From owner-freebsd-hackers Tue Oct 17 13:39:42 1995 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA06813 for hackers-outgoing; Tue, 17 Oct 1995 13:39:42 -0700 Received: from ki1.chemie.fu-berlin.de (ki1.Chemie.FU-Berlin.DE [160.45.24.21]) by freefall.freebsd.org (8.6.12/8.6.6) with SMTP id NAA06805 for ; Tue, 17 Oct 1995 13:39:34 -0700 Received: by ki1.chemie.fu-berlin.de (Smail3.1.28.1) from mail.hanse.de (134.100.239.2) with smtp id ; Tue, 17 Oct 95 21:38 MET Received: from wavehh.UUCP by mail.hanse.de with UUCP for gillham@andrews.EDU id ; Tue, 17 Oct 95 21:14 MET Received: by wavehh.hanse.de (4.1/SMI-4.1) id AA15624; Tue, 17 Oct 95 18:53:16 +0100 Date: Tue, 17 Oct 95 18:53:16 +0100 From: cracauer@wavehh.hanse.de (Martin Cracauer) Message-Id: <9510171753.AA15624@wavehh.hanse.de> To: gillham@andrews.EDU Cc: FreeBSD-Hackers@freebsd.org Subject: Re: Stability questions... Newsgroups: hanse-ml.freebsd.hackers References: <9510171512.AA24607@edmund> Reply-To: cracauer@wavehh.hanse.de Sender: owner-hackers@freebsd.org Precedence: bulk gillham@andrews.EDU (Andrew Gillham) wrote: >Apologies if this is the wrong mailing list. >I installed FreeBSD 2.0.5-RELEASE over the weekend, and I have >been having problems with it. What I am seeing is "cc1 interrupted >by signal 11" or similar when I am trying to compile a kernel. >Restarting make will eventually get it all built though. Is this a >known problem with 2.0.5-RELEASE, or is my hardware flakey? I can reliable make by box behave that way by either playing with the RAM timing settings or by overclocking the CPU. So I'm sure it is a hardware problem. >I have been running NetBSD on this box without any problems. That is odd, yes. When I kick my Mainboard, both NetBSD and FreeBSD break. Maybe it's a possible explanation that NetBSD uses other memory banks to provide memory for the compiler by default. You have 32 MB and only about 8 are used in compilations. And I know that the AMD chips have some littel differences in the MMU hardware. Maybe the FreeBSD VM system does something else here than NetBSD. Try that: - Remove some of your RAM, try different SIMMs on their own. - Turn of external Cache (standard answer :-) - Maybe you can try an Intel CPU? - Try a different SCSI controller. Yes, sounds odd, but the FreeBSD VM/buffer-cache system can provide files much faster that NetBSD does. NetBSD has one extra level of copying in RAM. Maybe this shows up timing problems with your system. Try for example a 1542, which will make things slower. OK, it really sounds odd. And let us know what solves the problem! >Also, I supped freebsd-current onto my NetBSD NFS server, and tried >building it. I was able to build a kernel and /sbin/* (by restarting >make whenever cc1 got a SEGV), and rebooted with the new kernel. This >new kernel didn't appear to have the SEGV problem, but I'm not 100% >sure, as something else broke. After a few minutes of compiling (on the >NFS mount) the kernel says "Biodone: buffer already done" or something >and hangs my compile. (sorry for the not exact errors, but I'm at >work, and the PCs are at home) Accessing the NFS mountpoint will hang >my shell, but I can still halt the box from another console. >Any ideas? Should I be trying to update to 2.1-current instead of >2.2-current? (how would I do that? I just ask for 'current' via sup.) Absolutely. 2-2-current is lilely to have problems. It is not like NetBSD-current, it is much more "experimental". The 2-1 snapshots are the best solution for production systems that should be up-to-date, but stable. Martin -- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Martin Cracauer - Fax +49 40 522 85 36 BSD User Group Hamburg, Germany - No NeXTMail anymore, please. Copyright 1995. Redistribution via Microsoft Network is prohibited