Re: [RFC/PATCH 2/3] rt: PI-workqueue support

[Daniel Walker]

On Mon, 2007-10-22 at 14:15 +0200, Peter Zijlstra wrote:

> Daniel wrote this bit, but I tend to agree with him, but can't give his
> rationale. Mine is that worklets are typically asynchonous and thus its
> prio should not depend on temporal things like boosting.
> 
> OTOH it would probably make sense to allow it to depend on it through
> the barrier constructs, but for that I have to hook the completions into
> the PI chain. Something that needs more thought.

Yeah, I think Peter summarized it .. Since the task isn't waiting on
work when it's inserted it didn't seem right to use a priority that may
be boosted, since the work isn't preventing completion .. I think the
only time you would want to transfer the boosted priority is when a task
gets blocked, which does happen when you flush the workqueue.

Although, If there is one area of this code that needs attention I think
it's the PI stuff, it wasn't my first priority at the time .. I also
recall Oleg find some issue with it ..

Daniel

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