When in RS485 emulation mode, __do_stop_tx_rs485() calls
serial8250_clear_fifos().  This not only clears the FIFOs, but also sets
all bits in their control register (UART_FCR) to 0.

One of the effects of this is the disabling of the FIFOs, which turns
them into single-byte holding registers.  The rest of the driver doesn't
know this, which results in the lions share of characters passed into a
write call to be dropped.

(I can supply logic analyzer screenshots if necessary)

This fix replaces the serial8250_clear_fifos() call to
serial8250_clear_and_reinit_fifos() - this prevents the "dropped
characters" issue from manifesting again while retaining the requirement
of clearing the RX FIFO after transmission if the SER_RS485_RX_DURING_TX
flag is disabled.
Signed-off-by: NDaniel Jedrychowski <avistel@gmail.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Expose set_ldisc() function so that it can be overridden with a
platform specific implementation.
Signed-off-by: NEd Blake <ed.blake@imgtec.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The following fix of runtime PM use in DMA mode requires at least
serial8250_rpm_put_tx() to be available. Export both calls.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

NXP SC16C2552 requires that we always write a reset to the RX FIFO and
TX FIFO whenever we enable the FIFOs

Cc: xe-kernel@external.cisco.com
Signed-off-by: NSteve Shih <sshih@cisco.com>
Signed-off-by: NDavid Singleton <davsingl@cisco.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The Altera 16550 soft IP UART requires 2 additional registers for
TX FIFO threshold support. These 2 registers enable the TX FIFO
Low Watermark and set the TX FIFO Low Watermark.
Set the TX FIFO threshold to the FIFO size - tx_loadsz.
Signed-off-by: NThor Thayer <tthayer@opensource.altera.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

First loop in wait_for_xmitr could also trigger NMI
watchdog in case reading from the port is slow:

  PID: 0      TASK: ffffffff819c1460  CPU: 0   COMMAND: "swapper/0"
   #0 [ffff88019f405e58] crash_nmi_callback at ffffffff8104d382
   #1 [ffff88019f405e68] nmi_handle at ffffffff8168ead9
   #2 [ffff88019f405eb0] do_nmi at ffffffff8168ec53
   #3 [ffff88019f405ef0] end_repeat_nmi at ffffffff8168df13
      [exception RIP: delay_tsc+50]
      RIP: ffffffff81325642  RSP: ffff88019f403bb0  RFLAGS: 00000083
      RAX: 00000000000005c8  RBX: ffffffff81f83000  RCX: 0000024e4fb88a8b
      RDX: 0000024e4fb89053  RSI: 0000000000000000  RDI: 00000000000007d1
      RBP: ffff88019f403bb0   R8: 000000000000000a   R9: 0000000000000000
      R10: 0000000000000000  R11: ffff88019f403ad6  R12: 000000000000250f
      R13: 0000000000000020  R14: ffffffff81d360c7  R15: 0000000000000047
      ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
  --- <NMI exception stack> ---
   #4 [ffff88019f403bb0] delay_tsc at ffffffff81325642
   #5 [ffff88019f403bb8] __const_udelay at ffffffff813255a8
   #6 [ffff88019f403bc8] wait_for_xmitr at ffffffff81404390
   #7 [ffff88019f403bf0] serial8250_console_putchar at ffffffff8140455c
   #8 [ffff88019f403c10] uart_console_write at ffffffff813ff00a
   #9 [ffff88019f403c40] serial8250_console_write at ffffffff814044ae
  #10 [ffff88019f403c88] call_console_drivers.constprop.15 at ffffffff81086b01
  #11 [ffff88019f403cb0] console_unlock at ffffffff8108842f
  #12 [ffff88019f403ce8] vprintk_emit at ffffffff81088834
  #13 [ffff88019f403d58] vprintk_default at ffffffff81088ba9
  #14 [ffff88019f403d68] printk at ffffffff8167f034

Adding touch_nmi_watchdog call to the first loop as well.
Reported-by: NChunyu Hu <chuhu@redhat.com>
Signed-off-by: NJiri Olsa <jolsa@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Unify the check of em485 variable to be either (em485) or (!em485) instead of
the explicit comparison to NULL.

