[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <1461934714-18681-1-git-send-email-geert+renesas@glider.be>
Date: Fri, 29 Apr 2016 14:58:23 +0200
From: Geert Uytterhoeven <geert+renesas@...der.be>
To: Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Jiri Slaby <jslaby@...e.com>,
Peter Hurley <peter@...leysoftware.com>
Cc: Magnus Damm <magnus.damm@...il.com>,
Laurent Pinchart <laurent.pinchart+renesas@...asonboard.com>,
Yoshinori Sato <ysato@...rs.sourceforge.jp>,
linux-serial@...r.kernel.org, linux-renesas-soc@...r.kernel.org,
linux-sh@...r.kernel.org, linux-kernel@...r.kernel.org,
Geert Uytterhoeven <geert+renesas@...der.be>
Subject: [PATCH v2 00/11] serial: sh-sci: Hardware Flow Control Updates
Hi all,
This patch series contains updates to the Renesas SCI UART driver,
related to hardware flow control:
- Patches 1-2: Device Tree binding updates for GPIO-controlled modem
lines, and for dedicated RTS/CTS modem lines,
- Patches 3-4: Driver support for GPIO-controlled modem lines, using
the serial_mctrl_gpio helpers,
- New patches 5-11: Driver support for hardware-assisted (automatic)
hardware flow control using the dedicated RTS/CTS modem lines.
Changes compared to v1 (more details in the individual patches):
- Add Acked-by,
- Refer to the Generic Serial DT Bindings,
- Drop out[12]-gpios,
- Reject combining GPIO and dedicated RTS/CTS,
- Drop "serial: sh-sci: Replace SCIx_HAVE_RTSCTS by standard
UPF_HARD_FLOW",
- Add support for hardware-assisted automatic RTS/CTS control on
(H)SCIF, SCIFA, and SCIFB.
Dependencies:
- The patches apply to next-20160429,
- The DT bindings refer to the Generic Serial DT Bindings that were
added by "doc: DT: Add Generic Serial Device Tree Bindings", which
is in tty-testing.
This was tested on r8a7791/koelsch and r8a7740/armadillo, using UARTs
and GPIO pins on expansion connectors.
Expected behavior, using GPIO-based RTS/CTS as a reference:
- When CRTSCTS is disabled:
- Reading from a UART asserts the RTS line,
- Writing to a UART asserts the RTS line.
- When CRTSCTS is enabled:
- Reading from a UART asserts the RTS line,
- Writing to a UART asserts the RTS line.
- Writing is blocked until CTS is asserted.
Hardware-assisted hardware flow control with HSCIF and SCIFB on
r8a7791/koelsch behaves as expected when dedicated RTS/CTS modem lines
are enabled (beware the large transmission FIFOs on SCIFB ;-)
For SCIFA on r8a7740/armadillo, manually overriding (asserting) the RTS
signal doesn't seem to work. Hence when CRTSCTS is disabled, RTS is
never asserted. When CRTSCTS is enabled, hardware-assisted RTS/CTS does
control RTS as expected. As SCIFA on r8a7791 doesn't have dedicated
RTS/CTS modem lines, I cannot compare.
DT overlays for testing can be found in the topic/renesas-overlays
branch of
https://git.kernel.org/cgit/linux/kernel/git/geert/renesas-drivers.git.
Regression testing on platforms without DT and/or GPIOLIB support
(SuperH) would be appreciated.
Compile-tested on ecovec24_defconfig(GPIOLIB=y) and se7780_defconfig
(GPIOLIB=n).
Thanks!
Geert Uytterhoeven (11):
serial: sh-sci: Update DT binding documentation for GPIO modem lines
serial: sh-sci: Update DT binding documentation for dedicated RTS/CTS
serial: sh-sci: Always set TIOCM_CTS in .get_mctrl() callback
serial: sh-sci: Add support for GPIO-controlled modem lines
serial: sh-sci: Do not open-code sci_getreg()
serial: sh-sci: Add more Serial Port Register documentation
serial: sh-sci: Add more Serial Port Control/Data Register
documentation
serial: sh-sci: Correct pin initialization on (H)SCIF
serial: sh-sci: Add pin initialization for SCIFA/SCIFB
serial: sh-sci: Fix support for hardware-assisted RTS/CTS
serial: sh-sci: Add DT support for dedicated RTS/CTS
.../bindings/serial/renesas,sci-serial.txt | 4 +
drivers/tty/serial/Kconfig | 1 +
drivers/tty/serial/sh-sci.c | 177 ++++++++++++++++++---
drivers/tty/serial/sh-sci.h | 24 ++-
4 files changed, 173 insertions(+), 33 deletions(-)
--
1.9.1
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...ux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Powered by blists - more mailing lists