Ecosystem/xmrig
Ecosystem/xmrig
1
0
Fork 0
mirror of https://github.com/xmrig/xmrig.git synced 2025-01-05 18:39:57 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge pull request #2961 from SChernykh/derohe_cuda

Browse source 
Dero HE (astrobwt/v2) CUDA config generator
This commit is contained in:
xmrig 2022-03-10 06:37:30 +07:00 committed by GitHub
commit afd79e7537
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
3 changed files with 23 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
src/backend/cuda/CudaConfig_gen.h
View file

@ -139,7 +139,14 @@ size_t inline generate<Algorithm::RANDOM_X>(Threads<CudaThreads> &threads, const
template<> template<>
size_t inline generate<Algorithm::ASTROBWT>(Threads<CudaThreads> &threads, const std::vector<CudaDevice> &devices) size_t inline generate<Algorithm::ASTROBWT>(Threads<CudaThreads> &threads, const std::vector<CudaDevice> &devices)
{ {
return generate(Algorithm::kASTROBWT, threads, Algorithm::ASTROBWT_DERO, devices); size_t count = 0;
if (!threads.isExist(Algorithm::ASTROBWT_DERO_2)) {
count += threads.move(Algorithm::kASTROBWT_DERO_2, CudaThreads(devices, Algorithm::ASTROBWT_DERO_2));
}
count += generate(Algorithm::kASTROBWT, threads, Algorithm::ASTROBWT_DERO, devices);
return count;
} }
#endif #endif

16
src/backend/cuda/runners/CudaAstroBWTRunner.cpp
View file

@ -32,11 +32,17 @@
constexpr uint32_t xmrig::CudaAstroBWTRunner::BWT_DATA_STRIDE; constexpr uint32_t xmrig::CudaAstroBWTRunner::BWT_DATA_STRIDE;
xmrig::CudaAstroBWTRunner::CudaAstroBWTRunner(size_t index, const CudaLaunchData &data) : xmrig::CudaAstroBWTRunner::CudaAstroBWTRunner(size_t index, const CudaLaunchData &data)
CudaBaseRunner(index, data) : CudaBaseRunner(index, data)
, m_algorithm(data.algorithm)
{ {
m_intensity = m_data.thread.threads() * m_data.thread.blocks(); m_intensity = m_data.thread.threads() * m_data.thread.blocks();
m_intensity -= m_intensity % 32; m_intensity -= m_intensity % 32;
// Dero HE has very fast blocks, so we can't use high intensity
if ((m_algorithm == Algorithm::ASTROBWT_DERO_2) && (m_intensity > 4096)) {
m_intensity = 4096;
}
} }
@ -58,10 +64,14 @@ bool xmrig::CudaAstroBWTRunner::set(const Job &job, uint8_t *blob)
size_t xmrig::CudaAstroBWTRunner::roundSize() const size_t xmrig::CudaAstroBWTRunner::roundSize() const
{ {
if (m_algorithm == Algorithm::ASTROBWT_DERO_2) {
return m_intensity;
}
constexpr uint32_t STAGE1_SIZE = 147253; constexpr uint32_t STAGE1_SIZE = 147253;
constexpr uint32_t STAGE1_DATA_STRIDE = (STAGE1_SIZE + 256 + 255) & ~255U; constexpr uint32_t STAGE1_DATA_STRIDE = (STAGE1_SIZE + 256 + 255) & ~255U;
const uint32_t BATCH2_SIZE = m_intensity; const uint32_t BATCH2_SIZE = static_cast<uint32_t>(m_intensity);
const uint32_t BWT_ALLOCATION_SIZE = BATCH2_SIZE * BWT_DATA_STRIDE; const uint32_t BWT_ALLOCATION_SIZE = BATCH2_SIZE * BWT_DATA_STRIDE;
const uint32_t BATCH1_SIZE = (BWT_ALLOCATION_SIZE / STAGE1_DATA_STRIDE) & ~255U; const uint32_t BATCH1_SIZE = (BWT_ALLOCATION_SIZE / STAGE1_DATA_STRIDE) & ~255U;

2
src/backend/cuda/runners/CudaAstroBWTRunner.h
View file

@ -27,6 +27,7 @@
#include "backend/cuda/runners/CudaBaseRunner.h" #include "backend/cuda/runners/CudaBaseRunner.h"
#include "base/crypto/Algorithm.h"
namespace xmrig { namespace xmrig {
@ -50,6 +51,7 @@ protected:
private: private:
size_t m_intensity = 0; size_t m_intensity = 0;
Algorithm m_algorithm;
}; };