mirror of https://github.com/QMCPACK/qmcpack.git
Merge branch 'develop' into cxx17-by-default
This commit is contained in:
commit
12725377de
|
@ -710,7 +710,7 @@ ENDIF(QMC_CUDA OR ENABLE_CUDA)
|
|||
|
||||
OPTION(USE_NVTX_API "Enable/disable NVTX regions in CUDA code." OFF)
|
||||
IF(USE_NVTX_API)
|
||||
IF(HAVE_CUDA)
|
||||
IF(HAVE_CUDA OR ENABLE_OFFLOAD)
|
||||
FIND_LIBRARY(NVTX_API_LIB
|
||||
NAME nvToolsExt
|
||||
HINTS ${CUDA_TOOLKIT_ROOT_DIR}
|
||||
|
@ -720,7 +720,7 @@ IF(USE_NVTX_API)
|
|||
ENDIF(NOT NVTX_API_LIB)
|
||||
MESSAGE("CUDA nvToolsExt library: ${NVTX_API_LIB}")
|
||||
LINK_LIBRARIES(${NVTX_API_LIB})
|
||||
ENDIF(HAVE_CUDA)
|
||||
ENDIF(HAVE_CUDA OR ENABLE_OFFLOAD)
|
||||
ENDIF(USE_NVTX_API)
|
||||
|
||||
#-------------------------------------------------------------------
|
||||
|
|
|
@ -45,6 +45,10 @@ void TimerType<CLOCK>::start()
|
|||
__itt_task_begin(manager->task_domain, __itt_null, parent_task, task_name);
|
||||
#endif
|
||||
|
||||
#ifdef USE_NVTX_API
|
||||
nvtxRangePushA(name.c_str());
|
||||
#endif
|
||||
|
||||
bool is_true_master(true);
|
||||
for (int level = omp_get_level(); level > 0; level--)
|
||||
if (omp_get_ancestor_thread_num(level) != 0)
|
||||
|
@ -90,6 +94,10 @@ void TimerType<CLOCK>::stop()
|
|||
__itt_task_end(manager->task_domain);
|
||||
#endif
|
||||
|
||||
#ifdef USE_NVTX_API
|
||||
nvtxRangePop();
|
||||
#endif
|
||||
|
||||
bool is_true_master(true);
|
||||
for (int level = omp_get_level(); level > 0; level--)
|
||||
if (omp_get_ancestor_thread_num(level) != 0)
|
||||
|
|
|
@ -28,6 +28,10 @@
|
|||
#include <ittnotify.h>
|
||||
#endif
|
||||
|
||||
#ifdef USE_NVTX_API
|
||||
#include <nvToolsExt.h>
|
||||
#endif
|
||||
|
||||
#define USE_STACK_TIMERS
|
||||
|
||||
namespace qmcplusplus
|
||||
|
|
Loading…
Reference in New Issue