[Deepin-Kernel-SIG] [linux 6.6.y] [Upstream] [PATCH v3 00/10] mm: convert mm counter to take a folio

[opsiff]

Link: https://lore.kernel.org/all/20240111152429.3374566-1-willy@infradead.org/T/#u

Make sure all mm_counter() and mm_counter_file() callers have a folio,
then convert mm counter functions to take a folio, which saves some
compound_head() calls.

v3:

• Rebase on top of my similar patch to add pfn_swap_entry_folio()
• Various wordsmithing & notes about where we're actually saving
  calls to compound_head().

v2:

• rebase on v6.7, since most of mm counter callers already with a folio,
  drop some unneeded changes.

v1:

• rename should_zap_page() to should_zap_folio(), per Matthew Wilcox
• pass page to page_remove_rmap/page_try_dup_anon_rmap for
  device private page, per Matthew Wilcox

Kefeng Wang (7):
s390: use pfn_swap_entry_folio() in ptep_zap_swap_entry()
mm: use pfn_swap_entry_folio() in __split_huge_pmd_locked()
mm: use pfn_swap_entry_to_folio() in zap_huge_pmd()
mm: use pfn_swap_entry_folio() in copy_nonpresent_pte()
mm: Convert to should_zap_page() to should_zap_folio()
mm: convert mm_counter() to take a folio
mm: convert mm_counter_file() to take a folio

Matthew Wilcox (Oracle) (3):
mm: Add pfn_swap_entry_folio()
proc: Use pfn_swap_entry_folio where obvious
mprotect: Use pfn_swap_entry_folio

arch/s390/mm/pgtable.c | 4 ++--
fs/proc/task_mmu.c | 4 ++--
include/linux/mm.h | 12 +++++-----
include/linux/swapops.h | 13 +++++++++++
kernel/events/uprobes.c | 2 +-
mm/filemap.c | 2 +-
mm/huge_memory.c | 23 ++++++++++---------
mm/khugepaged.c | 4 ++--
mm/memory.c | 49 ++++++++++++++++++++++-------------------
mm/mprotect.c | 4 ++--
mm/rmap.c | 10 ++++-----
mm/userfaultfd.c | 2 +-
12 files changed, 74 insertions(+), 55 deletions(-)

--
2.43.0

Summary by Sourcery

Convert memory management accounting and swap migration helpers to operate on folios instead of pages, and propagate these changes through mm, arch, and proc code paths to reduce compound_head() usage and align with folio-based APIs.

Enhancements:

• Introduce pfn_swap_entry_folio() for accessing folios from PFN-based swap entries with migration-entry safety checks.
• Update mm_counter() and mm_counter_file() and all their callers to take folio arguments, standardizing RSS accounting on folios.
• Rename and adapt should_zap_page() to should_zap_folio() and adjust zap paths and migration handling to use folio-aware checks.
• Replace pfn_swap_entry_to_page() usage with pfn_swap_entry_folio() in huge page, rmap, mprotect, userfaultfd, proc, uprobe, and s390 MM code paths for folio-centric handling.

[sourcery-ai]

Reviewer's Guide

This PR introduces folio-aware mm counter helpers and a pfn_swap_entry_folio() helper, then systematically updates migration, zap, rmap, THP, userfaultfd, proc, uprobes and s390 code paths to operate on folios instead of pages, reducing compound_head() usage and aligning swap/migration handling with folio-centric APIs.

Flow diagram for folio-based mm counter and swap entry handling

flowchart LR
  entry[swp_entry_t entry]
  pfn["pfn_swap_entry_folio(entry)"]
  folio[folio]
  anon["folio_test_anon(folio)"]
  swapb["folio_test_swapbacked(folio)"]
  mcf["mm_counter_file(folio)"]
  mc["mm_counter(folio)"]
  anonrss[MM_ANONPAGES]
  filepages[MM_FILEPAGES]
  shmem[MM_SHMEMPAGES]

  entry --> pfn --> folio
  folio --> anon
  anon -- true --> anonrss
  anon -- false --> swapb
  swapb -- true --> mcf --> shmem
  swapb -- false --> mcf --> filepages
  folio --> mc --> anonrss
  folio --> mc --> mcf

Loading

File-Level Changes

Change	Details	Files
Introduce pfn_swap_entry_folio() and update swap/migration users to work on folios rather than pages.	Add pfn_swap_entry_folio() helper that returns a folio from a pfn-backed swap entry and asserts the folio is locked for migration entries. Replace pfn_swap_entry_to_page()+page_folio() sequences with pfn_swap_entry_folio() in migration and swap-handling paths. Update checks like PageAnon(...) to folio_test_anon(...) where swap entries are converted to folios.	include/linux/swapops.h mm/huge_memory.c mm/mprotect.c mm/filemap.c mm/memory.c mm/khugepaged.c mm/rmap.c fs/proc/task_mmu.c arch/s390/mm/pgtable.c
Convert mm_counter() and mm_counter_file() to take a struct folio, and adjust all call sites in mm and related subsystems.	Redefine mm_counter_file() and mm_counter() to take a folio and use folio_test_swapbacked()/folio_test_anon() internally. Update inc_mm_counter(), dec_mm_counter(), add_mm_counter(), and RSS accounting sites to pass folios instead of pages. Clean up call sites that previously needed PageAnon()/PageSwapBacked() on struct page and now rely on folio helpers.	include/linux/mm.h mm/memory.c mm/huge_memory.c mm/khugepaged.c mm/rmap.c mm/userfaultfd.c kernel/events/uprobes.c arch/s390/mm/pgtable.c
Rename should_zap_page() to should_zap_folio() and refactor zap paths to operate purely on folios for decision-making and RSS accounting.	Introduce should_zap_folio() that accepts a folio pointer (possibly NULL) and bases decisions on folio_test_anon(). Adjust zap_pte_range() and related zap logic to obtain folios up front and call should_zap_folio() instead of should_zap_page(). Update RSS updates in zap paths to use mm_counter(folio) and remove now-unnecessary page-to-folio conversions.	mm/memory.c

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[deepin-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please ask for approval from opsiff. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• deepin/OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[sourcery-ai]

Hey - I've left some high level feedback:

• In zap_huge_pmd(), the non-present huge PMD path when !thp_migration_supported() leaves folio as NULL but still calls folio_test_anon(folio) and mm_counter_file(folio), which will dereference NULL; this path should either return early, set folio appropriately, or guard those calls.
• The new pfn_swap_entry_folio() BUG_ON(!folio_test_locked(folio)) tightens the locking requirements for migration entries; please double-check all call sites (e.g. pagemap_* helpers, migration_entry_wait_on_locked, zap_pte_range) to ensure the folio is guaranteed locked there or consider weakening/annotating the invariant if that’s not universally true.

Prompt for AI Agents

Please address the comments from this code review:

## Overall Comments
- In zap_huge_pmd(), the non-present huge PMD path when !thp_migration_supported() leaves folio as NULL but still calls folio_test_anon(folio) and mm_counter_file(folio), which will dereference NULL; this path should either return early, set folio appropriately, or guard those calls.
- The new pfn_swap_entry_folio() BUG_ON(!folio_test_locked(folio)) tightens the locking requirements for migration entries; please double-check all call sites (e.g. pagemap_* helpers, migration_entry_wait_on_locked, zap_pte_range) to ensure the folio is guaranteed locked there or consider weakening/annotating the invariant if that’s not universally true.

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}