linux-kernel - [PATCH 0/3] Fix the workqueue and lockdep annotation issue

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: <1504578554-4137-1-git-send-email-byungchul.park@lge.com>
Date:   Tue,  5 Sep 2017 11:29:11 +0900
From:   Byungchul Park <byungchul.park@....com>
To:     tj@...nel.org, johannes.berg@...el.com, peterz@...radead.org,
        mingo@...nel.org
Cc:     tglx@...utronix.de, oleg@...hat.com, david@...morbit.com,
        linux-kernel@...r.kernel.org, kernel-team@....com
Subject: [PATCH 0/3] Fix the workqueue and lockdep annotation issue

We introduced the following commit to detect deadlocks caused by
wait_for_completion() in flush_{workqueue, work}() and other locks. But
now LOCKDEP_COMPLETIONS is introduced, such works are automatically done
by LOCKDEP_COMPLETIONS.

   commit 4e6045f134784f4b158b3c0f7a282b04bd816887
   workqueue: debug flushing deadlocks with lockdep

However, LOCKDEP_COMPLETIONS has limitation that it cannot report
deadlocks at the real time. So we have to leave some annotations for now,
but some redundant acquisitions can be removed. And we need to make
workqueue code co-work with LOCKDEP_COMPLETIONS.

Byungchul Park (3):
  lockdep: Use enum type on hlock->read instead of magic number
  lockdep: Introduce lock_acquire_might()
  lockdep: Remove unnecessary acquisitions wrt workqueue flush

 include/linux/lockdep.h  |  2 ++
 kernel/locking/lockdep.c | 56 ++++++++++++++++++++++++++++++++++--------------
 kernel/workqueue.c       | 18 +++++++---------
 3 files changed, 50 insertions(+), 26 deletions(-)

-- 
1.9.1