From: Alistair Popple <apopple@nvidia.com>
To: Tal Zussman <tz2294@columbia.edu>
Cc: "Gregory Price (Meta)" <gourry@gourry.net>,
"Andrew Morton" <akpm@linux-foundation.org>,
"Mike Rapoport" <rppt@kernel.org>, "Peter Xu" <peterx@redhat.com>,
"James Houghton" <jthoughton@google.com>,
"Andrea Arcangeli" <aarcange@redhat.com>,
"Suren Baghdasaryan" <surenb@google.com>,
"Balbir Singh" <balbirs@nvidia.com>,
"Liam R. Howlett" <liam@infradead.org>,
"Lorenzo Stoakes" <ljs@kernel.org>,
"Vlastimil Babka" <vbabka@kernel.org>,
"Jann Horn" <jannh@google.com>,
"Pedro Falcato" <pfalcato@suse.de>,
"Michał Mirosław" <mirq-linux@rere.qmqm.pl>,
"Andrei Vagin" <avagin@gmail.com>,
"Stephen Rothwell" <sfr@canb.auug.org.au>,
"Arnd Bergmann" <arnd@arndb.de>,
"Muhammad Usama Anjum" <usama.anjum@arm.com>,
"David Hildenbrand" <david@kernel.org>, "Zi Yan" <ziy@nvidia.com>,
"Matthew Brost" <matthew.brost@intel.com>,
"Joshua Hahn" <joshua.hahnjy@gmail.com>,
"Byungchul Park" <byungchul@sk.com>,
"Ying Huang" <ying.huang@linux.alibaba.com>,
"Baolin Wang" <baolin.wang@linux.alibaba.com>,
"Nico Pache" <nico.pache@linux.dev>,
"Ryan Roberts" <ryan.roberts@arm.com>,
"Dev Jain" <dev.jain@arm.com>, "Barry Song" <baohua@kernel.org>,
"Lance Yang" <lance.yang@linux.dev>,
"Usama Arif" <usama.arif@linux.dev>,
"Kiryl Shutsemau" <kas@kernel.org>,
"Chris Down" <chris@chrisdown.name>,
"Muchun Song" <muchun.song@linux.dev>,
"Oscar Salvador" <osalvador@suse.de>,
"Mikołaj Lenczewski" <miko.lenczewski@arm.com>,
"Michal Hocko" <mhocko@suse.com>,
"Mike Kravetz" <mike.kravetz@oracle.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH 08/12] mm/migrate_device: do not migrate a uffd-protected zero page
Date: Mon, 16 Nov 2026 11:11:33 +1100 [thread overview]
Message-ID: <avpHe8sGgnzx6UwG@nvdebian.thelocal> (raw)
In-Reply-To: <c3a51a9c-083d-4c76-9a62-b7a6fe69a121@columbia.edu>
On 2026-11-14 at 09:26 +1100, Tal Zussman <tz2294@columbia.edu> wrote...
> On 11/13/26 1:04 AM, Gregory Price (Meta) wrote:
> > On 2026-11-11 12:55 -0500, Tal Zussman wrote:
> >> migrate_vma_setup() installs a migration entry for each page it
> >> collects, which carries the uffd bit over to the new PTE. However, a PTE
> >> that maps the zero page is left in place without a migration entry.
> >> migrate_vma_insert_page() later overwrites it with a new PTE that is
> >> built only from the VMA flags, so the uffd-wp bit or the RWP PAGE_NONE
> >> protection of the old PTE is lost. The same applies to the huge zero
> >> PMD.
> >>
> >> Once a driver has migrated such a zero page, writes and accesses to it
> >> no longer fault into userfaultfd, so the manager misses them. With async
> >> tracking, PAGEMAP_SCAN unconditionally reports the page as written or
> >> accessed.
> >>
> >> This requires a driver that migrates memory to a device.
> >>
> >> To fix it, do not mark a zero page for migration if its entry has the
> >> uffd bit set in a VMA registered for WP or RWP. The driver sees an
> >> address that was not migrated, which drivers already have to deal with.
> >>
> >> The uffd bit can also be set after the zero page was collected, as
> >> write-protecting only takes the mmap lock for read. So check again under
> >> the page table lock in migrate_vma_insert_page() and
> >> migrate_vma_insert_huge_pmd_page().
