Hello, On Tue, Jan 24, 2017 at 03:00:15PM -0800, Tim Harvey wrote: > I also have boards that uses a GPIO as an active-high enable to an > RS485 transceiver. In my device-trees I define as 'rts-gpios = <&gpio7 > 1 GPIO_ACTIVE_HIGH>'. I do find that the polarity is inverted on > master and your patch does flip the polarity. However, I share your > concern that this may not be the right way to fix it. Ack, this doesn't look right. > my understanding is that the TIOCSRS485 ioctl should have 'flags = > SER_RS485_ENABLED | SER_RS485_RTS_ON_SEND' to assert RTS on transmit. > The thing I find confusing is that in imx_start_tx the driver calls > imx_port_rts_inactive if SER_RS485_ENABLED and in imx_stop_tx the > driver calls imx_port_rts_active. To me that is what seems inverted, > not the implementation of those functions. Right, that looks fishy. It was introduced in 58362d5be3521 which keeps the behaviour of the hw rts. I.e. if SER_RS485_RTS_AFTER_SEND is set it clears UCR2_CTS. This means "The CTS pin is high (inactive)". This is wrong, too, isn't it? So the wrong commit is 17b8f2a3fdca and 58362d5be3521 only consistently expanded it. > It's confusing because the driver accounts for two options: > SER_RS485_RTS_ON_SEND (logic level for RTS pin when sending) and > SER_RS485_RTS_AFTER_SEND (logic level for RTS pin after sending). I'm > not sure why there needed to be two options here as they both seem to > indicate the same thing with an opposite polarity. I assume this is some standard that is implemented here. Maybe there are situations where you want to have RTS active all the time? Best regards Uwe -- Pengutronix e.K. | Uwe Kleine-König | Industrial Linux Solutions | http://www.pengutronix.de/ | -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
