From: Shakeel Butt <shakeelb@xxxxxxxxxx>
Subject: mm/rmap: always do TTU_IGNORE_ACCESS
Since commit 369ea8242c0f ("mm/rmap: update to new mmu_notifier semantic
v2"), the code to check the secondary MMU's page table access bit is
broken for !(TTU_IGNORE_ACCESS) because the page is unmapped from the
secondary MMU's page table before the check. More specifically for those
secondary MMUs which unmap the memory in
mmu_notifier_invalidate_range_start() like kvm.
However memory reclaim is the only user of !(TTU_IGNORE_ACCESS) or the
absence of TTU_IGNORE_ACCESS and it explicitly performs the page table
access check before trying to unmap the page. So, at worst the reclaim
will miss accesses in a very short window if we remove page table access
check in unmapping code.
There is an unintented consequence of !(TTU_IGNORE_ACCESS) for the memcg
reclaim. From memcg reclaim the page_referenced() only account the
accesses from the processes which are in the same memcg of the target page
but the unmapping code is considering accesses from all the processes, so,
decreasing the effectiveness of memcg reclaim.
The simplest solution is to always assume TTU_IGNORE_ACCESS in unmapping
code.
Link: https://lkml.kernel.org/r/20201104231928.1494083-1-shakeelb@xxxxxxxxxx
Fixes: 369ea8242c0f ("mm/rmap: update to new mmu_notifier semantic v2")
Signed-off-by: Shakeel Butt <shakeelb@xxxxxxxxxx>
Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Jerome Glisse <jglisse@xxxxxxxxxx>
Cc: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/rmap.h | 1 -
mm/huge_memory.c | 2 +-
mm/memory-failure.c | 2 +-
mm/memory_hotplug.c | 2 +-
mm/migrate.c | 8 +++-----
mm/rmap.c | 9 ---------
mm/vmscan.c | 14 +++++---------
7 files changed, 11 insertions(+), 27 deletions(-)
--- a/include/linux/rmap.h~mm-rmap-always-do-ttu_ignore_access
+++ a/include/linux/rmap.h
@@ -91,7 +91,6 @@ enum ttu_flags {
TTU_SPLIT_HUGE_PMD = 0x4, /* split huge PMD if any */
TTU_IGNORE_MLOCK = 0x8, /* ignore mlock */
- TTU_IGNORE_ACCESS = 0x10, /* don't age */
TTU_IGNORE_HWPOISON = 0x20, /* corrupted page is recoverable */
TTU_BATCH_FLUSH = 0x40, /* Batch TLB flushes where possible
* and caller guarantees they will
--- a/mm/huge_memory.c~mm-rmap-always-do-ttu_ignore_access
+++ a/mm/huge_memory.c
@@ -2321,7 +2321,7 @@ void vma_adjust_trans_huge(struct vm_are
static void unmap_page(struct page *page)
{
- enum ttu_flags ttu_flags = TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS |
+ enum ttu_flags ttu_flags = TTU_IGNORE_MLOCK |
TTU_RMAP_LOCKED | TTU_SPLIT_HUGE_PMD;
bool unmap_success;
--- a/mm/memory-failure.c~mm-rmap-always-do-ttu_ignore_access
+++ a/mm/memory-failure.c
@@ -989,7 +989,7 @@ static int get_hwpoison_page(struct page
static bool hwpoison_user_mappings(struct page *p, unsigned long pfn,
int flags, struct page **hpagep)
{
- enum ttu_flags ttu = TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS;
+ enum ttu_flags ttu = TTU_IGNORE_MLOCK;
struct address_space *mapping;
LIST_HEAD(tokill);
bool unmap_success = true;
--- a/mm/memory_hotplug.c~mm-rmap-always-do-ttu_ignore_access
+++ a/mm/memory_hotplug.c
@@ -1304,7 +1304,7 @@ do_migrate_range(unsigned long start_pfn
if (WARN_ON(PageLRU(page)))
isolate_lru_page(page);
if (page_mapped(page))
- try_to_unmap(page, TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS);
+ try_to_unmap(page, TTU_IGNORE_MLOCK);
