1) Give write permission to own user on /opt:
$chown -R /opt
2) Install required packages (for a Ubuntu 11.04 system they are listed link)
3) Download the framework from this page
4) Unpack the tar.gz in a user directory
$tar xvf panda-0.9.0.tar.gz
5) Move into the PandA directory
$cd panda-0.9.0
6) In case flopoco library will be used run the following script:
$./prepare_ext.sh
7) Prepare the configuration scripts:
$make -f Makefile.init
8 ) Create the directory for the building:
$mkdir obj $cd obj
9) Configure PandA:
$../configure [parameters] --prefix=/opt/panda
Currently accepted parameters are:
- –enable-flopoco: add floating point support by exploiting FloPoCo Library link/
- –enable-xilinx: enable Xilinx RTL synthesis and simulation tools integration
- –enable-altera: enable Altera RTL synthesis tool integration
- –enable-design-compiler: enable Synopsys Design Compiler integration
- –enable-modelsim: enable Modelsim based simulation
- –with-mentor-license=<license-file/license-server>
- –enable-icarus: enable Icarus based simulation link
- –enable-debug: add debugging symbol to the binaries
10 ) Compile the tool:
$make
11) Install the tool:
$make install
At the end, if the installation has been successful, you can run the tool:
$/opt/panda/bin/bambu
which prints its help message.
Create documentation
Documentation of the PandA Project is based on Doxygen and can be created in different way. To create all the documentation of the project give the following command from the build root directory:
$ make documentation
The documentation browsing can start from the file: obj/doc/PANDA_html/index.html.
If for only a sub-project is required the documentation, please move to the associated directory and perform the previous command. The standard behavior of Doxygen has been modified in order to allow the separated building of each sub-project documentation.
Testcases
In the directory examples, you can find two different testcases, with the
bash scripts for their execution.
To run them, you have to perform the following command:
$cd examples/arf $./bambu.sh
or
$cd examples/crc $./bambu.sh
These scripts automatically creates subdirectories containing all the results
of the execution. You can start from them to build your own hardware
accelerators with bambu.