> Am I reading your patch wrong, though, or does it in fact still have a
> minimum allocation size of a page? 

Yes. But I'll address that. See my previous reply and Jason's remark.


> Allocating less than a page was the reason I was attacking this using
> malloc and realloc instead of a private allocator.  Is there a real
> performance benefit to not using malloc, given that you're still using
> kernel_map?

Cf. Jason's remark. Since we don't allocate/release buffers from an
interrupt context, using kernel_map seems the obvious choice. I also
wanted to keep the option open to map the buffers non-wired someday;
I've yet to look at yamt's patches to see what it can do in this respect.

>  One thing I'd caution against would
> be eliminating the use of the AGE list without careful measurement on a
> number of different systems; it implements a pseudo-generational behaviour
> that is not strict LRU and may perform better for some workloads.

Ok, noted.

-pk