On 2 April 2013 17:34, Thomas Renninger trenn@suse.de wrote:
On Friday, March 29, 2013 10:40:38 PM Rafael J. Wysocki wrote:
On Friday, March 29, 2013 07:56:39 PM Viresh Kumar wrote:
printf(_(" CPUs which run at the same hardware frequency: "));
printf(_(" All (Online & Offline) CPUs that run at the same hardware frequency: "));This one is not worth changing IMO, in the end it tells the user more or less the same and as this stuff is translated, I'd not change it.
Okay.
printf(_(" CPUs which need to have their frequency coordinated by software: "));
printf(_(" Online CPUs that run at the same hardware frequency: ")); while (cpus->next) {I agree that this message is more developer than user oriented, but cpupower is more for the end-user. So this message is not perfect.
From what I can see of current code with patch aa77a52764a92216b61a6c8079b5c01937c046cd all related_cpus users are gone and related-cpus does not have any meaning at all anymore?
No, that's wrong. We need to set policy->cpus correctly (with online + offline cpus) and cpufreq core will take care of setting related_cpus with everything from policy->cpus and policy->cpus will be modified to keep only online cpus.
I haven't gone through your latest changes, but will at least give them a test on a AMD K10 multi socket machine which iirc where using related_cpus. I try to catch up with latest cpufreq changes as well, but wow... no idea when this will happen.
:)
For now I would just leave it (cpupower messages/manpage) as it is, there is nothing critical which must get fixed immediately.
Yes its not really critical but it must be fixed to reflect the right stuff.