Skip to contentSkip to conversation

mm: COW fixes part 2: reliable GUP pins of anonymous pages

PATCH Superseded David Hildenbrand 4 participants 38 messages Mar 15, 2022 – Mar 31, 2022

Subsystems (14)

v2 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

v2 cover mm: COW fixes part 2: reliable GUP pins of anonymous pages

This series is the result of the discussion on the previous approach [2]. More information on the general COW issues can be found there. It is based on v5.17-rc8 and [1], which resides in -mm and -next:
[PATCH v3 0/9] mm: COW fixes part 1: fix the COW security issue for THP and swap

v2 is located at:
https://github.com/davidhildenbrand/linux/tree/cow_fixes_part_2_v2

The plan is to rebase on top of v5.18-rc1 (there will be some collisions with folio reworks), pull it early into -next, to then eventually have it in v5.19. This should give it some good test coverage in -next. We could then decide, whether we want to backport into -stable trees.

This series fixes memory corruptions when a GUP pin (FOLL_PIN) was taken on an anonymous page and COW logic fails to detect exclusivity of the page to then replacing the anonymous page by a copy in the page table: The GUP pin lost synchronicity with the pages mapped into the page tables.

This issue, including other related COW issues, has been summarized in [3] under 3):
"
3. Intra Process Memory Corruptions due to Wrong COW (FOLL_PIN)

page_maybe_dma_pinned() is used to check if a page may be pinned for DMA (using FOLL_PIN instead of FOLL_GET). While false positives are tolerable, false negatives are problematic: pages that are pinned for DMA must not be added to the swapcache. If it happens, the (now pinned) page could be faulted back from the swapcache into page tables read-only. Future write-access would detect the pinning and COW the page, losing synchronicity. For the interested reader, this is nicely documented in feb889fb40fa ("mm: don't put pinned pages into the swap cache").

Peter reports [8] that page_maybe_dma_pinned() as used is racy in some cases and can result in a violation of the documented semantics:
giving false negatives because of the race.

There are cases where we call it without properly taking a per-process sequence lock, turning the usage of page_maybe_dma_pinned() racy. While one case (clear_refs SOFTDIRTY tracking, see below) seems to be easy to handle, there is especially one rmap case (shrink_page_list) that's hard to fix: in the rmap world, we're not limited to a single process.

The shrink_page_list() issue is really subtle. If we race with someone pinning a page, we can trigger the same issue as in the FOLL_GET case. See the detail section at the end of this mail on a discussion how bad this can bite us with VFIO or other FOLL_PIN user.

It's harder to reproduce, but I managed to modify the O_DIRECT reproducer to use io_uring fixed buffers [15] instead, which ends up using FOLL_PIN | FOLL_WRITE | FOLL_LONGTERM to pin buffer pages and can similarly trigger a loss of synchronicity and consequently a memory corruption.

Again, the root issue is that a write-fault on a page that has additional references results in a COW and thereby a loss of synchronicity and consequently a memory corruption if two parties believe they are referencing the same page.
"

This series makes GUP pins (R/O and R/W) on anonymous pages fully reliable, especially also taking care of concurrent pinning via GUP-fast, for example, also fully fixing an issue reported regarding NUMA balancing [4] recently. While doing that, it further reduces "unnecessary COWs", especially when we don't fork()/KSM and don't swapout, and fixes the COW security for hugetlb for FOLL_PIN.

In summary, we track via a pageflag (PG_anon_exclusive) whether a mapped anonymous page is exclusive. Exclusive anonymous pages that are mapped R/O can directly be mapped R/W by the COW logic in the write fault handler. Exclusive anonymous pages that want to be shared (fork(), KSM) first have to be marked shared -- which will fail if there are GUP pins on the page. GUP is only allowed to take a pin on anonymous pages that are exclusive. The PT lock is the primary mechanism to synchronize modifications of PG_anon_exclusive. We syncronize against GUP-fast either via the src_mm->write_protect_seq (during fork()) or via clear/invalidate+flush of the relevant page table entry.

Special care has to be taken about swap, migration, and THPs (whereby a PMD-mapping can be converted to a PTE mapping and we have to track information for subpages). Besides these, we let the rmap code handle most magic. For reliable R/O pins of anonymous pages, we need FAULT_FLAG_UNSHARE logic as part of our previous approach [2], however, it's now 100% mapcount free and I further simplified it a bit.

#1 is a fix
#3-#9 are mostly rmap preparations for PG_anon_exclusive handling #10 introduces PG_anon_exclusive
#11 uses PG_anon_exclusive and make R/W pins of anonymous pages
reliable
#12 is a preparation for reliable R/O pins
#13 and #14 is reused/modified GUP-triggered unsharing for R/O GUP pins
make R/O pins of anonymous pages reliable
#15 adds sanity check when (un)pinning anonymous pages

Patch #11 contains excessive explanations and the main logic for R/W pins. #12 and #13 resemble what we proposed in the previous approach [2]. I consider the general approach of #15 very nice and helpful, and I remember Linus even envisioning something like that for finding BUGs.

It passes my growing set of tests for "wrong COW" and "missed COW", including the ones in [3] -- I'd really appreciate some experienced eyes to take a close look at corner cases.

I'll be sending out part 3 next, that will remember PG_anon_exclusive for ordinary swap entries: this will make FOLL_GET | FOLL_WRITE references reliable and fix the memory corruptions for O_DIRECT -- as described in
[3] under 2) -- as well, as long as there is no fork().

The long term goal should be to convert relevant users of FOLL_GET to FOLL_PIN, however, with part 3 it wouldn't be required to fix the obvious memory corruptions we are aware of.

[1] https://lkml.kernel.org/r/[email protected]
[2] https://lkml.kernel.org/r/[email protected]
[3] https://lore.kernel.org/r/[email protected]
[4] https://bugzilla.kernel.org/show_bug.cgi?id=215616

v1 -> v2:
* Tested on aarch64, ppc64, s390x and x86_64
* "mm/page-flags: reuse PG_mappedtodisk as PG_anon_exclusive for PageAnon() pages"
-> Use PG_mappedtodisk instead of PG_slab (thanks Willy!), this simlifies
the patch and necessary handling a lot. Add safety BUG_ON's
-> Move most documentation to the patch description, to be placed in a
proper documentation doc in the future, once everything's in place
* ""mm: remember exclusively mapped anonymous pages with PG_anon_exclusive -> Skip check+clearing in page_try_dup_anon_rmap(), otherwise we might trigger a wrong VM_BUG_ON() for KSM pages in ClearPageAnonExclusive() -> In __split_huge_pmd_locked(), call page_try_share_anon_rmap() only for "anon_exclusive", otherwise we might trigger a wrong VM_BUG_ON() -> In __split_huge_page_tail(), drop any remaining PG_anon_exclusive on tail pages, and document why that is fine

RFC -> v1:
* Rephrased/extended some patch descriptions+comments
* Tested on aarch64, ppc64 and x86_64
* "mm/rmap: convert RMAP flags to a proper distinct rmap_t type" -> Added
* "mm/rmap: drop "compound" parameter from page_add_new_anon_rmap()" -> Added
* "mm: remember exclusively mapped anonymous pages with PG_anon_exclusive" -> Fixed __do_huge_pmd_anonymous_page() to recheck after temporarily dropping the PT lock.
-> Use "reuse" label in __do_huge_pmd_anonymous_page()
-> Slightly simplify logic in hugetlb_cow()
-> In remove_migration_pte(), remove unrelated changes around
page_remove_rmap()
* "mm: support GUP-triggered unsharing of anonymous pages"
-> In handle_pte_fault(), trigger pte_mkdirty() only with
FAULT_FLAG_WRITE
-> In __handle_mm_fault(), extend comment regarding anonymous PUDs
* "mm/gup: trigger FAULT_FLAG_UNSHARE when R/O-pinning a possibly shared anonymous page"
-> Added unsharing logic to gup_hugepte() and gup_huge_pud() -> Changed return logic in __follow_hugetlb_must_fault(), making sure
that "unshare" is always set
* "mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning"
-> Slightly simplified sanity_check_pinned_pages()

David Hildenbrand (15):
mm/rmap: fix missing swap_free() in try_to_unmap() after
arch_unmap_one() failed
mm/hugetlb: take src_mm->write_protect_seq in
copy_hugetlb_page_range()
mm/memory: slightly simplify copy_present_pte()
mm/rmap: split page_dup_rmap() into page_dup_file_rmap() and
page_try_dup_anon_rmap()
mm/rmap: convert RMAP flags to a proper distinct rmap_t type mm/rmap: remove do_page_add_anon_rmap()
mm/rmap: pass rmap flags to hugepage_add_anon_rmap() mm/rmap: drop "compound" parameter from page_add_new_anon_rmap() mm/rmap: use page_move_anon_rmap() when reusing a mapped PageAnon()
page exclusively
mm/page-flags: reuse PG_mappedtodisk as PG_anon_exclusive for
PageAnon() pages
mm: remember exclusively mapped anonymous pages with PG_anon_exclusive mm/gup: disallow follow_page(FOLL_PIN)
mm: support GUP-triggered unsharing of anonymous pages mm/gup: trigger FAULT_FLAG_UNSHARE when R/O-pinning a possibly shared
anonymous page
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are
exclusive when (un)pinning

 include/linux/mm.h         |  46 +++++++-
 include/linux/mm_types.h   |   8 ++
 include/linux/page-flags.h |  39 ++++++-
 include/linux/rmap.h       | 116 +++++++++++++++++--
 include/linux/swap.h       |  15 ++-
 include/linux/swapops.h    |  25 ++++
 kernel/events/uprobes.c    |   2 +-
 mm/gup.c                   | 103 +++++++++++++++-
 mm/huge_memory.c           | 124 +++++++++++++++-----
 mm/hugetlb.c               | 135 ++++++++++++++-------
 mm/khugepaged.c            |   2 +-
 mm/ksm.c                   |  15 ++-
 mm/memory.c                | 232 +++++++++++++++++++++++--------------
 mm/memremap.c              |   9 ++
 mm/migrate.c               |  40 ++++++-
 mm/mprotect.c              |   8 +-
 mm/rmap.c                  |  95 +++++++++++----
 mm/swapfile.c              |   8 +-
 mm/userfaultfd.c           |   2 +-
 tools/vm/page-types.c      |   8 +-
 20 files changed, 815 insertions(+), 217 deletions(-)
Signature · 1 line
2.35.1
#1 lore raw

PATCH v2 1/15 mm/rmap: fix missing swap_free() in try_to_unmap() after arch_unmap_one() failed

In case arch_unmap_one() fails, we already did a swap_duplicate(). let's undo that properly via swap_free().

Fixes ca827d55ebaa ("mm, swap: Add infrastructure for saving page metadata on swap")
---
 mm/rmap.c | 1 +
 1 file changed, 1 insertion(+)
mm/rmap.c +1 -0
index 6a1e8c7f6213..f825aeef61ca 100644
@@ -1625,6 +1625,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
 				break;
 			}
 			if (arch_unmap_one(mm, vma, address, pteval) < 0) {
+				swap_free(entry);
 				set_pte_at(mm, address, pvmw.pte, pteval);
 				ret = false;
 				page_vma_mapped_walk_done(&pvmw);
Signature · 1 line
2.35.1
#2 lore raw

Re: [PATCH v2 01/15] mm/rmap: fix missing swap_free() in try_to_unmap() after arch_unmap_one() failed

Quoted 15 lines · David Hildenbrand

In case arch_unmap_one() fails, we already did a swap_duplicate(). let's undo that properly via swap_free().

Fixes ca827d55ebaa ("mm, swap: Add infrastructure for saving page metadata on swap")
---
 mm/rmap.c | 1 +
 1 file changed, 1 insertion(+)
mm/rmap.c +0 -0
index 6a1e8c7f6213..f825aeef61ca 100644

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1631

Quoted 7 lines
mm/rmap.c +1 -0
@@ -1625,6 +1625,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
 				break;
 			}
 			if (arch_unmap_one(mm, vma, address, pteval) < 0) {
+				swap_free(entry);
 				set_pte_at(mm, address, pvmw.pte, pteval);
 				ret = false;
 				page_vma_mapped_walk_done(&pvmw);

Hi Khalid,

I'm a bit confused about the semantics if arch_unmap_one(), I hope you can clarify.

See patch #11 in this series, were we can fail unmapping after arch_unmap_one() succeeded. E.g.,

@@ -1623,6 +1634,24 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
page_vma_mapped_walk_done(&pvmw);
break;
}
+ if (anon_exclusive &&
+ page_try_share_anon_rmap(subpage)) {
+ swap_free(entry);
+ set_pte_at(mm, address, pvmw.pte, pteval);
+ ret = false;
+ page_vma_mapped_walk_done(&pvmw);
+ break;
+ }
+ /*
+ * Note: We don't remember yet if the page was mapped
+ * exclusively in the swap entry, so swapin code has
+ * to re-determine that manually and might detect the
+ * page as possibly shared, for example, if there are
+ * other references on the page or if the page is under
+ * writeback. We made sure that there are no GUP pins
+ * on the page that would rely on it, so for GUP pins
+ * this is fine.
+ */
if (list_empty(&mm->mmlist)) {
spin_lock(&mmlist_lock);
if (list_empty(&mm->mmlist))

For now, I was under the impression that we don't have to undo anything after arch_unmap_one() succeeded, because we seem to not do anything for two cases below. But looking into arch_unmap_one() and how it allocates stuff I do wonder what we would actually want to do here -- I'd assume we'd want to trigger the del_tag_store() somehow?

arch_unmap_one() calls adi_save_tags(), which allocates memory. adi_restore_tags()->del_tag_store() reverts that operation and ends up freeing memory conditionally; However, it's only
called from arch_do_swap_page().

Here is why I have to scratch my head:

a) arch_do_swap_page() is only called from do_swap_page(). We don't do anything similar for mm/swapfile.c:unuse_pte(), aren't we missing something?

b) try_to_migrate_one() does the arch_unmap_one(), but who will do the restore+free after migration succeeded or failed, aren't we missing something?

I assume that we won't be properly freeing the tag space in case of a), and won't be properly restoring/migrating the tags in case of a) and b).

I'll send out v3 of this series today, and I'll keep ignoring arch_unmap_one() for now, because I have no clue what to do and it looks incomplete already, unless I am missing something important.

Signature · 3 lines
Thanks,

David / dhildenb
#3 lore raw

Re: [PATCH v2 01/15] mm/rmap: fix missing swap_free() in try_to_unmap() after arch_unmap_one() failed

Quoted 58 lines · David Hildenbrand
Quoted 15 lines · David Hildenbrand

In case arch_unmap_one() fails, we already did a swap_duplicate(). let's undo that properly via swap_free().

Fixes ca827d55ebaa ("mm, swap: Add infrastructure for saving page metadata on swap")
---
  mm/rmap.c | 1 +
  1 file changed, 1 insertion(+)
mm/rmap.c +0 -0
index 6a1e8c7f6213..f825aeef61ca 100644

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1631

Quoted 7 lines
mm/rmap.c +1 -0
@@ -1625,6 +1625,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
  				break;
  			}
  			if (arch_unmap_one(mm, vma, address, pteval) < 0) {
+				swap_free(entry);
  				set_pte_at(mm, address, pvmw.pte, pteval);
  				ret = false;
  				page_vma_mapped_walk_done(&pvmw);

Hi Khalid,

I'm a bit confused about the semantics if arch_unmap_one(), I hope you can clarify.

See patch #11 in this series, were we can fail unmapping after arch_unmap_one() succeeded. E.g.,

(quoted diff) +18 -0
@@ -1623,6 +1634,24 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
  				page_vma_mapped_walk_done(&pvmw);
  				break;
  			}
+			if (anon_exclusive &&
+			    page_try_share_anon_rmap(subpage)) {
+				swap_free(entry);
+				set_pte_at(mm, address, pvmw.pte, pteval);
+				ret = false;
+				page_vma_mapped_walk_done(&pvmw);
+				break;
+			}
+			/*
+			 * Note: We don't remember yet if the page was mapped
+			 * exclusively in the swap entry, so swapin code has
+			 * to re-determine that manually and might detect the
+			 * page as possibly shared, for example, if there are
+			 * other references on the page or if the page is under
+			 * writeback. We made sure that there are no GUP pins
+			 * on the page that would rely on it, so for GUP pins
+			 * this is fine.
+			 */
  			if (list_empty(&mm->mmlist)) {
  				spin_lock(&mmlist_lock);
  				if (list_empty(&mm->mmlist))
Quoted 5 lines

For now, I was under the impression that we don't have to undo anything after arch_unmap_one() succeeded, because we seem to not do anything for two cases below. But looking into arch_unmap_one() and how it allocates stuff I do wonder what we would actually want to do here -- I'd assume we'd want to trigger the del_tag_store() somehow?

Hi David,

Currently once arch_unmap_one() completes successfully, we are at the point of no return for this pte. It will be replaced by swap pte soon thereafter. Patch 11 adds another case where we may return without replacing current pte with swap pte. For now could you resolve this by moving the above code block in patch 11 to before the call to arch_unmap_one(). That still leaves open the issue having the flexibility of undoing what arch_unmap_one() does for some other reason in future. That will require coming up with a properly architected way to do it.

Quoted 11 lines

arch_unmap_one() calls adi_save_tags(), which allocates memory. adi_restore_tags()->del_tag_store() reverts that operation and ends up freeing memory conditionally; However, it's only
called from arch_do_swap_page().

Here is why I have to scratch my head:

a) arch_do_swap_page() is only called from do_swap_page(). We don't do anything similar for mm/swapfile.c:unuse_pte(), aren't we missing something?

My understanding of this code path maybe flawed, so do correct me if this does not sound right. unused_pte() is called upon user turning off swap on a device. unused_pte() is called by unused_pte_range() which swaps the page back in from swap device before calling unuse_pte(). Once the page is read back in from swap, ultimately access to the va for the page will result in call to __handle_mm_fault() which in turn will call handle_pte_fault() to insert a new pte for this mapping and handle_pte_fault() will call arch_do_swap_page() which will restore the tags.

Quoted 3 lines

b) try_to_migrate_one() does the arch_unmap_one(), but who will do the restore+free after migration succeeded or failed, aren't we missing something?

try_to_migrate_one() replaces the current pte with a migration pte after calling arch_unmap_one(). This causes __handle_mm_fault() to be called when a reference to the va covered by migration pte is made. This will in turn finally result in a call to arch_do_swap_page() which restores the tags.

Quoted 4 lines

I assume that we won't be properly freeing the tag space in case of a), and won't be properly restoring/migrating the tags in case of a) and b).

I think both cases are covered correctly unless I am misunderstand the code.

--
Khalid

#4 lore raw

Re: [PATCH v2 01/15] mm/rmap: fix missing swap_free() in try_to_unmap() after arch_unmap_one() failed

Quoted 59 lines · Khalid Aziz
Quoted 58 lines · David Hildenbrand
Quoted 15 lines · David Hildenbrand

In case arch_unmap_one() fails, we already did a swap_duplicate(). let's undo that properly via swap_free().

Fixes ca827d55ebaa ("mm, swap: Add infrastructure for saving page metadata on swap")
---
  mm/rmap.c | 1 +
  1 file changed, 1 insertion(+)
mm/rmap.c +0 -0
index 6a1e8c7f6213..f825aeef61ca 100644

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1631

Quoted 7 lines
mm/rmap.c +1 -0
@@ -1625,6 +1625,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
  				break;
  			}
  			if (arch_unmap_one(mm, vma, address, pteval) < 0) {
+				swap_free(entry);
  				set_pte_at(mm, address, pvmw.pte, pteval);
  				ret = false;
  				page_vma_mapped_walk_done(&pvmw);

Hi Khalid,

I'm a bit confused about the semantics if arch_unmap_one(), I hope you can clarify.

See patch #11 in this series, were we can fail unmapping after arch_unmap_one() succeeded. E.g.,

(quoted diff) +18 -0
@@ -1623,6 +1634,24 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
  				page_vma_mapped_walk_done(&pvmw);
  				break;
  			}
+			if (anon_exclusive &&
+			    page_try_share_anon_rmap(subpage)) {
+				swap_free(entry);
+				set_pte_at(mm, address, pvmw.pte, pteval);
+				ret = false;
+				page_vma_mapped_walk_done(&pvmw);
+				break;
+			}
+			/*
+			 * Note: We don't remember yet if the page was mapped
+			 * exclusively in the swap entry, so swapin code has
+			 * to re-determine that manually and might detect the
+			 * page as possibly shared, for example, if there are
+			 * other references on the page or if the page is under
+			 * writeback. We made sure that there are no GUP pins
+			 * on the page that would rely on it, so for GUP pins
+			 * this is fine.
+			 */
  			if (list_empty(&mm->mmlist)) {
  				spin_lock(&mmlist_lock);
  				if (list_empty(&mm->mmlist))
Quoted 8 lines
Quoted 5 lines

For now, I was under the impression that we don't have to undo anything after arch_unmap_one() succeeded, because we seem to not do anything for two cases below. But looking into arch_unmap_one() and how it allocates stuff I do wonder what we would actually want to do here -- I'd assume we'd want to trigger the del_tag_store() somehow?

Hi David,

Hi,

thanks for your fast reply.

Quoted 5 lines

Currently once arch_unmap_one() completes successfully, we are at the point of no return for this pte. It will be replaced by swap pte soon thereafter. Patch 11 adds another case where we may return without replacing current pte with swap pte. For now could you resolve this by moving the above code block in patch 11 to before the call to arch_unmap_one(). That still leaves open the issue having the flexibility of undoing what arch_unmap_one() does for some other reason in future. That will require coming up with a properly architected way to do it.

I really want clearing PG_anon_exclusive be the last action, without eventually having to set it again and overcomplicating PG_anon_exclusive/rmap handling. Ideally, we'd have a "arch_remap_one()" that just reverts what arch_unmap_one() did.

Quoted 10 lines
Quoted 10 lines

arch_unmap_one() calls adi_save_tags(), which allocates memory. adi_restore_tags()->del_tag_store() reverts that operation and ends up freeing memory conditionally; However, it's only
called from arch_do_swap_page().

Here is why I have to scratch my head:

Quoted 8 lines
Quoted 2 lines

a) arch_do_swap_page() is only called from do_swap_page(). We don't do anything similar for mm/swapfile.c:unuse_pte(), aren't we missing something?

My understanding of this code path maybe flawed, so do correct me if this does not sound right. unused_pte() is called upon user turning off swap on a device. unused_pte() is called by unused_pte_range() which swaps the page back in from swap device before calling unuse_pte(). Once the page is read back in from swap, ultimately access to the va for the page will result in call to __handle_mm_fault() which in turn will call handle_pte_fault() to insert a new pte for this mapping and handle_pte_fault() will call arch_do_swap_page() which will restore the tags.

unuse_pte() will replace a swap pte directly by a proper, present pte, just like do_swap_page() would. You won't end up in do_swap_page() anymore and arch_do_swap_page() won't be called, because there is no swap PTE anymore.

Quoted 8 lines
Quoted 4 lines

b) try_to_migrate_one() does the arch_unmap_one(), but who will do the restore+free after migration succeeded or failed, aren't we missing something?

try_to_migrate_one() replaces the current pte with a migration pte after calling arch_unmap_one(). This causes __handle_mm_fault() to be called when a reference to the va covered by migration pte is made. This will in turn finally result in a call to arch_do_swap_page() which restores the tags.

Migration PTEs are restore via mm/migrate.c:remove_migration_ptes(). arch_do_swap_page() won't be called.

What you mention is if someone accesses the migration PTE while migration is active and the migration PTEs have not been removed yet. While we'll end up in do_swap_page(), we'll do a migration_entry_wait(), followed by an effective immediate "return 0;". arch_do_swap_page() won't get called.

So in essence, I think this doesn't work as expected yet. In the best case we don't immediately free memory. In the worst case we lose the tags.

Signature · 3 lines
Thanks,

David / dhildenb
#5 lore raw

Re: [PATCH v2 01/15] mm/rmap: fix missing swap_free() in try_to_unmap() after arch_unmap_one() failed

Quoted 79 lines · David Hildenbrand
Quoted 59 lines · Khalid Aziz
Quoted 58 lines · David Hildenbrand
Quoted 15 lines · David Hildenbrand

In case arch_unmap_one() fails, we already did a swap_duplicate(). let's undo that properly via swap_free().

Fixes ca827d55ebaa ("mm, swap: Add infrastructure for saving page metadata on swap")
---
   mm/rmap.c | 1 +
   1 file changed, 1 insertion(+)
mm/rmap.c +0 -0
index 6a1e8c7f6213..f825aeef61ca 100644

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1631

Quoted 7 lines
mm/rmap.c +1 -0
@@ -1625,6 +1625,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
   				break;
   			}
   			if (arch_unmap_one(mm, vma, address, pteval) < 0) {
+				swap_free(entry);
   				set_pte_at(mm, address, pvmw.pte, pteval);
   				ret = false;
   				page_vma_mapped_walk_done(&pvmw);

Hi Khalid,

I'm a bit confused about the semantics if arch_unmap_one(), I hope you can clarify.

See patch #11 in this series, were we can fail unmapping after arch_unmap_one() succeeded. E.g.,

(quoted diff) +18 -0
@@ -1623,6 +1634,24 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
   				page_vma_mapped_walk_done(&pvmw);
   				break;
   			}
+			if (anon_exclusive &&
+			    page_try_share_anon_rmap(subpage)) {
+				swap_free(entry);
+				set_pte_at(mm, address, pvmw.pte, pteval);
+				ret = false;
+				page_vma_mapped_walk_done(&pvmw);
+				break;
+			}
+			/*
+			 * Note: We don't remember yet if the page was mapped
+			 * exclusively in the swap entry, so swapin code has
+			 * to re-determine that manually and might detect the
+			 * page as possibly shared, for example, if there are
+			 * other references on the page or if the page is under
+			 * writeback. We made sure that there are no GUP pins
+			 * on the page that would rely on it, so for GUP pins
+			 * this is fine.
+			 */
   			if (list_empty(&mm->mmlist)) {
   				spin_lock(&mmlist_lock);
   				if (list_empty(&mm->mmlist))
Quoted 8 lines
Quoted 5 lines

For now, I was under the impression that we don't have to undo anything after arch_unmap_one() succeeded, because we seem to not do anything for two cases below. But looking into arch_unmap_one() and how it allocates stuff I do wonder what we would actually want to do here -- I'd assume we'd want to trigger the del_tag_store() somehow?

Hi David,

Hi,

thanks for your fast reply.

