Messages in this thread | | | Date | Wed, 17 Jul 2013 09:36:14 +0200 | From | Lukasz Majewski <> | Subject | Re: [PATCH v5 7/7] Documentation:cpufreq:boost: Update BOOST documentation |
| |
On Wed, 17 Jul 2013 10:52:18 +0530 Viresh Kumar viresh.kumar@linaro.org wrote, > On 16 July 2013 17:03, Lukasz Majewski <l.majewski@samsung.com> wrote: > > As fair as I've understood our previous discussion (at [*]) we have > > agreed about this. We only export boost attribute when it is > > supported by cpufreq_driver. Rafael was very clear about exporting > > boost attribute: > > > > "Simple: Export it only when supported." > > > > [*]: Re: [PATCH v4 2/7] cpufreq: Add boost frequency support in core > > I don't want to search in the Archives but I am sure what we decided > earlier.
At v4 there was the old acpi-cpufreq.c behaviour preserved (with always exporting boost - when not supported ro, when supported rw).
Due to Rafael and Dirk comments it has been rewritten at v5:
http://thread.gmane.org/gmane.linux.kernel/1511831/match=patch+v4+2+7+cpufreq+add+boost+frequency+support+core
> > Rafael Said: > "Simple: Export it only when supported." [*]
> > AND > > "Don't change behavior of acpi-cpufreq driver" [**] > > If you see acpi-cpufreq driver carefully, it always creates "boost" > sysfs entry. If its not supported then it creates a read only entry.
For me those two statements [*] and [**] contradict:
At v5: 1. ARM - export "boost" only when supported (rw) 2. x86 - export boost only when x86 supports it (as rw). When x86 doesn't support HW boost - DO NOT export it at all.
At v4: 1. ARM - export "boost" only when supported (rw) 2. x86 - always export boost - no matter if supported or not. If not supported, then export it as ro only.
> > So, please preserve that.
-- Best regards,
Lukasz Majewski
Samsung R&D Institute Poland (SRPOL) | Linux Platform Group
| |