bug#31901: Incorrect make-network-process + nowait state handling for non-existing unix sockets in emacs-26.1

[Eli Zaretskii]

Ping!  Ping!

> Date: Fri, 06 Jul 2018 11:59:35 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: mk.fraggod@gmail.com, 31901@debbugs.gnu.org
> 
> Ping!
> 
> Lars, could you please look into this?
> 
> > From: Noam Postavsky <npostavs@gmail.com>
> > Date: Sun, 01 Jul 2018 11:21:06 -0400
> > Cc: Lars Ingebrigtsen <larsi@gnus.org>, 31901@debbugs.gnu.org
> > 
> > tags 31901 + confirmed
> > quit
> > 
> > Mike Kazantsev <mk.fraggod@gmail.com> writes:
> > 
> > > As far as I can tell, this is not documented behavior, definitely
> > > unexpected, breaks all old code that used unix sockets, and probably
> > > unintentional, i.e. a bug.
> > 
> > It seems to be due to this is_non_blocking_client check in
> > connect_network_socket:
> > 
> >   if (s < 0)
> >     {
> >       /* If non-blocking got this far - and failed - assume non-blocking is
> >          not supported after all.  This is probably a wrong assumption, but
> >          the normal blocking calls to open-network-stream handles this error
> >          better.  */
> >       if (p->is_non_blocking_client)
> >         return;
> > 
> >       report_file_errno ((p->is_server
> >                           ? "make server process failed"
> >                           : "make client process failed"),
> >                          contact, xerrno);
> >     }
> > 
> > In Emacs 25, this check was directly in Fmake_network_process before the
> > process object creation code, so we would just return nil in that case.
> > Seems to have changed, I assume accidentally, in [1: e09c0972c3].
> > 
> > [1: e09c0972c3]: 2016-01-28 23:50:47 +0100
> >   Refactor make_network_process
> >   
> > https://git.savannah.gnu.org/cgit/emacs.git/commit/?id=e09c0972c350e9411683b509414fc598cbf387d3
> > 
> > 
> > 
> > 
> 
> 
> 
>