linux-kernel - [PATCH AUTOSEL 4.18 003/126] locking/lockdep: Fix debug

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

Message-Id: <20181031230754.29029-3-sashal@kernel.org>
Date:   Wed, 31 Oct 2018 19:05:51 -0400
From:   Sasha Levin <sashal@...nel.org>
To:     stable@...r.kernel.org, linux-kernel@...r.kernel.org
Cc:     Waiman Long <longman@...hat.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        "Paul E . McKenney" <paulmck@...ux.vnet.ibm.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Will Deacon <will.deacon@....com>,
        Ingo Molnar <mingo@...nel.org>, Sasha Levin <sashal@...nel.org>
Subject: [PATCH AUTOSEL 4.18 003/126] locking/lockdep: Fix debug_locks off performance problem

From: Waiman Long <longman@...hat.com>

[ Upstream commit 9506a7425b094d2f1d9c877ed5a78f416669269b ]

It was found that when debug_locks was turned off because of a problem
found by the lockdep code, the system performance could drop quite
significantly when the lock_stat code was also configured into the
kernel. For instance, parallel kernel build time on a 4-socket x86-64
server nearly doubled.

Further analysis into the cause of the slowdown traced back to the
frequent call to debug_locks_off() from the __lock_acquired() function
probably due to some inconsistent lockdep states with debug_locks
off. The debug_locks_off() function did an unconditional atomic xchg
to write a 0 value into debug_locks which had already been set to 0.
This led to severe cacheline contention in the cacheline that held
debug_locks.  As debug_locks is being referenced in quite a few different
places in the kernel, this greatly slow down the system performance.

To prevent that trashing of debug_locks cacheline, lock_acquired()
and lock_contended() now checks the state of debug_locks before
proceeding. The debug_locks_off() function is also modified to check
debug_locks before calling __debug_locks_off().

Signed-off-by: Waiman Long <longman@...hat.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Paul E. McKenney <paulmck@...ux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@...radead.org>
Cc: Thomas Gleixner <tglx@...utronix.de>
Cc: Will Deacon <will.deacon@....com>
Link: http://lkml.kernel.org/r/1539913518-15598-1-git-send-email-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@...nel.org>
Signed-off-by: Sasha Levin <sashal@...nel.org>
---
 kernel/locking/lockdep.c | 4 ++--
 lib/debug_locks.c        | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c
index 5fa4d3138bf1..aa6ebb799f16 100644
--- a/kernel/locking/lockdep.c
+++ b/kernel/locking/lockdep.c
@@ -4148,7 +4148,7 @@ void lock_contended(struct lockdep_map *lock, unsigned long ip)
 {
 	unsigned long flags;

-	if (unlikely(!lock_stat))
+	if (unlikely(!lock_stat || !debug_locks))
 		return;

 	if (unlikely(current->lockdep_recursion))
@@ -4168,7 +4168,7 @@ void lock_acquired(struct lockdep_map *lock, unsigned long ip)
 {
 	unsigned long flags;

-	if (unlikely(!lock_stat))
+	if (unlikely(!lock_stat || !debug_locks))
 		return;

 	if (unlikely(current->lockdep_recursion))
diff --git a/lib/debug_locks.c b/lib/debug_locks.c
index 96c4c633d95e..124fdf238b3d 100644
--- a/lib/debug_locks.c
+++ b/lib/debug_locks.c
@@ -37,7 +37,7 @@ EXPORT_SYMBOL_GPL(debug_locks_silent);
  */
 int debug_locks_off(void)
 {
-	if (__debug_locks_off()) {
+	if (debug_locks && __debug_locks_off()) {
 		if (!debug_locks_silent) {
 			console_verbose();
 			return 1;
-- 
2.17.1