X-Spam-Check-By: sourceware.org
Message-ID: <n2m-g.dr8ma5.3vvcg5j.1@buzzy-box.bavag>
From: "Buzz" <cygwin DOT buzz AT bavag DOT tmfweb DOT nl>
Subject: Re: Shell (bash, (pd)ksh, zsh, /not/ ash) + exec + here-doc + redirect  == trouble!
References: <n2m-g DOT dr7491 DOT 3vv7b8v DOT 1 AT buzzy-box DOT bavag> <43D78E29 DOT 4060504 AT byu DOT net>
Reply-To: cygwin mailing-list <cygwin AT cygwin DOT com>
Organisation: Ehm...
User-Agent: slrn/0.9.8.1 (Win32) Hamster/2.1.0.0 KorrNews/4.2.1
To: cygwin AT cygwin DOT com
In-Reply-To: <43D78E29.4060504@byu.net>
Lines: 85
Date: Wed, 25 Jan 2006 20:24:25 +0100 (MET)
X-IsSubscribed: yes
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Precedence: bulk
List-Unsubscribe: <mailto:cygwin-unsubscribe-archive-cygwin=delorie DOT com AT cygwin DOT com>
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com

Op Wed, 25 Jan 2006 07:41:45 -0700 schreef Eric Blake
in <43D78E29.4060504<at>byu.net>:
:
:  According to Bas van Gompel on 1/24/2006 10:11 PM:
[script]

:  First line: This could be rewritten "exec /bin/ksh 5<&0 <<EOSH".  Either
:  way, you are replacing the current shell with an invocation of /bin/ksh,
:  and with fd 5 set to your current stdin, and then with fd 0 set to a pipe
:  supplied by the contents of the here-doc.

So far so good...

:  Now ksh gets through the following input on stdin:
:  echo "First exec: Done."
:  exec 0<&5

:  At which point, the second exec says take fd 5 and dup it into fd 0 (in
:  other words, restore your original stdin from before the first exec back
:  to ksh's stdin).  POSIX requires that ksh reads its input from stdin a
:  line at a time, so the very next line it reads will be whatever you type
:  (if the original stdin happened to be a terminal).  Only ash is buggy here
:  (no big surprise, ash is usually the least POSIX-compliant of the shells);
:  ksh, zsh, and bash are all doing what is required of them.

I (now) understand what's happening. I think it's undesirable, though.

I think the shells should redirect the input for the executed programs,
/not/ for itself. (Could this be an exploitable vulnerability?)

[...]
: > In ash it prints ''

: >   First exec: Done.
: >   Second exec: Done.

: > '', as I expected. Compare p.e.

:  Your expectation was wrong.

Apparently.

: > === begin testexec2.sh ===
: > #!/bin/bash
: >
: > echo 'echo "First exec: Done."
: > exec 0<&5
: > echo "Second exec: Done."
: > exit 0' |exec 5<&0 /bin/bash
: >
: > ==== end testexec2.sh ====
:
:  Here, you are doing something slightly different.  The last line is a
:  pipeline, which must be applied before redirections.  Which means that you
:  are exec'ing /bin/bash, with fd 0 set to the pipeline from the echo, then
:  fd 5 copied from fd 0.  When the second "exec 0<&5" is reached, you copy
:  fd 5 back to fd 0, but since it was the same fd, it was effectively a
:  no-op.  Therefore, /bin/bash continues to read the next line from fd 0,
:  and successfully prints "Second exec: Done."

Right, my bad. Following shows the same (odd, IMO) behavior.

=== begin testexec2b.sh ===
#!/bin/bash

exec 5<&0
echo 'echo "First exec: Done."
exec 0<&5
echo "Second exec: Done."
exit 0' |exec /bin/bash

==== end testexec2b.sh ====


I can work around this effect using -c, but thought it best to report
anyway...


L8r,

Buzz.
-- 
  ) |  | ---/ ---/  Yes, this | This message consists of true | I do not
--  |  |   /    /   really is |   and false bits entirely.    | mail for
  ) |  |  /    /    a 72 by 4 +-------------------------------+ any1 but
--  \--| /--- /---  .sigfile. |   |perl -pe "s.u(z)\1.as."    | me. 4^re

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/