On Fri, 26 Nov 2004 19:48:49 +1100
Benjamin Herrenschmidt [off-list ref] wrote:

I suppose there is a good reason we can't just use the rtnl_sem for
these guys, though why isn't dev_base_lock a read/write semaphore
instead of a spinlock ? At least on ppc, I don't think there's any
overhead in the normal path, and this is not on a very critical path
anyway, is it ?

It can't be a semphore because it is taken in packet processing,
and thus softint handling, paths.