• Introduction
• CVS Repositories
• Installing L1Calo Software
• Using L1Calo Software

Introduction

This page goes through the steps necessary to get the L1Calo software on your system.

Before starting, ensure that you have a fully configured Linux PC which includes all our prerequisites. We also have pages describing the system setup and a list of useful RPMs.
Sorry, those lists are now really out of date.

One of the main prerequisites is the ATLAS Online and DataFlow Software which must be installed and configured on your system. Note that in a distributed system, the Online software requires that at least some of its files are available to all nodes via a common filesystem.

Another important prerequisite is CMT which we use to build our software.

Packages and CVS Repositories

All our packages are presently held in a CVS repository at CERN. You will need to ask for an account on our CVS server to access them.

Installation

First, establish a work area for installing the software. We normally use ${HOME}/l1calo. If your default shell is bash you must use this work area.

Basic Environment

Before anything else you need to

• setup access to our CVS repository at CERN:
  export CVSROOT=:kserver:atdaq-sw.cern.ch:/local/reps/atlastdaq
kinit yourusername@CERN.CH
• source the CMT setup script

Fetching the Scripts Package

A few useful scripts are available as a package from our CERN repository. Getting these first will create the l1calo directory for you.

  cd
  cvs co LVL1/l1calo/scripts
  ln -s LVL1/l1calo l1calo

Fetching L1Calo Packages

Our software consists of a large number of CMT packages. For convenience, there is a little script to get them all. The -h option below gets the HEAD version of each one. To get the latest tagged version, omit the -h.

  cd ${HOME}/l1calo
  ./scripts/getallpkg -h

After the above, you should find a number of packages have been checked out into your working directory.

Customising Login Scripts

Before building or using the L1Calo software, you need to customise your login script(s), eg .bashrc.

You will find a small example .bashrc file in ${HOME}/l1calo/L1CaloRelease/data/shell. You should merge this with any existing ${HOME}/.bashrc script you have. You may need more work if you use another shell.

The essential code in .bashrc just sources another file .l1calo_bashrc which sets some environment variables including the CVSROOT and sources scripts to setup CMT, the Online software and the L1Calo software.

After customising your ${HOME}/.bashrc, you should logout and login again to ensure everything is setup correctly. NB new shells in KDE (and GNOME?) do not run your login scripts so you really need to logout or source the ${HOME}/.bashrc script by hand.

Building L1Calo Packages

One advantage of CMT is its ability to describe the "use" relationships between all the packages and it is convenient to create a simple "release" package which uses all our other packages. Our version is the L1CaloRelease package. We can use this to build all the others.

  cd ${HOME}/l1calo/L1CaloRelease/cmt
  cmt broadcast cmt config
  cmt broadcast buildpkg       # NB this will take some time...

The buildpkg script is a convenient wrapper for make; make install. It can take a number of arguments and is described in more detail here.

Using L1Calo Software

Once the cmt broadcast buildpkg command has finished (without error) you should have the complete L1Calo software installed in your area and ready for use.

However if you are installing the software on a new system, there will undoubtedly be further customisation required. Eg you need to create/modify a database describing your local system. Unfortunately that is beyond the scope of this web page.