Title: | Digital Ladebug debugger |
Moderator: | TLE::LUCIA |
Created: | Fri Feb 28 1992 |
Last Modified: | Fri Jun 06 1997 |
Last Successful Update: | Fri Jun 06 1997 |
Number of topics: | 969 |
Total number of notes: | 3959 |
Hello a customer (Telekurs Payserv, Switzerland) is experiencing a problem with decladebug 4.0-14A on Digital UNIX 3.2g. it dumps core when debugging some of their applications. I included a logfile of the debugging session at the end of the note. The same happens with ladebug 4.0-30 on Digital UNIX 3.2g and Digital UNIX V4.0b ... the problem appears only when the application is compiled with symbol table (cxx -g), without symbol table, ladebug runs fine, but what's debugging without symbol table ? the application links in stuff from RTR (librtr.so), RogueWave Tools H++ and DB Tools (librwtool.so, librwdb.so). I received part of the source and precompiled binaries for the reproduction of the problem. Thanks for your help ! Miguel ------------------------------------------------------- Script started on Fri Apr 04 16:26:01 1997 furka:21 > furka:21 > furka:21 > furka:21 > uname -a OSF1 furka.zuo.dec.com V3.2 62 alpha furka:22 > furka:22 > furka:22 > furka:22 > setld -i | grep CXX ATMCXXSHLIB350 C++ Shared Libraries for Atom (Software Development) CXXBASE550 installed DEC C++ (cxx) for Digital UNIX CXXBASEA510 DEC C++ (cxx) for Digital UNIX CXXLIB550 installed DEC C++ static class libraries CXXLIBA510 DEC C++ static class libraries CXXMAN550 installed DEC C++ and class library manual pages CXXMANA510 DEC C++ manual pages: cxx, demangler and library routines CXXOLDA510 DEC C++ old packages: oldc89 and ild CXXSHRDA306 DEC C++ Class Library, Run-Time Support CXXSHRDA307 installed DEC C++ Class Library, Run-Time Support CXXV3HDR550 installed DEC C++ header files for Digital UNIX V3.x furka:23 > furka:23 > setld -i | grep LDB LDBBASE407 DECladebug Debugger Version 4.0-7 LDBBASE414 Ladebug Debugger Version 4.0-14A LDBBASE426 Ladebug Debugger Version 4.0-26 LDBBASE430 installed Ladebug Debugger Version 4.0-30 LDBDOC407 DECladebug Debugger Version 4.0-7 Release Notes LDBDOC426 Ladebug Debugger Version 4.0-26 Release Notes LDBDOC430 installed Ladebug Debugger Version 4.0-30 Release Notes LDBGUI414 Ladebug Debugger Window Interface LDBGUI426 Ladebug Debugger Window Interface LDBGUI430 installed Ladebug Debugger Window Interface LDBSRV426 Ladebug Debugger remote server furka:24 > furka:24 > furka:24 > furka:24 > furka:24 > ls Makefile ladebug.log libpnutil.a librwtool.so persontest.cxx pstorcli.h README libPBC.a librefgen.a libtypes.a persontest.v0 pstorrdb.cxx a.out libclntsh.so.1.0 librtr.so mnyrelrt.cxx persontest.v1 pstorrdb.h cxx_repository libperson.a librwdb.so mnyrelrt.h persstor.cat pstorsrv.cxx include libpersstor.a.v1 librwora.so mnyrelvt.cxx pstorcli.cxx pstorsrv.h furka:25 > furka:25 > furka:25 > furka:25 > furka:25 > furka:25 > furka:25 > furka:25 > furka:25 > furka:25 > ladebug persontest.v0 Welcome to the Ladebug Debugger Version 4.0-30 ------------------ object file name: persontest.v0 Reading symbolic information ...Warning: The /proc file system is not mounted, limited debugging support is provided. Please mount /proc on your system. Ladebug Debugger Version 4.0-30 caught signal "Segmentation fault" (11). This is an unexpected condition and may indicate the presence of a defect. If you wish to report this, please include the stack trace that follows. Diagnostic stack trace ... 0x1239db98 0x124d875c 0x124303cc 0x12422688 0x12431e74 0x12430d18 0x124334c8 0x1242f700 0x1242227c 0x12431c28 0x12430d18 0x124334c8 0x1242f700 0x1242227c 0x12431c28 0x12430d18 0x124334c8 0x1242f700 0x124279ac 0x12421fc4 0x1241f2fc 0x1241ed00 0x1241e39c 0x1241d150 0x125329d4 0x12519258 0x12380cf8 0x12380f3c 0x1235b55c 0x122142e8 0x122181b4 0x1235cc40 0x122025e0 0x12204c60 0x121f316c 0x121f02a4 end of diagnostic stack trace. Could not start process for persontest.v0 No image loaded ... Recovering ... (ladebug) quit Unaligned access pid=13856 <ladebug> va=6f6974636944693b pc=124086bc ra=124af124 type=ldl Segmentation fault (core dumped) furka:26 > file core core: core dump, generated