This is the list of configure variables to control a Score-P measurement.
SCOREP_ENABLE_PROFILING
- Enable profiling
- Type:
- Boolean
- Default:
- true
SCOREP_ENABLE_TRACING
- Enable tracing
- Type:
- Boolean
- Default:
- false
SCOREP_VERBOSE
- Be verbose
- Type:
- Boolean
- Default:
- false
SCOREP_TOTAL_MEMORY
- Total memory in bytes per process for the measurement system
- Type:
- Number with size suffixes
- Default:
- 16000k
SCOREP_PAGE_SIZE
- Memory page size in bytes
- Type:
- Number with size suffixes
- Default:
- 8k
TOTAL_MEMORY will be split up into pages of size PAGE_SIZE.
SCOREP_EXPERIMENT_DIRECTORY
- Name of the experiment directory
- Type:
- String
- Default:
When no experiment name is given (the default)Score-P names the
experiment directory `scorep-measurement-tmp' and renames this
after a successful measurement to a generated name based on the
current time.
SCOREP_OVERWRITE_EXPERIMENT_DIRECTORY
- Overwrite an existing experiment directory
- Type:
- Boolean
- Default:
- true
If you specified a specific experiment directory name, but this name
is already given, you can force overwriting it with this flag.
The previous experiment directory will be renamed.
SCOREP_MACHINE_NAME
- The machine name used in profile and trace output.
- Type:
- String
- Default:
- Linux
We suggest using a unique name, e.g., the fully qualified domain name.
The default machine name was set at configure time (see the INSTALL
file for customization options).
SCOREP_EXECUTABLE
- Executable of the application
- Type:
- String
- Default:
File name, preferrrably with full path, of the application's executable.
It is used for evaluating the symbol table of the application, which is
required by some compiler adapters.
SCOREP_DEBUG_UNIFY
- Writes the pre-unified definitions also in the local definition trace files.
- Type:
- Boolean
- Default:
- true
SCOREP_PROFILING_TASK_EXCHANGE_NUM
- Number of foreign task objects that are collected before they are put into the common task object exchange buffer.
- Type:
- Number
- Default:
- 1024
The profiling creates a record for every task instance that is running. To avoid locking, the required memory is taken from a preallocated memory block. Each thread has its own memory block. On task completion, the created object can be reused by other tasks. However, if tasks migrate, the data structure migrates with them. Thus, if there is an imbalance in the migration from a source thread that starts the execution of tasks towards a sink thread that completes the tasks, the source thread may continually creating new task objects while in the sink, released task objects are collected. Thus, if the sink collected a certain number of tasks it should trigger a backflow of its collected task objects. However, this requires locking which should be avoided as much as possible. Thus, we do not want the locking happen on every migrated task, but only if a certain imbalance occurs. This environment variable determines the number of igrated task instances that must be collected before the backflow is triggered.
SCOREP_PROFILING_MAX_CALLPATH_DEPTH
- Maximum depth of the calltree
- Type:
- Number
- Default:
- 30
SCOREP_PROFILING_BASE_NAME
- Base for construction of the profile filename
- Type:
- String
- Default:
- profile
String which is used as based to create the filenames for the profile files.
SCOREP_PROFILING_FORMAT
- Profile output format
- Type:
- Set
- Default:
- default
Sets the output format for the profile.
The following formats are supported:
none: No profile output. This does not disable profile recording.
tau_snapshot: Tau snapshot format
cube4: Stores the sum for every metric per callpath in Cube4 format.
cube_tuple: Stores an extended set of statistics in Cube4 format.
default: Default format. If Cube4 is supported, Cube4 is the default
else the Tau snapshot format is default
SCOREP_PROFILING_ENABLE_CLUSTERING
- Enable clustering
- Type:
- Boolean
- Default:
- true
SCOREP_PROFILING_CLUSTER_COUNT
- maximum cluster count for iteration clustering.
- Type:
- Number with size suffixes
- Default:
- 64
maximum cluster count for iteration clustering.
SCOREP_PROFILING_CLUSTERING_MODE
- Specifies the level of strictness when comparing call trees for equivalence.
- Type:
- Number with size suffixes
- Default:
- 1
Specifies the level of strictness when comparing call trees for equivalence.
SCOREP_PROFILING_CLUSTERED_REGION
- Name of the clustered region
- Type:
- String
- Default:
The clustering can only cluster one dynamic region. If more than one dynamic region are defined by the user, the region is clustered which is exited first. If another region should be clustered instead you can specify the region name in this variable. If the variable is unset or empty, the first exited dynamic region is clustered.
SCOREP_PROFILING_ENABLE_CORE_FILES
- Write .core files if an error occured.
- Type:
- Boolean
- Default:
- false
If an error occures inside the profiling system, the profiling is disabled. For dubugging reasons, it might be feasable to get the state of the local local at these points. It is not recommended to enable this feature for large scale measurements.
SCOREP_TRACING_USE_SION
- Whether or not to use libsion as OTF2 substrate
- Type:
- Boolean
- Default:
- false
SCOREP_TRACING_MAX_PROCS_PER_SION_FILE
- Maximum number of processes that share one sion file (must be > 0)
- Type:
- Number
- Default:
- 1024
All processes are than evenly distributed over the number of needed files to fulfill this constraint. E.g., having 4 processes and setting the maximum to 3 would result in 2 files each holding 2 processes.
