|
| #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 | qudaMemcpyAsync(dst, src, count, kind, stream) ::quda::qudaMemcpyAsync_(dst, src, count, kind, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaMemcpy2D(dst, dpitch, src, spitch, width, height, kind) |
| |
| #define | qudaMemcpy2DAsync(dst, dpitch, src, spitch, width, height, kind, stream) |
| |
| #define | qudaMemset(ptr, value, count) ::quda::qudaMemset_(ptr, value, count, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaMemset2D(ptr, pitch, value, width, height) ::quda::qudaMemset2D_(ptr, pitch, value, width, height, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaMemsetAsync(ptr, value, count, stream) ::quda::qudaMemsetAsync_(ptr, value, count, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaMemset2DAsync(ptr, pitch, value, width, height, stream) |
| |
| #define | qudaMemPrefetchAsync(ptr, count, mem_space, stream) |
| |
| #define | qudaEventQuery(event) ::quda::qudaEventQuery_(event, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaEventRecord(event, stream) ::quda::qudaEventRecord_(event, stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaStreamWaitEvent(stream, event, flags) ::quda::qudaStreamWaitEvent_(stream, event, flags, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaEventSynchronize(event) ::quda::qudaEventSynchronize_(event, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaStreamSynchronize(stream) ::quda::qudaStreamSynchronize_(stream, __func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
| #define | qudaDeviceSynchronize() ::quda::qudaDeviceSynchronize_(__func__, quda::file_name(__FILE__), __STRINGIFY__(__LINE__)) |
| |
|
| qudaError_t | quda::qudaLaunchKernel (const void *func, const TuneParam &tp, void **args, qudaStream_t stream) |
| | Wrapper around cudaLaunchKernel. More...
|
| |
| template<typename T , typename... Arg> |
| qudaError_t | quda::qudaLaunchKernel (T *func, const TuneParam &tp, qudaStream_t stream, const Arg &... arg) |
| | Templated wrapper around qudaLaunchKernel which can accept a templated kernel, and expects a kernel with a single Arg argument. More...
|
| |
| 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 or driver API equivalent. More...
|
| |
| void | quda::qudaMemcpyAsync_ (void *dst, const void *src, size_t count, cudaMemcpyKind kind, const qudaStream_t &stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemcpyAsync or driver API equivalent. More...
|
| |
| void | quda::qudaMemcpy2D_ (void *dst, size_t dpitch, const void *src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemcpy2DAsync or driver API equivalent. More...
|
| |
| void | quda::qudaMemcpy2DAsync_ (void *dst, size_t dpitch, const void *src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind, const qudaStream_t &stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemcpy2DAsync or driver API equivalent. More...
|
| |
| void | quda::qudaMemset_ (void *ptr, int value, size_t count, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemset or driver API equivalent. More...
|
| |
| void | quda::qudaMemset2D_ (void *ptr, size_t pitch, int value, size_t width, size_t height, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemset2D or driver API equivalent. More...
|
| |
| void | quda::qudaMemsetAsync_ (void *ptr, int value, size_t count, const qudaStream_t &stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemsetAsync or driver API equivalent. More...
|
| |
| void | quda::qudaMemset2DAsync_ (void *ptr, size_t pitch, int value, size_t width, size_t height, const qudaStream_t &stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemsetAsync or driver API equivalent. More...
|
| |
| void | quda::qudaMemPrefetchAsync_ (void *ptr, size_t count, QudaFieldLocation mem_space, const qudaStream_t &stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaMemPrefetchAsync or driver API equivalent. More...
|
| |
| bool | quda::qudaEventQuery_ (cudaEvent_t &event, const char *func, const char *file, const char *line) |
| | Wrapper around cudaEventQuery or cuEventQuery with built-in error checking. More...
|
| |
| void | quda::qudaEventRecord_ (cudaEvent_t &event, qudaStream_t stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaEventRecord or cuEventRecord with built-in error checking. More...
|
| |
| void | quda::qudaStreamWaitEvent_ (qudaStream_t stream, cudaEvent_t event, unsigned int flags, const char *func, const char *file, const char *line) |
| | Wrapper around cudaStreamWaitEvent or cuStreamWaitEvent with built-in error checking. More...
|
| |
| void | quda::qudaEventSynchronize_ (cudaEvent_t &event, const char *func, const char *file, const char *line) |
| | Wrapper around cudaEventSynchronize or cuEventSynchronize with built-in error checking. More...
|
| |
| void | quda::qudaStreamSynchronize_ (qudaStream_t &stream, const char *func, const char *file, const char *line) |
| | Wrapper around cudaStreamSynchronize or cuStreamSynchronize with built-in error checking. More...
|
| |
| void | quda::qudaDeviceSynchronize_ (const char *func, const char *file, const char *line) |
| | Wrapper around cudaDeviceSynchronize or cuDeviceSynchronize with built-in error checking. 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_api.h.
