[btrfs-devel][patch]fix for btrfs_truncate_in_trans

[Chris Mason]

On Tue, 18 Sep 2007 21:31:54 +0800
"Yan Zheng" <yanzheng at 21cn.com> wrote:

> 2007/9/18, Chris Mason <chris.mason at oracle.com>:
> > >
> > > Thanks for the clarification
> > >
> > > Yes, I'm wrong.  But during test the non-exist bug, I have a new
> > > discovery. It's seem there is a race between the truncate
> > > operation and the delay allocation.  The output of 'ls -s' looks
> > > strange, when the truncate operation is before allocation really
> > > happen.
> > >
> > Odd, the i_blocks changes happen inside file_write, not at delalloc
> > time.  Could you please try to narrow the problem down?
> > -chris
> >
> When allocation doesn't really happen, there is no extent item. So
> btrfs_truncate_in_trans finds nothing and i_blocks isn't updated.

Ok, that makes sense.  Christoph, what fits best in with your delalloc
plans for XFS?  Should extent_invalidatepage return some information
about the number of delalloc bytes it found so that
btrfs_invalidatepage can drop i_blocks?

At the start of btrfs_truncate, we call btrfs_drop_extent_cache.  That
could also send back information on the number of delalloc extent maps
that were found.

-chris