This patch adds RXBS register access support for zynqmp.
To avoid the corner error conditions it will consider only
RXBS[2:0] bits while checking the error conditions
(Parity,Framing and BRAK).
Signed-off-by: NNava kishore Manne <navam@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Check for uart_ops structures that are only stored in the ops field of a
uart_port structure.  This field is declared const, so uart_ops structures
that have this property can be declared as const also.

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@r disable optional_qualifier@
identifier i;
position p;
@@
static struct uart_ops i@p = { ... };

@ok@
identifier r.i;
struct uart_port e;
position p;
@@
e.ops = &i@p;

@bad@
position p != {r.p,ok.p};
identifier r.i;
struct uart_ops e;
@@
e@i@p

@depends on !bad disable optional_qualifier@
identifier r.i;
@@
static
+const
 struct uart_ops i = { ... };
// </smpl>
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently the clocks are enabled at probe and disabled
at remove. Instead enable the clocks when used.
Signed-off-by: NShubhrajyoti Datta <shubhraj@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Support early console setup via DT for all listed compatible strings.
Remove EARLYCON_DECLARE which was done by:
"Use common framework for earlycon declarations"
(sha1: 2eaa7909)
when OF_EARLYCON_DECLARE is defined.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

start_tx and the ISR used largely identical code to transmit data.
Consolidate that in one place.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Remove the _OFFSET suffix from all register defines which makes code a
little easier to read and avoids a few line breaks.
Suggested-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Handling magic sysrq included dropping a lock to avoid a deadlock that
happened when cdns_uart_console_write tried to acquire a lock in the
from the sysrq code path. By making the acquisition of the lock in
cdns_uart_console_write depending on port->sysrq, cdns_uart_handle_rx can be
simplified to simply call uart_handle_sysrq.
Suggested-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Convert an if-else into the more common early return on error, reducing
the indent level of the happy path.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The system could deadlock handling RX IRQs when RX-related IRQ
conditions became true while the receiver was disabled. To avoid this,
enable/disable the RX/TX IRQs together with the receiver/transmitter.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Request_irq() should be _after_ h/w programming, otherwise an
interrupt could be triggered and in-progress before the h/w has been
setup.
Reported-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Move RX-related IRQ handling into a helper function.
Fixes a problem where every char received after a parity or frame error
in the current isr will also be tagged as a parity or frame error.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Shutting down the UART port can happen while console operations are in
progress. Holding the port lock serializes these operations and avoids
the UART HW to be disabled in the middle of console prints.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The RX path in the interrupt handler released a lock unnecessarily.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The startup function is supposed to initialize the UART for receiving.
Hence, don't enable the TX part. Also, protect HW accesses with the port
lock.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When shutting down the UART, clear the interrupt status register. Bits
in the ISR are cleared by writing them as '1'.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Reviewed-by: NMoritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Restarting the transmitter even if the circ buffer is empty may be
necessary to push out remaining data when the port is restarted after
being stopped.

Cc: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Instead of disabling the IRQ, use the spin lock to serialize accesses to
the HW. This protects the driver from interference of non-IRQ callbacks
with each other and makes the driver more consistent in its
serialization method.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Non-functional, formatting changes to ease reading the code.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Reviewed-by: NMoritz Fischer <moritz.fischer@ettus.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Early console functions are only used during the early boot stage.
This change just saves a small amount of memory footprint.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

For systems with CONFIG_SERIAL_OF_PLATFORM=y and device_type =
"serial"; property in DT of_serial.c driver maps and unmaps IRQ (because
driver probe fails). Then a driver is called but irq mapping is not
created that's why driver is failing again in again on request_irq().
Based on this use platform_get_irq() instead of platform_get_resource()
which is doing irq_desc allocation and driver itself can request IRQ.

Fix both xilinx serial drivers in the tree.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
CC: <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

cdns_uart_console_setup() checks port->mapbase != 0, but the port may
not be initialized yet even if this condition is met [e.g., ioremap()
may have failed]. Check port->membase != NULL instead, similar to
cdns_early_console_setup().
Signed-off-by: NThomas Betker <thomas.betker@rohde-schwarz.com>
Reviewed-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Get rid of cdns_uart_readl() and cdns_uart_writel() and just call
readl() and writel() directly.

Most of the patch was created by search-and-replace (I had to convert a
few lines manually, and break some lines longer than 80 columns):
* s/cdns_uart_readl(/readl(port->membase + /g
* s/cdns_uart_writel(\([^,]*\),/writel(\1, port->membase +/g
Signed-off-by: NThomas Betker <thomas.betker@rohde-schwarz.com>
Acked-by: NSören Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

cdns_uart_get_port() sets port->iobase = 1 to "mark port in use", but
this "in use" condition is not checked anywhere else in the code. So
remove the line, keeping port->iobase = 0 (which also makes more sense).
Signed-off-by: NThomas Betker <thomas.betker@rohde-schwarz.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The code assumes that the array cdns_uart_port[] has dimension
CDNS_UART_NR_PORTS, so let us define it this way.
Signed-off-by: NThomas Betker <thomas.betker@rohde-schwarz.com>
Reviewed-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The register space size is 0x1000, and this value [not 0xfff] should be
provided to request_mem_region(), ioremap(), etc.
Signed-off-by: NThomas Betker <thomas.betker@rohde-schwarz.com>
Reviewed-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

of_device_id is always used as const.
(See driver.of_match_table and open firmware functions)
Signed-off-by: NFabian Frederick <fabf@skynet.be>
Acked-by: NPeter Korsgaard <peter@korsgaard.com>
Acked-by: NTimur Tabi <timur@tabi.org>
Acked-by: NPatrice Chotard <patrice.chotard@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

