On Sat, Apr 20, 2002 at 05:32:19PM -0300, Rik van Riel wrote:
> Better SMP scalability, first batch of lock breakup work. Still
> experimental, but testing is very much welcome...
> My big TODO items for a next release are:
> - O(1) page launder - currently functional but slow, needs to be tuned
> - pte-highmem
If I can clarify a bit, I'm deferring submission of the subsequent bits
of the lock breakups in order to to expose the various stages of it to
wider testing individually. One can't be too careful with these changes.
Also, Martin Bligh is the original author of the per-page pte_chain
More to come,