On Tuesday 26 February 2008, Peter Zijlstra wrote:
>
> On Mon, 2008-02-25 at 14:33 -0800, David Brownell wrote:
>
> > > +#ifdef CONFIG_LOCKDEP
> > > +
> > > +/* tell lockdep that this IRQ's locks and its parent's locks are in
> > > + * different categories, so that it won't detect false recursion
On Mon, 2008-02-25 at 14:33 -0800, David Brownell wrote:
> > +#ifdef CONFIG_LOCKDEP
> > +
> > +/* tell lockdep that this IRQ's locks and its parent's locks are in
> > + * different categories, so that it won't detect false recursion.
> > + */
> > +static struct lock_class_key gpio_lock_class;
> >
> So I'm think that the reason this only _changes_ the false
> recursion notification ...
Whoops, it's because of the following typo:
> --- a/arch/arm/plat-omap/gpio.c 2008-02-24 19:02:32.0 -0800
> +++ b/arch/arm/plat-omap/gpio.c 2008-02-25 10:54:29.0 -0800
> @@ -1332,
> > I thought the way to use the *_nested() calls was "consistently"!
>
> Very much depends on your view of consistent :-)
>
> > That is, if one instance of a lock access uses it, they all should,
> > since that's the only way lockdep learns about equivalence classes.
> > Also, locks shouldn't move
On Mon, 2008-02-25 at 03:21 -0800, David Brownell wrote:
> > > > > > ==> LOCKDEP feature is evidently missing:
> > > > > > spin_lock_irq_nested(lock_ptr, lock_class)
> > > > >
> > > > > This rant is more lines than adding the API :-/ the reason for it not
> > > > > being there is
> > > > > ==> LOCKDEP feature is evidently missing:
> > > > > spin_lock_irq_nested(lock_ptr, lock_class)
> > > >
> > > > This rant is more lines than adding the API :-/ the reason for it not
> > > > being there is simple, it wasn't needed up until now.
> > >
> > > I suspected that
On Sun, 2008-02-24 at 20:33 -0800, David Brownell wrote:
> > > > ==> LOCKDEP feature is evidently missing:
> > > > spin_lock_irq_nested(lock_ptr, lock_class)
> > >
> > > This rant is more lines than adding the API :-/ the reason for it not
> > > being there is simple, it wasn't ne
> > > ==> LOCKDEP feature is evidently missing:
> > > spin_lock_irq_nested(lock_ptr, lock_class)
> >
> > This rant is more lines than adding the API :-/ the reason for it not
> > being there is simple, it wasn't needed up until now.
>
> I suspected that was the case, but for all I
8 matches
Mail list logo
