Commit e490c9144c ("tty: Add software emulated RS485 support for 8250")
introduced support to use RTS as an rs485 Transmit Enable signal if data
is transmitted through the tty layer.
Console messages bypass the tty layer and instead are emitted via
serial8250_console_write(). Amend that function to drive RTS as well,
allowing for a console on rs485 ports.
Note that serial8250_console_write() may be called concurrently to the
tty layer accessing the port. The two protect their accesses with the
port lock, but serial8250_console_write() may find RTS still being
asserted by the tty layer, in which case it shouldn't be deasserted
after the console message has been printed. Recognize such situations
by checking the em485->tx_stopped flag.
If a delay_rts_before_send or delay_rts_after_send has been specified,
serial8250_console_write() busy-waits for its duration. Optimizations
for those wait times are conceivable: E.g. if RTS is already asserted,
we could check whether em485->start_tx_timer is active and wait only
for the remaining expire time. But this would require calling into
the hrtimer infrastructure, which involves acquiring locks and
potentially reprogramming timer hardware. Such operations seem too
risky in the context of console printout, which needs to work even when
the kernel has crashed and emits a BUG splat. So I've gone with a
simplistic solution which just always waits for the full delay.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Link: https://lore.kernel.org/r/65edffce4670a19e598015c03cbe46f1ffd93e43.1582895077.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Amend 8250_bcm2835aux.c to support rs485 as introduced for 8250_omap.c
by commit e490c9144c ("tty: Add software emulated RS485 support for
8250").
The bcm2835aux differs from omap chips by inverting the meaning of RTS
in the MCR register: If the bit is clear, RTS is high. With omap, it's
apparently the other way round.
Moreover, omap achieves half-duplex mode by disabling the UART_IER_RDI
interrupt and clearing the RX FIFO when TX stops. This approach doesn't
work on bcm2835aux because the UART_LSR_DR bit is set even when
UART_IER_RDI is disabled. Consequently, serial8250_handle_irq() invokes
serial8250_rx_chars() to empty the FIFO and characters are received even
though the user requested half-duplex. Solve by disabling the receiver
using the non-standard CNTL register.
Cache that register in the driver's private data for performance. Set
the private data pointer before calling serial8250_register_8250_port()
to prevent a null pointer deref in case one of the rs485 callbacks is
invoked immediately after port registration.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Link: https://lore.kernel.org/r/dd86460e20a8f979b7272a0bde73640312b902b1.1582895077.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit e490c9144c ("tty: Add software emulated RS485 support for 8250")
introduced support to use RTS as an rs485 Transmit Enable signal.
So far the only drivers taking advantage of it are 8250_omap.c and
8250_of.c.
We're about to make use of the feature in 8250_bcm2835aux.c as well.
The bcm2835aux differs from omap chips by inverting the meaning of RTS
in the MCR register. Moreover, omap achieves half-duplex mode by
disabling the RX interrupt and clearing the RX FIFO when TX stops.
The bcm2835aux requires disabling the receiver instead.
Support these behavioral differences by generalizing the rs485 emulation:
Introduce ->rs485_start_tx() and ->rs485_stop_tx() callbacks in struct
uart_8250_port, provide generic implementations containing the existing
code and use them as callbacks in 8250_omap.c and 8250_of.c.
start_tx_rs485() is idempotent in that it recognizes whether RTS is
already asserted. Achieve the same by introducing a tx_stopped flag in
struct uart_8250_em485. This may even perform a little better on arches
where memory access is faster than mmio access.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Link: https://lore.kernel.org/r/5ac0464ae4414708e723a1e0d52b0c1b2bd41b9b.1582895077.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When rs485 transmission over an 8250 port stops, __stop_tx() assigns
active_timer = NULL before calling __stop_tx_rs485().
That function in turn either assigns active_timer = stop_tx_timer and
rearms the timer (in case a delay_rts_after_send needs to be observed)
or directly calls __do_stop_tx_rs485().
Move the assignment active_timer = NULL to __stop_tx_rs485() into the
branch which directly calls __do_stop_tx_rs485(), thereby avoiding a
duplicate assignment and simplifying the code.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Link: https://lore.kernel.org/r/bca638405550eaf92f0c6060b553b687f35885e0.1582895077.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit e490c9144c ("tty: Add software emulated RS485 support for 8250")
introduced support to use RTS as an rs485 Transmit Enable signal.
Drivers opt in to the feature by calling serial8250_em485_init() from
their ->rs485_config() callback.
So far there are two drivers doing that, 8250_omap.c and 8250_of.c.
Both use an identical callback. We're about to add a third user of that
callback, therefore deduplicate it and move it to 8250_port.c.
Drivers now opt in to rs485 software emulation by assigning the generic
serial8250_rs485_config() callback introduced herein to their
.rs485_config struct member. This change allows unexporting
serial8250_em485_init() and declaring it static.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Cc: Heiko Schocher <hs@denx.de>
Link: https://lore.kernel.org/r/fcef63642dc4eae41ae7842d23747b2bf5d40285.1582895077.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
serial8250_do_set_mctrl() currently allows modifying the RTS modem
control line even when RTS is used as an rs485 Transmit Enable signal.
