Re: [PATCH v9 4/4] mm/madvise: optimize lazyfreeing with mTHP in madvise_free

[Lance Yang]

On Thu, Apr 18, 2024 at 8:44 PM David Hildenbrand <david@xxxxxxxxxx> wrote:
>
> On 18.04.24 14:33, Lance Yang wrote:
> > On Thu, Apr 18, 2024 at 8:03 PM David Hildenbrand <david@redhatcom> wrote:
> >>
> >> On 18.04.24 12:57, Lance Yang wrote:
> >>> This patch optimizes lazyfreeing with PTE-mapped mTHP[1]
> >>> (Inspired by David Hildenbrand[2]). We aim to avoid unnecessary folio
> >>> splitting if the large folio is fully mapped within the target range.
> >>>
> >>> If a large folio is locked or shared, or if we fail to split it, we just
> >>> leave it in place and advance to the next PTE in the range. But note that
> >>> the behavior is changed; previously, any failure of this sort would cause
> >>> the entire operation to give up. As large folios become more common,
> >>> sticking to the old way could result in wasted opportunities.
> >>>
> >>> On an Intel I5 CPU, lazyfreeing a 1GiB VMA backed by PTE-mapped folios of
> >>> the same size results in the following runtimes for madvise(MADV_FREE) in
> >>> seconds (shorter is better):
> >>>
> >>> Folio Size |   Old    |   New    | Change
> >>> ------------------------------------------
> >>>         4KiB | 0.590251 | 0.590259 |    0%
> >>>        16KiB | 2.990447 | 0.185655 |  -94%
> >>>        32KiB | 2.547831 | 0.104870 |  -95%
> >>>        64KiB | 2.457796 | 0.052812 |  -97%
> >>>       128KiB | 2.281034 | 0.032777 |  -99%
> >>>       256KiB | 2.230387 | 0.017496 |  -99%
> >>>       512KiB | 2.189106 | 0.010781 |  -99%
> >>>      1024KiB | 2.183949 | 0.007753 |  -99%
> >>>      2048KiB | 0.002799 | 0.002804 |    0%
> >>>
> >>> [1] https://lkml.kernel.org/r/20231207161211.2374093-5-ryan.roberts@xxxxxxx
> >>> [2] https://lore.kernel.org/linux-mm/20240214204435.167852-1-david@xxxxxxxxxx
> >>>
> >>> Reviewed-by: Ryan Roberts <ryan.roberts@xxxxxxx>
> >>> Signed-off-by: Lance Yang <ioworker0@xxxxxxxxx>
> >>> ---
> >>>    mm/madvise.c | 85 +++++++++++++++++++++++++++-------------------------
> >>>    1 file changed, 44 insertions(+), 41 deletions(-)
> >>>
> >>> diff --git a/mm/madvise.c b/mm/madvise.c
> >>> index 4597a3568e7e..375ab3234603 100644
> >>> --- a/mm/madvise.c
> >>> +++ b/mm/madvise.c
> >>> @@ -643,6 +643,7 @@ static int madvise_free_pte_range(pmd_t *pmd, unsigned long addr,
> >>>                                unsigned long end, struct mm_walk *walk)
> >>>
> >>>    {
> >>> +     const cydp_t cydp_flags = CYDP_CLEAR_YOUNG | CYDP_CLEAR_DIRTY;
> >>>        struct mmu_gather *tlb = walk->private;
> >>>        struct mm_struct *mm = tlb->mm;
> >>>        struct vm_area_struct *vma = walk->vma;
> >>> @@ -697,44 +698,57 @@ static int madvise_free_pte_range(pmd_t *pmd, unsigned long addr,
> >>>                        continue;
> >>>
> >>>                /*
> >>> -              * If pmd isn't transhuge but the folio is large and
> >>> -              * is owned by only this process, split it and
> >>> -              * deactivate all pages.
> >>> +              * If we encounter a large folio, only split it if it is not
> >>> +              * fully mapped within the range we are operating on. Otherwise
> >>> +              * leave it as is so that it can be marked as lazyfree. If we
> >>> +              * fail to split a folio, leave it in place and advance to the
> >>> +              * next pte in the range.
> >>>                 */
> >>>                if (folio_test_large(folio)) {
> >>> -                     int err;
> >>> +                     bool any_young, any_dirty;
> >>>
> >>> -                     if (folio_likely_mapped_shared(folio))
> >>> -                             break;
> >>> -                     if (!folio_trylock(folio))
> >>> -                             break;
> >>> -                     folio_get(folio);
> >>> -                     arch_leave_lazy_mmu_mode();
> >>> -                     pte_unmap_unlock(start_pte, ptl);
> >>> -                     start_pte = NULL;
> >>> -                     err = split_folio(folio);
> >>> -                     folio_unlock(folio);
> >>> -                     folio_put(folio);
> >>> -                     if (err)
> >>> -                             break;
> >>> -                     start_pte = pte =
> >>> -                             pte_offset_map_lock(mm, pmd, addr, &ptl);
> >>> -                     if (!start_pte)
> >>> -                             break;
> >>> -                     arch_enter_lazy_mmu_mode();
> >>> -                     pte--;
> >>> -                     addr -= PAGE_SIZE;
> >>> -                     continue;
> >>> +                     nr = madvise_folio_pte_batch(addr, end, folio, pte,
> >>> +                                                  ptent, &any_young, NULL);
> >>> +
> >>> +                     if (nr < folio_nr_pages(folio)) {
> >>> +                             int err;
> >>> +
> >>> +                             if (folio_likely_mapped_shared(folio))
> >>> +                                     continue;
> >>> +                             if (!folio_trylock(folio))
> >>> +                                     continue;
> >>> +                             folio_get(folio);
> >>> +                             arch_leave_lazy_mmu_mode();
> >>> +                             pte_unmap_unlock(start_pte, ptl);
> >>> +                             start_pte = NULL;
> >>> +                             err = split_folio(folio);
> >>> +                             folio_unlock(folio);
> >>> +                             folio_put(folio);
> >>> +                             start_pte = pte =
> >>> +                                     pte_offset_map_lock(mm, pmd, addr, &ptl);
> >>
> >> I'd just put it on a single line.
> >
> > start_pte = pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
> >
> > I suddenly realized that putting it on a single line would exceed the
> > 80-char limit.
>
> Which is fine according to Documentation/process/coding-style.rst
>
> ... as long as it aids readability.
>
> Alternatively, the following might do:
>
> pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
> start_pte = pte;

Yep, I understood.

Thanks,
Lance

>
> --
> Cheers,
>
> David / dhildenb
>