Quoted 5 lines

Currently once arch_unmap_one() completes successfully, we are at the point of no return for this pte. It will be replaced by swap pte soon thereafter. Patch 11 adds another case where we may return without replacing current pte with swap pte. For now could you resolve this by moving the above code block in patch 11 to before the call to arch_unmap_one(). That still leaves open the issue having the flexibility of undoing what arch_unmap_one() does for some other reason in future. That will require coming up with a properly architected way to do it.

Quoted 4 lines

I really want clearing PG_anon_exclusive be the last action, without eventually having to set it again and overcomplicating PG_anon_exclusive/rmap handling. Ideally, we'd have a "arch_remap_one()" that just reverts what arch_unmap_one() did.

Hi David,

arch_remap_one() sounds reasonable. Would you like to add that in your patch series, or would you like me to create a separate patch for adding this on top of your patch series?

Quoted 44 lines
Quoted 11 lines
Quoted 11 lines

arch_unmap_one() calls adi_save_tags(), which allocates memory. adi_restore_tags()->del_tag_store() reverts that operation and ends up freeing memory conditionally; However, it's only
called from arch_do_swap_page().

Here is why I have to scratch my head:

Quoted 8 lines
Quoted 2 lines

a) arch_do_swap_page() is only called from do_swap_page(). We don't do anything similar for mm/swapfile.c:unuse_pte(), aren't we missing something?

My understanding of this code path maybe flawed, so do correct me if this does not sound right. unused_pte() is called upon user turning off swap on a device. unused_pte() is called by unused_pte_range() which swaps the page back in from swap device before calling unuse_pte(). Once the page is read back in from swap, ultimately access to the va for the page will result in call to __handle_mm_fault() which in turn will call handle_pte_fault() to insert a new pte for this mapping and handle_pte_fault() will call arch_do_swap_page() which will restore the tags.

unuse_pte() will replace a swap pte directly by a proper, present pte, just like do_swap_page() would. You won't end up in do_swap_page() anymore and arch_do_swap_page() won't be called, because there is no swap PTE anymore.

Quoted 8 lines
Quoted 4 lines

b) try_to_migrate_one() does the arch_unmap_one(), but who will do the restore+free after migration succeeded or failed, aren't we missing something?

try_to_migrate_one() replaces the current pte with a migration pte after calling arch_unmap_one(). This causes __handle_mm_fault() to be called when a reference to the va covered by migration pte is made. This will in turn finally result in a call to arch_do_swap_page() which restores the tags.

Migration PTEs are restore via mm/migrate.c:remove_migration_ptes(). arch_do_swap_page() won't be called.

What you mention is if someone accesses the migration PTE while migration is active and the migration PTEs have not been removed yet. While we'll end up in do_swap_page(), we'll do a migration_entry_wait(), followed by an effective immediate "return 0;". arch_do_swap_page() won't get called.

Quoted 3 lines

So in essence, I think this doesn't work as expected yet. In the best case we don't immediately free memory. In the worst case we lose the tags.

I see what you mean. I can work on fixing these issues up.

Thanks,
Khalid

#6 lore raw

Re: [PATCH v2 01/15] mm/rmap: fix missing swap_free() in try_to_unmap() after arch_unmap_one() failed

Quoted 1 line

Hi David,

Hi,

Quoted 3 lines

arch_remap_one() sounds reasonable. Would you like to add that in your patch series, or would you like me to create a separate patch for adding this on top of your patch series?

Let's handle it separately, because I think there is still a lot to be clarified. I also feel like the terminology ("arch_unmap_one()") is a little too generic, if we really only care about anonymous pages (do we?).

Quoted 45 lines
Quoted 45 lines
Quoted 12 lines
Quoted 12 lines

arch_unmap_one() calls adi_save_tags(), which allocates memory. adi_restore_tags()->del_tag_store() reverts that operation and ends up freeing memory conditionally; However, it's only
called from arch_do_swap_page().

Here is why I have to scratch my head:

Quoted 8 lines
Quoted 2 lines

a) arch_do_swap_page() is only called from do_swap_page(). We don't do anything similar for mm/swapfile.c:unuse_pte(), aren't we missing something?

My understanding of this code path maybe flawed, so do correct me if this does not sound right. unused_pte() is called upon user turning off swap on a device. unused_pte() is called by unused_pte_range() which swaps the page back in from swap device before calling unuse_pte(). Once the page is read back in from swap, ultimately access to the va for the page will result in call to __handle_mm_fault() which in turn will call handle_pte_fault() to insert a new pte for this mapping and handle_pte_fault() will call arch_do_swap_page() which will restore the tags.

unuse_pte() will replace a swap pte directly by a proper, present pte, just like do_swap_page() would. You won't end up in do_swap_page() anymore and arch_do_swap_page() won't be called, because there is no swap PTE anymore.

Quoted 8 lines
Quoted 4 lines

b) try_to_migrate_one() does the arch_unmap_one(), but who will do the restore+free after migration succeeded or failed, aren't we missing something?

try_to_migrate_one() replaces the current pte with a migration pte after calling arch_unmap_one(). This causes __handle_mm_fault() to be called when a reference to the va covered by migration pte is made. This will in turn finally result in a call to arch_do_swap_page() which restores the tags.

Migration PTEs are restore via mm/migrate.c:remove_migration_ptes(). arch_do_swap_page() won't be called.

What you mention is if someone accesses the migration PTE while migration is active and the migration PTEs have not been removed yet. While we'll end up in do_swap_page(), we'll do a migration_entry_wait(), followed by an effective immediate "return 0;". arch_do_swap_page() won't get called.

Quoted 5 lines
Quoted 3 lines

So in essence, I think this doesn't work as expected yet. In the best case we don't immediately free memory. In the worst case we lose the tags.

I see what you mean. I can work on fixing these issues up.

Ideally, I think we'd handle migration differently: migrate the tags directly instead of temporarily storing them

I also wonder, how to handle migration of THP (which uses migration PMDs) and THP splitting (which also uses migration PTEs); maybe they don't apply on sparc?

Further, I wonder if you have to handle zapping of swap/migration PTEs, and if you'd also have to cleanup+freeup any allcoated tag memory? E.g., zap_pte_range() can simply zap swap and migration entries, wouldn't

Last but not least, I do wonder if mremap() -- e.g., move_pte() -- and fork() -- e.g., copy_pte_range() -- are handled properly, where we can end up moving/copying swap+migration PTEs around.

Signature · 3 lines
Thanks,

David / dhildenb
#7 lore raw

PATCH v2 2/15 mm/hugetlb: take src_mm->write_protect_seq in copy_hugetlb_page_range()

In reply to David Hildenbrand

Let's do it just like copy_page_range(), taking the seqlock and making sure the mmap_lock is held in write mode.

This allows for add a VM_BUG_ON to page_needs_cow_for_dma() and properly synchronizes cocnurrent fork() with GUP-fast of hugetlb pages, which will be relevant for further changes.

---
 include/linux/mm.h | 4 ++++
 mm/hugetlb.c       | 8 ++++++--
 2 files changed, 10 insertions(+), 2 deletions(-)
include/linux/mm.h +4 -0
index 9530f950f156..391b950e919d 100644
@@ -1323,6 +1323,8 @@ static inline bool is_cow_mapping(vm_flags_t flags)
 /*
  * This should most likely only be called during fork() to see whether we
  * should break the cow immediately for a page on the src mm.
+ *
+ * The caller has to hold the PT lock and the vma->vm_mm->->write_protect_seq.
  */
 static inline bool page_needs_cow_for_dma(struct vm_area_struct *vma,
 					  struct page *page)
@@ -1330,6 +1332,8 @@ static inline bool page_needs_cow_for_dma(struct vm_area_struct *vma,
 	if (!is_cow_mapping(vma->vm_flags))
 		return false;
 
+	VM_BUG_ON(!(raw_read_seqcount(&vma->vm_mm->write_protect_seq) & 1));
+
 	if (!test_bit(MMF_HAS_PINNED, &vma->vm_mm->flags))
 		return false;
 
mm/hugetlb.c +6 -2
index f294db835f4b..d3ce89697855 100644
@@ -4710,6 +4710,8 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 					vma->vm_start,
 					vma->vm_end);
 		mmu_notifier_invalidate_range_start(&range);
+		mmap_assert_write_locked(src);
+		raw_write_seqcount_begin(&src->write_protect_seq);
 	} else {
 		/*
 		 * For shared mappings i_mmap_rwsem must be held to call
@@ -4842,10 +4844,12 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 		spin_unlock(dst_ptl);
 	}
 
-	if (cow)
+	if (cow) {
+		raw_write_seqcount_end(&src->write_protect_seq);
 		mmu_notifier_invalidate_range_end(&range);
-	else
+	} else {
 		i_mmap_unlock_read(mapping);
+	}
 
 	return ret;
 }
Signature · 1 line
2.35.1
#8 lore raw

PATCH v2 3/15 mm/memory: slightly simplify copy_present_pte()

In reply to David Hildenbrand

Let's move the pinning check into the caller, to simplify return code logic and prepare for further changes: relocating the page_needs_cow_for_dma() into rmap handling code.

While at it, remove the unused pte parameter and simplify the comments a bit.

No functional change intended.

---
 mm/memory.c | 53 ++++++++++++++++-------------------------------------
 1 file changed, 16 insertions(+), 37 deletions(-)
mm/memory.c +16 -37
index c6177d897964..accb72a3343d 100644
@@ -865,19 +865,11 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 }
 
 /*
- * Copy a present and normal page if necessary.
+ * Copy a present and normal page.
  *
- * NOTE! The usual case is that this doesn't need to do
- * anything, and can just return a positive value. That
- * will let the caller know that it can just increase
- * the page refcount and re-use the pte the traditional
- * way.
- *
- * But _if_ we need to copy it because it needs to be
- * pinned in the parent (and the child should get its own
- * copy rather than just a reference to the same page),
- * we'll do that here and return zero to let the caller
- * know we're done.
+ * NOTE! The usual case is that this isn't required;
+ * instead, the caller can just increase the page refcount
+ * and re-use the pte the traditional way.
  *
  * And if we need a pre-allocated page but don't yet have
  * one, return a negative error to let the preallocation
@@ -887,25 +879,10 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 static inline int
 copy_present_page(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
 		  pte_t *dst_pte, pte_t *src_pte, unsigned long addr, int *rss,
-		  struct page **prealloc, pte_t pte, struct page *page)
+		  struct page **prealloc, struct page *page)
 {
 	struct page *new_page;
-
-	/*
-	 * What we want to do is to check whether this page may
-	 * have been pinned by the parent process.  If so,
-	 * instead of wrprotect the pte on both sides, we copy
-	 * the page immediately so that we'll always guarantee
-	 * the pinned page won't be randomly replaced in the
-	 * future.
-	 *
-	 * The page pinning checks are just "has this mm ever
-	 * seen pinning", along with the (inexact) check of
-	 * the page count. That might give false positives for
-	 * for pinning, but it will work correctly.
-	 */
-	if (likely(!page_needs_cow_for_dma(src_vma, page)))
-		return 1;
+	pte_t pte;
 
 	new_page = *prealloc;
 	if (!new_page)
@@ -947,14 +924,16 @@ copy_present_pte(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
 	struct page *page;
 
 	page = vm_normal_page(src_vma, addr, pte);
-	if (page) {
-		int retval;
-
-		retval = copy_present_page(dst_vma, src_vma, dst_pte, src_pte,
-					   addr, rss, prealloc, pte, page);
-		if (retval <= 0)
-			return retval;
-
+	if (page && unlikely(page_needs_cow_for_dma(src_vma, page))) {
+		/*
+		 * If this page may have been pinned by the parent process,
+		 * copy the page immediately for the child so that we'll always
+		 * guarantee the pinned page won't be randomly replaced in the
+		 * future.
+		 */
+		return copy_present_page(dst_vma, src_vma, dst_pte, src_pte,
+					 addr, rss, prealloc, page);
+	} else if (page) {
 		get_page(page);
 		page_dup_rmap(page, false);
 		rss[mm_counter(page)]++;
Signature · 1 line
2.35.1
#9 lore raw

PATCH v2 4/15 mm/rmap: split page_dup_rmap() into page_dup_file_rmap() and page_try_dup_anon_rmap()

In reply to David Hildenbrand

... and move the special check for pinned pages into page_try_dup_anon_rmap() to prepare for tracking exclusive anonymous pages via a new pageflag, clearing it only after making sure that there are no GUP pins on the anonymous page.

We really only care about pins on anonymous pages, because they are prone to getting replaced in the COW handler once mapped R/O. For !anon pages in cow-mappings (!VM_SHARED && VM_MAYWRITE) we shouldn't really care about that, at least not that I could come up with an example.

Let's drop the is_cow_mapping() check from page_needs_cow_for_dma(), as we know we're dealing with anonymous pages. Also, drop the handling of pinned pages from copy_huge_pud() and add a comment if ever supporting anonymous pages on the PUD level.

This is a preparation for tracking exclusivity of anonymous pages in the rmap code, and disallowing marking a page shared (-> failing to duplicate) if there are GUP pins on a page.

RFC notes: if I'm missing something important for !anon pages, we could
similarly handle it via page_try_dup_file_rmap().

---
 include/linux/mm.h   |  5 +----
 include/linux/rmap.h | 48 +++++++++++++++++++++++++++++++++++++++++++-
 mm/huge_memory.c     | 27 ++++++++-----------------
 mm/hugetlb.c         | 16 ++++++++-------
 mm/memory.c          | 17 +++++++++++-----
 mm/migrate.c         |  2 +-
 6 files changed, 78 insertions(+), 37 deletions(-)
include/linux/mm.h +1 -4
index 391b950e919d..63ee06001189 100644
@@ -1322,16 +1322,13 @@ static inline bool is_cow_mapping(vm_flags_t flags)
 
 /*
  * This should most likely only be called during fork() to see whether we
- * should break the cow immediately for a page on the src mm.
+ * should break the cow immediately for an anon page on the src mm.
  *
  * The caller has to hold the PT lock and the vma->vm_mm->->write_protect_seq.
  */
 static inline bool page_needs_cow_for_dma(struct vm_area_struct *vma,
 					  struct page *page)
 {
-	if (!is_cow_mapping(vma->vm_flags))
-		return false;
-
 	VM_BUG_ON(!(raw_read_seqcount(&vma->vm_mm->write_protect_seq) & 1));
 
 	if (!test_bit(MMF_HAS_PINNED, &vma->vm_mm->flags))
include/linux/rmap.h +47 -1

58 lines not on this page · raw message

mm/huge_memory.c +8 -19
index cda88d8ac1bd..c126d728b8de 100644
@@ -1097,23 +1097,16 @@ int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 	src_page = pmd_page(pmd);
 	VM_BUG_ON_PAGE(!PageHead(src_page), src_page);
 
-	/*
-	 * If this page is a potentially pinned page, split and retry the fault
-	 * with smaller page size.  Normally this should not happen because the
-	 * userspace should use MADV_DONTFORK upon pinned regions.  This is a
-	 * best effort that the pinned pages won't be replaced by another
-	 * random page during the coming copy-on-write.
-	 */
-	if (unlikely(page_needs_cow_for_dma(src_vma, src_page))) {
+	get_page(src_page);
+	if (unlikely(page_try_dup_anon_rmap(src_page, true, src_vma))) {
+		/* Page maybe pinned: split and retry the fault on PTEs. */
+		put_page(src_page);
 		pte_free(dst_mm, pgtable);
 		spin_unlock(src_ptl);
 		spin_unlock(dst_ptl);
 		__split_huge_pmd(src_vma, src_pmd, addr, false, NULL);
 		return -EAGAIN;
 	}
-
-	get_page(src_page);
-	page_dup_rmap(src_page, true);
 	add_mm_counter(dst_mm, MM_ANONPAGES, HPAGE_PMD_NR);
 out_zero_page:
 	mm_inc_nr_ptes(dst_mm);
@@ -1217,14 +1210,10 @@ int copy_huge_pud(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 		/* No huge zero pud yet */
 	}
 
-	/* Please refer to comments in copy_huge_pmd() */
-	if (unlikely(page_needs_cow_for_dma(vma, pud_page(pud)))) {
-		spin_unlock(src_ptl);
-		spin_unlock(dst_ptl);
-		__split_huge_pud(vma, src_pud, addr);
-		return -EAGAIN;
-	}
-
+	/*
+	 * TODO: once we support anonymous pages, use page_try_dup_anon_rmap()
+	 * and split if duplicating fails.
+	 */
 	pudp_set_wrprotect(src_mm, addr, src_pud);
 	pud = pud_mkold(pud_wrprotect(pud));
 	set_pud_at(dst_mm, addr, dst_pud, pud);
mm/hugetlb.c +9 -7
index d3ce89697855..9fb990d95dab 100644
@@ -4781,15 +4781,18 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 			get_page(ptepage);
 
 			/*
-			 * This is a rare case where we see pinned hugetlb
-			 * pages while they're prone to COW.  We need to do the
-			 * COW earlier during fork.
+			 * Failing to duplicate the anon rmap is a rare case
+			 * where we see pinned hugetlb pages while they're
+			 * prone to COW. We need to do the COW earlier during
+			 * fork.
 			 *
 			 * When pre-allocating the page or copying data, we
 			 * need to be without the pgtable locks since we could
 			 * sleep during the process.
 			 */
-			if (unlikely(page_needs_cow_for_dma(vma, ptepage))) {
+			if (!PageAnon(ptepage)) {
+				page_dup_file_rmap(ptepage, true);
+			} else if (page_try_dup_anon_rmap(ptepage, true, vma)) {
 				pte_t src_pte_old = entry;
 				struct page *new;
 
@@ -4836,7 +4839,6 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 				entry = huge_pte_wrprotect(entry);
 			}
 
-			page_dup_rmap(ptepage, true);
 			set_huge_pte_at(dst, addr, dst_pte, entry);
 			hugetlb_count_add(npages, dst);
 		}
@@ -5514,7 +5516,7 @@ static vm_fault_t hugetlb_no_page(struct mm_struct *mm,
 		ClearHPageRestoreReserve(page);
 		hugepage_add_new_anon_rmap(page, vma, haddr);
 	} else
-		page_dup_rmap(page, true);
+		page_dup_file_rmap(page, true);
 	new_pte = make_huge_pte(vma, page, ((vma->vm_flags & VM_WRITE)
 				&& (vma->vm_flags & VM_SHARED)));
 	set_huge_pte_at(mm, haddr, ptep, new_pte);
@@ -5874,7 +5876,7 @@ int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm,
 		goto out_release_unlock;
 
 	if (vm_shared) {
-		page_dup_rmap(page, true);
+		page_dup_file_rmap(page, true);
 	} else {
 		ClearHPageRestoreReserve(page);
 		hugepage_add_new_anon_rmap(page, dst_vma, dst_addr);
mm/memory.c +12 -5
index accb72a3343d..b9602d41d907 100644
@@ -828,7 +828,8 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 		 */
 		get_page(page);
 		rss[mm_counter(page)]++;
-		page_dup_rmap(page, false);
+		/* Cannot fail as these pages cannot get pinned. */
+		BUG_ON(page_try_dup_anon_rmap(page, false, src_vma));
 
 		/*
 		 * We do not preserve soft-dirty information, because so
@@ -924,18 +925,24 @@ copy_present_pte(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
 	struct page *page;
 
 	page = vm_normal_page(src_vma, addr, pte);
-	if (page && unlikely(page_needs_cow_for_dma(src_vma, page))) {
+	if (page && PageAnon(page)) {
 		/*
 		 * If this page may have been pinned by the parent process,
 		 * copy the page immediately for the child so that we'll always
 		 * guarantee the pinned page won't be randomly replaced in the
 		 * future.
 		 */
-		return copy_present_page(dst_vma, src_vma, dst_pte, src_pte,
-					 addr, rss, prealloc, page);
+		get_page(page);
+		if (unlikely(page_try_dup_anon_rmap(page, false, src_vma))) {
+			/* Page maybe pinned, we have to copy. */
+			put_page(page);
+			return copy_present_page(dst_vma, src_vma, dst_pte, src_pte,
+						 addr, rss, prealloc, page);
+		}
+		rss[mm_counter(page)]++;
 	} else if (page) {
 		get_page(page);
-		page_dup_rmap(page, false);
+		page_dup_file_rmap(page, false);
 		rss[mm_counter(page)]++;
 	}
 
mm/migrate.c +1 -1
index c7da064b4781..524c2648ab36 100644
@@ -240,7 +240,7 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 			if (PageAnon(new))
 				hugepage_add_anon_rmap(new, vma, pvmw.address);
 			else
-				page_dup_rmap(new, true);
+				page_dup_file_rmap(new, true);
 			set_huge_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
 		} else
 #endif
Signature · 1 line
2.35.1
#10 lore raw

Re: [PATCH v2 04/15] mm/rmap: split page_dup_rmap() into page_dup_file_rmap() and page_try_dup_anon_rmap()

Quoted 128 lines · David Hildenbrand

... and move the special check for pinned pages into page_try_dup_anon_rmap() to prepare for tracking exclusive anonymous pages via a new pageflag, clearing it only after making sure that there are no GUP pins on the anonymous page.

We really only care about pins on anonymous pages, because they are prone to getting replaced in the COW handler once mapped R/O. For !anon pages in cow-mappings (!VM_SHARED && VM_MAYWRITE) we shouldn't really care about that, at least not that I could come up with an example.

Let's drop the is_cow_mapping() check from page_needs_cow_for_dma(), as we know we're dealing with anonymous pages. Also, drop the handling of pinned pages from copy_huge_pud() and add a comment if ever supporting anonymous pages on the PUD level.

This is a preparation for tracking exclusivity of anonymous pages in the rmap code, and disallowing marking a page shared (-> failing to duplicate) if there are GUP pins on a page.

RFC notes: if I'm missing something important for !anon pages, we could
similarly handle it via page_try_dup_file_rmap().

---
 include/linux/mm.h   |  5 +----
 include/linux/rmap.h | 48 +++++++++++++++++++++++++++++++++++++++++++-
 mm/huge_memory.c     | 27 ++++++++-----------------
 mm/hugetlb.c         | 16 ++++++++-------
 mm/memory.c          | 17 +++++++++++-----
 mm/migrate.c         |  2 +-
 6 files changed, 78 insertions(+), 37 deletions(-)
include/linux/mm.h +1 -4
index 391b950e919d..63ee06001189 100644
@@ -1322,16 +1322,13 @@ static inline bool is_cow_mapping(vm_flags_t flags)
 
 /*
  * This should most likely only be called during fork() to see whether we
- * should break the cow immediately for a page on the src mm.
+ * should break the cow immediately for an anon page on the src mm.
  *
  * The caller has to hold the PT lock and the vma->vm_mm->->write_protect_seq.
  */
 static inline bool page_needs_cow_for_dma(struct vm_area_struct *vma,
                                          struct page *page)
 {
-       if (!is_cow_mapping(vma->vm_flags))
-               return false;
-
        VM_BUG_ON(!(raw_read_seqcount(&vma->vm_mm->write_protect_seq) & 1));
 
        if (!test_bit(MMF_HAS_PINNED, &vma->vm_mm->flags))
include/linux/rmap.h +47 -1
index e704b1a4c06c..92c3585b8c6a 100644
@@ -180,11 +180,57 @@ void hugepage_add_anon_rmap(struct page *, struct vm_area_struct *,
 void hugepage_add_new_anon_rmap(struct page *, struct vm_area_struct *,
                                unsigned long);
 
-static inline void page_dup_rmap(struct page *page, bool compound)
+static inline void __page_dup_rmap(struct page *page, bool compound)
 {
        atomic_inc(compound ? compound_mapcount_ptr(page) : &page->_mapcount);
 }
 
+static inline void page_dup_file_rmap(struct page *page, bool compound)
+{
+       __page_dup_rmap(page, compound);
+}
+
+/**
+ * page_try_dup_anon_rmap - try duplicating a mapping of an already mapped
+ *                         anonymous page
+ * @page: the page to duplicate the mapping for
+ * @compound: the page is mapped as compound or as a small page
+ * @vma: the source vma
+ *
+ * The caller needs to hold the PT lock and the vma->vma_mm->write_protect_seq.
+ *
+ * Duplicating the mapping can only fail if the page may be pinned; device
+ * private pages cannot get pinned and consequently this function cannot fail.
+ *
+ * If duplicating the mapping succeeds, the page has to be mapped R/O into
+ * the parent and the child. It must *not* get mapped writable after this call.
+ *
+ * Returns 0 if duplicating the mapping succeeded. Returns -EBUSY otherwise.
+ */
+static inline int page_try_dup_anon_rmap(struct page *page, bool compound,
+                                        struct vm_area_struct *vma)
+{
+       VM_BUG_ON_PAGE(!PageAnon(page), page);
+
+       /*
+        * If this page may have been pinned by the parent process,
+        * don't allow to duplicate the mapping but instead require to e.g.,
+        * copy the page immediately for the child so that we'll always
+        * guarantee the pinned page won't be randomly replaced in the
+        * future on write faults.
+        */
+       if (likely(!is_device_private_page(page) &&
+           unlikely(page_needs_cow_for_dma(vma, page))))
+               return -EBUSY;
+
+       /*
+        * It's okay to share the anon page between both processes, mapping
+        * the page R/O into both processes.
+        */
+       __page_dup_rmap(page, compound);
+       return 0;
+}
+
 /*
  * Called from mm/vmscan.c to handle paging out
  */
mm/huge_memory.c +0 -2
index cda88d8ac1bd..c126d728b8de 100644
@@ -1097,23 +1097,16 @@ int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm,
        src_page = pmd_page(pmd);
        VM_BUG_ON_PAGE(!PageHead(src_page), src_page);
 
-       /*
-        * If this page is a potentially pinned page, split and retry the fault

mm/huge_memory.c int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm, · line 1103

Quoted 10 lines
mm/huge_memory.c +4 -6
-        * with smaller page size.  Normally this should not happen because the
-        * userspace should use MADV_DONTFORK upon pinned regions.  This is a
-        * best effort that the pinned pages won't be replaced by another
-        * random page during the coming copy-on-write.
-        */
-       if (unlikely(page_needs_cow_for_dma(src_vma, src_page))) {
+       get_page(src_page);
+       if (unlikely(page_try_dup_anon_rmap(src_page, true, src_vma))) {
+               /* Page maybe pinned: split and retry the fault on PTEs. */
+               put_page(src_page);

Do we have to do the get_page()/put_page() sequence? It seems we don't have to get the page before calling page_try_dup_anon_rmap(), right? We just need to bump page refcount when dupping rmap successfully. So we could do:

if (unlikely(page_try_dup_anon_rmap(src_page, true, src_vma))) {
    /* do something */
}

get_page(src_page)

