Ecosystem/xmrig
Ecosystem/xmrig
1
0
Fork 0
mirror of https://github.com/xmrig/xmrig.git synced 2024-11-18 10:01:06 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Added autoconfig for cn/gpu.

Browse source
This commit is contained in:
XMRig 2019-09-08 16:28:51 +07:00
parent 859626cbe3
commit 29790da63d
10 changed files with 157 additions and 33 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
src/backend/opencl/OclBackend.cpp
View file

@ -136,8 +136,8 @@ public:
device.printableName().data(),
device.clock(),
device.computeUnits(),
device.freeMem() / oneMiB,
device.globalMem() / oneMiB);
device.freeMemSize() / oneMiB,
device.globalMemSize() / oneMiB);
}
}

4
src/backend/opencl/OclConfig.cpp
View file

@ -44,7 +44,7 @@ static const char *kPlatform = "platform";
#ifdef XMRIG_ALGO_CN_GPU
//static const char *kCnGPU = "cn/gpu";
static const char *kCnGPU = "cn/gpu";
#endif
#ifdef XMRIG_ALGO_CN_LITE
@ -225,7 +225,7 @@ void xmrig::OclConfig::generate()
m_threads.move(kCn2, xmrig::generate(Algorithm::CN_2, devices));
# ifdef XMRIG_ALGO_CN_GPU
// m_threads.move(kCnGPU, xmrig::generate(Algorithm::CN_GPU, devices));
m_threads.move(kCnGPU, xmrig::generate(Algorithm::CN_GPU, devices));
# endif
# ifdef XMRIG_ALGO_CN_LITE

7
src/backend/opencl/OclThread.cpp
View file

@ -69,6 +69,9 @@ xmrig::OclThread::OclThread(const rapidjson::Value &value)
m_stridedIndex = std::min(si[0].GetUint(), 2u);
m_memChunk = std::min(si[1].GetUint(), 18u);
}
else {
m_fields.set(STRIDED_INDEX_FIELD, false);
}
const rapidjson::Value &threads = Json::getArray(value, kThreads);
if (threads.IsArray()) {
@ -112,10 +115,13 @@ rapidjson::Value xmrig::OclThread::toJSON(rapidjson::Document &doc) const
out.AddMember(StringRef(kIntensity), intensity(), allocator);
out.AddMember(StringRef(kWorksize), worksize(), allocator);
if (m_fields.test(STRIDED_INDEX_FIELD)) {
Value si(kArrayType);
si.Reserve(2, allocator);
si.PushBack(stridedIndex(), allocator);
si.PushBack(memChunk(), allocator);
out.AddMember(StringRef(kStridedIndex), si, allocator);
}
Value threads(kArrayType);
threads.Reserve(m_threads.size(), allocator);
@ -124,7 +130,6 @@ rapidjson::Value xmrig::OclThread::toJSON(rapidjson::Document &doc) const
threads.PushBack(thread, allocator);
}
out.AddMember(StringRef(kStridedIndex), si, allocator);
out.AddMember(StringRef(kThreads), threads, allocator);
out.AddMember(StringRef(kUnroll), unrollFactor(), allocator);

26
src/backend/opencl/OclThread.h
View file

@ -30,6 +30,7 @@
#include "rapidjson/fwd.h"
#include <bitset>
#include <vector>
@ -40,8 +41,7 @@ class OclThread
{
public:
OclThread() = delete;
OclThread(uint32_t index, uint32_t intensity, uint32_t worksize, uint32_t stridedIndex, uint32_t memChunk, uint32_t threads, uint32_t unrollFactor, const Algorithm &algorithm) :
m_algorithm(algorithm),
OclThread(uint32_t index, uint32_t intensity, uint32_t worksize, uint32_t stridedIndex, uint32_t memChunk, uint32_t threads, uint32_t unrollFactor) :
m_threads(threads, -1),
m_index(index),
m_memChunk(memChunk),
@ -52,6 +52,20 @@ public:
setIntensity(intensity);
}
# ifdef XMRIG_ALGO_CN_GPU
OclThread(uint32_t index, uint32_t intensity, uint32_t worksize, uint32_t threads, uint32_t unrollFactor) :
m_fields(0),
m_threads(threads, -1),
m_index(index),
m_memChunk(0),
m_stridedIndex(0),
m_unrollFactor(unrollFactor),
m_worksize(worksize)
{
setIntensity(intensity);
}
# endif
OclThread(const rapidjson::Value &value);
inline bool isValid() const { return m_intensity > 0; }
@ -73,10 +87,16 @@ public:
rapidjson::Value toJSON(rapidjson::Document &doc) const;
private:
enum Fields {
STRIDED_INDEX_FIELD,
RANDOMX_FIELDS,
FIELD_MAX
};
inline void setIntensity(uint32_t intensity) { m_intensity = intensity / m_worksize * m_worksize; }
Algorithm m_algorithm;
int m_datasetHost = -1;
std::bitset<FIELD_MAX> m_fields = 1;
std::vector<int64_t> m_threads;
uint32_t m_bfactor = 6;
uint32_t m_gcnAsm = 1;

