Re: [PATCH 22/36] serial: Sanitize rs485_struct

[Lino Sanfilippo]

On 25.06.22 at 22:12, Lukas Wunner wrote:
> On Mon, Jun 06, 2022 at 01:04:19PM +0300, Ilpo Järvinen wrote:
>> -	if (rs485->delay_rts_before_send > RS485_MAX_RTS_DELAY) {
>> +	if (!port->rs485_supported->delay_rts_before_send) {
>> +		if (rs485->delay_rts_before_send) {
>> +			dev_warn_ratelimited(port->dev,
>> +				"%s (%d): RTS delay before sending not supported\n",
>> +				port->name, port->line);
>> +		}
>> +		rs485->delay_rts_before_send = 0;
>> +	} else if (rs485->delay_rts_before_send > RS485_MAX_RTS_DELAY) {
>>  		rs485->delay_rts_before_send = RS485_MAX_RTS_DELAY;
>>  		dev_warn_ratelimited(port->dev,
>>  			"%s (%d): RTS delay before sending clamped to %u ms\n",
>>  			port->name, port->line, rs485->delay_rts_before_send);
>>  	}
>
> This series seems to set rs485_supported->delay_rts_before_send to 1
> in all drivers to indicate that a delay is supported.
>
> It would probably be smarter to define it as a maximum, i.e. drivers
> declare the supported maximum delay in their rs485_supported struct
> and the core can use that to clamp the value.  Initially, all drivers
> may use RS485_MAX_RTS_DELAY.  Some chips only support specific delays
> (multiples of the UART clock or baud clock).  We can amend their
> drivers later according to their capabilities.

Agreed.

Regards,
Lino