[U-Boot] [PATCH 20/31] mpc83xx, kmeter1: extract common I2C options in keymile header

[Kim Phillips <kim.phillips@freescale.com>]

On Wed, 28 Jan 2009 10:40:16 +0100
Heiko Schocher <hs@denx.de> wrote:

> Signed-off-by: Heiko Schocher <hs@denx.de>
> ---
>  include/configs/keymile-common.h |    8 ++++----
>  include/configs/kmeter1.h        |   19 -------------------

a lot of patches in this patchseries touch a lot of the same files -
can you please reorganize the series so that we don't have to review
something that's going to change again in a subsequent patch?

>  2 files changed, 4 insertions(+), 23 deletions(-)
> 
> diff --git a/include/configs/keymile-common.h b/include/configs/keymile-common.h
> index 99c3380..4ff6fb7 100644
> --- a/include/configs/keymile-common.h
> +++ b/include/configs/keymile-common.h
> @@ -42,8 +42,6 @@
>  #define CONFIG_CMD_IMMAP
>  #define CONFIG_CMD_MII
>  #define CONFIG_CMD_PING
> -
> -/* should go away, if kmeter I2C support is enabled */
>  #define CONFIG_CMD_DTT
>  #define CONFIG_CMD_EEPROM
>  #define CONFIG_CMD_I2C

the comment alludes that something should go away yet nothing's going
away but the comment itself.

If there is indeed nothing to take away, then these are the types of
things you need to state why in your commit message.

> @@ -101,7 +99,6 @@
>  #define CONFIG_SYS_SLOT_ID_OFF		(0x07)	/* register offset */
>  #define CONFIG_SYS_SLOT_ID_MASK		(0x3f)	/* mask for slot ID bits */
> 
> -#if defined(CONFIG_MGCOGE) || defined(CONFIG_MGSUVD)
>  #define CONFIG_I2C_MULTI_BUS	1
>  #define CONFIG_I2C_CMD_TREE	1
>  #define CONFIG_SYS_MAX_I2C_BUS		2
> @@ -109,7 +106,11 @@
>  #define CONFIG_I2C_MUX		1
> 
>  /* EEprom support */
> +#if defined(CONFIG_MGCOGE) || defined(CONFIG_MGSUVD)
>  #define CONFIG_SYS_I2C_EEPROM_ADDR_LEN	1
> +#else
> +#define CONFIG_SYS_I2C_EEPROM_ADDR_LEN	2
> +#endif

if this differs per board, shouldn't it be set in the board config, and
not the common config?

Kim