Re: [ROOT] SegFault via G__defined_tagname in Root 3.10

From: Rene Brun (Rene.Brun@cern.ch)
Date: Fri Jan 09 2004 - 16:18:34 MET


Hi Volker,

This looks like a problem when building the dictionaries.
It could be that you moved to a new version of ROOT and not recompiled
some of your dictionaries or classes.
Also check the consistency of your PATH/LD_LIBRARY_PATH variables.

Rene brun

Volker Hejny wrote:
> 
> Dear Rene,
> Dear Rooters,
> 
> I have a strange problem and I'd like to ask for some advice,
> where to continue looking for the reason.
> 
> I had to face the problem when I was necessary to run our analysis
> software on Debian 3.0 (Woody) using gcc-2.95 instead of the Debian
> unstable or a recent SuSE with gcc 3.x. On the latter systems everythings
> runs without problem. We are using Root 3.10/02 as default in order
> to gain from recent developments. However, using Root version 3.05/07
> the problem described below does _not_ occur.
> 
> On Debian 3.0 all stand-alone programs in our package crash
> before entering the main function with the following traceback:
> 
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 1024 (LWP 9893)]
> 0x4195cda1 in strstr () from /lib/libc.so.6
> #1  0x406e78a3 in G__defined_tagname (tagname=0x0, noerror=2)
>     at cint/src/struct.c:425
> 425       }
> Current language:  auto; currently c
> #2  0x406e83c9 in G__search_tagname (tagname=0x0, type=0)
>     at cint/src/struct.c:748
> 748       i = G__defined_tagname(tagname,2);
> #3  0x406a7e8e in G__get_linked_tagnum (p=0x404988cc)
>     at cint/src/newlink.c:1780
> 1780         p->tagnum = G__search_tagname(p->tagname,p->tagtype);
> #4  0x402b04d8 in G__cpp_setup_tagtableG__Base1 ()
>     from /home/hejny/root/lib/libCore.so
> #5  0x402b1606 in G__cpp_setupG__Base1 () from /home/hejny/root/lib/libCore.so
> #6  0x40697fc8 in G__call_setup_funcs () at cint/src/init.c:142
> 142           (G__setup_func_list[i]->func)();
> #7  0x402b4445 in G__cpp_setup_initG__Base1::G__cpp_setup_initG__Base1 ()
>    from /home/hejny/root/lib/libCore.so
> #8  0x402b196b in __static_initialization_and_destruction_0 ()
>    from /home/hejny/root/lib/libCore.so
> #9  0x402b19ae in global constructors keyed to ROOT::GenerateInitInstance ()
>    from /home/hejny/root/lib/libCore.so
> #10 0x4042cda7 in __do_global_ctors_aux () from /home/hejny/root/lib/libCore.so
> #11 0x400f9d42 in _init () from /home/hejny/root/lib/libCore.so
> #12 0x40009cc7 in call_init () from /lib/ld-linux.so.2
> #13 0x40009e1d in _dl_init () from /lib/ld-linux.so.2
> 
> Well, the reason itself has to be in our code, because leaving out all
> own classes from a program makes this vanish. However, I have no idea
> where to start looking.
> 
> Moreover, when I start a root session and load our shared libraries
> within this session, there is neither a crash nor an error message.
> 
> Is there any important change from 3.05 to 3.10, which I missed?
> Does anyone has a clue where to continue debugging?
> 
> Best regards,
> Volker
> 
> --
> Dr. Volker Hejny                Tel: 02461/616853                      **
> Institut f. Kernphysik          Fax: 02461/613930                     **
> ---------------------------------------------------------------- **  ** ---
> Forschungszentrum Juelich GmbH, D-52425 Juelich                    **



This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:05 MET