); \
|
||||
v1 = __shfl_sync(0xFFFFFFFF, v1, t + 3, 4); \
|
||||
v2 = __shfl_sync(0xFFFFFFFF, v2, t + 2, 4); \
|
||||
v3 = __shfl_sync(0xFFFFFFFF, v3, t + 1, 4); \
|
||||
} while ((void)0, 0)
|
||||
|
||||
__constant__ uint64_t blake2b_IV[8] = {
|
||||
0x6A09E667F3BCC908, 0xBB67AE8584CAA73B,
|
||||
0x3C6EF372FE94F82B, 0xA54FF53A5F1D36F1,
|
||||
0x510E527FADE682D1, 0x9B05688C2B3E6C1F,
|
||||
0x1F83D9ABFB41BD6B, 0x5BE0CD19137E2179
|
||||
};
|
||||
|
||||
__constant__ uint32_t blake2b_sigma[12][16] = {
|
||||
{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15},
|
||||
{14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3},
|
||||
{11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4},
|
||||
{7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8},
|
||||
{9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13},
|
||||
{2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9},
|
||||
{12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11},
|
||||
{13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10},
|
||||
{6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5},
|
||||
{10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13, 0},
|
||||
{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15},
|
||||
{14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3},
|
||||
};
|
||||
|
||||
__device__ uint64_t rotr64(uint64_t x, uint32_t n)
|
||||
{
|
||||
return (x >> n) | (x << (64 - n));
|
||||
}
|
||||
|
||||
__device__ __forceinline__ void blake2b_compress(uint64_t *h, uint64_t *m, uint64_t f0, int thr_id)
|
||||
{
|
||||
uint64_t v0, v1, v2, v3;
|
||||
|
||||
v0 = h[thr_id];
|
||||
v1 = h[thr_id + 4];
|
||||
v2 = blake2b_IV[thr_id];
|
||||
v3 = blake2b_IV[thr_id + 4];
|
||||
|
||||
if(thr_id == 0) v3 ^= h[8];
|
||||
if(thr_id == 1) v3 ^= h[9];
|
||||
if(thr_id == 2) v3 ^= f0;
|
||||
|
||||
ROUND(m, thr_id, 0);
|
||||
ROUND(m, thr_id, 1);
|
||||
ROUND(m, thr_id, 2);
|
||||
ROUND(m, thr_id, 3);
|
||||
ROUND(m, thr_id, 4);
|
||||
ROUND(m, thr_id, 5);
|
||||
ROUND(m, thr_id, 6);
|
||||
ROUND(m, thr_id, 7);
|
||||
ROUND(m, thr_id, 8);
|
||||
ROUND(m, thr_id, 9);
|
||||
ROUND(m, thr_id, 10);
|
||||
ROUND(m, thr_id, 11);
|
||||
|
||||
h[thr_id] ^= v0 ^ v2;
|
||||
h[thr_id + 4] ^= v1 ^ v3;
|
||||
}
|
||||
|
||||
__device__ __forceinline__ void blake2b_compress_static(uint64_t *h, uint64_t m, uint64_t f0, int thr_id)
|
||||
{
|
||||
uint64_t v0, v1, v2, v3;
|
||||
|
||||
v0 = h[thr_id];
|
||||
v1 = h[thr_id + 4];
|
||||
v2 = blake2b_IV[thr_id];
|
||||
v3 = blake2b_IV[thr_id + 4];
|
||||
|
||||
if(thr_id == 0) v3 ^= h[8];
|
||||
if(thr_id == 1) v3 ^= h[9];
|
||||
if(thr_id == 2) v3 ^= f0;
|
||||
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
ROUND_S(m, thr_id);
|
||||
|
||||
h[thr_id] ^= v0 ^ v2;
|
||||
h[thr_id + 4] ^= v1 ^ v3;
|
||||
}
|
||||
|
||||
__device__ __forceinline__ int blake2b_init(uint64_t *h, int out_len, int thr_id)
|
||||
{
|
||||
h[thr_id * 2] = blake2b_IV[thr_id * 2];
|
||||
h[thr_id * 2 + 1] = blake2b_IV[thr_id * 2 + 1];
|
||||
|
||||
if(thr_id == 0) {
|
||||
h[8] = h[9] = 0;
|
||||
h[0] = 0x6A09E667F3BCC908 ^ ((out_len * 4) | (1 << 16) | (1 << 24));
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
__device__ __forceinline__ void blake2b_incrementCounter(uint64_t *h, int inc)
|
||||
{
|
||||
h[8] += (inc * 4);
|
||||
h[9] += (h[8] < (inc * 4));
|
||||
}
|
||||
|
||||
__device__ __forceinline__ int blake2b_update(uint32_t *in, int in_len, uint64_t *h, uint32_t *buf, int buf_len, int thr_id)
|
||||
{
|
||||
uint32_t *cursor_in = in;
|
||||
uint32_t *cursor_out = buf + buf_len;
|
||||
|
||||
if (buf_len + in_len > BLOCK_BYTES) {
|
||||
int left = BLOCK_BYTES - buf_len;
|
||||
|
||||
for(int i=0; i < (left >> 2); i++, cursor_in += 4, cursor_out += 4) {
|
||||
cursor_out[thr_id] = cursor_in[thr_id];
|
||||
}
|
||||
|
||||
if(thr_id == 0) {
|
||||
for (int i = 0; i < (left % 4); i++) {
|
||||
cursor_out[i] = cursor_in[i];
|
||||
}
|
||||
blake2b_incrementCounter(h, BLOCK_BYTES);
|
||||
}
|
||||
|
||||
blake2b_compress(h, (uint64_t*)buf, 0, thr_id);
|
||||
|
||||
buf_len = 0;
|
||||
|
||||
in_len -= left;
|
||||
in += left;
|
||||
|
||||
while (in_len > BLOCK_BYTES) {
|
||||
if(thr_id == 0)
|
||||
blake2b_incrementCounter(h, BLOCK_BYTES);
|
||||
|
||||
cursor_in = in;
|
||||
cursor_out = buf;
|
||||
|
||||
for(int i=0; i < (BLOCK_BYTES / 4); i++, cursor_in += 4, cursor_out += 4) {
|
||||
cursor_out[thr_id] = cursor_in[thr_id];
|
||||
}
|
||||
|
||||
blake2b_compress(h, (uint64_t *)buf, 0, thr_id);
|
||||
|
||||
in_len -= BLOCK_BYTES;
|
||||
in += BLOCK_BYTES;
|
||||
}
|
||||
}
|
||||
|
||||
cursor_in = in;
|
||||
cursor_out = buf + buf_len;
|
||||
|
||||
for(int i=0; i < (in_len >> 2); i++, cursor_in += 4, cursor_out += 4) {
|
||||
cursor_out[thr_id] = cursor_in[thr_id];
|
||||
}
|
||||
|
||||
if(thr_id == 0) {
|
||||
for (int i = 0; i < (in_len % 4); i++) {
|
||||
cursor_out[i] = cursor_in[i];
|
||||
}
|
||||
}
|
||||
|
||||
return buf_len + in_len;
|
||||
}
|
||||
|
||||
__device__ __forceinline__ int blake2b_update_static(uint32_t in, int in_len, uint64_t *h, uint32_t *buf, int buf_len, int thr_id)
|
||||
{
|
||||
uint64_t in64 = in;
|
||||
in64 = in64 << 32;
|
||||
in64 = in64 | in;
|
||||
|
||||
uint32_t *cursor_out = buf + buf_len;
|
||||
|
||||
if (buf_len + in_len > BLOCK_BYTES) {
|
||||
int left = BLOCK_BYTES - buf_len;
|
||||
|
||||
for(int i=0; i < (left >> 2); i++, cursor_out += 4) {
|
||||
cursor_out[thr_id] = in;
|
||||
}
|
||||
|
||||
if(thr_id == 0) {
|
||||
for (int i = 0; i < (left % 4); i++) {
|
||||
cursor_out[i] = in;
|
||||
}
|
||||
blake2b_incrementCounter(h, BLOCK_BYTES);
|
||||
}
|
||||
|
||||
blake2b_compress(h, (uint64_t*)buf, 0, thr_id);
|
||||
|
||||
buf_len = 0;
|
||||
|
||||
in_len -= left;
|
||||
|
||||
while (in_len > BLOCK_BYTES) {
|
||||
if(thr_id == 0)
|
||||
blake2b_incrementCounter(h, BLOCK_BYTES);
|
||||
|
||||
blake2b_compress_static(h, in64, 0, thr_id);
|
||||
|
||||
in_len -= BLOCK_BYTES;
|
||||
}
|
||||
}
|
||||
|
||||
cursor_out = buf + buf_len;
|
||||
|
||||
for(int i=0; i < (in_len >> 2); i++, cursor_out += 4) {
|
||||
cursor_out[thr_id] = in;
|
||||
}
|
||||
|
||||
if(thr_id == 0) {
|
||||
for (int i = 0; i < (in_len % 4); i++) {
|
||||
cursor_out[i] = in;
|
||||
}
|
||||
}
|
||||
|
||||
return buf_len + in_len;
|
||||
}
|
||||
|
||||
__device__ __forceinline__ void blake2b_final(uint32_t *out, int out_len, uint64_t *h, uint32_t *buf, int buf_len, int thr_id)
|
||||
{
|
||||
int left = BLOCK_BYTES - buf_len;
|
||||
uint32_t *cursor_out = buf + buf_len;
|
||||
|
||||
for(int i=0; i < (left >> 2); i++, cursor_out += 4) {
|
||||
cursor_out[thr_id] = 0;
|
||||
}
|
||||
|
||||
if(thr_id == 0) {
|
||||
for (int i = 0; i < (left % 4); i++) {
|
||||
cursor_out[i] = 0;
|
||||
}
|
||||
blake2b_incrementCounter(h, buf_len);
|
||||
