On 2015-08-19 23:48, Felix Deichmann wrote:
2015-08-20 7:03 GMT+02:00 John Klos <john%ziaspace.com@localhost>:While putting some test code in sys/arch/vax/vax/ka48.c, though, I see that the 6K of cache is never enabled. Wow! So that's why these machines are so pokey with NetBSD!Maybe everyone thought it's normal for those machines to be that slow and never noticed. :) Can you tell if ka660_cache_enable() looks correct, as you mention the KA660 (being SOC, too)? Test this code instead?
No, not really. We had this discussion a year ago in April. "NetBSD speed curiosities" So it is known, but nobody seems to know how to fix it? Cheers