[omniORB] Segfault from libomniOrb

[Marcin Cieslak]

On Tue, 25 Mar 2025, Krishnamohan, Srivathsan via omniORB-list wrote:

Additionally, you can enable some tracing options (that might
generate lots of output, so be prepared):

https://omniorb.sourceforge.io/omni42/omniORB/omniORB004.html#sec59

> We already have the full stack trace of where libomniORB crashes. Every time it crashes in the same place with same stack trace.

The following you have posted is not a full stack trace:

> #5  0x00007fe124f1a907 in omni::giopStream::inputMessage() () from
> /home/user/libs_gcc92/libomniORB4.so.2
> #6  0x00007fe124f2fe53 in omni::giopImpl12::inputNewServerMessage(omni::giopStream*) ()
> from /home/user/libs_gcc92/libomniORB4.so.2
> #7  0x00007fe124f30d38 in omni::giopImpl12::inputMessageBegin(omni::giopStream*, void
> (*)(omni::giopStream*)) ()

This is a similar stacktrace from the omniORB 4.3, when compiled with -g:

#0  omni::giopStream::inputMessage (this=0x801c00008) at giopStream.cc:837
#1  0x000000080044f79d in omni::giopImpl12::inputNewServerMessage (
     g=0x801c00008) at giopImpl12.cc:443
#2  0x0000000800450105 in omni::giopImpl12::inputMessageBegin (
     g=0x801c00008,
     unmarshalHeader=0x800450390 <omni::giopImpl12::unmarshalWildCardRequestHeader(omni::giopStream*)>) at giopImpl12.cc:653
#3  0x000000080043ca1b in omni::GIOP_S::dispatcher (this=0x801c00000)
     at GIOP_S.cc:135
#4  0x0000000800438ae8 in omni::giopWorker::execute (this=0x801813000)
     at giopWorker.cc:77
#5  0x00000008003c59ae in omniAsyncWorker::real_run (this=0x80180d0a0)
     at invoker.cc:578
#6  0x00000008003c5d08 in omniServerWorkerInfo::run (this=0x7fffdfbfbd10)
     at invoker.cc:682
#7  0x00000008003c7224 in omniAsyncPoolServer::workerRun (this=0x800c270a0,
     worker=0x80180d0a0) at invoker.cc:328
#8  0x00000008003c53d6 in omniAsyncWorker::mid_run (this=0x80180d0a0)
     at invoker.cc:511
#9  0x00000008003c5c98 in omniAsyncWorkerInfo::run (this=0x7fffdfbfbf30)
     at invoker.cc:662
#10 0x00000008003c763b in omniAsyncWorker::run (this=0x80180d0a0)
     at invoker.cc:126


> How will recompiling with -g help?

It won't help with the crash.

You should see additional information like arguments passed and the line numbers.
Also would be good to have a core dump file for analysis (please
check if your system is creating these and where they end up).

> It might prevent the crash from happening.

It should not. If it does, this is also an interesting information.

All the best,

Marcin
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3706 bytes
Desc: S/MIME Cryptographic Signature
URL: <https://www.omniorb-support.com/pipermail/omniorb-list/attachments/20250325/d15268c0/attachment.p7s>