xfs
[Top] [All Lists]

Re: [PATCH 2/7] xfs: remove xfs_buf_get / xfs_buf_read

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: [PATCH 2/7] xfs: remove xfs_buf_get / xfs_buf_read
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Mon, 16 Nov 2009 06:03:31 -0500
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
In-reply-to: <20091116031512.GC9467@xxxxxxxxxxxxxxxx>
References: <20091114161717.233372584@xxxxxxxxxxxxxxxxxxxxxx> <20091114161801.411213050@xxxxxxxxxxxxxxxxxxxxxx> <20091116031512.GC9467@xxxxxxxxxxxxxxxx>
User-agent: Mutt/1.5.19 (2009-01-05)
On Mon, Nov 16, 2009 at 02:15:12PM +1100, Dave Chinner wrote:
> On Sat, Nov 14, 2009 at 11:17:19AM -0500, Christoph Hellwig wrote:
> > The two interfaces are highl confusing as they do have a flags argument
> > which gets ignored and replaced with default flags.  Switch the few 
> > remaining
> > callers to use xfs_buf_get_flags and xfs_buf_read_flags instead and remove
> > these wrappers.
> > 
> > Signed-off-by: Christoph Hellwig <hch@xxxxxx>
> 
> Personally I would have dropped the *_flags from the function names
> as well (so everything calls xfs_buf_get() or xfs_buf_read()) but
> it doesn't really matter....

Maybe.  The issue would be if we forward-port some code that expects it
can pass random crap in the flags field will get different behaviour
now.  Then again it'll blow up fast enough to notice..

> > +                           XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)),
> > +                           BTOBB(mp->m_sb.sb_blocksize), XBF_LOCK | 
> > XBF_MAPPED);
> 
> That looks wider than 80 columns. Same for the other conversions,
> too.

I'll fix it.

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