|
QUDA
0.9.0
|


Go to the source code of this file.
Namespaces | |
| quda | |
Macros | |
| #define | STRINGIFY__(x) #x |
| #define | __STRINGIFY__(x) STRINGIFY__(x) |
| #define | qudaMemcpy(dst, src, count, kind) ::quda::qudaMemcpy_(dst, src, count, kind, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)); |
| #define | STRINGIFY__(x) #x |
| #define | __STRINGIFY__(x) STRINGIFY__(x) |
| #define | qudaMemcpyAsync(dst, src, count, kind, stream) ::quda::qudaMemcpyAsync_(dst, src, count, kind, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)); |
| #define | STRINGIFY__(x) #x |
| #define | __STRINGIFY__(x) STRINGIFY__(x) |
| #define | qudaMemcpy2DAsync(dst, dpitch, src, spitch, width, height, kind, stream) ::quda::qudaMemcpy2DAsync_(dst, dpitch, src, spitch, width, height, kind, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)); |
Functions | |
| void | quda::qudaMemcpy_ (void *dst, const void *src, size_t count, cudaMemcpyKind kind, const char *func, const char *file, const char *line) |
| Wrapper around cudaMemcpy used for auto-profiling. Do not call directly, rather call macro below which will grab the location of the call. More... | |
| void | quda::qudaMemcpyAsync_ (void *dst, const void *src, size_t count, cudaMemcpyKind kind, const cudaStream_t &stream, const char *func, const char *file, const char *line) |
| Wrapper around cudaMemcpyAsync or driver API equivalent Potentially add auto-profiling support. More... | |
| void | quda::qudaMemcpy2DAsync_ (void *dst, size_t dpitch, const void *src, size_t spitch, size_t width, size_t hieght, cudaMemcpyKind kind, const cudaStream_t &stream, const char *func, const char *file, const char *line) |
| Wrapper around cudaMemcpy2DAsync or driver API equivalent Potentially add auto-profiling support. More... | |
| cudaError_t | quda::qudaLaunchKernel (const void *func, dim3 gridDim, dim3 blockDim, void **args, size_t sharedMem, cudaStream_t stream) |
| Wrapper around cudaLaunchKernel. More... | |
| cudaError_t | quda::qudaEventQuery (cudaEvent_t &event) |
| Wrapper around cudaEventQuery or cuEventQuery. More... | |
| cudaError_t | quda::qudaEventRecord (cudaEvent_t &event, cudaStream_t stream=0) |
| Wrapper around cudaEventRecord or cuEventRecord. More... | |
| cudaError_t | quda::qudaStreamWaitEvent (cudaStream_t stream, cudaEvent_t event, unsigned int flags) |
| Wrapper around cudaEventRecord or cuEventRecord. More... | |
| cudaError_t | quda::qudaStreamSynchronize (cudaStream_t &stream) |
| Wrapper around cudaStreamSynchronize or cuStreamSynchronize. More... | |
| cudaError_t | quda::qudaEventSynchronize (cudaEvent_t &event) |
| Wrapper around cudaEventSynchronize or cuEventSynchronize. More... | |
| cudaError_t | quda::qudaDeviceSynchronize () |
| Wrapper around cudaDeviceSynchronize or cuDeviceSynchronize. More... | |
| void | quda::printAPIProfile () |
| Print out the timer profile for CUDA API calls. More... | |
Wrappers around CUDA API function calls allowing us to easily profile and switch between using the CUDA runtime and driver APIs.
Definition in file quda_cuda_api.h.
| #define __STRINGIFY__ | ( | x | ) | STRINGIFY__(x) |
Definition at line 41 of file quda_cuda_api.h.
| #define __STRINGIFY__ | ( | x | ) | STRINGIFY__(x) |
Definition at line 41 of file quda_cuda_api.h.
| #define __STRINGIFY__ | ( | x | ) | STRINGIFY__(x) |
Definition at line 41 of file quda_cuda_api.h.
| #define qudaMemcpy | ( | dst, | |
| src, | |||
| count, | |||
| kind | |||
| ) | ::quda::qudaMemcpy_(dst, src, count, kind, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)); |
Definition at line 32 of file quda_cuda_api.h.
Referenced by quda::cublas::BatchInvertMatrix(), quda::blas::copy_ns::copy(), quda::cudaCloverField::copy(), quda::cudaGaugeField::copy(), quda::cpuGaugeField::copy(), quda::Transfer::createGeoMap(), quda::GaugeField::exchange(), quda::ColorSpinorField::exchange(), quda::cudaGaugeField::exchangeExtendedGhost(), quda::cudaGaugeField::exchangeGhost(), quda::cudaGaugeField::injectGhost(), loadCloverQuda(), quda::cudaColorSpinorField::loadSpinorField(), multiReduceLaunch(), reduceLaunch(), quda::cudaCloverField::saveCPUField(), quda::cudaGaugeField::saveCPUField(), and quda::cudaColorSpinorField::saveSpinorField().
| #define qudaMemcpy2DAsync | ( | dst, | |
| dpitch, | |||
| src, | |||
| spitch, | |||
| width, | |||
| height, | |||
| kind, | |||
| stream | |||
| ) | ::quda::qudaMemcpy2DAsync_(dst, dpitch, src, spitch, width, height, kind, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)); |
Definition at line 42 of file quda_cuda_api.h.
Referenced by quda::cudaColorSpinorField::sendGhost().
| #define qudaMemcpyAsync | ( | dst, | |
| src, | |||
| count, | |||
| kind, | |||
| stream | |||
| ) | ::quda::qudaMemcpyAsync_(dst, src, count, kind, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)); |
Definition at line 37 of file quda_cuda_api.h.
Referenced by quda::cudaColorSpinorField::sendGhost(), and quda::cudaColorSpinorField::unpackGhost().
Definition at line 40 of file quda_cuda_api.h.
Definition at line 40 of file quda_cuda_api.h.
Definition at line 40 of file quda_cuda_api.h.
1.8.14