Date: Sun, 28 Nov 1999 22:42:40 +0100 (CET) From: Andrzej Bialecki <abial@webgiro.com> To: Poul-Henning Kamp <phk@critter.freebsd.dk> Cc: cvs-committers@freebsd.org, cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/conf options src/sys/dev/md md.c src/sys/i386/conf GENERIC LINT PCCARD src/sys/kern vfs_conf.c src/sys/sys conf.h src/sys/ufs/mfs mfs_vfsops.c Message-ID: <Pine.BSF.4.05.9911282233270.82278-100000@freja.webgiro.com> In-Reply-To: <18050.943824574@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 28 Nov 1999, Poul-Henning Kamp wrote: > MFS is a gross hack (Kirks words when in polite company). One of these > hacks is hacking up a rootvp. We *really* want to have a rootvp, > ie a vnode for the root device, but for a compiled in or preloaded > image MFS doesn't have a vnode for "the thing it mounts". > > Md is a true memory-disk, which has a dev_t for the device and therefore > a valid vnode can be synthesized. > > more features and doc for the md driver will be forthcoming. (Wishes > for features are also welcome!) Ok, that was one thing. What about its impact on VM resources? I had a look at the sources, and the look quite simple compared to MFS. From my cursory look it seems, though, that MD will truely reserve all memory tha you requested at compile time, and it's not pageable, whereas MFS could use it up as needed (except that it wouldn't give it back :-)... Andrzej Bialecki // <abial@webgiro.com> WebGiro AB, Sweden (http://www.webgiro.com) // ------------------------------------------------------------------- // ------ FreeBSD: The Power to Serve. http://www.freebsd.org -------- // --- Small & Embedded FreeBSD: http://www.freebsd.org/~picobsd/ ---- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.05.9911282233270.82278-100000>