It is thus possible for user space to interfere with rs485 communication
by invoking a TIOCMSET ioctl().
Ignore such change requests and retain the current RTS polarity when in
rs485 mode. Note that serial8250_set_mctrl() is always called with
port->lock held, so there's no risk that RTS is changed concurrently.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matwey V. Kornilov <matwey@sai.msu.ru>
Link: https://lore.kernel.org/r/b1ce34ca9bc4d7bdc6e9852fcf30b1f4e37c8a80.1582895077.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If a driver exposes early consoles with EARLYCON_DECLARE() and
OF_EARLYCON_DECLARE(), pefer the non-OF variant if the user specifies it
by
earlycon=<driver>,<options>
The rationale behind this is that some drivers register multiple setup
functions under the same driver name. Eg.
OF_EARLYCON_DECLARE(lpuart, "fsl,vf610-lpuart", lpuart_early_console_setup);
OF_EARLYCON_DECLARE(lpuart32, "fsl,ls1021a-lpuart", lpuart32_early_console_setup);
OF_EARLYCON_DECLARE(lpuart32, "fsl,imx7ulp-lpuart", lpuart32_imx_early_console_setup);
EARLYCON_DECLARE(lpuart, lpuart_early_console_setup);
EARLYCON_DECLARE(lpuart32, lpuart32_early_console_setup);
It depends on the order of the entries which console_setup() actually
gets called. To make things worse, I guess it also depends on the
compiler how these are ordered. Thus always prefer the EARLYCON_DECLARE()
ones.
Signed-off-by: Michael Walle <michael@walle.cc>
Link: https://lore.kernel.org/r/20200220174607.24285-1-michael@walle.cc
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since commit 3bc3206e1c ("serial: fsl_lpuart: Remove the alias node
dependence") the port line number can also be allocated by IDA, but in
case of an error the ID will no be removed again. More importantly, any
ID will be freed in remove(), even if it wasn't allocated but instead
fetched by of_alias_get_id(). If it was not allocated by IDA there will
be a warning:
WARN(1, "ida_free called for id=%d which is not allocated.\n", id);
Move the ID allocation more to the end of the probe() so that we still
can use plain return in the first error cases.
Fixes: 3bc3206e1c ("serial: fsl_lpuart: Remove the alias node dependence")
Signed-off-by: Michael Walle <michael@walle.cc>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200303174306.6015-3-michael@walle.cc
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This reverts commit a659652f61.
This broke the earlycon on LS1021A processors because the order of the
earlycon_setup() functions were changed. Before the commit the normal
lpuart32_early_console_setup() was called. After the commit the
lpuart32_imx_early_console_setup() is called instead.
Fixes: a659652f61 ("tty: serial: fsl_lpuart: drop EARLYCON_DECLARE")
Signed-off-by: Michael Walle <michael@walle.cc>
Link: https://lore.kernel.org/r/20200303174306.6015-2-michael@walle.cc
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The peripheral has support for inverting its input and/or output
signals. This is useful if the hardware flips polarity of the
peripheral's signal, such as swapped +/- pins on an RS-422 transceiver,
or an inverting level shifter. Add support for these control registers
via the device tree binding.
As part of this change, make the writes of the various registers more
uniform by moving the UCR3 block up near the other registers' blocks,
since the INVT bit must be set before enabling the peripheral.
Signed-off-by: George Hilliard <ghilliard@kopismobile.com>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20200226222319.18383-3-ghilliard@kopismobile.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If we're unlucky enough that this drivers binds to a mrvl,mmp-uart device
on a MMP3, the port type gets detected as 16550A instead of XScale, and it
won't work. Other drivers that may bind to the same hardware are 8250_of
and, god forbid, serial_pxa.
Force the port type, we know it's a PORT_XSCALE.
Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Link: https://lore.kernel.org/r/20200219080130.4334-1-lkundrak@v3.sk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Christophe reports that powerpc 8xx silently fails to 5.6-rc1. It turns
out I was wrong about nobody relying on the lazy initialization of the
cpm/qe muram in commit b6231ea2b3 (soc: fsl: qe: drop broken lazy
call of cpm_muram_init()).
Rather than reinstating the somewhat dubious lazy call (initializing a
currently held spinlock, and implicitly doing a GFP_KERNEL under that
spinlock), make sure that cpm_muram_init() is called early enough - I
thought the calls from the subsys_initcalls were good enough, but when
used by console drivers, that's obviously not the
case. cpm_muram_init() is safe to call twice (there's an early return
if it is already initialized), so keep the call from cpm_init() - in
case SERIAL_CPM_CONSOLE=n.
