This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
RE: notice_signals
- To: "'Michael Snyder'" <msnyder at redhat dot com>, "Brethour, Tanya (tonic)" <tonic at sequent dot com>
- Subject: RE: notice_signals
- From: "Brethour, Tanya (tonic)" <tonic at sequent dot com>
- Date: Mon, 31 Jul 2000 12:56:59 -0700
- Cc: "'gdb-patches at sourceware dot cygnus dot com'" <gdb-patches at sourceware dot cygnus dot com>
Michael,
Well there is a function called child_wait (ie hpux_thread_wait) which
essentially does the exact same thing. It handles all events the occur (ie.
system exit, signalled, creation of a thread). Do you know why this function
would be necessary.. and if it is.. why is it only used by solaris? All other
targets (at least from what I see doing a grep notice_signal) all set it to
zero.
So.. if anyone knows what makes this function important.. and needed aside
from child_wait... then let me know.
Thanks!
Tanya
> -----Original Message-----
> From: Michael Snyder [SMTP:msnyder@redhat.com]
> Sent: Monday, July 31, 2000 11:08 AM
> To: Brethour, Tanya (tonic)
> Cc: 'gdb-patches@sourceware.cygnus.com'
> Subject: Re: notice_signals
>
> Brethour, Tanya (tonic) wrote:
> >
> > Hello..
> >
> > I was curious if anyone knew what the notice_signals function (ie.
> > hpux_thread_notice_signals, child_ops_to.notice_signals,
> > procfs_notice_signals) is trying to accomplish. Is this needed for ALL
> > platforms.. or what is the criteria to tell if it is needed?
>
>
> Hi,
>
> "notice_signals" basically tells GDB what it should do
> when the child process gets a signal. Ie stop and
> prompt the user, ignore the signal and continue,
> let the child receive the signal, etc. I would think
> it is necessary for any target that uses signals.
>
> Michael
>