Re: [PATCH v2] mm/shmem: avoid zeroing new folios twice with init_on_alloc=1

From: David Hildenbrand (Arm)

Date: Thu Nov 19 2026 - 12:06:32 EST


On 11/18/26 21:28, Samridh via B4 Relay wrote:
> From: Samridh <samridh@xxxxxxxxx>
>
> shmem_get_folio_gfp() clears a new folio before marking it uptodate.
> With init_on_alloc=1 the page allocator has already zeroed it, so the
> folio ends up being zeroed twice.
>
> Skip the clear for a folio that was just allocated when
> user_alloc_needs_zeroing() says it is not needed, like anonymous memory
> does. Folios that are already in the page cache but not uptodate, such
> as those left by fallocate(), are still cleared, and so is every folio
> on CPUs with cache aliasing.
>
> Time to write one byte to each page of a new 4 GiB memfd mapping, with
> shmem_enabled=always and init_on_alloc=1 (median of 24 runs on 6.18.55,
> x86_64 VMs):
>
> before after
> Xeon 6975P-C 0.625s 0.420s
> EPYC 9R14 0.216s 0.160s
>
> With 4K folios the times change by 3% or less, and init_on_alloc=0 is
> not affected. We found this while running gVisor, which keeps sandbox
> memory in a memfd.
>
> I wrote this patch with help from an AI coding assistant (Claude Code),
> then reviewed and tested it.
>
> Signed-off-by: Samridh <samridh@xxxxxxxxx>
> ---
> Tested on 6.18.55 with this change backported, on Intel and AMD x86_64
> VMs, and build-tested on mm/core for-next. A simple test fills memory
> with data, frees it, and then checks that new memfd pages read as zero,
> with MADV_HUGEPAGE and MADV_NOHUGEPAGE on the fault and fallocate paths.
> It passes on both machines with init_on_alloc=1.
> ---
> Changes in v2:
> - Use user_alloc_needs_zeroing() instead of __GFP_ZERO (Baolin).
> - Link to v1: https://patch.msgid.link/20261117-samridh-shmem-send-v1-1-1b897ff562d9@xxxxxxxxx
> ---
> mm/shmem.c | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/mm/shmem.c b/mm/shmem.c
> index c769d0bc7df9..1a78d15a768e 100644
> --- a/mm/shmem.c
> +++ b/mm/shmem.c
> @@ -2810,8 +2810,11 @@ static int shmem_get_folio_gfp(struct inode *inode, pgoff_t index,
> if (sgp != SGP_WRITE && !folio_test_uptodate(folio)) {
> long i, n = folio_nr_pages(folio);
>
> - for (i = 0; i < n; i++)
> - clear_highpage(folio_page(folio, i));
> + /* init_on_alloc already zeroed a new folio. */
> + if (!alloced || user_alloc_needs_zeroing()) {
> + for (i = 0; i < n; i++)
> + clear_highpage(folio_page(folio, i));
> + }
> flush_dcache_folio(folio);

I do wonder whether that could go into the loop, because I faintly remember that
user_alloc_needs_zeroing() is disabled when CPUs have dcache aliasing.

In general LGTM, but I'll wait for more feedback from shmem maintainers.

--
Cheers,

David