Quoted 143 lines
mm/huge_memory.c +4 -11
                pte_free(dst_mm, pgtable);
                spin_unlock(src_ptl);
                spin_unlock(dst_ptl);
                __split_huge_pmd(src_vma, src_pmd, addr, false, NULL);
                return -EAGAIN;
        }
-
-       get_page(src_page);
-       page_dup_rmap(src_page, true);
        add_mm_counter(dst_mm, MM_ANONPAGES, HPAGE_PMD_NR);
 out_zero_page:
        mm_inc_nr_ptes(dst_mm);
@@ -1217,14 +1210,10 @@ int copy_huge_pud(struct mm_struct *dst_mm, struct mm_struct *src_mm,
                /* No huge zero pud yet */
        }
 
-       /* Please refer to comments in copy_huge_pmd() */
-       if (unlikely(page_needs_cow_for_dma(vma, pud_page(pud)))) {
-               spin_unlock(src_ptl);
-               spin_unlock(dst_ptl);
-               __split_huge_pud(vma, src_pud, addr);
-               return -EAGAIN;
-       }
-
+       /*
+        * TODO: once we support anonymous pages, use page_try_dup_anon_rmap()
+        * and split if duplicating fails.
+        */
        pudp_set_wrprotect(src_mm, addr, src_pud);
        pud = pud_mkold(pud_wrprotect(pud));
        set_pud_at(dst_mm, addr, dst_pud, pud);
mm/hugetlb.c +9 -7
index d3ce89697855..9fb990d95dab 100644
@@ -4781,15 +4781,18 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
                        get_page(ptepage);
 
                        /*
-                        * This is a rare case where we see pinned hugetlb
-                        * pages while they're prone to COW.  We need to do the
-                        * COW earlier during fork.
+                        * Failing to duplicate the anon rmap is a rare case
+                        * where we see pinned hugetlb pages while they're
+                        * prone to COW. We need to do the COW earlier during
+                        * fork.
                         *
                         * When pre-allocating the page or copying data, we
                         * need to be without the pgtable locks since we could
                         * sleep during the process.
                         */
-                       if (unlikely(page_needs_cow_for_dma(vma, ptepage))) {
+                       if (!PageAnon(ptepage)) {
+                               page_dup_file_rmap(ptepage, true);
+                       } else if (page_try_dup_anon_rmap(ptepage, true, vma)) {
                                pte_t src_pte_old = entry;
                                struct page *new;
 
@@ -4836,7 +4839,6 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
                                entry = huge_pte_wrprotect(entry);
                        }
 
-                       page_dup_rmap(ptepage, true);
                        set_huge_pte_at(dst, addr, dst_pte, entry);
                        hugetlb_count_add(npages, dst);
                }
@@ -5514,7 +5516,7 @@ static vm_fault_t hugetlb_no_page(struct mm_struct *mm,
                ClearHPageRestoreReserve(page);
                hugepage_add_new_anon_rmap(page, vma, haddr);
        } else
-               page_dup_rmap(page, true);
+               page_dup_file_rmap(page, true);
        new_pte = make_huge_pte(vma, page, ((vma->vm_flags & VM_WRITE)
                                && (vma->vm_flags & VM_SHARED)));
        set_huge_pte_at(mm, haddr, ptep, new_pte);
@@ -5874,7 +5876,7 @@ int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm,
                goto out_release_unlock;
 
        if (vm_shared) {
-               page_dup_rmap(page, true);
+               page_dup_file_rmap(page, true);
        } else {
                ClearHPageRestoreReserve(page);
                hugepage_add_new_anon_rmap(page, dst_vma, dst_addr);
mm/memory.c +12 -5
index accb72a3343d..b9602d41d907 100644
@@ -828,7 +828,8 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
                 */
                get_page(page);
                rss[mm_counter(page)]++;
-               page_dup_rmap(page, false);
+               /* Cannot fail as these pages cannot get pinned. */
+               BUG_ON(page_try_dup_anon_rmap(page, false, src_vma));
 
                /*
                 * We do not preserve soft-dirty information, because so
@@ -924,18 +925,24 @@ copy_present_pte(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
        struct page *page;
 
        page = vm_normal_page(src_vma, addr, pte);
-       if (page && unlikely(page_needs_cow_for_dma(src_vma, page))) {
+       if (page && PageAnon(page)) {
                /*
                 * If this page may have been pinned by the parent process,
                 * copy the page immediately for the child so that we'll always
                 * guarantee the pinned page won't be randomly replaced in the
                 * future.
                 */
-               return copy_present_page(dst_vma, src_vma, dst_pte, src_pte,
-                                        addr, rss, prealloc, page);
+               get_page(page);
+               if (unlikely(page_try_dup_anon_rmap(page, false, src_vma))) {
+                       /* Page maybe pinned, we have to copy. */
+                       put_page(page);
+                       return copy_present_page(dst_vma, src_vma, dst_pte, src_pte,
+                                                addr, rss, prealloc, page);
+               }
+               rss[mm_counter(page)]++;
        } else if (page) {
                get_page(page);
-               page_dup_rmap(page, false);
+               page_dup_file_rmap(page, false);
                rss[mm_counter(page)]++;
        }
 
mm/migrate.c +1 -2
index c7da064b4781..524c2648ab36 100644
@@ -240,7 +240,7 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
                        if (PageAnon(new))
                                hugepage_add_anon_rmap(new, vma, pvmw.address);
                        else
-                               page_dup_rmap(new, true);
+                               page_dup_file_rmap(new, true);
                        set_huge_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
                } else
 #endif
--

2.35.1

#11 lore raw

Re: [PATCH v2 04/15] mm/rmap: split page_dup_rmap() into page_dup_file_rmap() and page_try_dup_anon_rmap()

Quoted 145 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

... and move the special check for pinned pages into page_try_dup_anon_rmap() to prepare for tracking exclusive anonymous pages via a new pageflag, clearing it only after making sure that there are no GUP pins on the anonymous page.

We really only care about pins on anonymous pages, because they are prone to getting replaced in the COW handler once mapped R/O. For !anon pages in cow-mappings (!VM_SHARED && VM_MAYWRITE) we shouldn't really care about that, at least not that I could come up with an example.

Let's drop the is_cow_mapping() check from page_needs_cow_for_dma(), as we know we're dealing with anonymous pages. Also, drop the handling of pinned pages from copy_huge_pud() and add a comment if ever supporting anonymous pages on the PUD level.

This is a preparation for tracking exclusivity of anonymous pages in the rmap code, and disallowing marking a page shared (-> failing to duplicate) if there are GUP pins on a page.

RFC notes: if I'm missing something important for !anon pages, we could
similarly handle it via page_try_dup_file_rmap().

---
 include/linux/mm.h   |  5 +----
 include/linux/rmap.h | 48 +++++++++++++++++++++++++++++++++++++++++++-
 mm/huge_memory.c     | 27 ++++++++-----------------
 mm/hugetlb.c         | 16 ++++++++-------
 mm/memory.c          | 17 +++++++++++-----
 mm/migrate.c         |  2 +-
 6 files changed, 78 insertions(+), 37 deletions(-)
include/linux/mm.h +1 -4
index 391b950e919d..63ee06001189 100644
@@ -1322,16 +1322,13 @@ static inline bool is_cow_mapping(vm_flags_t flags)
 
 /*
  * This should most likely only be called during fork() to see whether we
- * should break the cow immediately for a page on the src mm.
+ * should break the cow immediately for an anon page on the src mm.
  *
  * The caller has to hold the PT lock and the vma->vm_mm->->write_protect_seq.
  */
 static inline bool page_needs_cow_for_dma(struct vm_area_struct *vma,
                                          struct page *page)
 {
-       if (!is_cow_mapping(vma->vm_flags))
-               return false;
-
        VM_BUG_ON(!(raw_read_seqcount(&vma->vm_mm->write_protect_seq) & 1));
 
        if (!test_bit(MMF_HAS_PINNED, &vma->vm_mm->flags))
include/linux/rmap.h +47 -1
index e704b1a4c06c..92c3585b8c6a 100644
@@ -180,11 +180,57 @@ void hugepage_add_anon_rmap(struct page *, struct vm_area_struct *,
 void hugepage_add_new_anon_rmap(struct page *, struct vm_area_struct *,
                                unsigned long);
 
-static inline void page_dup_rmap(struct page *page, bool compound)
+static inline void __page_dup_rmap(struct page *page, bool compound)
 {
        atomic_inc(compound ? compound_mapcount_ptr(page) : &page->_mapcount);
 }
 
+static inline void page_dup_file_rmap(struct page *page, bool compound)
+{
+       __page_dup_rmap(page, compound);
+}
+
+/**
+ * page_try_dup_anon_rmap - try duplicating a mapping of an already mapped
+ *                         anonymous page
+ * @page: the page to duplicate the mapping for
+ * @compound: the page is mapped as compound or as a small page
+ * @vma: the source vma
+ *
+ * The caller needs to hold the PT lock and the vma->vma_mm->write_protect_seq.
+ *
+ * Duplicating the mapping can only fail if the page may be pinned; device
+ * private pages cannot get pinned and consequently this function cannot fail.
+ *
+ * If duplicating the mapping succeeds, the page has to be mapped R/O into
+ * the parent and the child. It must *not* get mapped writable after this call.
+ *
+ * Returns 0 if duplicating the mapping succeeded. Returns -EBUSY otherwise.
+ */
+static inline int page_try_dup_anon_rmap(struct page *page, bool compound,
+                                        struct vm_area_struct *vma)
+{
+       VM_BUG_ON_PAGE(!PageAnon(page), page);
+
+       /*
+        * If this page may have been pinned by the parent process,
+        * don't allow to duplicate the mapping but instead require to e.g.,
+        * copy the page immediately for the child so that we'll always
+        * guarantee the pinned page won't be randomly replaced in the
+        * future on write faults.
+        */
+       if (likely(!is_device_private_page(page) &&
+           unlikely(page_needs_cow_for_dma(vma, page))))
+               return -EBUSY;
+
+       /*
+        * It's okay to share the anon page between both processes, mapping
+        * the page R/O into both processes.
+        */
+       __page_dup_rmap(page, compound);
+       return 0;
+}
+
 /*
  * Called from mm/vmscan.c to handle paging out
  */
mm/huge_memory.c +0 -2
index cda88d8ac1bd..c126d728b8de 100644
@@ -1097,23 +1097,16 @@ int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm,
        src_page = pmd_page(pmd);
        VM_BUG_ON_PAGE(!PageHead(src_page), src_page);
 
-       /*
-        * If this page is a potentially pinned page, split and retry the fault

mm/huge_memory.c int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm, · line 1103

Quoted 10 lines
mm/huge_memory.c +4 -6
-        * with smaller page size.  Normally this should not happen because the
-        * userspace should use MADV_DONTFORK upon pinned regions.  This is a
-        * best effort that the pinned pages won't be replaced by another
-        * random page during the coming copy-on-write.
-        */
-       if (unlikely(page_needs_cow_for_dma(src_vma, src_page))) {
+       get_page(src_page);
+       if (unlikely(page_try_dup_anon_rmap(src_page, true, src_vma))) {
+               /* Page maybe pinned: split and retry the fault on PTEs. */
+               put_page(src_page);

Do we have to do the get_page()/put_page() sequence? It seems we don't have to get the page before calling page_try_dup_anon_rmap(), right? We just need to bump page refcount when dupping rmap successfully. So we could do:

Quoted 5 lines
if (unlikely(page_try_dup_anon_rmap(src_page, true, src_vma))) {
    /* do something */
}

get_page(src_page)

I refrained from doing that in this series because I wasn't able to convince myself that it is completely sane.

Increasing the refcount before the mapcount here results in an interesting property (when unmap is done similarly, in reverse order):
reading a snapshot of page_count()+mapcount() at a random point in time guarantees that page_count() >= mapcount.

Now, the issue I see with changing that is the following: We seem to have some code that tries detecting additional references on a mapped page. While such checks can deal with false positives ("there is an additional reference"), false negatives ("these is no additional reference") could be bad.

One example I stumbled over is KSM code:

		/*
		 * Check that no O_DIRECT or similar I/O is in progress on the
		 * page
		 */
		if (page_mapcount(page) + 1 + swapped != page_count(page)) {
			set_pte_at(mm, pvmw.address, pvmw.pte, entry);
			goto out_unlock;
		}

Imagine there is an additional reference but another process sharing the page did the mapcount++ but not the refcount++ yet. You'd get a false negative.

Now, I'm not sure if there is anything else protecting us from this, however, it rather made me be very careful about changing the order of things here. So in the context of this patch, I won't be messing with the order.

Thanks!

Signature · 3 lines
Thanks,

David / dhildenb
#12 lore raw

PATCH v2 5/15 mm/rmap: convert RMAP flags to a proper distinct rmap_t type

In reply to David Hildenbrand

We want to pass the flags to more than one anon rmap function, getting rid of special "do_page_add_anon_rmap()". So let's pass around a distinct __bitwise type and refine documentation.

---
 include/linux/rmap.h | 22 ++++++++++++++++++----
 mm/memory.c          |  6 +++---
 mm/rmap.c            |  7 ++++---
 3 files changed, 25 insertions(+), 10 deletions(-)
include/linux/rmap.h +18 -4
index 92c3585b8c6a..49f6b208938c 100644
@@ -158,9 +158,23 @@ static inline void anon_vma_merge(struct vm_area_struct *vma,
 
 struct anon_vma *page_get_anon_vma(struct page *page);
 
-/* bitflags for do_page_add_anon_rmap() */
-#define RMAP_EXCLUSIVE 0x01
-#define RMAP_COMPOUND 0x02
+/* RMAP flags, currently only relevant for some anon rmap operations. */
+typedef int __bitwise rmap_t;
+
+/*
+ * No special request: if the page is a subpage of a compound page, it is
+ * mapped via a PTE. The mapped (sub)page is possibly shared between processes.
+ */
+#define RMAP_NONE		((__force rmap_t)0)
+
+/* The (sub)page is exclusive to a single process. */
+#define RMAP_EXCLUSIVE		((__force rmap_t)BIT(0))
+
+/*
+ * The compound page is not mapped via PTEs, but instead via a single PMD and
+ * should be accounted accordingly.
+ */
+#define RMAP_COMPOUND		((__force rmap_t)BIT(1))
 
 /*
  * rmap interfaces called when adding or removing pte of page
@@ -169,7 +183,7 @@ void page_move_anon_rmap(struct page *, struct vm_area_struct *);
 void page_add_anon_rmap(struct page *, struct vm_area_struct *,
 		unsigned long, bool);
 void do_page_add_anon_rmap(struct page *, struct vm_area_struct *,
-			   unsigned long, int);
+			   unsigned long, rmap_t);
 void page_add_new_anon_rmap(struct page *, struct vm_area_struct *,
 		unsigned long, bool);
 void page_add_file_rmap(struct page *, bool);
mm/memory.c +3 -3
index b9602d41d907..bbce3ca72974 100644
@@ -3515,10 +3515,10 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 	struct vm_area_struct *vma = vmf->vma;
 	struct page *page = NULL, *swapcache;
 	struct swap_info_struct *si = NULL;
+	rmap_t rmap_flags = RMAP_NONE;
 	swp_entry_t entry;
 	pte_t pte;
 	int locked;
-	int exclusive = 0;
 	vm_fault_t ret = 0;
 	void *shadow = NULL;
 
@@ -3693,7 +3693,7 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 		pte = maybe_mkwrite(pte_mkdirty(pte), vma);
 		vmf->flags &= ~FAULT_FLAG_WRITE;
 		ret |= VM_FAULT_WRITE;
-		exclusive = RMAP_EXCLUSIVE;
+		rmap_flags |= RMAP_EXCLUSIVE;
 	}
 	flush_icache_page(vma, page);
 	if (pte_swp_soft_dirty(vmf->orig_pte))
@@ -3709,7 +3709,7 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 		page_add_new_anon_rmap(page, vma, vmf->address, false);
 		lru_cache_add_inactive_or_unevictable(page, vma);
 	} else {
-		do_page_add_anon_rmap(page, vma, vmf->address, exclusive);
+		do_page_add_anon_rmap(page, vma, vmf->address, rmap_flags);
 	}
 
 	set_pte_at(vma->vm_mm, vmf->address, vmf->pte, pte);
mm/rmap.c +4 -3
index f825aeef61ca..3d7028d100ea 100644
@@ -1139,7 +1139,8 @@ static void __page_check_anon_rmap(struct page *page,
 void page_add_anon_rmap(struct page *page,
 	struct vm_area_struct *vma, unsigned long address, bool compound)
 {
-	do_page_add_anon_rmap(page, vma, address, compound ? RMAP_COMPOUND : 0);
+	do_page_add_anon_rmap(page, vma, address,
+			      compound ? RMAP_COMPOUND : RMAP_NONE);
 }
 
 /*
@@ -1148,7 +1149,7 @@ void page_add_anon_rmap(struct page *page,
  * Everybody else should continue to use page_add_anon_rmap above.
  */
 void do_page_add_anon_rmap(struct page *page,
-	struct vm_area_struct *vma, unsigned long address, int flags)
+	struct vm_area_struct *vma, unsigned long address, rmap_t flags)
 {
 	bool compound = flags & RMAP_COMPOUND;
 	bool first;
@@ -1189,7 +1190,7 @@ void do_page_add_anon_rmap(struct page *page,
 	/* address might be in next vma when migration races vma_adjust */
 	if (first)
 		__page_set_anon_rmap(page, vma, address,
-				flags & RMAP_EXCLUSIVE);
+				     !!(flags & RMAP_EXCLUSIVE));
 	else
 		__page_check_anon_rmap(page, vma, address);
 }
Signature · 1 line
2.35.1
#13 lore raw

PATCH v2 6/15 mm/rmap: remove do_page_add_anon_rmap()

In reply to David Hildenbrand

... and instead convert page_add_anon_rmap() to accept flags.

Passing flags instead of bools is usually nicer either way, and we want to more often also pass RMAP_EXCLUSIVE in follow up patches when detecting that an anonymous page is exclusive: for example, when restoring an anonymous page from a writable migration entry.

This is a preparation for marking an anonymous page inside page_add_anon_rmap() as exclusive when RMAP_EXCLUSIVE is passed.

---
 include/linux/rmap.h |  4 +---
 mm/huge_memory.c     |  2 +-
 mm/ksm.c             |  2 +-
 mm/memory.c          |  2 +-
 mm/migrate.c         |  3 ++-
 mm/rmap.c            | 14 +-------------
 mm/swapfile.c        |  2 +-
 7 files changed, 8 insertions(+), 21 deletions(-)
include/linux/rmap.h +1 -3
index 49f6b208938c..a269f80fbeef 100644
@@ -181,9 +181,7 @@ typedef int __bitwise rmap_t;
  */
 void page_move_anon_rmap(struct page *, struct vm_area_struct *);
 void page_add_anon_rmap(struct page *, struct vm_area_struct *,
-		unsigned long, bool);
-void do_page_add_anon_rmap(struct page *, struct vm_area_struct *,
-			   unsigned long, rmap_t);
+		unsigned long, rmap_t);
 void page_add_new_anon_rmap(struct page *, struct vm_area_struct *,
 		unsigned long, bool);
 void page_add_file_rmap(struct page *, bool);
mm/huge_memory.c +1 -1
index c126d728b8de..2ca137e01e84 100644
@@ -3115,7 +3115,7 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new)
 
 	flush_cache_range(vma, mmun_start, mmun_start + HPAGE_PMD_SIZE);
 	if (PageAnon(new))
-		page_add_anon_rmap(new, vma, mmun_start, true);
+		page_add_anon_rmap(new, vma, mmun_start, RMAP_COMPOUND);
 	else
 		page_add_file_rmap(new, true);
 	set_pmd_at(mm, mmun_start, pvmw->pmd, pmde);
mm/ksm.c +1 -1
index c20bd4d9a0d9..9ff28097bc0a 100644
@@ -1153,7 +1153,7 @@ static int replace_page(struct vm_area_struct *vma, struct page *page,
 	 */
 	if (!is_zero_pfn(page_to_pfn(kpage))) {
 		get_page(kpage);
-		page_add_anon_rmap(kpage, vma, addr, false);
+		page_add_anon_rmap(kpage, vma, addr, RMAP_NONE);
 		newpte = mk_pte(kpage, vma->vm_page_prot);
 	} else {
 		newpte = pte_mkspecial(pfn_pte(page_to_pfn(kpage),
mm/memory.c +1 -1
index bbce3ca72974..e2d8e55c55c0 100644
@@ -3709,7 +3709,7 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 		page_add_new_anon_rmap(page, vma, vmf->address, false);
 		lru_cache_add_inactive_or_unevictable(page, vma);
 	} else {
-		do_page_add_anon_rmap(page, vma, vmf->address, rmap_flags);
+		page_add_anon_rmap(page, vma, vmf->address, rmap_flags);
 	}
 
 	set_pte_at(vma->vm_mm, vmf->address, vmf->pte, pte);
mm/migrate.c +2 -1
index 524c2648ab36..907052b72581 100644
@@ -246,7 +246,8 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 #endif
 		{
 			if (PageAnon(new))
-				page_add_anon_rmap(new, vma, pvmw.address, false);
+				page_add_anon_rmap(new, vma, pvmw.address,
+						   RMAP_NONE);
 			else
 				page_add_file_rmap(new, false);
 			set_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
mm/rmap.c +1 -13
index 3d7028d100ea..97bcc50c3489 100644
@@ -1129,7 +1129,7 @@ static void __page_check_anon_rmap(struct page *page,
  * @page:	the page to add the mapping to
  * @vma:	the vm area in which the mapping is added
  * @address:	the user virtual address mapped
- * @compound:	charge the page as compound or small page
+ * @flags:	the rmap flags
  *
  * The caller needs to hold the pte lock, and the page must be locked in
  * the anon_vma case: to serialize mapping,index checking after setting,
@@ -1137,18 +1137,6 @@ static void __page_check_anon_rmap(struct page *page,
  * (but PageKsm is never downgraded to PageAnon).
  */
 void page_add_anon_rmap(struct page *page,
-	struct vm_area_struct *vma, unsigned long address, bool compound)
-{
-	do_page_add_anon_rmap(page, vma, address,
-			      compound ? RMAP_COMPOUND : RMAP_NONE);
-}
-
-/*
- * Special version of the above for do_swap_page, which often runs
- * into pages that are exclusively owned by the current process.
- * Everybody else should continue to use page_add_anon_rmap above.
- */
-void do_page_add_anon_rmap(struct page *page,
 	struct vm_area_struct *vma, unsigned long address, rmap_t flags)
 {
 	bool compound = flags & RMAP_COMPOUND;
mm/swapfile.c +1 -1
index a5183315dc58..41ba8238d16b 100644
@@ -1800,7 +1800,7 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 	inc_mm_counter(vma->vm_mm, MM_ANONPAGES);
 	get_page(page);
 	if (page == swapcache) {
-		page_add_anon_rmap(page, vma, addr, false);
+		page_add_anon_rmap(page, vma, addr, RMAP_NONE);
 	} else { /* ksm created a completely new copy */
 		page_add_new_anon_rmap(page, vma, addr, false);
 		lru_cache_add_inactive_or_unevictable(page, vma);
Signature · 1 line
2.35.1
#14 lore raw

PATCH v2 7/15 mm/rmap: pass rmap flags to hugepage_add_anon_rmap()

In reply to David Hildenbrand

Let's prepare for passing RMAP_EXCLUSIVE, similarly as we do for page_add_anon_rmap() now. RMAP_COMPOUND is implicit for hugetlb pages and ignored.

---
 include/linux/rmap.h | 2 +-
 mm/migrate.c         | 3 ++-
 mm/rmap.c            | 9 ++++++---
 3 files changed, 9 insertions(+), 5 deletions(-)
include/linux/rmap.h +1 -1
index a269f80fbeef..94ee38829c63 100644
@@ -188,7 +188,7 @@ void page_add_file_rmap(struct page *, bool);
 void page_remove_rmap(struct page *, bool);
 
 void hugepage_add_anon_rmap(struct page *, struct vm_area_struct *,
-			    unsigned long);
+			    unsigned long, rmap_t);
 void hugepage_add_new_anon_rmap(struct page *, struct vm_area_struct *,
 				unsigned long);
 
mm/migrate.c +2 -1
index 907052b72581..e6b3cb3d148b 100644
@@ -238,7 +238,8 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 			pte = pte_mkhuge(pte);
 			pte = arch_make_huge_pte(pte, shift, vma->vm_flags);
 			if (PageAnon(new))
-				hugepage_add_anon_rmap(new, vma, pvmw.address);
+				hugepage_add_anon_rmap(new, vma, pvmw.address,
+						       RMAP_NONE);
 			else
 				page_dup_file_rmap(new, true);
 			set_huge_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
mm/rmap.c +6 -3
index 97bcc50c3489..7162689203fc 100644
@@ -2409,9 +2409,11 @@ void rmap_walk_locked(struct page *page, struct rmap_walk_control *rwc)
  * The following two functions are for anonymous (private mapped) hugepages.
  * Unlike common anonymous pages, anonymous hugepages have no accounting code
  * and no lru code, because we handle hugepages differently from common pages.
+ *
+ * RMAP_COMPOUND is ignored.
  */
-void hugepage_add_anon_rmap(struct page *page,
-			    struct vm_area_struct *vma, unsigned long address)
+void hugepage_add_anon_rmap(struct page *page, struct vm_area_struct *vma,
+			    unsigned long address, rmap_t flags)
 {
 	struct anon_vma *anon_vma = vma->anon_vma;
 	int first;
@@ -2421,7 +2423,8 @@ void hugepage_add_anon_rmap(struct page *page,
 	/* address might be in next vma when migration races vma_adjust */
 	first = atomic_inc_and_test(compound_mapcount_ptr(page));
 	if (first)
-		__page_set_anon_rmap(page, vma, address, 0);
+		__page_set_anon_rmap(page, vma, address,
+				     flags & RMAP_EXCLUSIVE);
 }
 
 void hugepage_add_new_anon_rmap(struct page *page,
Signature · 1 line
2.35.1
#15 lore raw

PATCH v2 8/15 mm/rmap: drop "compound" parameter from page_add_new_anon_rmap()

In reply to David Hildenbrand

New anonymous pages are always mapped natively: only THP/khugepagd code maps a new compound anonymous page and passes "true". Otherwise, we're just dealing with simple, non-compound pages.

Let's give the interface clearer semantics and document these.

