On 18.06.24 01:11, Barry Song wrote:
From: Barry Song <v-songbaohua@xxxxxxxx>
For the !folio_test_anon(folio) case, we can now invoke folio_add_new_anon_rmap()
with the rmap flags set to either EXCLUSIVE or non-EXCLUSIVE. This action will
suppress the VM_WARN_ON_FOLIO check within __folio_add_anon_rmap() while initiating
the process of bringing up mTHP swapin.
static __always_inline void __folio_add_anon_rmap(struct folio *folio,
struct page *page, int nr_pages, struct vm_area_struct *vma,
unsigned long address, rmap_t flags, enum rmap_level level)
{
...
if (unlikely(!folio_test_anon(folio))) {
VM_WARN_ON_FOLIO(folio_test_large(folio) &&
level != RMAP_LEVEL_PMD, folio);
}
...
}
It also improves the code’s readability. Currently, all new anonymous
folios calling folio_add_anon_rmap_ptes() are order-0. This ensures
that new folios cannot be partially exclusive; they are either entirely
exclusive or entirely shared.
Suggested-by: David Hildenbrand <david@xxxxxxxxxx>
Signed-off-by: Barry Song <v-songbaohua@xxxxxxxx>
Tested-by: Shuai Yuan <yuanshuai@xxxxxxxx>
---
mm/memory.c | 8 ++++++++
mm/swapfile.c | 13 +++++++++++--
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/mm/memory.c b/mm/memory.c
index 1f24ecdafe05..620654c13b2f 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -4339,6 +4339,14 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
if (unlikely(folio != swapcache && swapcache)) {
folio_add_new_anon_rmap(folio, vma, address, RMAP_EXCLUSIVE);
folio_add_lru_vma(folio, vma);
+ } else if (!folio_test_anon(folio)) {
+ /*
+ * We currently only expect small !anon folios, for which we now
+ * that they are either fully exclusive or fully shared. If we
+ * ever get large folios here, we have to be careful.
+ */
+ VM_WARN_ON_ONCE(folio_test_large(folio));
+ folio_add_new_anon_rmap(folio, vma, address, rmap_flags);
} else {
folio_add_anon_rmap_ptes(folio, page, nr_pages, vma, address,
rmap_flags);
diff --git a/mm/swapfile.c b/mm/swapfile.c
index ae1d2700f6a3..69efa1a57087 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -1908,8 +1908,17 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
VM_BUG_ON_FOLIO(folio_test_writeback(folio), folio);
if (pte_swp_exclusive(old_pte))
rmap_flags |= RMAP_EXCLUSIVE;
-
- folio_add_anon_rmap_pte(folio, page, vma, addr, rmap_flags);
+ /*
+ * We currently only expect small !anon folios, for which we now that
+ * they are either fully exclusive or fully shared. If we ever get
+ * large folios here, we have to be careful.
+ */
+ if (!folio_test_anon(folio)) {
+ VM_WARN_ON_ONCE(folio_test_large(folio));
(comment applies to both cases)
Thinking about Hugh's comment, we should likely add here:
VM_WARN_ON_FOLIO(!folio_test_locked(folio), folio);
[the check we are removing from __folio_add_anon_rmap()]
and document for folio_add_new_anon_rmap() in patch #1, that when
dealing with folios that might be mapped concurrently by others, the
folio lock must be held.
+ folio_add_new_anon_rmap(folio, vma, addr, rmap_flags);
+ } else {
+ folio_add_anon_rmap_pte(folio, page, vma, addr, rmap_flags);
+ }
} else { /* ksm created a completely new copy */
folio_add_new_anon_rmap(folio, vma, addr, RMAP_EXCLUSIVE);
folio_add_lru_vma(folio, vma);
--
Cheers,
David / dhildenb