continue;
}
--- a/mm/migrate.c~mm-rmap-always-do-ttu_ignore_access
+++ a/mm/migrate.c
@@ -1122,8 +1122,7 @@ static int __unmap_and_move(struct page
/* Establish migration ptes */
VM_BUG_ON_PAGE(PageAnon(page) && !PageKsm(page) && !anon_vma,
page);
- try_to_unmap(page,
- TTU_MIGRATION|TTU_IGNORE_MLOCK|TTU_IGNORE_ACCESS);
+ try_to_unmap(page, TTU_MIGRATION|TTU_IGNORE_MLOCK);
page_was_mapped = 1;
}
@@ -1329,8 +1328,7 @@ static int unmap_and_move_huge_page(new_
if (page_mapped(hpage)) {
bool mapping_locked = false;
- enum ttu_flags ttu = TTU_MIGRATION|TTU_IGNORE_MLOCK|
- TTU_IGNORE_ACCESS;
+ enum ttu_flags ttu = TTU_MIGRATION|TTU_IGNORE_MLOCK;
if (!PageAnon(hpage)) {
/*
@@ -2688,7 +2686,7 @@ static void migrate_vma_prepare(struct m
*/
static void migrate_vma_unmap(struct migrate_vma *migrate)
{
- int flags = TTU_MIGRATION | TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS;
+ int flags = TTU_MIGRATION | TTU_IGNORE_MLOCK;
const unsigned long npages = migrate->npages;
const unsigned long start = migrate->start;
unsigned long addr, i, restore = 0;
--- a/mm/rmap.c~mm-rmap-always-do-ttu_ignore_access
+++ a/mm/rmap.c
@@ -1533,15 +1533,6 @@ static bool try_to_unmap_one(struct page
goto discard;
}
- if (!(flags & TTU_IGNORE_ACCESS)) {
- if (ptep_clear_flush_young_notify(vma, address,
- pvmw.pte)) {
- ret = false;
- page_vma_mapped_walk_done(&pvmw);
- break;
- }
- }
-
/* Nuke the page table entry. */
flush_cache_page(vma, address, pte_pfn(*pvmw.pte));
if (should_defer_flush(mm, flags)) {
--- a/mm/vmscan.c~mm-rmap-always-do-ttu_ignore_access
+++ a/mm/vmscan.c
@@ -1072,7 +1072,6 @@ static void page_check_dirty_writeback(s
static unsigned int shrink_page_list(struct list_head *page_list,
struct pglist_data *pgdat,
struct scan_control *sc,
- enum ttu_flags ttu_flags,
struct reclaim_stat *stat,
bool ignore_references)
{
@@ -1297,7 +1296,7 @@ static unsigned int shrink_page_list(str
* processes. Try to unmap it here.
*/
if (page_mapped(page)) {
- enum ttu_flags flags = ttu_flags | TTU_BATCH_FLUSH;
+ enum ttu_flags flags = TTU_BATCH_FLUSH;
bool was_swapbacked = PageSwapBacked(page);
if (unlikely(PageTransHuge(page)))
@@ -1514,7 +1513,7 @@ unsigned int reclaim_clean_pages_from_li
}
nr_reclaimed = shrink_page_list(&clean_pages, zone->zone_pgdat, &sc,
- TTU_IGNORE_ACCESS, &stat, true);
+ &stat, true);
list_splice(&clean_pages, page_list);
mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE,
-(long)nr_reclaimed);
@@ -1958,8 +1957,7 @@ shrink_inactive_list(unsigned long nr_to
if (nr_taken == 0)
return 0;
- nr_reclaimed = shrink_page_list(&page_list, pgdat, sc, 0,
- &stat, false);
+ nr_reclaimed = shrink_page_list(&page_list, pgdat, sc, &stat, false);
spin_lock_irq(&pgdat->lru_lock);
@@ -2131,8 +2129,7 @@ unsigned long reclaim_pages(struct list_
nr_reclaimed += shrink_page_list(&node_page_list,
NODE_DATA(nid),
- &sc, 0,
- &dummy_stat, false);
+ &sc, &dummy_stat, false);
while (!list_empty(&node_page_list)) {
page = lru_to_page(&node_page_list);
list_del(&page->lru);
@@ -2145,8 +2142,7 @@ unsigned long reclaim_pages(struct list_
if (!list_empty(&node_page_list)) {
nr_reclaimed += shrink_page_list(&node_page_list,
NODE_DATA(nid),
- &sc, 0,
- &dummy_stat, false);
+ &sc, &dummy_stat, false);
while (!list_empty(&node_page_list)) {
page = lru_to_page(&node_page_list);
list_del(&page->lru);
_
