[PATCH 3/3] cpufreq: exynos: Add exynos5420 cpufreq driver

[Lukasz Majewski]

Hi Arun,

> Hi Lukasz,
> 
> On Tue, Dec 17, 2013 at 1:30 PM, Lukasz Majewski
> <l.majewski at samsung.com> wrote:
> > Hi Viresh,
> >
> >> On 6 December 2013 15:18, Arun Kumar K <arun.kk at samsung.com> wrote:
> >> > From: "Arjun.K.V" <arjun.kv at samsung.com>
> >> >
> >> > The patch adds cpufreq driver for exynos5420.
> >> >
> >> > Signed-off-by: Arjun.K.V <arjun.kv at samsung.com>
> >> > Signed-off-by: Andrew Bresticker <abrestic at chromium.org>
> >> > Signed-off-by: Arun Kumar K <arun.kk at samsung.com>
> >> > ---
> >> >  drivers/cpufreq/Kconfig.arm          |   11 ++
> >> >  drivers/cpufreq/Makefile             |    1 +
> >> >  drivers/cpufreq/exynos-cpufreq.c     |    2 +
> >> >  drivers/cpufreq/exynos-cpufreq.h     |    8 +
> >> >  drivers/cpufreq/exynos5420-cpufreq.c |  346
> >> > ++++++++++++++++++++++++++++++++++ 5 files changed, 368
> >> > insertions(+) create mode 100644
> >> > drivers/cpufreq/exynos5420-cpufreq.c
> >>
> >> Hmm.. That's the second big LITTLE SoC after TC2..
> >> And we already have a generic enough driver for all big LITTLE
> >> platforms: arm_big_little.c
> >>
> >> Please use that with a glue clk and regulator driver if required.
> >> We don't want to get any new driver for big LITTLE unless there is
> >> a real limitation.
> >>
> >> Also, the generic big LITTLE driver supports IKS (In Kernel
> >> Switcher) which may or maynot be used by Exynos.
> >>
> >> @Lukasz: We can't use cpu0 driver here as we have multiple clusters
> >> here A7 and A15, and so there are two clock domains instead of one.
> >> That's why we have written big LITTLE driver.
> >
> > When I first looked into the patch - it was almost identical to the
> > exynos4210-cpufreq.c or exynos4x12-cpufreq.c files.
> >
> > Also I didn't spot the code characteristic to IKS. It seems like the
> > code in this file is not aware if it runs on A7 or A15.
> >
> 
> This initial version was only for A15s and I was planning for an
> incremental patch for A7s along with common code for the  same. Now
> that is not needed with the generic implementation in place.

Ok. I see.

> 
> Regards
> Arun



-- 
Best regards,

Lukasz Majewski

Samsung R&D Institute Poland (SRPOL) | Linux Platform Group