Date: Sun, 15 Jul 2007 23:00:52 -0400 From: christos@zoulas.com (Christos Zoulas) To: Giorgos Keramidas <keramida@freebsd.org> Cc: tcsh-bugs@mx.gw.com, current@freebsd.org Subject: Re: tcsh backtick hang info Message-ID: <20070716030052.1B43D56407@rebar.astron.com> In-Reply-To: <20070714005559.GB6661@kobe.laptop> from Giorgos Keramidas (Jul 14, 3:55am)
next in thread | previous in thread | raw e-mail | index | archive | help
On Jul 14, 3:55am, keramida@freebsd.org (Giorgos Keramidas) wrote:
-- Subject: Re: tcsh backtick hang info
| On 2007-07-12 14:56, Christos Zoulas <christos@zoulas.com> wrote:
| > On Jul 12, 11:48am, dwhite@gumbysoft.com (Doug White) wrote:
| > -- Subject: Re: tcsh backtick hang info
| >
| > | Thanks for the good words, Christos! If you can confirm this will be the
| > | official patch then we can apply it to our vendor sources. If there are
| > | any other fixes for nasty bugs in 6.15.00 that you're hanging on to it'd
| > | be good to get those out and imported as well.
| >
| > You are welcome!
| > This will be the official patch, and yes I have another 9 bug fixes since
| > 6.15.00. In general, I don't make 6.15.XX generally available, but the
| > severity of some of those problems fixed is high enough to possibly justify
| > a 6.16.00 sooner than later. Here's the list of fixes:
| >
| > 10. kill `foo` got stuck because sigchld was disabled too soon (Mark Peek)
| > 9. Avoid null pointer dereference in proc cwd (Kurt Miller)
| > 8. eval "foreach a b c" exits (Anthony Menasse)
| > 7. Quoting was broken in substitutions (Joe Wells)
| > 6. QNX patches via pkgsrc
| > 5. cd - twice from a directory that contained a glob pattern,
| > expands the glob twice (Mark Santcroos)
| > 4. MidnightBsd support (Lucas Holt)
| > 3. Fix history substitution core-dump with no history entries
| > 2. Merge two character tables that are the same (Martin Kraemer)
| > 1. On ancient 7 bit locales, punctuation characters are used to
| > denote special characters such as umlaut, adiaresis, etc.
| > These characters return true for isalpha/isalnum. Ignore them
| > because they break parsing (Martin Kraemer)
|
| Hi Christos,
|
| Can we persuade you to include the following local patch I keep for
| autologout detection with /dev/pts/XXX ptys in FreeBSD?
|
| %%%
| Fix pty detection logic of tcsh autologout initialization.
|
| Noticed by: kris
|
| diff --git a/contrib/tcsh/sh.c b/contrib/tcsh/sh.c
| --- a/contrib/tcsh/sh.c
| +++ b/contrib/tcsh/sh.c
| @@ -457,7 +457,7 @@ main(int argc, char **argv)
| if (*cp) {
| /* only for login shells or root and we must have a tty */
| if ((cp2 = Strrchr(cp, (Char) '/')) != NULL) {
| - cp = cp2 + 1;
| + cp2 = cp2 + 1;
| }
| else
| cp2 = cp;
| %%%
The code is wrong but I think that the fix is not exactly right... The
following should work:
/* only for login shells or root and we must have a tty */
if ((cp2 = Strrchr(cp, (Char) '/')) != NULL) {
cp2 = cp2 + 1;
}
else
cp2 = cp;
if (!(((Strncmp(cp2, STRtty, 3) == 0) && Isalpha(cp2[3])) ||
Strstr(cp, Strslptssl) != NULL)) {
if (getenv("DISPLAY") == NULL) {
/* NOT on X window shells */
setcopy(STRautologout, STRdefautologout, VAR_READWRITE);
}
}
Thanks for the bug report. wonder how come nobody complained for years!
christos
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070716030052.1B43D56407>