---
 include/linux/rmap.h    |  2 +-
 kernel/events/uprobes.c |  2 +-
 mm/huge_memory.c        |  2 +-
 mm/khugepaged.c         |  2 +-
 mm/memory.c             | 10 +++++-----
 mm/migrate.c            |  2 +-
 mm/rmap.c               |  9 ++++++---
 mm/swapfile.c           |  2 +-
 mm/userfaultfd.c        |  2 +-
 9 files changed, 18 insertions(+), 15 deletions(-)
include/linux/rmap.h +1 -1
index 94ee38829c63..51953bace0a3 100644
@@ -183,7 +183,7 @@ void page_move_anon_rmap(struct page *, struct vm_area_struct *);
 void page_add_anon_rmap(struct page *, struct vm_area_struct *,
 		unsigned long, rmap_t);
 void page_add_new_anon_rmap(struct page *, struct vm_area_struct *,
-		unsigned long, bool);
+		unsigned long);
 void page_add_file_rmap(struct page *, bool);
 void page_remove_rmap(struct page *, bool);
 
kernel/events/uprobes.c +1 -1
index 6357c3580d07..b6fdb23fb3ea 100644
@@ -184,7 +184,7 @@ static int __replace_page(struct vm_area_struct *vma, unsigned long addr,
 
 	if (new_page) {
 		get_page(new_page);
-		page_add_new_anon_rmap(new_page, vma, addr, false);
+		page_add_new_anon_rmap(new_page, vma, addr);
 		lru_cache_add_inactive_or_unevictable(new_page, vma);
 	} else
 		/* no new page, just dec_mm_counter for old_page */
mm/huge_memory.c +1 -1
index 2ca137e01e84..c1f7eaba23ff 100644
@@ -647,7 +647,7 @@ static vm_fault_t __do_huge_pmd_anonymous_page(struct vm_fault *vmf,
 
 		entry = mk_huge_pmd(page, vma->vm_page_prot);
 		entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma);
-		page_add_new_anon_rmap(page, vma, haddr, true);
+		page_add_new_anon_rmap(page, vma, haddr);
 		lru_cache_add_inactive_or_unevictable(page, vma);
 		pgtable_trans_huge_deposit(vma->vm_mm, vmf->pmd, pgtable);
 		set_pmd_at(vma->vm_mm, haddr, vmf->pmd, entry);
mm/khugepaged.c +1 -1
index a325a646be33..96cc903c4788 100644
@@ -1183,7 +1183,7 @@ static void collapse_huge_page(struct mm_struct *mm,
 
 	spin_lock(pmd_ptl);
 	BUG_ON(!pmd_none(*pmd));
-	page_add_new_anon_rmap(new_page, vma, address, true);
+	page_add_new_anon_rmap(new_page, vma, address);
 	lru_cache_add_inactive_or_unevictable(new_page, vma);
 	pgtable_trans_huge_deposit(mm, pmd, pgtable);
 	set_pmd_at(mm, address, pmd, _pmd);
mm/memory.c +5 -5
index e2d8e55c55c0..00c45b3a9576 100644
@@ -896,7 +896,7 @@ copy_present_page(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma
 	*prealloc = NULL;
 	copy_user_highpage(new_page, page, addr, src_vma);
 	__SetPageUptodate(new_page);
-	page_add_new_anon_rmap(new_page, dst_vma, addr, false);
+	page_add_new_anon_rmap(new_page, dst_vma, addr);
 	lru_cache_add_inactive_or_unevictable(new_page, dst_vma);
 	rss[mm_counter(new_page)]++;
 
@@ -3052,7 +3052,7 @@ static vm_fault_t wp_page_copy(struct vm_fault *vmf)
 		 * some TLBs while the old PTE remains in others.
 		 */
 		ptep_clear_flush_notify(vma, vmf->address, vmf->pte);
-		page_add_new_anon_rmap(new_page, vma, vmf->address, false);
+		page_add_new_anon_rmap(new_page, vma, vmf->address);
 		lru_cache_add_inactive_or_unevictable(new_page, vma);
 		/*
 		 * We call the notify macro here because, when using secondary
@@ -3706,7 +3706,7 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 
 	/* ksm created a completely new copy */
 	if (unlikely(page != swapcache && swapcache)) {
-		page_add_new_anon_rmap(page, vma, vmf->address, false);
+		page_add_new_anon_rmap(page, vma, vmf->address);
 		lru_cache_add_inactive_or_unevictable(page, vma);
 	} else {
 		page_add_anon_rmap(page, vma, vmf->address, rmap_flags);
@@ -3856,7 +3856,7 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
 	}
 
 	inc_mm_counter_fast(vma->vm_mm, MM_ANONPAGES);
-	page_add_new_anon_rmap(page, vma, vmf->address, false);
+	page_add_new_anon_rmap(page, vma, vmf->address);
 	lru_cache_add_inactive_or_unevictable(page, vma);
 setpte:
 	set_pte_at(vma->vm_mm, vmf->address, vmf->pte, entry);
@@ -4033,7 +4033,7 @@ void do_set_pte(struct vm_fault *vmf, struct page *page, unsigned long addr)
 	/* copy-on-write page */
 	if (write && !(vma->vm_flags & VM_SHARED)) {
 		inc_mm_counter_fast(vma->vm_mm, MM_ANONPAGES);
-		page_add_new_anon_rmap(page, vma, addr, false);
+		page_add_new_anon_rmap(page, vma, addr);
 		lru_cache_add_inactive_or_unevictable(page, vma);
 	} else {
 		inc_mm_counter_fast(vma->vm_mm, mm_counter_file(page));
mm/migrate.c +1 -1
index e6b3cb3d148b..fd9eba33b34a 100644
@@ -2725,7 +2725,7 @@ static void migrate_vma_insert_page(struct migrate_vma *migrate,
 		goto unlock_abort;
 
 	inc_mm_counter(mm, MM_ANONPAGES);
-	page_add_new_anon_rmap(page, vma, addr, false);
+	page_add_new_anon_rmap(page, vma, addr);
 	if (!is_zone_device_page(page))
 		lru_cache_add_inactive_or_unevictable(page, vma);
 	get_page(page);
mm/rmap.c +6 -3
index 7162689203fc..ebe7140c4493 100644
@@ -1184,19 +1184,22 @@ void page_add_anon_rmap(struct page *page,
 }
 
 /**
- * page_add_new_anon_rmap - add pte mapping to a new anonymous page
+ * page_add_new_anon_rmap - add mapping to a new anonymous page
  * @page:	the page to add the mapping to
  * @vma:	the vm area in which the mapping is added
  * @address:	the user virtual address mapped
- * @compound:	charge the page as compound or small page
+ *
+ * If it's a compound page, it is accounted as a compound page. As the page
+ * is new, it's assume to get mapped exclusively by a single process.
  *
  * Same as page_add_anon_rmap but must only be called on *new* pages.
  * This means the inc-and-test can be bypassed.
  * Page does not have to be locked.
  */
 void page_add_new_anon_rmap(struct page *page,
-	struct vm_area_struct *vma, unsigned long address, bool compound)
+	struct vm_area_struct *vma, unsigned long address)
 {
+	const bool compound = PageCompound(page);
 	int nr = compound ? thp_nr_pages(page) : 1;
 
 	VM_BUG_ON_VMA(address < vma->vm_start || address >= vma->vm_end, vma);
mm/swapfile.c +1 -1
index 41ba8238d16b..7edc8e099b22 100644
@@ -1802,7 +1802,7 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 	if (page == swapcache) {
 		page_add_anon_rmap(page, vma, addr, RMAP_NONE);
 	} else { /* ksm created a completely new copy */
-		page_add_new_anon_rmap(page, vma, addr, false);
+		page_add_new_anon_rmap(page, vma, addr);
 		lru_cache_add_inactive_or_unevictable(page, vma);
 	}
 	set_pte_at(vma->vm_mm, addr, pte,
mm/userfaultfd.c +1 -1
index 0780c2a57ff1..4ca854ce14f0 100644
@@ -98,7 +98,7 @@ int mfill_atomic_install_pte(struct mm_struct *dst_mm, pmd_t *dst_pmd,
 	if (page_in_cache)
 		page_add_file_rmap(page, false);
 	else
-		page_add_new_anon_rmap(page, dst_vma, dst_addr, false);
+		page_add_new_anon_rmap(page, dst_vma, dst_addr);
 
 	/*
 	 * Must happen after rmap, as mm_counter() checks mapping (via
Signature · 1 line
2.35.1
#16 lore raw

PATCH v2 9/15 mm/rmap: use page_move_anon_rmap() when reusing a mapped PageAnon() page exclusively

In reply to David Hildenbrand

We want to mark anonymous pages exclusive, and when using page_move_anon_rmap() we know that we are the exclusive user, as properly documented. This is a preparation for marking anonymous pages exclusive in page_move_anon_rmap().

In both instances, we're holding page lock and are sure that we're the exclusive owner (page_count() == 1). hugetlb already properly uses page_move_anon_rmap() in the write fault handler.

Note that in case of a PTE-mapped THP, we'll only end up calling this function if the whole THP is only referenced by the single PTE mapping a single subpage (page_count() == 1); consequently, it's fine to modify the compound page mapping inside page_move_anon_rmap().

---
 mm/huge_memory.c | 2 ++
 mm/memory.c      | 1 +
 2 files changed, 3 insertions(+)
mm/huge_memory.c +2 -0
index c1f7eaba23ff..0b6fb409b9e4 100644
@@ -1317,6 +1317,8 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
 		try_to_free_swap(page);
 	if (page_count(page) == 1) {
 		pmd_t entry;
+
+		page_move_anon_rmap(page, vma);
 		entry = pmd_mkyoung(orig_pmd);
 		entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma);
 		if (pmdp_set_access_flags(vma, haddr, vmf->pmd, entry, 1))
mm/memory.c +1 -0
index 00c45b3a9576..7b32f422798d 100644
@@ -3307,6 +3307,7 @@ static vm_fault_t do_wp_page(struct vm_fault *vmf)
 		 * and the page is locked, it's dark out, and we're wearing
 		 * sunglasses. Hit it.
 		 */
+		page_move_anon_rmap(page, vma);
 		unlock_page(page);
 		wp_page_reuse(vmf);
 		return VM_FAULT_WRITE;
Signature · 1 line
2.35.1
#17 lore raw

PATCH v2 10/15 mm/page-flags: reuse PG_mappedtodisk as PG_anon_exclusive for PageAnon() pages

In reply to David Hildenbrand

The basic question we would like to have a reliable and efficient answer to is: is this anonymous page exclusive to a single process or might it be shared? We need that information for ordinary/single pages, hugetlb pages, and possibly each subpage of a THP.

Introduce a way to mark an anonymous page as exclusive, with the ultimate goal of teaching our COW logic to not do "wrong COWs", whereby GUP pins lose consistency with the pages mapped into the page table, resulting in reported memory corruptions.

Most pageflags already have semantics for anonymous pages, however, PG_mappedtodisk should never apply to pages in the swapcache, so let's reuse that flag.

As PG_has_hwpoisoned also uses that flag on the second tail page of a compound page, convert it to PG_error instead, which is marked as PF_NO_TAIL, so never used for tail pages.

Use custom page flag modification functions such that we can do additional sanity checks. The semantics we'll put into some kernel doc in the future are:

"
PG_anon_exclusive is *usually* only expressive in combination with a page table entry. Depending on the page table entry type it might store the following information:

Is what's mapped via this page table entry exclusive to the single process and can be mapped writable without further checks? If not, it might be shared and we might have to COW.

For now, we only expect PTE-mapped THPs to make use of PG_anon_exclusive in subpages. For other anonymous compound folios (i.e., hugetlb), only the head page is logically mapped and holds this information.

For example, an exclusive, PMD-mapped THP only has PG_anon_exclusive set on the head page. When replacing the PMD by a page table full of PTEs, PG_anon_exclusive, if set on the head page, will be set on all tail pages accordingly. Note that converting from a PTE-mapping to a PMD mapping using the same compound page is currently not possible and consequently doesn't require care.

If GUP wants to take a reliable pin (FOLL_PIN) on an anonymous page, it should only pin if the relevant PG_anon_bit is set. In that case, the pin will be fully reliable and stay consistent with the pages mapped into the page table, as the bit cannot get cleared (e.g., by fork(), KSM) while the page is pinned. For anonymous pages that are mapped R/W, PG_anon_exclusive can be assumed to always be set because such pages cannot possibly be shared.

The page table lock protecting the page table entry is the primary synchronization mechanism for PG_anon_exclusive; GUP-fast that does not take the PT lock needs special care when trying to clear the flag.

Page table entry types and PG_anon_exclusive:
* Present: PG_anon_exclusive applies.
* Swap: the information is lost. PG_anon_exclusive was cleared.
* Migration: the entry holds this information instead. PG_anon_exclusive was cleared.
* Device private: PG_anon_exclusive applies.
* Device exclusive: PG_anon_exclusive applies.
* HW Poison: PG_anon_exclusive is stale and not changed.

If the page may be pinned (FOLL_PIN), clearing PG_anon_exclusive is not allowed and the flag will stick around until the page is freed and folio->mapping is cleared.
"

We won't be clearing PG_anon_exclusive on destructive unmapping (i.e., zapping) of page table entries, page freeing code will handle that when also invalidate page->mapping to not indicate PageAnon() anymore. Letting information about exclusivity stick around will be an important property when adding sanity checks to unpinning code.

Note that we properly clear the flag in free_pages_prepare() via PAGE_FLAGS_CHECK_AT_PREP for each individual subpage of a compound page, so there is no need to manually clear the flag.

---
 include/linux/page-flags.h | 39 +++++++++++++++++++++++++++++++++++++-
 mm/hugetlb.c               |  2 ++
 mm/memory.c                | 11 +++++++++++
 mm/memremap.c              |  9 +++++++++
 mm/swapfile.c              |  4 ++++
 tools/vm/page-types.c      |  8 +++++++-
 6 files changed, 71 insertions(+), 2 deletions(-)
include/linux/page-flags.h +38 -1
index 1c3b6e5c8bfd..e905247dd984 100644
@@ -142,6 +142,15 @@ enum pageflags {
 
 	PG_readahead = PG_reclaim,
 
+	/*
+	 * Depending on the way an anonymous folio can be mapped into a page
+	 * table (e.g., single PMD/PUD/CONT of the head page vs. PTE-mapped
+	 * THP), PG_anon_exclusive may be set only for the head page or for
+	 * tail pages of an anonymous folio. For now, we only expect it to be
+	 * set on tail pages for PTE-mapped THP.
+	 */
+	PG_anon_exclusive = PG_mappedtodisk,
+
 	/* Filesystems */
 	PG_checked = PG_owner_priv_1,
 
@@ -176,7 +185,7 @@ enum pageflags {
 	 * Indicates that at least one subpage is hwpoisoned in the
 	 * THP.
 	 */
-	PG_has_hwpoisoned = PG_mappedtodisk,
+	PG_has_hwpoisoned = PG_error,
 #endif
 
 	/* non-lru isolated movable page */
@@ -920,6 +929,34 @@ extern bool is_free_buddy_page(struct page *page);
 
 __PAGEFLAG(Isolated, isolated, PF_ANY);
 
+static __always_inline int PageAnonExclusive(struct page *page)
+{
+	VM_BUG_ON_PGFLAGS(!PageAnon(page), page);
+	VM_BUG_ON_PGFLAGS(PageHuge(page) && !PageHead(page), page);
+	return test_bit(PG_anon_exclusive, &PF_ANY(page, 1)->flags);
+}
+
+static __always_inline void SetPageAnonExclusive(struct page *page)
+{
+	VM_BUG_ON_PGFLAGS(!PageAnon(page) || PageKsm(page), page);
+	VM_BUG_ON_PGFLAGS(PageHuge(page) && !PageHead(page), page);
+	set_bit(PG_anon_exclusive, &PF_ANY(page, 1)->flags);
+}
+
+static __always_inline void ClearPageAnonExclusive(struct page *page)
+{
+	VM_BUG_ON_PGFLAGS(!PageAnon(page) || PageKsm(page), page);
+	VM_BUG_ON_PGFLAGS(PageHuge(page) && !PageHead(page), page);
+	clear_bit(PG_anon_exclusive, &PF_ANY(page, 1)->flags);
+}
+
+static __always_inline void __ClearPageAnonExclusive(struct page *page)
+{
+	VM_BUG_ON_PGFLAGS(!PageAnon(page), page);
+	VM_BUG_ON_PGFLAGS(PageHuge(page) && !PageHead(page), page);
+	__clear_bit(PG_anon_exclusive, &PF_ANY(page, 1)->flags);
+}
+
 #ifdef CONFIG_MMU
 #define __PG_MLOCKED		(1UL << PG_mlocked)
 #else
mm/hugetlb.c +2 -0
index 9fb990d95dab..1ff0b9e1e28e 100644
@@ -1669,6 +1669,8 @@ void free_huge_page(struct page *page)
 	VM_BUG_ON_PAGE(page_mapcount(page), page);
 
 	hugetlb_set_page_subpool(page, NULL);
+	if (PageAnon(page))
+		__ClearPageAnonExclusive(page);
 	page->mapping = NULL;
 	restore_reserve = HPageRestoreReserve(page);
 	ClearHPageRestoreReserve(page);
mm/memory.c +11 -0
index 7b32f422798d..d01fab481134 100644
@@ -3671,6 +3671,17 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 		goto out_nomap;
 	}
 
+	/*
+	 * PG_anon_exclusive reuses PG_mappedtodisk for anon pages. A swap pte
+	 * must never point at an anonymous page in the swapcache that is
+	 * PG_anon_exclusive. Sanity check that this holds and especially, that
+	 * no filesystem set PG_mappedtodisk on a page in the swapcache. Sanity
+	 * check after taking the PT lock and making sure that nobody
+	 * concurrently faulted in this page and set PG_anon_exclusive.
+	 */
+	BUG_ON(!PageAnon(page) && PageMappedToDisk(page));
+	BUG_ON(PageAnon(page) && PageAnonExclusive(page));
+
 	/*
 	 * Remove the swap entry and conditionally try to free up the swapcache.
 	 * We're already holding a reference on the page but haven't mapped it
mm/memremap.c +9 -0
index 6aa5f0c2d11f..160ea92e4e17 100644
@@ -478,6 +478,15 @@ void free_devmap_managed_page(struct page *page)
 
 	mem_cgroup_uncharge(page_folio(page));
 
+	/*
+	 * Note: we don't expect anonymous compound pages yet. Once supported
+	 * and we could PTE-map them similar to THP, we'd have to clear
+	 * PG_anon_exclusive on all tail pages.
+	 */
+	VM_BUG_ON_PAGE(PageAnon(page) && PageCompound(page), page);
+	if (PageAnon(page))
+		__ClearPageAnonExclusive(page);
+
 	/*
 	 * When a device_private page is freed, the page->mapping field
 	 * may still contain a (stale) mapping value. For example, the
mm/swapfile.c +4 -0
index 7edc8e099b22..493acb967b7a 100644
@@ -1796,6 +1796,10 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 		goto out;
 	}
 
+	/* See do_swap_page() */
+	BUG_ON(!PageAnon(page) && PageMappedToDisk(page));
+	BUG_ON(PageAnon(page) && PageAnonExclusive(page));
+
 	dec_mm_counter(vma->vm_mm, MM_SWAPENTS);
 	inc_mm_counter(vma->vm_mm, MM_ANONPAGES);
 	get_page(page);
tools/vm/page-types.c +7 -1
index b1ed76d9a979..381dcc00cb62 100644
@@ -80,9 +80,10 @@
 #define KPF_SOFTDIRTY		40
 #define KPF_ARCH_2		41
 
-/* [48-] take some arbitrary free slots for expanding overloaded flags
+/* [47-] take some arbitrary free slots for expanding overloaded flags
  * not part of kernel API
  */
+#define KPF_ANON_EXCLUSIVE	47
 #define KPF_READAHEAD		48
 #define KPF_SLOB_FREE		49
 #define KPF_SLUB_FROZEN		50
@@ -138,6 +139,7 @@ static const char * const page_flag_names[] = {
 	[KPF_SOFTDIRTY]		= "f:softdirty",
 	[KPF_ARCH_2]		= "H:arch_2",
 
+	[KPF_ANON_EXCLUSIVE]	= "d:anon_exclusive",
 	[KPF_READAHEAD]		= "I:readahead",
 	[KPF_SLOB_FREE]		= "P:slob_free",
 	[KPF_SLUB_FROZEN]	= "A:slub_frozen",
@@ -472,6 +474,10 @@ static int bit_mask_ok(uint64_t flags)
 
 static uint64_t expand_overloaded_flags(uint64_t flags, uint64_t pme)
 {
+	/* Anonymous pages overload PG_mappedtodisk */
+	if ((flags & BIT(ANON)) && (flags & BIT(MAPPEDTODISK)))
+		flags ^= BIT(MAPPEDTODISK) | BIT(ANON_EXCLUSIVE);
+
 	/* SLOB/SLUB overload several page flags */
 	if (flags & BIT(SLAB)) {
 		if (flags & BIT(PRIVATE))
Signature · 1 line
2.35.1
#18 lore raw

PATCH v2 11/15 mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

In reply to David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

When splitting a compound anonymous folio (THP), the information about exclusivity is implicitly handled via the migration entries: no need to replicate PG_anon_exclusive manually.

I.7. fork() handling

fork() handling is relatively easy, because PG_anon_exclusive is only expressive for some page table entry types.

a) Present anonymous pages

page_try_dup_anon_rmap() will mark the given subpage shared -- which will fail if the page is pinned. If it failed, we have to copy (or PTE-map a PMD to handle it on the PTE level).

Note that device exclusive entries are just a pointer at a PageAnon() page. fork() will first convert a device exclusive entry to a present page table and handle it just like present anonymous pages.

b) Device private entry

Device private entries point at PageAnon() pages that cannot be mapped directly and, therefore, cannot get pinned.

page_try_dup_anon_rmap() will mark the given subpage shared, which cannot fail because they cannot get pinned.

c) HW poison entries

PG_anon_exclusive will remain untouched and is stale -- the page table entry is just a placeholder after all.

d) Migration entries

Writable and readable-exclusive entries are converted to readable
entries: possibly shared.

I.8. mprotect() handling

mprotect() only has to properly handle the new readable-exclusive migration entry:

When write-protecting a migration entry that points at an anonymous page, remember the information about exclusivity via the "readable-exclusive" migration entry type.

II. Migration and GUP-fast

Whenever replacing a present page table entry that maps an exclusive anonymous page by a migration entry, we have to mark the page possibly shared and synchronize against GUP-fast by a proper clear/invalidate+flush to make the following scenario impossible:

1. try_to_migrate() places a migration entry after checking for GUP pins and marks the page possibly shared.
2. GUP-fast pins the page due to lack of synchronization
3. fork() converts the "writable/readable-exclusive" migration entry into a readable migration entry
4. Migration fails due to the GUP pin (failing to freeze the refcount)
5. Migration entries are restored. PG_anon_exclusive is lost

-> We have a pinned page that is not marked exclusive anymore.

Note that we move information about exclusivity from the page to the migration entry as it otherwise highly overcomplicates fork() and PTE-mapping a THP.

III. Swapout and GUP-fast

Whenever replacing a present page table entry that maps an exclusive anonymous page by a swap entry, we have to mark the page possibly shared and synchronize against GUP-fast by a proper clear/invalidate+flush to make the following scenario impossible:

1. try_to_unmap() places a swap entry after checking for GUP pins and clears exclusivity information on the page.
2. GUP-fast pins the page due to lack of synchronization.

-> We have a pinned page that is not marked exclusive anymore.

If we'd ever store information about exclusivity in the swap entry, similar to migration handling, the same considerations as in II would apply. This is future work.

