[PATCH v2 00/19] maple_tree: lock checking and clean ups

[Liam R. Howlett (Oracle)]

The goals of this series are:
1. Lock issue detection
  A number of syzbot reports are incorrectly pointing to the mm exit as
  a source of the locking error.  The first three patches attempt to
  help users detect errors in their locking - but they still have to use
  LOCKDEP.  I guess it's still down to hope and prayers.

2. Documentation fixes
  The documentation was lacking clarity, there are updates to try and
  help the users, especially around the erase() cases.

3. Two benign issues
  The cyclic allocator may have a race, although no in-kernel user can
  hit it.
  The erase functions may cause allocation issues if used with the
  incorrect locking type, but none are present in-tree.

Beyond these goals there are some test fixes, some general speed-up
patches targeting extra work and cycles, and dropping dead code.

v1: https://lore.kernel.org/all/20260629144145.1572283-1-liam@infradead.org/
Changes since v1:
 - Relocated mas_set_parent_slots() lower so that mt_slots_locked() for rcu dereferencing - Thanks kernel build bot
 - Made mas_lockdep_map() static - Thanks kernel build bot
 - Added mas_lock_check() to mas_may_activate() when restoring a node to active - Thanks sashiko
 - Changed mas_lock_check() aliasing to use mt_locked() - Thanks sashiko
 - Updated __lock_unpin_lock() underflow check - Thanks sashiko
 - Updated checking the lock sequence based on the lockdep map - Thanks sashiko
 - Add a comment to mas_wr_node_store() to explain why the last slot isn't cleared. - Thanks sashiko
 - Drop mas_set_parent() function as it is no longer used - Thanks sashiko
 - Clear the mas_alloc_cyclic() wrap flag after storage - Thanks sashiko
 - Add might_allocate() to the mtree_erase() call - Thanks Jason
 - Added one more cleanup for making maple state walkable.

Liam R. Howlett (Oracle) (19):
  maple_tree: Add rcu locking check when LOCKDEP is enabled
  locking/lockdep: Add sequence counter to held_lock
  maple_tree: Add write lock checking with lockdep sequence numbers
  maple_tree: Documentation fix
  maple_tree: Drop dead code from mas_extend_spanning_null()
  maple_tree: Drop MAPLE_ALLOC_SLOTS
  maple_tree: Clarify comments on mas_nomem()
  maple_tree: Use prefetched value in mas_wr_store_type()
  maple_tree: Optimise mas_wr_node_store() when not in rcu mode
  maple_tree: micro optimisation of mas_wr_store_type()
  maple_tree: Add bulk parent set helper
  maple_tree: Catch race in mas_alloc_cyclic()
  maple_tree: Document that erase may use GFP_KERNEL for allocations
  maple_tree: WARN_ON_ONCE when allocations fail
  maple_tree: Document erase and allocations better
  maple_tree: Change two GFP flags in tests
  maple_tree: Fix argument name in header
  maple_tree: Avoid extra gap calculation
  maple_tree: Add helper mas_make_walkable()

 Documentation/core-api/maple_tree.rst |  21 +-
 include/linux/lockdep.h               |   3 +
 include/linux/lockdep_types.h         |   3 +-
 include/linux/maple_tree.h            |  10 +-
 include/linux/sched.h                 |   1 +
 kernel/locking/lockdep.c              |  58 ++++-
 lib/maple_tree.c                      | 353 ++++++++++++++++++--------
 tools/testing/radix-tree/maple.c      |   4 +-
 8 files changed, 326 insertions(+), 127 deletions(-)

-- 
2.47.3