[RFC 6/6] mm/migrate_pages: batch flushing TLB

[Huang Ying]

The TLB flushing will cost quite some CPU cycles during the page
migration in some situations.  For example, when migrate a page of a
process with multiple active threads that run on multiple CPUs.  After
batching the _unmap and _move in migrate_pages(), the TLB flushing can
be batched easily with the existing TLB flush batching mechanism.
This patch implements that.

We use the following test case to test the patch.

On a 2-socket Intel server,

- Run pmbench memory accessing benchmark

- Run `migratepages` to migrate pages of pmbench between node 0 and
  node 1 back and forth.

With the patch, the TLB flushing IPI reduces 99.1% during the test and
the number of pages migrated successfully per second increases 291.7%.

Signed-off-by: "Huang, Ying" <ying.huang@xxxxxxxxx>
Cc: Zi Yan <ziy@xxxxxxxxxx>
Cc: Yang Shi <shy828301@xxxxxxxxx>
Cc: Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>
Cc: Oscar Salvador <osalvador@xxxxxxx>
Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx>
---
 mm/migrate.c |  4 +++-
 mm/rmap.c    | 24 ++++++++++++++++++++----
 2 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/mm/migrate.c b/mm/migrate.c
index 042fa147f302..a0de0d9b4d41 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -1179,7 +1179,7 @@ static int migrate_page_unmap(new_page_t get_new_page, free_page_t put_new_page,
 		/* Establish migration ptes */
 		VM_BUG_ON_PAGE(PageAnon(page) && !PageKsm(page) && !anon_vma,
 				page);
-		try_to_migrate(folio, 0);
+		try_to_migrate(folio, TTU_BATCH_FLUSH);
 		page_was_mapped = 1;
 	}
 
@@ -1647,6 +1647,8 @@ static int migrate_pages_batch(struct list_head *from, new_page_t get_new_page,
 	nr_thp_failed += thp_retry;
 	nr_failed_pages += nr_retry_pages;
 move:
+	try_to_unmap_flush();
+
 	retry = 1;
 	thp_retry = 1;
 	for (pass = 0; pass < 10 && (retry || thp_retry); pass++) {
diff --git a/mm/rmap.c b/mm/rmap.c
index 93d5a6f793d2..ab88136720dc 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1960,8 +1960,24 @@ static bool try_to_migrate_one(struct folio *folio, struct vm_area_struct *vma,
 			pteval = huge_ptep_clear_flush(vma, address, pvmw.pte);
 		} else {
 			flush_cache_page(vma, address, pte_pfn(*pvmw.pte));
-			/* Nuke the page table entry. */
-			pteval = ptep_clear_flush(vma, address, pvmw.pte);
+			/*
+			 * Nuke the page table entry.
+			 */
+			if (should_defer_flush(mm, flags)) {
+				/*
+				 * We clear the PTE but do not flush so potentially
+				 * a remote CPU could still be writing to the folio.
+				 * If the entry was previously clean then the
+				 * architecture must guarantee that a clear->dirty
+				 * transition on a cached TLB entry is written through
+				 * and traps if the PTE is unmapped.
+				 */
+				pteval = ptep_get_and_clear(mm, address, pvmw.pte);
+
+				set_tlb_ubc_flush_pending(mm, pte_dirty(pteval));
+			} else {
+				pteval = ptep_clear_flush(vma, address, pvmw.pte);
+			}
 		}
 
 		/* Set the dirty flag on the folio now the pte is gone. */
@@ -2128,10 +2144,10 @@ void try_to_migrate(struct folio *folio, enum ttu_flags flags)
 
 	/*
 	 * Migration always ignores mlock and only supports TTU_RMAP_LOCKED and
-	 * TTU_SPLIT_HUGE_PMD and TTU_SYNC flags.
+	 * TTU_SPLIT_HUGE_PMD, TTU_SYNC and TTU_BATCH_FLUSH flags.
 	 */
 	if (WARN_ON_ONCE(flags & ~(TTU_RMAP_LOCKED | TTU_SPLIT_HUGE_PMD |
-					TTU_SYNC)))
+					TTU_SYNC | TTU_BATCH_FLUSH)))
 		return;
 
 	if (folio_is_zone_device(folio) &&
-- 
2.35.1