On Sun, 19 Apr 2026 12:57:42 -0600 Nico Pache <[email protected]> wrote:

> Pass an order and offset to collapse_huge_page to support collapsing anon
> memory to arbitrary orders within a PMD. order indicates what mTHP size we
> are attempting to collapse to, and offset indicates were in the PMD to
> start the collapse attempt.
> 
> For non-PMD collapse we must leave the anon VMA write locked until after
> we collapse the mTHP-- in the PMD case all the pages are isolated, but in
> the mTHP case this is not true, and we must keep the lock to prevent
> access/changes to the page tables. This can happen if the rmap walkers hit
> a pmd_none while the PMD entry is currently unavailable due to being
> temporarily removed during the collapse phase.
> 
> Signed-off-by: Nico Pache <[email protected]>
> ---
>  mm/khugepaged.c | 103 +++++++++++++++++++++++++++---------------------
>  1 file changed, 57 insertions(+), 46 deletions(-)
> 
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index 283bb63854a5..ff6f9f1883ed 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -1198,42 +1198,36 @@ static enum scan_result alloc_charge_folio(struct 
> folio **foliop, struct mm_stru
>       return SCAN_SUCCEED;
>  }
>  
> -static enum scan_result collapse_huge_page(struct mm_struct *mm, unsigned 
> long address,
> -             int referenced, int unmapped, struct collapse_control *cc)
> +static enum scan_result collapse_huge_page(struct mm_struct *mm, unsigned 
> long start_addr,
> +             int referenced, int unmapped, struct collapse_control *cc,
> +             unsigned int order)
>  {
>       LIST_HEAD(compound_pagelist);
>       pmd_t *pmd, _pmd;
> -     pte_t *pte;
> +     pte_t *pte = NULL;
>       pgtable_t pgtable;
>       struct folio *folio;
>       spinlock_t *pmd_ptl, *pte_ptl;
>       enum scan_result result = SCAN_FAIL;
>       struct vm_area_struct *vma;
>       struct mmu_notifier_range range;
> +     bool anon_vma_locked = false;
> +     const unsigned long pmd_addr = start_addr & HPAGE_PMD_MASK;
> +     const unsigned long end_addr = start_addr + (PAGE_SIZE << order);
>  
> -     VM_BUG_ON(address & ~HPAGE_PMD_MASK);
> -
> -     /*
> -      * Before allocating the hugepage, release the mmap_lock read lock.
> -      * The allocation can take potentially a long time if it involves
> -      * sync compaction, and we do not need to hold the mmap_lock during
> -      * that. We will recheck the vma after taking it again in write mode.
> -      */
> -     mmap_read_unlock(mm);
> -

My understanding now is that the caller will need to drop the mmap_read lock?

This needs explicit documentation at the start of the function IMO. I think
there are callers in later patches and its very easy to miss this.


