linux-kernel - [GIT PULL] regmap fixes for v6.5-rc2

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20230721193227.C8F05C433C9@smtp.kernel.org>
Date:   Fri, 21 Jul 2023 20:32:19 +0100
From:   Mark Brown <broonie@...nel.org>
To:     Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     linux-kernel@...r.kernel.org, Mark Brown <broonie@...nel.org>
Subject: [GIT PULL] regmap fixes for v6.5-rc2

The following changes since commit fdf0eaf11452d72945af31804e2a1048ee1b574c:

  Linux 6.5-rc2 (2023-07-16 15:10:37 -0700)

are available in the Git repository at:

  https://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap.git tags/regmap-fix-v6.5-rc2

for you to fetch changes up to a9e26169cfda651802f88262a315146fbe4bc74c:

  regmap: Disable locking for RBTREE and MAPLE unit tests (2023-07-20 13:57:01 +0100)

----------------------------------------------------------------
regmap: Fixes for v6.5

Three fixes here:

 - The issues with accounting for register and padding length on raw
   buses turn out to be quite widespread in custom buses, in order to
   avoid disturbing anything drop the initial fixes and fall back to
   a point fix in the SMBus code where the issue was originally noticed,
   a more substantial refactoring of the API which ensures that all
   buses make the same assumptions will follow.

 - The generic regcache code had been forcing on async I/O which did
   not work with the new maple tree sync code when used with SPI.  Since
   that was mainly for the rbtree cache and the assumptions about hardware
   that drove the choice are probably not true any more fix this by pushing
   the enablement of async down into the rbtree code, probably this also
   makes cache syncs for systems faster though it's not the point.

 - The test code was triggering use of the rbtree and maple tree caches
   with dynamic allocation of nodes since all the testing is with RAM
   backed caches with no I/O performance issues.  Just disable the
   locking in the tests to avoid triggering warnings when allocation
   debugging is turned on, it's not really what's being tested.

----------------------------------------------------------------
Guenter Roeck (1):
      regmap: Disable locking for RBTREE and MAPLE unit tests

Mark Brown (3):
      regmap: Drop initial version of maximum transfer length fixes
      regmap: Account for register length in SMBus I/O limits
      regcache: Push async I/O request down into the rbtree cache

 drivers/base/regmap/regcache-rbtree.c | 4 ++++
 drivers/base/regmap/regcache.c        | 3 ---
 drivers/base/regmap/regmap-i2c.c      | 8 ++++----
 drivers/base/regmap/regmap-kunit.c    | 5 +++++
 drivers/base/regmap/regmap-spi-avmm.c | 2 +-
 drivers/base/regmap/regmap.c          | 6 ++----
 6 files changed, 16 insertions(+), 12 deletions(-)