FYI, I've switched my primary desktop to AMD64.

[Peter Wemm]

Adriaan de Groot wrote:
> On Friday 17 October 2003 19:17, Peter Wemm wrote:

> > real memory  = 1072955392 (1023 MB)
> > avail memory = 1011191808 (964 MB)
> 
> There isn't anything out there that supports more than 4G memory, is there? 

sledge.freebsd.org has 8GB of ram..  Its a Rioworks/Arima HDAMA motherboard
with 1GB dimms in each of the 8 slots.  All the server boards should work
fine with this.

CPU: AMD Opteron(tm) Processor 244 (1804.10-MHz K8-class CPU)
  Origin = "AuthenticAMD"  Id = 0xf51  Stepping = 1
  Features=0x78bfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CM
OV,PAT,PSE36,CLFLUSH,MMX,FXSR,SSE,SSE2>
  AMD Features=0xe0500000<NX,MMX+,LM,3DNow!+,3DNow!>
real memory  = 8589934592 (8192 MB)
avail memory = 8130809856 (7754 MB)

> > fxp0: <Intel 82559 Pro/100 Ethernet> port 0xdf00-0xdf3f mem
> > 0xfc800000-0xfc8fffff,0xfc9fe000-0xfc9fefff irq 11 at device 5.0 on pci1
> > fxp0: Ethernet address 00:d0:b7:21:31:94
> 
> I take it this means that the nforce3 integrated nvidia NIC is still not 
> working? This was one of the reasons for me to go with a VIA chipset board.

Correct.  There is a binary driver stub for Linux/i386 for the nForce/
nForce2 MCP ethernet.  One enterprising chap has written a FreeBSD/i386
shim for it, but I never got around to finding out if it also worked with
the nForce3.  I'm not expecting it to, because the nForce2 is 10/100 and
the nForce3 is 10/100/1000 - obviously the hardware is different.

Under amd64 native mode though, even the binary stub is useless to us.  We
can't use it.

> > atapci1: <Promise PDC20378 SATA150 controller> port
> > 0xdc00-0xdc7f,0xdf60-0xdf6f,0xde80-0xdebf mem
> > 0xfc9a0000-0xfc9bffff,0xfc9fd000-0xfc9fdfff irq 11 at device 8.0 on pci1
> 
> I think the K8V also uses promise sata, so that ought to go well for me too.

The K8V has a Silicon Image ship on it right next to the SATA connectors.
I've seen Soren add support for several of those recently.

> > pcib2: <ACPI PCI-PCI bridge> at device 11.0 on pci0
> > pcib2: could not get PCI interrupt routing table for \\_SB_.PCI0.P0P2 -
> > AE_NOT_FOUND pci2: <ACPI PCI bus> on pcib2
> > pci2: <display, VGA> at device 0.0 (no driver attached)
> 
> That's you matrox card?

Correct.  X uses it in userland.  All that would attach in the kernel to
a vga card would be a DRI driver.  I have not tried it.

> > acpi_button0: <Power Button> on acpi0
> >   uteval-0226: *** Error: Return object type is incorrect
> > [\\_SB_.LATA._CRS] (Node 0xffffff0000a24730), AE_TYPE can't fetch resources
> > for \\_SB_.LATA - AE_TYPE
> 
> Isn't that vaguely disconcerting, to have runtime type errors in ACPI code?

Yes, to say the least.  Fortunately it isn't a problem.  My Tyan thunder K8 at
work gives about 10 pages of ACPI errors but still works.

> One last question, about the wifi on the K8V Deluxe. Do you have any idea wha
    t 
> that is? Whether it's supported? I'll find out automatically, I guess :)

I have no idea.  My cynical side would bet it would be something like the
realtek 802.11b chipset which we do not support (no programming info).

Cheers,
-Peter
--
Peter Wemm - peter at wemm.org; peter at FreeBSD.org; peter at yahoo-inc.com
"All of this is for nothing if we don't go to the stars" - JMS/B5