▼Nlikwid | Likwid Fortran90 Module for embedding the Marker API into Fortran applications In the basic configuration the module is compiled with the Intel Fortran Compiler |
Clikwid_markerClose | Close the Likwid Marker API Close the Likwid Marker API and write measured results to temporary file for evaluation done by likwid-perfctr |
Clikwid_markerGetRegion | Get accumulated measurement results for a code region Get the accumulated data in the measurement group identified by regionTag for the current thread |
Clikwid_markerInit | Initialize the Likwid Marker API This routine initializes the Marker API for Fortran. It reads some environment commonly set by likwid-perfctr |
Clikwid_markerNextGroup | Setup performance counters for the next event set If multiple groups should be measured this function switches to the next group in a round robin fashion. Each call reprogramms the performance counters for the current CPU, |
Clikwid_markerRegisterRegion | Register a code region Initializes the hash table with an empty entry to reduce the overhead at likwid_markerStartRegion() |
Clikwid_markerStartRegion | Start the measurement for a code region Reads the currently running event set and store the results as start values. for the measurement group identified by regionTag |
Clikwid_markerStopRegion | Stop the measurement for a code region Reads the currently running event set and accumulate the difference between stop and start data in the measurement group identified by regionTag |
Clikwid_markerThreadInit | Add current thread to Likwid for Marker API measurements This routine adds the current thread to Likwid that it performs measurements for this thread. If using the daemon access mode, it starts a deamon for the current thread |
CAffinityDomain | The AffinityDomain data structure describes a single domain in the current system |
CAffinityDomains | The AffinityDomains data structure holds different count variables describing the various system layers |
CCacheLevel | Structure describing a cache level |
CCpuInfo | Structure with general CPU information |
CCpuTopology | Structure describing the topology of the HW threads in the system |
CHWThread | Structure with IDs of a HW thread |
CLikwid_Configuration | Structure holding values of the configuration file |
CNumaNode | CPUs in NUMA node and general information about a NUMA domain |
CNumaTopology | The NumaTopology structure describes all NUMA nodes in the current system |
CPerfmonCounter | Structure describing performance monitoring counter data |
CPerfmonEvent | Structure specifying an performance monitoring event |
CPerfmonEventOption | Structure specifying event/counter options and their value |
CPerfmonEventSet | Structure specifying an performance monitoring event group |
CPerfmonEventSetEntry | Structure specifying an performance monitoring event |
CPerfmonGroupSet | Structure specifying all performance monitoring event groups |
CPerfmonThread | Structure specifying thread to CPU relation |
CPowerData | Power measurement data for start/stop measurements |
CPowerDomain | Structure describing an RAPL power domain |
CPowerInfo | Information structure of CPU's power measurement facility |
CTimerData | Struct defining the start and stop time of a time interval |
CtreeNode | Structure of a tree node |
CTscCounter | Struct defining the start and stop time of a time interval |
CTurboBoost | Information structure of CPU's turbo mode |