Re: [PATCH 08/13] mm: swap in PMD swap entries as whole THPs during swapoff
From: Usama Arif
Date: Fri May 29 2026 - 11:09:17 EST
On 26/05/2026 20:44, Alexandre Ghiti wrote:
> Hi Usama,
>
> On 4/27/26 12:01, Usama Arif wrote:
>> Add unuse_pmd() and call it from unuse_pmd_range() to swap in
>> PMD-level swap entries as whole THPs during swapoff. This mirrors
>> the existing unuse_pte_range() but operates at PMD granularity.
>>
>> If the PMD-order folio cannot be allocated, the cached folio is no
>> longer PMD-sized (e.g. split in the swap cache by
>> deferred_split_scan() or memory_failure() while the PMD swap entry
>> was installed), or the folio is not uptodate, the PMD swap entry is
>> split into PTE-level entries via __split_huge_pmd() and a non-zero
>> error is returned so unuse_pmd_range() falls through to
>> unuse_pte_range(), which handles the individual entries at order-0.
>>
>> swapin_alloc_pmd_folio() is a separate function in swap_state.c
>> as it will be reused in swapin in a later patch.
>>
>> Signed-off-by: Usama Arif <usama.arif@xxxxxxxxx>
>> ---
>> mm/swap.h | 7 +++
>> mm/swap_state.c | 35 +++++++++++++
>> mm/swapfile.c | 137 ++++++++++++++++++++++++++++++++++++++++++++++++
>> 3 files changed, 179 insertions(+)
>>
>> diff --git a/mm/swap.h b/mm/swap.h
>> index a77016f2423b..76752df71693 100644
>> --- a/mm/swap.h
>> +++ b/mm/swap.h
>> @@ -301,6 +301,7 @@ struct folio *swap_cluster_readahead(swp_entry_t entry, gfp_t flag,
>> struct folio *swapin_readahead(swp_entry_t entry, gfp_t flag,
>> struct vm_fault *vmf);
>> struct folio *swapin_folio(swp_entry_t entry, struct folio *folio);
>> +struct folio *swapin_alloc_pmd_folio(swp_entry_t entry, struct mm_struct *mm);
>> void swap_update_readahead(struct folio *folio, struct vm_area_struct *vma,
>> unsigned long addr);
>> @@ -438,6 +439,12 @@ static inline struct folio *swapin_folio(swp_entry_t entry, struct folio *folio)
>> return NULL;
>> }
>> +static inline struct folio *swapin_alloc_pmd_folio(swp_entry_t entry,
>> + struct mm_struct *mm)
>> +{
>> + return NULL;
>> +}
>> +
>> static inline void swap_update_readahead(struct folio *folio,
>> struct vm_area_struct *vma, unsigned long addr)
>> {
>> diff --git a/mm/swap_state.c b/mm/swap_state.c
>> index 1415a5c54a43..c2e8c76658f5 100644
>> --- a/mm/swap_state.c
>> +++ b/mm/swap_state.c
>> @@ -584,6 +584,41 @@ struct folio *swapin_folio(swp_entry_t entry, struct folio *folio)
>> return swapcache;
>> }
>> +#ifdef CONFIG_THP_SWAP
>> +/**
>> + * swapin_alloc_pmd_folio - allocate, charge, and read a PMD-sized swap folio.
>> + * @entry: starting swap entry to swap in
>> + * @mm: mm to charge for the swap-in
>> + *
>> + * Allocate a HPAGE_PMD_ORDER folio, charge it to @mm's memcg for @entry, and
>> + * issue the swap-in via swapin_folio(). Used by callers that need to map a
>> + * PMD swap entry as a whole THP (PMD swapoff).
>> + *
>> + * Return: the swapped-in folio, or NULL on alloc/charge/swapin failure (in
>> + * which case the caller should fall back to splitting the PMD).