While here, remove redundant check in __do_stop_tx_rs485() and
__stop_tx_rs485() since the functions ain't called with NULL value of em485
variable.

Cc: "Matwey V. Kornilov" <matwey@sai.msu.ru>
Tested-by: NYegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

There are calls to serial8250_rpm_{get|put}() in __do_stop_tx_rs485() that are
certainly placed in a wrong location. I dunno how it had been tested with
runtime PM enabled because it is obvious "sleep in atomic context" error.

Besides that serial8250_rpm_get() is called immediately after an IO just
happened. It implies that the device is already powered on, see implementation
of serial8250_em485_rts_after_send() and serial8250_clear_fifos() for the
details.

There is no bug have been seen due to, as I can guess, use of auto suspend mode
when scheduled transaction to suspend is invoked quite lately than it's needed
for a few writes to the port. It might be possible to trigger a warning if
stop_tx_timer fires when device is suspended.

Refactor the code to use runtime PM only in case of timer function.

Fixes: 0c66940d ("tty/serial/8250: fix RS485 half-duplex RX")
Cc: "Matwey V. Kornilov" <matwey@sai.msu.ru>
Tested-by: NYegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Intel Quark has 16550A compatible UART with autoflow feature enabled. It has
only 16 bytes of FIFO. Currently serial8250_do_set_termios() prevents to enable
autoflow since the minimum requirement of 32 bytes of FIFO size.

Drop a FIFO size limitation to allow autoflow control be enabled on such UARTs.

While here, comment out UART_CAP_AFE for PORT_AR7 since it wasn't working and
it will be not a good idea to use it in conjunction with trigger level of 1
byte.
Suggested-by: NPeter Hurley <peter@hurleysoftware.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: NBryan O'Donoghue <pure.logic@nexus-software.ie>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

mediatek can support baud rate up to 4M.
the 'uart_get_baud_rate' function will limit the max baud rate.
Modify max baud to remove the limit.
Signed-off-by: NLong Cheng <long.cheng@mediatek.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Serial console is broken in v4.8-rcX. Mika and I independently bisected down to
commit 4ef03d32 ("tty/serial/8250: use mctrl_gpio helpers").

Since neither author nor anyone else didn't propose a solution we better revert
it for now.

This reverts commit 4ef03d32.

Link: https://lkml.kernel.org/r/20160809130229.GN1729@lahna.fi.intel.comSigned-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: NHeikki Krogerus <heikki.krogerus@linux.intel.com>
Tested-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch permits the usage for GPIOs to control
the CTS/RTS/DTR/DSR/DCD/RI signals.
Signed-off-by: NYegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Introduce serial8250_out_MCR() and serial8250_in_MCR() routines, that
replace following calls:

serial_out(port, UART_MCR, val)
serial_port_out(up, UART_MCR, val)
serial_in(port, UART_MCR)

This patch is needed in order to integrate reading/writing of MCR
signals via SERIAL_MCTRL_GPIO infrastructure later.
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NYegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The compiler complains about variables that are set, but never used:
* intX variables in exar_handle_irq
  drivers/tty/serial/8250/8250_port.c:1864:34: warning: variable ‘int3’ set but not used [-Wunused-but-set-variable]
* val variable in pci_quatech_wqopr
  drivers/tty/serial/8250/8250_pci.c:1139:10: warning: variable ‘val’ set but not used [-Wunused-but-set-variable]

And about a shadow variable:
* tmout in wait_for_xmitr is defined twice with the same type. Both of
  them are also initialized before use.

