hi again,

I implemented a couple more optimizations and the results now look like:

fork+exit
orig	0.025u 0.979s 0:09.22 10.7%     0+0k 0+1io 0pf+0w
chuq	0.021u 0.629s 0:06.04 10.5%     0+0k 0+0io 0pf+0w
improvement	34%


fork+exec+exit
perf kernel
orig	0.076u 2.232s 0:07.57 30.3%     0+0k 0+3io 0pf+0w
chuq	0.126u 1.645s 0:05.74 30.6%     0+0k 0+2io 0pf+0w
improvement	24%


the main improvement is pre-loading the TSB in pmap_enter().
new diff in the usual place.

(I'll get back to the pmap API discussion this weekend...)

-Chuck