SCOREP_TRACING_COMPRESS
- Whether or not to compress traces with libz
- Type:
- Boolean
- Default:
- false
SCOREP_ONLINEACCESS_ENABLE
- Enable online access interface
- Type:
- Boolean
- Default:
- false
SCOREP_ONLINEACCESS_REG_PORT
- Online access registry service port
- Type:
- Number
- Default:
- 50100
SCOREP_ONLINEACCESS_REG_HOST
- Online access registry service hostname
- Type:
- String
- Default:
- localhost
SCOREP_ONLINEACCESS_BASE_PORT
- Base port for online access server
- Type:
- Number
- Default:
- 50010
SCOREP_ONLINEACCESS_APPL_NAME
- Application name to be registered
- Type:
- String
- Default:
- appl
SCOREP_FILTERING_FILE
- A file name which contain the filter rules
- Type:
- String
- Default:
SCOREP_METRIC_PAPI
- PAPI metric names to measure
- Type:
- String
- Default:
List of requested metric names that will be collected during program run.
SCOREP_METRIC_PAPI_PER_PROCESS
- PAPI metric names to measure per-process
- Type:
- String
- Default:
List of requested metric names that will be recorded only by first thread of a process.
SCOREP_METRIC_PAPI_SEP
- Separator of metric names
- Type:
- String
- Default:
- ,
Character that separates metric names in SCOREP_METRIC_PAPI and SCOREP_METRIC_PAPI_PER_PROCESS.
SCOREP_METRIC_RUSAGE
- Resource usage metric names to measure
- Type:
- String
- Default:
List of requested resource usage metric names that will be collected during program run.
SCOREP_METRIC_RUSAGE_PER_PROCESS
- Resource usage metric names to measure per-process
- Type:
- String
- Default:
List of requested resource usage metric names that will be recorded only by first thread of a process.
SCOREP_METRIC_RUSAGE_SEP
- Separator of resource usage metric names.
- Type:
- String
- Default:
- ,
Character that separates metric names in SCOREP_METRIC_RUSAGE and SCOREP_METRIC_RUSAGE_PER_PROCESS.
SCOREP_METRIC_PLUGINS
- Specify list of used plugins
- Type:
- String
- Default:
List of requested metric plugin names that will be used during program run.
SCOREP_METRIC_PLUGINS_SEP
- Separator of plugin names
- Type:
- String
- Default:
- ,
Character that separates plugin names in SCOREP_METRIC_PLUGINS.
SCOREP_SELECTIVE_CONFIG_FILE
- A file name which configures selective recording
- Type:
- String
- Default:
SCOREP_MPI_MAX_COMMUNICATORS
- Determines the number of concurrently used communicators per process
- Type:
- Number
- Default:
- 50
SCOREP_MPI_MAX_WINDOWS
- Determines the number of concurrently used windows for MPI one-sided communication per process
- Type:
- Number
- Default:
- 50
SCOREP_MPI_MAX_ACCESS_EPOCHS
- Maximum amount of concurrently active access or exposure epochs per process
- Type:
- Number
- Default:
- 50
SCOREP_MPI_MAX_GROUPS
- Maximum number of concurrently used MPI groups per process
- Type:
- Number
- Default:
- 50
SCOREP_MPI_ENABLE_GROUPS
- The names of the function groups which are measured
- Type:
- Set
- Default:
- default
Other functions are not measured.
Possible groups are:
all: All MPI functions
cg: Communicator and group management
coll: Collective functions
default: Default configuration
env: Environmental management
err: MPI Error handling
ext: External interface functions
io: MPI file I/O
misc: Miscellaneous
perf: PControl
p2p: Peer-to-peer communication
rma: One sided communication
spawn: Process management
topo: Topology
type: MPI datatype functions
xnonblock: Extended non-blocking events
xreqtest: Test events for uncompleted requests
SCOREP_MPI_ONLINE_ANALYSIS
- Enable online mpi wait states analysis
- Type:
- Boolean
- Default:
- false
SCOREP_CUDA_ENABLE
- CUDA measurement features
- Type:
- Set
- Default:
- no
Sets the CUDA measurement mode to capture:
runtime: CUDA runtime API
driver: CUDA driver API
kernel: CUDA kernels
kernel_serial: Serialized kernel recording.
kernel_counter: Fixed CUDA kernel metrics.
idle: GPU compute idle time
pure_idle: GPU idle time (memory copies are not idle)
memcpy: CUDA memory copies
sync: Record implicit and explicit CUDA synchronization
gpumemusage: Record CUDA memory (de)allocations as a counter
references: Record references between CUDA activities
flushatexit: Flush CUDA activity buffer at program exit
default/yes/1: CUDA runtime API and GPU activities
no: Disable CUDA measurement
SCOREP_CUDA_BUFFER
- Total memory in bytes for the CUDA record buffer
- Type:
- Number with size suffixes
- Default:
- 1M
SCOREP_CUDA_BUFFER_CHUNK
- Chunk size in bytes for the CUDA record buffer (ignored for CUDA 5.5 and earlier)
- Type:
- Number with size suffixes
- Default:
- 8k
SCOREP_OPENCL_ENABLE
- OpenCL measurement features
- Type:
- Set
- Default:
- no
Sets the OpenCL measurement mode to capture:
api: OpenCL runtime API
kernel: OpenCL kernels
memcpy: OpenCL buffer reads/writes
memusage: Record OpenCL memory (de)allocations as a counter
default/yes/1: OpenCL API and GPU activities
no: Disable OpenCL measurement
SCOREP_OPENCL_BUFFER
- Total memory in bytes for the OpenCL record buffer
- Type:
- Number with size suffixes
- Default:
- 1M
SCOREP_OPENCL_BUFFER_QUEUE
- Memory in bytes for the OpenCL command queue buffer
- Type:
- Number with size suffixes
- Default:
- 8k