From owner-freebsd-hackers  Sun May 11 01:31:41 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id BAA18257
          for hackers-outgoing; Sun, 11 May 1997 01:31:41 -0700 (PDT)
Received: from labs.usn.blaze.net.au (labs.usn.blaze.net.au [203.17.53.30])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA18252
          for <freebsd-hackers@freebsd.org>; Sun, 11 May 1997 01:31:36 -0700 (PDT)
Received: from labs.usn.blaze.net.au (local [127.0.0.1])
	by labs.usn.blaze.net.au (8.8.5/8.8.5) with ESMTP id SAA04202;
	Sun, 11 May 1997 18:29:45 +1000 (EST)
Message-Id: <199705110829.SAA04202@labs.usn.blaze.net.au>
X-Mailer: exmh version 2.0gamma 1/27/96
To: "Marty Leisner" <leisner@sdsp.mc.xerox.com>
cc: freebsd-hackers@freebsd.org
Subject: Re: socketpair() 
In-reply-to: Your message of "Sat, 10 May 1997 13:44:57 PDT."
             <9705102044.AA01420@gnu.sdsp.mc.xerox.com> 
X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P<Wc^}f75I|456ez/M
 E3OlD+ZL@c>#N+Ec~Z&@;'LL!;3?y<m[SK`'\R~Oe|(t-x+U@a%p!1_fZP0!Tb^?'%M1|"n9
 Fl{Ao~%X[Ec|bMlaN-3#x.e')nZ2/HHYSa2D[:bk#pj0(}\HpKeE}/GqUb3-1}lnUm6Et4?(
 gB5[*Z3Mu6o{Og.EJ))beY9oq9wcVU7_n}:I2JvCaa[5=&PI;x__["DL`uI}=s>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Sun, 11 May 1997 18:29:43 +1000
From: David Nugent <davidn@labs.usn.blaze.net.au>
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>  >I remember some discussion a while ago about pipe() in FreeBSD
>  >having been implemented (until recently?) using socketpair(),
>  >so perhaps there's no difference. What about with regards to
>  >portability and so forth?
>
>  I think socketpair gives you bidirectional pipes...

pipe() should too, I think. I've never come across an implementation
that doesn't. Or, at least it doesn't seem to matter which returned
handle you use for reading and which for writing.

[popen() of course doesn't, but that's a different story].

I had seen socketpair() used in BSDI's authentication code, which is
why I asked. Previously we were using popen(), but perhaps your
comments are relevent. Their new authentication api allows writing
to the pipe as well before starting to listen on it for the
authentication data, so it /has/ to be bidirectional. I just didn't
know that pipe had that limitation, assuming it does.

Regards,

David Nugent - Unique Computing Pty Ltd - Melbourne, Australia
Voice +61-3-9791-9547  Data/BBS +61-3-9792-3507  3:632/348@fidonet
davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/