---
 include/linux/rmap.h    | 40 +++++++++++++++++++++
 include/linux/swap.h    | 15 +++++---
 include/linux/swapops.h | 25 +++++++++++++
 mm/huge_memory.c        | 77 +++++++++++++++++++++++++++++++++++++----
 mm/hugetlb.c            | 15 +++++---
 mm/ksm.c                | 13 ++++++-
 mm/memory.c             | 33 +++++++++++++-----
 mm/migrate.c            | 34 ++++++++++++++++--
 mm/mprotect.c           |  8 +++--
 mm/rmap.c               | 59 ++++++++++++++++++++++++++++---
 10 files changed, 285 insertions(+), 34 deletions(-)
include/linux/rmap.h +40 -0
index 51953bace0a3..1aef834e1d60 100644
@@ -224,6 +224,13 @@ static inline int page_try_dup_anon_rmap(struct page *page, bool compound,
 {
 	VM_BUG_ON_PAGE(!PageAnon(page), page);
 
+	/*
+	 * No need to check+clear for already shared pages, including KSM
+	 * pages.
+	 */
+	if (!PageAnonExclusive(page))
+		goto dup;
+
 	/*
 	 * If this page may have been pinned by the parent process,
 	 * don't allow to duplicate the mapping but instead require to e.g.,
@@ -235,14 +242,47 @@ static inline int page_try_dup_anon_rmap(struct page *page, bool compound,
 	    unlikely(page_needs_cow_for_dma(vma, page))))
 		return -EBUSY;
 
+	ClearPageAnonExclusive(page);
 	/*
 	 * It's okay to share the anon page between both processes, mapping
 	 * the page R/O into both processes.
 	 */
+dup:
 	__page_dup_rmap(page, compound);
 	return 0;
 }
 
+/**
+ * page_try_share_anon_rmap - try marking an exclusive anonymous page possibly
+ *			      shared to prepare for KSM or temporary unmapping
+ * @page: the exclusive anonymous page to try marking possibly shared
+ *
+ * The caller needs to hold the PT lock and has to have the page table entry
+ * cleared/invalidated+flushed, to properly sync against GUP-fast.
+ *
+ * This is similar to page_try_dup_anon_rmap(), however, not used during fork()
+ * to duplicate a mapping, but instead to prepare for KSM or temporarily
+ * unmapping a page (swap, migration) via page_remove_rmap().
+ *
+ * Marking the page shared can only fail if the page may be pinned; device
+ * private pages cannot get pinned and consequently this function cannot fail.
+ *
+ * Returns 0 if marking the page possibly shared succeeded. Returns -EBUSY
+ * otherwise.
+ */
+static inline int page_try_share_anon_rmap(struct page *page)
+{
+	VM_BUG_ON_PAGE(!PageAnon(page) || !PageAnonExclusive(page), page);
+
+	/* See page_try_dup_anon_rmap(). */
+	if (likely(!is_device_private_page(page) &&
+	    unlikely(page_maybe_dma_pinned(page))))
+		return -EBUSY;
+
+	ClearPageAnonExclusive(page);
+	return 0;
+}
+
 /*
  * Called from mm/vmscan.c to handle paging out
  */
include/linux/swap.h +11 -4
index b546e4bd5c5a..422765d1141c 100644
@@ -78,12 +78,19 @@ static inline int current_is_kswapd(void)
 #endif
 
 /*
- * NUMA node memory migration support
+ * Page migration support.
+ *
+ * SWP_MIGRATION_READ_EXCLUSIVE is only applicable to anonymous pages and
+ * indicates that the referenced (part of) an anonymous page is exclusive to
+ * a single process. For SWP_MIGRATION_WRITE, that information is implicit:
+ * (part of) an anonymous page that are mapped writable are exclusive to a
+ * single process.
  */
 #ifdef CONFIG_MIGRATION
-#define SWP_MIGRATION_NUM 2
-#define SWP_MIGRATION_READ	(MAX_SWAPFILES + SWP_HWPOISON_NUM)
-#define SWP_MIGRATION_WRITE	(MAX_SWAPFILES + SWP_HWPOISON_NUM + 1)
+#define SWP_MIGRATION_NUM 3
+#define SWP_MIGRATION_READ (MAX_SWAPFILES + SWP_HWPOISON_NUM)
+#define SWP_MIGRATION_READ_EXCLUSIVE (MAX_SWAPFILES + SWP_HWPOISON_NUM + 1)
+#define SWP_MIGRATION_WRITE (MAX_SWAPFILES + SWP_HWPOISON_NUM + 2)
 #else
 #define SWP_MIGRATION_NUM 0
 #endif
include/linux/swapops.h +25 -0
index d356ab4047f7..06280fc1c99b 100644
@@ -194,6 +194,7 @@ static inline bool is_writable_device_exclusive_entry(swp_entry_t entry)
 static inline int is_migration_entry(swp_entry_t entry)
 {
 	return unlikely(swp_type(entry) == SWP_MIGRATION_READ ||
+			swp_type(entry) == SWP_MIGRATION_READ_EXCLUSIVE ||
 			swp_type(entry) == SWP_MIGRATION_WRITE);
 }
 
@@ -202,11 +203,26 @@ static inline int is_writable_migration_entry(swp_entry_t entry)
 	return unlikely(swp_type(entry) == SWP_MIGRATION_WRITE);
 }
 
+static inline int is_readable_migration_entry(swp_entry_t entry)
+{
+	return unlikely(swp_type(entry) == SWP_MIGRATION_READ);
+}
+
+static inline int is_readable_exclusive_migration_entry(swp_entry_t entry)
+{
+	return unlikely(swp_type(entry) == SWP_MIGRATION_READ_EXCLUSIVE);
+}
+
 static inline swp_entry_t make_readable_migration_entry(pgoff_t offset)
 {
 	return swp_entry(SWP_MIGRATION_READ, offset);
 }
 
+static inline swp_entry_t make_readable_exclusive_migration_entry(pgoff_t offset)
+{
+	return swp_entry(SWP_MIGRATION_READ_EXCLUSIVE, offset);
+}
+
 static inline swp_entry_t make_writable_migration_entry(pgoff_t offset)
 {
 	return swp_entry(SWP_MIGRATION_WRITE, offset);
@@ -224,6 +240,11 @@ static inline swp_entry_t make_readable_migration_entry(pgoff_t offset)
 	return swp_entry(0, 0);
 }
 
+static inline swp_entry_t make_readable_exclusive_migration_entry(pgoff_t offset)
+{
+	return swp_entry(0, 0);
+}
+
 static inline swp_entry_t make_writable_migration_entry(pgoff_t offset)
 {
 	return swp_entry(0, 0);
@@ -244,6 +265,10 @@ static inline int is_writable_migration_entry(swp_entry_t entry)
 {
 	return 0;
 }
+static inline int is_readable_migration_entry(swp_entry_t entry)
+{
+	return 0;
+}
 
 #endif
 
mm/huge_memory.c +70 -7

176 lines not on this page · raw message

mm/hugetlb.c +11 -4
index 1ff0b9e1e28e..c0e19ea5ebb5 100644
@@ -4765,7 +4765,7 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 				    is_hugetlb_entry_hwpoisoned(entry))) {
 			swp_entry_t swp_entry = pte_to_swp_entry(entry);
 
-			if (is_writable_migration_entry(swp_entry) && cow) {
+			if (!is_readable_migration_entry(swp_entry) && cow) {
 				/*
 				 * COW mappings require pages in both
 				 * parent and child to be set to read.
@@ -5165,6 +5165,8 @@ static vm_fault_t hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma,
 		set_huge_ptep_writable(vma, haddr, ptep);
 		return 0;
 	}
+	VM_BUG_ON_PAGE(PageAnon(old_page) && PageAnonExclusive(old_page),
+		       old_page);
 
 	/*
 	 * If the process that created a MAP_PRIVATE mapping is about to
@@ -6160,12 +6162,17 @@ unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
 		}
 		if (unlikely(is_hugetlb_entry_migration(pte))) {
 			swp_entry_t entry = pte_to_swp_entry(pte);
+			struct page *page = pfn_swap_entry_to_page(entry);
 
-			if (is_writable_migration_entry(entry)) {
+			if (!is_readable_migration_entry(entry)) {
 				pte_t newpte;
 
-				entry = make_readable_migration_entry(
-							swp_offset(entry));
+				if (PageAnon(page))
+					entry = make_readable_exclusive_migration_entry(
+								swp_offset(entry));
+				else
+					entry = make_readable_migration_entry(
+								swp_offset(entry));
 				newpte = swp_entry_to_pte(entry);
 				set_huge_swap_pte_at(mm, address, ptep,
 						     newpte, huge_page_size(h));
mm/ksm.c +12 -1
index 9ff28097bc0a..350b46432d65 100644
@@ -866,6 +866,7 @@ static inline struct stable_node *page_stable_node(struct page *page)
 static inline void set_page_stable_node(struct page *page,
 					struct stable_node *stable_node)
 {
+	VM_BUG_ON_PAGE(PageAnon(page) && PageAnonExclusive(page), page);
 	page->mapping = (void *)((unsigned long)stable_node | PAGE_MAPPING_KSM);
 }
 
@@ -1041,6 +1042,7 @@ static int write_protect_page(struct vm_area_struct *vma, struct page *page,
 	int swapped;
 	int err = -EFAULT;
 	struct mmu_notifier_range range;
+	bool anon_exclusive;
 
 	pvmw.address = page_address_in_vma(page, vma);
 	if (pvmw.address == -EFAULT)
@@ -1058,9 +1060,10 @@ static int write_protect_page(struct vm_area_struct *vma, struct page *page,
 	if (WARN_ONCE(!pvmw.pte, "Unexpected PMD mapping?"))
 		goto out_unlock;
 
+	anon_exclusive = PageAnonExclusive(page);
 	if (pte_write(*pvmw.pte) || pte_dirty(*pvmw.pte) ||
 	    (pte_protnone(*pvmw.pte) && pte_savedwrite(*pvmw.pte)) ||
-						mm_tlb_flush_pending(mm)) {
+	    anon_exclusive || mm_tlb_flush_pending(mm)) {
 		pte_t entry;
 
 		swapped = PageSwapCache(page);
@@ -1088,6 +1091,12 @@ static int write_protect_page(struct vm_area_struct *vma, struct page *page,
 			set_pte_at(mm, pvmw.address, pvmw.pte, entry);
 			goto out_unlock;
 		}
+
+		if (anon_exclusive && page_try_share_anon_rmap(page)) {
+			set_pte_at(mm, pvmw.address, pvmw.pte, entry);
+			goto out_unlock;
+		}
+
 		if (pte_dirty(entry))
 			set_page_dirty(page);
 
@@ -1146,6 +1155,8 @@ static int replace_page(struct vm_area_struct *vma, struct page *page,
 		pte_unmap_unlock(ptep, ptl);
 		goto out_mn;
 	}
+	VM_BUG_ON_PAGE(PageAnonExclusive(page), page);
+	VM_BUG_ON_PAGE(PageAnon(kpage) && PageAnonExclusive(kpage), kpage);
 
 	/*
 	 * No need to check ksm_use_zero_pages here: we can only have a
mm/memory.c +25 -8
index d01fab481134..222aaf277af4 100644
@@ -720,6 +720,8 @@ static void restore_exclusive_pte(struct vm_area_struct *vma,
 	else if (is_writable_device_exclusive_entry(entry))
 		pte = maybe_mkwrite(pte_mkdirty(pte), vma);
 
+	VM_BUG_ON(pte_write(pte) && !(PageAnon(page) && PageAnonExclusive(page)));
+
 	/*
 	 * No need to take a page reference as one was already
 	 * created when the swap entry was made.
@@ -799,11 +801,12 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 
 		rss[mm_counter(page)]++;
 
-		if (is_writable_migration_entry(entry) &&
+		if (!is_readable_migration_entry(entry) &&
 				is_cow_mapping(vm_flags)) {
 			/*
-			 * COW mappings require pages in both
-			 * parent and child to be set to read.
+			 * COW mappings require pages in both parent and child
+			 * to be set to read. A previously exclusive entry is
+			 * now shared.
 			 */
 			entry = make_readable_migration_entry(
 							swp_offset(entry));
@@ -954,6 +957,7 @@ copy_present_pte(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
 		ptep_set_wrprotect(src_mm, addr, src_pte);
 		pte = pte_wrprotect(pte);
 	}
+	VM_BUG_ON(page && PageAnon(page) && PageAnonExclusive(page));
 
 	/*
 	 * If it's a shared mapping, mark it clean in
@@ -2943,6 +2947,9 @@ static inline void wp_page_reuse(struct vm_fault *vmf)
 	struct vm_area_struct *vma = vmf->vma;
 	struct page *page = vmf->page;
 	pte_t entry;
+
+	VM_BUG_ON(PageAnon(page) && !PageAnonExclusive(page));
+
 	/*
 	 * Clear the pages cpupid information as the existing
 	 * information potentially belongs to a now completely
@@ -3277,6 +3284,13 @@ static vm_fault_t do_wp_page(struct vm_fault *vmf)
 	if (PageAnon(vmf->page)) {
 		struct page *page = vmf->page;
 
+		/*
+		 * If the page is exclusive to this process we must reuse the
+		 * page without further checks.
+		 */
+		if (PageAnonExclusive(page))
+			goto reuse;
+
 		/*
 		 * We have to verify under page lock: these early checks are
 		 * just an optimization to avoid locking the page and freeing
@@ -3309,6 +3323,7 @@ static vm_fault_t do_wp_page(struct vm_fault *vmf)
 		 */
 		page_move_anon_rmap(page, vma);
 		unlock_page(page);
+reuse:
 		wp_page_reuse(vmf);
 		return VM_FAULT_WRITE;
 	} else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) ==
@@ -3700,11 +3715,12 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 	 * that are certainly not shared because we just allocated them without
 	 * exposing them to the swapcache.
 	 */
-	if ((vmf->flags & FAULT_FLAG_WRITE) && !PageKsm(page) &&
-	    (page != swapcache || page_count(page) == 1)) {
-		pte = maybe_mkwrite(pte_mkdirty(pte), vma);
-		vmf->flags &= ~FAULT_FLAG_WRITE;
-		ret |= VM_FAULT_WRITE;
+	if (!PageKsm(page) && (page != swapcache || page_count(page) == 1)) {
+		if (vmf->flags & FAULT_FLAG_WRITE) {
+			pte = maybe_mkwrite(pte_mkdirty(pte), vma);
+			vmf->flags &= ~FAULT_FLAG_WRITE;
+			ret |= VM_FAULT_WRITE;
+		}
 		rmap_flags |= RMAP_EXCLUSIVE;
 	}
 	flush_icache_page(vma, page);
@@ -3724,6 +3740,7 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 		page_add_anon_rmap(page, vma, vmf->address, rmap_flags);
 	}
 
+	VM_BUG_ON(!PageAnon(page) || (pte_write(pte) && !PageAnonExclusive(page)));
 	set_pte_at(vma->vm_mm, vmf->address, vmf->pte, pte);
 	arch_do_swap_page(vma->vm_mm, vma, vmf->address, pte, vmf->orig_pte);
 
mm/migrate.c +31 -3
index fd9eba33b34a..7f440d2103ce 100644
@@ -188,6 +188,8 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 
 	VM_BUG_ON_PAGE(PageTail(page), page);
 	while (page_vma_mapped_walk(&pvmw)) {
+		rmap_t rmap_flags = RMAP_NONE;
+
 		if (PageKsm(page))
 			new = page;
 		else
@@ -217,6 +219,9 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 		else if (pte_swp_uffd_wp(*pvmw.pte))
 			pte = pte_mkuffd_wp(pte);
 
+		if (PageAnon(new) && !is_readable_migration_entry(entry))
+			rmap_flags |= RMAP_EXCLUSIVE;
+
 		if (unlikely(is_device_private_page(new))) {
 			if (pte_write(pte))
 				entry = make_writable_device_private_entry(
@@ -239,7 +244,7 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 			pte = arch_make_huge_pte(pte, shift, vma->vm_flags);
 			if (PageAnon(new))
 				hugepage_add_anon_rmap(new, vma, pvmw.address,
-						       RMAP_NONE);
+						       rmap_flags);
 			else
 				page_dup_file_rmap(new, true);
 			set_huge_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
@@ -248,7 +253,7 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 		{
 			if (PageAnon(new))
 				page_add_anon_rmap(new, vma, pvmw.address,
-						   RMAP_NONE);
+						   rmap_flags);
 			else
 				page_add_file_rmap(new, false);
 			set_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
@@ -379,6 +384,10 @@ int folio_migrate_mapping(struct address_space *mapping,
 		/* No turning back from here */
 		newfolio->index = folio->index;
 		newfolio->mapping = folio->mapping;
+		/*
+		 * Note: PG_anon_exclusive is always migrated via migration
+		 * entries.
+		 */
 		if (folio_test_swapbacked(folio))
 			__folio_set_swapbacked(newfolio);
 
@@ -2302,15 +2311,34 @@ static int migrate_vma_collect_pmd(pmd_t *pmdp,
 		 * set up a special migration page table entry now.
 		 */
 		if (trylock_page(page)) {
+			bool anon_exclusive;
 			pte_t swp_pte;
 
+			anon_exclusive = PageAnon(page) && PageAnonExclusive(page);
+			if (anon_exclusive) {
+				flush_cache_page(vma, addr, pte_pfn(*ptep));
+				ptep_clear_flush(vma, addr, ptep);
+
+				if (page_try_share_anon_rmap(page)) {
+					set_pte_at(mm, addr, ptep, pte);
+					unlock_page(page);
+					put_page(page);
+					mpfn = 0;
+					goto next;
+				}
+			} else {
+				ptep_get_and_clear(mm, addr, ptep);
+			}
+
 			migrate->cpages++;
-			ptep_get_and_clear(mm, addr, ptep);
 
 			/* Setup special migration page table entry */
 			if (mpfn & MIGRATE_PFN_WRITE)
 				entry = make_writable_migration_entry(
 							page_to_pfn(page));
+			else if (anon_exclusive)
+				entry = make_readable_exclusive_migration_entry(
+							page_to_pfn(page));
 			else
 				entry = make_readable_migration_entry(
 							page_to_pfn(page));
mm/mprotect.c +6 -2
index 2887644fd150..9441675e993c 100644
@@ -141,6 +141,7 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
 			pages++;
 		} else if (is_swap_pte(oldpte)) {
 			swp_entry_t entry = pte_to_swp_entry(oldpte);
+			struct page *page = pfn_swap_entry_to_page(entry);
 			pte_t newpte;
 
 			if (is_writable_migration_entry(entry)) {
@@ -148,8 +149,11 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
 				 * A protection check is difficult so
 				 * just be safe and disable write
 				 */
-				entry = make_readable_migration_entry(
-							swp_offset(entry));
+				if (PageAnon(page))
+					entry = make_readable_exclusive_migration_entry(
+							     swp_offset(entry));
+				else
+					entry = make_readable_migration_entry(swp_offset(entry));
 				newpte = swp_entry_to_pte(entry);
 				if (pte_swp_soft_dirty(oldpte))
 					newpte = pte_swp_mksoft_dirty(newpte);
mm/rmap.c +54 -5

156 lines not on this page · raw message

Signature · 1 line
2.35.1
#19 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

Quoted 935 lines

When splitting a compound anonymous folio (THP), the information about exclusivity is implicitly handled via the migration entries: no need to replicate PG_anon_exclusive manually.

I.7. fork() handling

fork() handling is relatively easy, because PG_anon_exclusive is only expressive for some page table entry types.

a) Present anonymous pages

page_try_dup_anon_rmap() will mark the given subpage shared -- which will fail if the page is pinned. If it failed, we have to copy (or PTE-map a PMD to handle it on the PTE level).

Note that device exclusive entries are just a pointer at a PageAnon() page. fork() will first convert a device exclusive entry to a present page table and handle it just like present anonymous pages.

b) Device private entry

Device private entries point at PageAnon() pages that cannot be mapped directly and, therefore, cannot get pinned.

page_try_dup_anon_rmap() will mark the given subpage shared, which cannot fail because they cannot get pinned.

c) HW poison entries

PG_anon_exclusive will remain untouched and is stale -- the page table entry is just a placeholder after all.

d) Migration entries

Writable and readable-exclusive entries are converted to readable
entries: possibly shared.

I.8. mprotect() handling

mprotect() only has to properly handle the new readable-exclusive migration entry:

When write-protecting a migration entry that points at an anonymous page, remember the information about exclusivity via the "readable-exclusive" migration entry type.

II. Migration and GUP-fast

Whenever replacing a present page table entry that maps an exclusive anonymous page by a migration entry, we have to mark the page possibly shared and synchronize against GUP-fast by a proper clear/invalidate+flush to make the following scenario impossible:

1. try_to_migrate() places a migration entry after checking for GUP pins and marks the page possibly shared.
2. GUP-fast pins the page due to lack of synchronization
3. fork() converts the "writable/readable-exclusive" migration entry into a readable migration entry
4. Migration fails due to the GUP pin (failing to freeze the refcount)
5. Migration entries are restored. PG_anon_exclusive is lost

-> We have a pinned page that is not marked exclusive anymore.

Note that we move information about exclusivity from the page to the migration entry as it otherwise highly overcomplicates fork() and PTE-mapping a THP.

III. Swapout and GUP-fast

Whenever replacing a present page table entry that maps an exclusive anonymous page by a swap entry, we have to mark the page possibly shared and synchronize against GUP-fast by a proper clear/invalidate+flush to make the following scenario impossible:

1. try_to_unmap() places a swap entry after checking for GUP pins and clears exclusivity information on the page.
2. GUP-fast pins the page due to lack of synchronization.

-> We have a pinned page that is not marked exclusive anymore.

If we'd ever store information about exclusivity in the swap entry, similar to migration handling, the same considerations as in II would apply. This is future work.

---
 include/linux/rmap.h    | 40 +++++++++++++++++++++
 include/linux/swap.h    | 15 +++++---
 include/linux/swapops.h | 25 +++++++++++++
 mm/huge_memory.c        | 77 +++++++++++++++++++++++++++++++++++++----
 mm/hugetlb.c            | 15 +++++---
 mm/ksm.c                | 13 ++++++-
 mm/memory.c             | 33 +++++++++++++-----
 mm/migrate.c            | 34 ++++++++++++++++--
 mm/mprotect.c           |  8 +++--
 mm/rmap.c               | 59 ++++++++++++++++++++++++++++---
 10 files changed, 285 insertions(+), 34 deletions(-)
include/linux/rmap.h +40 -0
index 51953bace0a3..1aef834e1d60 100644
@@ -224,6 +224,13 @@ static inline int page_try_dup_anon_rmap(struct page *page, bool compound,
 {
        VM_BUG_ON_PAGE(!PageAnon(page), page);
 
+       /*
+        * No need to check+clear for already shared pages, including KSM
+        * pages.
+        */
+       if (!PageAnonExclusive(page))
+               goto dup;
+
        /*
         * If this page may have been pinned by the parent process,
         * don't allow to duplicate the mapping but instead require to e.g.,
@@ -235,14 +242,47 @@ static inline int page_try_dup_anon_rmap(struct page *page, bool compound,
            unlikely(page_needs_cow_for_dma(vma, page))))
                return -EBUSY;
 
+       ClearPageAnonExclusive(page);
        /*
         * It's okay to share the anon page between both processes, mapping
         * the page R/O into both processes.
         */
+dup:
        __page_dup_rmap(page, compound);
        return 0;
 }
 
+/**
+ * page_try_share_anon_rmap - try marking an exclusive anonymous page possibly
+ *                           shared to prepare for KSM or temporary unmapping
+ * @page: the exclusive anonymous page to try marking possibly shared
+ *
+ * The caller needs to hold the PT lock and has to have the page table entry
+ * cleared/invalidated+flushed, to properly sync against GUP-fast.
+ *
+ * This is similar to page_try_dup_anon_rmap(), however, not used during fork()
+ * to duplicate a mapping, but instead to prepare for KSM or temporarily
+ * unmapping a page (swap, migration) via page_remove_rmap().
+ *
+ * Marking the page shared can only fail if the page may be pinned; device
+ * private pages cannot get pinned and consequently this function cannot fail.
+ *
+ * Returns 0 if marking the page possibly shared succeeded. Returns -EBUSY
+ * otherwise.
+ */
+static inline int page_try_share_anon_rmap(struct page *page)
+{
+       VM_BUG_ON_PAGE(!PageAnon(page) || !PageAnonExclusive(page), page);
+
+       /* See page_try_dup_anon_rmap(). */
+       if (likely(!is_device_private_page(page) &&
+           unlikely(page_maybe_dma_pinned(page))))
+               return -EBUSY;
+
+       ClearPageAnonExclusive(page);
+       return 0;
+}
+
 /*
  * Called from mm/vmscan.c to handle paging out
  */
include/linux/swap.h +11 -4
index b546e4bd5c5a..422765d1141c 100644
@@ -78,12 +78,19 @@ static inline int current_is_kswapd(void)
 #endif
 
 /*
- * NUMA node memory migration support
+ * Page migration support.
+ *
+ * SWP_MIGRATION_READ_EXCLUSIVE is only applicable to anonymous pages and
+ * indicates that the referenced (part of) an anonymous page is exclusive to
+ * a single process. For SWP_MIGRATION_WRITE, that information is implicit:
+ * (part of) an anonymous page that are mapped writable are exclusive to a
+ * single process.
  */
 #ifdef CONFIG_MIGRATION
-#define SWP_MIGRATION_NUM 2
-#define SWP_MIGRATION_READ     (MAX_SWAPFILES + SWP_HWPOISON_NUM)
-#define SWP_MIGRATION_WRITE    (MAX_SWAPFILES + SWP_HWPOISON_NUM + 1)
+#define SWP_MIGRATION_NUM 3
+#define SWP_MIGRATION_READ (MAX_SWAPFILES + SWP_HWPOISON_NUM)
+#define SWP_MIGRATION_READ_EXCLUSIVE (MAX_SWAPFILES + SWP_HWPOISON_NUM + 1)
+#define SWP_MIGRATION_WRITE (MAX_SWAPFILES + SWP_HWPOISON_NUM + 2)
 #else
 #define SWP_MIGRATION_NUM 0
 #endif
include/linux/swapops.h +25 -0
index d356ab4047f7..06280fc1c99b 100644
@@ -194,6 +194,7 @@ static inline bool is_writable_device_exclusive_entry(swp_entry_t entry)
 static inline int is_migration_entry(swp_entry_t entry)
 {
        return unlikely(swp_type(entry) == SWP_MIGRATION_READ ||
+                       swp_type(entry) == SWP_MIGRATION_READ_EXCLUSIVE ||
                        swp_type(entry) == SWP_MIGRATION_WRITE);
 }
 
@@ -202,11 +203,26 @@ static inline int is_writable_migration_entry(swp_entry_t entry)
        return unlikely(swp_type(entry) == SWP_MIGRATION_WRITE);
 }
 
+static inline int is_readable_migration_entry(swp_entry_t entry)
+{
+       return unlikely(swp_type(entry) == SWP_MIGRATION_READ);
+}
+
+static inline int is_readable_exclusive_migration_entry(swp_entry_t entry)
+{
+       return unlikely(swp_type(entry) == SWP_MIGRATION_READ_EXCLUSIVE);
+}
+
 static inline swp_entry_t make_readable_migration_entry(pgoff_t offset)
 {
        return swp_entry(SWP_MIGRATION_READ, offset);
 }
 
+static inline swp_entry_t make_readable_exclusive_migration_entry(pgoff_t offset)
+{
+       return swp_entry(SWP_MIGRATION_READ_EXCLUSIVE, offset);
+}
+
 static inline swp_entry_t make_writable_migration_entry(pgoff_t offset)
 {
        return swp_entry(SWP_MIGRATION_WRITE, offset);
@@ -224,6 +240,11 @@ static inline swp_entry_t make_readable_migration_entry(pgoff_t offset)
        return swp_entry(0, 0);
 }
 
+static inline swp_entry_t make_readable_exclusive_migration_entry(pgoff_t offset)
+{
+       return swp_entry(0, 0);
+}
+
 static inline swp_entry_t make_writable_migration_entry(pgoff_t offset)
 {
        return swp_entry(0, 0);
@@ -244,6 +265,10 @@ static inline int is_writable_migration_entry(swp_entry_t entry)
 {
        return 0;
 }
+static inline int is_readable_migration_entry(swp_entry_t entry)
+{
+       return 0;
+}
 
 #endif
 
mm/huge_memory.c +70 -7
index 0b6fb409b9e4..4872e7120ee1 100644
@@ -1054,7 +1054,7 @@ int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm,
                swp_entry_t entry = pmd_to_swp_entry(pmd);
 
                VM_BUG_ON(!is_pmd_migration_entry(pmd));
-               if (is_writable_migration_entry(entry)) {
+               if (!is_readable_migration_entry(entry)) {
                        entry = make_readable_migration_entry(
                                                        swp_offset(entry));
                        pmd = swp_entry_to_pmd(entry);
@@ -1292,6 +1292,10 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
        page = pmd_page(orig_pmd);
        VM_BUG_ON_PAGE(!PageHead(page), page);
 
+       /* Early check when only holding the PT lock. */
+       if (PageAnonExclusive(page))
+               goto reuse;
+
        if (!trylock_page(page)) {
                get_page(page);
                spin_unlock(vmf->ptl);
@@ -1306,6 +1310,12 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
                put_page(page);
        }
 
