Date: Wed, 15 Jul 1998 16:46:13 +0100 From: "Alton, Matthew" <Matthew.Alton@anheuser-busch.com> To: "Hancock, Michael (Internet)" <michaelh@cet.co.jp> Cc: "'FreeBSD-fs@FreeBSD.ORG'" <FreeBSD-fs@FreeBSD.ORG> Subject: RE: LFS Hacking Message-ID: <31B3F0BF1C40D11192A700805FD48BF90177660E@STLABCEXG011>
next in thread | raw e-mail | index | archive | help
> -----Original Message----- > From: Hancock, Michael (Internet) > Sent: Wednesday, July 15, 1998 12:54 AM > To: Alton, Matthew > Cc: 'FreeBSD-fs@FreeBSD.ORG' > Subject: Re: LFS Hacking > > On Tue, 14 Jul 1998, Alton, Matthew wrote: > > > OK people, I'm hip-deep in LFS code and the vfs/vnode interface > > and all the other goodies we need to make LFS happen. Are there > > any wish-list type requests or design deltas or anything before I > > move into the 'you should have said something earlier' phase? > > Right now, I'm just trying to stabilize the existing code. Once > this > > done, I'll apply lessons learned to logging metadata only to make > > a journaling filesystem. > > Just go to it. If you get LFS working that'd be excellent. > > You do understand that a log structured filesystem is quite a > different > kettle of fish than a meta-data journaling filesystem, right? [Alton, Matthew] Yupper. For the metadata-only one I would like to use ufs-style fixed-position inodes in conjunction with asyncronous I/O and do pending/commit flags in a IBM-JFS type fixed-size log. The 'lessons learned' will be general vfs-oid techniques &etc. I'm still not overly happy with the vfs/vnode structure but the chapter from Vahalla's _UNIX Internals_ has shed considerable light on the subject. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-fs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?31B3F0BF1C40D11192A700805FD48BF90177660E>
