[v2 12/16] mm: handle PMD swap entries in MADV_WILLNEED

[Usama Arif]

swapin_walk_pmd_entry() walks PTEs and skips non-present PMDs, so
MADV_WILLNEED is a no-op on a PMD swap entry. Read the whole 2 MB
folio in at PMD order via swapin_sync(BIT(HPAGE_PMD_ORDER)) so the
subsequent fault hits do_huge_pmd_swap_page() and restores the THP
mapping; an order-0 read-ahead would force the fault to split.

Signed-off-by: Usama Arif <usama.arif@xxxxxxxxx>
---
 mm/madvise.c | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/mm/madvise.c b/mm/madvise.c
index 00539022f804..25f40542b951 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -193,6 +193,46 @@ static int swapin_walk_pmd_entry(pmd_t *pmd, unsigned long start,
 	spinlock_t *ptl;
 	unsigned long addr;
 
+	ptl = pmd_trans_huge_lock(pmd, vma);
+	if (ptl) {
+		pmd_t pmdval = *pmd;
+
+		if (pmd_is_swap_entry(pmdval)) {
+			softleaf_t entry = softleaf_from_pmd(pmdval);
+			struct vm_fault vmf = {
+				.vma = vma,
+				.address = start,
+				.real_address = start,
+				.pmd = pmd,
+			};
+			struct swap_info_struct *si;
+			struct folio *folio;
+
+			/*
+			 * Pin the swap device under the PMD lock so the
+			 * lookup is atomic with the PMD-swap-entry
+			 * observation; swapin_sync() requires its caller to
+			 * keep the device valid for the duration of the call.
+			 */
+			si = get_swap_device(entry);
+			spin_unlock(ptl);
+			if (!si) {
+				cond_resched();
+				return 0;
+			}
+
+			folio = swapin_sync(entry, GFP_HIGHUSER_MOVABLE,
+					    BIT(HPAGE_PMD_ORDER), &vmf,
+					    NULL, 0);
+			if (!IS_ERR_OR_NULL(folio))
+				folio_put(folio);
+			put_swap_device(si);
+			cond_resched();
+			return 0;
+		}
+		spin_unlock(ptl);
+	}
+
 	for (addr = start; addr < end; addr += PAGE_SIZE) {
 		pte_t pte;
 		softleaf_t entry;
-- 
2.52.0