GitLab

Fixes #56 - OpalRing components throw segmentation fault and #59 - Unit tests throw segmentation fault; partial implementation of #68 DumpEmFields

 moving startTimer/stopTimer outside loop over particles and - avoiding string
 comparison inside loops over particles

ParallelTTracker. Particle-Matter interaction doesn't work either at the moment.

I hope that I didn't create too much of a mess when merging!

In case of a multi-bunch simulation with

MBMODE=FORCE (in OPAL input file)

the npart_mb, q_mb, p_mb, m_mb and r_mb are not set. Leading to a segfault.
I did as provisional fix in ParallelCyclotronTracker::Tracker_Generic()
(A 10 bunch simulation worked without errors).

modified:   src/Algorithms/ParallelCyclotronTracker.cpp

allow 0 in dump frequency (frequ is maxint) and supress h5 writes of number of partcles are 1 and 2 in OPAL_cycl

- ParallelCyclotronTracker::Tracker_RK4(): add missing curly brackets

Tried to merge master and local changes to ParallelCyclotronTracker and MGPoissonSolver...hope it worked :) -DW

Revert "In case a distribution has OFFSET, in OPAL-cycl the referernce particle has to be adjusted."

This reverts commit 20d8e10d.

Need this for OPAL-t, OPAL-cycl coupling in lieu of gantry simulations (Jozef).

fixing restart: can now (again) restart from any step in any H5Part file that was written for/with OPAL; SDDS file and load balance file are rewound to an s-position less or equal as in the H5file

removing 'using namespace', 'using somenamespace::function' and 'using somenamespace::class' from headerfiles

Brought back saving in local frame for cyclotron tracker. Reverted the way mean E is calculated in PartBunch. Hopefully, restarts in local as well as in global frame now work in the cyclotron tracker.

- moving the following classes over from OPAL to CLASSIC
  o OpalField       -> ClassicField
  o OpalRing        -> Ring
  o OpalRingSection -> RingSection
  o LossDataSink
- clean up
- splitting Options into classic and opal options

This allows to specify collimator losses as an objective. CHK wrote the original part, I did 
minor modifications. I tested with BeamLine-1 from the regressiontests so far only with one core.
 

- putting own header file first
  o for reason see http://www.cplusplus.com/forum/articles/10627/#msg49679 and
                   http://www.cplusplus.com/forum/articles/10627/#msg49680