On 2020/03/28 17:31, hch@infradead.org wrote:

On Sat, Mar 28, 2020 at 08:27:09AM +0000, Damien Le Moal wrote:

quoted

On 2020/03/28 17:22, hch@infradead.org wrote:

quoted

On Sat, Mar 28, 2020 at 08:13:26AM +0000, Damien Le Moal wrote:

quoted

In any case, I think that the patch has value as a nice cleanup of the reset vs
reset all request operations (the latter going away). The side effect of it
being that open/close/finish all come for free with it. I would like to get it
in just for this nice cleanup.

Well, this keeps about the same amount of core code (and I'm not sure
it really is cleaner) while adding significantly more code in sd and
null_blk.  Not my classic definition of a cleanup.

sd is simplified (one less argument for setup of zone mgmt commands). But yes,
agreed that null_blk does need more code.

OK. Let's drop this for now then.

FYI, if you think the flag is cleaner than a new op you could resubmit
this but keep the flag reset specific for now.  That won't require new
code in null_blk for a use cases no callers uses anyway.

OK. I can do that.


-- 
Damien Le Moal
Western Digital Research