On Tue, May 25, 2010 at 12:40:52PM -0400, Christoph Hellwig wrote:
> Looks good, but some minor nits on the comments below:
> Reviewed-by: Christoph Hellwig <hch@xxxxxx>
> > /*
> > + * Now we've locked out tail pushing and flushing by locking
> > + * the buffer, look for each inode in memory and attempt to
> > + * lock it. Any inode we get the locks on add it to the inode
> > + * buffer and set it up for being staled on buffer IO
> > + * completion.
> This comment reads a bit odd. The first thing we do in the loop is
> locking the buffer, so the "Now" at the beginning of the comment feels
> rather out of place. What about:
> * For each inode in memory attempt to add it to the inode
> * buffer and set it up for being staled on buffer IO
> * completion. This is safe as we've locked out tail
> * pushing and flushing by locking the buffer.
> * We have already marked every inode that was part of
> * a transaction stale above, which means there is no
> * point in even trying to lock them.
That reads better. I'll update it to match.