powerd freeze with amd 5000 X2 but not with lower cpus

[JoaoBR]

On Saturday 28 July 2007 10:46:15 Michael Nottebrock wrote:
> >>
> >> Good to hear. Unfortunately my Mobo's BIOS is already up-to-date, the
> >> CPU is officially supported and Cool'n'Quiet works dandy in Windows XP
> >> ... =/
> >
> > my ok msg was too fast, after some time my video starts flickering and
> > stays so and the PC is freezed up, disabling powerd and works stable
>
> Meanwhile I found a workaround for my system: I had SCHED_ULE configured
> in my kernel - switching to SCHED_4BSD gets rid of the freezes. Should
> have thought of that sooner, ISTR having problems with powerd and
> SCHED_ULE even on single cpu P4s.
>


I spend some time here on different MBs with the same 5000+ cpu, comparing 
i386 and amd64

I cvsuped both sources and compiled world and kernel an hour ago

firstable, smp (either 4BSD or ULE) with powerd_enable (cpufreq in kernel) 
works as long as no xorg is started (GENERIC, no sysctl, no loader.conf 
options)

nevertheless strange, on all MBs, same CPU:

i386:
dev.cpu.0.freq_levels: 2600/100000 2400/85596 2200/72544 2000/60778 1800/50237 
1575/43957 1350/37677 1125/31398 1000/25535 875/22343 750/19151 625/15959 
500/12767 375/9575 250/6383 125/3191

amd64:dev.cpu.0.freq_levels: 2600/100000 2400/85596 2200/72544 2000/60778 
1800/50237 1000/25535


both, amd64 and i386 with same ports and up to date

wether 4BSD or ULE on i386 freeze soon as xorg starts when powerd is enabled

ULE freeze on xorg start before coming up and 4BSD comes up and works for a 
certain time on amd64 (5-30 minutes) 


all kernels are compiled with make.conf option "CPUTYPE?=athlon64" beeing 
translated appearently to the same "athlon-mp" option when available

somebody knows why I get different frequencies under i386 and amd64 for the 
same hardware?



-- 

João







A mensagem foi scaneada pelo sistema de e-mail e pode ser considerada segura.
Service fornecido pelo Datacenter Matik  https://datacenter.matik.com.br