>I see this commit:
>  Change an if_qflush() to an IFQ_PURGE() to deal with ALTQ correctly.
>and I find myself wondering `why does if_qflush() even continue to
>exist when it won't work with ALTQ'?

	to migrate to ALTQ gradually, we need to kee if_qflush() until
	all the drivers are rewritten.  if_qflush() certainly works for
	non-ALTQ drivers (correct me if i'm wrong).

>But I've also been wondering why IFQ_PURGE() is inlined at all.  It's
>not as if it ever occurs in a critical path.  Shouldn't the ALTQ
>support just go in if_qflush(), and all the users do a function call
>instead?
>Is there any high-level design here, or is this all being done seat of
>the pants?

	not sure i understand the concern, but IFQ_PURGE() may be
	expanded into ALTQ_PURGE().

itojun