Andrew

Happy anniversary old PR.

8mb matrices are probably less than caches by magnitude.

Your data set is too small for all 64 caches.

there are some badly modelled areas, if input+temp+output fits in one cache one core is optimal, arbitrary above that it switches to all core threads and there is observable glitch...

I got your idea, you are talking scoreboard, not lock. Currently OpenBLAS uses one or all threads, if it could be tuned to gradually rise threads based on size (vs...

Could be useful to allow dhcp for 5 guest nets.... Similar thing is done in one place parse_reflection_source / fw4.uc - synthetic rules are generated as needed for reflection_src.

Yeah it is cool, but it needs to be done for every rule type at least few lines...