Skip to contentSkip to conversation

mm-optimize-mprotect-for-mm_cp_prot_numa-by-batch-skipping-ptes.patch removed from -mm tree

merged mm-stable Andrew Morton 1 participant 1 message Jul 25, 2025

lore
Share & tools
lore

Andrew Morton

author tree notice Applied
#0 lore raw

The quilt patch titled
Subject: mm: optimize mprotect() for MM_CP_PROT_NUMA by batch-skipping PTEs
has been removed from the -mm tree. Its filename was
mm-optimize-mprotect-for-mm_cp_prot_numa-by-batch-skipping-ptes.patch

This patch was dropped because it was merged into the mm-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

------------------------------------------------------
From: Dev Jain <[email protected]>
Subject: mm: optimize mprotect() for MM_CP_PROT_NUMA by batch-skipping PTEs
Date: Fri, 18 Jul 2025 14:32:39 +0530

For the MM_CP_PROT_NUMA skipping case, observe that, if we skip an iteration due to the underlying folio satisfying any of the skip conditions, then for all subsequent ptes which map the same folio, the iteration will be skipped for them too. Therefore, we can optimize by using folio_pte_batch() to batch skip the iterations.

Use prot_numa_skip() introduced in the previous patch to determine whether we need to skip the iteration. Change its signature to have a double pointer to a folio, which will be used by mprotect_folio_pte_batch() to determine the number of iterations we can safely skip.

Cc Anshuman Khandual <[email protected]>
Cc Barry Song <[email protected]>
Cc Catalin Marinas <[email protected]>
Cc Christophe Leroy <[email protected]>
Cc David Hildenbrand <[email protected]>
Cc Hugh Dickins <[email protected]>
Cc Jann Horn <[email protected]>
Cc Joey Gouly <[email protected]>
Cc Kevin Brodsky <[email protected]>
Cc Lance Yang <[email protected]>
Cc Liam Howlett <[email protected]>
Cc Matthew Wilcox (Oracle) <[email protected]>
Cc Peter Xu <[email protected]>
Cc Vlastimil Babka <[email protected]>
Cc Will Deacon <[email protected]>
Cc Yang Shi <[email protected]>
Cc Yicong Yang <[email protected]>
Cc Zhenhua Huang <[email protected]>

---

 mm/mprotect.c |   55 ++++++++++++++++++++++++++++++++++++------------
 1 file changed, 42 insertions(+), 13 deletions(-)
mm/mprotect.c renamed from mm/mprotect.c~mm-optimize-mprotect-for-mm_cp_prot_numa-by-batch-skipping-ptes +42 -13
@@ -83,28 +83,43 @@ bool can_change_pte_writable(struct vm_a
 	return pte_dirty(pte);
 }
 
+static int mprotect_folio_pte_batch(struct folio *folio, pte_t *ptep,
+				    pte_t pte, int max_nr_ptes)
+{
+	/* No underlying folio, so cannot batch */
+	if (!folio)
+		return 1;
+
+	if (!folio_test_large(folio))
+		return 1;
+
+	return folio_pte_batch(folio, ptep, pte, max_nr_ptes);
+}
+
 static bool prot_numa_skip(struct vm_area_struct *vma, unsigned long addr,
-			   pte_t oldpte, pte_t *pte, int target_node)
+			   pte_t oldpte, pte_t *pte, int target_node,
+			   struct folio **foliop)
 {
-	struct folio *folio;
+	struct folio *folio = NULL;
+	bool ret = true;
 	bool toptier;
 	int nid;
 
 	/* Avoid TLB flush if possible */
 	if (pte_protnone(oldpte))
-		return true;
+		goto skip;
 
 	folio = vm_normal_folio(vma, addr, oldpte);
 	if (!folio)
-		return true;
+		goto skip;
 
 	if (folio_is_zone_device(folio) || folio_test_ksm(folio))
-		return true;
+		goto skip;
 
 	/* Also skip shared copy-on-write pages */
 	if (is_cow_mapping(vma->vm_flags) &&
 	    (folio_maybe_dma_pinned(folio) || folio_maybe_mapped_shared(folio)))
-		return true;
+		goto skip;
 
 	/*
 	 * While migration can move some dirty pages,
@@ -112,7 +127,7 @@ static bool prot_numa_skip(struct vm_are
 	 * context.
 	 */
 	if (folio_is_file_lru(folio) && folio_test_dirty(folio))
-		return true;
+		goto skip;
 
 	/*
 	 * Don't mess with PTEs if page is already on the node
@@ -120,7 +135,7 @@ static bool prot_numa_skip(struct vm_are
 	 */
 	nid = folio_nid(folio);
 	if (target_node == nid)
-		return true;
+		goto skip;
 
 	toptier = node_is_toptier(nid);
 
@@ -129,11 +144,15 @@ static bool prot_numa_skip(struct vm_are
 	 * balancing is disabled
 	 */
 	if (!(sysctl_numa_balancing_mode & NUMA_BALANCING_NORMAL) && toptier)
-		return true;
+		goto skip;
 
+	ret = false;
 	if (folio_use_access_time(folio))
 		folio_xchg_access_time(folio, jiffies_to_msecs(jiffies));
-	return false;
+
+skip:
+	*foliop = folio;
+	return ret;
 }
 
 static long change_pte_range(struct mmu_gather *tlb,
@@ -147,6 +166,7 @@ static long change_pte_range(struct mmu_
 	bool prot_numa = cp_flags & MM_CP_PROT_NUMA;
 	bool uffd_wp = cp_flags & MM_CP_UFFD_WP;
 	bool uffd_wp_resolve = cp_flags & MM_CP_UFFD_WP_RESOLVE;
+	int nr_ptes;
 
 	tlb_change_page_size(tlb, PAGE_SIZE);
 	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
@@ -161,8 +181,11 @@ static long change_pte_range(struct mmu_
 	flush_tlb_batched_pending(vma->vm_mm);
 	arch_enter_lazy_mmu_mode();
 	do {
+		nr_ptes = 1;
 		oldpte = ptep_get(pte);
 		if (pte_present(oldpte)) {
+			int max_nr_ptes = (end - addr) >> PAGE_SHIFT;
+			struct folio *folio;
 			pte_t ptent;
 
 			/*
@@ -170,9 +193,15 @@ static long change_pte_range(struct mmu_
 			 * pages. See similar comment in change_huge_pmd.
 			 */
 			if (prot_numa) {
-				if (prot_numa_skip(vma, addr, oldpte, pte,
-						   target_node))
+				int ret = prot_numa_skip(vma, addr, oldpte, pte,
+							 target_node, &folio);
+				if (ret) {
+
+					/* determine batch to skip */
+					nr_ptes = mprotect_folio_pte_batch(folio,
+						  pte, oldpte, max_nr_ptes);
 					continue;
+				}
 			}
 
 			oldpte = ptep_modify_prot_start(vma, addr, pte);
@@ -289,7 +318,7 @@ static long change_pte_range(struct mmu_
 				pages++;
 			}
 		}
-	} while (pte++, addr += PAGE_SIZE, addr != end);
+	} while (pte += nr_ptes, addr += nr_ptes * PAGE_SIZE, addr != end);
 	arch_leave_lazy_mmu_mode();
 	pte_unmap_unlock(pte - 1, ptl);
 

_

Patches currently in -mm which might be from [email protected] are

khugepaged-optimize-__collapse_huge_page_copy_succeeded-by-pte-batching.patch khugepaged-optimize-collapse_pte_mapped_thp-by-pte-batching.patch