On Tue, Oct 08, 2019 at 12:42:09PM +0200, Jan Kara wrote:
> On Thu 03-10-19 21:33:45, Matthew Bobrowski wrote:
> The patch looks good to me. You can add:
>
> Reviewed-by: Jan Kara <jack@xxxxxxx>
Thanks Jan! :)
> It would just need small adjustments if you change patch 1 as I suggested:
I will await what you say in response to what my thoughts were aronud
ext4_set_iomap() before doing any updates here.
> > +static u16 ext4_iomap_check_delalloc(struct inode *inode,
> > + struct ext4_map_blocks *map)
> > +{
> > + struct extent_status es;
> > + ext4_lblk_t end = map->m_lblk + map->m_len - 1;
> > +
> > + ext4_es_find_extent_range(inode, &ext4_es_is_delayed, map->m_lblk,
> > + end, &es);
> > +
> > + /* Entire range is a hole */
> > + if (!es.es_len || es.es_lblk > end)
> > + return IOMAP_HOLE;
> > + if (es.es_lblk <= map->m_lblk) {
> > + ext4_lblk_t offset = 0;
> > +
> > + if (es.es_lblk < map->m_lblk)
> > + offset = map->m_lblk - es.es_lblk;
> > + map->m_lblk = es.es_lblk + offset;
> > + map->m_len = es.es_len - offset;
> > + return IOMAP_DELALLOC;
> > + }
> > +
> > + /* Range starts with a hole */
> > + map->m_len = es.es_lblk - map->m_lblk;
> > + return IOMAP_HOLE;
> > +}
>
> This function would then be IMO better off to directly update 'iomap' as
> needed after ext4_set_iomap() sets hole there.
As mentioned in 1/8, it would be nice to leave all iomap setting up to
ext4_set_iomap(), but if we're strongly against passing 'type', then
I'm happy to change it and update this to pass an 'iomap'.
--<M>--
