Date: Wed, 14 Dec 2011 16:44:11 +1100 (EST) From: Bruce Evans <brde@optusnet.com.au> To: Bruce Evans <brde@optusnet.com.au> Cc: =?ISO-8859-1?Q?Olivier_Cochard=2DLabb=E9?= <olivier@cochard.me>, freebsd-bugs@freebsd.org Subject: Re: bin/162659: Can' t install FreeBSD-amd64-9.0-RC2 on disk less than 6Go (/mnt: out of inodes) Message-ID: <20111214152109.J2567@besplex.bde.org> In-Reply-To: <20111214125712.P1950@besplex.bde.org> References: <201112131030.pBDAUDCI065085@freefall.freebsd.org> <20111214125712.P1950@besplex.bde.org>
next in thread | previous in thread | raw e-mail | index | archive | help
This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --0-1046583486-1323841451=:2567 Content-Type: TEXT/PLAIN; charset=X-UNKNOWN; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE On Wed, 14 Dec 2011, Bruce Evans wrote: > On Tue, 13 Dec 2011, [ISO-8859-1] Olivier Cochard-Labb=E9 wrote: > >> The following reply was made to PR bin/162659; it has been noted by GNAT= S. > > THe following reply probably won't be noted by gnats, since gnats > created an incomplete Cc list as usual. > >> I've did some tests: The number of inode after a fresh install of >> 9.0-RC2 is half the number of inode after the fresh install of a 8.2. >> I've install a 6Gio hard drive, MBR, no swap partition and all in / >> (under VirtualBox). > > 9.0 has the dubious change of increasing the default block size in > newfs from 16K to 32K (and, correspondingly, increasing the default > fragment size from 2K to 4K). It is missing the corresponding change > of increasing the number of fragments per inode (NFPI) from 4 to 8. That should be "reducing ... (NFPI) from 4 to 2". > Thus if you blindly accept the newfs defaults, you get half as many > inodes as before. I forget to mention than newfs on a regular file to verify this is quite broken. You now have to write a label to the regular file, since the usual mechanisms for specifying the geometry when there is no label (-S sector-size -s size or even -Tfoo) no longer work. I want to use the usual mechanisms and not mess around with labels. There are relatively minor documentation bugs related to this. -S, -s and -T are still documented, but after removal of many geometry- related options near -S, there are plural references to the single -S option and other anachronisms related to geometry, which together result in the description of -S being mostly wrong. Also, -s is arguably a geometry option but it is not in the geometry section. The geometry section with only -S in it makes -S unsorted relative to the other -[AZ] options. Bruce --0-1046583486-1323841451=:2567--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20111214152109.J2567>