Re: Add a new work-queue for destructing stations?

[Johannes Berg <johannes@sipsolutions.net>]

On Thu, 2012-12-13 at 11:00 -0800, Ben Greear wrote:
> On 12/13/2012 10:53 AM, Johannes Berg wrote:
> > On Thu, 2012-12-13 at 19:49 +0100, Johannes Berg wrote:
> >
> >>>> Well the concern there is that we expose the workqueue to the driver,
> >>>> and it might even put sdata-referencing work items on it and expect them
> >>>> to be flushed out before stop_interface() is called?
> >
> > Wait actually, that isn't a concern. The flush_workqueue() was new in
> > commit b22cfcfca, before that drivers couldn't rely on it. It seems very
> > unlikely that in the meantime drivers have come to rely on it and we
> > didn't document this, so ... with the proposed change to sta_free_wk we
> > should be fine.
> 
> We still need to cancel the other mac80211 work items though, right?

Well, no, because we didn't add any work items since we added the
flush_workqueue() there.

> I didn't understand your proposal for the sta_free_wk changes,
> but I can at least work on tracking down the mac80211 work items
> if that helps...

I'll get a patch out later or tomorrow.

johannes