linux-kernel - Re: [PATCH 1/9] serial: core: move RS485 configuration tasks from drivers into core

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 14 Feb 2022 17:13:20 +0100
From:   Lino Sanfilippo <LinoSanfilippo@....de>
To:     Jiri Slaby <jirislaby@...nel.org>, gregkh@...uxfoundation.org
Cc:     linux@...linux.org.uk, richard.genoud@...il.com,
        nicolas.ferre@...rochip.com, alexandre.belloni@...tlin.com,
        ludovic.desroches@...rochip.com, shawnguo@...nel.org,
        s.hauer@...gutronix.de, kernel@...gutronix.de, festevam@...il.com,
        linux-imx@....com, mcoquelin.stm32@...il.com,
        alexandre.torgue@...s.st.com, linux-serial@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-stm32@...md-mailman.stormreply.com, lukas@...ner.de
Subject: Re: [PATCH 1/9] serial: core: move RS485 configuration tasks from
 drivers into core


Hi,

On 14.02.22 at 06:41, Jiri Slaby wrote:
> On 13. 02. 22, 23:27, Lino Sanfilippo wrote:
>> Several drivers that support setting the RS485 configuration via userspace
>> implement on or more of the following tasks:
>>
>> - in case of an invalid RTS configuration (both RTS after send and RTS on
>>    send set or both unset) fall back to enable RTS on send and disable RTS
>>    after send
>>
>> - nullify the padding field of the returned serial_rs485 struct
>>
>> - copy the configuration into the uart port struct
>>
>> - limit RTS delays to 100 ms
>>
>> Move these tasks into the serial core to make them generic and to provide
>> a consistent beheviour among all drivers.
>>
>> Signed-off-by: Lino Sanfilippo <LinoSanfilippo@....de>
>> ---
>>   drivers/tty/serial/serial_core.c | 13 +++++++++++++
>>   1 file changed, 13 insertions(+)
>>
>> diff --git a/drivers/tty/serial/serial_core.c b/drivers/tty/serial/serial_core.c
>> index 846192a7b4bf..3fab4070359c 100644
>> --- a/drivers/tty/serial/serial_core.c
>> +++ b/drivers/tty/serial/serial_core.c
>> @@ -1282,8 +1282,21 @@ static int uart_set_rs485_config(struct uart_port *port,
>>       if (copy_from_user(&rs485, rs485_user, sizeof(*rs485_user)))
>>           return -EFAULT;
>>   +    /* pick sane settings if the user hasn't */
>> +    if (!(rs485.flags & SER_RS485_RTS_ON_SEND) ==
>> +        !(rs485.flags & SER_RS485_RTS_AFTER_SEND)) {
>> +        rs485.flags |= SER_RS485_RTS_ON_SEND;
>> +        rs485.flags &= ~SER_RS485_RTS_AFTER_SEND;
>> +    }
>> +    /* clamp the delays to [0, 100ms] */
>> +    rs485.delay_rts_before_send = min(rs485.delay_rts_before_send, 100U);
>> +    rs485.delay_rts_after_send = min(rs485.delay_rts_after_send, 100U);
>
> Why is this magic 100?

The only drivers that seem to care about a max value for the RTS delays use 100 ms
(omap-serial, amba pl011, 8250) so I chose this to stay compatible with the current
driver implementations. 100 ms also seems large enough to be used as a general max value.

> Can we have that number somehow documented? You should define a macro for that anyway.

Ok, I will do so.

>
>> +    memset(rs485.padding, 0, sizeof(rs485.padding));
>
> What is this memset good for?

Drivers like max310x, amba-pl011, 8250_pci, 8250_fintek, 8250_lpc18xx seem to care about
returning a serial_rs485 struct with cleared padding field to userspace. So they all clear
that field on their own. Although not really necessary, to me this seems to be a good
default behavior, so I added it to the serial core.

>
>>       spin_lock_irqsave(&port->lock, flags);
>>       ret = port->rs485_config(port, &rs485);
>> +    if (!ret)
>> +        port->rs485 = rs485;
>>       spin_unlock_irqrestore(&port->lock, flags);
>>       if (ret)
>>           return ret;
>
> thanks,

Thanks for the review!

Regards,
Lino