lkml.org 
[lkml]   [2011]   [Jan]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
DateFri, 7 Jan 2011 13:32:02 -0500
FromChristoph Hellwig <>
SubjectRe: [patch 8/8] fs: add i_op->sync_inode
On Fri, Jan 07, 2011 at 01:30:47PM -0500, Ted Ts'o wrote:
> > the write_inode in XFS is already trying to do it, it's just the
> > caller not copying with it:
> >
> > - if we can't get locks for a non-blocking ->write_inode we return
> > EAGAIN, and the callers sets the dirty bits again.
>
> I like that solution; it might be one of the easier ways to maintain
> backwards compatibility. Especially since (correct me if I am wrong)
> the simpler file systems which always write out the inode in the case
> of a non-blocking write_inode, say, like say the fat file system, are
> immune from this specific problem, right?

Yes. Also one of the patches in Nick's series actually implements
this already. Looks like we lost that fact when arguing about other
things..


\
 
 \ /
  Last update: 2011-01-07 19:35    [from the cache]
©2003-2010