Re: SIGIO and the NS port

[Alan Third <alan@idiocy.org>]

On Sun, May 21, 2017 at 12:46:53AM -0700, Paul Eggert wrote:
> Alan Third wrote:
> > I’ve stopped the signal handler from responding to SIGIO,
> 
> I'm not following, as sigaction (SIGIO, ...) is still being called, so the
> signal handler deliver_input_available_signal is still responding to SIGIO,
> and is setting pending_signals to true.
> 
> The changes you made affect SIGUSR1 and SIGUSR2 handling, since they patch
> handle_user_signal: they cause the SIGUSR1 and SIGUSR2 handler to not
> process SIGIO events. Also, they prevent SIGIO from being masked out during
> non-SIGIO signal handlers. Is that what you intended? I'm not getting the
> connection between these non-SIGIO handlers and interrupt_input.

I’ve probably misunderstood what the code was doing. Would it be best
to just leave the signal handlers as‐is, but replace the calls to
raise(SIGIO) with direct calls as below?

> > and the two
> > places where the NS GUI used to raise SIGIO have been replaced with
> > direct calls to the SIGIO handler code.
> 
> That sounds like a reasonable thing to do in any event. I don't see why the
> code needs to do a raise (SIGIO) when it can do a
> handle_input_available_signal (SIGIO);

Thanks!
-- 
Alan Third