+       /* Recheck after temporarily dropping the PT lock. */
+       if (PageAnonExclusive(page)) {
+               unlock_page(page);
+               goto reuse;
+       }
+
        /*
         * See do_wp_page(): we can only map the page writable if there are
         * no additional references. Note that we always drain the LRU
@@ -1319,11 +1329,12 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
                pmd_t entry;
 
                page_move_anon_rmap(page, vma);
+               unlock_page(page);
+reuse:
                entry = pmd_mkyoung(orig_pmd);
                entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma);
                if (pmdp_set_access_flags(vma, haddr, vmf->pmd, entry, 1))
                        update_mmu_cache_pmd(vma, vmf->address, vmf->pmd);
-               unlock_page(page);
                spin_unlock(vmf->ptl);
                return VM_FAULT_WRITE;
        }
@@ -1741,6 +1752,7 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd,
 #ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION
        if (is_swap_pmd(*pmd)) {
                swp_entry_t entry = pmd_to_swp_entry(*pmd);
+               struct page *page = pfn_swap_entry_to_page(entry);
 
                VM_BUG_ON(!is_pmd_migration_entry(*pmd));
                if (is_writable_migration_entry(entry)) {
@@ -1749,8 +1761,10 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd,
                         * A protection check is difficult so
                         * just be safe and disable write
                         */
-                       entry = make_readable_migration_entry(
-                                                       swp_offset(entry));
+                       if (PageAnon(page))
+                               entry = make_readable_exclusive_migration_entry(swp_offset(entry));
+                       else
+                               entry = make_readable_migration_entry(swp_offset(entry));
                        newpmd = swp_entry_to_pmd(entry);
                        if (pmd_swp_soft_dirty(*pmd))
                                newpmd = pmd_swp_mksoft_dirty(newpmd);
@@ -1959,6 +1973,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
        pgtable_t pgtable;
        pmd_t old_pmd, _pmd;
        bool young, write, soft_dirty, pmd_migration = false, uffd_wp = false;
+       bool anon_exclusive = false;
        unsigned long addr;
        int i;
 
@@ -2040,6 +2055,8 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
                entry = pmd_to_swp_entry(old_pmd);
                page = pfn_swap_entry_to_page(entry);
                write = is_writable_migration_entry(entry);
+               if (PageAnon(page))
+                       anon_exclusive = is_readable_exclusive_migration_entry(entry);
                young = false;
                soft_dirty = pmd_swp_soft_dirty(old_pmd);
                uffd_wp = pmd_swp_uffd_wp(old_pmd);
@@ -2051,6 +2068,23 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
                young = pmd_young(old_pmd);
                soft_dirty = pmd_soft_dirty(old_pmd);
                uffd_wp = pmd_uffd_wp(old_pmd);
+
+               /*
+                * Without "freeze", we'll simply split the PMD, propagating the
+                * PageAnonExclusive() flag for each PTE by setting it for
+                * each subpage -- no need to (temporarily) clear.
+                *
+                * With "freeze" we want to replace mapped pages by
+                * migration entries right away. This is only possible if we
+                * managed to clear PageAnonExclusive() -- see
+                * set_pmd_migration_entry().
+                *
+                * In case we cannot clear PageAnonExclusive(), split the PMD
+                * only and let try_to_migrate_one() fail later.
+                */
+               anon_exclusive = PageAnon(page) && PageAnonExclusive(page);
+               if (freeze && anon_exclusive && page_try_share_anon_rmap(page))
+                       freeze = false;
        }
        VM_BUG_ON_PAGE(!page_count(page), page);
        page_ref_add(page, HPAGE_PMD_NR - 1);
@@ -2074,6 +2108,9 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
                        if (write)
                                swp_entry = make_writable_migration_entry(
                                                        page_to_pfn(page + i));
+                       else if (anon_exclusive)
+                               swp_entry = make_readable_exclusive_migration_entry(
+                                                       page_to_pfn(page + i));
                        else
                                swp_entry = make_readable_migration_entry(
                                                        page_to_pfn(page + i));
@@ -2085,6 +2122,8 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
                } else {
                        entry = mk_pte(page + i, READ_ONCE(vma->vm_page_prot));
                        entry = maybe_mkwrite(entry, vma);
+                       if (anon_exclusive)
+                               SetPageAnonExclusive(page + i);
                        if (!write)
                                entry = pte_wrprotect(entry);
                        if (!young)
@@ -2315,6 +2354,13 @@ static void __split_huge_page_tail(struct page *head, int tail,
         *
         * After successful get_page_unless_zero() might follow flags change,
         * for example lock_page() which set PG_waiters.
+        *
+        * Note that for mapped sub-pages of an anonymous THP,
+        * PG_anon_exclusive has been cleared in unmap_page() and is stored in
+        * the migration entry instead from where remap_page() will restore it.
+        * We can still have PG_anon_exclusive set on effectively unmapped and
+        * unreferenced sub-pages of an anonymous THP: we can simply drop
+        * PG_anon_exclusive (-> PG_mappedtodisk) for these here.
         */
        page_tail->flags &= ~PAGE_FLAGS_CHECK_AT_PREP;
        page_tail->flags |= (head->flags &
@@ -3070,6 +3116,7 @@ void set_pmd_migration_entry(struct page_vma_mapped_walk *pvmw,
        struct vm_area_struct *vma = pvmw->vma;
        struct mm_struct *mm = vma->vm_mm;
        unsigned long address = pvmw->address;
+       bool anon_exclusive;
        pmd_t pmdval;
        swp_entry_t entry;
        pmd_t pmdswp;
@@ -3079,10 +3126,19 @@ void set_pmd_migration_entry(struct page_vma_mapped_walk *pvmw,
 
        flush_cache_range(vma, address, address + HPAGE_PMD_SIZE);
        pmdval = pmdp_invalidate(vma, address, pvmw->pmd);
+
+       anon_exclusive = PageAnon(page) && PageAnonExclusive(page);
+       if (anon_exclusive && page_try_share_anon_rmap(page)) {
+               set_pmd_at(mm, address, pvmw->pmd, pmdval);
+               return;
+       }
+
        if (pmd_dirty(pmdval))
                set_page_dirty(page);
        if (pmd_write(pmdval))
                entry = make_writable_migration_entry(page_to_pfn(page));
+       else if (anon_exclusive)
+               entry = make_readable_exclusive_migration_entry(page_to_pfn(page));
        else
                entry = make_readable_migration_entry(page_to_pfn(page));
        pmdswp = swp_entry_to_pmd(entry);
@@ -3116,10 +3172,17 @@ void remove_migration_pmd(struct page_vma_mapped_walk *pvmw, struct page *new)
                pmde = pmd_wrprotect(pmd_mkuffd_wp(pmde));
 
        flush_cache_range(vma, mmun_start, mmun_start + HPAGE_PMD_SIZE);
-       if (PageAnon(new))
-               page_add_anon_rmap(new, vma, mmun_start, RMAP_COMPOUND);
-       else
+       if (PageAnon(new)) {
+               int rmap_flags = RMAP_COMPOUND;
+
+               if (!is_readable_migration_entry(entry))
+                       rmap_flags |= RMAP_EXCLUSIVE;
+
+               page_add_anon_rmap(new, vma, mmun_start, rmap_flags);
+       } else {
                page_add_file_rmap(new, true);
+       }
+       VM_BUG_ON(pmd_write(pmde) && PageAnon(new) && !PageAnonExclusive(new));
        set_pmd_at(mm, mmun_start, pvmw->pmd, pmde);
        if ((vma->vm_flags & VM_LOCKED) && !PageDoubleMap(new))
                mlock_vma_page(new);
mm/hugetlb.c +11 -4
index 1ff0b9e1e28e..c0e19ea5ebb5 100644
@@ -4765,7 +4765,7 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
                                    is_hugetlb_entry_hwpoisoned(entry))) {
                        swp_entry_t swp_entry = pte_to_swp_entry(entry);
 
-                       if (is_writable_migration_entry(swp_entry) && cow) {
+                       if (!is_readable_migration_entry(swp_entry) && cow) {
                                /*
                                 * COW mappings require pages in both
                                 * parent and child to be set to read.
@@ -5165,6 +5165,8 @@ static vm_fault_t hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma,
                set_huge_ptep_writable(vma, haddr, ptep);
                return 0;
        }
+       VM_BUG_ON_PAGE(PageAnon(old_page) && PageAnonExclusive(old_page),
+                      old_page);
 
        /*
         * If the process that created a MAP_PRIVATE mapping is about to
@@ -6160,12 +6162,17 @@ unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
                }
                if (unlikely(is_hugetlb_entry_migration(pte))) {
                        swp_entry_t entry = pte_to_swp_entry(pte);
+                       struct page *page = pfn_swap_entry_to_page(entry);
 
-                       if (is_writable_migration_entry(entry)) {
+                       if (!is_readable_migration_entry(entry)) {
                                pte_t newpte;
 
-                               entry = make_readable_migration_entry(
-                                                       swp_offset(entry));
+                               if (PageAnon(page))
+                                       entry = make_readable_exclusive_migration_entry(
+                                                               swp_offset(entry));
+                               else
+                                       entry = make_readable_migration_entry(
+                                                               swp_offset(entry));
                                newpte = swp_entry_to_pte(entry);
                                set_huge_swap_pte_at(mm, address, ptep,
                                                     newpte, huge_page_size(h));
mm/ksm.c +12 -1
index 9ff28097bc0a..350b46432d65 100644
@@ -866,6 +866,7 @@ static inline struct stable_node *page_stable_node(struct page *page)
 static inline void set_page_stable_node(struct page *page,
                                        struct stable_node *stable_node)
 {
+       VM_BUG_ON_PAGE(PageAnon(page) && PageAnonExclusive(page), page);
        page->mapping = (void *)((unsigned long)stable_node | PAGE_MAPPING_KSM);
 }
 
@@ -1041,6 +1042,7 @@ static int write_protect_page(struct vm_area_struct *vma, struct page *page,
        int swapped;
        int err = -EFAULT;
        struct mmu_notifier_range range;
+       bool anon_exclusive;
 
        pvmw.address = page_address_in_vma(page, vma);
        if (pvmw.address == -EFAULT)
@@ -1058,9 +1060,10 @@ static int write_protect_page(struct vm_area_struct *vma, struct page *page,
        if (WARN_ONCE(!pvmw.pte, "Unexpected PMD mapping?"))
                goto out_unlock;
 
+       anon_exclusive = PageAnonExclusive(page);
        if (pte_write(*pvmw.pte) || pte_dirty(*pvmw.pte) ||
            (pte_protnone(*pvmw.pte) && pte_savedwrite(*pvmw.pte)) ||
-                                               mm_tlb_flush_pending(mm)) {
+           anon_exclusive || mm_tlb_flush_pending(mm)) {
                pte_t entry;
 
                swapped = PageSwapCache(page);
@@ -1088,6 +1091,12 @@ static int write_protect_page(struct vm_area_struct *vma, struct page *page,
                        set_pte_at(mm, pvmw.address, pvmw.pte, entry);
                        goto out_unlock;
                }
+
+               if (anon_exclusive && page_try_share_anon_rmap(page)) {
+                       set_pte_at(mm, pvmw.address, pvmw.pte, entry);
+                       goto out_unlock;
+               }
+
                if (pte_dirty(entry))
                        set_page_dirty(page);
 
@@ -1146,6 +1155,8 @@ static int replace_page(struct vm_area_struct *vma, struct page *page,
                pte_unmap_unlock(ptep, ptl);
                goto out_mn;
        }
+       VM_BUG_ON_PAGE(PageAnonExclusive(page), page);
+       VM_BUG_ON_PAGE(PageAnon(kpage) && PageAnonExclusive(kpage), kpage);
 
        /*
         * No need to check ksm_use_zero_pages here: we can only have a
mm/memory.c +25 -8
index d01fab481134..222aaf277af4 100644
@@ -720,6 +720,8 @@ static void restore_exclusive_pte(struct vm_area_struct *vma,
        else if (is_writable_device_exclusive_entry(entry))
                pte = maybe_mkwrite(pte_mkdirty(pte), vma);
 
+       VM_BUG_ON(pte_write(pte) && !(PageAnon(page) && PageAnonExclusive(page)));
+
        /*
         * No need to take a page reference as one was already
         * created when the swap entry was made.
@@ -799,11 +801,12 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 
                rss[mm_counter(page)]++;
 
-               if (is_writable_migration_entry(entry) &&
+               if (!is_readable_migration_entry(entry) &&
                                is_cow_mapping(vm_flags)) {
                        /*
-                        * COW mappings require pages in both
-                        * parent and child to be set to read.
+                        * COW mappings require pages in both parent and child
+                        * to be set to read. A previously exclusive entry is
+                        * now shared.
                         */
                        entry = make_readable_migration_entry(
                                                        swp_offset(entry));
@@ -954,6 +957,7 @@ copy_present_pte(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
                ptep_set_wrprotect(src_mm, addr, src_pte);
                pte = pte_wrprotect(pte);
        }
+       VM_BUG_ON(page && PageAnon(page) && PageAnonExclusive(page));
 
        /*
         * If it's a shared mapping, mark it clean in
@@ -2943,6 +2947,9 @@ static inline void wp_page_reuse(struct vm_fault *vmf)
        struct vm_area_struct *vma = vmf->vma;
        struct page *page = vmf->page;
        pte_t entry;
+
+       VM_BUG_ON(PageAnon(page) && !PageAnonExclusive(page));
+
        /*
         * Clear the pages cpupid information as the existing
         * information potentially belongs to a now completely
@@ -3277,6 +3284,13 @@ static vm_fault_t do_wp_page(struct vm_fault *vmf)
        if (PageAnon(vmf->page)) {
                struct page *page = vmf->page;
 
+               /*
+                * If the page is exclusive to this process we must reuse the
+                * page without further checks.
+                */
+               if (PageAnonExclusive(page))
+                       goto reuse;
+
                /*
                 * We have to verify under page lock: these early checks are
                 * just an optimization to avoid locking the page and freeing
@@ -3309,6 +3323,7 @@ static vm_fault_t do_wp_page(struct vm_fault *vmf)
                 */
                page_move_anon_rmap(page, vma);
                unlock_page(page);
+reuse:
                wp_page_reuse(vmf);
                return VM_FAULT_WRITE;
        } else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) ==
@@ -3700,11 +3715,12 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
         * that are certainly not shared because we just allocated them without
         * exposing them to the swapcache.
         */
-       if ((vmf->flags & FAULT_FLAG_WRITE) && !PageKsm(page) &&
-           (page != swapcache || page_count(page) == 1)) {
-               pte = maybe_mkwrite(pte_mkdirty(pte), vma);
-               vmf->flags &= ~FAULT_FLAG_WRITE;
-               ret |= VM_FAULT_WRITE;
+       if (!PageKsm(page) && (page != swapcache || page_count(page) == 1)) {
+               if (vmf->flags & FAULT_FLAG_WRITE) {
+                       pte = maybe_mkwrite(pte_mkdirty(pte), vma);
+                       vmf->flags &= ~FAULT_FLAG_WRITE;
+                       ret |= VM_FAULT_WRITE;
+               }
                rmap_flags |= RMAP_EXCLUSIVE;
        }
        flush_icache_page(vma, page);
@@ -3724,6 +3740,7 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
                page_add_anon_rmap(page, vma, vmf->address, rmap_flags);
        }
 
+       VM_BUG_ON(!PageAnon(page) || (pte_write(pte) && !PageAnonExclusive(page)));
        set_pte_at(vma->vm_mm, vmf->address, vmf->pte, pte);
        arch_do_swap_page(vma->vm_mm, vma, vmf->address, pte, vmf->orig_pte);
 
mm/migrate.c +31 -3
index fd9eba33b34a..7f440d2103ce 100644
@@ -188,6 +188,8 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
 
        VM_BUG_ON_PAGE(PageTail(page), page);
        while (page_vma_mapped_walk(&pvmw)) {
+               rmap_t rmap_flags = RMAP_NONE;
+
                if (PageKsm(page))
                        new = page;
                else
@@ -217,6 +219,9 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
                else if (pte_swp_uffd_wp(*pvmw.pte))
                        pte = pte_mkuffd_wp(pte);
 
+               if (PageAnon(new) && !is_readable_migration_entry(entry))
+                       rmap_flags |= RMAP_EXCLUSIVE;
+
                if (unlikely(is_device_private_page(new))) {
                        if (pte_write(pte))
                                entry = make_writable_device_private_entry(
@@ -239,7 +244,7 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
                        pte = arch_make_huge_pte(pte, shift, vma->vm_flags);
                        if (PageAnon(new))
                                hugepage_add_anon_rmap(new, vma, pvmw.address,
-                                                      RMAP_NONE);
+                                                      rmap_flags);
                        else
                                page_dup_file_rmap(new, true);
                        set_huge_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
@@ -248,7 +253,7 @@ static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma,
                {
                        if (PageAnon(new))
                                page_add_anon_rmap(new, vma, pvmw.address,
-                                                  RMAP_NONE);
+                                                  rmap_flags);
                        else
                                page_add_file_rmap(new, false);
                        set_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte);
@@ -379,6 +384,10 @@ int folio_migrate_mapping(struct address_space *mapping,
                /* No turning back from here */
                newfolio->index = folio->index;
                newfolio->mapping = folio->mapping;
+               /*
+                * Note: PG_anon_exclusive is always migrated via migration
+                * entries.
+                */
                if (folio_test_swapbacked(folio))
                        __folio_set_swapbacked(newfolio);
 
@@ -2302,15 +2311,34 @@ static int migrate_vma_collect_pmd(pmd_t *pmdp,
                 * set up a special migration page table entry now.
                 */
                if (trylock_page(page)) {
+                       bool anon_exclusive;
                        pte_t swp_pte;
 
+                       anon_exclusive = PageAnon(page) && PageAnonExclusive(page);
+                       if (anon_exclusive) {
+                               flush_cache_page(vma, addr, pte_pfn(*ptep));
+                               ptep_clear_flush(vma, addr, ptep);
+
+                               if (page_try_share_anon_rmap(page)) {
+                                       set_pte_at(mm, addr, ptep, pte);
+                                       unlock_page(page);
+                                       put_page(page);
+                                       mpfn = 0;
+                                       goto next;
+                               }
+                       } else {
+                               ptep_get_and_clear(mm, addr, ptep);
+                       }
+
                        migrate->cpages++;
-                       ptep_get_and_clear(mm, addr, ptep);
 
                        /* Setup special migration page table entry */
                        if (mpfn & MIGRATE_PFN_WRITE)
                                entry = make_writable_migration_entry(
                                                        page_to_pfn(page));
+                       else if (anon_exclusive)
+                               entry = make_readable_exclusive_migration_entry(
+                                                       page_to_pfn(page));
                        else
                                entry = make_readable_migration_entry(
                                                        page_to_pfn(page));
mm/mprotect.c +6 -2
index 2887644fd150..9441675e993c 100644
@@ -141,6 +141,7 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
                        pages++;
                } else if (is_swap_pte(oldpte)) {
                        swp_entry_t entry = pte_to_swp_entry(oldpte);
+                       struct page *page = pfn_swap_entry_to_page(entry);
                        pte_t newpte;
 
                        if (is_writable_migration_entry(entry)) {
@@ -148,8 +149,11 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
                                 * A protection check is difficult so
                                 * just be safe and disable write
                                 */
-                               entry = make_readable_migration_entry(
-                                                       swp_offset(entry));
+                               if (PageAnon(page))
+                                       entry = make_readable_exclusive_migration_entry(
+                                                            swp_offset(entry));
+                               else
+                                       entry = make_readable_migration_entry(swp_offset(entry));
                                newpte = swp_entry_to_pte(entry);
                                if (pte_swp_soft_dirty(oldpte))
                                        newpte = pte_swp_mksoft_dirty(newpte);
mm/rmap.c +54 -6
index ebe7140c4493..9d2a7e11e8cc 100644
@@ -1048,6 +1048,7 @@ EXPORT_SYMBOL_GPL(folio_mkclean);
 void page_move_anon_rmap(struct page *page, struct vm_area_struct *vma)
 {
        struct anon_vma *anon_vma = vma->anon_vma;
+       struct page *subpage = page;
 
        page = compound_head(page);
 
@@ -1061,6 +1062,7 @@ void page_move_anon_rmap(struct page *page, struct vm_area_struct *vma)
         * PageAnon()) will not see one without the other.
         */
        WRITE_ONCE(page->mapping, (struct address_space *) anon_vma);
+       SetPageAnonExclusive(subpage);
 }
 
 /**
@@ -1078,7 +1080,7 @@ static void __page_set_anon_rmap(struct page *page,
        BUG_ON(!anon_vma);
 
        if (PageAnon(page))
-               return;
+               goto out;
 
        /*
         * If the page isn't exclusively mapped into this vma,
@@ -1097,6 +1099,9 @@ static void __page_set_anon_rmap(struct page *page,
        anon_vma = (void *) anon_vma + PAGE_MAPPING_ANON;
        WRITE_ONCE(page->mapping, (struct address_space *) anon_vma);
        page->index = linear_page_index(vma, address);
+out:
+       if (exclusive)
+               SetPageAnonExclusive(page);
 }
 
 /**
@@ -1156,6 +1161,8 @@ void page_add_anon_rmap(struct page *page,
        } else {
                first = atomic_inc_and_test(&page->_mapcount);
        }
+       VM_BUG_ON_PAGE(!first && (flags & RMAP_EXCLUSIVE), page);
+       VM_BUG_ON_PAGE(!first && PageAnonExclusive(page), page);
 
        if (first) {
                int nr = compound ? thp_nr_pages(page) : 1;
@@ -1422,7 +1429,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
        };
        pte_t pteval;
        struct page *subpage;
-       bool ret = true;
+       bool anon_exclusive, ret = true;
        struct mmu_notifier_range range;
        enum ttu_flags flags = (enum ttu_flags)(long)arg;
 
@@ -1485,6 +1492,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
 
                subpage = page - page_to_pfn(page) + pte_pfn(*pvmw.pte);
                address = pvmw.address;
+               anon_exclusive = PageAnon(page) && PageAnonExclusive(subpage);
 
                if (PageHuge(page) && !PageAnon(page)) {
                        /*
@@ -1520,9 +1528,12 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
                        }
                }
 
-               /* Nuke the page table entry. */
+               /*
+                * Nuke the page table entry. When having to clear
+                * PageAnonExclusive(), we always have to flush.
+                */
                flush_cache_page(vma, address, pte_pfn(*pvmw.pte));
-               if (should_defer_flush(mm, flags)) {
+               if (should_defer_flush(mm, flags) && !anon_exclusive) {
                        /*
                         * We clear the PTE but do not flush so potentially
                         * a remote CPU could still be writing to the page.
@@ -1623,6 +1634,24 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               swap_free(entry);
+                               set_pte_at(mm, address, pvmw.pte, pteval);
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }
+                       /*
+                        * Note: We *don't* remember yet if the page was mapped
+                        * exclusively in the swap entry, so swapin code has
+                        * to re-determine that manually and might detect the
+                        * page as possibly shared, for example, if there are
+                        * other references on the page or if the page is under
+                        * writeback. We made sure that there are no GUP pins
+                        * on the page that would rely on it, so for GUP pins
+                        * this is fine.
+                        */
                        if (list_empty(&mm->mmlist)) {
                                spin_lock(&mmlist_lock);
                                if (list_empty(&mm->mmlist))
@@ -1723,7 +1752,7 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
        };
        pte_t pteval;
        struct page *subpage;
-       bool ret = true;
+       bool anon_exclusive, ret = true;
        struct mmu_notifier_range range;
        enum ttu_flags flags = (enum ttu_flags)(long)arg;
 
@@ -1782,6 +1811,7 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
 
                subpage = page - page_to_pfn(page) + pte_pfn(*pvmw.pte);
                address = pvmw.address;
+               anon_exclusive = PageAnon(page) && PageAnonExclusive(subpage);
 
                if (PageHuge(page) && !PageAnon(page)) {
                        /*
@@ -1833,6 +1863,9 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                        swp_entry_t entry;
                        pte_t swp_pte;
 
+                       if (anon_exclusive)
+                               BUG_ON(page_try_share_anon_rmap(subpage));
+
                        /*
                         * Store the pfn of the page in a special migration
                         * pte. do_swap_page() will wait until the migration
@@ -1841,6 +1874,8 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                        entry = pte_to_swp_entry(pteval);
                        if (is_writable_device_private_entry(entry))
                                entry = make_writable_migration_entry(pfn);
+                       else if (anon_exclusive)
+                               entry = make_readable_exclusive_migration_entry(pfn);
                        else
                                entry = make_readable_migration_entry(pfn);
                        swp_pte = swp_entry_to_pte(entry);
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }
 
                        /*
                         * Store the pfn of the page in a special migration
@@ -1912,6 +1956,9 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                        if (pte_write(pteval))
                                entry = make_writable_migration_entry(
                                                        page_to_pfn(subpage));
+                       else if (anon_exclusive)
+                               entry = make_readable_exclusive_migration_entry(
+                                                       page_to_pfn(subpage));
                        else
                                entry = make_readable_migration_entry(
                                                        page_to_pfn(subpage));
@@ -2425,6 +2472,8 @@ void hugepage_add_anon_rmap(struct page *page, struct vm_area_struct *vma,
        BUG_ON(!anon_vma);
        /* address might be in next vma when migration races vma_adjust */
        first = atomic_inc_and_test(compound_mapcount_ptr(page));
+       VM_BUG_ON_PAGE(!first && (flags & RMAP_EXCLUSIVE), page);
+       VM_BUG_ON_PAGE(!first && PageAnonExclusive(page), page);
        if (first)
                __page_set_anon_rmap(page, vma, address,
                                     flags & RMAP_EXCLUSIVE);
--

2.35.1

#20 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 5 lines
Quoted 5 lines
mm/rmap.c +1 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
Quoted 8 lines

mm/rmap.c static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma, · line 1949

Quoted 8 lines
mm/rmap.c +8 -0
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Signature · 3 lines
Thanks,

David / dhildenb
#21 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 157 lines · David Hildenbrand
Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 9 lines
Quoted 9 lines
mm/rmap.c +5 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
Quoted 8 lines
Quoted 4 lines

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1644

Quoted 4 lines
mm/rmap.c +4 -0
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Thanks, I missed that. Yes, the page will still be mapped. This should trigger the VM_WARN_ON_ONCE in unmap_page(), if this change will make this happen more often, we may consider removing that warning even though it is "once" since seeing a mapped page may become a normal case (once DIO is switched to FOLL_PIN, it may be more often). Anyway we don't have to remove it right now.

Quoted 6 lines

--
Thanks,

David / dhildenb

#22 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 167 lines · Yang Shi
Quoted 157 lines · David Hildenbrand
Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 9 lines
Quoted 9 lines
mm/rmap.c +5 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
Quoted 8 lines
Quoted 4 lines

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1644

Quoted 4 lines
mm/rmap.c +4 -0
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Quoted 6 lines

Thanks, I missed that. Yes, the page will still be mapped. This should trigger the VM_WARN_ON_ONCE in unmap_page(), if this change will make this happen more often, we may consider removing that warning even though it is "once" since seeing a mapped page may become a normal case (once DIO is switched to FOLL_PIN, it may be more often). Anyway we don't have to remove it right now.

