[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: DFS speed control on netbsd-5 (and netbsd-6 issues)
I can certainly submit a pullup - presumably this would be for:
I've just tried to build a GENERIC kernel with the pullup - if someone
would like to test its at:
I just booted that kernel on this 1.5 GHz PowerBook G4:
NetBSD 5.2_STABLE (GENERIC) #0: Wed Aug 21 23:57:31 BST 2013
cpu0 at mainbus0: 7447A (Revision 1.1), ID 0 (primary)
cpu0: HID0 8450c0bc<EMCP,EICE,NAP,DPM,ICE,DCE,SGE,BTIC,FBIOB,ABE,BHT>,
cpu0: 1504.64 MHz, 512KB L2 cache no parity parity enabled
However, I see no sysctl for speed:
sysctl -a | grep speed
hw.fwmem.speed = 2
hw.sbp.max_speed = -1
hw.spdmem0.speed = 333
hw.spdmem1.speed = 333
And the speed is 750 MHz (calculating 1,000,000 primes takes 39 seconds
compared with 19 seconds on a 1.5 GHz 7447a Mac mini).
I tried some other patches which I thought were tested with netbsd-5 from
but while I have a machdep.cpu_speed (which defaults to zero), any attempt
to write to it locks up the laptop, so apparently I've applied the wrong
patches. I'll look into it more tonight.
Do we have a PR for the netbsd-6 / current issues?
Good question. We have plenty of discussion about the problems. Three
problems are summarized by Donald Lee here:
I personally loaded netbsd-6 onto a Mac mini, saw it crash twice in a
week, read about those issues and went with netbsd-5. My PowerMac 9600 was
in heavy use with netbsd-5 and had two years of uptime. A system that
crashes twice a week isn't usable comparatively.
I don't think anyone made any PRs, though. Donald?
Main Index |
Thread Index |