> >>
> >
> > Similar to last patch, this changelog could be improved. My best
> > attempt at summarizing:
> >
> > """
> > Device migration replaces a zero-page mapping with a fresh entry,
> > losing its userfaultfd WP or RWP protection.
> >
> > Don't migrate the uffd-protected zero pages.
> >
> > Skip protected zero pages during collection. Check again under the
> > page-table lock before insertion, because another thread can apply
> > protection after collection finishes.
> >
> > The alternative ...
> > """
> >
>
> Thanks, I'll trim it along these lines.
>
> > I have concerns with this:
> >
> >> The alternative would be to carry the uffd state over in
> >> migrate_vma_insert_page() and migrate_vma_insert_huge_pmd_page(). Not
> >> migrating the zero page is simpler, and a zero page that is protected by
> >> userfaultfd in a range that a driver migrates should be rare.
> >
> > This basically force-fails a migration of a uffd protected zero page.
> >
> > Wouldn't we expect this to cause drivers to fail to make progress if
> > they require that page to be resident on the device? Not sure we want
> > to force drivers to try to handle this special case (if they even can).
> >
> > We'll need some input from the migrate_device/GPU folks.
Thanks, I was halfway through responding last week but Microslop had other
ideas.
Failing the migration is fine - drivers can never *require* a page to be
resident on the device, as migration can always fail for a host of reasons.
In these cases drivers should fallback to remote mapping pages (eg. using
hmm_range_fault()).
Now if userfaultfd+device memory was a commonly used feature then it would
be worth fixing this as always failing would have a large performance impact.
However I've never tested uffd+device memory and have never been confident the
implementation was correct anyway (both features were merged around the same
time). So I think the simple fix is fine:
Reviewed-by: Alistair Popple <apopple@nvidia.com>
> Yeah, makes sense. I can implement the alternative approach, but testing
> for this is a little tricky without the right devices... I might drop this
> patch and send it separately with the GPU folks on Cc.
It probably is worth copying other GPU folk as well though in case they have
different use-cases. Matthew Brost (Intel) has done a lot of work in this area
in DRM, but I notice he is already on CC.
- Alistair
> > --
> > ~Gregory
> >
>
next prev parent reply other threads:[~2026-11-16 0:11 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-11 17:55 [PATCH 00/12] userfaultfd: assorted fixes (part 1) Tal Zussman
2026-11-11 17:55 ` [PATCH 01/12] userfaultfd: use anonymous memory for UFFDIO_ZEROPAGE on private file mappings Tal Zussman
2026-11-12 18:05 ` Lorenzo Stoakes (ARM)
2026-11-11 17:55 ` [PATCH 02/12] userfaultfd: refuse UFFDIO_MOVE into a non-present huge destination PMD Tal Zussman
2026-11-12 17:56 ` sashiko-bot
2026-11-12 18:07 ` Lorenzo Stoakes (ARM)
2026-11-13 4:25 ` Gregory Price (Meta)
2026-11-11 17:55 ` [PATCH 03/12] fs/proc/task_mmu: downgrade writable migration entries in PAGEMAP_SCAN write-protect Tal Zussman
2026-11-12 18:34 ` Lorenzo Stoakes (ARM)
2026-11-13 4:55 ` Gregory Price (Meta)
2026-11-11 17:55 ` [PATCH 04/12] mm/migrate: keep the uffd bit when restoring a writable migration entry Tal Zussman
2026-11-12 18:44 ` Lorenzo Stoakes (ARM)
2026-11-13 21:59 ` Tal Zussman
