SMP/UP kernel S01E02

Bill Hacker wbh at conducive.org
Sun Mar 25 02:35:12 PDT 2007


Thomas E. Spanjaard wrote:
Dmitri Nikulin wrote:
On 3/24/07, Gergo Szakal <bastyaelvtars at gmail.com> wrote:
Well, I've been thinking of the SMP/UP issue many times, and the best 
solution would be if SMP kernels could boot on UP machines. I have 
plenty UP boxes and a few SMP boxes to test on. I am not claiming 
anyhing, this is just 'food for thought'.
Don't they? I thought the only problem for most cases is that the SMP
kernels do more heavyweight locking and are therefore slower, and
without the gain of parallelism on UP it's a pure loss.
DragonFly SMP expects multiple processors, and will die if it doesn't 
find any, iirc when it tries to detect other APICs. As an example of 
working SMP-kernel-on-UP-hardware, see NetBSD (which doesn't involve 
boottime binary patching, btw).

Cheers,
Haven't looked into *how* FreeBSD 7-CURRENT i386 (not - 64 or AMD -64) does it, 
but it does it, as did 6.2-PRE AMD-64. (Tyan MB, two different Celeron, one 
single, one dual-core).

7-CURRENT boots unaltered on MSI-PM8M2V with Celeron single-core and Asus 
P5VDC-MX with P4 Core-D dual-core.

All Socket 775 gear.

/var/run/dmesg.boot tells the tale, and 'top' adds the column for which CPU is 
in use automagically.

Dragonfly 1.8, installed with the UNI MB,  is not (so far) willing to boot on 
the SMP MB. Will try a re-install later.

Unless there is a switch I can edit from the live-CD...  ??

Bill





More information about the Kernel mailing list