From: Andrew Morton <akpm@linux-foundation.org>
To: mpenttil@redhat.com
Cc: linux-mm@kvack.org, dri-devel@lists.freedesktop.org,
intel-xe@lists.freedesktop.org, linux-kernel@vger.kernel.org,
David Hildenbrand <david@kernel.org>,
Jason Gunthorpe <jgg@nvidia.com>,
Leon Romanovsky <leonro@nvidia.com>,
Alistair Popple <apopple@nvidia.com>,
Balbir Singh <balbirs@nvidia.com>, Zi Yan <ziy@nvidia.com>,
Matthew Brost <matthew.brost@intel.com>,
Lorenzo Stoakes <ljs@kernel.org>,
"Liam R. Howlett" <Liam.Howlett@oracle.com>,
Vlastimil Babka <vbabka@suse.cz>, Mike Rapoport <rppt@kernel.org>,
Suren Baghdasaryan <surenb@google.com>,
Michal Hocko <mhocko@suse.com>
Subject: Re: [PATCH 00/12] [PATCH v14 00/12] migrate on fault for device pages
Date: Tue, 22 Sep 2026 19:27:26 -0700 [thread overview]
Message-ID: <20260922192726.90dec3c4c77b8c9063dbf218@linux-foundation.org> (raw)
In-Reply-To: <20260922053421.4092027-1-mpenttil@redhat.com>
On Tue, 22 Sep 2026 08:34:09 +0300 mpenttil@redhat.com wrote:
> From: Mika Penttilä <mpenttil@redhat.com>
>
> Currently, the way device page faulting and migration works
> is not optimal, if you want to do both fault handling and
> migration at once.
>
> Being able to migrate not present pages (or pages mapped with incorrect
> permissions, eg. COW) to the GPU requires doing either of the
> following sequences:
>
> 1. hmm_range_fault() - fault in non-present pages with correct permissions, etc.
> 2. migrate_vma_*() - migrate the pages
>
> Or:
>
> 1. migrate_vma_*() - migrate present pages
> 2. If non-present pages detected by migrate_vma_*():
> a) call hmm_range_fault() to fault pages in
> b) call migrate_vma_*() again to migrate now present pages
>
> The problem with the first sequence is that you always have to do two
> page walks even when most of the time the pages are present or zero page
> mappings so the common case takes a performance hit.
>
> The second sequence is better for the common case, but far worse if
> pages aren't present because now you have to walk the page tables three
> times (once to find the page is not present, once so hmm_range_fault()
> can find a non-present page to fault in and once again to setup the
> migration). It is also tricky to code correctly. One page table walk
> could costs over 1000 cpu cycles on X86-64, which is a significant hit.
>
> We should be able to walk the page table once, faulting
> pages in as required and replacing them with migration entries if
> requested.
Sounds sensible.
> Tested in X86-64 VM with HMM test device, passing the selftests.
> For performance, the migrate throughput tests from the selftests
> show similar numbers (within error margin) as unmodified kernel.
But no performance benefits are demonstrated?
next prev parent reply other threads:[~2026-09-23 2:27 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-22 5:34 mpenttil
2026-09-22 5:34 ` [PATCH v14 01/12] mm/Kconfig: changes for " mpenttil
2026-09-22 22:27 ` Balbir Singh
2026-09-23 5:42 ` Mika Penttilä
2026-09-22 5:34 ` [PATCH v14 02/12] mm: add helper to convert HMM pfn to migrate pfn mpenttil
2026-09-22 5:34 ` [PATCH v14 03/12] mm/hmm: preparations for HMM to participate in migration mpenttil
2026-09-22 5:34 ` [PATCH v14 04/12] mm/hmm: do the plumbing " mpenttil
2026-09-22 5:34 ` [PATCH v14 05/12] mm/hmm: implement folio split for migrate needs in HMM pagewalk mpenttil
2026-09-22 5:34 ` [PATCH v14 06/12] mm/hmm: migrate collection in HMM pagewalk - pte level mpenttil
2026-09-22 5:34 ` [PATCH v14 07/12] mm/hmm: migrate collection in HMM pagewalk - pmd level mpenttil
2026-09-22 5:34 ` [PATCH v14 08/12] mm/hmm: add lazy MMU mode support for migration in HMM pagewalk mpenttil
2026-09-22 5:34 ` [PATCH v14 09/12] mm/hmm: implement rollback for device page " mpenttil
2026-09-22 5:34 ` [PATCH v14 10/12] mm: enable device page migration from " mpenttil
2026-09-22 5:34 ` [PATCH v14 11/12] lib/test_hmm: add a new testcase for the migrate on fault mpenttil
2026-09-22 5:34 ` [PATCH v14 12/12] Documentation/mm/hmm: document migration through hmm_range_fault() mpenttil
2026-09-23 2:27 ` Andrew Morton [this message]
2026-09-23 5:29 ` [PATCH 00/12] [PATCH v14 00/12] migrate on fault for device pages Mika Penttilä
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=20260922192726.90dec3c4c77b8c9063dbf218@linux-foundation.org \
--to=akpm@linux-foundation.org \
--cc=Liam.Howlett@oracle.com \
--cc=apopple@nvidia.com \
--cc=balbirs@nvidia.com \
--cc=david@kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=intel-xe@lists.freedesktop.org \
--cc=jgg@nvidia.com \
--cc=leonro@nvidia.com \
--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=mpenttil@redhat.com \
--cc=rppt@kernel.org \
--cc=surenb@google.com \
--cc=vbabka@suse.cz \
--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
all inboxes | Powered by JetHome®