>> + */
>> +struct folio *swapin_alloc_pmd_folio(swp_entry_t entry, struct mm_struct *mm)
>> +{
>> + struct folio *folio;
>> +
>> + folio = folio_alloc(GFP_HIGHUSER_MOVABLE, HPAGE_PMD_ORDER);
>
>
> Don't you think we should back off early by passing __GFP_NORETRY? No need to keep trying under memory pressure no?
Yes, I think that makes sense, either that or GFP_TRANSHUGE_LIGHT.
I will rebase on latest, and probably use Kairuis function for allocating folios.
Thanks!
>
>
>> + if (!folio)
>> + return NULL;
>> +
>> + if (mem_cgroup_swapin_charge_folio(folio, mm, GFP_KERNEL, entry)) {
>> + folio_put(folio);
>> + return NULL;
>> + }
>> +
>> + if (!swapin_folio(entry, folio)) {
>> + folio_put(folio);
>> + return NULL;
>> + }
>> +
>> + return folio;
>> +}
>> +#endif /* CONFIG_THP_SWAP */
>> +
>> /*
>> * Locate a page of swap in physical memory, reserving swap cache space
>> * and reading the disk if it is not already cached.
>> diff --git a/mm/swapfile.c b/mm/swapfile.c
>> index 390f191be9a6..7256edf4ce66 100644
>> --- a/mm/swapfile.c
>> +++ b/mm/swapfile.c
>> @@ -42,6 +42,7 @@
>> #include <linux/suspend.h>
>> #include <linux/zswap.h>
>> #include <linux/plist.h>
>> +#include <linux/huge_mm.h>
>> #include <asm/tlbflush.h>
>> #include <linux/leafops.h>
>> @@ -2519,6 +2520,130 @@ static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
>> return 0;
>> }
>> +/*
>> + * unuse_pmd - Map a locked folio at PMD granularity during swapoff.
>> + *
>> + * The caller provides a locked, swapped-in folio. Returns 0 on success
>> + * (PMD was mapped). Returns -EAGAIN if the swap cache folio no longer
>> + * matches the entry or the PMD changed under the lock (try_to_unuse will
>> + * rescan). Returns -EIO if the folio is not uptodate; in that case the
>> + * PMD is split so unuse_pte_range() can handle individual pages.
>> + */
>> +static int unuse_pmd(struct vm_area_struct *vma, pmd_t *pmd,
>> + unsigned long addr, softleaf_t entry,
>> + struct folio *folio)
>> +{
>> + struct mm_struct *mm = vma->vm_mm;
>> + struct page *page;
>> + pmd_t new_pmd, old_pmd;
>> + spinlock_t *ptl;
>> + rmap_t rmap_flags = RMAP_NONE;
>> + bool exclusive;
>> +
>> + if (unlikely(!folio_matches_swap_entry(folio, entry)))
>> + return -EAGAIN;
>> +
>> + if (unlikely(!folio_test_uptodate(folio))) {
>> + __split_huge_pmd(vma, pmd, addr, false);
>> + return -EIO;
>> + }
>> +
>> + page = folio_page(folio, 0);
>> +
>> + ptl = pmd_lock(mm, pmd);
>> + old_pmd = pmdp_get(pmd);
>> +
>> + if (!pmd_is_swap_entry(old_pmd) ||
>> + softleaf_from_pmd(old_pmd).val != entry.val) {
>> + spin_unlock(ptl);
>> + return -EAGAIN;
>> + }
>> +
>> + exclusive = pmd_swp_exclusive(old_pmd);
>> +
>> + /*
>> + * Some architectures may have to restore extra metadata to the folio
>> + * when reading from swap. This metadata may be indexed by swap entry
>> + * so this must be called before folio_put_swap().
