Project

General

Profile

Geant4 EM Physics Code Review Weekly Phone Meeting

Meeting

Date: October 25, 2013 at 10.00 AM US CDT (Fermilab local time)
Place: 1-866-740-1260 (ReadyTalk line) (Host : Krzysztof)

Participants:

Andrea Dotti
Krzysztof Genser
Soon Yung Jun
Boyana Norris

Discussion

(Summary by Krzysztof)

The reference release 10 is out. It does have some
changes in the em code (e.g. in the G4PhysicsVector a
protection in find bin functions and usage of g4log and g4exp
and others see release notes and the code itself)
so we should switch to it.
Soon will place it in the repository and is profiling it now.
The new reports should have the csv tables in them as well.
Note two new data files:
http://geant4.cern.ch/support/source/G4EMLOW.6.34.tar.gz
http://geant4.cern.ch/support/source/G4NEUTRONXS.1.4.tar.gz
(and a possibly the third one, if you do not have it yet)
http://geant4.cern.ch/support/source/G4NDL.4.4.tar.gz

Krzysztof and Soon started reviewing G4VEmProcess and G4VEnergyLossProcess
but those are a very large classes, so it will take some time to review them.

Soon looked at a possibility of using the Gooda profiler for the review, but concluded that
it was not contributing significantly new information.

He also looked the Geant4 coverity results as obtained at the code server at CERN, but since
those results are usually taken into account during standard Geant4 code maintenance
there was not much new information relevant for this review there.

Andrea suggested that we should use gooda to make sure that our
conclusions obtained with the sequential code still hold for the MT version.

(e.g. in g4physicsvector::value()the caching attempt is actually
causing higher cache misses)

The plan for next week is to work on the items from the todo
list from the previous weeks.

We shall meet again on Friday (10am Central).