[Top] [All Lists]

Re: fs/xfs/xfs_dmapi.c vanished from CVS?

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: fs/xfs/xfs_dmapi.c vanished from CVS?
From: Nathan Scott <nathans@xxxxxxx>
Date: Fri, 17 Feb 2006 06:24:12 +1100
Cc: Eric Sandeen <sandeen@xxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <20060216190022.GA17824@infradead.org>; from hch@infradead.org on Thu, Feb 16, 2006 at 07:00:22PM +0000
References: <20060216144537.GA12409@infradead.org> <43F4C979.3080802@sgi.com> <20060216190022.GA17824@infradead.org>
Sender: linux-xfs-bounce@xxxxxxxxxxx
User-agent: Mutt/1.2.5i
On Thu, Feb 16, 2006 at 07:00:22PM +0000, Christoph Hellwig wrote:
> On Thu, Feb 16, 2006 at 12:50:33PM -0600, Eric Sandeen wrote:
> > Christoph Hellwig wrote:
> > >I tried to rediff and send an older patch, but fs/xfs/xfs_dmapi.c in the
> > >CVS tree vanished.  It's still in the SuSE kernel, though including the
> > >related hacks.  What's the story here?
> > >
> > 
> > http://oss.sgi.com/cgi-bin/cvsweb.cgi/xfs-linux/xfs_rw.h.diff?r1=text&tr1=1.78&r2=text&tr2=1.77&f=h
> > xfs_dmapi.c - 1.142 - renamed to dmapi/xfs_dm.c 1.1
> > 
> > from nathan's most recent change...
> A thanks, must be the 1^16th location of the dmapi code in the Linux tree :)

Heh.  The problem was xfs_dmapi.c calling dmapi.ko symbols directly,
which meant the dmapi module had to be loaded all the time, which was
reported as an issue from a certain distro.

> Use xfs_vtoi and xfs_vfstom in the dmapi code instead of opencoding
> behaviour lookups.

Thanks Christoph, I'll get that merged.



<Prev in Thread] Current Thread [Next in Thread>