}
|
||||
|
||||
blake2b_compress(h, (uint64_t*)buf, 0xFFFFFFFFFFFFFFFF, thr_id);
|
||||
|
||||
uint32_t *cursor_in = (uint32_t *)h;
|
||||
cursor_out = out;
|
||||
|
||||
for(int i=0; i < (out_len >> 2); i++, cursor_in += 4, cursor_out += 4) {
|
||||
cursor_out[thr_id] = cursor_in[thr_id];
|
||||
}
|
||||
|
||||
if(thr_id == 0) {
|
||||
for (int i = 0; i < (out_len % 4); i++) {
|
||||
cursor_out[i] = cursor_in[i];
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
__device__ void blake2b_digestLong(uint32_t *out, int out_len, uint32_t *in, int in_len, int thr_id, uint32_t *shared)
|
||||
{
|
||||
uint64_t *h = (uint64_t*)shared;
|
||||
uint32_t *buf = (uint32_t*)&h[10];
|
||||
uint32_t *out_buffer = &buf[32];
|
||||
int buf_len;
|
||||
|
||||
if(thr_id == 0) buf[0] = (out_len * 4);
|
||||
buf_len = 1;
|
||||
|
||||
if (out_len <= OUT_BYTES) {
|
||||
blake2b_init(h, out_len, thr_id);
|
||||
buf_len = blake2b_update(in, in_len, h, buf, buf_len, thr_id);
|
||||
blake2b_final(out, out_len, h, buf, buf_len, thr_id);
|
||||
} else {
|
||||
uint32_t *cursor_in = out_buffer;
|
||||
uint32_t *cursor_out = out;
|
||||
|
||||
blake2b_init(h, OUT_BYTES, thr_id);
|
||||
buf_len = blake2b_update(in, in_len, h, buf, buf_len, thr_id);
|
||||
blake2b_final(out_buffer, OUT_BYTES, h, buf, buf_len, thr_id);
|
||||
|
||||
for(int i=0; i < (OUT_BYTES / 8); i++, cursor_in += 4, cursor_out += 4) {
|
||||
cursor_out[thr_id] = cursor_in[thr_id];
|
||||
}
|
||||
|
||||
out += OUT_BYTES / 2;
|
||||
|
||||
int to_produce = out_len - OUT_BYTES / 2;
|
||||
while (to_produce > OUT_BYTES) {
|
||||
buf_len = blake2b_init(h, OUT_BYTES, thr_id);
|
||||
buf_len = blake2b_update(out_buffer, OUT_BYTES, h, buf, buf_len, thr_id);
|
||||
blake2b_final(out_buffer, OUT_BYTES, h, buf, buf_len, thr_id);
|
||||
|
||||
cursor_out = out;
|
||||
cursor_in = out_buffer;
|
||||
for(int i=0; i < (OUT_BYTES / 8); i++, cursor_in += 4, cursor_out += 4) {
|
||||
cursor_out[thr_id] = cursor_in[thr_id];
|
||||
}
|
||||
|
||||
out += OUT_BYTES / 2;
|
||||
to_produce -= OUT_BYTES / 2;
|
||||
}
|
||||
|
||||
buf_len = blake2b_init(h, to_produce, thr_id);
|
||||
buf_len = blake2b_update(out_buffer, OUT_BYTES, h, buf, buf_len, thr_id);
|
||||
blake2b_final(out, to_produce, h, buf, buf_len, thr_id);
|
||||
}
|
||||
}
|
||||
340
src/crypto/argon2_hasher/hash/gpu/cuda/cuda_hasher.cpp
Normal file
340
src/crypto/argon2_hasher/hash/gpu/cuda/cuda_hasher.cpp
Normal file
|
|
@ -0,0 +1,340 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#include <crypto/Argon2_constants.h>
|
||||
|
||||
#include "../../../common/common.h"
|
||||
|
||||
#include "crypto/argon2_hasher/hash/Hasher.h"
|
||||
#include "crypto/argon2_hasher/hash/argon2/Argon2.h"
|
||||
|
||||
#if defined(WITH_CUDA)
|
||||
|
||||
#include <cuda_runtime.h>
|
||||
#include <driver_types.h>
|
||||
|
||||
#include "cuda_hasher.h"
|
||||
#include "../../../common/DLLExport.h"
|
||||
|
||||
cuda_hasher::cuda_hasher() {
|
||||
m_type = "GPU";
|
||||
m_subType = "CUDA";
|
||||
m_shortSubType = "NVD";
|
||||
m_intensity = 0;
|
||||
m_description = "";
|
||||
m_computingThreads = 0;
|
||||
}
|
||||
|
||||
|
||||
cuda_hasher::~cuda_hasher() {
|
||||
this->cleanup();
|
||||
}
|
||||
|
||||
bool cuda_hasher::initialize(xmrig::Algo algorithm, xmrig::Variant variant) {
|
||||
cudaError_t error = cudaSuccess;
|
||||
string error_message;
|
||||
|
||||
m_profile = getArgon2Profile(algorithm, variant);
|
||||
|
||||
__devices = __query_cuda_devices(error, error_message);
|
||||
|
||||
if(error != cudaSuccess) {
|
||||
m_description = "No compatible GPU detected: " + error_message;
|
||||
return false;
|
||||
}
|
||||
|
||||
if (__devices.empty()) {
|
||||
m_description = "No compatible GPU detected.";
|
||||
return false;
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
vector<cuda_device_info *> cuda_hasher::__query_cuda_devices(cudaError_t &error, string &error_message) {
|
||||
vector<cuda_device_info *> devices;
|
||||
int devCount = 0;
|
||||
error = cudaGetDeviceCount(&devCount);
|
||||
|
||||
if(error != cudaSuccess) {
|
||||
error_message = "Error querying CUDA device count.";
|
||||
return devices;
|
||||
}
|
||||
|
||||
if(devCount == 0)
|
||||
return devices;
|
||||
|
||||
for (int i = 0; i < devCount; ++i)
|
||||
{
|
||||
cuda_device_info *dev = __get_device_info(i);
|
||||
if(dev == NULL)
|
||||
continue;
|
||||
if(dev->error != cudaSuccess) {
|
||||
error = dev->error;
|
||||
error_message = dev->error_message;
|
||||
continue;
|
||||
}
|
||||
devices.push_back(dev);
|
||||
}
|
||||
return devices;
|
||||
}
|
||||
|
||||
cuda_device_info *cuda_hasher::__get_device_info(int device_index) {
|
||||
cuda_device_info *device_info = new cuda_device_info();
|
||||
device_info->error = cudaSuccess;
|
||||
device_info->cuda_index = device_index;
|
||||
|
||||
device_info->error = cudaSetDevice(device_index);
|
||||
if(device_info->error != cudaSuccess) {
|
||||
device_info->error_message = "Error setting current device.";
|
||||
return device_info;
|
||||
}
|
||||
|
||||
cudaDeviceProp devProp;
|
||||
device_info->error = cudaGetDeviceProperties(&devProp, device_index);
|
||||
if(device_info->error != cudaSuccess) {
|
||||
device_info->error_message = "Error setting current device.";
|
||||
return device_info;
|
||||
}
|
||||
|
||||
device_info->device_string = devProp.name;
|
||||
|
||||
size_t freemem, totalmem;
|
||||
device_info->error = cudaMemGetInfo(&freemem, &totalmem);
|
||||
if(device_info->error != cudaSuccess) {
|
||||
device_info->error_message = "Error setting current device.";
|
||||
return device_info;
|
||||
}
|
||||
|
||||
device_info->free_mem_size = freemem;
|
||||
device_info->max_allocable_mem_size = freemem / 4;
|
||||
|
||||
double mem_in_gb = totalmem / 1073741824.0;
|
||||
stringstream ss;
|
||||
ss << setprecision(2) << mem_in_gb;
|
||||
device_info->device_string += (" (" + ss.str() + "GB)");
|
||||
|
||||
return device_info;
|
||||
}
|
||||
|
||||
bool cuda_hasher::configure(xmrig::HasherConfig &config) {
|
||||
int index = config.getGPUCardsCount();
|
||||
double intensity = 0;
|
||||
|
||||
int total_threads = 0;
|
||||
intensity = config.getAverageGPUIntensity();
|
||||
|
||||
if (intensity == 0) {
|
||||
m_intensity = 0;
|
||||
m_description = "Status: DISABLED - by user.";
|
||||
return false;
|
||||
}
|
||||
|
||||
bool cards_selected = false;
|
||||
intensity = 0;
|
||||
|
||||
for(vector<cuda_device_info *>::iterator d = __devices.begin(); d != __devices.end(); d++, index++) {
|
||||
stringstream ss;
|
||||
ss << "["<< (index + 1) << "] " << (*d)->device_string;
|
||||
string device_description = ss.str();
|
||||
(*d)->device_index = index;
|
||||
(*d)->profile_info.profile = m_profile;
|
||||
|
||||
if(config.gpuFilter().size() > 0) {
|
||||
bool found = false;
|
||||