2026-11-16 14:41 ` Lorenzo Stoakes (ARM)
2026-11-13 4:59 ` Gregory Price (Meta)
2026-11-15 23:57 ` Alistair Popple
2026-11-11 17:55 ` [PATCH 05/12] fs/proc/task_mmu: write-protect device-private THPs in PAGEMAP_SCAN Tal Zussman
2026-11-12 19:34 ` Lorenzo Stoakes (ARM)
2026-11-13 5:08 ` Gregory Price (Meta)
2026-11-16 14:46 ` Lorenzo Stoakes (ARM)
2026-11-13 22:16 ` Tal Zussman
2026-11-16 14:50 ` Lorenzo Stoakes (ARM)
2026-11-13 5:10 ` Gregory Price (Meta)
2026-11-11 17:55 ` [PATCH 06/12] mm/huge_memory: build the moved huge zero PMD from the destination VMA Tal Zussman
2026-11-11 19:12 ` Kiryl Shutsemau
2026-11-12 17:57 ` sashiko-bot
2026-11-13 5:19 ` Gregory Price (Meta)
2026-11-11 17:55 ` [PATCH 07/12] mm/mremap: resolve uffd state when the context is dropped, not during the move Tal Zussman
2026-11-11 19:20 ` Kiryl Shutsemau
2026-11-11 22:39 ` Tal Zussman
2026-11-12 11:44 ` Lorenzo Stoakes (ARM)
2026-11-12 16:00 ` Tal Zussman
2026-11-12 17:57 ` sashiko-bot
2026-11-13 5:46 ` Gregory Price (Meta)
2026-11-11 17:55 ` [PATCH 08/12] mm/migrate_device: do not migrate a uffd-protected zero page Tal Zussman
2026-11-13 6:04 ` Gregory Price (Meta)
2026-11-13 22:26 ` Tal Zussman
2026-11-16 0:11 ` Alistair Popple [this message]
2026-11-11 17:55 ` [PATCH 09/12] mm/madvise: let a guard marker carry the uffd-wp bit Tal Zussman
2026-11-12 11:21 ` Lorenzo Stoakes (ARM)
2026-11-12 16:06 ` Tal Zussman
2026-11-11 17:55 ` [PATCH 10/12] mm: set the uffd-wp bit on a guard marker when write-protecting Tal Zussman
2026-11-12 11:43 ` Lorenzo Stoakes (ARM)
2026-11-11 17:55 ` [PATCH 11/12] mm/madvise: send UFFD_EVENT_REMOVE when installing guard regions Tal Zussman
2026-11-11 19:25 ` Kiryl Shutsemau
2026-11-11 23:21 ` Tal Zussman
2026-11-12 9:40 ` Jose A. Perez de Azpillaga
2026-11-12 11:57 ` Kiryl Shutsemau
2026-11-12 11:03 ` Lorenzo Stoakes (ARM)
2026-11-12 11:06 ` Lorenzo Stoakes (ARM)
2026-11-12 15:56 ` Tal Zussman
2026-11-12 17:57 ` sashiko-bot
2026-11-11 17:55 ` [PATCH 12/12] mm/userfaultfd: check hugetlb end alignment in UFFDIO_UNREGISTER pre-scan Tal Zussman
2026-11-12 19:52 ` Lorenzo Stoakes (ARM)
2026-11-13 22:34 ` Tal Zussman
2026-11-16 14:42 ` Lorenzo Stoakes (ARM)
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=avpHe8sGgnzx6UwG@nvdebian.thelocal \
--to=apopple@nvidia.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=avagin@gmail.com \
--cc=balbirs@nvidia.com \
--cc=baohua@kernel.org \
--cc=baolin.wang@linux.alibaba.com \
--cc=byungchul@sk.com \
--cc=chris@chrisdown.name \
--cc=david@kernel.org \
--cc=dev.jain@arm.com \
--cc=gourry@gourry.net \
--cc=jannh@google.com \
--cc=joshua.hahnjy@gmail.com \
--cc=jthoughton@google.com \
--cc=kas@kernel.org \
--cc=lance.yang@linux.dev \
--cc=liam@infradead.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=ljs@kernel.org \
--cc=matthew.brost@intel.com \
--cc=mhocko@suse.com \
--cc=mike.kravetz@oracle.com \
--cc=miko.lenczewski@arm.com \
--cc=mirq-linux@rere.qmqm.pl \
--cc=muchun.song@linux.dev \
--cc=nico.pache@linux.dev \
--cc=osalvador@suse.de \
--cc=peterx@redhat.com \
--cc=pfalcato@suse.de \
--cc=rppt@kernel.org \
--cc=ryan.roberts@arm.com \
--cc=sfr@canb.auug.org.au \
--cc=surenb@google.com \
--cc=tz2294@columbia.edu \
--cc=usama.anjum@arm.com \
--cc=usama.arif@linux.dev \
--cc=vbabka@kernel.org \
--cc=ying.huang@linux.alibaba.com \
--cc=ziy@nvidia.com \
/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