powernow-k7 for XP-M 2400+ CPU
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 31 lines · Harald Milz
Quoted 2 lines · Bruno Ducrot
Also the flags from the bios indicate that there is a desktop voltage regulator, and AFAIK it's not possible to change voltage from that.
That's okay with me. It would be better to be able to change voltages as well but as long as I can set the clock rate it's okay. :-)
Quoted 3 lines
I suspect more that the BIOS is not able to says if an Athlon is mobile, or not. Actually there is very few differences (in detection code, I mean) to sort a MP from a mobile one.
Yes, but if desktop boards were willing to detect mobiles they could. It's just a marketing issue - which I am not willing to accept :-))
just a marketing issue - which I am not willing to accept :-))
I gave
dump_psb.ca try:PSB version: 12 flags: 1 settling time 100us res1 0 num pst 1
PST 1 cpuid 761 fsb 100 mfid 2 svid c numberstates 80 500000kHz FID 04 ( 5.0) VID 0c (1400mV) 1 600000kHz FID 06 ( 6.0) VID 0c (1400mV) 2 700000kHz FID 08 ( 7.0) VID 0c (1400mV) 3 800000kHz FID 0a ( 8.0) VID 0c (1400mV) 4 900000kHz FID 0c ( 9.0) VID 0c (1400mV) 5 1000000kHz FID 0e (10.0) VID 0c (1400mV) 6 1100000kHz FID 00 (11.0) VID 0c (1400mV) 7 1200000kHz FID 02 (12.0) VID 0c (1400mV)But this is exactly what I got by changing line 352 in
powernow-k7.c:
Quoted 7 lines
/*
if ((etuple == pst->cpuid) && (maxfid==pst->maxf id) && (startvid==pst->startvid)) */ if (1==1)So - nothing to write home about. The BIOS knows only about cpuid
0x761-- does anybody know what it is? I suspect it's a TB mobile, right? See also
No. Those values are indeed for a low powered processor. Not a mobile. Plus, those values are for a FSB at 100MHz whereas yours is at 133MHz.
Quoted 2 lines
http://www.uwsg.iu.edu/hypermail/linux/kernel/0302.2/0063.html (interesting Dave thought about this exact patch as well last year).
That patch does exist btw. But it never go to mainline, and I doubt it will work now due to some recent changes in powernow-k7.
Quoted 3 lines
Just found http://seclists.org/lists/linux-kernel/2004/May/1600.html
(Google is my friend) - and Erik Meitner has the PST data on
http://emeitner.f2o.org/nx9005/files/dmesg.txt :-))
Yes, but those values are wrong unfortunately (well, SGTC at least is not OK, it should be at least 13333 IMO). If I were you, I will keep the multiplier, but not the VIDs. Something like that (a little bit modified though):
FID VID
04 ( 5.0) 0c
06 ( 6.0) 0c
08 ( 7.0) 0c
0a ( 8.0) 0c
0c ( 9.0) 0c
0e (10.0) 0c
00 (11.0) 0c
02 (12.0) 0c
15 (13.5) 0c I don't think it's possible to change VID as I said before.
And the table in order to convert VID to voltage may need to be modified (but you know that already). It's purely cosmetic, but..
Quoted 2 lines
It's bad that my combo doesn't even grok the ACPI P states... but I suspect the ACPI BIOS doesn't know anything about mobile CPUs.
Well, it's possible that a BIOS upgrade is necessary in order to get that fixed, who knows. But I do have some doubt though. The only way to detect if there is a mobile processor is to know if the VRM part is able to change the voltage of the processor at run time. IMHO, you should also printk the SFID, MFID, SVID and MVID at boot time:
arch/i386/kernel/cpu/cpufreq/powernow-k7.c +3 -0
@@ -557,6 +557,9 @@ static int __init powernow_cpu_init (strreturn -EINVAL ; } dprintk(KERN_INFO PFX "FSB: %3d.%03d MHz\n", fsb/1000, fsb%1000); + dprintk(KERN_INFO PFX "SVID: %.2x MVID: %.2x SFID: %.2x MFID: %.2x\n", + fidvidstatus.bits.SVID, fidvidstatus.bits.MVID, + fidvidstatus.bits.SFID, fidvidstatus.bits.MFID);if ((dmi_broken &BROKEN_CPUFREQ ) || acpi_force) { printk (KERN_INFO PFX "PSB/PST known to be broken. Trying ACPI instead\n");
If SVID (start VID) and MVID (max VID) are not the same, then I'm wrong, and it's possible to change voltage.
Quoted 3 lines
Anyway - I'll try making up a couple of FIDs and patch the module. If it works (or not) I'll post the patch.
Good luck!
Signature · 4 lines
Bruno Ducrot -- Which is worse: ignorance or apathy? -- Don't know. Don't care.