Remove all of them.
Signed-off-by: NJiri Slaby <jslaby@suse.cz>
Cc: Matt Schulte <matts@commtech-fastcom.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Convert DEBUG_INTR to pr_debug:
* defined semantics (DEBUG, DYNAMIC_DEBUG)
* KERN_DEBUG level instead of KERN_DEFAULT
* emit __func__ and \n
* verified 'fmt' even when !DEBUG

I wonder if anybody ever used that or whether we should just drop the
lines.
Signed-off-by: NJiri Slaby <jslaby@suse.cz>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The 8250_fintek driver advertises as the PNP0501 driver; however this
conflicts with the standard 16550A uart PNP0501. The conflict causes
the 8250_fintek driver to load with _every_ PNP0501, but never probe,
and causing the entire 8250 driver stack to unload if the 8250_fintek
driver is unloaded (modprobe doesn't know that 8250_pnp rather than
8250_fintek claimed the resource).

This patch merges the Fintek driver into 8250_base. On autoconfig_16550
the device is probed to verify if it is a FINTEK device or not.

This custom probing can be disabled completely via configuration. When a
Fintek device is not probed it will behave as a standard 16550A device,
with no RS485 capabilities.
Reported-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NRicardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Using fake IIR values to perform rx dma operations unnecessarily
conflates separate operations, stopping in-progress dma with starting
new dma.

Introduce serial8250_rx_dma_flush() to handle stopping in-progress dma
[omap8250 already has equivalent omap_8250_rx_dma_flush()].
Replace rx_dma(UART_IIR_RX_TIMEOUT) with the equivalent *_rx_dma_flush(),
and rx_dma(0) with the equivalent *_rx_dma(). Handle IIR steering
in the irq handler with handle_rx_dma() helper.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Reviewed-by: NHeikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

If !up->dma == F, then up->dma == T and can be elided from the
(up->dma && up->dma->tx_err) sub-expression.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

dma_err is not used other than for boolean evaluation;
substitute the actual operation directly.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Exposes get_mctrl() function so that it can be overriden with platform
specific implementation.
Signed-off-by: NWan Ahmad Zainie <wan.ahmad.zainie.wan.mohamad@intel.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NHeikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When in half-duplex mode RX will be disabled before TX, but not
enabled after deactivating transmitter. This patch enables
UART_IER_RLSI and UART_IER_RDI interrupts after TX is over.

Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Signed-off-by: NYegor Yefremov <yegorslists@googlemail.com>
Fixes: e490c914 ("tty: Add software emulated RS485 support for 8250")
Acked-by: NMatwey V. Kornilov <matwey@sai.msu.ru>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

up_to_u8250p() casts struct uart_port * to struct uart_8250_port *. Update code
to use it instead of open coded variant.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

serial8250_em485_init() is supposed to be protected with
p->port.lock spinlock.
This may lead to issues when kmalloc sleeps, so it is better to use
GFP_ATOMIC in this spinlocked context.

Fixes: e490c914 ("tty: Add software emulated RS485 support for 8250")
Reported-by: NИльяс Гасанов <torso.nafi@gmail.com>
Signed-off-by: NMatwey V. Kornilov <matwey@sai.msu.ru>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Checkpatch complains about incorrect indentation of switch/case statements.
This patch fixes the corresponding warnings. Additionally some indentation
is changed to match the correct format specified in the Linux Kernel
Coding Style.
Signed-off-by: NAnton Würfel <anton.wuerfel@fau.de>
Signed-off-by: NPhillip Raffeck <phillip.raffeck@fau.de>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: linux-kernel@i4.cs.fau.de
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch replaces printk by the corresponding variant of dev_* in order
to fix checkpatch warnings. If no suitable device pointer is present, the
corresponding pr_* variant is used.
Signed-off-by: NPhillip Raffeck <phillip.raffeck@fau.de>
Signed-off-by: NAnton Würfel <anton.wuerfel@fau.de>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: linux-kernel@i4.cs.fau.de
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Checkpatch outputs some warnings about incorrect comment style,
which is fixed by this patch.
Signed-off-by: NAnton Würfel <anton.wuerfel@fau.de>
Signed-off-by: NPhillip Raffeck <phillip.raffeck@fau.de>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: linux-kernel@i4.cs.fau.de
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Indentation is changed to match the correct format of using tabs instead
of spaces wherever possible.
Signed-off-by: NAnton Würfel <anton.wuerfel@fau.de>
Signed-off-by: NPhillip Raffeck <phillip.raffeck@fau.de>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Cc: linux-kernel@i4.cs.fau.de
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch fixes whitespace errors reported by checkpatch to increase
readability. Main focus is on missing spaces after commas in
function headers and macros (like foo,bar edited to foo, bar).
Signed-off-by: NAnton Würfel <anton.wuerfel@fau.de>
Signed-off-by: NPhillip Raffeck <phillip.raffeck@fau.de>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: linux-kernel@i4.cs.fau.de
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Some UART HW has a single register combining UART_DLL/UART_DLM
(this was probably forgotten in the change that introduced the
callbacks, commit b32b19b8)

