From mboxrd@z Thu Jan 1 00:00:00 1970 From: baruch@tkos.co.il (Baruch Siach) Date: Fri, 15 Apr 2016 08:41:07 +0300 Subject: [PATCH] ARM: dts: imx: ventana: add RS485 txen gpio support In-Reply-To: References: <1460639949-26392-1-git-send-email-tharvey@gateworks.com> <20160414133237.GE2092@tarshish> Message-ID: <20160415054107.GH2092@tarshish> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Hi Tim, On Thu, Apr 14, 2016 at 07:39:40AM -0700, Tim Harvey wrote: > On Thu, Apr 14, 2016 at 6:32 AM, Baruch Siach wrote: > > On Thu, Apr 14, 2016 at 06:19:09AM -0700, Tim Harvey wrote: > >> The GW52xx/GW53xx/GW54xx have an on-board RS485 transceiver for half-duplex > >> RS485 using uart1. The active-high TXEN is GPIO7__IO1 which we can configure > >> as the rts-gpio as long as we specify it as active-low to invert the polarity > >> managed by mctrl_gpio helpers. This allows for RS485 to be used from > >> userspace by setting flags to SER_RS485_RTS_ON_SEND in the serial_rs485 > >> struct when using the TIOCGRS485 ioctl. > > > > Do you use http://article.gmane.org/gmane.linux.ports.arm.kernel/482641 for > > that? > > No - that wasn't in my tree, but I will test it and respond to that thread. I see that the MAX14840 transceiver has built-in Rx disable on Tx logic, so you don't strictly need this patch. But since my patch changes the behaviour when SER_RS485_RX_DURING_TX is not set, it would be nice if you give it a spin to verify that it doesn't break you use case. I think though that a SER_RS485_DISABLE_RX_ON_TX flag that inverts the logic of SER_RS485_RX_DURING_TX would make much more sense, both on serial drivers side, and on userspace. But that's history now since we can't break the ABI. Thanks, baruch -- http://baruch.siach.name/blog/ ~. .~ Tk Open Systems =}------------------------------------------------ooO--U--Ooo------------{= - baruch at tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il -