On Wed, Nov 03, 2021 at 06:09:57PM +0000, Jane Chu wrote:

This is clearer, I've looked at your 'dax-devirtualize' patch which 
removes pmem_copy_to/from_iter, and as you mentioned before,
a separate API for poison-clearing is needed. So how about I go ahead
rebase my earlier patch
 
https://lore.kernel.org/lkml/20210914233132.3680546-2-jane.chu@oracle.com/ (local)
on 'dax-devirtualize', provide dm support for clear-poison?
That way, the non-dax 99% of the pwrite use-cases aren't impacted at all
and we resolve the urgent pmem poison-clearing issue?

FYI, I really do like the in-kernel interface in that series.  And
now that we discussed all the effects I also think that automatically
doing the clear on EIO is a good idea.

I'll go back to the series and will reply with a few nitpicks.