Date: Tue, 7 Jul 1998 10:36:22 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: abial@nask.pl (Andrzej Bialecki) Cc: tlambert@primenet.com, mike@smith.net.au, freebsd-current@FreeBSD.ORG Subject: Re: FreeBSD in less than 4MB RAM Message-ID: <199807071036.DAA00212@usr01.primenet.com> In-Reply-To: <Pine.NEB.3.95.980707100257.20507B-100000@korin.warman.org.pl> from "Andrzej Bialecki" at Jul 7, 98 10:06:10 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > > All you have to do is add the mountroot stuff to MSDOSFS - nobody would > > > complain if you did. > > > > I would. > > > > The "mountroot" and "mount" disctinctions are relatively stupid. > > > > The implementaiton of this is rather simple: > > [...but way above my head... :-( ] BS. I could do an ASCII art of it, if I weren't about to hit the sack... it's very easy to visualize. The hardest part about explaining it is explaining the byzantine way it currently works so you can see what's changing and how it's better off changed. > > I have implemented this three times (so far) in various versions > > of FreeBSD. > > Wow... So, what are the obstacles? As you present it, it has > only advantages... The first time, it was USL. After USL? It touches a lot of code at once. Every FS. And everything that got touched would need to get tested. It also changes the VFS interface, damaging compatability with other 4.4-derived OS's; I don't think this is as big a problem as it was the last time, before the Lite2 release was integrated, though, since FreeBSD is diverging quite a bit now. Also, many times, when this testing happens, you see bugs that were there all along (NFS, especially), but were not revealed until the testing. And then, of course, the last change is the obvious culprit. 8-). Part of the changes made it in last time; the root/non-root mount unification was a step towards it, since it documents the code distinction much better. Moving the SETMOUNTON and the vn covering is pretty obvious, if you read the mount code for two or more FS's, side-by-side. Just avoid NFS and CD9660 as documentation, and move part of the FFS mount to common code, and another part of it to a seperate VFSOP, and FFS is pretty much done. I might do it this coming weekend, if I can get g++ 2.8.1 to behave correctly before then (the headers are either installed in the wrong place, or they are incorrectly referenced by the compiler, and you end up with the g++ 2.7.2 includes and a g++ 2.8.1 compiler, which is a bad mix -- the port is only partly happy). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199807071036.DAA00212>