[omniORB] Process compiled with omniORB3_0_4 dumps on Solaris

[Venkateswara Rao Kanaparthi]

Hi,

>> I am trying to use omniORB 3_0_4 on Solaris 5.7, but facing a strange 
>> problem when I run my process.
>> It ends up dumping radomly. However the top portion of the stack remain 
>> same.
>
>
>Crashes within  malloc are almost certainly due to you having
>corrupted the heap, by freeing something more than once, or falling
>off the end of a heap-allocated block. Try running your application
>under Purify or similar.
>

Earlier we were statically linking with omniORB libs (*.a's), as we had 
problems in compiling *.so's on Solaris due to some python issues. We 
had PURIFIED our process then and could fix all the leaks.

Now since we have installed python, thought of using *.so's 
(libomniORB3.so.0.4, libomnithread.so* ...). Only after that we started 
getting this crash in malloc()/new.

Any guesses why this could suddenly happen?

Thanks,
KV