On Fri, 2018-01-26 at 10:06 -0700, Jason Gunthorpe wrote:
> On Fri, Jan 26, 2018 at 12:00:11PM -0500, Doug Ledford wrote:
> 
> > device, there is that window from our first ib_query_pkey call until we
> > complete the ib_register_event_handler() call for the pkey to change. 
> > Instead of throwing the flush regardless, it might be nicer to do:
> 
> Can we move the ib_register_event_handler sooner?

I didn't do a full audit, but I'm not sure it's safe to, simply because
we haven't fully completed our device setup and if our event handler is
live, we can get event triggered activities that might expect everything
to be done.  It would take a complete audit to know for sure.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
    GPG KeyID: B826A3330E572FDD
    Key fingerprint = AE6B 1BDA 122B 23B4 265B  1274 B826 A333 0E57 2FDD