From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from gatech.edu (gatech.edu [130.207.244.244]) by werple.mira.net.au (8.6.12/8.6.9) with SMTP id FAA20328 for ; Sat, 8 Jul 1995 05:09:14 +1000 Received: from math (math.skiles.gatech.edu) by gatech.edu with SMTP id AA28237 (5.65c/Gatech-10.0-IDA for ); Fri, 7 Jul 1995 15:09:59 -0400 Received: by math (5.x/SMI-SVR4) id AA00313; Fri, 7 Jul 1995 15:06:44 -0400 Resent-Date: Fri, 7 Jul 1995 21:05:01 +0100 (MET DST) Old-Return-Path: From: hzoli@cs.elte.hu (Zoltan Hidvegi) Message-Id: <9507071905.AA21584@turan.elte.hu> Subject: Re: Serious HP-UX problem ! Shell hangs ! To: chen@adi.com Date: Fri, 7 Jul 1995 21:05:01 +0100 (MET DST) In-Reply-To: <9507071358.AA04372@grunt> from "chen@adi.com" at Jul 7, 95 09:58:41 am X-Mailer: ELM [version 2.4 PL21] Content-Type: text Sender: hzoli@cs.elte.hu Resent-Message-Id: <"Ry0hv1.0.p4.3NO_l"@math> Resent-From: zsh-workers@math.gatech.edu X-Mailing-List: archive/latest/180 X-Loop: zsh-workers@math.gatech.edu Precedence: list Resent-Sender: zsh-workers-request@math.gatech.edu Franklin Chen wrote: > Richard Coleman writes: > > > > This is a *serious* HP-UX 9 problem of the zsh, which it didn't have > > > > with previous versions (not with beta6 at least). > > > > > > > > Situation : > > > > > > > > $ sleep 10 (any process will do) > > > > [Ctrl-Z] > > > > zsh: 16763 suspended sleep 10 > > > > $ sleep 5 (any process) > > > > *hang* > > > > > > > > Fiddling with gdb shows that we are hanging in sigsuspend(), which is > > > > called exactly once: signals.c with POSIX_SIGNALS defined. > > > > > > This is a quite old problem (it is present at least since beta6). It is OS > > > dependent, and it is also present on ULTRIX. As I remember it appeared after > > > Richard rewrote the signals code. He promised to fix that. > > > > I don't have any HP or DEC machines, so it is difficult for me to > > track down. Does this still happen if you change POSIX_SIGNALS to > > BSD_SIGNALS? It might be bugs in the Posix code for these machines. > > If this fixes it, I can just force configure to use BSD signals on > > these machines. > > I have been unable to reproduce this problem. > > 6 ~ chen@gar$ uname -a > HP-UX gar A.09.03 A 9000/720 2000519265 two-user license > 7 ~ chen@gar$ echo $ZSH_VERSION > 2.6-beta10 > 8 ~ chen@gar$ sleep 10 > > zsh: 11651 suspended sleep 10 > 9 ~ chen@gar$ sleep 5 > 10 ~ chen@gar$ jobs > [1] + suspended sleep 10 > 11 ~ chen@gar$ fg > [1] + continued sleep 10 > 12 ~ chen@gar$ jobs > 13 ~ chen@gar$ > > > > I build zsh (and all other autoconf-based software) with > CC="c89 -D_HPUX_SOURCE" configure This bug seems to be compiler dependent. Here is a quotion from my BUGS file: ------------------------------------------------------------------------ On ULTRIX 4.2, if NOTIFY is unset, zsh gets SEGV each time a background job terminates if compiled with gcc-2.6.3 or c89. It does not happen with cc (non-ANSI compiler, ansi2knr is used). Terminal suspension does not seem work regardless of the compiler. HP-UX seems to have similar problems. ------------------------------------------------------------------------ And with gcc 2.6.3 on ultrix 4.2 ^Z does not work. The job won'n continue after ^Z but the shell also hangs. Bye, Zoltan