Concurrency problems

Ignasi Villagrasa gri@netcom.es
Thu Sep 13 03:25:00 GMT 2001


I have a problem in a Client / Server structure using Cygwin. This 
problem is solved using Watcom 10.5, the original used compiler for this 
projects.

I have been investigating several days, but I don't know exactly where 
the access violation (That's the error) is:


MCASE-STAT-N            Estado
Nueva conexión del usuario 'NTNUEVO         ' por una sesion.
Nueva conexión del usuario 'GRIX            ' por una sesion.

Program received signal SIGSEGV, Segmentation fault.
[Switching to thread 1348.0x4dc]
0x42ed37 in McPetHndlr::MessageHandler (this=0x28dea0, DatSz=64, SockIn=0x0)
   at \server\src\mc_msg.cpp:356
356         return (RespSz);
(gdb) info threads
  14 thread 1348.0x4f0  0x784627e8 in ?? ()
* 13 thread 1348.0x4dc  0x42ed37 in McPetHndlr::MessageHandler 
(this=0x28dea0,
   DatSz=64, SockIn=0x0) at \server\src\mc_msg.cpp:356
  12 thread 1348.0x11c  0x784627e8 in ?? ()
  11 thread 1348.0x360  0x784627e8 in ?? ()
  10 thread 1348.0x538  0x784627e8 in ?? ()
  9 thread 1348.0x184  0x784627e8 in ?? ()
  8 thread 1348.0x4e8  0x78462829 in ?? ()
  7 thread 1348.0x3fc  0x784627e8 in ?? ()
  6 thread 1348.0x410  0x784627e8 in ?? ()
  4 thread 1348.0x1b4  0x78462f27 in ?? ()
  3 thread 1348.0x340  0x784630cf in ?? ()
  2 thread 1348.0x470  0x78462231 in ?? ()
  1 process 1348  0x0 in ?? ()
(gdb)

------------------------------------------------------------------------

The two threads in conflict are 13 and 14. But I can see thread number 
14 is at 0x784627e8. I supose that's a stdlib function memory address, 
but I don't know which one and how to look for it. Additionally, I don't 
know if I'm going wrong using any compiler or linking option. At first 
glance, I think I configured the options as documented.

I'd like anyone could give me a hint to follow my search.

Remember the same program is running properly using Watcom 10.5. So I  
think perhaps the problem relates to a configuration issue, and not to a 
source code problem.

Because the features of the program, it's not easy to simplify the 
problem,  trying to test a smaller program without changing the 
conditions that provoke the error.

Thanks in advance. Ignasi Villagrasa.





--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/



More information about the Cygwin mailing list