Re: vfs: Add MS_FLUSHONFSYNC mount flag

["Fernando Luis Vázquez Cao" <fernando@oss.ntt.co.jp>]

On Sat, 2009-02-14 at 22:03 +0900, Fernando Luis Vázquez Cao wrote:
> > > 2- Modify file_fsync() so that it checks whether FLUSHONFSYNC is set and
> > > flushes the underlying device accordingly. With this we would cover all
> > > filesystems that use the vfs-provided file_fsync() as their fsync method
> > > (commonly used filesystems such as fat fall in this group).
> > 
> > Just make it flush the block device.
> 
> I wrote a patch that does exactly that but, in addition, it checks
> whether FLUSHONFSYNC is set to avoid sending unnecessary flushes down
> the block layer (this patch is not included in this patch-set, but I
> will add it in the next iteration).

I probably should have mentioned in my proposal all filesystems would
just call the helper function block_flush_device() to flush the
underlying block device, unconditionally.

This helper function looks like this:

+/* Issue flush of write caches on the block device */
+int block_flush_device(struct super_block *sb)
+{
+       int ret = 0;
+
+       if (!(sb->s_flags & MS_FLUSHONFSYNC))
+               return ret;
+
+       ret = blkdev_issue_flush(sb->s_bdev, NULL);
+
+       if (ret == -EOPNOTSUPP)
+               return 0;
+
+       return ret;
+}

As you can see the check for flushonfsync is done here, so changing my
patches along the lines you suggest would be a trivial two lines patch.

- Fernando

> As I mentioned above, if everyone thinks this small optimization not
> elegant or an undesirable layering violation I will remove it.