Re: [PATCH v2] lockdep: Let lock_is_held_type() detect recursive read as read

[Waiman Long]

On 9/7/21 10:16 PM, Boqun Feng wrote:
> On Fri, Sep 03, 2021 at 10:40:01AM +0200, Sebastian Andrzej Siewior wrote:
> > lock_is_held_type(, 1) detects acquired read locks. It only recognized
> > locks acquired with lock_acquire_shared(). Read locks acquired with
> > lock_acquire_shared_recursive() are not recognized because a `2' is
> > stored as the read value.
> >
> > Rework the check to additionally recognise lock's read value one and two
> > as a read held lock.
> >
> > Fixes: e918188611f07 ("locking: More accurate annotations for read_lock()")
> > Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
> > ---
> > v1…v2:
> >    - simplify the read check to !!read as suggested by Waiman Long.
> >
> >   kernel/locking/lockdep.c |    2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > --- a/kernel/locking/lockdep.c
> > +++ b/kernel/locking/lockdep.c
> > @@ -5366,7 +5366,7 @@ int __lock_is_held(const struct lockdep_
> >   		struct held_lock *hlock = curr->held_locks + i;
> >   
> >   		if (match_held_lock(hlock, lock)) {
> > -			if (read == -1 || hlock->read == read)
> > +			if (read == -1 || hlock->read == !!read)
> I think this should be:
>
> 	!!hlock->read == read
>
> With that,
>
> Acked-by: Boqun Feng <boqun.feng@xxxxxxxxx>
You are right. It should be the other way around. read can only be -1, 
0, 1 while hlock->read can be 0, 1, 2.

Cheers,
Longman