Re: [PATCH v3] mm: use per_vma lock for MADV_DONTNEED

[Lorenzo Stoakes <lorenzo.stoakes@oracle.com>]

On Sat, Jun 07, 2025 at 08:22:45AM +0100, Lorenzo Stoakes wrote:
> A quick build/selftest run reveals nothing concerning.
>
> On Sat, Jun 07, 2025 at 12:46:23PM +1200, Barry Song wrote:
> > From: Barry Song <v-songbaohua@oppo.com>
> >
> > Certain madvise operations, especially MADV_DONTNEED, occur far more
> > frequently than other madvise options, particularly in native and Java
> > heaps for dynamic memory management.
> >
> > Currently, the mmap_lock is always held during these operations, even when
> > unnecessary. This causes lock contention and can lead to severe priority
> > inversion, where low-priority threads—such as Android's HeapTaskDaemon—
> > hold the lock and block higher-priority threads.
> >
> > This patch enables the use of per-VMA locks when the advised range lies
> > entirely within a single VMA, avoiding the need for full VMA traversal. In
> > practice, userspace heaps rarely issue MADV_DONTNEED across multiple VMAs.
> >
> > Tangquan’s testing shows that over 99.5% of memory reclaimed by Android
> > benefits from this per-VMA lock optimization. After extended runtime,
> > 217,735 madvise calls from HeapTaskDaemon used the per-VMA path, while
> > only 1,231 fell back to mmap_lock.
> >
> > To simplify handling, the implementation falls back to the standard
> > mmap_lock if userfaultfd is enabled on the VMA, avoiding the complexity of
> > userfaultfd_remove().
> >
> > Many thanks to Lorenzo's work[1] on:
> > "Refactor the madvise() code to retain state about the locking mode
> > utilised for traversing VMAs.
> >
> > Then use this mechanism to permit VMA locking to be done later in the
> > madvise() logic and also to allow altering of the locking mode to permit
> > falling back to an mmap read lock if required."
> >
> > One important point, as pointed out by Jann[2], is that
> > untagged_addr_remote() requires holding mmap_lock. This is because
> > address tagging on x86 and RISC-V is quite complex.
> >
> > Until untagged_addr_remote() becomes atomic—which seems unlikely in
> > the near future—we cannot support per-VMA locks for remote processes.
> > So for now, only local processes are supported.
>
> Nice commit message!
>
> >
> > Link: https://lore.kernel.org/all/0b96ce61-a52c-4036-b5b6-5c50783db51f@lucifer.local/ [1]
> > Link: https://lore.kernel.org/all/CAG48ez11zi-1jicHUZtLhyoNPGGVB+ROeAJCUw48bsjk4bbEkA@mail.gmail.com/ [2]
> > Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
> > Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> > Cc: David Hildenbrand <david@redhat.com>
> > Cc: Vlastimil Babka <vbabka@suse.cz>
> > Cc: Jann Horn <jannh@google.com>
> > Cc: Suren Baghdasaryan <surenb@google.com>
> > Cc: Lokesh Gidra <lokeshgidra@google.com>
> > Cc: Tangquan Zheng <zhengtangquan@oppo.com>
> > Cc: Qi Zheng <zhengqi.arch@bytedance.com>
> > Signed-off-by: Barry Song <v-songbaohua@oppo.com>
>
> Since I proposed the approach I am somewhat biased (you should get review from
> others too! :) and I would really like Jann to confirm the untagged addr ting is
> fine, but LGTM other than a thought/suggestion below, so:
>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Whoops, I should add R-b only applies with the fix-patch proposed in [0].

[0]: https://lore.kernel.org/all/CAGsJ_4x1RbQ+GKKc1rrTaNA8Xd+W8K-Zu6-kwVYNKzB0OWiowQ@mail.gmail.com/

Barry - Probably worth respinning a v4 to pick that up to make things
clear, you can propagate my tag with that.

Cheers, Lorenzo

[snip]