lkml.org 
[lkml]   [2018]   [Mar]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [RFC tip/locking/lockdep v5 07/17] lockdep: Adjust check_redundant() for recursive read change
On Thu, Feb 22, 2018 at 06:29:06PM +0100, Peter Zijlstra wrote:
> On Thu, Feb 22, 2018 at 03:08:54PM +0800, Boqun Feng wrote:
> > As we have four kinds of dependencies now, check_redundant() should only
> > report redundant if we have a dependency path which is equal or
> > _stronger_ than the current dependency. For example if in
> > check_prev_add() we have:
> >
> > prev->read == 2 && next->read != 2
> >
> > , we should only report redundant if we find a path like:
> >
> > prev--(RN)-->....--(*N)-->next
> >
> > and if we have:
> >
> > prev->read == 2 && next->read == 2
> >
> > , we could report redundant if we find a path like:
> >
> > prev--(RN)-->....--(*N)-->next
> >
> > or
> >
> > prev--(RN)-->....--(*R)-->next
> >
> > To do so, we need to pass the recursive-read status of @next into
> > check_redundant().
>
> Very hard to read that.
>

Right.. and I find a bug about this, let me explain below..

> > Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
> > ---
> > kernel/locking/lockdep.c | 13 ++++++++-----
> > 1 file changed, 8 insertions(+), 5 deletions(-)
> >
> > diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c
> > index e1be088a34c4..0b0ad3db78b4 100644
> > --- a/kernel/locking/lockdep.c
> > +++ b/kernel/locking/lockdep.c
> > @@ -1338,9 +1338,12 @@ print_circular_bug_header(struct lock_list *entry, unsigned int depth,
> > return 0;
> > }
> >
> > -static inline int class_equal(struct lock_list *entry, void *data)
> > +static inline int hlock_equal(struct lock_list *entry, void *data)
> > {
> > - return entry->class == data;
> > + struct held_lock *hlock = (struct held_lock *)data;
> > +
> > + return hlock_class(hlock) == entry->class &&
> > + (hlock->read == 2 || !entry->is_rr);
> > }
>
> So I guess @data = @next, and we're checking if @prev -> @next already
> exists.
>
> Since we only care about forward dependencies, @next->read==2 means *R
> and if so, any existing link is equal or stronger. If @next->read!=2, it
> means *N and we must regard *R as weaker and not match.
>

Yep, the idea if we find a path @prev -> .. -> @next is RN, and @prev ->
@next is RR, then we treat RR as weaker and redundant. Basically, if we
find a strong path that could replace the about-to-add dependency
(the path is stronger than or equal to the dependency), we report
redundant (a match).

But I miss something here, as you may see both the start and end of the
path are important, so say we find a strong path RN, but the
about-to-add dependency is NR, we can not report it as redundant,
because the path can not replace the dependency.

To make sure we find a path whose start point is stronger than @prev, we
need a trick, we should initialize the ->only_xr (or ->have_xr) of the
root (start point) of __bfs() to be @prev->read != 2, therefore if @prev
is N, __bfs() will pick N* for the first dependency, otherwise, __bfs()
can pick N* or R* for the first dependency.

I use a similar setup before check_noncircular(), which sets the initial
->only_xr to be @next->read == 2, because we need @prev -> @next ->
<path> to be strong. But I should use a opposite setup for
check_redundant() as I describe above.

Anyway, I will fix this and prove (hopefully) enough comments for those
tricks.

Regards,
Boqun

> OK, that seems to be fine, but again, that function _really_ could do
> with a comment.
>
>
[unhandled content-type:application/pgp-signature]
\
 
 \ /
  Last update: 2018-03-16 09:17    [W:0.096 / U:1.648 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site