for(xmrig::GPUFilter fit : config.gpuFilter()) {
|
||||
if(device_description.find(fit.filter) != string::npos) {
|
||||
found = true;
|
||||
break;
|
||||
}
|
||||
}
|
||||
if(!found) {
|
||||
(*d)->profile_info.threads = 0;
|
||||
ss << " - DISABLED" << endl;
|
||||
m_description += ss.str();
|
||||
continue;
|
||||
}
|
||||
else {
|
||||
cards_selected = true;
|
||||
}
|
||||
}
|
||||
else {
|
||||
cards_selected = true;
|
||||
}
|
||||
|
||||
ss << endl;
|
||||
|
||||
double device_intensity = config.getGPUIntensity((*d)->device_index);
|
||||
|
||||
m_description += ss.str();
|
||||
|
||||
if(!(__setup_device_info((*d), device_intensity))) {
|
||||
m_description += (*d)->error_message;
|
||||
m_description += "\n";
|
||||
continue;
|
||||
};
|
||||
|
||||
DeviceInfo device;
|
||||
|
||||
char bus_id[100];
|
||||
if(cudaDeviceGetPCIBusId(bus_id, 100, (*d)->cuda_index) == cudaSuccess) {
|
||||
device.bus_id = bus_id;
|
||||
int domain_separator = device.bus_id.find(":");
|
||||
if(domain_separator != string::npos) {
|
||||
device.bus_id.erase(0, domain_separator + 1);
|
||||
}
|
||||
}
|
||||
|
||||
device.name = (*d)->device_string;
|
||||
device.intensity = device_intensity;
|
||||
storeDeviceInfo((*d)->device_index, device);
|
||||
|
||||
__enabledDevices.push_back(*d);
|
||||
|
||||
total_threads += (*d)->profile_info.threads;
|
||||
intensity += device_intensity;
|
||||
}
|
||||
|
||||
config.addGPUCardsCount(index - config.getGPUCardsCount());
|
||||
|
||||
if(!cards_selected) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - no card enabled because of filtering.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (total_threads == 0) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - not enough resources.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if(!buildThreadData())
|
||||
return false;
|
||||
|
||||
m_intensity = intensity / __enabledDevices.size();
|
||||
m_computingThreads = __enabledDevices.size() * 2; // 2 computing threads for each device
|
||||
m_description += "Status: ENABLED - with " + to_string(total_threads) + " threads.";
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
void cuda_hasher::cleanup() {
|
||||
for(vector<cuda_device_info *>::iterator d = __devices.begin(); d != __devices.end(); d++) {
|
||||
cuda_free(*d);
|
||||
}
|
||||
}
|
||||
|
||||
bool cuda_hasher::__setup_device_info(cuda_device_info *device, double intensity) {
|
||||
device->profile_info.threads_per_chunk = (uint32_t)(device->max_allocable_mem_size / device->profile_info.profile->memSize);
|
||||
size_t chunk_size = device->profile_info.threads_per_chunk * device->profile_info.profile->memSize;
|
||||
|
||||
if(chunk_size == 0) {
|
||||
device->error = cudaErrorInitializationError;
|
||||
device->error_message = "Not enough memory on GPU.";
|
||||
return false;
|
||||
}
|
||||
|
||||
uint64_t usable_memory = device->free_mem_size;
|
||||
double chunks = (double)usable_memory / (double)chunk_size;
|
||||
|
||||
uint32_t max_threads = (uint32_t)(device->profile_info.threads_per_chunk * chunks);
|
||||
|
||||
if(max_threads == 0) {
|
||||
device->error = cudaErrorInitializationError;
|
||||
device->error_message = "Not enough memory on GPU.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->profile_info.threads = (uint32_t)(max_threads * intensity / 100.0);
|
||||
device->profile_info.threads = (device->profile_info.threads / 2) * 2; // make it divisible by 2 to allow for parallel kernel execution
|
||||
if(max_threads > 0 && device->profile_info.threads == 0 && intensity > 0)
|
||||
device->profile_info.threads = 2;
|
||||
|
||||
chunks = (double)device->profile_info.threads / (double)device->profile_info.threads_per_chunk;
|
||||
|
||||
cuda_allocate(device, chunks, chunk_size);
|
||||
|
||||
if(device->error != cudaSuccess)
|
||||
return false;
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
bool cuda_hasher::buildThreadData() {
|
||||
__thread_data = new cuda_gpumgmt_thread_data[__enabledDevices.size() * 2];
|
||||
|
||||
for(int i=0; i < __enabledDevices.size(); i++) {
|
||||
cuda_device_info *device = __enabledDevices[i];
|
||||
for(int threadId = 0; threadId < 2; threadId ++) {
|
||||
cuda_gpumgmt_thread_data &thread_data = __thread_data[i * 2 + threadId];
|
||||
thread_data.device = device;
|
||||
thread_data.thread_id = threadId;
|
||||
|
||||
cudaStream_t stream;
|
||||
device->error = cudaStreamCreate(&stream);
|
||||
if(device->error != cudaSuccess) {
|
||||
LOG("Error running kernel: (" + to_string(device->error) + ") cannot create cuda stream.");
|
||||
return false;
|
||||
}
|
||||
|
||||
thread_data.device_data = stream;
|
||||
|
||||
#ifdef PARALLEL_CUDA
|
||||
if(threadId == 0) {
|
||||
thread_data.threads_idx = 0;
|
||||
thread_data.threads = device->profile_info.threads / 2;
|
||||
}
|
||||
else {
|
||||
thread_data.threads_idx = device->profile_info.threads / 2;
|
||||
thread_data.threads = device->profile_info.threads - thread_data.threads_idx;
|
||||
}
|
||||
#else
|
||||
thread_data.threads_idx = 0;
|
||||
thread_data.threads = device->profile_info.threads;
|
||||
#endif
|
||||
|
||||
thread_data.argon2 = new Argon2(cuda_kernel_prehasher, cuda_kernel_filler, cuda_kernel_posthasher,
|
||||
nullptr, &thread_data);
|
||||
thread_data.argon2->setThreads(thread_data.threads);
|
||||
thread_data.hashData.outSize = xmrig::ARGON2_HASHLEN + 4;
|
||||
}
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
int cuda_hasher::compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output) {
|
||||
cuda_gpumgmt_thread_data &threadData = __thread_data[threadIdx];
|
||||
|
||||
cudaSetDevice(threadData.device->cuda_index);
|
||||
|
||||
threadData.hashData.input = input;
|
||||
threadData.hashData.inSize = size;
|
||||
threadData.hashData.output = output;
|
||||
int hashCount = threadData.argon2->generateHashes(*m_profile, threadData.hashData);
|
||||
if(threadData.device->error != cudaSuccess) {
|
||||
LOG("Error running kernel: (" + to_string(threadData.device->error) + ")" + threadData.device->error_message);
|
||||
return 0;
|
||||
}
|
||||
|
||||
uint32_t *nonce = ((uint32_t *)(((uint8_t*)threadData.hashData.input) + 39));
|
||||
(*nonce) += threadData.threads;
|
||||
|
||||
return hashCount;
|
||||
|
||||
}
|
||||
|
||||
size_t cuda_hasher::parallelism(int workerIdx) {
|
||||
cuda_gpumgmt_thread_data &threadData = __thread_data[workerIdx];
|
||||
return threadData.threads;
|
||||
}
|
||||
|
||||
size_t cuda_hasher::deviceCount() {
|
||||
return __enabledDevices.size();
|
||||
}
|
||||
|
||||
REGISTER_HASHER(cuda_hasher);
|
||||
|
||||
#endif //WITH_CUDA
|
||||
126
src/crypto/argon2_hasher/hash/gpu/cuda/cuda_hasher.h
Normal file
126
src/crypto/argon2_hasher/hash/gpu/cuda/cuda_hasher.h
Normal file
|
|
@ -0,0 +1,126 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 18/09/2018.
