Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 04 Oct 1996 14:51:30 -0600
From:      Steve Passe <smp@csn.net>
To:        Rob Miracle <rwm@MPGN.COM>
Cc:        freebsd-smp@freebsd.org
Subject:   Re: SMP processor utilization question 
Message-ID:  <199610042051.OAA24658@clem.systemsix.com>
In-Reply-To: Your message of "Fri, 04 Oct 1996 09:14:47 EDT." <3.0b24.32.19961004091446.006a680c@central.TanSoft.COM> 

next in thread | previous in thread | raw e-mail | index | archive | help
Hi,

>send.  I did get an error when I redirected the output to a file, but did
>not get it when piping it to less:
>
>	Warning: EBDA support is BROKEN!!
This message is because the variable I need is way down low where the
kernel doesn't map memory, so I can't get at it from /dev/kmem in mptable.
The kernel itself doesn't have this problem as it can directly read it
during boot.  This only appears to affect EISA and MCA boards, ie only
they use EBDA.

As to getting lost, it goes to stderr while the rest goes to stdout.

---------------------------------------------------------------------------
Several people have noticed that the output of mptable says the CPU APIC
version is 17 (0x11), while the boot probe says its 16 (0x10).  I now have
info for the case of a P6 from your mailing, specifically:

mptable output for your dual P6:
 ...
Processor
 apic ID: 1, version: 17

dmesg output of your dual P6:
 ...
FreeBSD/SMP: Multiprocessor motherboard
 BSP APIC ID: 1, version: 0x00040011
 AP APIC IDs: 0
 IO APIC IDs: 2, version: 0x00170011
 # of APs available: 1, # of APs started: 1

In this case (ie P6) THEY MATCH. Also note the fact that the '4' in the 5th
(from the right) digit is correct for a P6.

---
mptable output for my P5:
 ...
Processor
 apic ID: 0, version: 17

dmesg output of my dual P5 (text diffs from slightly newer kernel):

FreeBSD/SMP: Multiprocessor motherboard
 cpu0 (BSP): apic id: 0, version: 0x00030010
 cpu1 (AP):  apic id: 1, version: 0x00030010
 io0 (APIC): apic id: 2, version: 0x00170011

In this case (ie P5) THEY DON'T MATCH. Note the fact that the '3' in the 5th
field is correct for a P5.

My conclusion is that the Intel MP BIOS  extensions report this wrong, or
perhaps it is a "feature" that I haven't decoded out of the spec yet...

---
Keep those mptable reports coming, folks, they are useful.  If you have the
MP kernel running also include the SMP specific dmesg output.

--
Steve Passe	| powered by
smp@csn.net	|            FreeBSD




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199610042051.OAA24658>