From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from euclid.skiles.gatech.edu (list@euclid.skiles.gatech.edu [130.207.146.50]) by melb.werple.net.au (8.7.5/8.7.3/2) with ESMTP id DAA17224 for ; Thu, 25 Jul 1996 03:39:26 +1000 (EST) Received: (from list@localhost) by euclid.skiles.gatech.edu (8.7.3/8.7.3) id NAA12862; Wed, 24 Jul 1996 13:33:39 -0400 (EDT) Resent-Date: Wed, 24 Jul 1996 13:33:39 -0400 (EDT) Sender: joda@pdc.kth.se To: Bruce Stephens Cc: zsh-workers@math.gatech.edu Subject: Re: 3.0-pre3 fails on CRAY-YMP References: <199607241417.QAA00042@cantecler.math.ruu.nl> Mime-Version: 1.0 (generated by tm-edit 7.70) Content-Type: multipart/mixed; boundary="Multipart_Wed_Jul_24_19:32:31_1996-1" Content-Transfer-Encoding: 7bit From: joda@pdc.kth.se (Johan Danielsson) Date: 24 Jul 1996 19:32:31 +0200 In-Reply-To: joda@pdc.kth.se's message of 24 Jul 1996 16:49:19 +0200 Message-ID: X-Mailer: Gnus v5.2.32/Emacs 19.31 Resent-Message-ID: <"LiYt2.0.u83.ovbzn"@euclid> Resent-From: zsh-workers@math.gatech.edu X-Mailing-List: archive/latest/1759 X-Loop: zsh-workers@math.gatech.edu Precedence: list Resent-Sender: zsh-workers-request@math.gatech.edu --Multipart_Wed_Jul_24_19:32:31_1996-1 Content-Type: text/plain; charset=US-ASCII > > The second is that there are a couple of calls to setlimits in > > exec.c which aren't protected with #ifdef's for getrlimits, which > > the Cray appears to lack. Ok, this is a new bug in pre2 or pre3. This is the quick patch: --Multipart_Wed_Jul_24_19:32:31_1996-1 Content-Type: text/plain; charset=US-ASCII RCS file: RCS/exec.c,v retrieving revision 1.1 diff -u -r1.1 exec.c --- exec.c 1996/07/24 17:16:34 1.1 +++ exec.c 1996/07/24 17:16:46 @@ -1686,8 +1686,10 @@ } if (type == SIMPLE) { closem(1); +#ifdef HAVE_GETRLIMIT if (!forked) setlimits(NULL); +#endif execute((Cmdnam) hn, cflags & BINF_DASH); } else { /* ( ... ) */ list_pipe = 0; --Multipart_Wed_Jul_24_19:32:31_1996-1 Content-Type: text/plain; charset=US-ASCII However, shouldn't the ifdef stuff be in zsetlimit? This way you don't have to add ifdefs in every place you want to change limits. > One other problem I have is that it fails to change the terminal > settings (with stty) - it looks like they get reset back when stty > exits (and no I don't have that option on). This is the symptoms. I haven't had time to look into this, it is a bit hard to debug these things without a working debugger :-) $ stty speed 9600 baud; -parity hupcl -cread rows = 32; columns = 80; ypixels = 0; xpixels = 0; erase = ^h; brkint -inpck icrnl -ixany onlcr tab3 echo echoe echok iexten $ stty erase '^?' $ stty speed 9600 baud; -parity hupcl -cread rows = 32; columns = 80; ypixels = 0; xpixels = 0; erase = ^h; brkint -inpck icrnl -ixany onlcr tab3 echo echoe echok iexten $ exec /bin/bash $ stty erase $ stty speed 9600 baud; -parity hupcl -cread rows = 32; columns = 80; ypixels = 0; xpixels = 0; erase = ^h; brkint -inpck icrnl -ixany onlcr tab3 echo echoe echok iexten $ stty erase '^?' $ stty # note that the erase character has changed speed 9600 baud; -parity hupcl -cread rows = 32; columns = 80; ypixels = 0; xpixels = 0; brkint -inpck icrnl -ixany onlcr tab3 echo echoe echok iexten $ exec ./zsh /scratch/joda/zsh-3.0-pre3/Src>stty speed 9600 baud; -parity hupcl -cread rows = 32; columns = 80; ypixels = 0; xpixels = 0; brkint -inpck icrnl -ixany onlcr tab3 echo echoe echok iexten $ stty erase '^h' $ stty speed 9600 baud; -parity hupcl -cread rows = 32; columns = 80; ypixels = 0; xpixels = 0; brkint -inpck icrnl -ixany onlcr tab3 echo echoe echok iexten /Johan --Multipart_Wed_Jul_24_19:32:31_1996-1--