|
||||
//
|
||||
|
||||
#ifndef ARGON2_CUDA_HASHER_H
|
||||
#define ARGON2_CUDA_HASHER_H
|
||||
|
||||
#if defined(WITH_CUDA)
|
||||
|
||||
struct cuda_kernel_arguments {
|
||||
void *memory_chunk_0;
|
||||
void *memory_chunk_1;
|
||||
void *memory_chunk_2;
|
||||
void *memory_chunk_3;
|
||||
void *memory_chunk_4;
|
||||
void *memory_chunk_5;
|
||||
|
||||
uint32_t *refs;
|
||||
uint32_t *idxs;
|
||||
uint32_t *segments;
|
||||
|
||||
uint32_t *preseed_memory[2];
|
||||
uint32_t *seed_memory[2];
|
||||
uint32_t *out_memory[2];
|
||||
uint32_t *hash_memory[2];
|
||||
|
||||
uint32_t *host_seed_memory[2];
|
||||
};
|
||||
|
||||
struct argon2profile_info {
|
||||
argon2profile_info() {
|
||||
threads = 0;
|
||||
threads_per_chunk = 0;
|
||||
}
|
||||
uint32_t threads;
|
||||
uint32_t threads_per_chunk;
|
||||
Argon2Profile *profile;
|
||||
};
|
||||
|
||||
struct cuda_device_info {
|
||||
cuda_device_info() {
|
||||
device_index = 0;
|
||||
device_string = "";
|
||||
free_mem_size = 0;
|
||||
max_allocable_mem_size = 0;
|
||||
|
||||
error = cudaSuccess;
|
||||
error_message = "";
|
||||
}
|
||||
|
||||
int device_index;
|
||||
int cuda_index;
|
||||
|
||||
string device_string;
|
||||
uint64_t free_mem_size;
|
||||
uint64_t max_allocable_mem_size;
|
||||
|
||||
argon2profile_info profile_info;
|
||||
cuda_kernel_arguments arguments;
|
||||
|
||||
mutex device_lock;
|
||||
|
||||
cudaError_t error;
|
||||
string error_message;
|
||||
};
|
||||
|
||||
struct cuda_gpumgmt_thread_data {
|
||||
void lock() {
|
||||
#ifndef PARALLEL_CUDA
|
||||
device->device_lock.lock();
|
||||
#endif
|
||||
}
|
||||
|
||||
void unlock() {
|
||||
#ifndef PARALLEL_CUDA
|
||||
device->device_lock.unlock();
|
||||
#endif
|
||||
}
|
||||
|
||||
int thread_id;
|
||||
cuda_device_info *device;
|
||||
Argon2 *argon2;
|
||||
HashData hashData;
|
||||
|
||||
void *device_data;
|
||||
|
||||
int threads;
|
||||
int threads_idx;
|
||||
};
|
||||
|
||||
class cuda_hasher : public Hasher {
|
||||
public:
|
||||
cuda_hasher();
|
||||
~cuda_hasher();
|
||||
|
||||
virtual bool initialize(xmrig::Algo algorithm, xmrig::Variant variant);
|
||||
virtual bool configure(xmrig::HasherConfig &config);
|
||||
virtual void cleanup();
|
||||
virtual int compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output);
|
||||
virtual size_t parallelism(int workerIdx);
|
||||
virtual size_t deviceCount();
|
||||
|
||||
private:
|
||||
cuda_device_info *__get_device_info(int device_index);
|
||||
bool __setup_device_info(cuda_device_info *device, double intensity);
|
||||
vector<cuda_device_info*> __query_cuda_devices(cudaError_t &error, string &error_message);
|
||||
bool buildThreadData();
|
||||
|
||||
vector<cuda_device_info*> __devices;
|
||||
vector<cuda_device_info*> __enabledDevices;
|
||||
cuda_gpumgmt_thread_data *__thread_data;
|
||||
|
||||
Argon2Profile *m_profile;
|
||||
};
|
||||
|
||||
// CUDA kernel exports
|
||||
extern void cuda_allocate(cuda_device_info *device, double chunks, size_t chunk_size);
|
||||
extern void cuda_free(cuda_device_info *device);
|
||||
extern bool cuda_kernel_prehasher(void *memory, int threads, Argon2Profile *profile, void *user_data);
|
||||
extern void *cuda_kernel_filler(int threads, Argon2Profile *profile, void *user_data);
|
||||
extern bool cuda_kernel_posthasher(void *memory, int threads, Argon2Profile *profile, void *user_data);
|
||||
// end CUDA kernel exports
|
||||
|
||||
#endif //WITH_CUDA
|
||||
|
||||
#endif //ARGON2_CUDA_HASHER_H
|
||||
1132
src/crypto/argon2_hasher/hash/gpu/cuda/cuda_kernel.cu
Normal file
1132
src/crypto/argon2_hasher/hash/gpu/cuda/cuda_kernel.cu
Normal file
File diff suppressed because it is too large
Load diff
888
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLHasher.cpp
Executable file
888
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLHasher.cpp
Executable file
|
|
@ -0,0 +1,888 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#include <crypto/Argon2_constants.h>
|
||||
#include "../../../common/common.h"
|
||||
|
||||
#include "crypto/argon2_hasher/hash/Hasher.h"
|
||||
#include "crypto/argon2_hasher/hash/argon2/Argon2.h"
|
||||
|
||||
#include "OpenCLHasher.h"
|
||||
#include "OpenCLKernel.h"
|
||||
|
||||
#include "crypto/argon2_hasher/common/DLLExport.h"
|
||||
|
||||
#if defined(WITH_OPENCL)
|
||||
|
||||
#ifndef CL_DEVICE_BOARD_NAME_AMD
|
||||
#define CL_DEVICE_BOARD_NAME_AMD 0x4038
|
||||
#endif
|
||||
#ifndef CL_DEVICE_TOPOLOGY_AMD
|
||||
#define CL_DEVICE_TOPOLOGY_AMD 0x4037
|
||||
#endif
|
||||
#ifndef CL_DEVICE_PCI_BUS_ID_NV
|
||||
#define CL_DEVICE_PCI_BUS_ID_NV 0x4008
|
||||
#endif
|
||||
#ifndef CL_DEVICE_PCI_SLOT_ID_NV
|
||||
#define CL_DEVICE_PCI_SLOT_ID_NV 0x4009
|
||||
#endif
|
||||
|
||||
typedef union
|
||||
{
|
||||
struct { cl_uint type; cl_uint data[5]; } raw;
|
||||
struct { cl_uint type; cl_char unused[17]; cl_char bus; cl_char device; cl_char function; } pcie;
|
||||
} device_topology_amd;
|
||||
|
||||
#define KERNEL_WORKGROUP_SIZE 32
|
||||
|
||||
opencl_hasher::opencl_hasher() {
|
||||
m_type = "GPU";
|
||||
m_subType = "OPENCL";
|
||||
m_shortSubType = "OCL";
|
||||
m_intensity = 0;
|
||||
m_description = "";
|
||||
m_computingThreads = 0;
|
||||
}
|
||||
|
||||
opencl_hasher::~opencl_hasher() {
|
||||
// this->cleanup();
|
||||
}
|
||||
|
||||
bool opencl_hasher::initialize(xmrig::Algo algorithm, xmrig::Variant variant) {
|
||||
cl_int error = CL_SUCCESS;
|
||||
string error_message;
|
||||
|
||||
m_profile = getArgon2Profile(algorithm, variant);
|
||||
|
||||
__devices = __query_opencl_devices(error, error_message);
|
||||
if(error != CL_SUCCESS) {
|
||||
m_description = "No compatible GPU detected: " + error_message;
|
||||
return false;
|
||||
}
|
||||
|
||||
if (__devices.empty()) {
|
||||
m_description = "No compatible GPU detected.";
|
||||
return false;
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
vector<opencl_device_info*> opencl_hasher::__query_opencl_devices(cl_int &error, string &error_message) {
|
||||
cl_int err;
|
||||
|
||||
cl_uint platform_count = 0;
|
||||
cl_uint device_count = 0;
|
||||
|
||||
vector<opencl_device_info*> result;
|
||||
|
||||
clGetPlatformIDs(0, NULL, &platform_count);
|
||||
if(platform_count == 0) {
|
||||
return result;
|
||||
}
|
||||
|
||||
cl_platform_id *platforms = (cl_platform_id*)malloc(platform_count * sizeof(cl_platform_id));
|
||||
|
||||
err=clGetPlatformIDs(platform_count, platforms, &platform_count);
|
||||
if(err != CL_SUCCESS) {
|
||||
free(platforms);
|
||||
error = err;
|
||||
error_message = "Error querying for opencl platforms.";
|
||||
return result;
|
||||
}
|
||||
|
||||
int counter = 0;
|
||||
|
||||
for(uint32_t i=0; i < platform_count; i++) {
|
||||
device_count = 0;
|
||||
clGetDeviceIDs(platforms[i], CL_DEVICE_TYPE_GPU, 0, NULL, &device_count);
|
||||
if(device_count == 0) {
|
||||
continue;
|
||||
}
|
||||
|
||||
cl_device_id * devices = (cl_device_id*)malloc(device_count * sizeof(cl_device_id));
|
||||
err=clGetDeviceIDs(platforms[i], CL_DEVICE_TYPE_GPU, device_count, devices, &device_count);
|
||||
|
||||
if(err != CL_SUCCESS) {
|
||||
free(devices);
|
||||
error = err;
|
||||
error_message = "Error querying for opencl devices.";
|
||||
continue;
|
||||
}
|
||||
|
||||
for(uint32_t j=0; j < device_count; j++) {
|
||||
opencl_device_info *info = __get_device_info(platforms[i], devices[j]);
|
||||
if(info->error != CL_SUCCESS) {
|
||||
error = info->error;
|
||||
error_message = info->error_message;
|
||||
}
|
||||
else {
|
||||
info->device_index = counter;
|
||||
result.push_back(info);
|
||||
counter++;
|
||||
}
|
||||
}
|
||||
|
||||
free(devices);
|
||||
}
|
||||
|
||||
free(platforms);
|
||||
|
||||
return result;
|
||||
}
|
||||
|
||||
opencl_device_info *opencl_hasher::__get_device_info(cl_platform_id platform, cl_device_id device) {
|
||||
opencl_device_info *device_info = new opencl_device_info(CL_SUCCESS, "");
|
||||
|
||||
device_info->platform = platform;
|
||||
device_info->device = device;
