[U-Boot] [PATCH 4/7] Enable rkclk init function

From: Heiko Stuebner <heiko@sntech.de>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH 4/7] Enable rkclk init function
Date: Sun, 02 Apr 2017 13:52:07 +0200	[thread overview]
Message-ID: <12862552.kjCuFfrPQm@phil> (raw)
In-Reply-To: <1491057777-1640-5-git-send-email-eric.gao@rock-chips.com>

Hi Eric,

Am Samstag, 1. April 2017, 22:42:54 CEST schrieb eric.gao at rock-chips.com:
> From: "eric.gao" <eric.gao@rock-chips.com>
> 
> Signed-off-by: eric.gao <eric.gao@rock-chips.com>
> ---
> 
>  drivers/clk/rockchip/clk_rk3399.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/clk/rockchip/clk_rk3399.c b/drivers/clk/rockchip/clk_rk3399.c
> index 922ce7e..cfcd1e4 100644
> --- a/drivers/clk/rockchip/clk_rk3399.c
> +++ b/drivers/clk/rockchip/clk_rk3399.c
> @@ -20,6 +20,10 @@
>  
>  DECLARE_GLOBAL_DATA_PTR;
>  
> +#ifndef CONFIG_SPL_BUILD
> +#define CONFIG_SPL_BUILD
> +#endif
> +

That looks very strange. In all cases your commit message should explain
why a change is necessary, i.e. what you want to achieve.

In this special case, also rkclk_init runs from SPL context and inits the
clocks already, the full u-boot then finds the clocks already set up as
necessary and should have no need to for rkclk_init at all itself.

Heiko