kevinlekiller/amdctl

Amdctl reports incorrect frequencies

Closed this issue · 11 comments

echaskaris commented

I have AMD FX-4170. Top frequency is 8400 mhz, but the CPU is rated for 4200.

kevinlekiller commented

It might be setting the wrong FSB speed, some bulldozer CPU's are 100 MHz, some are 200, it multiplies the FSB by the multiplier to get the frequency (ex 42 * 100 = 4200).

echaskaris commented

What field is the revision in /proc/cpuinfo?

echaskaris commented

/proc/cpuinfo contents:
processor : 0
vendor_id : AuthenticAMD
cpu family : 21
model : 1
model name : AMD FX(tm)-4170 Quad-Core Processor
stepping : 2
microcode : 0x600063e
cpu MHz : 1630.357
cache size : 2048 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 2
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt fma4 nodeid_msr topoext perfctr_core perfctr_nb cpb hw_pstate ssbd ibpb vmmcall arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
bugs : fxsave_leak sysret_ss_attrs null_seg spectre_v1 spectre_v2 spec_store_bypass
bogomips : 8437.44
TLB size : 1536 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 48 bits physical, 48 bits virtual
power management: ts ttp tm 100mhzsteps hwpstate cpb

kevinlekiller commented

Yeah, you're in between 0 and 0f (model : 1), I wasn't sure about the math on those, because the manual says they have a FSB of 200 MHz, if you look through, search for cpudid / cpu fid / nbdid / nbfid / corecof, you will find some explanations of how the CPU frequency is calculated, https://www.amd.com/system/files/TechDocs/42301_15h_Mod_00h-0Fh_BKDG.pdf

If I have time this weekend, I will look to see how it is done on these models.

echaskaris commented

Thanks!

kevinlekiller commented

You could try commenting out this section https://github.com/kevinlekiller/amdctl/blob/master/amdctl.c#L390-L392 and see if any values look off.

echaskaris commented

They look okay without that section. (4200 base, 4300 boost)

kevinlekiller commented

Alright, the only thing I wonder is if it affects the north bridge clock calculations, maybe it requires the math on 200 MHz for north bridge calculations and 100 MHz for core clock calculations ?

Referring to this block of code I mean (which shows north bridge information - if refclck is now 100, do those numbers output wrong?): https://github.com/kevinlekiller/amdctl/blob/master/amdctl.c#L318-L328

kevinlekiller commented

If so, this line might need to be changed: https://github.com/kevinlekiller/amdctl/blob/master/amdctl.c#L328

(((cpuModel >= 0x00 && cpuModel <= 0x0f) ? REFCLK * 2 : REFCLK) * (nbdid + 0x4) / pow(2, nbfid))

priand1 commented

Hi Kevin, thanks for amdctl. It works quite well. I also have the problem that echaskaris has, but with an FX-8320.
I was able to fix it by making the two changes mentioned above in amdctl.c:
#14 (comment)

and
#14 (comment)

My /proc/cpuinfo:
processor : 7
vendor_id : AuthenticAMD
cpu family : 21
model : 2
model name : AMD FX(tm)-8320 Eight-Core Processor
stepping : 0
microcode : 0x6000852
cpu MHz : 1399.187
cache size : 2048 KB
physical id : 0
siblings : 8
core id : 7
cpu cores : 4
apicid : 23
initial apicid : 7
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 popcnt aes xsave avx f16c lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 tce nodeid_msr tbm topoext perfctr_core perfctr_nb cpb hw_pstate ssbd ibpb vmmcall bmi1 arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
bugs : fxsave_leak sysret_ss_attrs null_seg spectre_v1 spectre_v2 spec_store_bypass
bogomips : 6999.84
TLB size : 1536 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 48 bits physical, 48 bits virtual
power management: ts ttp tm 100mhzsteps hwpstate cpb eff_freq_ro

Thank you, Andreas