|
||||
|
||||
char *buffer;
|
||||
size_t sz;
|
||||
|
||||
// device name
|
||||
string device_vendor;
|
||||
sz = 0;
|
||||
clGetDeviceInfo(device, CL_DEVICE_VENDOR, 0, NULL, &sz);
|
||||
buffer = (char *)malloc(sz + 1);
|
||||
device_info->error = clGetDeviceInfo(device, CL_DEVICE_VENDOR, sz, buffer, &sz);
|
||||
if(device_info->error != CL_SUCCESS) {
|
||||
free(buffer);
|
||||
device_info->error_message = "Error querying device vendor.";
|
||||
return device_info;
|
||||
}
|
||||
else {
|
||||
buffer[sz] = 0;
|
||||
device_vendor = buffer;
|
||||
free(buffer);
|
||||
}
|
||||
|
||||
string device_name;
|
||||
cl_device_info query_type = CL_DEVICE_NAME;
|
||||
|
||||
if(device_vendor.find("Advanced Micro Devices") != string::npos)
|
||||
query_type = CL_DEVICE_BOARD_NAME_AMD;
|
||||
|
||||
sz = 0;
|
||||
clGetDeviceInfo(device, query_type, 0, NULL, &sz);
|
||||
buffer = (char *) malloc(sz + 1);
|
||||
device_info->error = clGetDeviceInfo(device, query_type, sz, buffer, &sz);
|
||||
if (device_info->error != CL_SUCCESS) {
|
||||
free(buffer);
|
||||
device_info->error_message = "Error querying device name.";
|
||||
return device_info;
|
||||
} else {
|
||||
buffer[sz] = 0;
|
||||
device_name = buffer;
|
||||
free(buffer);
|
||||
}
|
||||
|
||||
string device_version;
|
||||
sz = 0;
|
||||
clGetDeviceInfo(device, CL_DEVICE_VERSION, 0, NULL, &sz);
|
||||
buffer = (char *)malloc(sz + 1);
|
||||
device_info->error = clGetDeviceInfo(device, CL_DEVICE_VERSION, sz, buffer, &sz);
|
||||
if(device_info->error != CL_SUCCESS) {
|
||||
free(buffer);
|
||||
device_info->error_message = "Error querying device version.";
|
||||
return device_info;
|
||||
}
|
||||
else {
|
||||
buffer[sz] = 0;
|
||||
device_version = buffer;
|
||||
free(buffer);
|
||||
}
|
||||
|
||||
device_info->device_string = device_vendor + " - " + device_name/* + " : " + device_version*/;
|
||||
|
||||
device_info->error = clGetDeviceInfo(device, CL_DEVICE_GLOBAL_MEM_SIZE, sizeof(device_info->max_mem_size), &(device_info->max_mem_size), NULL);
|
||||
if(device_info->error != CL_SUCCESS) {
|
||||
device_info->error_message = "Error querying device global memory size.";
|
||||
return device_info;
|
||||
}
|
||||
|
||||
device_info->error = clGetDeviceInfo(device, CL_DEVICE_MAX_MEM_ALLOC_SIZE, sizeof(device_info->max_allocable_mem_size), &(device_info->max_allocable_mem_size), NULL);
|
||||
if(device_info->error != CL_SUCCESS) {
|
||||
device_info->error_message = "Error querying device max memory allocation.";
|
||||
return device_info;
|
||||
}
|
||||
|
||||
double mem_in_gb = device_info->max_mem_size / 1073741824.0;
|
||||
stringstream ss;
|
||||
ss << setprecision(2) << mem_in_gb;
|
||||
device_info->device_string += (" (" + ss.str() + "GB)");
|
||||
|
||||
return device_info;
|
||||
}
|
||||
|
||||
bool opencl_hasher::configure(xmrig::HasherConfig &config) {
|
||||
int index = config.getGPUCardsCount();
|
||||
double intensity = 0;
|
||||
|
||||
int total_threads = 0;
|
||||
intensity = config.getAverageGPUIntensity();
|
||||
|
||||
if (intensity == 0) {
|
||||
m_intensity = 0;
|
||||
m_description = "Status: DISABLED - by user.";
|
||||
return false;
|
||||
}
|
||||
|
||||
bool cards_selected = false;
|
||||
|
||||
intensity = 0;
|
||||
|
||||
for(vector<opencl_device_info *>::iterator d = __devices.begin(); d != __devices.end(); d++, index++) {
|
||||
stringstream ss;
|
||||
ss << "["<< (index + 1) << "] " << (*d)->device_string;
|
||||
string device_description = ss.str();
|
||||
(*d)->device_index = index;
|
||||
(*d)->profile_info.profile = m_profile;
|
||||
|
||||
if(config.gpuFilter().size() > 0) {
|
||||
bool found = false;
|
||||
for(xmrig::GPUFilter fit : config.gpuFilter()) {
|
||||
if(device_description.find(fit.filter) != string::npos) {
|
||||
found = true;
|
||||
break;
|
||||
}
|
||||
}
|
||||
if(!found) {
|
||||
(*d)->profile_info.threads = 0;
|
||||
ss << " - DISABLED" << endl;
|
||||
m_description += ss.str();
|
||||
continue;
|
||||
}
|
||||
else {
|
||||
cards_selected = true;
|
||||
}
|
||||
}
|
||||
else {
|
||||
cards_selected = true;
|
||||
}
|
||||
|
||||
ss << endl;
|
||||
|
||||
double device_intensity = config.getGPUIntensity((*d)->device_index);
|
||||
|
||||
m_description += ss.str();
|
||||
|
||||
if(!(__setup_device_info((*d), device_intensity))) {
|
||||
m_description += (*d)->error_message;
|
||||
m_description += "\n";
|
||||
continue;
|
||||
};
|
||||
|
||||
DeviceInfo device;
|
||||
|
||||
if((*d)->device_string.find("Advanced Micro Devices") != string::npos) {
|
||||
device_topology_amd amdtopo;
|
||||
if(clGetDeviceInfo((*d)->device, CL_DEVICE_TOPOLOGY_AMD, sizeof(amdtopo), &amdtopo, NULL) == CL_SUCCESS) {
|
||||
char bus_id[50];
|
||||
sprintf(bus_id, "%02x:%02x.%x", amdtopo.pcie.bus, amdtopo.pcie.device, amdtopo.pcie.function);
|
||||
device.bus_id = bus_id;
|
||||
}
|
||||
}
|
||||
else if((*d)->device_string.find("NVIDIA") != string::npos) {
|
||||
cl_uint bus;
|
||||
cl_uint slot;
|
||||
|
||||
if(clGetDeviceInfo ((*d)->device, CL_DEVICE_PCI_BUS_ID_NV, sizeof(bus), &bus, NULL) == CL_SUCCESS) {
|
||||
if(clGetDeviceInfo ((*d)->device, CL_DEVICE_PCI_SLOT_ID_NV, sizeof(slot), &slot, NULL) == CL_SUCCESS) {
|
||||
char bus_id[50];
|
||||
sprintf(bus_id, "%02x:%02x.0", bus, slot);
|
||||
device.bus_id = bus_id;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
device.name = (*d)->device_string;
|
||||
device.intensity = device_intensity;
|
||||
storeDeviceInfo((*d)->device_index, device);
|
||||
|
||||
__enabledDevices.push_back(*d);
|
||||
|
||||
total_threads += (*d)->profile_info.threads;
|
||||
intensity += device_intensity;
|
||||
}
|
||||
|
||||
config.addGPUCardsCount(index - config.getGPUCardsCount());
|
||||
|
||||
if(!cards_selected) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - no card enabled because of filtering.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (total_threads == 0) {
|
||||
m_intensity = 0;
|
||||
m_description += "Status: DISABLED - not enough resources.";
|
||||
return false;
|
||||
}
|
||||
|
||||
buildThreadData();
|
||||
|
||||
m_intensity = intensity / __enabledDevices.size();
|
||||
m_computingThreads = __enabledDevices.size() * 2; // 2 computing threads for each device
|
||||
m_description += "Status: ENABLED - with " + to_string(total_threads) + " threads.";
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
bool opencl_hasher::__setup_device_info(opencl_device_info *device, double intensity) {
|
||||
cl_int error;
|
||||
|
||||
cl_context_properties properties[] = {
|
||||
CL_CONTEXT_PLATFORM, (cl_context_properties) device->platform,
|
||||
0};
|
||||
|
||||
device->context = clCreateContext(properties, 1, &(device->device), NULL, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error getting device context.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->queue = clCreateCommandQueue(device->context, device->device, CL_QUEUE_PROFILING_ENABLE, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error getting device command queue.";
|
||||
return false;
|
||||
}
|
||||
|
||||
const char *srcptr[] = {OpenCLKernel.c_str()};
|
||||
size_t srcsize = OpenCLKernel.size();
|
||||
|
||||
device->program = clCreateProgramWithSource(device->context, 1, srcptr, &srcsize, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating opencl program for device.";
|
||||
return false;
|
||||
}
|
||||
|
||||
error = clBuildProgram(device->program, 1, &device->device, "", NULL, NULL);
|
||||
if (error != CL_SUCCESS) {
|
||||
size_t log_size;
|
||||
clGetProgramBuildInfo(device->program, device->device, CL_PROGRAM_BUILD_LOG, 0, NULL, &log_size);
|
||||
char *log = (char *) malloc(log_size + 1);
|
||||
clGetProgramBuildInfo(device->program, device->device, CL_PROGRAM_BUILD_LOG, log_size, log, NULL);
|
||||
log[log_size] = 0;
|
||||
string build_log = log;
|
||||
free(log);
|
||||
|
||||
device->error = error;
|
||||
device->error_message = "Error building opencl program for device: " + build_log;
