linux-kernel - Re: [PATCH linux-next] cpufreq: conservative: Fix sampling_down

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <51365247.5030005@semaphore.gr>
Date:	Tue, 05 Mar 2013 22:15:03 +0200
From:	Stratos Karafotis <stratosk@...aphore.gr>
To:	Viresh Kumar <viresh.kumar@...aro.org>
CC:	"Rafael J. Wysocki" <rjw@...k.pl>, cpufreq@...r.kernel.org,
	linux-pm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH linux-next] cpufreq: conservative: Fix sampling_down_factor
 functionality

On 03/05/2013 09:34 AM, Viresh Kumar wrote:
> On 5 March 2013 13:22, Stratos Karafotis <stratosk@...aphore.gr> wrote:
> I misread it here when i looked at this mail for the first time. :)
> I strongly believe that we need a full stop (.) before "Every sampling_rate",
> otherwise it looks like we check for down_factor while increasing freq :)

I agree. I will do that.

> Even now we aren't checking this 80% thing, right? And so in your patch we can
> actually fix the patch too with the right logic of code.. And
> documentation too :)

In my opinion the logic was initially correct. It was broken in the same 
commit that broke also sampling_down_factor.

Now we check if load < (cs_tuners.down_threshold - 10) to decrease freq.
Down threshold is 20, so we actually check the 80% idle.

I think the subtraction of 10 from down_threshold is wrong. It seems 
similar with ondemand but there is no logic for this in conservative.
User can simply select the down_threshold and the load will be compared 
with user's value. No need to alter user's selection.

I will prepare a patchset for these changes.

Regards,
Stratos
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/