Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker

Joris Mooij
The following reply was made to PR gdb/2031; it has been noted by GNATS.

From: [hidden email]
To: [hidden email]
Cc:  
Subject: Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker
Date: Tue, 9 May 2006 10:50:28 +0200

 Hi,
 
 I got the same error message after quiting from a running program using
 gdb in Debian sarge. It was apparently caused by changing some methods
 to virtual (and adding a virtual destructor) in combination with a
 *partial* rebuild of the code. The segmentation fault (which, when
 analyzed with gdb, leaded to the BFD error) went away after a complete
 rebuild of the code, so the problem may have been caused by partially
 incorrect debug information in the executable. Just a guess.
 
 If anybody is interested, I could try to reproduce it (unfortunately, I
 overwrote the offending executable) and check what happens with a CVS
 HEAD snapshot of gdb.
 
 Regards, Joris Mooij
 
 -----
 
 GNU gdb 6.3-debian
 Copyright 2004 Free Software Foundation, Inc.
 GDB is free software, covered by the GNU General Public License, and you are
 welcome to change it and/or distribute copies of it under certain conditions.
 Type "show copying" to see the conditions.
 There is absolutely no warranty for GDB.  Type "show warranty" for details.
 This GDB was configured as "i386-linux"...Using host libthread_db library "/lib/tls/libthread_db.so.1".
 
 (gdb) run
 Starting program: /home/jorism/svn/network/marginal
 
 Program received signal SIGSEGV, Segmentation fault.
 _Rb_tree (this=0x80c552c, __x=@0x80cc984) at stl_tree.h:648
 648           _M_root() const { return (_Link_type&)
 _M_header->_M_parent; }
 (gdb) bt
 #0  _Rb_tree (this=0x80c552c, __x=@0x80cc984) at stl_tree.h:648
 #1  0x080601ec in __uninitialized_copy_aux<__gnu_cxx::__normal_iterator<const Factor*, std::vector<Factor, std::allocator<Factor> > >, Factor*> (__first=
       {<std::iterator<std::random_access_iterator_tag, TFactor<double>, int, TFactor<double> const*, TFactor<double> const&>> = {<No data fields>}, _M_current = 0x80cc984}, __result=0x80c5510) at stl_set.h:131
 #2  0x080742a0 in BipartiteGraph (this=0xbffff850, x=@0xbffffc60) at stl_uninitialized.h:109
 #3  0x080a058e in JunctionTree (this=0xbffffb40, fg=@0xbffffc60, method=135034832, verbose=135034832) at factorgraph.h:45
 #4  0x0804b333 in main () at marginal.cpp:179
 (gdb) q
 The program is running.  Exit anyway? (y or n) y
 BFD: BFD 2.15.93 20041018 internal error, aborting at
 /nevyn/local/gdb/gdb-6.3/bfd/cache.c line 495 in bfd_cache_lookup_worker
 
 BFD: Please report this bug.
 
Reply | Threaded
Open this post in threaded view
|

Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker

jimb (Bugzilla)
Synopsis: gdb - aborting - bfd_cache_lookup_worker

Responsible-Changed-From-To: unassigned->jimb
Responsible-Changed-By: jimb
Responsible-Changed-When: Tue May  9 17:37:30 2006
Responsible-Changed-Why:
    I'll take it.
State-Changed-From-To: open->feedback
State-Changed-By: jimb
State-Changed-When: Tue May  9 17:37:30 2006
State-Changed-Why:
    I believe this is fixed on HEAD.  Please let us know.

http://sources.redhat.com/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gdb&pr=2031
Reply | Threaded
Open this post in threaded view
|

Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker

Joris Mooij
In reply to this post by Joris Mooij
The following reply was made to PR gdb/2031; it has been noted by GNATS.

From: [hidden email]
To: [hidden email]
Cc:  
Subject: Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker
Date: Wed, 10 May 2006 15:20:06 +0200

 Unfortunately, I could only reproduce the segfault (in a slightly
 different executable this time), but not the BFD internal error...
 sorry.
 
 Regards, Joris
Reply | Threaded
Open this post in threaded view
|

Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker

jimb (Bugzilla)
In reply to this post by Joris Mooij
Synopsis: gdb - aborting - bfd_cache_lookup_worker

State-Changed-From-To: feedback->closed
State-Changed-By: jimb
State-Changed-When: Wed May 10 15:04:55 2006
State-Changed-Why:
    I think that means that Daniel's fix is working.  Thanks!

http://sources.redhat.com/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gdb&pr=2031
Reply | Threaded
Open this post in threaded view
|

Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker

Joris Mooij
In reply to this post by Joris Mooij
The following reply was made to PR gdb/2031; it has been noted by GNATS.

From: [hidden email]
To: [hidden email]
Cc:  
Subject: Re: gdb/2031: gdb - aborting - bfd_cache_lookup_worker
Date: Fri, 12 May 2006 14:18:40 +0200

 No, it means that I cannot reproduce the bug using the *unfixed* (old) version
 of GDB that originally gave the internal error message.
 
 I apologize for not being entirely clear.
 
 Next time I'll remember to keep a backup of the executable, so that I
 can hopefully reproduce the bug.
 
 By the way, it seems there are different PR's in which this same
 internal error message is encountered.
 
 Regards, Joris