Re: [PATCH] xfs: optimise away log forces on timestamp updates for fdatasync

[Dave Chinner <david@xxxxxxxxxxxxx>]

On Mon, Oct 26, 2015 at 07:49:29AM -0400, Brian Foster wrote:
> On Mon, Oct 26, 2015 at 04:07:20PM +1100, Dave Chinner wrote:
> > > Also, is it me or are we sending an unconditional flush in the hunk
> > > following the log force call in xfs_file_fsync() (even if we've skipped
> > > the log force)?
> > 
> > The flush is needed - fdatasync needs to guarantee the data is
> > on stable storage even if no metadata needs to be written to the
> > journal.
> > 
> 
> Ok. Well it's too bad we don't get any feedback about what was written
> from the filemap_write_and_wait_range() call. As it is, we send a flush
> even if there's nothing to write back.

Direct IO still needs cache flushes.

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx