From owner-freebsd-bugs  Tue May  9 23:25:58 1995
Return-Path: bugs-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id XAA10737
          for bugs-outgoing; Tue, 9 May 1995 23:25:58 -0700
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id XAA10730
          for <freebsd-bugs@FreeBSD.org>; Tue, 9 May 1995 23:25:47 -0700
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id QAA01050; Wed, 10 May 1995 16:22:23 +1000
Date: Wed, 10 May 1995 16:22:23 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199505100622.QAA01050@godzilla.zeta.org.au>
To: Bernard.Steiner@Germany.EU.net, terry@cs.weber.edu
Subject: Re: strange symlinks
Cc: freebsd-bugs@FreeBSD.org
Sender: bugs-owner@FreeBSD.org
Precedence: bulk

>> The next one may be a general 4.4 problem...
>> assumption: /tmp/foo does not exist, /tmp/bar is a symlink to /tmp/foo.
>> chdir("/tmp/bar") fails with ENOENT, but at the same time
>> mkdir("/tmp/bar", 0x777) fails with EEXIST.

>I think this is a generic result of the path component item evaluation
>order.  I guess the only thing to say is that "according to the code,
>this is correct behaviour"  8-).

It's because chdir follows symlinks but mkdir doesn't.  chdir operates on
directories while mkdir operates on directory entries.  The directory
/tmp/foo doesn't exist so the error for chdir is ENOENT but the directory
entry /tmp/bar exists so the error for mkdir is EEXIST.

>> Third, whatever happened to the fchdir() syscall that I vaguely remember
>> having had in (at least) 386BSD0.1 ?

>It had *better* still be there!  I've been using it with a couple of

It hasn't changed.  See `man fchdir'.

Bruce