From owner-freebsd-questions Mon Jul 15 00:36:01 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id AAA10926 for questions-outgoing; Mon, 15 Jul 1996 00:36:01 -0700 (PDT) Received: from gdi.uoregon.edu ([128.223.186.250]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id AAA10898 for ; Mon, 15 Jul 1996 00:35:59 -0700 (PDT) Received: (from dwhite@localhost) by gdi.uoregon.edu (8.7.5/8.6.12) id AAA00259; Mon, 15 Jul 1996 00:35:59 -0700 (PDT) Date: Mon, 15 Jul 1996 00:35:59 -0700 (PDT) From: Doug White Reply-To: dwhite@resnet.uoregon.edu To: questions@freebsd.org Subject: First look: 2.1.5 install Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-questions@freebsd.org X-Loop: FreeBSD.org Precedence: bulk (Pointers for a more appropriate forum for this are greatly appreciated) Just finished going through the 2.1->2.1.5 upgrade. Software copyover (via ftp from a local machine) was error-free. I like the new bottom-level status display! Nice to know how fast the system is processing info. Phase 2 didn't go as well. It hosed sysconfig and services _BEFORE_ copying them to /usr/tmp/etc!! Luckily I didn't trust myself and made a separate backup of /etc so I was able to resurrect the proper /etc/services from the hacked-down version and found the proper settings in sysconfig. Here are the date/timestamps off of the "backups" (note they haven't been touched since mid-June): -rw-rw-r-- 1 root wheel 137 Jul 13 21:02 services -rw-r--r-- 1 root wheel 8409 Jul 13 19:49 sysconfig Hm. /etc/services should be more like: -rw-r--r-- 1 root wheel 59967 Jul 14 22:39 services So Jordan, please fix that!! I think I got bit by this last time because I remember hunting through the archives looking for a services file. Looking at timestamps protocols may also be affected. I don't use it so I wouldn't know if it gets hosed or not. When I came back up (with the wrong sysconfig) there were some errors but nothing catastrophic. Once the configuration was correctly adjusted then all was back to normal. The only goofy thing was that my ed0 ( a nonamer NE2000) locked up for the first time *ever* while I was writing this mail message before. And it went 9 months without a burp before. :-/ Question: what is the new 'package startup' bit? I didn't see any explanative text or sample rc.d's regarding that. I noticed that the hooks for the popular software (apache, pcnfsd) had been removed. I've hacked them into rc.local until I understand the package startup section. (yeah, I should read CHANGES, but....) Poking through LINT I see some odd lines sprinkled in: First, an odd place for the ccd pseudo-device: # # BOUNCE_BUFFERS provides support for ISA DMA on machines with more # than 16 megabytes of memory. It doesn't hurt on other machines. # Some broken EISA and VLB hardware may need this, too. pseudo-device ccd 4 #Concatenated disk driver # # DUMMY_NOPS disables extra delays for some bus operations. The delays # are mostly for older systems and aren't used consistently. Probably # works OK on most EISA bus machines. # And then the infamous npx0 is in the middle of nowhere with no explanative text: # # MAXCONS is maximum number of virtual consoles, no more than 16 # default value: 12 # options "MAXCONS=16" device npx0 at isa? port "IO_NPX" irq 13 vector npxintr # # Optional ISA and EISA devices: Just some stuff I noticed. In all other cases, smooth transition. Hey, even the backported qcam driver compiled without a squeak. Thanks for the great release guys! Keep it up! Doug White | University of Oregon Internet: dwhite@resnet.uoregon.edu | Residence Networking Assistant http://gladstone.uoregon.edu/~dwhite | Computer Science Major