SIGUSR1 using omniORB2.4 with egcs-1.0.1 and Redhat 5.0

[Jörg Narr]

Dear experts,

I am actually facing a problem I can't solve on my own. I am using the
above configuration
to write my multithreaded applications on Linux.

After error-free compilation of the applications I receive a
SIGUSR1-signal from the
ORB-implementation which looks like this:

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

0x40113412      mutex.c:112: No such file or directory.
(gdb) si
0x40113418      mutex.c:112: No such file or directory.
(gdb) s
0x80a7f45 in omni_mutex::lock () at select.C:255
255     }
Current language:  auto; currently c++
(gdb) si
0x80a7f48       255     }
(gdb) si
0x80a7f49       255     }
(gdb) si
0x804f9f4 in pthread_mutex_lock ()
(gdb) si
0x40113400 in __pthread_mutex_lock () at mutex.c:112
mutex.c:112: No such file or directory.
Current language:  auto; currently c
(gdb) si
0x40113401      mutex.c:112: No such file or directory.
(gdb) s
0x80a7f45 in omni_mutex::lock () at select.C:255
255     }
Current language:  auto; currently c++
(gdb) s

Program received signal SIGUSR1, User defined signal 1.
0x401e3bb4 in __syscall_sigsuspend ()

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

I am quite new to egcs and omniORB so please forgive this newby
question: One of the main-
tainers of the omniORB-list outlined that egcs-1.00 only works with
thread-safe patches
as discribed in: http://www.orl.co.uk/omniORB/omniORB_240/README.egcs
Since parts
of these patches are written for egcs-1.0 I am actually not sure if I
would have to apply them
to egcs-1.0.1, too. Could anybody of you tell me what exactly I need to
do to get egcs-1.0.1
working with multithreaded applications wich make intensive use of
exceptions? Maybe one
of you has egcs-1.0.1 properly working with omniORB. The examples work
most of the time,
only complex applications have random crashes caused by the problem with
the threads.

Thank you very much in advance.

Cheers,

Joerg