hpcviewer is the presentation layer of HPCToolkit which is a suite of tools
for measurement and analysis of program performance.
It interactively presents program performance in a top-down fashion and also
visualizes trace data generated by hpcrun if the flag -t
is specified.
For static linked program, the variable environment HPCRUN_TRACE
has to be set.
If there is no GTK+ 3.20 or newer installed in the system, you may install it via spack
:
spack install gtkplus
spack load gtkplus
- Download and install Maven (if not available on the systems) at https://maven.apache.org/
- Recommended: install via spack
spack install maven; spack load maven
- Recommended: install via spack
Run the build script from the project root:
./build.sh
The script generates five hpcviewer-<release>-<platform>.[zip|tgz]
files:
Windows, Mac (x86_64 and Arm), and Linux (x86_64, ppcle64, and Arm).
-
untar
orunzip
the file based according to the platform. -
ONLY for Linux platform, need to run the installation script:
./install.sh <directory>
where <directory>
is the installation root for hpcviewer binary.
Build directly with the Maven script:
mvnw.cmd clean package
This will compile and create hpcviewer packages for 4 platforms: Linux x86_64 and ppcle64, Windows and Mac with Eclipse 4.30 (the default). Example of the output:
...
[INFO] Building zip: <hpcviewer.e4>/edu.rice.cs.hpcviewer.product/target/products/edu.rice.cs.hpcviewer-win32.win32.x86_64.zip
Unzip edu.rice.cs.hpcviewer-win32.win32.x86_64.zip
to another folder.
It isn't recommended to overwrite the existing folder.
Requirements:
- Recommended: Eclipse 2023.12 RCP or newer.
- Warning: May not work properly with older versions of Eclipse.
Recommended:
- Source code for hpcdata
- Source code for graphbuilder (math parser)
- Source code for remote database:
git clone https://gitlab.com/hpctoolkit/hpcdata
git clone https://gitlab.com/hpctoolkit/graphbuilder
- Start Eclipse
- Open the Import window via the menu File > Import
- Select the import wizard General > Existing Projects into Workspace and click Next >
- In Select root directory, select the directory where you have downloaded (the Git root)
- In the Options section of the window, activate Search for nested projects
- Click Finish
To run hpcviewer, it requires Eclipse bundles and some external libraries such as Nebula NatTable, Eclipse Collections, JCraft and SLF4J. The set of bundles that are available is defined by the bundles in the Eclipse workspace, and additionally the bundles in the active target platform The first Eclipse starts after the installation, the target platform only includes the bundles that are installed in the workspace which doesn't include the external libraries.
The bundles that hpcviewer needs are defined in a custom target platform definition project, which is located in the target.platform
directory:
- Open file
target-platform.target
intarget.platform project
. - Click the "Set as Active Target Platform" link at the top-right panel.
- Open product configuration
hpcviewer.product
atedu.rice.cs.hpcviewer.product
- To run: Click
Launch an Eclipse application
Recommended coding and formatting style: