Re: [RFC] generic iMX-spi driver

[Holger Schurig <hs4233-x6+DxXLjN1AJvtFkdXX2Hg4jNU5vUVPG@public.gmane.org>]

> +	int (*exit)(struct platform_device *pdev);

void (*exit)(struct platform_device *pdev);

> +struct spi_core_version {
> +	u32 rxdata;
> +	u32 txdata;
> +	u32 ctrl;
> +	u32 int_en;
> +	u32 status;
> +	u32 test;
> +	u32 period;
> +	u32 dma;
> +	u32 reset;
> +
> +	u32 int_te;
> +	u32 int_th;
> +	u32 int_ro;
> +	u32 int_te_en;
> +	u32 int_th_en;
> +	u32 int_ro_en;
> +	u32 xch;
> +
> +	u8 bitcount_shift;
> +	u8 sspol_shift;
> +	u8 ssctl_shift;
> +	u8 pha_shift;
> +	u8 pol_shift;
> +	u8 datarate_shift;
> +	u8 reset_val;
> +
> +	u32 max_divider;
> +	u32 max_bitcount;
> +	u32 default_ctrl;
> +	u32 (*calc_datarate)(struct driver_data *drv_data, u32
>           speed_hz, u32 *val); 
> +}; 

A number of u8's in the middle of a structure where later
32 bit long entries are comming?

Move that to the end or change it to 32 (I yet have to verify if
the compiler creates stupid or good code when accessing u8 out of
a struct, but I assume that on ARM the code is more-or-less stupid).


> +	/* Setup any GPIO active */
> +	if (platform_info->init)
> +		platform_info->init(pdev);

This function can fail, so you need to handle the error case.

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H