Fixes: b32b19b8 ("[SERIAL] 8250: set divisor register correctly ...")
Signed-off-by: NSebastian Frias <sf84@laposte.net>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Implementation of software emulation of RS485 direction handling is based
on omap_serial driver.
Before and after transmission RTS is set to the appropriate value.

Note that before calling serial8250_em485_init() the caller has to
ensure that UART will interrupt when shift register empty. Otherwise,
emultaion cannot be used.

Both serial8250_em485_init() and serial8250_em485_destroy() are
idempotent functions.
Signed-off-by: NMatwey V. Kornilov <matwey@sai.msu.ru>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Software RS485 emultaion is to be added in the following commit.
serial8250_start_tx() will need to refer serial8250_stop_rx().
Move serial8250_stop_rx() in front of serial8250_start_tx() in order
to avoid function forward declaration.
Signed-off-by: NMatwey V. Kornilov <matwey@sai.msu.ru>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

After masking all interrupts, wait for the irq handler to complete
before continuing shutdown.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The defunct low_latency input steering executed flush_to_ldisc()
directly from interrupt context so dropping the port lock was
necessary to avoid deadlock. That steering was removed by

    commit a9c3f68f
    Author: Peter Hurley <peter@hurleysoftware.com>
    Date:   Sat Feb 22 07:31:21 2014 -0500

    tty: Fix low_latency BUG
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When max_count is reached, the rx loop exits. However, UART_LSR has
already been read so those char flags are lost, and subsequent rx
status will be for the wrong byte until the rx fifo drains.
Reported-by: NGeorge Spelvin <linux@horizon.com>
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor the read/process one char inner loop to a separate helper
function serial8250_read_char(). No functional change.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently, 8-bit (MMIO) and 32-bit (MMIO32) register interfaces are
supported for the 8250 console, but the 16-bit (MMIO16) is not.
The 8250 UART device on my board is connected to a 16-bit bus and
my main motivation is to use earlycon with it.
(Refer to arch/arm/boot/dts/uniphier-support-card.dtsi)
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This adds an entry to the uart_config table for PORT_RT2880
enabling rx/tx FIFOs.  The UART is actually a Palmchip BK-3103
which is found in several devices from Alchemy/RMI, Ralink, and
Sigma Designs.
Signed-off-by: NMans Rullgard <mans@mansr.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Move h/w reinit of serial console restore-from-suspend into
standalone helper function.

No functional change.
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When the UART clock is set slightly under 1.8432MHz, the 8250 driver
core doesn't permit the 115200 baud rate since it calculates the maximum
frequency to pass to uart_get_baud_rate by simply dividing the uart
clock by 16 which yields a value slightly under 115200, even though the
frequency is close enough for the UART to operate reliably.

Therefore add some tolerance in the calculation of the maximum baud
rate. 1% tolerance allows for marginally slower uart clk than nominal
without introducing transmission errors.
Signed-off-by: NJames Hogan <james.hogan@imgtec.com>
[pjh: Forward-port & refactor original patch; change tolerance to 1%]
Signed-off-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>