> -     result = alloc_charge_folio(&folio, mm, cc, HPAGE_PMD_ORDER);
> +     result = alloc_charge_folio(&folio, mm, cc, order);
>       if (result != SCAN_SUCCEED)
>               goto out_nolock;
>  
>       mmap_read_lock(mm);
> -     result = hugepage_vma_revalidate(mm, address, true, &vma, cc,
> -                                      HPAGE_PMD_ORDER);
> +     result = hugepage_vma_revalidate(mm, pmd_addr, /*expect_anon=*/ true,
> +                                      &vma, cc, order);
>       if (result != SCAN_SUCCEED) {
>               mmap_read_unlock(mm);
>               goto out_nolock;
>       }
>  
> -     result = find_pmd_or_thp_or_none(mm, address, &pmd);
> +     result = find_pmd_or_thp_or_none(mm, pmd_addr, &pmd);
>       if (result != SCAN_SUCCEED) {
>               mmap_read_unlock(mm);
>               goto out_nolock;
> @@ -1245,8 +1239,8 @@ static enum scan_result collapse_huge_page(struct 
> mm_struct *mm, unsigned long a
>                * released when it fails. So we jump out_nolock directly in
>                * that case.  Continuing to collapse causes inconsistency.
>                */
> -             result = __collapse_huge_page_swapin(mm, vma, address, pmd,
> -                                                  referenced, 
> HPAGE_PMD_ORDER);
> +             result = __collapse_huge_page_swapin(mm, vma, start_addr, pmd,
> +                                                  referenced, order);
>               if (result != SCAN_SUCCEED)
>                       goto out_nolock;
>       }
> @@ -1261,20 +1255,21 @@ static enum scan_result collapse_huge_page(struct 
> mm_struct *mm, unsigned long a
>        * mmap_lock.
>        */
>       mmap_write_lock(mm);
> -     result = hugepage_vma_revalidate(mm, address, true, &vma, cc,
> -                                      HPAGE_PMD_ORDER);
> +     result = hugepage_vma_revalidate(mm, pmd_addr, /*expect_anon=*/ true,
> +                                      &vma, cc, order);
>       if (result != SCAN_SUCCEED)
>               goto out_up_write;
>       /* check if the pmd is still valid */
>       vma_start_write(vma);
> -     result = check_pmd_still_valid(mm, address, pmd);
> +     result = check_pmd_still_valid(mm, pmd_addr, pmd);
>       if (result != SCAN_SUCCEED)
>               goto out_up_write;
>  
>       anon_vma_lock_write(vma->anon_vma);
> +     anon_vma_locked = true;
>  
> -     mmu_notifier_range_init(&range, MMU_NOTIFY_CLEAR, 0, mm, address,
> -                             address + HPAGE_PMD_SIZE);
> +     mmu_notifier_range_init(&range, MMU_NOTIFY_CLEAR, 0, mm, start_addr,
> +                             end_addr);
>       mmu_notifier_invalidate_range_start(&range);
>  
>       pmd_ptl = pmd_lock(mm, pmd); /* probably unnecessary */
> @@ -1286,26 +1281,23 @@ static enum scan_result collapse_huge_page(struct 
> mm_struct *mm, unsigned long a
>        * Parallel GUP-fast is fine since GUP-fast will back off when
>        * it detects PMD is changed.
>        */
> -     _pmd = pmdp_collapse_flush(vma, address, pmd);
> +     _pmd = pmdp_collapse_flush(vma, pmd_addr, pmd);


For an mTHP collapse covering, say, 64KiB of a 2MiB PMD, the patch still
flushes the entire PMD via pmdp_collapse_flush and tlb_remove_table_sync_one.
That triggers cross-CPU TLB shootdowns for ~1.94MiB of unrelated mappings on
every successful sub-PMD collapse. Probably acceptable as a first cut?


>       spin_unlock(pmd_ptl);
>       mmu_notifier_invalidate_range_end(&range);
>       tlb_remove_table_sync_one();
>  
> -     pte = pte_offset_map_lock(mm, &_pmd, address, &pte_ptl);
> +     pte = pte_offset_map_lock(mm, &_pmd, start_addr, &pte_ptl);
>       if (pte) {
> -             result = __collapse_huge_page_isolate(vma, address, pte, cc,
> -                                                   HPAGE_PMD_ORDER,
> -                                                   &compound_pagelist);
> +             result = __collapse_huge_page_isolate(vma, start_addr, pte, cc,
> +                                                   order, 
> &compound_pagelist);
>               spin_unlock(pte_ptl);
>       } else {
>               result = SCAN_NO_PTE_TABLE;
>       }
>  
>       if (unlikely(result != SCAN_SUCCEED)) {
> -             if (pte)
> -                     pte_unmap(pte);
>               spin_lock(pmd_ptl);
> -             BUG_ON(!pmd_none(*pmd));
> +             WARN_ON_ONCE(!pmd_none(*pmd));

Why was this turned into WARN_ON_ONCE? Would be good to add to commit message
what the reason is if it has been discussed earlier.

The next line writes a PMD entry over an existing one — that leaks the
previous page table or PMD-mapped folio and can corrupt VA mappings. BUG_ON
failed loudly and safely; WARN_ON_ONCE continues into the corruption and falls
silent after the first hit.


