From owner-freebsd-hackers Thu Nov 2 06:15:31 1995 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id GAA01151 for hackers-outgoing; Thu, 2 Nov 1995 06:15:31 -0800 Received: from Root.COM (implode.Root.COM [198.145.90.17]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id GAA01146 for ; Thu, 2 Nov 1995 06:15:27 -0800 Received: from corbin.Root.COM (corbin [198.145.90.50]) by Root.COM (8.6.12/8.6.5) with ESMTP id GAA02341; Thu, 2 Nov 1995 06:15:22 -0800 Received: from localhost (localhost [127.0.0.1]) by corbin.Root.COM (8.6.12/8.6.5) with SMTP id GAA02819; Thu, 2 Nov 1995 06:12:17 -0800 Message-Id: <199511021412.GAA02819@corbin.Root.COM> To: "Danny J. Zerkel" cc: freebsd-hackers@freebsd.org Subject: Re: More nits In-reply-to: Your message of "Thu, 02 Nov 95 07:57:11 EST." <199511021257.HAA10396@feephi.phofarm.com> From: David Greenman Reply-To: davidg@Root.COM Date: Thu, 02 Nov 1995 06:12:15 -0800 Sender: owner-hackers@freebsd.org Precedence: bulk >>>> Ummm, actually, I'm the one who made the changes to /etc/rc. I don't care >>>> about the CDROM, but I do think it's extremely important that the system not >>>> come up if some filesystems fail to mount. >>> >>>Well, for those of us that *do* care about the CDROM, how about >>>making these changes a little less draconian? Perhaps a two stage >>>mount? Everything but the CD and DOS for the first, non-optional stage, >>>then a second mount of CDs and DOS partitions that's allowed to >>>fail? >> >> I think it would be better to add yet another fstab option to specify that >>it's okay for it to fail. The offending parition that was so marked would not >>cause mount to exit with bad status. Perhaps the option could be called >>"optional". > >If you made the change to cause it to fail, perhaps the option should be >that not mounting causes a failure, instead of changing the default behavior >which so many people are used to. > >I think, as someone else pointed out in a later response, that adding more >nonstandard options is probably a bad idea, especially if you need to use >them to get standard behavior. Okay, I think the solution is starting to become clear. Basically, mount should return a failure code only in the case of it failing to mount a filesystem because it is dirty (which is the entire reason behind the check in /etc/rc in the first place). It should not fail if the only problem is that there is no media in a drive. I'll have to look into this further to see how difficult this is going to be to implement (I don't know if the error codes returned by the kernel are unique enough, for instance, so it may require kernel changes, too). Will this make people happy? -DG