A question: if you issue a wait on the process, then how do you read
the pipe output?
Bernard
On 8/30/06, kim aldis <kim(at)aldis.org.uk> wrote:
They're zombie processes and they're what happens when your calling process
doesn't issue a wait.
http://aplawrence.com/SCOFAQ/FAQ_scotec6cantkill.html
> -----Original Message-----
> From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On
> Behalf Of Bernard Lebel
> Sent: 30 August 2006 14:50
> To: XSI(at)Softimage.COM
> Subject: Re: Killing XSIBATCH with a signal
>
> Perhaps I was not clear enough in what I'm looking for. "Signals"
> might not be the perfect term for what I'm after.
>
> I was looking more of way like if you hit the Cancel button when you do
> a render in XSI. Instead of targetting the XSIBATCH process with
> termination signals, I was wondering if there is a way to pass XSIBATCH
> an "abort" instruction, either through a pipe or else.
>
> The purpose of that is that when using the SIGTERM family of signals
> for processes, there are cases where XSIBATCH doesn't exit cleanly and
> may stay indefinitely as a <defunct> process. Using killall -9 can even
> cause semaphore problems. Since XSIBATCH is running in a child thread
> of the Python program and there is a pipe between the two programs,
> perhaps it was possible to use that pipe to tell XSIBATCH to abort the
> render?
>
>
> Thanks
> Bernard
>
>
>
> On 8/30/06, Alan Jones <skyphyr(at)gmail.com> wrote:
> > SIGSTOP also can't be caught - it's signal 7.
> >
> > Cheers,
> >
> > Alan.
> >
> > On 8/30/06, kim aldis <kim(at)aldis.org.uk> wrote:
> > > The man pages for signal should give you a bunch of options, but it
> > > would depend on what you mean by clean and how XSI does it's
> > > cleanup. In theory 9 is the only uncatchable signal so you should
> be
> > > able to send it, say, an 11 and XSI would trap the signal and exit
> > > cleanly. This was how we would kill si3d back on IRIX but last time
> I tried it in XSI it didn't work so well.
> > > That's not to say don't give it a go.
> > >
> > > There are many advantages to writing your own frame loop for
> > > rendering and one of those might be that you could add something
> > > that checks on a socket at the end of each frame for a kill command
> of your own. Just a thought.
> > >
> > > > -----Original Message-----
> > > > From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On
> > > > Behalf Of Bernard Lebel
> > > > Sent: 29 August 2006 19:55
> > > > To: XSI(at)Softimage.COM
> > > > Subject: Killing XSIBATCH with a signal
> > > >
> > > > Hello,
> > > >
> > > > Anyone knows if there is a way to kill the XSIBATCH process with
> a
> > > > signal to XSIBATCH?
> > > >
> > > > Right now we kill it by terminating the process explicitely,
> using
> > > > killall -9 XSIBATCH and the likes. However I'd like to know if
> you
> > > > can send a signal to XSIBATCH so it can exit cleanly (with a
> return value).
> > > >
> > > >
> > > > Thanks
> > > > Bernard
> > > > ---
> > > > Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text
> > > > in
> > > > body:
> > > > unsubscribe xsi
> > >
> > >
> > > ---
> > > Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text
> in body:
> > > unsubscribe xsi
> > >
> > ---
> > Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in
> body:
> > unsubscribe xsi
> >
> ---
> Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in
> body:
> unsubscribe xsi
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
