On Jun 8, 6:01 pm, hyc wrote:
> I just replaced my X2 3800+ with an Opteron 185 and now it appears
> that the powernow driver no longer works. The relevant boot messages:
>
> ACPI: Power Button (FF) [PWRF]
> ACPI: Power Button (CM) [PWRB]
> ACPI: Sleep Button (CM) [SLPB]
> powernow-k8: Found 2 Dual Core AMD Opteron(tm) Processor 185
> processors (version 2.00.00)
> powernow-k8: MP systems not supported by PSB BIOS structure
> powernow-k8: MP systems not supported by PSB BIOS structure
> acpi_processor-0742 [00] processor_preregister_: Error while parsing
> _PSD domain information. Assuming no coordination
>
> This is on an Asus A8V-Deluxe with the latest (beta) bios (1018) and
> the 2.6.18 kernel. Any ideas on how to get this working? Everything
> was fine with the X2 3800+.

It seems that I may be able to get around this by hacking my DSDT and
setting up the _PSS and _PCT tables myself. Does anyone have an
Opteron 185 running with PowerNow supported, who can disassemble their
current DSDT and send these tables to me? I can construct them from
the AMD docs but it's a bit of a pain. In my current DSDT the tables
are named APSS and APCT and contain invalid data, I guess that's what
the BIOS does when it doesn't recognize the processor model. (And of
course, there's actually a syntax error in there when I recompile the
disassembled table. Haven't figured out how to fix that yet.)

On Jun 9, 6:33 pm, hyc wrote:
> It seems that I may be able to get around this by hacking my DSDT and
> setting up the _PSS and _PCT tables myself. Does anyone have an
> Opteron 185 running with PowerNow supported, who can disassemble their
> current DSDT and send these tables to me? I can construct them from
> the AMD docs but it's a bit of a pain. In my current DSDT the tables
> are named APSS and APCT and contain invalid data, I guess that's what
> the BIOS does when it doesn't recognize the processor model. (And of
> course, there's actually a syntax error in there when I recompile the
> disassembled table. Haven't figured out how to fix that yet.)

The syntax error reported by Intel's acpica 20050309 (and older) was
apparently a compiler bug, it's not a problem in the 20061109 release.

For future reference, here's the set of tables I got from an Opteron
285 that I modified for my system:

Processor (CPU1, 0x01, 0x00005010, 0x06)
{
Name (_PCT, Package (0x02)
{
ResourceTemplate ()
{
Register (FFixedHW, 0x00, 0x00,
0x0000000000000000)
},

ResourceTemplate ()
{
Register (FFixedHW, 0x00, 0x00,
0x0000000000000000)
}
})
Name (_PSS, Package (0x06)
{
Package (0x06)
{
0x0A28,
0x00017318,
0x64,
0x07,
0xE0202A12,
0x0212
},

Package (0x06)
{
0x0960,
0x000160BC,
0x64,
0x07,
0xE0202A90,
0x0290
},

Package (0x06)
{
0x0898,
0x000129A8,
0x64,
0x07,
0xE0202B0E,
0x030E
},

Package (0x06)
{
0x07D0,
0xF938,
0x64,
0x07,
0xE0202B8C,
0x038C
},

Package (0x06)
{
0x0708,
0xCFD0,
0x64,
0x07,
0xE0202B8A,
0x038A
},

Package (0x06)
{
0x03E8,
0x87F0,
0x64,
0x07,
0xE0202B82,
0x0382
}
})
Name (PSXE, Buffer (0x18)
{
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00
})
Name (PSXD, Buffer (0x18)
{
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00
})
Name (PSXC, Buffer (0x18)
{
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00
})
Name (PSXB, Buffer (0x18)
{
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00
})
Method (_PPC, 0, NotSerialized)
{
Return (Zero)
}
}

I don't have any idea what those PSX? buffers are all about. Can't
seem to find a reference in any revision of the ACPI spec...