|
||||
return false;
|
||||
}
|
||||
|
||||
device->kernel_prehash = clCreateKernel(device->program, "prehash", &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating opencl prehash kernel for device.";
|
||||
return false;
|
||||
}
|
||||
device->kernel_fill_blocks = clCreateKernel(device->program, "fill_blocks", &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating opencl main kernel for device.";
|
||||
return false;
|
||||
}
|
||||
device->kernel_posthash = clCreateKernel(device->program, "posthash", &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating opencl posthash kernel for device.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->profile_info.threads_per_chunk = (uint32_t) (device->max_allocable_mem_size / device->profile_info.profile->memSize);
|
||||
size_t chunk_size = device->profile_info.threads_per_chunk * device->profile_info.profile->memSize;
|
||||
|
||||
if (chunk_size == 0) {
|
||||
device->error = -1;
|
||||
device->error_message = "Not enough memory on GPU.";
|
||||
return false;
|
||||
}
|
||||
|
||||
uint64_t usable_memory = device->max_mem_size;
|
||||
double chunks = (double) usable_memory / (double) chunk_size;
|
||||
|
||||
uint32_t max_threads = (uint32_t) (device->profile_info.threads_per_chunk * chunks);
|
||||
|
||||
if (max_threads == 0) {
|
||||
device->error = -1;
|
||||
device->error_message = "Not enough memory on GPU.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->profile_info.threads = (uint32_t) (max_threads * intensity / 100.0);
|
||||
device->profile_info.threads = (device->profile_info.threads / 4) * 4; // make it divisible by 4
|
||||
if (max_threads > 0 && device->profile_info.threads == 0 && intensity > 0)
|
||||
device->profile_info.threads = 4;
|
||||
|
||||
double counter = (double) device->profile_info.threads / (double) device->profile_info.threads_per_chunk;
|
||||
size_t allocated_mem_for_current_chunk = 0;
|
||||
|
||||
if (counter > 0) {
|
||||
if (counter > 1) {
|
||||
allocated_mem_for_current_chunk = chunk_size;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = (size_t) ceil(chunk_size * counter);
|
||||
}
|
||||
counter -= 1;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = 1;
|
||||
}
|
||||
device->arguments.memory_chunk_0 = clCreateBuffer(device->context, CL_MEM_READ_WRITE,
|
||||
allocated_mem_for_current_chunk, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (counter > 0) {
|
||||
if (counter > 1) {
|
||||
allocated_mem_for_current_chunk = chunk_size;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = (size_t) ceil(chunk_size * counter);
|
||||
}
|
||||
counter -= 1;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = 1;
|
||||
}
|
||||
device->arguments.memory_chunk_1 = clCreateBuffer(device->context, CL_MEM_READ_WRITE,
|
||||
allocated_mem_for_current_chunk, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (counter > 0) {
|
||||
if (counter > 1) {
|
||||
allocated_mem_for_current_chunk = chunk_size;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = (size_t) ceil(chunk_size * counter);
|
||||
}
|
||||
counter -= 1;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = 1;
|
||||
}
|
||||
device->arguments.memory_chunk_2 = clCreateBuffer(device->context, CL_MEM_READ_WRITE,
|
||||
allocated_mem_for_current_chunk, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (counter > 0) {
|
||||
if (counter > 1) {
|
||||
allocated_mem_for_current_chunk = chunk_size;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = (size_t) ceil(chunk_size * counter);
|
||||
}
|
||||
counter -= 1;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = 1;
|
||||
}
|
||||
device->arguments.memory_chunk_3 = clCreateBuffer(device->context, CL_MEM_READ_WRITE,
|
||||
allocated_mem_for_current_chunk, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (counter > 0) {
|
||||
if (counter > 1) {
|
||||
allocated_mem_for_current_chunk = chunk_size;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = (size_t) ceil(chunk_size * counter);
|
||||
}
|
||||
counter -= 1;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = 1;
|
||||
}
|
||||
device->arguments.memory_chunk_4 = clCreateBuffer(device->context, CL_MEM_READ_WRITE,
|
||||
allocated_mem_for_current_chunk, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (counter > 0) {
|
||||
if (counter > 1) {
|
||||
allocated_mem_for_current_chunk = chunk_size;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = (size_t) ceil(chunk_size * counter);
|
||||
}
|
||||
counter -= 1;
|
||||
} else {
|
||||
allocated_mem_for_current_chunk = 1;
|
||||
}
|
||||
device->arguments.memory_chunk_5 = clCreateBuffer(device->context, CL_MEM_READ_WRITE,
|
||||
allocated_mem_for_current_chunk, NULL, &error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.refs = clCreateBuffer(device->context, CL_MEM_READ_ONLY,
|
||||
device->profile_info.profile->blockRefsSize * sizeof(uint32_t), NULL,
|
||||
&error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
if (device->profile_info.profile->succesiveIdxs == 1) {
|
||||
device->arguments.idxs = NULL;
|
||||
}
|
||||
else {
|
||||
device->arguments.idxs = clCreateBuffer(device->context, CL_MEM_READ_ONLY,
|
||||
device->profile_info.profile->blockRefsSize * sizeof(uint32_t), NULL,
|
||||
&error);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
}
|
||||
|
||||
device->arguments.segments = clCreateBuffer(device->context, CL_MEM_READ_ONLY, device->profile_info.profile->segCount * 3 * sizeof(uint32_t), NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
size_t preseed_memory_size = device->profile_info.profile->pwdLen * 4;
|
||||
size_t seed_memory_size = device->profile_info.threads * (device->profile_info.profile->thrCost * 2) * ARGON2_BLOCK_SIZE;
|
||||
size_t out_memory_size = device->profile_info.threads * ARGON2_BLOCK_SIZE;
|
||||
size_t hash_memory_size = device->profile_info.threads * (xmrig::ARGON2_HASHLEN + 4);
|
||||
|
||||
device->arguments.preseed_memory[0] = clCreateBuffer(device->context, CL_MEM_READ_ONLY, preseed_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.preseed_memory[1] = clCreateBuffer(device->context, CL_MEM_READ_ONLY, preseed_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.seed_memory[0] = clCreateBuffer(device->context, CL_MEM_READ_WRITE, seed_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.seed_memory[1] = clCreateBuffer(device->context, CL_MEM_READ_WRITE, seed_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.out_memory[0] = clCreateBuffer(device->context, CL_MEM_READ_WRITE, out_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.out_memory[1] = clCreateBuffer(device->context, CL_MEM_READ_WRITE, out_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.hash_memory[0] = clCreateBuffer(device->context, CL_MEM_WRITE_ONLY, hash_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
device->arguments.hash_memory[1] = clCreateBuffer(device->context, CL_MEM_WRITE_ONLY, hash_memory_size, NULL, &error);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error creating memory buffer.";
|
||||
return false;
|
||||
}
|
||||
|
||||
//optimise address sizes
|
||||
uint32_t *refs = (uint32_t *)malloc(device->profile_info.profile->blockRefsSize * sizeof(uint32_t));
|
||||
for(int i=0;i<device->profile_info.profile->blockRefsSize;i++) {
|
||||
refs[i] = device->profile_info.profile->blockRefs[i*3 + 1];
|
||||
}
|
||||
|
||||
error=clEnqueueWriteBuffer(device->queue, device->arguments.refs, CL_TRUE, 0, device->profile_info.profile->blockRefsSize * sizeof(uint32_t), refs, 0, NULL, NULL);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error writing to gpu memory.";
|
||||
return false;
|
||||
}
|
||||
|
||||
free(refs);
|
||||
|
||||
if(device->profile_info.profile->succesiveIdxs == 0) {
|
||||
uint32_t *idxs = (uint32_t *) malloc(device->profile_info.profile->blockRefsSize * sizeof(uint32_t));
|
||||
for (int i = 0; i < device->profile_info.profile->blockRefsSize; i++) {
|
||||
idxs[i] = device->profile_info.profile->blockRefs[i * 3];
|
||||
