Re: Phyr Starter

From: John Hubbard <jhubbard@nvidia.com>
To: Christoph Hellwig <hch@lst.de>
Cc: Matthew Wilcox <willy@infradead.org>,
	linux-kernel@vger.kernel.org, Jason Gunthorpe <jgg@nvidia.com>,
	Joao Martins <joao.m.martins@oracle.com>,
	Logan Gunthorpe <logang@deltatee.com>,
	Ming Lei <ming.lei@redhat.com>,
	linux-block@vger.kernel.org, netdev@vger.kernel.org,
	linux-mm@kvack.org, linux-rdma@vger.kernel.org,
	dri-devel@lists.freedesktop.org, nvdimm@lists.linux.dev
Subject: Re: Phyr Starter
Date: Thu, 20 Jan 2022 13:35:23 -0800	[thread overview]
Message-ID: <a6b65260-669b-65d8-c20f-0d75e0393200@nvidia.com> (raw)
In-Reply-To: <20220120141219.GB11707@lst.de>

On 1/20/22 6:12 AM, Christoph Hellwig wrote:
> On Tue, Jan 11, 2022 at 12:17:18AM -0800, John Hubbard wrote:
>> Zooming in on the pinning aspect for a moment: last time I attempted to
>> convert O_DIRECT callers from gup to pup, I recall wanting very much to
>> record, in each bio_vec, whether these pages were acquired via FOLL_PIN,
>> or some non-FOLL_PIN method. Because at the end of the IO, it is not
>> easy to disentangle which pages require put_page() and which require
>> unpin_user_page*().
> 
> I don't think that is a problem.  Pinning only need to happen for
> ITER_IOVEC, and the only non-user pages there is the ZERO_PAGE added
> for padding that can be special cased.

I am really glad to hear you say that. Because I just worked through it
again in detail yesterday (including your and others' old emails about
this), and tentatively reached the same conclusion from seeing the call 
paths. But I wanted to confirm with someone who actually knows this code 
well, and that's not me. :)

Things like dio_refill_pages() are mixing in the zero page, but like you 
say, that can be handled. I have a few ideas for that.

Now that the goal is a considerably narrower as compared to in 2019 
("convert DIO callers to pup", instead of "convert the world to pup", 
ha), this looks quite feasible after all.

thanks,
-- 
John Hubbard
NVIDIA