>               /*
>                * We can only use set_pmd_at when establishing
>                * hugepmds and never for establishing regular pmds that
> @@ -1313,21 +1305,24 @@ static enum scan_result collapse_huge_page(struct 
> mm_struct *mm, unsigned long a
>                */
>               pmd_populate(mm, pmd, pmd_pgtable(_pmd));
>               spin_unlock(pmd_ptl);
> -             anon_vma_unlock_write(vma->anon_vma);
>               goto out_up_write;
>       }
>  
>       /*
> -      * All pages are isolated and locked so anon_vma rmap
> -      * can't run anymore.
> +      * For PMD collapse all pages are isolated and locked so anon_vma
> +      * rmap can't run anymore. For mTHP collapse the PMD entry has been
> +      * removed and not all pages are isolated and locked, so we must hold
> +      * the lock to prevent neighboring folios from attempting to access
> +      * this PMD until its reinstalled.
>        */
> -     anon_vma_unlock_write(vma->anon_vma);
> +     if (is_pmd_order(order)) {
> +             anon_vma_unlock_write(vma->anon_vma);
> +             anon_vma_locked = false;
> +     }
>  
>       result = __collapse_huge_page_copy(pte, folio, pmd, _pmd,
> -                                        vma, address, pte_ptl,
> -                                        HPAGE_PMD_ORDER,
> -                                        &compound_pagelist);
> -     pte_unmap(pte);
> +                                        vma, start_addr, pte_ptl,
> +                                        order, &compound_pagelist);
>       if (unlikely(result != SCAN_SUCCEED))
>               goto out_up_write;
>  
> @@ -1337,18 +1332,27 @@ static enum scan_result collapse_huge_page(struct 
> mm_struct *mm, unsigned long a
>        * write.
>        */
>       __folio_mark_uptodate(folio);
> -     pgtable = pmd_pgtable(_pmd);
> -
>       spin_lock(pmd_ptl);
> -     BUG_ON(!pmd_none(*pmd));
> -     pgtable_trans_huge_deposit(mm, pmd, pgtable);
> -     map_anon_folio_pmd_nopf(folio, pmd, vma, address);
> +     WARN_ON_ONCE(!pmd_none(*pmd));
> +     if (is_pmd_order(order)) { /* PMD collapse */
> +             pgtable = pmd_pgtable(_pmd);
> +             pgtable_trans_huge_deposit(mm, pmd, pgtable);
> +             map_anon_folio_pmd_nopf(folio, pmd, vma, pmd_addr);
> +     } else { /* mTHP collapse */
> +             map_anon_folio_pte_nopf(folio, pte, vma, start_addr, 
> /*uffd_wp=*/ false);

map_anon_folio_pte_nopf calls set_ptes and modifies pagetable, while holding
pmd_ptl only. It should be safe as we expect pmd_none. But I think you should
put a comment about this?


> +             smp_wmb(); /* make PTEs visible before PMD. See pmd_install() */
> +             pmd_populate(mm, pmd, pmd_pgtable(_pmd));
> +     }
>       spin_unlock(pmd_ptl);
>  
>       folio = NULL;
>  
>       result = SCAN_SUCCEED;
>  out_up_write:
> +     if (anon_vma_locked)
> +             anon_vma_unlock_write(vma->anon_vma);
> +     if (pte)
> +             pte_unmap(pte);
>       mmap_write_unlock(mm);
>  out_nolock:
>       if (folio)
> @@ -1525,8 +1529,15 @@ static enum scan_result collapse_scan_pmd(struct 
> mm_struct *mm,
>  out_unmap:
>       pte_unmap_unlock(pte, ptl);
>       if (result == SCAN_SUCCEED) {
> +             /*
> +              * Before allocating the hugepage, release the mmap_lock read 
> lock.
> +              * The allocation can take potentially a long time if it 
> involves
> +              * sync compaction, and we do not need to hold the mmap_lock 
> during
> +              * that. We will recheck the vma after taking it again in write 
> mode.
> +              */
> +             mmap_read_unlock(mm);
>               result = collapse_huge_page(mm, start_addr, referenced,
> -                                         unmapped, cc);
> +                                         unmapped, cc, HPAGE_PMD_ORDER);
>               /* collapse_huge_page will return with the mmap_lock released */
>               *lock_dropped = true;
>       }
> -- 
> 2.53.0
> 
> 

Reply via email to