Oh, very good catch! I wasn't able to trigger that warning in my testing so far. Interestingly, arch_unmap_one() could theoretically make this fail already and trigger the warning.

Apart from that warning, split_huge_page_to_list() should work as
expected: freezing the refcount will fail if still mapped and we'll remap.

I'll include a separate patch to just remove that VM_WARN_ON_ONCE -- thanks!

Signature · 3 lines
Thanks,

David / dhildenb
#23 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 175 lines · David Hildenbrand
Quoted 167 lines · Yang Shi
Quoted 157 lines · David Hildenbrand
Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 9 lines
Quoted 9 lines
(quoted diff) +5 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
Quoted 8 lines
Quoted 4 lines

Review static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma, · line 1949

Quoted 4 lines
(quoted diff) +4 -0
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Quoted 6 lines

Thanks, I missed that. Yes, the page will still be mapped. This should trigger the VM_WARN_ON_ONCE in unmap_page(), if this change will make this happen more often, we may consider removing that warning even though it is "once" since seeing a mapped page may become a normal case (once DIO is switched to FOLL_PIN, it may be more often). Anyway we don't have to remove it right now.

Quoted 9 lines

Oh, very good catch! I wasn't able to trigger that warning in my testing so far. Interestingly, arch_unmap_one() could theoretically make this fail already and trigger the warning.

Apart from that warning, split_huge_page_to_list() should work as
expected: freezing the refcount will fail if still mapped and we'll remap.

I'll include a separate patch to just remove that VM_WARN_ON_ONCE -- thanks!

From e6e983d841cd2aa2a9c8dc71779211881cf0d96f Mon Sep 17 00:00:00 2001
From: David Hildenbrand <[email protected]>
Date: Sat, 19 Mar 2022 11:49:39 +0100
Subject: [PATCH] mm/huge_memory: remove outdated VM_WARN_ON_ONCE_PAGE from
unmap_page()

We can already theoretically fail to unmap (still having page_mapped()) in case arch_unmap_one() fails. Although this applies only to anonymous pages for now, get rid of the VM_WARN_ON_ONCE_PAGE() completely: the caller -- split_huge_page_to_list() -- will fail to freeze the refcount and remap the page via remap_page(). So the caller can handle unmap errors just fine.

This is a preparation for making try_to_migrate() fail on anonymous pages with GUP pins.

---
 mm/huge_memory.c | 2 --
 1 file changed, 2 deletions(-)
mm/huge_memory.c +0 -2
index 0b6fb409b9e4..0fe0ab3ec3fc 100644
@@ -2263,8 +2263,6 @@ static void unmap_page(struct page *page)
 		try_to_migrate(page, ttu_flags);
 	else
 		try_to_unmap(page, ttu_flags | TTU_IGNORE_MLOCK);
-
-	VM_WARN_ON_ONCE_PAGE(page_mapped(page), page);
 }
 
 static void remap_page(struct page *page, unsigned int nr)
Signature · 7 lines
2.35.1


-- 
Thanks,

David / dhildenb
#24 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 200 lines · David Hildenbrand
Quoted 175 lines · David Hildenbrand
Quoted 167 lines · Yang Shi
Quoted 157 lines · David Hildenbrand
Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 9 lines
Quoted 9 lines
(quoted diff) +5 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
Quoted 8 lines
Quoted 4 lines

Review static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma, · line 1949

Quoted 4 lines
(quoted diff) +4 -0
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Quoted 6 lines

Thanks, I missed that. Yes, the page will still be mapped. This should trigger the VM_WARN_ON_ONCE in unmap_page(), if this change will make this happen more often, we may consider removing that warning even though it is "once" since seeing a mapped page may become a normal case (once DIO is switched to FOLL_PIN, it may be more often). Anyway we don't have to remove it right now.

Quoted 9 lines

Oh, very good catch! I wasn't able to trigger that warning in my testing so far. Interestingly, arch_unmap_one() could theoretically make this fail already and trigger the warning.

Apart from that warning, split_huge_page_to_list() should work as
expected: freezing the refcount will fail if still mapped and we'll remap.

I'll include a separate patch to just remove that VM_WARN_ON_ONCE -- thanks!

From e6e983d841cd2aa2a9c8dc71779211881cf0d96f Mon Sep 17 00:00:00 2001
From: David Hildenbrand <[email protected]>
Date: Sat, 19 Mar 2022 11:49:39 +0100
Subject: [PATCH] mm/huge_memory: remove outdated VM_WARN_ON_ONCE_PAGE from
unmap_page()

We can already theoretically fail to unmap (still having page_mapped()) in case arch_unmap_one() fails. Although this applies only to anonymous pages for now, get rid of the VM_WARN_ON_ONCE_PAGE() completely: the caller -- split_huge_page_to_list() -- will fail to freeze the refcount and remap the page via remap_page(). So the caller can handle unmap errors just fine.

Quoted 2 lines

This is a preparation for making try_to_migrate() fail on anonymous pages with GUP pins.

Better to mention "making try_to_migrate() fail on pinned anonymous pages may make this happen more usual".

Quoted 3 lines

Fine to me. Reviewed-by: Yang Shi <[email protected]>

Quoted 26 lines
---
 mm/huge_memory.c | 2 --
 1 file changed, 2 deletions(-)
mm/huge_memory.c +0 -3
index 0b6fb409b9e4..0fe0ab3ec3fc 100644
@@ -2263,8 +2263,6 @@ static void unmap_page(struct page *page)
                try_to_migrate(page, ttu_flags);
        else
                try_to_unmap(page, ttu_flags | TTU_IGNORE_MLOCK);
-
-       VM_WARN_ON_ONCE_PAGE(page_mapped(page), page);
 }
 
 static void remap_page(struct page *page, unsigned int nr)
--

2.35.1

--
Thanks,

David / dhildenb

author Reviewed-by Yang Shi
#25 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

Quoted 204 lines · Yang Shi
Quoted 200 lines · David Hildenbrand
Quoted 175 lines · David Hildenbrand
Quoted 167 lines · Yang Shi
Quoted 157 lines · David Hildenbrand
Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 9 lines
Quoted 9 lines
(quoted diff) +5 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
Quoted 8 lines
Quoted 4 lines

Review static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma, · line 1949

Quoted 4 lines
(quoted diff) +4 -0
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Quoted 6 lines

Thanks, I missed that. Yes, the page will still be mapped. This should trigger the VM_WARN_ON_ONCE in unmap_page(), if this change will make this happen more often, we may consider removing that warning even though it is "once" since seeing a mapped page may become a normal case (once DIO is switched to FOLL_PIN, it may be more often). Anyway we don't have to remove it right now.

Quoted 9 lines

Oh, very good catch! I wasn't able to trigger that warning in my testing so far. Interestingly, arch_unmap_one() could theoretically make this fail already and trigger the warning.

Apart from that warning, split_huge_page_to_list() should work as
expected: freezing the refcount will fail if still mapped and we'll remap.

I'll include a separate patch to just remove that VM_WARN_ON_ONCE -- thanks!

From e6e983d841cd2aa2a9c8dc71779211881cf0d96f Mon Sep 17 00:00:00 2001
From: David Hildenbrand <[email protected]>
Date: Sat, 19 Mar 2022 11:49:39 +0100
Subject: [PATCH] mm/huge_memory: remove outdated VM_WARN_ON_ONCE_PAGE from
unmap_page()

We can already theoretically fail to unmap (still having page_mapped()) in case arch_unmap_one() fails. Although this applies only to anonymous pages for now, get rid of the VM_WARN_ON_ONCE_PAGE() completely: the caller -- split_huge_page_to_list() -- will fail to freeze the refcount and remap the page via remap_page(). So the caller can handle unmap errors just fine.

Quoted 2 lines

This is a preparation for making try_to_migrate() fail on anonymous pages with GUP pins.

Quoted 8 lines

Better to mention "making try_to_migrate() fail on pinned anonymous pages may make this happen more usual".

Quoted 3 lines

Fine to me. Reviewed-by: Yang Shi <[email protected]>

Thanks, updated patch:

From 4f68aa1067e43caa370a62b930094a4d07492f92 Mon Sep 17 00:00:00 2001
From: David Hildenbrand <[email protected]>
Date: Sat, 19 Mar 2022 11:49:39 +0100
Subject: [PATCH] mm/huge_memory: remove outdated VM_WARN_ON_ONCE_PAGE from
unmap_page()

We can already theoretically fail to unmap (still having page_mapped()) in case arch_unmap_one() fails, which can happen on sparc. Failures to unmap are handled gracefully, just as if there are other references on the target page: freezing the refcount in split_huge_page_to_list() will fail if still mapped and we'll simply remap.

In commit 504e070dc08f ("mm: thp: replace DEBUG_VM BUG with VM_WARN when unmap fails for split") we already converted to VM_WARN_ON_ONCE_PAGE, let's get rid of it completely now.

This is a preparation for making try_to_migrate() fail on anonymous pages with GUP pins, which will make this VM_WARN_ON_ONCE_PAGE trigger more frequently.

---
 mm/huge_memory.c | 2 --
 1 file changed, 2 deletions(-)
mm/huge_memory.c +0 -2
index 0b6fb409b9e4..0fe0ab3ec3fc 100644
@@ -2263,8 +2263,6 @@ static void unmap_page(struct page *page)
 		try_to_migrate(page, ttu_flags);
 	else
 		try_to_unmap(page, ttu_flags | TTU_IGNORE_MLOCK);
-
-	VM_WARN_ON_ONCE_PAGE(page_mapped(page), page);
 }
 
 static void remap_page(struct page *page, unsigned int nr)
Signature · 8 lines
2.35.1



-- 
Thanks,

David / dhildenb
#26 lore raw

Re: [PATCH v2 11/15] mm: remember exclusively mapped anonymous pages with PG_anon_exclusive

In reply to David Hildenbrand
Quoted 176 lines · David Hildenbrand
Quoted 167 lines · Yang Shi
Quoted 157 lines · David Hildenbrand
Quoted 129 lines · Yang Shi
Quoted 128 lines · David Hildenbrand

Let's mark exclusively mapped anonymous pages with PG_anon_exclusive as exclusive, and use that information to make GUP pins reliable and stay consistent with the page mapped into the page table even if the page table entry gets write-protected.

With that information at hand, we can extend our COW logic to always reuse anonymous pages that are exclusive. For anonymous pages that might be shared, the existing logic applies.

As already documented, PG_anon_exclusive is usually only expressive in combination with a page table entry. Especially PTE vs. PMD-mapped anonymous pages require more thought, some examples: due to mremap() we can easily have a single compound page PTE-mapped into multiple page tables exclusively in a single process -- multiple page table locks apply. Further, due to MADV_WIPEONFORK we might not necessarily write-protect all PTEs, and only some subpages might be pinned. Long story short: once PTE-mapped, we have to track information about exclusivity per sub-page, but until then, we can just track it for the compound page in the head page and not having to update a whole bunch of subpages all of the time for a simple PMD mapping of a THP.

For simplicity, this commit mostly talks about "anonymous pages", while it's for THP actually "the part of an anonymous folio referenced via a page table entry".

To not spill PG_anon_exclusive code all over the mm code-base, we let the anon rmap code to handle all PG_anon_exclusive logic it can easily handle.

If a writable, present page table entry points at an anonymous (sub)page, that (sub)page must be PG_anon_exclusive. If GUP wants to take a reliably pin (FOLL_PIN) on an anonymous page references via a present page table entry, it must only pin if PG_anon_exclusive is set for the mapped (sub)page.

This commit doesn't adjust GUP, so this is only implicitly handled for FOLL_WRITE, follow-up commits will teach GUP to also respect it for FOLL_PIN without !FOLL_WRITE, to make all GUP pins of anonymous pages fully reliable.

Whenever an anonymous page is to be shared (fork(), KSM), or when temporarily unmapping an anonymous page (swap, migration), the relevant PG_anon_exclusive bit has to be cleared to mark the anonymous page possibly shared. Clearing will fail if there are GUP pins on the page:
* For fork(), this means having to copy the page and not being able to share it. fork() protects against concurrent GUP using the PT lock and the src_mm->write_protect_seq.
* For KSM, this means sharing will fail. For swap this means, unmapping will fail, For migration this means, migration will fail early. All three cases protect against concurrent GUP using the PT lock and a proper clear/invalidate+flush of the relevant page table entry.

This fixes memory corruptions reported for FOLL_PIN | FOLL_WRITE, when a pinned page gets mapped R/O and the successive write fault ends up replacing the page instead of reusing it. It improves the situation for O_DIRECT/vmsplice/... that still use FOLL_GET instead of FOLL_PIN, if fork() is *not* involved, however swapout and fork() are still problematic. Properly using FOLL_PIN instead of FOLL_GET for these GUP users will fix the issue for them.

I. Details about basic handling

I.1. Fresh anonymous pages

page_add_new_anon_rmap() and hugepage_add_new_anon_rmap() will mark the given page exclusive via __page_set_anon_rmap(exclusive=1). As that is the mechanism fresh anonymous pages come into life (besides migration code where we copy the page->mapping), all fresh anonymous pages will start out as exclusive.

I.2. COW reuse handling of anonymous pages

When a COW handler stumbles over a (sub)page that's marked exclusive, it simply reuses it. Otherwise, the handler tries harder under page lock to detect if the (sub)page is exclusive and can be reused. If exclusive, page_move_anon_rmap() will mark the given (sub)page exclusive.

Note that hugetlb code does not yet check for PageAnonExclusive(), as it still uses the old COW logic that is prone to the COW security issue because hugetlb code cannot really tolerate unnecessary/wrong COW as huge pages are a scarce resource.

I.3. Migration handling

try_to_migrate() has to try marking an exclusive anonymous page shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. migrate_vma_collect_pmd() and __split_huge_pmd_locked() are handled similarly.

Writable migration entries implicitly point at shared anonymous pages. For readable migration entries that information is stored via a new "readable-exclusive" migration entry, specific to anonymous pages.

When restoring a migration entry in remove_migration_pte(), information about exlusivity is detected via the migration entry type, and RMAP_EXCLUSIVE is set accordingly for page_add_anon_rmap()/hugepage_add_anon_rmap() to restore that information.

I.4. Swapout handling

try_to_unmap() has to try marking the mapped page possibly shared via page_try_share_anon_rmap(). If it fails because there are GUP pins on the page, unmap fails. For now, information about exclusivity is lost. In the future, we might want to remember that information in the swap entry in some cases, however, it requires more thought, care, and a way to store that information in swap entries.

I.5. Swapin handling

do_swap_page() will never stumble over exclusive anonymous pages in the swap cache, as try_to_migrate() prohibits that. do_swap_page() always has to detect manually if an anonymous page is exclusive and has to set RMAP_EXCLUSIVE for page_add_anon_rmap() accordingly.

I.6. THP handling

__split_huge_pmd_locked() has to move the information about exclusivity from the PMD to the PTEs.

a) In case we have a readable-exclusive PMD migration entry, simply insert readable-exclusive PTE migration entries.

b) In case we have a present PMD entry and we don't want to freeze ("convert to migration entries"), simply forward PG_anon_exclusive to all sub-pages, no need to temporarily clear the bit.

Quoted 5 lines
Quoted 5 lines

c) In case we have a present PMD entry and want to freeze, handle it similar to try_to_migrate(): try marking the page shared first. In case we fail, we ignore the "freeze" instruction and simply split ordinarily. try_to_migrate() will properly fail because the THP is still mapped via PTEs.

Hi,

thanks for the review!

Quoted 4 lines

How come will try_to_migrate() fail? The afterward pvmw will find those PTEs then convert them to migration entries anyway IIUC.

It will run into that code:

Quoted 9 lines
Quoted 9 lines
mm/rmap.c +5 -0
@@ -1903,6 +1938,15 @@ static bool try_to_migrate_one(struct page *page, struct vm_area_struct *vma,
                                page_vma_mapped_walk_done(&pvmw);
                                break;
                        }
+                       VM_BUG_ON_PAGE(pte_write(pteval) && PageAnon(page) &&
+                                      !anon_exclusive, page);
+                       if (anon_exclusive &&
+                           page_try_share_anon_rmap(subpage)) {
+                               set_pte_at(mm, address, pvmw.pte, pteval);
Quoted 8 lines
Quoted 4 lines

mm/rmap.c static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, · line 1644

Quoted 4 lines
mm/rmap.c +4 -0
+                               ret = false;
+                               page_vma_mapped_walk_done(&pvmw);
+                               break;
+                       }

and similarly fail the page_try_share_anon_rmap(), at which point try_to_migrate() stops and the caller will still observe a "page_mapped() == true".

Quoted 6 lines

Thanks, I missed that. Yes, the page will still be mapped. This should trigger the VM_WARN_ON_ONCE in unmap_page(), if this change will make this happen more often, we may consider removing that warning even though it is "once" since seeing a mapped page may become a normal case (once DIO is switched to FOLL_PIN, it may be more often). Anyway we don't have to remove it right now.

Quoted 3 lines

Oh, very good catch! I wasn't able to trigger that warning in my testing so far. Interestingly, arch_unmap_one() could theoretically make this fail already and trigger the warning.

It should be very rare to trigger that warning. I just saw one report before. And arch_unmap_one() actually can't fail except for Sparc. So that warning was intended to catch some unusual issues, but this COW patchset may make it much more usual than before.

Quoted 3 lines

Apart from that warning, split_huge_page_to_list() should work as
expected: freezing the refcount will fail if still mapped and we'll remap.

Yes, that was why it was changed to a WARN from a BUG, please see commit 504e070dc08f ("mm: thp: replace DEBUG_VM BUG with VM_WARN when unmap fails for split")

Quoted 2 lines

I'll include a separate patch to just remove that VM_WARN_ON_ONCE -- thanks!

Thanks, I'm fine to remove it.

Quoted 6 lines

--
Thanks,

David / dhildenb

#27 lore raw

PATCH v2 12/15 mm/gup: disallow follow_page(FOLL_PIN)

In reply to David Hildenbrand

We want to change the way we handle R/O pins on anonymous pages that might be shared: if we detect a possibly shared anonymous page -- mapped R/O and not !PageAnonExclusive() -- we want to trigger unsharing via a page fault, resulting in an exclusive anonymous page that can be pinned reliably without getting replaced via COW on the next write fault.

However, the required page fault will be problematic for follow_page():
in contrast to ordinary GUP, follow_page() doesn't trigger faults internally. So we would have to end up failing a R/O pin via follow_page(), although there is something mapped R/O into the page table, which might be rather surprising.

We don't seem to have follow_page(FOLL_PIN) users, and it's a purely internal MM function. Let's just make our life easier and the semantics of follow_page() clearer by just disallowing FOLL_PIN for follow_page() completely.

---
 mm/gup.c     | 3 +++
 mm/hugetlb.c | 8 +++++---
 2 files changed, 8 insertions(+), 3 deletions(-)
mm/gup.c +3 -0
index 7bc1ba9ce440..ae1aa2e93b0f 100644
@@ -875,6 +875,9 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
 	if (vma_is_secretmem(vma))
 		return NULL;
 
+	if (foll_flags & FOLL_PIN)
+		return NULL;
+
 	page = follow_page_mask(vma, address, foll_flags, &ctx);
 	if (ctx.pgmap)
 		put_dev_pagemap(ctx.pgmap);
mm/hugetlb.c +5 -3
index c0e19ea5ebb5..640364a11336 100644
@@ -6692,9 +6692,11 @@ follow_huge_pmd(struct mm_struct *mm, unsigned long address,
 	spinlock_t *ptl;
 	pte_t pte;
 
-	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
-	if (WARN_ON_ONCE((flags & (FOLL_PIN | FOLL_GET)) ==
-			 (FOLL_PIN | FOLL_GET)))
+	/*
+	 * FOLL_PIN is not supported for follow_page(). Ordinary GUP goes via
+	 * follow_hugetlb_page().
+	 */
+	if (WARN_ON_ONCE(flags & FOLL_PIN))
 		return NULL;
 
 retry:
Signature · 1 line
2.35.1
#28 lore raw

PATCH v2 13/15 mm: support GUP-triggered unsharing of anonymous pages

In reply to David Hildenbrand

Whenever GUP currently ends up taking a R/O pin on an anonymous page that might be shared -- mapped R/O and !PageAnonExclusive() -- any write fault on the page table entry will end up replacing the mapped anonymous page due to COW, resulting in the GUP pin no longer being consistent with the page actually mapped into the page table.

The possible ways to deal with this situation are:
(1) Ignore and pin -- what we do right now.
(2) Fail to pin -- which would be rather surprising to callers and could break user space.
(3) Trigger unsharing and pin the now exclusive page -- reliable R/O pins.

We want to implement 3) because it provides the clearest semantics and allows for checking in unpin_user_pages() and friends for possible BUGs:
when trying to unpin a page that's no longer exclusive, clearly something went very wrong and might result in memory corruptions that might be hard to debug. So we better have a nice way to spot such issues.

To implement 3), we need a way for GUP to trigger unsharing:
FAULT_FLAG_UNSHARE. FAULT_FLAG_UNSHARE is only applicable to R/O mapped anonymous pages and resembles COW logic during a write fault. However, in contrast to a write fault, GUP-triggered unsharing will, for example, still maintain the write protection.

Let's implement FAULT_FLAG_UNSHARE by hooking into the existing write fault handlers for all applicable anonymous page types: ordinary pages, THP and hugetlb.

* If FAULT_FLAG_UNSHARE finds a R/O-mapped anonymous page that has been marked exclusive in the meantime by someone else, there is nothing to do.
* If FAULT_FLAG_UNSHARE finds a R/O-mapped anonymous page that's not marked exclusive, it will try detecting if the process is the exclusive owner. If exclusive, it can be set exclusive similar to reuse logic during write faults via page_move_anon_rmap() and there is nothing else to do; otherwise, we either have to copy and map a fresh, anonymous exclusive page R/O (ordinary pages, hugetlb), or split the THP.

This commit is heavily based on patches by Andrea.

---
 include/linux/mm_types.h |   8 +++
 mm/huge_memory.c         |  10 +++-
 mm/hugetlb.c             |  56 ++++++++++++--------
 mm/memory.c              | 107 +++++++++++++++++++++++++++------------
 4 files changed, 126 insertions(+), 55 deletions(-)
include/linux/mm_types.h +8 -0
index 0f549870da6a..7e1c7d03dcd8 100644
@@ -794,6 +794,9 @@ typedef struct {
  * @FAULT_FLAG_REMOTE: The fault is not for current task/mm.
  * @FAULT_FLAG_INSTRUCTION: The fault was during an instruction fetch.
  * @FAULT_FLAG_INTERRUPTIBLE: The fault can be interrupted by non-fatal signals.
+ * @FAULT_FLAG_UNSHARE: The fault is an unsharing request to unshare (and mark
+ *                      exclusive) a possibly shared anonymous page that is
+ *                      mapped R/O.
  *
  * About @FAULT_FLAG_ALLOW_RETRY and @FAULT_FLAG_TRIED: we can specify
  * whether we would allow page faults to retry by specifying these two
@@ -813,6 +816,10 @@ typedef struct {
  * continuous faults with flags (b).  We should always try to detect pending
  * signals before a retry to make sure the continuous page faults can still be
  * interrupted if necessary.
+ *
+ * The combination FAULT_FLAG_WRITE|FAULT_FLAG_UNSHARE is illegal.
+ * FAULT_FLAG_UNSHARE is ignored and treated like an ordinary read fault when
+ * no existing R/O-mapped anonymous page is encountered.
  */
 enum fault_flag {
 	FAULT_FLAG_WRITE =		1 << 0,
@@ -825,6 +832,7 @@ enum fault_flag {
 	FAULT_FLAG_REMOTE =		1 << 7,
 	FAULT_FLAG_INSTRUCTION =	1 << 8,
 	FAULT_FLAG_INTERRUPTIBLE =	1 << 9,
+	FAULT_FLAG_UNSHARE =		1 << 10,
 };
 
 #endif /* _LINUX_MM_TYPES_H */
mm/huge_memory.c +9 -1
index 4872e7120ee1..e7addf1aeabd 100644
@@ -1271,6 +1271,7 @@ void huge_pmd_set_accessed(struct vm_fault *vmf)
 
 vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
 {
+	const bool unshare = vmf->flags & FAULT_FLAG_UNSHARE;
 	struct vm_area_struct *vma = vmf->vma;
 	struct page *page;
 	unsigned long haddr = vmf->address & HPAGE_PMD_MASK;
@@ -1279,6 +1280,9 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
 	vmf->ptl = pmd_lockptr(vma->vm_mm, vmf->pmd);
 	VM_BUG_ON_VMA(!vma->anon_vma, vma);
 
+	VM_BUG_ON(unshare && (vmf->flags & FAULT_FLAG_WRITE));
+	VM_BUG_ON(!unshare && !(vmf->flags & FAULT_FLAG_WRITE));
+
 	if (is_huge_zero_pmd(orig_pmd))
 		goto fallback;
 
@@ -1317,7 +1321,7 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
 	}
 
 	/*
-	 * See do_wp_page(): we can only map the page writable if there are
+	 * See do_wp_page(): we can only reuse the page exclusively if there are
 	 * no additional references. Note that we always drain the LRU
 	 * pagevecs immediately after adding a THP.
 	 */
@@ -1331,6 +1335,10 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf)
 		page_move_anon_rmap(page, vma);
 		unlock_page(page);
 reuse:
+		if (unlikely(unshare)) {
+			spin_unlock(vmf->ptl);
+			return 0;
+		}
 		entry = pmd_mkyoung(orig_pmd);
 		entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma);
 		if (pmdp_set_access_flags(vma, haddr, vmf->pmd, entry, 1))
mm/hugetlb.c +35 -21

138 lines not on this page · raw message

mm/memory.c +74 -33

218 lines not on this page · raw message

Signature · 1 line
2.35.1
#29 lore raw

Re: [PATCH v2 13/15] mm: support GUP-triggered unsharing of anonymous pages

Quoted 20 lines · David Hildenbrand

Whenever GUP currently ends up taking a R/O pin on an anonymous page that might be shared -- mapped R/O and !PageAnonExclusive() -- any write fault on the page table entry will end up replacing the mapped anonymous page due to COW, resulting in the GUP pin no longer being consistent with the page actually mapped into the page table.

The possible ways to deal with this situation are:
(1) Ignore and pin -- what we do right now.
(2) Fail to pin -- which would be rather surprising to callers and could break user space.
(3) Trigger unsharing and pin the now exclusive page -- reliable R/O pins.

