From: Christoph Hellwig <hch@infradead.org>
To: Christopher Lameter <cl@linux.com>
Cc: john.hubbard@gmail.com, Matthew Wilcox <willy@infradead.org>,
Michal Hocko <mhocko@kernel.org>, Jason Gunthorpe <jgg@ziepe.ca>,
Dan Williams <dan.j.williams@intel.com>, Jan Kara <jack@suse.cz>,
linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>,
linux-rdma <linux-rdma@vger.kernel.org>,
John Hubbard <jhubbard@nvidia.com>
Subject: Re: [PATCH 0/2] mm: gup: don't unmap or drop filesystem buffers
Date: Mon, 18 Jun 2018 01:10:03 -0700 [thread overview]
Message-ID: <20180618081003.GA20927@infradead.org> (raw)
In-Reply-To: <010001640fbe0dd8-f999e7f6-7b6e-4deb-b073-0c572006727d-000000@email.amazonses.com>
On Sun, Jun 17, 2018 at 09:54:31PM +0000, Christopher Lameter wrote:
> On Sat, 16 Jun 2018, john.hubbard@gmail.com wrote:
>
> > I've come up with what I claim is a simple, robust fix, but...I'm
> > presuming to burn a struct page flag, and limit it to 64-bit arches, in
> > order to get there. Given that the problem is old (Jason Gunthorpe noted
> > that RDMA has been living with this problem since 2005), I think it's
> > worth it.
> >
> > Leaving the new page flag set "nearly forever" is not great, but on the
> > other hand, once the page is actually freed, the flag does get cleared.
> > It seems like an acceptable tradeoff, given that we only get one bit
> > (and are lucky to even have that).
>
> This is not robust. Multiple processes may register a page with the RDMA
> subsystem. How do you decide when to clear the flag? I think you would
> need an additional refcount for the number of times the page was
> registered.
And it's not just RDMA that is using get_user_pages. We have tons of
users that do short, spurious get_user_pages do do zero copy operations.
We can't leave the page in a wrecked state after that.
> I still think the cleanest solution here is to require mmu notifier
> callbacks and to not pin the page in the first place. If a NIC does not
> support a hardware mmu then it can still simulate it in software by
> holding off the ummapping the mmu notifier callback until any pending
> operation is complete and then invalidate the mapping so that future
> operations require a remapping (or refaulting).
Sounds ok for RDMA, not going to help for most other users.
prev parent reply other threads:[~2018-06-18 8:10 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-06-17 1:25 [PATCH 0/2] mm: gup: don't unmap or drop filesystem buffers john.hubbard
2018-06-17 1:25 ` [PATCH 1/2] consolidate get_user_pages error handling john.hubbard
2018-06-17 1:25 ` [PATCH 2/2] mm: set PG_dma_pinned on get_user_pages*() john.hubbard
2018-06-17 19:53 ` Dan Williams
2018-06-17 20:04 ` Jason Gunthorpe
2018-06-17 20:10 ` Dan Williams
2018-06-17 20:28 ` John Hubbard
2018-06-18 8:12 ` Christoph Hellwig
2018-06-18 17:50 ` John Hubbard
2018-06-18 17:56 ` Dan Williams
2018-06-18 18:14 ` John Hubbard
2018-06-18 19:21 ` Dan Williams
2018-06-18 19:31 ` Jason Gunthorpe
2018-06-18 20:04 ` Dan Williams
2018-06-18 21:36 ` John Hubbard
2018-06-19 8:29 ` Jan Kara
2018-06-19 9:02 ` Matthew Wilcox
2018-06-19 10:41 ` Jan Kara
2018-06-19 18:11 ` John Hubbard
2018-06-20 1:24 ` Dan Williams
2018-06-20 1:34 ` John Hubbard
2018-06-20 1:57 ` Dan Williams
2018-06-20 2:03 ` John Hubbard
2018-06-20 12:08 ` Jan Kara
2018-06-20 22:55 ` John Hubbard
2018-06-21 16:30 ` Jan Kara
2018-06-25 15:21 ` Jan Kara
2018-06-25 19:03 ` John Hubbard
2018-06-26 7:52 ` Jan Kara
2018-06-26 6:31 ` John Hubbard
2018-06-26 11:48 ` Jan Kara
2018-06-26 13:47 ` Michal Hocko
2018-06-26 16:48 ` Jan Kara
2018-06-27 11:32 ` Michal Hocko
2018-06-27 11:53 ` Jan Kara
2018-06-27 11:59 ` Michal Hocko
2018-06-27 12:42 ` Jan Kara
2018-06-27 14:57 ` Jason Gunthorpe
2018-06-27 17:02 ` Jan Kara
2018-06-28 2:42 ` John Hubbard
2018-06-28 9:17 ` Jan Kara
2018-07-02 5:52 ` Leon Romanovsky
2018-07-02 6:10 ` John Hubbard
2018-07-02 6:34 ` Leon Romanovsky
2018-07-02 6:41 ` John Hubbard
2018-07-02 10:36 ` Michal Hocko
2018-07-02 7:02 ` Jan Kara
2018-07-02 14:48 ` Michal Hocko
2018-07-02 6:58 ` Jan Kara
2018-06-18 8:11 ` Christoph Hellwig
2018-06-19 6:15 ` Leon Romanovsky
2018-06-17 22:19 ` John Hubbard
2018-06-18 7:56 ` Christoph Hellwig
2018-06-18 17:44 ` John Hubbard
2018-06-17 21:54 ` [PATCH 0/2] mm: gup: don't unmap or drop filesystem buffers Christopher Lameter
2018-06-17 22:23 ` John Hubbard
2018-06-18 8:10 ` Christoph Hellwig [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180618081003.GA20927@infradead.org \
--to=hch@infradead.org \
--cc=cl@linux.com \
--cc=dan.j.williams@intel.com \
--cc=jack@suse.cz \
--cc=jgg@ziepe.ca \
--cc=jhubbard@nvidia.com \
--cc=john.hubbard@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-rdma@vger.kernel.org \
--cc=mhocko@kernel.org \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).