6
src/backend/opencl/generators/ocl_generic_cn_generator.cpp
View file

@ -50,7 +50,7 @@ static inline uint32_t getPossibleIntensity(const OclDevice &device, const Algor
{
const uint32_t maxThreads = getMaxThreads(device, algorithm);
const size_t minFreeMem = (maxThreads == 40000u ? 512u : 128u) * oneMiB;
const size_t availableMem = device.freeMem() - minFreeMem;
const size_t availableMem = device.freeMemSize() - minFreeMem;
const size_t perThread = algorithm.l3() + 224u;
const auto maxIntensity = static_cast<uint32_t>(availableMem / perThread);
@ -104,9 +104,9 @@ bool ocl_generic_cn_generator(const OclDevice &device, const Algorithm &algorith
return false;
}
const uint32_t threadCount = ((device.globalMem() - intensity * 2 * algorithm.l3()) > 128 * oneMiB) ? 2 : 1;
const uint32_t threadCount = ((device.globalMemSize() - intensity * 2 * algorithm.l3()) > 128 * oneMiB) ? 2 : 1;
threads.add(OclThread(device.index(), intensity, 8, getStridedIndex(device, algorithm), 2, threadCount, 8, algorithm));
threads.add(OclThread(device.index(), intensity, 8, getStridedIndex(device, algorithm), 2, threadCount, 8));
return true;
}

84
src/backend/opencl/generators/ocl_generic_cn_gpu_generator.cpp Normal file
View file

@ -0,0 +1,84 @@
/* XMRig
* Copyright 2010 Jeff Garzik <jgarzik@pobox.com>
* Copyright 2012-2014 pooler <pooler@litecoinpool.org>
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018 Lee Clagett <https://github.com/vtnerd>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "backend/opencl/OclThreads.h"
#include "backend/opencl/wrappers/OclDevice.h"
#include "crypto/common/Algorithm.h"
namespace xmrig {
constexpr const size_t oneMiB = 1024u * 1024u;
bool ocl_generic_cn_gpu_generator(const OclDevice &device, const Algorithm &algorithm, OclThreads &threads)
{
if (algorithm != Algorithm::CN_GPU) {
return false;
}
uint32_t worksize = 8;
uint32_t numThreads = 1u;
size_t minFreeMem = 128u * oneMiB;
if (device.type() == OclDevice::Vega_10 || device.type() == OclDevice::Vega_20) {
minFreeMem = oneMiB;
worksize = 16;
}
else if (device.type() == OclDevice::Navi_10) {
numThreads = 2u;
}
else if (device.name() == "Fiji") {
worksize = 16;
}
size_t maxThreads = device.computeUnits() * 6 * 8;
const size_t maxAvailableFreeMem = device.freeMemSize() - minFreeMem;
const size_t memPerThread = std::min(device.maxMemAllocSize(), maxAvailableFreeMem);
size_t memPerHash = algorithm.l3() + 240u;
size_t maxIntensity = memPerThread / memPerHash;
size_t possibleIntensity = std::min(maxThreads, maxIntensity);
size_t intensity = 0;
size_t cuUtilization = ((possibleIntensity * 100) / (worksize * device.computeUnits())) % 100;
if (cuUtilization >= 75) {
intensity = (possibleIntensity / worksize) * worksize;
}
else {
intensity = (possibleIntensity / (worksize * device.computeUnits())) * device.computeUnits() * worksize;
}
threads.add(OclThread(device.index(), intensity, worksize, numThreads, 1));
return true;
}
} // namespace xmrig

