Date: Sun, 3 Apr 2005 09:53:17 -0700 From: Alfred Perlstein <alfred@freebsd.org> To: Scott Long <scottl@freebsd.org> Cc: cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/ufs/ffs ffs_extern.h ffs_softdep.c ffs_vfsops.c Message-ID: <20050403165317.GS60345@elvis.mu.org> In-Reply-To: <20050403074335.A35686@pooker.samsco.org> References: <200504031029.j33ATtAX021544@repoman.freebsd.org> <20050403074335.A35686@pooker.samsco.org>
next in thread | previous in thread | raw e-mail | index | archive | help
* Scott Long <scottl@freebsd.org> [050403 06:41] wrote: > On Sun, 3 Apr 2005, Jeff Roberson wrote: > >jeff 2005-04-03 10:29:55 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/ufs/ffs ffs_extern.h ffs_softdep.c ffs_vfsops.c > > Log: > > - Move the contents of softdep_disk_prewrite into ffs_geom_strategy to > > fix > > two bugs. > > - ffs_disk_prewrite was pulling the vp from the buf and checking for > > COPYONWRITE, when really it wanted the vp from the bufobj that we're > > writing to, which is the devvp. This lead to us skipping the copy on > > write to all file data, which significantly broke snapshots for the > > last few months. > > - When the SOFTUPDATES option was not included in the kernel config we > > would also skip the copy on write check, which would effectively > > disable > > snapshots. > > I thought that snapshots required softupdates? Not really, it's just the combination of the two that allows for bgfsck. Actually I'm trying to think of the situation where softdep is actually required because sync metadata updates should do the same thing as softdeps in theory (just slower). -- - Alfred Perlstein - Research Engineering Development Inc. - email: bright@mu.org cell: 408-480-4684
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050403165317.GS60345>