Re: [PATCH v3 3/4] dm: add infrastructure for DAX support

[Mike Snitzer]

On Thu, Jun 23 2016 at  7:36pm -0400,
Kani, Toshimitsu <toshi.kani@xxxxxxx> wrote:

> On Thu, 2016-06-23 at 17:05 -0400, Mike Snitzer wrote:
>  :
> > +static int device_supports_dax(struct dm_target *ti, struct dm_dev *dev,
> > +			       sector_t start, sector_t len, void *data)
> > +{
> > +	struct request_queue *q = bdev_get_queue(dev->bdev);
> > +
> > +	return q && blk_queue_dax(q);
> > +}
> > +
> > +static bool dm_table_supports_dax(struct dm_table *t)
> > +{
> > +	struct dm_target *ti;
> > +	unsigned i = 0;
> > +
> > +	/* Ensure that all targets support DAX. */
> > +	while (i < dm_table_get_num_targets(t)) {
> > +		ti = dm_table_get_target(t, i++);
> > +
> > +		if (!ti->type->direct_access)
> > +			return false;
> > +
> > +		if (!ti->type->iterate_devices ||
> > +		    !ti->type->iterate_devices(ti, device_supports_dax,
> > NULL))
> > +			return false;
> > +	}
> > +
> > +	return true;
> > +}
> > +
> 
> Hi Mike,
> 
> Thanks for the update.  I have a question about the above change.  Targets may
> have their own parameters.  For instance, dm-stripe has 'chunk_size', which is
> checked in stripe_ctr().  DAX adds additional restriction that chunk_size
> needs to be aligned by page size.  So, I think we need to keep target
> responsible to verify if DAX can be supported.  What do you think?

We've never had to concern the dm-stripe target with hardware
specific chunk_size validation.  The user is able to specify the
chunk_size via lvm2's lvcreate -I argument.  Yes this gives users enough
rope to hang themselves but it is very easy to configure a dm-stripe
device with the appropriate chunk size (PAGE_SIZE) from userspace.

But lvm2 could even be trained to make sure the chunk_size is a factor
of physical_block_size (PAGE_SIZE in the case of pmem) if the underlying
devices export queue/dax=1