from 'persontest.v0' furka:27 > furka:27 > furka:27 > furka:27 > furka:27 > furka:27 > ladebug persontest.v1 Welcome to the Ladebug Debugger Version 4.0-30 ------------------ object file name: persontest.v1 Reading symbolic information ...Warning: The /proc file system is not mounted, limited debugging support is provided. Please mount /proc on your system. done (ladebug) (ladebug) (ladebug) stop in main [#1: stop in int main(void) ] (ladebug) stop in ManyRelationRCT<Person,Person>::toStream Ladebug Debugger Version 4.0-30 caught signal "Segmentation fault" (11). This is an unexpected condition and may indicate the presence of a defect. If you wish to report this, please include the stack trace that follows. Diagnostic stack trace ... 0x1239db98 0x124d875c 0x124303cc 0x12422688 0x12431e74 0x12430d18 0x124334c8 0x1242f700 0x1242227c 0x12431c28 0x12430d18 0x124334c8 0x1242f700 0x124272d8 0x1241f564 0x1241ed00 0x1241e39c 0x1241d150 0x125329d4 0x12519258 0x12380cf8 0x12381160 0x12519974 0x12448f94 0x1244b6a0 0x1244bc54 0x1244b8b4 0x1244bc30 0x1244b8b4 0x1244bc30 0x122ac974 0x122a1d28 0x1229fd90 0x1224ea4c 0x12217a84 0x12203074 0x121f31b4 0x121f02a4 end of diagnostic stack trace. Fatal error: Segmentation fault furka:28 > furka:28 > cat README The following binaries cause a segmentation violation in the decladebug debugger System: OSF1 leopard V3.2 62 alpha Cxx: DEC C++ V5.5-010 on Digital UNIX (Alpha) Note: setenv LD_LIBRARY_PATH . persontest.v0: Built in our development tree, compiled all files in library with -g option same libpersstor.a.v1 used as in persontest.v1 Debugger crashes when reading symbol table persontest.v1: Built locally (with shorter -I paths) , compiled all files in library with -g option Debugger crashes when setting a breakpoint in ManyRelationRCT<Person,Person>::toStream furka:28 > furka:28 > furka:28 > exit furka:29 > script done on Fri Apr 04 16:30:33 1997 [Posted by WWW Notes gateway]
T.R | Title | User | Personal Name | Date | Lines |
---|---|---|---|---|---|
910.1 | TLE::SHAMIM | Fri Apr 04 1997 19:56 | 11 | ||
Miguel, From the ladebug log, it looks like the /proc file system is not mounted on your system. This may not be the cause of your ladebug's core dump, but you can get much better debugging support if /proc is mounted. See man page of proc to get more information. thanks shamim | |||||
910.2 | /proc mounted, same effect | NETRIX::"[email protected]" | Miguel Mena | Mon Apr 07 1997 04:38 | 11 |
thanks shamim for your reply the logfile was written during a session on digital unix 3.2g. the same happens on digital unix 4.0 with /proc mounted. i can put the customers binaries, source and libs somewhere you can copy it from. let me know ... thanks ! Miguel [Posted by WWW Notes gateway] | |||||
910.3 | TLE::SHAMIM | Mon Apr 07 1997 13:53 | 4 | ||
Yes, please send me a pointer to customer binaries, etc. thanks shamim | |||||
910.4 | here it is ... | NETRIX::"[email protected]" | Miguel Mena | Tue Apr 08 1997 04:07 | 12 |
shamim, find the compresses (gzip) tar file of the source/binary I received at this location: ftp://migix.zuo.dec.com/pub/telekurs/telekurs.tar.gz (~10 MBytes) Thanks ! Miguel [Posted by WWW Notes gateway] | |||||
910.5 | ladebug V4.0-35 works ! | FURKA::"[email protected]" | Miguel Mena | Wed Apr 09 1997 04:47 | 14 |
Shamim, I tested it again with the latest ladebug kit (4-35) on 3.2g and 4.0 and it works ! But :( the cover letter says, that I'm not allowed to give it to customers ... :( Regards, Miguel [Posted by WWW Notes gateway] | |||||
910.6 | TLE::BRETT | Wed Apr 09 1997 08:48 | 9 | ||
If this customer already has ladebug, and this is a serious problem for them, then go ahead and give it to them with the understanding that this is an unsupported internal kit. You should also tell them that there is another internal kit coming out at the end of April that will have more fixes in it, and that they may want to wait for that one. /Bevin | |||||
910.7 | ok, thanks ! | NETRIX::"[email protected]" | Miguel Mena | Wed Apr 09 1997 14:22 | 10 |
I'll do that. thanks Shamim, thanks Bevin ! Best regards, Miguel [Posted by WWW Notes gateway] |