Re: [RFC PATCH v2 1/3] mm: Don't pin ZERO_PAGE in pin_user_pages()

To: David Howells <dhowells@xxxxxxxxxx>
Subject: Re: [RFC PATCH v2 1/3] mm: Don't pin ZERO_PAGE in pin_user_pages()
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Fri, 26 May 2023 01:22:54 -0700
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, David Hildenbrand <david@xxxxxxxxxx>, Jens Axboe <axboe@xxxxxxxxx>, Al Viro <viro@xxxxxxxxxxxxxxxxxx>, Matthew Wilcox <willy@xxxxxxxxxxxxx>, Jan Kara <jack@xxxxxxx>, Jeff Layton <jlayton@xxxxxxxxxx>, Jason Gunthorpe <jgg@xxxxxxxxxx>, Logan Gunthorpe <logang@xxxxxxxxxxxx>, Hillf Danton <hdanton@xxxxxxxx>, Christian Brauner <brauner@xxxxxxxxxx>, Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>, linux-fsdevel@xxxxxxxxxxxxxxx, linux-block@xxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, linux-mm@xxxxxxxxx, Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
In-reply-to: <20230525223953.225496-2-dhowells@redhat.com>

Shouldn't unpin_user_pages and bio_release_page also be updated to
skip a zero page here?

Follow-Ups:
- Re: [RFC PATCH v2 1/3] mm: Don't pin ZERO_PAGE in pin_user_pages()
  - From: Lorenzo Stoakes

References:
- [RFC PATCH v2 0/3] block: Make old dio use iov_iter_extract_pages() and page pinning
  - From: David Howells
- [RFC PATCH v2 1/3] mm: Don't pin ZERO_PAGE in pin_user_pages()
  - From: David Howells