GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » PANDA » PandaRoot » Bugs, Fixes, Releases » Segfault in PndMvdStripClusterBuilder
Segfault in PndMvdStripClusterBuilder [message #10035] Mon, 25 January 2010 10:52 Go to next message
Ralf Kliemt is currently offline  Ralf Kliemt
Messages: 507
Registered: May 2007
Location: GSI, Darmstadt
first-grade participant

From: *cb.uni-bonn.de
Hi everybody,

I get a quite evil segfault in my code which did not appear last year.
Maybe someone more experienced with stl can help here.

I run the 3 "runMvd...C" macros from pandaroot/macros/mvd. on the third I get the following messages from gdb. Note that I checked all variables with a cout in the lines before.
Quote:

**Sarting PndMvdStripClusterTask::Exec()**
PndMvdStripClusterBuilder::AddDigi(): Adding a Digi with name 1_1/40_0/66_1/65_1/63_1/62_6/61_1/60_1/57_1/56_1/
time 0, number 355, side 0 forom point 0

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: 13 at address: 0x0000000000000000
0x00007fff80509a24 in std::string::compare ()


Toggle Spoiler



Thanks for any help here, Ralf.

Re: Segfault in PndMvdStripClusterBuilder [message #10043 is a reply to message #10035] Mon, 25 January 2010 20:07 Go to previous messageGo to next message
Johan Messchendorp is currently offline  Johan Messchendorp
Messages: 693
Registered: April 2007
Location: University of Groningen
first-grade participant

From: *xs4all.nl
Hi Ralf,

I was just playing with your macros, and I already got stuck with the digitization macro. It crashes with a floating point exception, which can easily be traced back to line 14 in PndMvdCalcFePixel.cxx where there is a division by zero (fNrows=0). My educated guess would be that something goes wrong with reading/writing the digitisation parameters (mvd.digi.par?). Could this be related somehow with your segfault.... ?!?

Greetings,

Johan.

Quote:

Frontend type name is = APV25
Clusterfinder Mode = 0
Clusterfinder Search Radius: Channels = 2
Clusterfinder Search Radius: Time = 0
Top/Bottom Charge correlation cut = 2.50941e-319
-I- PndMvdNoiseProducer: Intialisation successfull

*** Break *** floating point exception



===========================================================
There was a crash.
This is the entire stack trace of all threads:
===========================================================
#0 0x00007fd93192d5e5 in waitpid () from /lib64/libc.so.6
#1 0x00007fd9318c9861 in ?? () from /lib64/libc.so.6
#2 0x00007fd9339fd64a in TUnixSystem::StackTrace() ()
from /opt/exp_soft/panda/fairroot/fairsoft_jan10/tools/root/lib/libCore.so.5. 26
#3 0x00007fd9339fe045 in TUnixSystem::DispatchSignals(ESignals) ()
from /opt/exp_soft/panda/fairroot/fairsoft_jan10/tools/root/lib/libCore.so.5. 26
#4 <signal handler called>
#5 0x00007fd9277efec8 in PndMvdCalcFePixel::CalcFEHits (this=0x7fff32bc7910)
at /home/panda/pandaroot/trunk/mvd/MvdDigi/PndMvdCalcFePixel.cxx:13
#6 0x00007fd9277fbb2f in PndMvdCalcFePixel::CalcFEHits (this=0x7fff32bc7910, SensorHits=
{<std::_Vector_base<PndMvdPixel, std::allocator<PndMvdPixel> >> = {_M_impl = {<std::allocator<PndMvdPixel>> = {<__gnu_cxx::new_allocator<PndMvdPixel>> = {<No data fields>}, <No data fields>}, _M_start = 0x7fff32bc7b90, _M_finish = 0x7fff32bc7910, _M_end_of_storage = 0xdb0210}}, <No data fields>}) at /home/panda/pandaroot/trunk/mvd/MvdDigi/PndMvdCalcFePixel.h:48

[Updated on: Mon, 25 January 2010 21:17]

Report message to a moderator

Re: Segfault in PndMvdStripClusterBuilder [message #10044 is a reply to message #10035] Mon, 25 January 2010 23:05 Go to previous messageGo to next message
Elwin Dijck
Messages: 16
Registered: June 2009
Location: Groningen, The Netherland...
occasional visitor
From: *mxp.dsl.internl.net
I don't know whether the errors are related, but I took a look at the segfault in runMvdReco.C. The crash occurs in PndMvdStripClusterBuilder::AddDigi(), where just some data is put into some nested std::maps. The stack trace showed that the memory of the map was somehow corrupted and the reason for this is that a PndMvdStripClusterBuilder object is used after it is deleted. Some extra debug output:

0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/63_1/62_5/54_3/52_4/49_1/48_1/"
0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/63_1/62_5/54_3/52_4/49_1/48_1/"
0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/63_1/62_5/54_3/52_4/49_1/48_1/"
0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/64_1/62_15/54_3/52_4/49_1/48_1/"
0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/64_1/62_15/54_3/52_4/49_1/48_1/"
0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/64_1/62_21/61_1/60_1/57_1/"
0xbfa10f0->DESTRUCTOR CALLED!
0xbfa10f0->AddDigi called for detName: "1_1/40_0/66_1/65_1/63_1/62_6/61_1/60_1/57_1/56_1/"

 *** Break *** segmentation violation
 Generating stack trace... etc.


I don't know why this happens exactly, but it looks like a bug in PndMvdStripClusterTask::Exec() from where AddDigi() is called.

Regards,
Elwin Dijck

[Updated on: Mon, 25 January 2010 23:06]

Report message to a moderator

Re: Segfault in PndMvdStripClusterBuilder [message #10046 is a reply to message #10043] Tue, 26 January 2010 10:22 Go to previous messageGo to next message
Johan Messchendorp is currently offline  Johan Messchendorp
Messages: 693
Registered: April 2007
Location: University of Groningen
first-grade participant

From: *KVI.nl
btw, Mohammad A. fixed the problem with the MvdDigi and the parameter file .... some mixup with Double_t and Int_t... The crash in reco still occurs.

Johan.

[Updated on: Tue, 26 January 2010 10:23]

Report message to a moderator

Re: Segfault in PndMvdStripClusterBuilder [message #10047 is a reply to message #10035] Tue, 26 January 2010 10:45 Go to previous message
Ralf Kliemt is currently offline  Ralf Kliemt
Messages: 507
Registered: May 2007
Location: GSI, Darmstadt
first-grade participant

From: *cb.uni-bonn.de
Hello,

Thanks a lot.
I fixed both issues.

Greetings, Ralf.
Previous Topic: libLheTrack.so: undefined symbol: _ZTV17PndMvdGeoHandling
Next Topic: update stable branch...
Goto Forum:
  


Current Time: Mon Dec 09 01:49:52 CET 2024

Total time taken to generate the page: 0.00754 seconds