[Catalyst] SIGINT triggers segfault in Oracle libraries

[Tomas Doran]

On 7 Sep 2010, at 21:27, E R wrote:

> If I hit Control-c after starting up the standalone catalyst server
> without the -r option, I'll get a segfault.
>
> Here is the backtrace:
>
> #0  0xb6240a07 in nsgetcinfo () from .../rnapps/oracle/libclntsh.so. 
> 10.1
> #1  0xb62658c8 in nioqih () from .../rnapps/oracle/libclntsh.so.10.1
> #2  0xb626bb49 in nigsuiint () from .../rnapps/oracle/libclntsh.so. 
> 10.1
> #3  0xb626bbd4 in nigsuihdlr () from .../rnapps/oracle/libclntsh.so. 
> 10.1
> #4  0xb6939181 in sslsshdlr2 () from .../rnapps/oracle/libclntsh.so. 
> 10.1
> #5  0xb6939038 in sslsshandler () from .../rnapps/oracle/ 
> libclntsh.so.10.1
> #6  <signal handler called>
> #7  0xb775e42e in __kernel_vsyscall ()
> #8  0xb7646a91 in accept () from /lib/tls/i686/cmov/libc.so.6
> #9  0x080c4614 in Perl_pp_accept ()
> #10 0x0808db13 in Perl_runops_standard ()
> #11 0x0808bdcb in perl_run ()
> #12 0x08062a2d in main ()
>
> If I run with -r I do not get a segfault. Any ideas?

This is an Oracle library thing, can't really help, sorry..

With the (newer) restarter the signal is caught by the parent  
(watcher) process, so you don't see this...

Cheers
t0m