From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <zsh-workers-request@math.gatech.edu>
Received: (qmail 2870 invoked from network); 20 Oct 1998 20:43:54 -0000
Received: from math.gatech.edu (list@130.207.146.50)
  by ns1.primenet.com.au with SMTP; 20 Oct 1998 20:43:54 -0000
Received: (from list@localhost)
	by math.gatech.edu (8.9.1/8.9.1) id QAA12031;
	Tue, 20 Oct 1998 16:36:19 -0400 (EDT)
Resent-Date: Tue, 20 Oct 1998 16:36:07 -0400 (EDT)
From: "Bart Schaefer" <schaefer@brasslantern.com>
Message-Id: <981020125449.ZM6543@candle.brasslantern.com>
Date: Tue, 20 Oct 1998 12:54:49 -0700
In-Reply-To: <981020095542.ZM5494@candle.brasslantern.com>
Comments: In reply to "Bart Schaefer" <schaefer@brasslantern.com>
        "Re: Backticks In zshenv" (Oct 20,  9:55am)
References: <362C9FB4.B4C08045@guildford.ericsson.se> 
	<981020095542.ZM5494@candle.brasslantern.com>
X-Mailer: Z-Mail (4.0b.820 20aug96)
To: Daniel Rigal <daniel.rigal@deep-thought.ericsson.se>,
        zsh-users@math.gatech.edu
Subject: PATCH: Re: Backticks In zshenv
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Resent-Message-ID: <"Kb--d.0.Gx2.tGFBs"@math>
Resent-From: zsh-users@math.gatech.edu
X-Mailing-List: <zsh-users@math.gatech.edu> archive/latest/1880
X-Loop: zsh-users@math.gatech.edu
X-Loop: zsh-workers@math.gatech.edu
Precedence: list
Resent-Sender: zsh-workers-request@math.gatech.edu

On Oct 20,  9:55am, Bart Schaefer wrote:
} Subject: Re: Backticks In zshenv
}
} I suspect zsh is getting some kind of a signal -- perhaps SIGWINCH? --
} which is interrupting the read on the command substitution.

The /etc/zshenv on the RH5.1 machine now looks (in part) like this:

##
for i in {23..31}; do trap "echo caught signal $i" $i; done

if [ `id -u` -eq 0 ]; then
  path=(/sbin /usr/sbin)
fi

# ... some other stuff ...

trap - {23..31}
##

(BTW, DON'T try this without the "trap -", or zsh can go into an infinite
loop; I haven't tracked *that* down, but it seems to be SIGWINCH-related
as well.)

When I rlogin to that machine, I see:

caught signal 28
/etc/zshenv: parse error: condition expected: -eq [14]

I'm now 99% certain that the SIGWINCH is causing read() to fail with EINTR
during the fgetc() in readoutput() in exec.c.

I'm unable to adequately test the following patch at the moment, as it's
inconvenient to muck with the shells on the RH5.1 machine, but it doesn't
break anything obvious on my RH4.2 machine at home.  If anyone has time to
try this out, please tell the list whether it works.

Index: Src/exec.c
===================================================================
RCS file: /extra/cvsroot/zsh/zsh-3.0/Src/exec.c,v
retrieving revision 1.4
diff -c -r1.4 exec.c
*** exec.c	1997/06/27 19:57:19	1.4
--- exec.c	1998/10/20 19:46:11
***************
*** 2104,2110 ****
      fin = fdopen(in, "r");
      ret = newlinklist();
      ptr = buf = (char *) ncalloc(bsiz = 64);
!     while ((c = fgetc(fin)) != EOF) {
  	if (imeta(c)) {
  	    *ptr++ = Meta;
  	    c ^= 32;
--- 2104,2115 ----
      fin = fdopen(in, "r");
      ret = newlinklist();
      ptr = buf = (char *) ncalloc(bsiz = 64);
!     while ((c = fgetc(fin)) != EOF || errno == EINTR) {
! 	if (c == EOF) {
! 	    errno = 0;
! 	    clearerr(fin);
! 	    continue;
! 	}
  	if (imeta(c)) {
  	    *ptr++ = Meta;
  	    c ^= 32;


-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com