We want to implement 3) because it provides the clearest semantics and allows for checking in unpin_user_pages() and friends for possible BUGs:
when trying to unpin a page that's no longer exclusive, clearly something went very wrong and might result in memory corruptions that might be hard to debug. So we better have a nice way to spot such issues.

Quoted 5 lines

To implement 3), we need a way for GUP to trigger unsharing:
FAULT_FLAG_UNSHARE. FAULT_FLAG_UNSHARE is only applicable to R/O mapped anonymous pages and resembles COW logic during a write fault. However, in contrast to a write fault, GUP-triggered unsharing will, for example, still maintain the write protection.

Given the way this series has developed you might want to call this FAULT_FLAG_MAKE_ANON_EXCLUSIVE

Which strikes me as more directly connected to what it is trying to do.

Jason

#30 lore raw

Re: [PATCH v2 13/15] mm: support GUP-triggered unsharing of anonymous pages

Quoted 27 lines · Jason Gunthorpe
Quoted 20 lines · David Hildenbrand

Whenever GUP currently ends up taking a R/O pin on an anonymous page that might be shared -- mapped R/O and !PageAnonExclusive() -- any write fault on the page table entry will end up replacing the mapped anonymous page due to COW, resulting in the GUP pin no longer being consistent with the page actually mapped into the page table.

The possible ways to deal with this situation are:
(1) Ignore and pin -- what we do right now.
(2) Fail to pin -- which would be rather surprising to callers and could break user space.
(3) Trigger unsharing and pin the now exclusive page -- reliable R/O pins.

We want to implement 3) because it provides the clearest semantics and allows for checking in unpin_user_pages() and friends for possible BUGs:
when trying to unpin a page that's no longer exclusive, clearly something went very wrong and might result in memory corruptions that might be hard to debug. So we better have a nice way to spot such issues.

Quoted 5 lines

To implement 3), we need a way for GUP to trigger unsharing:
FAULT_FLAG_UNSHARE. FAULT_FLAG_UNSHARE is only applicable to R/O mapped anonymous pages and resembles COW logic during a write fault. However, in contrast to a write fault, GUP-triggered unsharing will, for example, still maintain the write protection.

Quoted 5 lines

Given the way this series has developed you might want to call this FAULT_FLAG_MAKE_ANON_EXCLUSIVE

Which strikes me as more directly connected to what it is trying to do.

I thought about something similar along those lines, and I think it would apply even when extending that mechanism to anything !anon inside a MAP_PRIVATE mapping.

The whole

const bool unshare = vmf->flags & FAULT_FLAG_UNSHARE;

would get a little more verbose, though

const bool make_anon_exclusive = vmf->flags &

				 FAULT_FLAG_MAKE_ANON_EXCLUSIVE;

No strong opinion, any other opinions or alternatives?

Signature · 3 lines
Thanks,

David / dhildenb
#31 lore raw

Re: [PATCH v2 13/15] mm: support GUP-triggered unsharing of anonymous pages

Quoted 34 lines · David Hildenbrand
Quoted 27 lines · Jason Gunthorpe
Quoted 20 lines · David Hildenbrand

Whenever GUP currently ends up taking a R/O pin on an anonymous page that might be shared -- mapped R/O and !PageAnonExclusive() -- any write fault on the page table entry will end up replacing the mapped anonymous page due to COW, resulting in the GUP pin no longer being consistent with the page actually mapped into the page table.

The possible ways to deal with this situation are:
(1) Ignore and pin -- what we do right now.
(2) Fail to pin -- which would be rather surprising to callers and could break user space.
(3) Trigger unsharing and pin the now exclusive page -- reliable R/O pins.

We want to implement 3) because it provides the clearest semantics and allows for checking in unpin_user_pages() and friends for possible BUGs:
when trying to unpin a page that's no longer exclusive, clearly something went very wrong and might result in memory corruptions that might be hard to debug. So we better have a nice way to spot such issues.

Quoted 5 lines

To implement 3), we need a way for GUP to trigger unsharing:
FAULT_FLAG_UNSHARE. FAULT_FLAG_UNSHARE is only applicable to R/O mapped anonymous pages and resembles COW logic during a write fault. However, in contrast to a write fault, GUP-triggered unsharing will, for example, still maintain the write protection.

Quoted 5 lines

Given the way this series has developed you might want to call this FAULT_FLAG_MAKE_ANON_EXCLUSIVE

Which strikes me as more directly connected to what it is trying to do.

Quoted 7 lines

I thought about something similar along those lines, and I think it would apply even when extending that mechanism to anything !anon inside a MAP_PRIVATE mapping.

The whole

const bool unshare = vmf->flags & FAULT_FLAG_UNSHARE;

I think the extra words are worthwhile, share makes me think about MAP_SHARED as we don't really use shared anywhere else FWICT..

Jason

#32 lore raw

Re: [PATCH v2 13/15] mm: support GUP-triggered unsharing of anonymous pages

Quoted 39 lines · Jason Gunthorpe
Quoted 38 lines · David Hildenbrand
Quoted 27 lines · Jason Gunthorpe
Quoted 20 lines · David Hildenbrand

Whenever GUP currently ends up taking a R/O pin on an anonymous page that might be shared -- mapped R/O and !PageAnonExclusive() -- any write fault on the page table entry will end up replacing the mapped anonymous page due to COW, resulting in the GUP pin no longer being consistent with the page actually mapped into the page table.

The possible ways to deal with this situation are:
(1) Ignore and pin -- what we do right now.
(2) Fail to pin -- which would be rather surprising to callers and could break user space.
(3) Trigger unsharing and pin the now exclusive page -- reliable R/O pins.

We want to implement 3) because it provides the clearest semantics and allows for checking in unpin_user_pages() and friends for possible BUGs:
when trying to unpin a page that's no longer exclusive, clearly something went very wrong and might result in memory corruptions that might be hard to debug. So we better have a nice way to spot such issues.

Quoted 5 lines

To implement 3), we need a way for GUP to trigger unsharing:
FAULT_FLAG_UNSHARE. FAULT_FLAG_UNSHARE is only applicable to R/O mapped anonymous pages and resembles COW logic during a write fault. However, in contrast to a write fault, GUP-triggered unsharing will, for example, still maintain the write protection.

Quoted 5 lines

Given the way this series has developed you might want to call this FAULT_FLAG_MAKE_ANON_EXCLUSIVE

Which strikes me as more directly connected to what it is trying to do.

I thought about something similar along those lines, and I think it would apply even when extending that mechanism to anything !anon inside a MAP_PRIVATE mapping.

Quoted 6 lines
Quoted 3 lines

The whole

const bool unshare = vmf->flags & FAULT_FLAG_UNSHARE;

I think the extra words are worthwhile, share makes me think about MAP_SHARED as we don't really use shared anywhere else FWICT..

Yeah, my point would be that you can only "unshare" in MAP_PRIVATE (!MAP_SHARED) :)

Signature · 3 lines
Thanks,

David / dhildenb
#33 lore raw

PATCH v2 14/15 mm/gup: trigger FAULT_FLAG_UNSHARE when R/O-pinning a possibly shared anonymous page

In reply to David Hildenbrand

Whenever GUP currently ends up taking a R/O pin on an anonymous page that might be shared -- mapped R/O and !PageAnonExclusive() -- any write fault on the page table entry will end up replacing the mapped anonymous page due to COW, resulting in the GUP pin no longer being consistent with the page actually mapped into the page table.

The possible ways to deal with this situation are:
(1) Ignore and pin -- what we do right now.
(2) Fail to pin -- which would be rather surprising to callers and could break user space.
(3) Trigger unsharing and pin the now exclusive page -- reliable R/O pins.

Let's implement 3) because it provides the clearest semantics and allows for checking in unpin_user_pages() and friends for possible BUGs:
when trying to unpin a page that's no longer exclusive, clearly something went very wrong and might result in memory corruptions that might be hard to debug. So we better have a nice way to spot such issues.

This change implies that whenever user space *wrote* to a private mapping (IOW, we have an anonymous page mapped), that GUP pins will always remain consistent: reliable R/O GUP pins of anonymous pages.

As a side note, this commit fixes the COW security issue for hugetlb with FOLL_PIN as documented in:
https://lore.kernel.org/r/[email protected]
The vmsplice reproducer still applies, because vmsplice uses FOLL_GET instead of FOLL_PIN.

Note that follow_huge_pmd() doesn't apply because we cannot end up in there with FOLL_PIN.

This commit is heavily based on prototype patches by Andrea.

---
 include/linux/mm.h | 39 +++++++++++++++++++++++++++++++++++++++
 mm/gup.c           | 42 +++++++++++++++++++++++++++++++++++++++---
 mm/huge_memory.c   |  3 +++
 mm/hugetlb.c       | 27 ++++++++++++++++++++++++---
 4 files changed, 105 insertions(+), 6 deletions(-)
include/linux/mm.h +39 -0
index 63ee06001189..45fe97213a27 100644
@@ -3003,6 +3003,45 @@ static inline int vm_fault_to_errno(vm_fault_t vm_fault, int foll_flags)
 	return 0;
 }
 
+/*
+ * Indicates for which pages that are write-protected in the page table,
+ * whether GUP has to trigger unsharing via FAULT_FLAG_UNSHARE such that the
+ * GUP pin will remain consistent with the pages mapped into the page tables
+ * of the MM.
+ *
+ * Temporary unmapping of PageAnonExclusive() pages or clearing of
+ * PageAnonExclusive() has to protect against concurrent GUP:
+ * * Ordinary GUP: Using the PT lock
+ * * GUP-fast and fork(): mm->write_protect_seq
+ * * GUP-fast and KSM or temporary unmapping (swap, migration):
+ *   clear/invalidate+flush of the page table entry
+ *
+ * Must be called with the (sub)page that's actually referenced via the
+ * page table entry, which might not necessarily be the head page for a
+ * PTE-mapped THP.
+ */
+static inline bool gup_must_unshare(unsigned int flags, struct page *page)
+{
+	/*
+	 * FOLL_WRITE is implicitly handled correctly as the page table entry
+	 * has to be writable -- and if it references (part of) an anonymous
+	 * folio, that part is required to be marked exclusive.
+	 */
+	if ((flags & (FOLL_WRITE | FOLL_PIN)) != FOLL_PIN)
+		return false;
+	/*
+	 * Note: PageAnon(page) is stable until the page is actually getting
+	 * freed.
+	 */
+	if (!PageAnon(page))
+		return false;
+	/*
+	 * Note that PageKsm() pages cannot be exclusive, and consequently,
+	 * cannot get pinned.
+	 */
+	return !PageAnonExclusive(page);
+}
+
 typedef int (*pte_fn_t)(pte_t *pte, unsigned long addr, void *data);
 extern int apply_to_page_range(struct mm_struct *mm, unsigned long address,
 			       unsigned long size, pte_fn_t fn, void *data);
mm/gup.c +39 -3

96 lines not on this page · raw message

mm/huge_memory.c +3 -0
index e7addf1aeabd..1896388b1e30 100644
@@ -1389,6 +1389,9 @@ struct page *follow_trans_huge_pmd(struct vm_area_struct *vma,
 	page = pmd_page(*pmd);
 	VM_BUG_ON_PAGE(!PageHead(page) && !is_zone_device_page(page), page);
 
+	if (!pmd_write(*pmd) && gup_must_unshare(flags, page))
+		return ERR_PTR(-EMLINK);
+
 	if (!try_grab_page(page, flags))
 		return ERR_PTR(-ENOMEM);
 
mm/hugetlb.c +24 -3
index 8a40f3f5867a..c40478b5bead 100644
@@ -5954,6 +5954,25 @@ static void record_subpages_vmas(struct page *page, struct vm_area_struct *vma,
 	}
 }
 
+static inline bool __follow_hugetlb_must_fault(unsigned int flags, pte_t *pte,
+					       bool *unshare)
+{
+	pte_t pteval = huge_ptep_get(pte);
+
+	*unshare = false;
+	if (is_swap_pte(pteval))
+		return true;
+	if (huge_pte_write(pteval))
+		return false;
+	if (flags & FOLL_WRITE)
+		return true;
+	if (gup_must_unshare(flags, pte_page(pteval))) {
+		*unshare = true;
+		return true;
+	}
+	return false;
+}
+
 long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
 			 struct page **pages, struct vm_area_struct **vmas,
 			 unsigned long *position, unsigned long *nr_pages,
@@ -5968,6 +5987,7 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
 	while (vaddr < vma->vm_end && remainder) {
 		pte_t *pte;
 		spinlock_t *ptl = NULL;
+		bool unshare = false;
 		int absent;
 		struct page *page;
 
@@ -6018,9 +6038,8 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
 		 * both cases, and because we can't follow correct pages
 		 * directly from any kind of swap entries.
 		 */
-		if (absent || is_swap_pte(huge_ptep_get(pte)) ||
-		    ((flags & FOLL_WRITE) &&
-		      !huge_pte_write(huge_ptep_get(pte)))) {
+		if (absent ||
+		    __follow_hugetlb_must_fault(flags, pte, &unshare)) {
 			vm_fault_t ret;
 			unsigned int fault_flags = 0;
 
@@ -6028,6 +6047,8 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
 				spin_unlock(ptl);
 			if (flags & FOLL_WRITE)
 				fault_flags |= FAULT_FLAG_WRITE;
+			else if (unshare)
+				fault_flags |= FAULT_FLAG_UNSHARE;
 			if (locked)
 				fault_flags |= FAULT_FLAG_ALLOW_RETRY |
 					FAULT_FLAG_KILLABLE;
Signature · 1 line
2.35.1
#34 lore raw

PATCH v2 15/15 mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning

In reply to David Hildenbrand

Let's verify when (un)pinning anonymous pages that we always deal with exclusive anonymous pages, which guarantees that we'll have a reliable PIN, meaning that we cannot end up with the GUP pin being inconsistent with he pages mapped into the page tables due to a COW triggered by a write fault.

When pinning pages, after conditionally triggering GUP unsharing of possibly shared anonymous pages, we should always only see exclusive anonymous pages. Note that anonymous pages that are mapped writable must be marked exclusive, otherwise we'd have a BUG.

When pinning during ordinary GUP, simply add a check after our conditional GUP-triggered unsharing checks. As we know exactly how the page is mapped, we know exactly in which page we have to check for PageAnonExclusive().

When pinning via GUP-fast we have to be careful, because we can race with fork(): verify only after we made sure via the seqcount that we didn't race with concurrent fork() that we didn't end up pinning a possibly shared anonymous page.

Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can result in random memory corruptions, which we really want to catch.

With only the pinned pages at hand and not the actual page table entries we have to be a bit careful: hugetlb pages are always mapped via a single logical page table entry referencing the head page and PG_anon_exclusive of the head page applies. Anon THP are a bit more complicated, because we might have obtained the page reference either via a PMD or a PTE -- depending on the mapping type we either have to check PageAnonExclusive of the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies: as we don't know and to make our life easier, check that either is set.

Take care to not verify in case we're unpinning during GUP-fast because we detected concurrent fork(): we might stumble over an anonymous page that is now shared.

---
 mm/gup.c         | 58 +++++++++++++++++++++++++++++++++++++++++++++++-
 mm/huge_memory.c |  3 +++
 mm/hugetlb.c     |  3 +++
 3 files changed, 63 insertions(+), 1 deletion(-)
mm/gup.c +57 -1
index 92dcd92f9d67..72e39b77da10 100644
@@ -45,6 +45,38 @@ static void hpage_pincount_sub(struct page *page, int refs)
 	atomic_sub(refs, compound_pincount_ptr(page));
 }
 
+static inline void sanity_check_pinned_pages(struct page **pages,
+					     unsigned long npages)
+{
+#ifdef CONFIG_DEBUG_VM
+	/*
+	 * We only pin anonymous pages if they are exclusive. Once pinned, we
+	 * can no longer turn them possibly shared and PageAnonExclusive() will
+	 * stick around until the page is freed.
+	 *
+	 * We'd like to verify that our pinned anonymous pages are still mapped
+	 * exclusively. The issue with anon THP is that we don't know how
+	 * they are/were mapped when pinning them. However, for anon
+	 * THP we can assume that either the given page (PTE-mapped THP) or
+	 * the head page (PMD-mapped THP) should be PageAnonExclusive(). If
+	 * neither is the case, there is certainly something wrong.
+	 */
+	for (; npages; npages--, pages++) {
+		struct page *page = *pages;
+		struct page *head = compound_head(page);
+
+		if (!PageAnon(head))
+			continue;
+		if (!PageCompound(head) || PageHuge(head))
+			VM_BUG_ON_PAGE(!PageAnonExclusive(head), page);
+		else
+			/* Either a PTE-mapped or a PMD-mapped THP. */
+			VM_BUG_ON_PAGE(!PageAnonExclusive(head) &&
+				       !PageAnonExclusive(page), page);
+	}
+#endif /* CONFIG_DEBUG_VM */
+}
+
 /* Equivalent to calling put_page() @refs times. */
 static void put_page_refs(struct page *page, int refs)
 {
@@ -250,6 +282,7 @@ bool __must_check try_grab_page(struct page *page, unsigned int flags)
  */
 void unpin_user_page(struct page *page)
 {
+	sanity_check_pinned_pages(&page, 1);
 	put_compound_head(compound_head(page), 1, FOLL_PIN);
 }
 EXPORT_SYMBOL(unpin_user_page);
@@ -340,6 +373,7 @@ void unpin_user_pages_dirty_lock(struct page **pages, unsigned long npages,
 		return;
 	}
 
+	sanity_check_pinned_pages(pages, npages);
 	for_each_compound_head(index, pages, npages, head, ntails) {
 		/*
 		 * Checking PageDirty at this point may race with
@@ -404,6 +438,21 @@ void unpin_user_page_range_dirty_lock(struct page *page, unsigned long npages,
 }
 EXPORT_SYMBOL(unpin_user_page_range_dirty_lock);
 
+static void unpin_user_pages_lockless(struct page **pages, unsigned long npages)
+{
+	unsigned long index;
+	struct page *head;
+	unsigned int ntails;
+
+	/*
+	 * Don't perform any sanity checks because we might have raced with
+	 * fork() and some anonymous pages might now actually be shared --
+	 * which is why we're unpinning after all.
+	 */
+	for_each_compound_head(index, pages, npages, head, ntails)
+		put_compound_head(head, ntails, FOLL_PIN);
+}
+
 /**
  * unpin_user_pages() - release an array of gup-pinned pages.
  * @pages:  array of pages to be marked dirty and released.
@@ -426,6 +475,7 @@ void unpin_user_pages(struct page **pages, unsigned long npages)
 	 */
 	if (WARN_ON(IS_ERR_VALUE(npages)))
 		return;
+	sanity_check_pinned_pages(pages, npages);
 
 	for_each_compound_head(index, pages, npages, head, ntails)
 		put_compound_head(head, ntails, FOLL_PIN);
@@ -572,6 +622,10 @@ static struct page *follow_page_pte(struct vm_area_struct *vma,
 		page = ERR_PTR(-EMLINK);
 		goto out;
 	}
+
+	VM_BUG_ON((flags & FOLL_PIN) && PageAnon(page) &&
+		  !PageAnonExclusive(page));
+
 	/* try_grab_page() does nothing unless FOLL_GET or FOLL_PIN is set. */
 	if (unlikely(!try_grab_page(page, flags))) {
 		page = ERR_PTR(-ENOMEM);
@@ -2876,8 +2930,10 @@ static unsigned long lockless_pages_from_mm(unsigned long start,
 	 */
 	if (gup_flags & FOLL_PIN) {
 		if (read_seqcount_retry(&current->mm->write_protect_seq, seq)) {
-			unpin_user_pages(pages, nr_pinned);
+			unpin_user_pages_lockless(pages, nr_pinned);
 			return 0;
+		} else {
+			sanity_check_pinned_pages(pages, nr_pinned);
 		}
 	}
 	return nr_pinned;
mm/huge_memory.c +3 -0
index 1896388b1e30..0cc34addd911 100644
@@ -1392,6 +1392,9 @@ struct page *follow_trans_huge_pmd(struct vm_area_struct *vma,
 	if (!pmd_write(*pmd) && gup_must_unshare(flags, page))
 		return ERR_PTR(-EMLINK);
 
+	VM_BUG_ON((flags & FOLL_PIN) && PageAnon(page) &&
+		  !PageAnonExclusive(page));
+
 	if (!try_grab_page(page, flags))
 		return ERR_PTR(-ENOMEM);
 
mm/hugetlb.c +3 -0
index c40478b5bead..8ba2fdd6f2e3 100644
@@ -6090,6 +6090,9 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
 		pfn_offset = (vaddr & ~huge_page_mask(h)) >> PAGE_SHIFT;
 		page = pte_page(huge_ptep_get(pte));
 
+		VM_BUG_ON((flags & FOLL_PIN) && PageAnon(page) &&
+			  !PageAnonExclusive(page));
+
 		/*
 		 * If subpage information not requested, update counters
 		 * and skip the same_page loop below.
Signature · 1 line
2.35.1
#35 lore raw

Re: [PATCH v2 15/15] mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning

Quoted 49 lines · David Hildenbrand

Let's verify when (un)pinning anonymous pages that we always deal with exclusive anonymous pages, which guarantees that we'll have a reliable PIN, meaning that we cannot end up with the GUP pin being inconsistent with he pages mapped into the page tables due to a COW triggered by a write fault.

When pinning pages, after conditionally triggering GUP unsharing of possibly shared anonymous pages, we should always only see exclusive anonymous pages. Note that anonymous pages that are mapped writable must be marked exclusive, otherwise we'd have a BUG.

When pinning during ordinary GUP, simply add a check after our conditional GUP-triggered unsharing checks. As we know exactly how the page is mapped, we know exactly in which page we have to check for PageAnonExclusive().

When pinning via GUP-fast we have to be careful, because we can race with fork(): verify only after we made sure via the seqcount that we didn't race with concurrent fork() that we didn't end up pinning a possibly shared anonymous page.

Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can result in random memory corruptions, which we really want to catch.

With only the pinned pages at hand and not the actual page table entries we have to be a bit careful: hugetlb pages are always mapped via a single logical page table entry referencing the head page and PG_anon_exclusive of the head page applies. Anon THP are a bit more complicated, because we might have obtained the page reference either via a PMD or a PTE -- depending on the mapping type we either have to check PageAnonExclusive of the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies: as we don't know and to make our life easier, check that either is set.

Take care to not verify in case we're unpinning during GUP-fast because we detected concurrent fork(): we might stumble over an anonymous page that is now shared.

 mm/gup.c         | 58 +++++++++++++++++++++++++++++++++++++++++++++++-
 mm/huge_memory.c |  3 +++
 mm/hugetlb.c     |  3 +++
 3 files changed, 63 insertions(+), 1 deletion(-)
mm/gup.c +1 -0
index 92dcd92f9d67..72e39b77da10 100644
+++ b/mm/gup.c

mm/gup.c static void hpage_pincount_sub(struct page *page, int refs) · line 51

Quoted 7 lines
mm/gup.c +4 -0
@@ -45,6 +45,38 @@ static void hpage_pincount_sub(struct page *page, int refs)
 	atomic_sub(refs, compound_pincount_ptr(page));
 }
 
+static inline void sanity_check_pinned_pages(struct page **pages,
+					     unsigned long npages)
+{
+#ifdef CONFIG_DEBUG_VM

Perhaps:

if (!IS_ENABLED(CONFIG_DEBUG_VM))
return;

So this gets compilation coverage

Jason

#36 lore raw

Re: [PATCH v2 15/15] mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning

Quoted 60 lines · Jason Gunthorpe
Quoted 49 lines · David Hildenbrand

Let's verify when (un)pinning anonymous pages that we always deal with exclusive anonymous pages, which guarantees that we'll have a reliable PIN, meaning that we cannot end up with the GUP pin being inconsistent with he pages mapped into the page tables due to a COW triggered by a write fault.

When pinning pages, after conditionally triggering GUP unsharing of possibly shared anonymous pages, we should always only see exclusive anonymous pages. Note that anonymous pages that are mapped writable must be marked exclusive, otherwise we'd have a BUG.

When pinning during ordinary GUP, simply add a check after our conditional GUP-triggered unsharing checks. As we know exactly how the page is mapped, we know exactly in which page we have to check for PageAnonExclusive().

When pinning via GUP-fast we have to be careful, because we can race with fork(): verify only after we made sure via the seqcount that we didn't race with concurrent fork() that we didn't end up pinning a possibly shared anonymous page.

Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can result in random memory corruptions, which we really want to catch.

With only the pinned pages at hand and not the actual page table entries we have to be a bit careful: hugetlb pages are always mapped via a single logical page table entry referencing the head page and PG_anon_exclusive of the head page applies. Anon THP are a bit more complicated, because we might have obtained the page reference either via a PMD or a PTE -- depending on the mapping type we either have to check PageAnonExclusive of the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies: as we don't know and to make our life easier, check that either is set.

Take care to not verify in case we're unpinning during GUP-fast because we detected concurrent fork(): we might stumble over an anonymous page that is now shared.

 mm/gup.c         | 58 +++++++++++++++++++++++++++++++++++++++++++++++-
 mm/huge_memory.c |  3 +++
 mm/hugetlb.c     |  3 +++
 3 files changed, 63 insertions(+), 1 deletion(-)
mm/gup.c +1 -0
index 92dcd92f9d67..72e39b77da10 100644
+++ b/mm/gup.c

mm/gup.c static void hpage_pincount_sub(struct page *page, int refs) · line 51

Quoted 7 lines
mm/gup.c +4 -0
@@ -45,6 +45,38 @@ static void hpage_pincount_sub(struct page *page, int refs)
 	atomic_sub(refs, compound_pincount_ptr(page));
 }
 
+static inline void sanity_check_pinned_pages(struct page **pages,
+					     unsigned long npages)
+{
+#ifdef CONFIG_DEBUG_VM

Perhaps:

Quoted 4 lines

if (!IS_ENABLED(CONFIG_DEBUG_VM))
return;

So this gets compilation coverage

Makes sense, that code should compile just fine without CONFIG_DEBUG_VM. Thanks!

Signature · 3 lines
Thanks,

David / dhildenb
#37 lore raw
In reply to David Hildenbrand
Quoted 4 lines · David Hildenbrand

This series fixes memory corruptions when a GUP pin (FOLL_PIN) was taken on an anonymous page and COW logic fails to detect exclusivity of the page to then replacing the anonymous page by a copy in the page table: The GUP pin lost synchronicity with the pages mapped into the page tables.

I don't know much about most of this part of the mm, but nothing here gave me serious pause. We are trying to run this series in our RDMA testing. Will let you know.

Thanks,
Jason