Date: Tue, 4 Jun 1996 13:00:46 +0200 (MET DST) From: grog@lemis.de (Greg Lehey) To: toor@dyson.iquest.net (John S. Dyson) Cc: FreeBSD-current@FreeBSD.ORG (FreeBSD current users) Subject: Re: Vm fixes NG Message-ID: <199606041100.NAA13089@allegro.lemis.de> In-Reply-To: <199606021959.OAA00644@dyson.iquest.net> from "John S. Dyson" at Jun 2, 96 02:59:17 pm
next in thread | previous in thread | raw e-mail | index | archive | help
John S. Dyson writes: > >> >> I had quite a few start-up crashes with Emacs as well. >> >> I'm currently compiling a ctm 2068 kernel. I'll post again if there >> are any new developments. >> > If you have pmap.c v1.97 you have the latest. There are a couple of > people (including you) who are helping immensely. Thanks for being > patient. The more info you can send me, the better the chance that > I can help (and you can help me.) If at all possible, of course a stack > traceback is minimum (and very helpful.) But sometimes the (p->pindex, > p->flags, p->queue, and maybe p->object->type) are all very helpful. I've now built a new kernel with ctm 2072 (pmap.c 1.98). I'm still getting dying Emacsen: === root@freebie (/dev/ttyp2) /usr/home/grog 1 -> emacs& [1] 220 === root@freebie (/dev/ttyp2) /usr/home/grog 2 -> Invalid function: [] [1]+ Exit 255 emacs === root@freebie (/dev/ttyp2) /usr/home/grog 2 -> emacs& [1] 221 === root@freebie (/dev/ttyp2) /usr/home/grog 3 -> emacs& [2] 222 Jun 4 12:44:19 freebie /kernel: pid 221 (emacs), uid 0: exited on signal 11 [1] Segmentation fault (core dumped) emacs === root@freebie (/dev/ttyp2) /usr/home/grog 4 -> emacs& [3] 223 [2] Abort trap (core dumped) emacs Jun 4 12:44:27 freebie /kernel: pid 222 (emacs), uid 0: exited on signal 6 PID 223 started normally and has been running ever since (well, 5 minutes, but it's always been my experience that if it runs this long, it'll continue to run). Another point of interest: this machine has 16 MB of memory and 32 MB swap, but lately when I build a kernel (with debug symbols) I get a number of Jun 4 12:27:25 freebie /kernel: swap_pager: out of swap space Jun 4 12:27:38 freebie /kernel: swap_pager: out of swap space I think these messages are bogus. I can build kernels on another machine (running -stable), with only 8 MB memory and 32 MB swap, and I don't have any trouble. Is there any way to monitor swap space usage? Another point that was mentioned: options DIAGNOSTIC. I have this in my kernel too. Is there any reason why this should make a difference? Anyway, just for completeness' sake I'm building a kernel without DIAGNOSTIC, and will report. Greg
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199606041100.NAA13089>