All instances of "ttyPS" use this macro except for this one. Convert
it.
Signed-off-by: NPeter Crosthwaite <peter.crosthwaite@xilinx.com>
Acked-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The implementation of flushing the RX FIFO breaks in a number of cases,
it is impossible to ensure an complete flush of the RX FIFO due to the
hardware not allowing the use of the FIFOs when the receiver is disabled
(Reading from the FIFO register does not remove it from the FIFO when
the RX_EN=0 or RX_DIS=1). Additionally during an initial set_termios
call where RX_DIS=1 causes a hang waiting forever for the RX FIFO to
empty. On top of this the FIFO will be cleared by the use of the RXRST
bits on the Control Register, making the RX flush pointless (as it does
not preserve the data read anyway).

Due to the TXRST the TX FIFO and transmitter can be interrupted during
frame trasmission, causing corruption and additionally data lost in the
FIFO. Most other serial drivers do not flush or clear the FIFOs during
a termios configuration change and as such do not have issues with
corruption. For this UART controller is it required that the TXRST/RXRST
bit be flagged during the change, this means that the data in the FIFO
will be dropped when changing configuration. In order to prevent data
loss and corruption of the transmitted data, wait until the TX FIFO is
empty before changing the configuration. The performance of this may
cause the set_termios call to take a longer amount of time especially
on lower baud rates, however it is comparable to the same performance
hit that a console_write call costs.
Signed-off-by: NNathan Rossi <nathan.rossi@xilinx.com>
Acked-by: NAnirudha Sarangi <anirudh@xilinx.com>
Acked-by: NHarini Katakam <harinik@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Add support for setting the state of the DTR and RTS signals.
Acked-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Add earlycon support for the cadence serial port.
This is based on recent patches:
"tty/serial: pl011: add generic earlycon support"
(sha1: 0d3c673e)
"tty/serial: add arm/arm64 semihosting earlycon"
(sha1: d50d7269)
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The tx_empty() callback currently checks the TXEMPTY bit in the interrupt
status register to decided whether the FIFO should be reported as empty or
not. The bit in this register gets set when the FIFO state transitions from
non-empty to empty but is cleared again in the interrupt handler. This means
it is not suitable to be used to decided whether the FIFO is currently empty
or not. Instead use the TXEMPTY bit from the status register which will be
set as long as the FIFO is empty.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

There is no need to init .owner field.

Based on the patch from Peter Griffin <peter.griffin@linaro.org>
"mmc: remove .owner field for drivers using module_platform_driver"

This patch removes the superflous .owner field for drivers which
use the module_platform_driver API, as this is overriden in
platform_driver_register anyway."
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NAlexander Shiyan <shc_work@mail.ru>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Zynq's UART is Cadence IP. Make this visible in the prompt in kconfig
and additional comments in the driver.
This also renames functions and symbols, as far as possible without
breaking user space API, to reflect the Cadence origin. This is achieved
through simple search and replace:
 - s/XUARTPS/CDNS_UART/g
 - s/xuartps/cdns_uart/g
The only exceptions are PORT_XUARTPS and the driver name, which stay as is,
due to their exposure to user space. As well as the - no legacy -
compatibility string 'xlnx,xuartps'
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Tested-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

A comment states, that, according to the data sheet, to enable
interrupts the disable register should be written, but the enable
register could be left untouched. And it suspsects a HW bug requiring
to write both.
Reviewing the data sheet, these statements seem wrong. Just as one would
expect. Writing to the enable/disable register enables/disables
interrupts.
Hence the misleading comment and needless write to the disable register
are removed from the enable sequence.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Tested-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

A lot of read-modify-write sequences used a one-line statement which
nests a readl() within a writel(). Convert this into code sequences that
make the three steps more obvious.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Tested-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Print a warning if the clock notifier rejects a clock frequency change
to facilitate debugging (see:
http://thread.gmane.org/gmane.linux.ports.arm.kernel/304329/focus=304379)
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Tested-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This is all white space and comment clean up. Mostly reformatting
comments.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Tested-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Register port numbers according to order in DT aliases.
If aliases are not defined, order in DT is used.
If aliases are defined, register port id based
on that.
This patch ensures proper ttyPS0/1 assignment.

[soren]: Combined integer declarations in probe(), removed warning message
if no alias is found.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

No functional changes.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>