From: Alistair Popple <apopple@nvidia.com>
To: Peter Xu <peterx@redhat.com>
Cc: <linux-mm@kvack.org>, <nouveau@lists.freedesktop.org>,
<bskeggs@redhat.com>, <akpm@linux-foundation.org>,
<linux-doc@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<dri-devel@lists.freedesktop.org>, <jhubbard@nvidia.com>,
<rcampbell@nvidia.com>, <jglisse@redhat.com>, <jgg@nvidia.com>,
<hch@infradead.org>, <daniel@ffwll.ch>, <willy@infradead.org>,
<bsingharora@gmail.com>, Christoph Hellwig <hch@lst.de>
Subject: Re: [PATCH v8 5/8] mm: Device exclusive memory access
Date: Wed, 19 May 2021 20:49:01 +1000 [thread overview]
Message-ID: <3859486.fHISG1RMxY@nvdebian> (raw)
In-Reply-To: <YKQutvAa3NlgGaMm@t490s>
On Wednesday, 19 May 2021 7:16:38 AM AEST Peter Xu wrote:
> External email: Use caution opening links or attachments
>
>
> On Wed, Apr 07, 2021 at 06:42:35PM +1000, Alistair Popple wrote:
>
> [...]
>
> > +static bool try_to_protect(struct page *page, struct mm_struct *mm,
> > + unsigned long address, void *arg)
> > +{
> > + struct ttp_args ttp = {
> > + .mm = mm,
> > + .address = address,
> > + .arg = arg,
> > + .valid = false,
> > + };
> > + struct rmap_walk_control rwc = {
> > + .rmap_one = try_to_protect_one,
> > + .done = page_not_mapped,
> > + .anon_lock = page_lock_anon_vma_read,
> > + .arg = &ttp,
> > + };
> > +
> > + /*
> > + * Restrict to anonymous pages for now to avoid potential writeback
> > + * issues.
> > + */
> > + if (!PageAnon(page))
> > + return false;
> > +
> > + /*
> > + * During exec, a temporary VMA is setup and later moved.
> > + * The VMA is moved under the anon_vma lock but not the
> > + * page tables leading to a race where migration cannot
> > + * find the migration ptes. Rather than increasing the
> > + * locking requirements of exec(), migration skips
> > + * temporary VMAs until after exec() completes.
> > + */
> > + if (!PageKsm(page) && PageAnon(page))
> > + rwc.invalid_vma = invalid_migration_vma;
> > +
> > + rmap_walk(page, &rwc);
> > +
> > + return ttp.valid && !page_mapcount(page);
> > +}
>
> I raised a question in the other thread regarding fork():
>
> https://lore.kernel.org/lkml/YKQjmtMo+YQGx%2FwZ@t490s/
>
> While I suddenly noticed that we may have similar issues even if we fork()
> before creating the ptes.
>
> In that case, we may see multiple read-only ptes pointing to the same page.
> We will convert all of them into device exclusive read ptes in rmap_walk()
> above, however how do we guarantee after all COW done in the parent and all
> the childs processes, the device owned page will be returned to the parent?
I assume you are talking about a fork() followed by a call to
make_device_exclusive()? I think this should be ok because
make_device_exclusive() always calls GUP with FOLL_WRITE both to break COW and
because a device performing atomic operations needs to write to the page. I
suppose a comment here highlighting the need to break COW to avoid this
scenario would be useful though.
> E.g., if parent accesses the page earlier than the children processes
> (actually, as long as not the last one), do_wp_page() will do COW for parent
> on this page because refcount(page)>1, then the page seems to get lost to a
> random child too..
>
> To resolve all these complexity, not sure whether try_to_protect() could
> enforce VM_DONTCOPY (needs madvise MADV_DONTFORK in the user app), meanwhile
> make sure mapcount(page)==1 before granting the page to the device, so that
> this will guarantee this mm owns this page forever, I think? It'll
> simplify fork() too as a side effect, since VM_DONTCOPY vma go away when
> fork.
>
> --
> Peter Xu
next prev parent reply other threads:[~2021-05-19 10:49 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-07 8:42 [PATCH v8 0/8] Add support for SVM atomics in Nouveau Alistair Popple
2021-04-07 8:42 ` [PATCH v8 1/8] mm: Remove special swap entry functions Alistair Popple
2021-05-18 2:17 ` Peter Xu
2021-05-18 11:58 ` Alistair Popple
2021-05-18 14:17 ` Peter Xu
2021-04-07 8:42 ` [PATCH v8 2/8] mm/swapops: Rework swap entry manipulation code Alistair Popple
2021-04-07 8:42 ` [PATCH v8 3/8] mm/rmap: Split try_to_munlock from try_to_unmap Alistair Popple
2021-05-18 20:04 ` Liam Howlett
2021-05-19 12:38 ` Alistair Popple
2021-05-20 20:24 ` Liam Howlett
2021-05-21 2:23 ` Alistair Popple
2021-04-07 8:42 ` [PATCH v8 4/8] mm/rmap: Split migration into its own function Alistair Popple
2021-04-07 8:42 ` [PATCH v8 5/8] mm: Device exclusive memory access Alistair Popple
2021-05-18 2:08 ` Peter Xu
2021-05-18 13:19 ` Alistair Popple
2021-05-18 17:27 ` Peter Xu
2021-05-18 17:33 ` Jason Gunthorpe
2021-05-18 18:01 ` Peter Xu
2021-05-18 19:45 ` Jason Gunthorpe
2021-05-18 20:29 ` Peter Xu
2021-05-18 23:03 ` Jason Gunthorpe
2021-05-18 23:45 ` Peter Xu
2021-05-19 11:04 ` Alistair Popple
2021-05-19 12:15 ` Peter Xu
2021-05-19 13:11 ` Alistair Popple
2021-05-19 14:04 ` Peter Xu
2021-05-19 13:28 ` Jason Gunthorpe
2021-05-19 14:09 ` Peter Xu
2021-05-19 18:11 ` Jason Gunthorpe
2021-05-19 11:35 ` Alistair Popple
2021-05-19 12:21 ` Peter Xu
2021-05-19 12:46 ` Alistair Popple
2021-05-21 6:53 ` Alistair Popple
2021-05-18 21:16 ` Peter Xu
2021-05-19 10:49 ` Alistair Popple [this message]
2021-05-19 12:24 ` Peter Xu
2021-05-19 12:46 ` Alistair Popple
2021-04-07 8:42 ` [PATCH v8 6/8] mm: Selftests for exclusive device memory Alistair Popple
2021-04-07 8:42 ` [PATCH v8 7/8] nouveau/svm: Refactor nouveau_range_fault Alistair Popple
2021-04-07 8:42 ` [PATCH v8 8/8] nouveau/svm: Implement atomic SVM access Alistair Popple
2021-05-21 4:04 ` Ben Skeggs
2021-05-06 7:43 ` [PATCH v8 0/8] Add support for SVM atomics in Nouveau Alistair Popple
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=3859486.fHISG1RMxY@nvdebian \
--to=apopple@nvidia.com \
--cc=akpm@linux-foundation.org \
--cc=bsingharora@gmail.com \
--cc=bskeggs@redhat.com \
--cc=daniel@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=hch@infradead.org \
--cc=hch@lst.de \
--cc=jgg@nvidia.com \
--cc=jglisse@redhat.com \
--cc=jhubbard@nvidia.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=nouveau@lists.freedesktop.org \
--cc=peterx@redhat.com \
--cc=rcampbell@nvidia.com \
--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).