![]() |
Cube GUI User Guide
(CubeGUI 4.9, revision 6e5e012c)
Introduction in Cube GUI and its usage
|
When the plugin is launched, the setup tab is displayed. There, the user can decide whether to load a former measurement or to start a new one. It is important to note that the plugin is only usable when there exists an installed Score-P version, and the plugin must be started in the build environment of the application the user wants to measure.
When the user clicks on the "Load Recent Measurement" button, the most recently saved measurement will be loaded. This will return the user to the measurement at the same point where they exited the plugin during the previous session.
The plugin lists all measurements that were submitted as a job during previous usage of the measurement. It displays the job ID, the status of the job and options to load or delete the specific measurement. If the user selects the option to load the measurement, they will be taken to the Measurement tab, where they can access various options depending on the Job Status.
The status of a job can be one of the following:
If the status of a job is not available, the user should try resubmitting the job or checking for any errors that may have occurred.
When the user selects the "Start New Measurement" button, the setup tab layout will change to show the steps required to select a Score-P installation.
The plugin offers multiple ways to select a Score-P version. If the user has a Score-P version available on their system's path, it will be automatically detected and listed in the Score-P version found in PATH section. Alternatively, if the user wants to load a Score-P module, they can use the Find Score-P versions Button to search for and select the desired version. Finally, if the Score-P version is not automatically detected or available as a module, the user can manually browse the system directory using the Browse Score-P Button to select the appropriate version.
The plugin lists all detected Score-P versions and provides additional information about each version, including whether PAPI counter and libunwind are available and whether it is a usable configuration. A usable configuration implies that the MPI and/or the compiler used to build the Score-P version matches the MPI and compiler specified for building the user's application.
There are two drop down menus to select the MPI and Compiler used to build the program being measured.
The user can choose between the following options:
The user can choose between the following options:
If a Score-P installation was found in the user's PATH variable, the plugin lists it with providing additional information. It gives information whether PAPI counter and libunwinded are available in this installation and if it is a usable configuration.
The usable configuration status indicates whether the selected Score-P version is compatible with the MPI and compiler used to build the user's application. The status can have one of three values:
Next to the status message, an info icon is displayed. If the user hovers over this icon with their mouse, a tooltip will appear containing information about the MPI and compiler used to build the selected Score-P installation.
This button allows the plugin to search for Score-P modules and display all available versions. If no Score-P module is found, an error message is displayed to the user.
When the "Browse Score-P" button is pressed, a file dialog box will appear for the user to select an installed Score-P version. Only visible after Find Score-P versions Button is clicked.
When the user has selected an appropriate Score-P installation, the "Proceed" button becomes enabled. Clicking this button will move the user to the Instrumentation tab and the selected Score-P version is used for the measurement.
If the user needs assistance in detecting a Score-P version, they can click on the "Help" button. This opens a pop-up window with information on how to detect a Score-P version.
As next, see Instrumentation
![]() |
Copyright © 1998 Forschungszentrum Jülich GmbH,
Jülich Supercomputing Centre
Copyright © 2009–2015 German Research School for Simulation Sciences GmbH, Laboratory for Parallel Programming |