Re: [PATCH 3.18 031/104] locking/lockdep: Add debug_locks check in __lock_downgrade()

From: Tetsuo Handa
Date: Wed Apr 24 2019 - 13:31:36 EST


On 2019/04/25 2:08, Greg Kroah-Hartman wrote:
> [ Upstream commit 71492580571467fb7177aade19c18ce7486267f5 ]
>
> Tetsuo Handa had reported he saw an incorrect "downgrading a read lock"
> warning right after a previous lockdep warning. It is likely that the
> previous warning turned off lock debugging causing the lockdep to have
> inconsistency states leading to the lock downgrade warning.
>
> Fix that by add a check for debug_locks at the beginning of
> __lock_downgrade().

Excuse me?

>
> Debugged-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
> Reported-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
> Reported-by: syzbot+53383ae265fb161ef488@xxxxxxxxxxxxxxxxxxxxxxxxx
> Signed-off-by: Waiman Long <longman@xxxxxxxxxx>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
> Cc: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> Cc: Will Deacon <will.deacon@xxxxxxx>
> Link: https://lkml.kernel.org/r/1547093005-26085-1-git-send-email-longman@xxxxxxxxxx
> Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
> ---
> kernel/locking/lockdep.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c
> index fb90ca3a296e..27de98428367 100644
> --- a/kernel/locking/lockdep.c
> +++ b/kernel/locking/lockdep.c
> @@ -3312,6 +3312,9 @@ __lock_set_class(struct lockdep_map *lock, const char *name,
> unsigned int depth;
> int i;
>
> + if (unlikely(!debug_locks))
> + return 0;
> +

This is __lock_set_class() function rather than __lock_downgrade() function.
__lock_downgrade() is available in 4.12+.

For 3.18-stable, downgrade_write() is in kernel/locking/rwsem.c . Therefore,
we should check whether adding this check into __lock_set_class() is what
we want to do...

> depth = curr->lockdep_depth;
> /*
> * This function is about (re)setting the class of a held lock,
>