[CFT] Athlon testers needed.

[CFT] Athlon testers needed.

[Dave Jones]

In some obscure documentation, I found out that recent Athlons
(model 8 stepping 1 and greater) behave better with certain MSRs
programmed slightly differently to those of earlier models.
It's likely a lot of BIOSen out there don't get this right
(Especially in the BIOS older than CPU case).

x86info[1] will be able to dump these registers (run as root
and use -m option), and mail me the results so I can confirm
my suspicions.  I've a patch pending, but I'd rather get some
test data before unleashing it on unsuspecting victims.

I'm particularly interested in hearing from folks who have
been experiencing problems with model 8 and above Athlons,
earlier models are less interesting for the moment.

		Dave

[1] http://www.codemonkey.org.uk/x86info/x86info-1.11.tar.gz


-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: [CFT] Athlon testers needed.

[erich]

Dave Jones <davej@codemonkey.org.uk> wrote:

> In some obscure documentation, I found out that recent Athlons
> (model 8 stepping 1 and greater) behave better with certain MSRs
> programmed slightly differently to those of earlier models.
> It's likely a lot of BIOSen out there don't get this right
> (Especially in the BIOS older than CPU case).
> 
> x86info[1] will be able to dump these registers (run as root
> and use -m option), and mail me the results so I can confirm
> my suspicions.  I've a patch pending, but I'd rather get some
> test data before unleashing it on unsuspecting victims.
> 
> I'm particularly interested in hearing from folks who have
> been experiencing problems with model 8 and above Athlons,
> earlier models are less interesting for the moment.
> 
> 		Dave
> 
> [1] http://www.codemonkey.org.uk/x86info/x86info-1.11.tar.gz


2 points:

  --  The link is:

	http://www.codemonkey.org.uk/x86info/x86info-1.11.tgz

  --  A note to those not in the know to load the "msr" driver
      first, for example on my RedHat system:

	insmod msr


In any case, here's the result from my recently acquired Athlon
XP 2600+:


			31       23       15       7 
MSR: 0x0000002a=0x00000000 : 00000000 00000000 00000000 00000000
MSR: 0xc0000080=0x00000000 : 00000000 00000000 00000000 00000000
MSR: 0xc0010010=0x00160604 : 00000000 00010110 00000110 00000100
MSR: 0xc0010015=0x080b9008 : 00001000 00001011 10010000 00001000
MSR: 0xc001001b=0x6003d22f : 01100000 00000011 11010010 00101111

Family: 6 Model: 8 Stepping: 1
CPU Model : Athlon XP (Thoroughbred)[B0]
Processor name string: AMD Athlon(tm) XP Processor

PowerNOW! Technology information
Available features:
	  Temperature sensing diode present.


For comparison, here is the output from an Athlon XP 1800+:

			31       23       15       7 
MSR: 0x0000002a=0x00000000 : 00000000 00000000 00000000 00000000
MSR: 0xc0000080=0x00000000 : 00000000 00000000 00000000 00000000
MSR: 0xc0010010=0x00168604 : 00000000 00010110 10000110 00000100
MSR: 0xc0010015=0x01001008 : 00000001 00000000 00010000 00001000
MSR: 0xc001001b=0x6003d22f : 01100000 00000011 11010010 00101111

Family: 6 Model: 6 Stepping: 2
CPU Model : Athlon MP (palomino)
Processor name string: AMD Athlon(TM) MP 1800+

PowerNOW! Technology information
Available features:
	  Temperature sensing diode present.


--
    Erich Stefan Boleyn     <erich@uruk.org>     http://www.uruk.org/
"Reality is truly stranger than fiction; Probably why fiction is so popular"

Re: [CFT] Athlon testers needed.

[Dave Jones]

On Wed, Nov 20, 2002 at 12:11:35PM -0800, erich@uruk.org wrote:
 > In any case, here's the result from my recently acquired Athlon
 > XP 2600+:
 > 
 > 			31       23       15       7 
 > MSR: 0x0000002a=0x00000000 : 00000000 00000000 00000000 00000000
 > MSR: 0xc0000080=0x00000000 : 00000000 00000000 00000000 00000000
 > MSR: 0xc0010010=0x00160604 : 00000000 00010110 00000110 00000100
 > MSR: 0xc0010015=0x080b9008 : 00001000 00001011 10010000 00001000
 > MSR: 0xc001001b=0x6003d22f : 01100000 00000011 11010010 00101111
 > 
 > Family: 6 Model: 8 Stepping: 1
 > CPU Model : Athlon XP (Thoroughbred)[B0]

Bingo, this has the bug.

 > For comparison, here is the output from an Athlon XP 1800+:
 > 
 > 			31       23       15       7 
 > MSR: 0x0000002a=0x00000000 : 00000000 00000000 00000000 00000000
 > MSR: 0xc0000080=0x00000000 : 00000000 00000000 00000000 00000000
 > MSR: 0xc0010010=0x00168604 : 00000000 00010110 10000110 00000100
 > MSR: 0xc0010015=0x01001008 : 00000001 00000000 00010000 00001000
 > MSR: 0xc001001b=0x6003d22f : 01100000 00000011 11010010 00101111
 > 
 > Family: 6 Model: 6 Stepping: 2
 > CPU Model : Athlon MP (palomino)

This one is ok.
The bug is that c001001b are being programmed the same way.
They need different programming on newer models, which older
BIOSes are unaware of.

I'll cook up a patch and send a complete description after a quick
compile test. Thanks for the report.

		Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: [CFT] Athlon testers needed.

[Petr Vandrovec]

On 20 Nov 02 at 18:53, Dave Jones wrote:
> and use -m option), and mail me the results so I can confirm
> my suspicions.  I've a patch pending, but I'd rather get some
> test data before unleashing it on unsuspecting victims.
> 
> [1] http://www.codemonkey.org.uk/x86info/x86info-1.11.tar.gz

.../x86info-1.11.tgz... Unfortunately, all my athlons are model 6.
I'll have to buy some new ;-)
                                            Petr Vandrovec