[PATCH v3 1/3] ARM: davinci: clk framework support for enable/disable functionality

[Sekhar Nori]

On 3/25/2013 3:07 PM, Philip Avinash wrote:
> Davinci clock framework currently not supporting clock enable/disable

This is incorrect. We do support enable disable of PSCs in davinci clock
implementation.

> functionality on clock nodes. In Davinci platform, EHRPWM module requires

s/In Davinci platform/On DA850 SoC/

> support for clock enable/disable for TBCLK support. Hence this patch
> adds support for enabling/disabling clocks depends on the availability
> of the functionality.

I reworded the path description as below. Taking care of above and
making some other minor changes:

    DaVinci clock implementation does not support clock enable/disable
    functionality on non-PSC clock nodes.

    On DA850 SoC, EHRPWM module requires support for enable/disable
    of TBCLK controlled using a system module register.

    This patch adds a method for enabling/disabling non-PSC clocks
    into DaVinci clock implementation.

> 
> Signed-off-by: Philip Avinash <avinashphilip at ti.com>
> ---
> Changes since v2:
> 	- start using "usecount" field in clk structure to
> 	  enable/disable clocks.
> 
> Changes since v1:
> 	- Add support for clock enable/disable functionality.
> 
>  arch/arm/mach-davinci/clock.c |   21 ++++++++++++++-------
>  arch/arm/mach-davinci/clock.h |    2 ++
>  2 files changed, 16 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/arm/mach-davinci/clock.c b/arch/arm/mach-davinci/clock.c
> index d458558..960c7a5 100644
> --- a/arch/arm/mach-davinci/clock.c
> +++ b/arch/arm/mach-davinci/clock.c
> @@ -35,19 +35,26 @@ static void __clk_enable(struct clk *clk)
>  {
>  	if (clk->parent)
>  		__clk_enable(clk->parent);
> -	if (clk->usecount++ == 0 && (clk->flags & CLK_PSC))
> -		davinci_psc_config(clk->domain, clk->gpsc, clk->lpsc,
> -				true, clk->flags);
> +	if (clk->usecount++ == 0) {
> +		if (clk->flags & CLK_PSC)
> +			davinci_psc_config(clk->domain, clk->gpsc, clk->lpsc,
> +					true, clk->flags);

This line break can be aligned to open parenthesis. This is reported by
checkpatch --strict. I fixed this locally.

With these changed, I am queuing this patch for v3.10

Thanks,
Sekhar