4
src/backend/opencl/generators/ocl_vega_cn_generator.cpp
View file

@ -61,7 +61,7 @@ static inline uint32_t getMaxThreads(const OclDevice &device, const Algorithm &a
static inline uint32_t getPossibleIntensity(const OclDevice &device, const Algorithm &algorithm)
{
const uint32_t maxThreads = getMaxThreads(device, algorithm);
const size_t availableMem = device.freeMem() - (128u * oneMiB);
const size_t availableMem = device.freeMemSize() - (128u * oneMiB);
const size_t perThread = algorithm.l3() + 224u;
const auto maxIntensity = static_cast<uint32_t>(availableMem / perThread);
@ -123,7 +123,7 @@ bool ocl_vega_cn_generator(const OclDevice &device, const Algorithm &algorithm,
const uint32_t worksize = getWorksize(algorithm);
const uint32_t memChunk = getMemChunk(algorithm);
threads.add(OclThread(device.index(), intensity, worksize, getStridedIndex(algorithm), memChunk, 2, 8, algorithm));
threads.add(OclThread(device.index(), intensity, worksize, getStridedIndex(algorithm), memChunk, 2, 8));
return true;
}

5
src/backend/opencl/opencl.cmake
View file

@ -5,8 +5,6 @@ if (WITH_OPENCL)
set(HEADERS_BACKEND_OPENCL
src/backend/opencl/cl/OclSource.h
src/backend/opencl/generators/ocl_generic_cn_generator.cpp
src/backend/opencl/generators/ocl_vega_cn_generator.cpp
src/backend/opencl/interfaces/IOclRunner.h
src/backend/opencl/kernels/Cn0Kernel.h
src/backend/opencl/kernels/Cn1Kernel.h
@ -35,6 +33,8 @@ if (WITH_OPENCL)
set(SOURCES_BACKEND_OPENCL
src/backend/opencl/cl/OclSource.cpp
src/backend/opencl/generators/ocl_generic_cn_generator.cpp
src/backend/opencl/generators/ocl_vega_cn_generator.cpp
src/backend/opencl/kernels/Cn0Kernel.cpp
src/backend/opencl/kernels/Cn1Kernel.cpp
src/backend/opencl/kernels/Cn2Kernel.cpp
@ -78,6 +78,7 @@ if (WITH_OPENCL)
)
list(APPEND SOURCES_BACKEND_OPENCL
src/backend/opencl/generators/ocl_generic_cn_gpu_generator.cpp
src/backend/opencl/kernels/Cn00RyoKernel.cpp
src/backend/opencl/kernels/Cn1RyoKernel.cpp
src/backend/opencl/kernels/Cn2RyoKernel.cpp

19
src/backend/opencl/wrappers/OclDevice.cpp
View file

@ -47,11 +47,18 @@ typedef union
namespace xmrig {
#ifdef XMRIG_ALGO_CN_GPU
extern bool ocl_generic_cn_gpu_generator(const OclDevice &device, const Algorithm &algorithm, OclThreads &threads);
#endif
extern bool ocl_vega_cn_generator(const OclDevice &device, const Algorithm &algorithm, OclThreads &threads);
extern bool ocl_generic_cn_generator(const OclDevice &device, const Algorithm &algorithm, OclThreads &threads);
ocl_gen_config_fun generators[] = {
# ifdef XMRIG_ALGO_CN_GPU
ocl_generic_cn_gpu_generator,
# endif
ocl_vega_cn_generator,
ocl_generic_cn_generator
};
@ -149,18 +156,24 @@ xmrig::OclDevice::OclDevice(uint32_t index, cl_device_id id, cl_platform_id plat
}
size_t xmrig::OclDevice::freeMem() const
size_t xmrig::OclDevice::freeMemSize() const
{
return std::min<size_t>(OclLib::getDeviceUlong(id(), CL_DEVICE_MAX_MEM_ALLOC_SIZE), globalMem());
return std::min(maxMemAllocSize(), globalMemSize());
}
size_t xmrig::OclDevice::globalMem() const
size_t xmrig::OclDevice::globalMemSize() const
{
return OclLib::getDeviceUlong(id(), CL_DEVICE_GLOBAL_MEM_SIZE);
}
size_t xmrig::OclDevice::maxMemAllocSize() const
{
return OclLib::getDeviceUlong(id(), CL_DEVICE_MAX_MEM_ALLOC_SIZE);
}
xmrig::String xmrig::OclDevice::printableName() const
{
const size_t size = m_board.size() + m_name.size() + 64;

5
src/backend/opencl/wrappers/OclDevice.h
View file

@ -62,8 +62,9 @@ public:
OclDevice() = delete;
OclDevice(uint32_t index, cl_device_id id, cl_platform_id platform);
size_t freeMem() const;
size_t globalMem() const;
size_t freeMemSize() const;
size_t globalMemSize() const;
size_t maxMemAllocSize() const;
String printableName() const;
uint32_t clock() const;
void generate(const Algorithm &algorithm, OclThreads &threads) const;