GSI Forum: Bugs, Fixes, Releases » segmentation violation in reco macro

Home » PANDA » PandaRoot » Bugs, Fixes, Releases » segmentation violation in reco macro

Show: Today's Messages :: Polls :: Message Navigator

segmentation violation in reco macro [message #13610]

Fri, 15 June 2012 13:11

MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany

first-grade participant

From: *physik.uni-giessen.de

Dear all,

after having set nmaxMvdPixelHitsInTrack from 30 to 60 in sttmvdtracking/PndSttMvdTracking.h I encountered another problem when running the reco macro from macro/pid in my simulation of X(3872) -> J/\psi \pi^+ \pi^- with the corrected VVpipi decay model on PandaRoot revision 15615 (Scientific Linux CERN SLC release 5.5 (Boron), 64 bit, fairroot: jan12, root session called with "root -q run_reco_sttcombi.C &> logreco2.log" in a screen session).

I have created a separate topic for this problem as I believe that this problem is not directly related to the nmaxMvdPixelHitsInTrack in the other thread.

Toggle Spoiler

[Updated on: Fri, 15 June 2012 18:38]

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13611 is a reply to message #13610]

Fri, 15 June 2012 13:38

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: 31.189.23*

Are you able to reproduce it, starting few events before?

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13612 is a reply to message #13611]

Fri, 15 June 2012 13:44

MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany

first-grade participant

From: *physik.uni-giessen.de

Thank you for your reply, Stefano.

I will try as you suggested. Is it correct to use fRun->Run(700, nEvents); in order to start with event 700?

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13615 is a reply to message #13612]

Fri, 15 June 2012 14:06

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *to.infn.it

Yes, try fRun->Run(700, 800) and without storing the log output ( >>> xxx.txt), simply catch what gdb threws.

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13616 is a reply to message #13615]

Fri, 15 June 2012 18:37

MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany

first-grade participant

From: *physik.uni-giessen.de

I have just tried it. It seems that the messages that I posted earlier are unrelated to the segmentation violation.

This time I got:

Toggle Spoiler

As a remark on the side:

When I put

fRun->Run(720, nEvents);

fRun loops through events 0 to 719. Is this meant to be that way or a bug?

Best regards,

Martin

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13617 is a reply to message #13616]

Fri, 15 June 2012 18:52

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: 31.189.23*

What did you put as nEvents? It should me the maximum event i.e. 800

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13618 is a reply to message #13617]

Fri, 15 June 2012 19:18

MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany

first-grade participant

From: *physik.uni-giessen.de

nEvents is set to 0.

I believed that represents "run until you reach the last event" (at least that is what fRun does when I call it with fRun->Run(0, nEvents); with nEvents set to 0).

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13619 is a reply to message #13618]

Fri, 15 June 2012 20:27

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *24-79-r.retail.telecomitalia.it

No, it does (firstEvent, lastEvent).
Please try.

Report message to a moderator

Re: segmentation violation in reco macro -- GEANT related? [message #13620 is a reply to message #13619]

Sun, 17 June 2012 15:09

MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany

first-grade participant

From: *physik.uni-giessen.de

You are right, it works correctly when the second argument is not equal to 0. I was just wondering about the behavior in case it actually is 0, but now that I know the behavior, it is not a problem for me any more.

Coming back to the crash that I encountered, it seems that on Friday I was not patient enough for root to give me the full error message. I have just tried to run the events 750 to 760 again and I get:

Toggle Spoiler

Maybe it would be interesting to investigate this crash, but it seems to be caused by me setting nmaxMvdPixelHitsInTrack from 30 to 60 in sttmvdtracking/PndSttMvdTracking.h. When I change this constant back to 30, there is no crash in event 752. (However, unfortunately I get a different crash in event 490.)

Report message to a moderator

Previous Topic:	Crash in reco macro: "PndSttMvdTracking, nMvdPixelHitsinTrack[4] is 58and it is > nmaxMvdPix
Next Topic:	[FIXED] Possibly a bug in MC Truth Forwarding (?) - no bug after all [SOLVED]

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Dec 21 12:51:24 CET 2024

Total time taken to generate the page: 0.00878 seconds