Re: remote debugging

[Cal Erickson]

Melissa,
The SIGPWR signal is what gdb interprets when a thread exits.
Threads do not send a SIGCHILD upon exit. They send a signal
32 which is caught by the thread manager who then knows
what to do.

At MontaVista we have fixed this problem so that gdb
understands more about threads and can interpret
the signals properly. We have also fixed a bug where
a thread exiting under gdb becomes a zombie process.
This fix is to the Linux kernel.

Hopefully these changes will get pushed back to the
community at large for their use.

We have also found that gdbserver does not attach
to threads or processes. We have a project in place
to address this in future releases.

Cal Erickson

Melissa wrote:

> I am using remote debugging, I run the gdbserver on my target machine
> and then run ddd on my machine.  This works fine, however, ddd
> interprets the creation of a pthread in linux using pthread_create as a
> SIGPWR.  This confuses me a great deal and I am open to any suggestions
> that may help be use ddd in this situation properly.
>
> Thanks for your time
> Melissa Donaldson
>
> _______________________________________________
> Ddd mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/ddd

--
===========================================================================
Cal Erickson                 MontaVista Software Inc.
Linux Consultant             1237 E. Arques Ave.
Phone (408) 328-0304         Sunnyvale CA 94085
Fax   (408) 328-9204         web http://www.mvista.com
===========================================================================