> > I was assuming that once you have the lock from the arbitrator you can
> > be sure that the other master is not active and, thus, you maybe don't
> > need to read the status register at all? But I am probably wrong here.
> 
> You are right in that when you have the lock from the arbtrator, you
> lock the other master(s) out, but the issue here is that the mux code
> releases the arbitrator lock after each xfer. Krzysztof needs to do
> two xfers with the arbitrator lock held over both so that another
> master cannot sneak in and change the world view.

Yes. My probably wrong assumption was that the first part of the
transfer (reading a status reg) was only to check if another master has
started the device and is running some operation. So, if we could avoid
that this happens by proper locking of the arbitrator, you could either
skip the first part of the transaction or just send both in one go.