Re: [PATCH 4.4 48/91] ext4: fix inline data error paths

[Greg Kroah-Hartman]

On Fri, Mar 10, 2017 at 04:48:52PM +0000, Ben Hutchings wrote:
> On Fri, 2017-03-10 at 10:08 +0100, Greg Kroah-Hartman wrote:
> > 4.4-stable review patch.  If anyone has any objections, please let me
> > know.
> > 
> > ------------------
> > 
> > From: Theodore Ts'o <tytso@xxxxxxx>
> > 
> > commit eb5efbcb762aee4b454b04f7115f73ccbcf8f0ef upstream.
> > 
> > The write_end() function must always unlock the page and drop its ref
> > count, even on an error.
> 
> This looks like a theoretical rather than a real issue, because I can't
> see how ext4_write_inline_data_end() ever returns an error code.

I'll leave that up to Ted to justify :)

> > Signed-off-by: Theodore Ts'o <tytso@xxxxxxx>
> > Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> > 
> > ---
> >  fs/ext4/inline.c |    9 ++++++++-
> >  fs/ext4/inode.c  |   20 +++++++++++++++-----
> >  2 files changed, 23 insertions(+), 6 deletions(-)
> > 
> > --- a/fs/ext4/inline.c
> > +++ b/fs/ext4/inline.c
> > @@ -933,8 +933,15 @@ int ext4_da_write_inline_data_end(struct
> > >  				  struct page *page)
> >  {
> >  	int i_size_changed = 0;
> > +	int ret;
> >  
> > -	copied = ext4_write_inline_data_end(inode, pos, len, copied, page);
> > +	ret = ext4_write_inline_data_end(inode, pos, len, copied, page);
> > +	if (ret < 0) {
> > +		unlock_page(page);
> > +		put_page(page);
> [...]
> 
> For 4.4 each put_page() should ideally be changed to
> page_cache_release().  It makes no practical difference but would be
> consistent with other paths.

As it's still the same logic, I'd prefer to stick to what newer kernels
do if at all possible.

thanks,

greg k-h