Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley
        Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United
        Kingdom.
        Registered in England and Wales under Company Registration No. 3798903
From:   David Howells <dhowells@redhat.com>
In-Reply-To: <de425aeb-4064-733a-52ed-e702c640c36f@redhat.com>
References: <de425aeb-4064-733a-52ed-e702c640c36f@redhat.com> <20230526214142.958751-1-dhowells@redhat.com> <20230526214142.958751-3-dhowells@redhat.com>
To:     David Hildenbrand <david@redhat.com>
Cc:     dhowells@redhat.com, Christoph Hellwig <hch@infradead.org>,
        Lorenzo Stoakes <lstoakes@gmail.com>,
        Jens Axboe <axboe@kernel.dk>,
        Al Viro <viro@zeniv.linux.org.uk>,
        Matthew Wilcox <willy@infradead.org>, Jan Kara <jack@suse.cz>,
        Jeff Layton <jlayton@kernel.org>,
        Jason Gunthorpe <jgg@nvidia.com>,
        Logan Gunthorpe <logang@deltatee.com>,
        Hillf Danton <hdanton@sina.com>,
        Christian Brauner <brauner@kernel.org>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        linux-fsdevel@vger.kernel.org, linux-block@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-mm@kvack.org,
        Andrew Morton <akpm@linux-foundation.org>,
        John Hubbard <jhubbard@nvidia.com>
Subject: Re: [PATCH v4 2/3] mm: Provide a function to get an additional pin on a page
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <510618.1685521238.1@warthog.procyon.org.uk>
Content-Transfer-Encoding: quoted-printable
Date:   Wed, 31 May 2023 09:20:38 +0100
Message-ID: <510619.1685521238@warthog.procyon.org.uk>
Precedence: bulk

David Hildenbrand <david@redhat.com> wrote:

> > Provide a function to get an additional pin on a page that we already =
have
> > a pin on.  This will be used in fs/direct-io.c when dispatching multip=
le
> > bios to a page we've extracted from a user-backed iter rather than red=
oing
> > the extraction.
> > =

> =

> I guess this function is only used for "replicating" an existing pin, an=
d not
> changing the semantics of an existing pin: something that was pinned
> !FOLL_LONGTERM cannot suddenly become effectively pinned FOLL_LONGTERM.
> =

> Out of curiosity, could we end up passing in an anonymous page, or is th=
is
> almost exclusively for pagecache+zero pages? (I rememebr John H. had a s=
imilar
> patch where he said it would not apply to anon pages)

It has to be able to handle anything in a process's VM that you can
legitimately use as the buffer for a DIO read/write.  If we can get a pin =
on
it from pin_user_pages_fast(), then we need to be able to get an additiona=
l
pin on it.  For the moment it's just in the old DIO stuff, but it will alm=
ost
certainly be necessary in the networking code too to handle splicing into
network packets.

David