>> + */
>> + arch_swap_restore(folio_swap(entry, folio), folio);
>> +
>> + add_mm_counter(mm, MM_ANONPAGES, HPAGE_PMD_NR);
>> + add_mm_counter(mm, MM_SWAPENTS, -HPAGE_PMD_NR);
>> +
>> + new_pmd = folio_mk_pmd(folio, vma->vm_page_prot);
>> + new_pmd = pmd_mkold(new_pmd);
>> + if (pmd_swp_soft_dirty(old_pmd))
>> + new_pmd = pmd_mksoft_dirty(new_pmd);
>> + if (pmd_swp_uffd_wp(old_pmd))
>> + new_pmd = pmd_mkuffd_wp(new_pmd);
>> +
>> + if (exclusive)
>> + rmap_flags |= RMAP_EXCLUSIVE;
>> +
>> + folio_get(folio);
>> + if (!folio_test_anon(folio))
>> + folio_add_new_anon_rmap(folio, vma, addr, rmap_flags);
>> + else
>> + folio_add_anon_rmap_pmd(folio, page, vma, addr, rmap_flags);
>> +
>> + set_pmd_at(mm, addr, pmd, new_pmd);
>> + folio_put_swap(folio, NULL);
>> +
>> + spin_unlock(ptl);
>> +
>> + folio_free_swap(folio);
>> + return 0;
>> +}
>> +
>> +/*
>> + * Try to swap in a PMD swap entry as a whole THP. Returns 0 on success.
>> + * Returns -ENOMEM if the PMD-order folio could not be allocated/charged,
>> + * -EIO if swap-in failed, or -EAGAIN if the cached folio is no longer
>> + * PMD-sized; in all of these the PMD is split so the caller can fall
>> + * back to unuse_pte_range(). Otherwise propagates the error from
>> + * unuse_pmd().
>> + */
>> +static int unuse_pmd_entry(struct vm_area_struct *vma, pmd_t *pmd,
>> + unsigned long addr, softleaf_t entry)
>> +{
>> + struct folio *folio;
>> + int ret;
>> +
>> + folio = swap_cache_get_folio(entry);
>> + if (!folio) {
>> + folio = swapin_alloc_pmd_folio(entry, vma->vm_mm);
>> + if (!folio) {
>> + ret = -ENOMEM;
>> + goto split_fallback;
>> + }
>> + }
>> +
>> + folio_lock(folio);
>> + folio_wait_writeback(folio);
>> + /*
>> + * If the cached folio is no longer PMD-sized (e.g. split in the
>> + * swap cache by deferred_split_scan() or memory_failure() while
>> + * the PMD swap entry was installed), the PMD swap entry no longer
>> + * maps a single contiguous folio. Split the PMD swap entry so
>> + * unuse_pte_range() can swap the per-slot folios in individually.
>> + */
>> + if (folio_nr_pages(folio) != HPAGE_PMD_NR) {
>> + folio_unlock(folio);
>> + folio_put(folio);
>> + ret = -EAGAIN;
>> + goto split_fallback;
>> + }
>> + ret = unuse_pmd(vma, pmd, addr, entry, folio);
>> + folio_unlock(folio);
>> + folio_put(folio);
>> + return ret;
>> +
>> +split_fallback:
>> + __split_huge_pmd(vma, pmd, addr, false);
>> + return ret;
>> +}
>> +
>> static inline int unuse_pmd_range(struct vm_area_struct *vma, pud_t *pud,
>> unsigned long addr, unsigned long end,
>> unsigned int type)
>> @@ -2531,6 +2656,18 @@ static inline int unuse_pmd_range(struct vm_area_struct *vma, pud_t *pud,
>> do {
>> cond_resched();
>> next = pmd_addr_end(addr, end);
>> +
>> + pmd_t pmdval = pmdp_get(pmd);
>> +
>> + if (pmd_is_swap_entry(pmdval)) {
>> + softleaf_t sl = softleaf_from_pmd(pmdval);
>> +
>> + if (swp_type(sl) == type) {
>> + if (!unuse_pmd_entry(vma, pmd, addr, sl))
>> + continue;
>> + }
>> + }
>> +
>> ret = unuse_pte_range(vma, pmd, addr, next, type);
>> if (ret)
>> return ret;