if (device->profile_info.profile->blockRefs[i * 3 + 2] == 1) {
|
||||
idxs[i] |= 0x80000000;
|
||||
}
|
||||
}
|
||||
|
||||
error=clEnqueueWriteBuffer(device->queue, device->arguments.idxs, CL_TRUE, 0, device->profile_info.profile->blockRefsSize * sizeof(uint32_t), idxs, 0, NULL, NULL);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error writing to gpu memory.";
|
||||
return false;
|
||||
}
|
||||
|
||||
free(idxs);
|
||||
}
|
||||
|
||||
error=clEnqueueWriteBuffer(device->queue, device->arguments.segments, CL_TRUE, 0, device->profile_info.profile->segCount * 3 * sizeof(uint32_t), device->profile_info.profile->segments, 0, NULL, NULL);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error writing to gpu memory.";
|
||||
return false;
|
||||
}
|
||||
|
||||
clSetKernelArg(device->kernel_fill_blocks, 0, sizeof(device->arguments.memory_chunk_0), &device->arguments.memory_chunk_0);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 1, sizeof(device->arguments.memory_chunk_1), &device->arguments.memory_chunk_1);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 2, sizeof(device->arguments.memory_chunk_2), &device->arguments.memory_chunk_2);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 3, sizeof(device->arguments.memory_chunk_3), &device->arguments.memory_chunk_3);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 4, sizeof(device->arguments.memory_chunk_4), &device->arguments.memory_chunk_4);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 5, sizeof(device->arguments.memory_chunk_5), &device->arguments.memory_chunk_5);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 8, sizeof(device->arguments.refs), &device->arguments.refs);
|
||||
if(device->profile_info.profile->succesiveIdxs == 0)
|
||||
clSetKernelArg(device->kernel_fill_blocks, 9, sizeof(device->arguments.idxs), &device->arguments.idxs);
|
||||
else
|
||||
clSetKernelArg(device->kernel_fill_blocks, 9, sizeof(cl_mem), NULL);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 10, sizeof(device->arguments.segments), &device->arguments.segments);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 11, sizeof(int32_t), &device->profile_info.profile->memSize);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 12, sizeof(int32_t), &device->profile_info.profile->thrCost);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 13, sizeof(int32_t), &device->profile_info.profile->segSize);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 14, sizeof(int32_t), &device->profile_info.profile->segCount);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 15, sizeof(int32_t), &device->profile_info.threads_per_chunk);
|
||||
|
||||
clSetKernelArg(device->kernel_prehash, 2, sizeof(int32_t), &device->profile_info.profile->memCost);
|
||||
clSetKernelArg(device->kernel_prehash, 3, sizeof(int32_t), &device->profile_info.profile->thrCost);
|
||||
int passes = device->profile_info.profile->segCount / (4 * device->profile_info.profile->thrCost);
|
||||
clSetKernelArg(device->kernel_prehash, 4, sizeof(int32_t), &passes);
|
||||
clSetKernelArg(device->kernel_prehash, 6, sizeof(int32_t), &device->profile_info.profile->saltLen);
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
bool opencl_kernel_prehasher(void *memory, int threads, Argon2Profile *profile, void *user_data) {
|
||||
opencl_gpumgmt_thread_data *gpumgmt_thread = (opencl_gpumgmt_thread_data *)user_data;
|
||||
opencl_device_info *device = gpumgmt_thread->device;
|
||||
|
||||
cl_int error;
|
||||
|
||||
int sessions = max(profile->thrCost * 2, (uint32_t)16);
|
||||
double hashes_per_block = sessions / (profile->thrCost * 2.0);
|
||||
|
||||
size_t total_work_items = sessions * 4 * ceil(threads / hashes_per_block);
|
||||
size_t local_work_items = sessions * 4;
|
||||
|
||||
device->device_lock.lock();
|
||||
|
||||
error = clEnqueueWriteBuffer(device->queue, device->arguments.preseed_memory[gpumgmt_thread->thread_id],
|
||||
CL_FALSE, 0, gpumgmt_thread->hashData.inSize, memory, 0, NULL, NULL);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error writing to gpu memory.";
|
||||
device->device_lock.unlock();
|
||||
return false;
|
||||
}
|
||||
|
||||
int inSizeInInt = gpumgmt_thread->hashData.inSize / 4;
|
||||
clSetKernelArg(device->kernel_prehash, 0, sizeof(device->arguments.preseed_memory[gpumgmt_thread->thread_id]), &device->arguments.preseed_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_prehash, 1, sizeof(device->arguments.seed_memory[gpumgmt_thread->thread_id]), &device->arguments.seed_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_prehash, 5, sizeof(int), &inSizeInInt);
|
||||
clSetKernelArg(device->kernel_prehash, 7, sizeof(int), &threads);
|
||||
clSetKernelArg(device->kernel_prehash, 8, sessions * sizeof(cl_ulong) * 76, NULL); // (preseed size is 16 ulongs = 128 bytes)
|
||||
|
||||
error=clEnqueueNDRangeKernel(device->queue, device->kernel_prehash, 1, NULL, &total_work_items, &local_work_items, 0, NULL, NULL);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error running the kernel.";
|
||||
device->device_lock.unlock();
|
||||
return false;
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
void *opencl_kernel_filler(int threads, Argon2Profile *profile, void *user_data) {
|
||||
opencl_gpumgmt_thread_data *gpumgmt_thread = (opencl_gpumgmt_thread_data *)user_data;
|
||||
opencl_device_info *device = gpumgmt_thread->device;
|
||||
|
||||
cl_int error;
|
||||
|
||||
size_t total_work_items = threads * KERNEL_WORKGROUP_SIZE * profile->thrCost;
|
||||
size_t local_work_items = KERNEL_WORKGROUP_SIZE * profile->thrCost;
|
||||
|
||||
size_t shared_mem = profile->thrCost * ARGON2_QWORDS_IN_BLOCK;
|
||||
|
||||
clSetKernelArg(device->kernel_fill_blocks, 6, sizeof(device->arguments.seed_memory[gpumgmt_thread->thread_id]), &device->arguments.seed_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 7, sizeof(device->arguments.out_memory[gpumgmt_thread->thread_id]), &device->arguments.out_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_fill_blocks, 16, sizeof(cl_ulong) * shared_mem, NULL);
|
||||
|
||||
error=clEnqueueNDRangeKernel(device->queue, device->kernel_fill_blocks, 1, NULL, &total_work_items, &local_work_items, 0, NULL, NULL);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error running the kernel.";
|
||||
device->device_lock.unlock();
|
||||
return NULL;
|
||||
}
|
||||
|
||||
return (void *)1;
|
||||
}
|
||||
|
||||
bool opencl_kernel_posthasher(void *memory, int threads, Argon2Profile *profile, void *user_data) {
|
||||
opencl_gpumgmt_thread_data *gpumgmt_thread = (opencl_gpumgmt_thread_data *)user_data;
|
||||
opencl_device_info *device = gpumgmt_thread->device;
|
||||
|
||||
cl_int error;
|
||||
|
||||
size_t total_work_items = threads * 4;
|
||||
size_t local_work_items = 4;
|
||||
|
||||
clSetKernelArg(device->kernel_posthash, 0, sizeof(device->arguments.hash_memory[gpumgmt_thread->thread_id]), &device->arguments.hash_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_posthash, 1, sizeof(device->arguments.out_memory[gpumgmt_thread->thread_id]), &device->arguments.out_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_posthash, 2, sizeof(device->arguments.preseed_memory[gpumgmt_thread->thread_id]), &device->arguments.preseed_memory[gpumgmt_thread->thread_id]);
|
||||
clSetKernelArg(device->kernel_posthash, 3, sizeof(cl_ulong) * 60, NULL);
|
||||
|
||||
error=clEnqueueNDRangeKernel(device->queue, device->kernel_posthash, 1, NULL, &total_work_items, &local_work_items, 0, NULL, NULL);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error running the kernel.";
|
||||
device->device_lock.unlock();
|
||||
return false;
|
||||
}
|
||||
|
||||
error = clEnqueueReadBuffer(device->queue, device->arguments.hash_memory[gpumgmt_thread->thread_id], CL_FALSE, 0, threads * (xmrig::ARGON2_HASHLEN + 4), memory, 0, NULL, NULL);
|
||||
if (error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error reading gpu memory.";
|
||||
device->device_lock.unlock();
|
||||
