Date: Thu, 21 Mar 1996 11:05:48 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: jkh@time.cdrom.com (Jordan K. Hubbard) Cc: jgreco@brasil.moneng.mei.com, j@uriah.heep.sax.de, freebsd-current@FreeBSD.org Subject: Re: lost+found ??? Message-ID: <199603211805.LAA00425@phaeton.artisoft.com> In-Reply-To: <14569.827428382@time.cdrom.com> from "Jordan K. Hubbard" at Mar 21, 96 09:13:02 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > They _will_ also create it at fsck time if it does not exist, but it looks > > like that's handled as an "unusual" condition. > > Interesting, looks like I've frequently had this as an unusal > condition.. :-) The problem, of course, is that deleting files in lost+found will truncate back the size of the directory, to a minimum of 512 bytes, if you remove the things from the lost+found. The lost+found creation makes an 8k directory, but the first time you use it and do something about the recovered files, it's back down to 512 bytes (the directory block size). Traditionally, the directory has been pegged at 8k of allocated blocks to ensure that there is some place to put at least 8k of directory entries. This is probably too small for modern file system sizes any more. 8-(. That's it's considered "unusual" is proven by the need to run fsck again (my fsck patch simply maintains the correct link count for the create case -- something that wasn't, apparently, tested ever before, since it was an '"unusual" condition'). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199603211805.LAA00425>