Date: Mon, 2 Dec 1996 01:11:47 -0500 (EST) From: "John S. Dyson" <toor@dyson.iquest.net> To: taob@io.org (Brian Tao) Cc: rnordier@iafrica.com, jkh@time.cdrom.com, hackers@FreeBSD.org Subject: Re: A simple way to crash your system. Message-ID: <199612020612.BAA04128@dyson.iquest.net> In-Reply-To: <Pine.BSF.3.95.961201205627.4096C-100000@nap.io.org> from "Brian Tao" at Dec 1, 96 09:02:04 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> On Tue, 26 Nov 1996, Robert Nordier wrote: > > > > *All* problems occurred with the DOS FS on a 64/63 IDE drive. FIPS > > was not necessarily used. In one case, the corrupted UFS fs was > > actually on another drive. > > Twice I've had ufs corruption with 2.2-ALPHA: > We have had at least one major bug in the vfs_bio code, that we weren't checking for the size of the block allocated by the filesystem. I had mistakenly assumed that MSDOSFS was mistakenly allocating 512 byte block buffers. MSDOSFS was doing the "right" thing all along allocating up to 32K or 64K buffers -- and that was breaking vfs_bio. Buffers were overrunning their allocated kva space of 16K. Bruce and/or Robert found this problem and brought it to my attention. The obvious solution of increasing the kva space in the buffer cache (per buffer) to 64K was not workable, but I rewrote part of the buffer allocation code to allow for more dynamic sizing of the buffer kva space (of course, managing the fragmentation issues, etc.) So, at least on MSDOSFS problem has been fixed, and this could explain at least one of the reasons for the filesystem corruptions associated with the usage of MSDOSFS. The corruptions would have happened whether or not one would write to the MSDOS filesystem. There are likely more bugs to be fixed, but at least this one should be gone now (in -current.) John dyson@freebsd.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199612020612.BAA04128>