bug#29454: wait_reading_process_output duplication

[Charles A. Roelli]

> Date: Sun, 26 Nov 2017 17:54:00 +0200
> From: Eli Zaretskii <eliz@gnu.org>
> CC: 29454@debbugs.gnu.org
> Reply-to: Eli Zaretskii <eliz@gnu.org>
> 
> > Date: Sun, 26 Nov 2017 10:45:53 +0100
> > From: charles@aurox.ch (Charles A. Roelli)
> > 
> > Looks like these two blocks are duplicated in
> > wait_reading_process_output in process.c:
> > 
> > 5628 #ifdef WINDOWSNT
> > 5629              /* FIXME: Is this special case still needed?  */
> > 5630              /* Note that we cannot distinguish between no input
> > 5631                 available now and a closed pipe.
> > 5632                 With luck, a closed pipe will be accompanied by
> > 5633                 subprocess termination and SIGCHLD.  */
> > 5634              else if (nread == 0 && !NETCONN_P (proc) && !SERIALCONN_P 
> > (proc)
> > 5635                       && !PIPECONN_P (proc))
> > 5636                ;
> > 5637 #endif
> > 
> > 5664              /* If we can detect process termination, don't consider 
> > the
> > 5665                 process gone just because its pipe is closed.  */
> > 5666              else if (nread == 0 && !NETCONN_P (proc) && !SERIALCONN_P 
> > (proc)
> > 5667                       && !PIPECONN_P (proc))
> > 5668                ;
> > 
> > Can the first be deleted?
> 
> Yes.  (The second was originally conditioned on SIGCHLD, which is why
> it was separate.)
> 
> Thanks.

Thanks, it's pushed to master.

  commit 6ec5d497b6623e612ca6936ac848234725d4fc61
  Date:   Sun Nov 26 20:18:54 2017 +0100

  * src/process.c (wait_reading_process_output): Deduplicate check.

  (Bug#29454)