| segmentation fault during digitization (wrong mapper?) [message #10727] |
Fri, 21 May 2010 18:24  |
Manuel Zambrana
Messages: 3
Registered: May 2010
|
occasional visitor |
From: *gsi.de
|
|
Dear all,
after a successful simulation test, we have tried digitization. We have used "case 17" for the emc geometry and we get segmentation fault when we have particles in the forward direction. We think it is a problem related to the mapper of the forward endcap or shashlyk. We used BoxGenerator and shoot photons in different directions. When we shoot in the backward region(no hits in forward endcap or shashlyk calorimeter) there is no error. When we shoot in the forward region we get the error.
Could anybody take a look to test, and eventually fix, the mapper used by the latest emc geometry?
Thank you very much.
cheers,
manuel
ps. this is an example of the message we get in the log file when problems are found:
===========================================================
#5 0x00007f87c6aba0aa in PndEmcTwoCoordIndex::XCoord (this=0x0)
at /home/zambrana/work/software/trunk/emc/EmcTools/PndEmcTwoCoordIndex.h:59
#6 0x00007f87c6ab9c0a in PndEmcDigi (this=0xd97b0a0, trackid=0, id=509000021, energy=0.00731491204, time=18,
hitIndex=109) at /home/zambrana/work/software/trunk/emc/EmcData/PndEmcDigi.cxx:54
#7 0x00007f87c6ad7cd6 in PndEmcWaveformToDigi::Exec (this=0x1862ee0, opt=0x7f87c88d9290 "")
at /home/zambrana/work/software/trunk/emc/EmcDigi/PndEmcWaveformToDigi.cxx: 133
#8 0x00007f87d3b86f90 in TTask::ExecuteTasks ()
from /local/raid0/work/khaneftd/fairsoft/tools/root/lib/libCore.so.5.26
#9 0x00007f87d3b862cb in TTask::ExecuteTask ()
from /local/raid0/work/khaneftd/fairsoft/tools/root/lib/libCore.so.5.26
#10 0x00007f87c8849537 in FairRunAna::Run (this=0x78de10, Ev_start=0, Ev_end=10)
at /home/zambrana/work/software/trunk/base/FairRunAna.cxx:271
===========================================================
|
|
|
|
|
|
|
|
| Re: segmentation fault during digitization (wrong mapper?) [message #10736 is a reply to message #10727] |
Tue, 25 May 2010 14:54  |
Dmitry Khaneft
Messages: 75
Registered: January 2009
|
continuous participant |
From: *gsi.de
|
|
Dear all,
I met the same problem even with case 15. But in addition there is another problem:
====== EMC:: ConstructRootGeometry() m4a ===
=============================================
fgeoName3:: /u/dkhaneft/programs/trunk/geometry/emc_module4_StraightGeo24.4.root
File name Bw1= /u/dkhaneft/programs/trunk/geometry/emc_module4_StraightGeo24.4.root
You do not provide a ROOT file
Somehow pandaroot isn't able to load backward endcap geometry in case 15 but still can do it in other cases. I checked case 11,12 and 17.
P.S. pandaroot v.8554
[Updated on: Tue, 25 May 2010 14:55] Report message to a moderator |
|
|
|
|
|
|
|
|
|
| Re: segmentation fault during digitization (wrong mapper?) [message #10741 is a reply to message #10727] |
Tue, 25 May 2010 16:59 |
Dmitry Khaneft
Messages: 75
Registered: January 2009
|
continuous participant |
From: *gsi.de
|
|
I looked for cout command and I have found it only in ConstructGeometry() function in PndEmc.cxx. I think that during an introduction of cases 16 and 17 some part of code has been changed. Probably ConstructGeometry() is called somewhere one more time without specifying a name of the detector which cause the message during simulation.
It should really tell you that something going wrong with your geometry but normally it should't print this message.
[Updated on: Tue, 25 May 2010 17:01] Report message to a moderator |
|
|
|
|
|
GSI Forum
Search
Help
Members
Login
Home
