Cpufreq and AMD mobile Athlon 4

Hi guys,

I've just tried the cpufreq modules that come with 2.4.21-ac1 and I'm having no luck getting them to work. I have a Compaq mx9005 based on the mobile Athlon 4 (2400+). When I load the powernow-k7 module I get the following:

powernow: AMD K7 CPU detected.
powernow: PowerNOW! Technology present. Can scale: frequency and voltage.
powernow: Found PSB header at c00f18c0
powernow: Table version: 0x12
powernow: Flags: 0x0 (Mobile voltage regulator)
powernow: Settling Time: 100 microseconds.
powernow: Has 14 PST tables. (Only dumping ones relevant to this CPU).

Nothing else, it doesn't seem to dump and PST tables (as other examples I've seen show the driver reporting extra info about the CPU).

If I execute the following command:

echo "5000000:17000000:performance" > /proc/cpufreq

Then the module panics and I get the following in dmesg

Unable to handle kernel NULL pointer dereference at virtual address

00000004
printing eip:
c019423c
*pde = 00000000

Oops: 0000
CPU:    0
EIP:    0010:[<c019423c>]    Not tainted
EFLAGS: 00010246
eax: 00000000   ebx: c0c45f10   ecx: 00000000   edx: 00000000
esi: 00000000   edi: ffffffff   ebp: 00000000   esp: c0c45ebc
ds: 0018   es: 0018   ss: 0018
Process bash (pid: 4881, stackpage=c0c45000)
Stack: 00000000 c02a2ee0 c0c45f10 c0c45f40 c0c45f10 ccd157e8 c0c45f10
00000000
       c012860e c0c45f10 c34767d8 0000000c 400153f0 0000001d 00000000
c0c45f40
       c0c45f10 c0194838 c0c45f10 c0c45f10 0000001d 00000000 00000000
00000000
Call Trace:    [<ccd157e8>] [<c012860e>] [<c0194838>] [<c015dce0>]

[<c013c1d3>]
[<c0108dc7>]

Code: 8b 45 04 31 c9 83 f8 fe 74 1b 83 f8 ff 74 0c 3b 43 04 72 5a

Do I need to add some kind of processor defintion? If so how do I do it?

Regards,

Matt

This typically means that of all the PST tables, none of them had the same cpuid. I should put in a check for this, and printk something. BIOS upgrade ? If there isn't one, let me know what manufacturer/model, and I'll ask AMD if they'll 'send the heavies around'.

Ouch, that shouldn't happen no matter what.

Can you run that oops through ksymoops to decode the addresses to symbols ?

Dave

 Dave Jones     http://www.codemonkey.org.uk

Hi,

I've got the same problem with my notebook. This is a

Fujitsu-Siemens Amilo A7600.

# cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 6
model           : 8
model name      : mobile AMD Athlon (tm) 2000+    
stepping        : 1
cpu MHz         : 1656.764
cache size      : 256 KB
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow
bogomips        : 3296.46

# uname -r
2.4.21
# modprobe powernow-k7
powernow: AMD K7 CPU detected.
powernow: PowerNOW! Technology present. Can scale: frequency and voltage.
powernow: Found PSB header at c00f0640
powernow: Table version: 0x12
powernow: Flags: 0x0 (Mobile voltage regulator)
powernow: Settling Time: 100 microseconds.
powernow: Has 14 PST tables. (Only dumping ones relevant to this CPU).
Nothing else.
But what to do?

Zoltan

noted. Can you mail me the dmidecode output too please ? (See http://people.redhat.com/arjanv/dmidecode.c)

Is there a BIOS update available you can try ?

I really wish I had the time to get sysfs overrides written..

Dave

 Dave Jones     http://www.codemonkey.org.uk

How to buy you some free time, I need that as well for my ugly Acer Laptop.

Bye, Stefan aka mETz

[email protected] | ICQ#51123152 | Moege der Pinguin mit euch sein

Hi guys,

See attached the decoded sys oops from my encounters with powernow-k7.

Regards,

Matt

Unable to handle kernel NULL pointer dereference at virtual address 00000004
c019423c
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0010:[<c019423c>]    Not tainted

Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax: 00000000 ebx: c3c5df10 ecx: 00000000 edx: 00000000
esi: 00000000 edi: ffffffff ebp: 00000000 esp: c3c5debc
ds: 0018 es: 0018 ss: 0018
Process bash (pid: 4857, stackpage=c3c5d000)

Stack: 00000000 c02a2ee0 c3c5df10 c3c5df40 c3c5df10 ccd007e8 c3c5df10 00000000
       c012860e c3c5df10 c2194898 0000000c c3c5df84 0000001b 00000000 c3c5df40
       c3c5df10 c0194838 c3c5df10 c3c5df10 0000001b 00000000 00000000 00000000
Call Trace:    [<ccd007e8>] [<c012860e>] [<c0194838>] [<c015dce0>] [<c013c1d3>]
  [<c0108dc7>]
Code: 8b 45 04 31 c9 83 f8 fe 74 1b 83 f8 ff 74 0c 3b 43 04 72 5a

Trace; ccd007e8 <[powernow-k7]powernow_verify+18/20> Trace; c012860e <cpufreq_set_policy+8e/270> Trace; c0194838 <cpufreq_proc_write+f8/140> Trace; c015dce0 <proc_file_write+40/50> Trace; c013c1d3 <sys_write+a3/130>
Trace; c0108dc7 <system_call+33/38>
Code; c019423c <cpufreq_frequency_table_verify+5c/f0> 00000000 <_EIP>:
Code; c019423c <cpufreq_frequency_table_verify+5c/f0> <=====
0: 8b 45 04 mov 0x4(%ebp),%eax <=====
Code; c019423f <cpufreq_frequency_table_verify+5f/f0>
3: 31 c9 xor %ecx,%ecx
Code; c0194241 <cpufreq_frequency_table_verify+61/f0>
5: 83 f8 fe cmp $0xfffffffe,%eax
Code; c0194244 <cpufreq_frequency_table_verify+64/f0>
8: 74 1b je 25 <_EIP+0x25>
Code; c0194246 <cpufreq_frequency_table_verify+66/f0>
a: 83 f8 ff cmp $0xffffffff,%eax
Code; c0194249 <cpufreq_frequency_table_verify+69/f0>
d: 74 0c je 1b <_EIP+0x1b>
Code; c019424b <cpufreq_frequency_table_verify+6b/f0>
f: 3b 43 04 cmp 0x4(%ebx),%eax
Code; c019424e <cpufreq_frequency_table_verify+6e/f0>
12: 72 5a jb 6e <_EIP+0x6e>

_______________________________________________
Cpufreq mailing list
[email protected]
http://www.linux.org.uk/mailman/listinfo/cpufreq

Hey guys,

Any more info on my problem, things I can do etc etc.

Please, any comments are welcome.

Matt