[U-Boot] [PATCH v2] spi: kirkwood: avoid divide by zero crash

[Heinrich Schuchardt]

On 11/12/2017 04:30 PM, Baruch Siach wrote:
> Calling .set_speed with zero speed is definitely a bug. Instead of crashing,
> set hz to 1 so that we get the lowest possible frequency rate.

Did this actually occur? Why?

> 
> Signed-off-by: Baruch Siach <baruch at tkos.co.il>
> ---
> v2: Don't fail; set lowest rate (Jagan)
> ---
>   drivers/spi/kirkwood_spi.c | 3 +++
>   1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/spi/kirkwood_spi.c b/drivers/spi/kirkwood_spi.c
> index 0c6bd295cde9..5744849a5ad9 100644
> --- a/drivers/spi/kirkwood_spi.c
> +++ b/drivers/spi/kirkwood_spi.c
> @@ -257,6 +257,9 @@ static int mvebu_spi_set_speed(struct udevice *bus, uint hz)
>   	struct kwspi_registers *reg = plat->spireg;
>   	u32 data;
>   
> +	/* Avoid divide by zero; instead, set the lowest possible speed */
> +	if (hz == 0)
> +		hz = 1;

Shouldn't we write a debug message here like in other drivers?

debug("%s: can't set 0 Hz (actual will be 1 Hz)\n", __func__);

Does any bus member support 1 Hz? Wouldn't 125 kHz be a more reasonable 
value?

Could a division by zero valued max_hz occur in spi_setup_slave too?

Best regards

Heinrich

>   	/* calculate spi clock prescaller using max_hz */
>   	data = ((CONFIG_SYS_TCLK / 2) / hz) + 0x10;
>   	data = data < KWSPI_CLKPRESCL_MIN ? KWSPI_CLKPRESCL_MIN : data;
>