Hello,

sorry, forgot to hold the mailinglist on CC:

-------- Original-Nachricht -------- Betreff: Re: [U-Boot] PATCH 2/8 Multi-adapter multi-bus I2C Datum: Tue, 10 Feb 2009 08:54:10 +0100 Von: Heiko Schocher hs@denx.de Antwort an: hs@denx.de Organisation: DENX GmbH An: ksi@koi8.net Referenzen: Pine.LNX.4.64ksi.0902061703330.14843@home-gw.koi8.net 4990175A.4020703@denx.de Pine.LNX.4.64ksi.0902091353360.32084@home-gw.koi8.net

Hello ksi,

ksi@koi8.net wrote:

On Mon, 9 Feb 2009, Heiko Schocher wrote:

...

Hello ksi,

ksi@koi8.net wrote: [...]

...

diff --git a/drivers/i2c/fsl_i2c.c b/drivers/i2c/fsl_i2c.c index ce646fd..f7998e3 100644 --- a/drivers/i2c/fsl_i2c.c

[...]

...

...

-void -i2c_init(int speed, int slaveadd)

+static void __i2c_init(int adap_no, int speed, int slaveadd) {

• struct fsl_i2c *dev; unsigned int temp;

• dev = (struct fsl_i2c *) (CONFIG_SYS_IMMR + CONFIG_SYS_I2C_OFFSET);

• writeb(0, &dev->cr); /* stop I2C controller */

• writeb(0, &i2c_dev[adap_no]->cr); /* stop I2C controller */

Why do you here substitute a "dev->" with "&i2c_dev[adap_no]"? Why not dev = &i2c_dev[adap_no] and you can let the dev-> unchanged.

This eliminates one variable, dev. Another reason is uniformity through the entire driver source.

But it is less readable (at least to me). And it distracts from the reason of the patch, making the driver multibus able. Again, please don't mix Codingstyle, Code optimization and functional changes.

Anyway, keeping that dev won't make any difference other than making the patch a tad shorter...

thanks.

bye Heiko