[PATCH v2 00/12] clk/exynos convert clock IDs to macros
Tomasz Figa
t.figa at samsung.com
Thu Sep 12 07:12:45 EDT 2013
Hi Kukjin,
On Thursday 12 of September 2013 19:50:01 Kukjin Kim wrote:
> Tomasz Figa wrote:
> > Hi Andrzej,
> >
> > On Friday 06 of September 2013 12:12:29 Andrzej Hajda wrote:
> > > Hi,
> > >
> > > This patch set adds headers with defines of exynos clocks.
> > > Then it converts dts files and drivers to use macros instead
> > > of magic numbers/enums to describe clock bindings.
> > >
> > > The patch set is based on the latest mturquette/clk-next branch.
> > >
> > > The patches are generated by script.
> > > I have verified random clocks by hand.
> > > I have also tested it on exynos4 based board.
> > >
> > > This is the 2nd version of the patchset.
> > > Changes:
> > > - corrected devicetree mailing list,
> > > - added comments to include/dt-bindings/clock/exynos4.h for
> > >
> > > clocks present only in particular chip,
> > >
> > > - added tab alignement in headers,
> > > - added comment to CLK_NR_CLKS,
> > > - added copyright headers,
> > > - split long lines in dts,
> > > - corrected example in bindings/clock/exynos5250-clock.txt, to point
> > >
> > > appropriate clocks.
> > >
> > > Regards
> > > Andrzej
> > >
> > > Andrzej Hajda (12):
> > > ARM: exynos4: create a DT header defining CLK IDs
> > > ARM: dts: exynos4: convert magic numbers to macros in clock
> > > bindings
> > > clk: exynos4: replace clock ID private enums with IDs from DT
> > > header
> > > ARM: exynos5250: create a DT header defining CLK IDs
> > > ARM: dts: exynos5250: convert magic numbers to macros in clock
> > >
> > > bindings
> > >
> > > clk: exynos5250: replace clock ID private enums with IDs from DT
> > >
> > > header
> > >
> > > ARM: exynos5420: create a DT header defining CLK IDs
> > > ARM: dts: exynos5420: convert magic numbers to macros in clock
> > >
> > > bindings
> > >
> > > clk: exynos5420: replace clock ID private enums with IDs from DT
> > >
> > > header
> > >
> > > ARM: exynos5440: create a DT header defining CLK IDs
> > > ARM: dts: exynos5440: convert magic numbers to macros in clock
> > >
> > > bindings
> > >
> > > clk: exynos5440: replace clock ID private enums with IDs from DT
> > >
> > > header
> > >
> > > .../devicetree/bindings/clock/exynos4-clock.txt | 259 +------
> > > .../devicetree/bindings/clock/exynos5250-clock.txt | 160 +---
> > > .../devicetree/bindings/clock/exynos5420-clock.txt | 184 +----
> > > .../devicetree/bindings/clock/exynos5440-clock.txt | 45 +-
> > > arch/arm/boot/dts/exynos4.dtsi | 55 +-
> > > arch/arm/boot/dts/exynos4210.dtsi | 6 +-
> > > arch/arm/boot/dts/exynos4412.dtsi | 2 +-
> > > arch/arm/boot/dts/exynos4x12.dtsi | 2 +-
> > > arch/arm/boot/dts/exynos5250.dtsi | 92 +--
> > > arch/arm/boot/dts/exynos5420.dtsi | 13 +-
> > > arch/arm/boot/dts/exynos5440.dtsi | 27 +-
> > > drivers/clk/samsung/clk-exynos4.c | 857
> > >
> > > ++++++++++----------- drivers/clk/samsung/clk-exynos5250.c
> > >
> > > | 501 ++++++------ drivers/clk/samsung/clk-exynos5420.c
> > > | |
> > >
> > > 648 ++++++++-------- drivers/clk/samsung/clk-exynos5440.c
> > >
> > > | 81 +-
> > >
> > > include/dt-bindings/clock/exynos4.h | 244 ++++++
> > > include/dt-bindings/clock/exynos5250.h | 157 ++++
> > > include/dt-bindings/clock/exynos5420.h | 188 +++++
> > > include/dt-bindings/clock/exynos5440.h | 42 +
> > > 19 files changed, 1734 insertions(+), 1829 deletions(-)
> > > create mode 100644 include/dt-bindings/clock/exynos4.h
> > > create mode 100644 include/dt-bindings/clock/exynos5250.h
> > > create mode 100644 include/dt-bindings/clock/exynos5420.h
> > > create mode 100644 include/dt-bindings/clock/exynos5440.h
> >
> > I believe this series should be rebased on top of my series (and
> > possibly remaining patches touching clock numbers queued on the list):
> > [PATCH v2 00/16] Exynos clock clean-up for 3.12
> > http://thread.gmane.org/gmane.linux.kernel.samsung-soc/22514
> >
> > Otherwise, it's really nice to finally see this being cleaned up, so:
> > Acked-by: Tomasz Figa <t.figa at samsung.com>
>
> Well...I'm not sure changing to use macro is better or not at this
> moment...
It was supposed to work like this from the beginning, but the preprocessing
was not available at the time our common clock framework support was
merged, so a temporary solution was to use the indices directly, which is
far from being easy to use and scalable.
Best regards,
Tomasz
More information about the linux-arm-kernel
mailing list