GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » PANDA » PandaRoot » Bugs, Fixes, Releases » Bug in base I/O?
icon4.gif  Bug in base I/O? [message #10628] Wed, 05 May 2010 18:20 Go to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant

From: *to.infn.it
Dear all,
I would like to report about the problem I have trying to plot variables from the root shell, or from the TBrowser.

I can run a simple simulation macro, i.e. macro/mdt/sim_muo.C, and open the output root file test.root.

Once I open test.root, I can open a TBrowser and try to plot MCTrack variables, such as MCTrack.fPz or MCTrack.GetMomentum().Pz(), everything works. I can try also to write the commands in a shell:

root [2] cbmsim.Draw("MCTrack.fPz")
root [3] cbmsim.Draw("MCTrack.GetMomentum().fZ")
root [4] cbmsim.Draw("MCTrack.GetMomentum().Pz()")


They work.

I can try to do the same with MdtHit TCA, a TClonesArray made of PndMdtPoint inheriting from FairMCPoint<-FairBasePoint<-FairMultiLinkedData<-FairLinkedData<-TObject.

I can plot MdtPoint.fX, fine, but when I try to plot something using functions, i.e. MdtPoint.GetSector(), I have the following crash:

Toggle Spoiler


Considering that MCTrack is not inheriting from the FairXXXLinkData, and that I have the problem with all the objects inheriting from that class, and that before the link implementation I was able to plot everything fine (mdt code is not touched since a while), I would think it is a problem of this Link implementation. But it can be also that the problem comes from some "base" implementation different from the Link.

I ahve trid with different data object and I found always the same crash. If one want to plot something like theta or momentum which come from functions, it is quite... nasty.

I would call for some help from base developers.

Many thanks
Re: Bug in base I/O? [message #10632 is a reply to message #10628] Thu, 06 May 2010 11:03 Go to previous messageGo to next message
Radoslaw Karabowicz is currently offline  Radoslaw Karabowicz
Messages: 108
Registered: June 2004
Location: GSI
continuous participant
From: *gsi.de
Dear Stefano, what machine are you using? I do not see problem like you reported neither on MAC nor on LINUX. In the root I can do whatever I like:

root [2] cbmsim->Draw("MCTrack.GetMomentum().Pz()")
<TCanvas::MakeDefCanvas>: created default TCanvas with name c1
root [3] cbmsim->Draw("MCTrack.GetMomentum().Pz()","")
(Long64_t)13
root [4] cbmsim->Draw("MdtPoint.GetSector()","")
(Long64_t)100
root [5] cbmsim->Draw("MdtPoint.GetModule()","")
(Long64_t)100
root [6] cbmsim->Draw("MdtPoint.fX","")
(Long64_t)100
root [7] cbmsim->Draw("MdtPoint.GetModule()","")
(Long64_t)100
root [8]


yours,
radek

ps. what pandaroot version are you using?
Re: Bug in base I/O? [message #10636 is a reply to message #10632] Thu, 06 May 2010 11:59 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant

From: *to.infn.it
Hi,
I am using the last svn version, clean installation.

I have tried in my PC, SL4.7 gcc 3.4.6, and also at GSI, lxg0442 machine, normal etch32, gcc 4.1.2.

SL 4.7:

Toggle Spoiler


etch32:

Toggle Spoiler


In both the cases I have segmentation violations.
Which linux version are you using? As far as I have understood, even Tobias and Ralf can see the same crash. Ralf is able to recover it by playing with rootlogon and open file, but his trick does not work in my case.
Re: Bug in base I/O? [message #10638 is a reply to message #10628] Thu, 06 May 2010 12:29 Go to previous messageGo to next message
Ralf Kliemt is currently offline  Ralf Kliemt
Messages: 507
Registered: May 2007
Location: GSI, Darmstadt
first-grade participant

From: *hiskp.uni-bonn.de
Hi,

Running on a Mac keeps me safe, but on the Linux machine I get the same error.
Just to notify.

Ralf.
Re: Bug in base I/O? [message #10639 is a reply to message #10636] Thu, 06 May 2010 14:01 Go to previous messageGo to next message
Radoslaw Karabowicz is currently offline  Radoslaw Karabowicz
Messages: 108
Registered: June 2004
Location: GSI
continuous participant
From: *gsi.de
On my MAC (gcc 4.2.1, root 5.26.00) there is no problems at all.
In GSI I am running on lxi013 (etchnhalf, gcc 4.1.2, root 5.26.00), and here
I have problems only when I exit root after drawing histograms
(have to type .q 3 times, and many lines of errors). But still the
histograms were drawn.

yours
radek
Re: Bug in base I/O? [message #10640 is a reply to message #10639] Thu, 06 May 2010 14:22 Go to previous message
Radoslaw Karabowicz is currently offline  Radoslaw Karabowicz
Messages: 108
Registered: June 2004
Location: GSI
continuous participant
From: *gsi.de
OOOOKKKK,

I have to get back my previous statements. It is crashing also for me... MOSTLY.
Both on MAC and on LINUX.

I can only tell, when it does not crashes.

When I plot any variable or function of the MCTrack, and only then
anything from the MdtPoint, then it DOES NOT crashes.

If I plot a function from MdtPoint immediately after loading the file, it crashes.
If I plot a function from MdtPoint after plotting a variable from MdtPoint, it crashes.
If I plot a function from MdtPoint after plotting a variable from MCTrack, it crashes.

I am lost.

yours
radek
Previous Topic: cannot control submit job in batch
Next Topic: Dashboard not killing timeout tests
Goto Forum:
  


Current Time: Sat May 25 18:39:34 CEST 2024

Total time taken to generate the page: 0.01033 seconds