Re: [PATCH mm-new v8 4/4] mm: khugepaged: skip lazy-free folios

[Barry Song]

On Sat, Feb 21, 2026 at 5:40 PM Vernon Yang <vernon2gm@xxxxxxxxx> wrote:
>
> From: Vernon Yang <yanglincheng@xxxxxxxxxx>
>
> For example, create three task: hot1 -> cold -> hot2. After all three
> task are created, each allocate memory 128MB. the hot1/hot2 task
> continuously access 128 MB memory, while the cold task only accesses
> its memory briefly and then call madvise(MADV_FREE). However, khugepaged
> still prioritizes scanning the cold task and only scans the hot2 task
> after completing the scan of the cold task.
>
> And if all folios in VM_DROPPABLE are lazyfree, Collapsing maintains
> that property, so we can just collapse and memory pressure in the future

I don’t think this is accurate. A VMA without VM_DROPPABLE
can still have all folios marked as lazyfree. Therefore, having
all folios lazyfree is not the reason why collapsing preserves
the property.

This raises a question: if a VMA without VM_DROPPABLE has
many contiguous lazyfree folios that can be collapsed, and
none of those folios are non-lazyfree, should we collapse
them and pass the lazyfree state to the new folio?

Currently, our approach skips the collapse, which also feels
a bit inconsistent.

> will free it up. In contrast, collapsing in !VM_DROPPABLE does not
> maintain that property, the collapsed folio will not be lazyfree and
> memory pressure in the future will not be able to free it up.
>
> So if the user has explicitly informed us via MADV_FREE that this memory
> will be freed, and this vma does not have VM_DROPPABLE flags, it is
> appropriate for khugepaged to skip it only, thereby avoiding unnecessary
> scan and collapse operations to reducing CPU wastage.
>
> Here are the performance test results:
> (Throughput bigger is better, other smaller is better)
>
> Testing on x86_64 machine:
>
> | task hot2           | without patch | with patch    |  delta  |
> |---------------------|---------------|---------------|---------|
> | total accesses time |  3.14 sec     |  2.93 sec     | -6.69%  |
> | cycles per access   |  4.96         |  2.21         | -55.44% |
> | Throughput          |  104.38 M/sec |  111.89 M/sec | +7.19%  |
> | dTLB-load-misses    |  284814532    |  69597236     | -75.56% |
>
> Testing on qemu-system-x86_64 -enable-kvm:
>
> | task hot2           | without patch | with patch    |  delta  |
> |---------------------|---------------|---------------|---------|
> | total accesses time |  3.35 sec     |  2.96 sec     | -11.64% |
> | cycles per access   |  7.29         |  2.07         | -71.60% |
> | Throughput          |  97.67 M/sec  |  110.77 M/sec | +13.41% |
> | dTLB-load-misses    |  241600871    |  3216108      | -98.67% |
>
> Signed-off-by: Vernon Yang <yanglincheng@xxxxxxxxxx>
> Acked-by: David Hildenbrand (arm) <david@xxxxxxxxxx>
> Reviewed-by: Lance Yang <lance.yang@xxxxxxxxx>
> ---

Overall, LGTM,

Reviewed-by: Barry Song <baohua@xxxxxxxxxx>

>  include/trace/events/huge_memory.h |  1 +
>  mm/khugepaged.c                    | 13 +++++++++++++
>  2 files changed, 14 insertions(+)
>
> diff --git a/include/trace/events/huge_memory.h b/include/trace/events/huge_memory.h
> index 384e29f6bef0..bcdc57eea270 100644
> --- a/include/trace/events/huge_memory.h
> +++ b/include/trace/events/huge_memory.h
> @@ -25,6 +25,7 @@
>         EM( SCAN_PAGE_LRU,              "page_not_in_lru")              \
>         EM( SCAN_PAGE_LOCK,             "page_locked")                  \
>         EM( SCAN_PAGE_ANON,             "page_not_anon")                \
> +       EM( SCAN_PAGE_LAZYFREE,         "page_lazyfree")                \
>         EM( SCAN_PAGE_COMPOUND,         "page_compound")                \
>         EM( SCAN_ANY_PROCESS,           "no_process_for_page")          \
>         EM( SCAN_VMA_NULL,              "vma_null")                     \
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index 61e25cf5424b..e792e9074b48 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -46,6 +46,7 @@ enum scan_result {
>         SCAN_PAGE_LRU,
>         SCAN_PAGE_LOCK,
>         SCAN_PAGE_ANON,
> +       SCAN_PAGE_LAZYFREE,
>         SCAN_PAGE_COMPOUND,
>         SCAN_ANY_PROCESS,
>         SCAN_VMA_NULL,
> @@ -574,6 +575,12 @@ static enum scan_result __collapse_huge_page_isolate(struct vm_area_struct *vma,
>                 folio = page_folio(page);
>                 VM_BUG_ON_FOLIO(!folio_test_anon(folio), folio);
>
> +               if (cc->is_khugepaged && !(vma->vm_flags & VM_DROPPABLE) &&
> +                   folio_test_lazyfree(folio) && !pte_dirty(pteval)) {

I would prefer to add a comment about VM_DROPPABLE here
rather than only mentioning it in the changelog.

> +                       result = SCAN_PAGE_LAZYFREE;
> +                       goto out;
> +               }
> +
>                 /* See hpage_collapse_scan_pmd(). */
>                 if (folio_maybe_mapped_shared(folio)) {
>                         ++shared;
> @@ -1326,6 +1333,12 @@ static enum scan_result hpage_collapse_scan_pmd(struct mm_struct *mm,
>                 }
>                 folio = page_folio(page);
>
> +               if (cc->is_khugepaged && !(vma->vm_flags & VM_DROPPABLE) &&
> +                   folio_test_lazyfree(folio) && !pte_dirty(pteval)) {
> +                       result = SCAN_PAGE_LAZYFREE;
> +                       goto out_unmap;
> +               }

As above.

> +
>                 if (!folio_test_anon(folio)) {
>                         result = SCAN_PAGE_ANON;
>                         goto out_unmap;
> --
> 2.51.0
>

Thanks
Barry