Re: [PATCH] fallout from atomic_long_t patch

[Evgeniy Polyakov <johnpol@2ka.mipt.ru>]

On Sun, Dec 17, 2006 at 09:24:30AM -0800, Linus Torvalds (torvalds@osdl.org) wrote:
> 
> 
> On Sun, 17 Dec 2006, Al Viro wrote:
> > -			if (likely(!test_bit(WORK_STRUCT_PENDING,
> > -					     &__cbq->work.work.management) &&
> > +			if (likely(!work_pending(&__cbq->work.work) &&
> 
> That should properly be
> 
> 	if (likely(!delayed_work_pending(&__cbq->work) && ...
> 
> and why the heck was it doing that open-coded int he first place?
>
> HOWEVER, looking even more, why is that thing a "delayed work" at all? All 
> the queuing seems to happen with a timeout of zero..
> 
> So I _think_ that the proper patch is actually the following, but somebody 
> who knows and uses the connector thing should double-check. Please?

Delayed work was used to play with different timeouts and thus allow to
smooth performance peaks, but then I dropped that idea, so timeout is always
zero.

I posted similar patch today to netdev@, which directly used
work_pending instead of delayed_work_pending(), but if you will figure
this out itself, I'm ok with proposed patch.


> 			Linus

-- 
	Evgeniy Polyakov
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/