return false;
|
||||
}
|
||||
|
||||
error=clFinish(device->queue);
|
||||
if(error != CL_SUCCESS) {
|
||||
device->error = error;
|
||||
device->error_message = "Error flushing GPU queue.";
|
||||
device->device_lock.unlock();
|
||||
return false;
|
||||
}
|
||||
|
||||
device->device_lock.unlock();
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
void opencl_hasher::buildThreadData() {
|
||||
__thread_data = new opencl_gpumgmt_thread_data[__enabledDevices.size() * 2];
|
||||
|
||||
for(int i=0; i < __enabledDevices.size(); i++) {
|
||||
opencl_device_info *device = __enabledDevices[i];
|
||||
for(int threadId = 0; threadId < 2; threadId ++) {
|
||||
opencl_gpumgmt_thread_data &thread_data = __thread_data[i * 2 + threadId];
|
||||
thread_data.device = device;
|
||||
thread_data.thread_id = threadId;
|
||||
thread_data.argon2 = new Argon2(opencl_kernel_prehasher, opencl_kernel_filler, opencl_kernel_posthasher,
|
||||
nullptr, &thread_data);
|
||||
thread_data.argon2->setThreads(device->profile_info.threads);
|
||||
thread_data.hashData.outSize = xmrig::ARGON2_HASHLEN + 4;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
int opencl_hasher::compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output) {
|
||||
opencl_gpumgmt_thread_data &threadData = __thread_data[threadIdx];
|
||||
threadData.hashData.input = input;
|
||||
threadData.hashData.inSize = size;
|
||||
threadData.hashData.output = output;
|
||||
int hashCount = threadData.argon2->generateHashes(*m_profile, threadData.hashData);
|
||||
if(threadData.device->error != CL_SUCCESS) {
|
||||
LOG("Error running kernel: (" + to_string(threadData.device->error) + ")" + threadData.device->error_message);
|
||||
return 0;
|
||||
}
|
||||
|
||||
uint32_t *nonce = ((uint32_t *)(((uint8_t*)threadData.hashData.input) + 39));
|
||||
(*nonce) += threadData.device->profile_info.threads;
|
||||
|
||||
return hashCount;
|
||||
}
|
||||
|
||||
void opencl_hasher::cleanup() {
|
||||
vector<cl_platform_id> platforms;
|
||||
|
||||
for(vector<opencl_device_info *>::iterator it=__devices.begin(); it != __devices.end(); it++) {
|
||||
if ((*it)->profile_info.threads != 0) {
|
||||
clReleaseMemObject((*it)->arguments.memory_chunk_0);
|
||||
clReleaseMemObject((*it)->arguments.memory_chunk_1);
|
||||
clReleaseMemObject((*it)->arguments.memory_chunk_2);
|
||||
clReleaseMemObject((*it)->arguments.memory_chunk_3);
|
||||
clReleaseMemObject((*it)->arguments.memory_chunk_4);
|
||||
clReleaseMemObject((*it)->arguments.memory_chunk_5);
|
||||
clReleaseMemObject((*it)->arguments.refs);
|
||||
clReleaseMemObject((*it)->arguments.segments);
|
||||
clReleaseMemObject((*it)->arguments.preseed_memory[0]);
|
||||
clReleaseMemObject((*it)->arguments.preseed_memory[1]);
|
||||
clReleaseMemObject((*it)->arguments.seed_memory[0]);
|
||||
clReleaseMemObject((*it)->arguments.seed_memory[1]);
|
||||
clReleaseMemObject((*it)->arguments.out_memory[0]);
|
||||
clReleaseMemObject((*it)->arguments.out_memory[1]);
|
||||
clReleaseMemObject((*it)->arguments.hash_memory[0]);
|
||||
clReleaseMemObject((*it)->arguments.hash_memory[1]);
|
||||
|
||||
clReleaseKernel((*it)->kernel_prehash);
|
||||
clReleaseKernel((*it)->kernel_fill_blocks);
|
||||
clReleaseKernel((*it)->kernel_posthash);
|
||||
clReleaseProgram((*it)->program);
|
||||
clReleaseCommandQueue((*it)->queue);
|
||||
clReleaseContext((*it)->context);
|
||||
}
|
||||
clReleaseDevice((*it)->device);
|
||||
delete (*it);
|
||||
}
|
||||
__devices.clear();
|
||||
}
|
||||
|
||||
size_t opencl_hasher::parallelism(int workerIdx) {
|
||||
// there are 2 computing threads per device, so divide by 2 to get device index
|
||||
workerIdx /= 2;
|
||||
|
||||
if(workerIdx < 0 || workerIdx > __enabledDevices.size())
|
||||
return 0;
|
||||
|
||||
return __enabledDevices[workerIdx]->profile_info.threads;
|
||||
}
|
||||
|
||||
size_t opencl_hasher::deviceCount() {
|
||||
return __enabledDevices.size();
|
||||
}
|
||||
|
||||
REGISTER_HASHER(opencl_hasher);
|
||||
|
||||
#endif // WITH_OPENCL
|
||||
110
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLHasher.h
Executable file
110
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLHasher.h
Executable file
|
|
@ -0,0 +1,110 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 03/08/2018.
|
||||
//
|
||||
|
||||
#ifndef ARGON2_OPENCL_HASHER_H
|
||||
#define ARGON2_OPENCL_HASHER_H
|
||||
|
||||
#if defined(WITH_OPENCL)
|
||||
|
||||
#define CL_USE_DEPRECATED_OPENCL_1_2_APIS
|
||||
|
||||
#if defined(__APPLE__) || defined(__MACOSX)
|
||||
#include <OpenCL/opencl.h>
|
||||
#else
|
||||
#include <CL/opencl.h>
|
||||
#endif // !__APPLE__
|
||||
|
||||
struct opencl_kernel_arguments {
|
||||
cl_mem memory_chunk_0;
|
||||
cl_mem memory_chunk_1;
|
||||
cl_mem memory_chunk_2;
|
||||
cl_mem memory_chunk_3;
|
||||
cl_mem memory_chunk_4;
|
||||
cl_mem memory_chunk_5;
|
||||
cl_mem refs;
|
||||
cl_mem idxs;
|
||||
cl_mem segments;
|
||||
cl_mem preseed_memory[2];
|
||||
cl_mem seed_memory[2];
|
||||
cl_mem out_memory[2];
|
||||
cl_mem hash_memory[2];
|
||||
};
|
||||
|
||||
struct argon2profile_info {
|
||||
argon2profile_info() {
|
||||
threads = 0;
|
||||
threads_per_chunk = 0;
|
||||
}
|
||||
|
||||
uint32_t threads;
|
||||
uint32_t threads_per_chunk;
|
||||
Argon2Profile *profile;
|
||||
};
|
||||
|
||||
struct opencl_device_info {
|
||||
opencl_device_info(cl_int err, const string &err_msg) {
|
||||
error = err;
|
||||
error_message = err_msg;
|
||||
}
|
||||
|
||||
cl_platform_id platform;
|
||||
cl_device_id device;
|
||||
cl_context context;
|
||||
cl_command_queue queue;
|
||||
|
||||
cl_program program;
|
||||
cl_kernel kernel_prehash;
|
||||
cl_kernel kernel_fill_blocks;
|
||||
cl_kernel kernel_posthash;
|
||||
|
||||
int device_index;
|
||||
|
||||
opencl_kernel_arguments arguments;
|
||||
argon2profile_info profile_info;
|
||||
|
||||
string device_string;
|
||||
uint64_t max_mem_size;
|
||||
uint64_t max_allocable_mem_size;
|
||||
|
||||
cl_int error;
|
||||
string error_message;
|
||||
|
||||
mutex device_lock;
|
||||
};
|
||||
|
||||
struct opencl_gpumgmt_thread_data {
|
||||
int thread_id;
|
||||
opencl_device_info *device;
|
||||
Argon2 *argon2;
|
||||
HashData hashData;
|
||||
};
|
||||
|
||||
class opencl_hasher : public Hasher {
|
||||
public:
|
||||
opencl_hasher();
|
||||
~opencl_hasher();
|
||||
|
||||
virtual bool initialize(xmrig::Algo algorithm, xmrig::Variant variant);
|
||||
virtual bool configure(xmrig::HasherConfig &config);
|
||||
virtual void cleanup();
|
||||
virtual int compute(int threadIdx, uint8_t *input, size_t size, uint8_t *output);
|
||||
virtual size_t parallelism(int workerIdx);
|
||||
virtual size_t deviceCount();
|
||||
|
||||
private:
|
||||
opencl_device_info *__get_device_info(cl_platform_id platform, cl_device_id device);
|
||||
bool __setup_device_info(opencl_device_info *device, double intensity);
|
||||
vector<opencl_device_info*> __query_opencl_devices(cl_int &error, string &error_message);
|
||||
void buildThreadData();
|
||||
|
||||
vector<opencl_device_info*> __devices;
|
||||
vector<opencl_device_info*> __enabledDevices;
|
||||
opencl_gpumgmt_thread_data *__thread_data;
|
||||
|
||||
Argon2Profile *m_profile;
|
||||
};
|
||||
|
||||
#endif //WITH_OPENCL
|
||||
|
||||
#endif //ARGON2_OPENCL_HASHER_H
|
||||
1085
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLKernel.cpp
Normal file
1085
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLKernel.cpp
Normal file
File diff suppressed because it is too large
Load diff
10
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLKernel.h
Normal file
10
src/crypto/argon2_hasher/hash/gpu/opencl/OpenCLKernel.h
Normal file
|
|
@ -0,0 +1,10 @@
|
|||
//
|
||||
// Created by Haifa Bogdan Adnan on 06/08/2018.
|
||||
//
|
||||
|
||||
#ifndef ARGON2_OPENCL_KERNEL_H
|
||||
#define ARGON2_OPENCL_KERNEL_H
|
||||
|
||||
extern string OpenCLKernel;
|
||||
|
||||
#endif //ARGON2_OPENCL_KERNEL_H
|
||||
Loading…
Add table
Add a link
Reference in a new issue