Fixes: b6231ea2b3 (soc: fsl: qe: drop broken lazy call of cpm_muram_init())
Reported-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Tested-by: Christophe Leroy <christophe.leroy@c-s.fr>
Link: https://lore.kernel.org/r/20200213114342.21712-1-linux@rasmusvillemoes.dk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Link: https://lore.kernel.org/r/20200213004426.GA7886@embeddedor.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Link: https://lore.kernel.org/r/20200213004611.GA8748@embeddedor.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
First, rename PM_QOS_CPU_DMA_LAT_DEFAULT_VALUE to
PM_QOS_CPU_LATENCY_DEFAULT_VALUE and update all of the code
referring to it accordingly.
Next, rename cpu_dma_constraints to cpu_latency_constraints, move
the definition of it closer to the functions referring to it and
update all of them accordingly. [While at it, add a comment to mark
the start of the code related to the CPU latency QoS.]
Finally, rename the pm_qos_power_*() family of functions and
pm_qos_power_fops to cpu_latency_qos_*() and cpu_latency_qos_fops,
respectively, and update the definition of cpu_latency_qos_miscdev.
[While at it, update the miscdev interface code start comment.]
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Amit Kucheria <amit.kucheria@linaro.org>
Tested-by: Amit Kucheria <amit.kucheria@linaro.org>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertenly introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Link: https://lore.kernel.org/r/20200212193700.GA29715@embeddedor
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The commit 54e53b2e80
("tty: serial: 8250: pass IRQ shared flag to UART ports")
nicely explained the problem:
---8<---8<---
On some systems IRQ lines between multiple UARTs might be shared. If so, the
irqflags have to be configured accordingly. The reason is: The 8250 port startup
code performs IRQ tests *before* the IRQ handler for that particular port is
registered. This is performed in serial8250_do_startup(). This function checks
whether IRQF_SHARED is configured and only then disables the IRQ line while
testing.
This test is performed upon each open() of the UART device. Imagine two UARTs
share the same IRQ line: On is already opened and the IRQ is active. When the
second UART is opened, the IRQ line has to be disabled while performing IRQ
tests. Otherwise an IRQ might handler might be invoked, but the IRQ itself
cannot be handled, because the corresponding handler isn't registered,
yet. That's because the 8250 code uses a chain-handler and invokes the
corresponding port's IRQ handling routines himself.
Unfortunately this IRQF_SHARED flag isn't configured for UARTs probed via device
tree even if the IRQs are shared. This way, the actual and shared IRQ line isn't
disabled while performing tests and the kernel correctly detects a spurious
IRQ. So, adding this flag to the DT probe solves the issue.
Note: The UPF_SHARE_IRQ flag is configured unconditionally. Therefore, the
IRQF_SHARED flag can be set unconditionally as well.
Example stack trace by performing `echo 1 > /dev/ttyS2` on a non-patched system:
|irq 85: nobody cared (try booting with the "irqpoll" option)
| [...]
|handlers:
|[<ffff0000080fc628>] irq_default_primary_handler threaded [<ffff00000855fbb8>] serial8250_interrupt
|Disabling IRQ #85
---8<---8<---
But unfortunately didn't fix the root cause. Let's try again here by moving
IRQ flag assignment from serial_link_irq_chain() to serial8250_do_startup().
This should fix the similar issue reported for 8250_pnp case.
Since this change we don't need to have custom solutions in 8250_aspeed_vuart
and 8250_of drivers, thus, drop them.
Fixes: 1c2f04937b ("serial: 8250: add IRQ trigger support")
Reported-by: Li RongQing <lirongqing@baidu.com>
Cc: Kurt Kanzenbach <kurt@linutronix.de>
Cc: Vikram Pandita <vikram.pandita@ti.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: stable <stable@vger.kernel.org>
Acked-by: Kurt Kanzenbach <kurt@linutronix.de>
Link: https://lore.kernel.org/r/20200211135559.85960-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It is possible to get an instant RX timeout or end-of-transfer interrupt
before RX DMA was started, if transaction is less than 16 bytes. Transfer
should be handled in PIO mode in this case because DMA can't handle it.
This patch brings back the original behaviour of the driver that was
changed by accident by a previous commit, it fixes occasional Bluetooth HW
initialization failures which I started to notice recently.
Fixes: d5e3fadb70 ("tty: serial: tegra: Activate RX DMA transfer by request")
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20200209164415.9632-1-digetx@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On AR934x this UART is usually not initialized by the bootloader
as it is only used as a secondary serial port while the primary
UART is a newly introduced NS16550-compatible.
In order to make use of the ar933x-uart on AR934x without RTS/CTS
hardware flow control, one needs to set the
UART_CS_{RX,TX}_READY_ORIDE bits as other than on AR933x where this
UART is used as primary/console, the bootloader on AR934x typically
doesn't set those bits.
Setting them explicitely on AR933x should not do any harm, so just
set them unconditionally.
Tested-by: Chuanhong Guo <gch981213@gmail.com>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Link: https://lore.kernel.org/r/20200207095335.GA179836@makrotopia.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
