At the default TX trigger level of 2 in non-DMA mode (meaning that an
interrupt is generated when less than 2 characters are left in the
FIFO), we have observed frequent buffer underruns at 115200 Baud on an
i.MX8M Nano. This can cause communication issues if the receiving side
expects a continuous transfer.
Increasing the level to 8 makes the UART trigger an interrupt earlier,
giving the kernel enough time to refill the FIFO, at the cost of
triggering one interrupt per ~24 instead of ~30 bytes of transmitted
data (as the i.MX UART has a 32 byte FIFO).
Signed-off-by: Michael Krummsdorf <michael.krummsdorf@tq-group.com>
Signed-off-by: Matthias Schiffer <matthias.schiffer@ew.tq-group.com>
Link: https://lore.kernel.org/r/20240508133744.35858-1-matthias.schiffer@ew.tq-group.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8250_pnp sets drvdata to line + 1 if the probe is successful. The users
of drvdata are in remove, suspend and resume callbacks, none of which
will be called if probe failed. The line acquired from drvdata can
never be zero in those functions and the checks for that can be
removed.
Eliminate also +/-1 step because all users of line subtract 1 from the
value.
These might have been leftover from legacy PM callbacks that could
be called without probe being successful.
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20240506121202.11253-1-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The hrtimer for RXDMA timeout was unconditionally restarted in the RXDMA
complete handler ignoring the fact that setting up DMA may fail and PIO
is used instead. Explicitly stop the timer when DMA is completed and
only restart it when setting up DMA was successful. This makes the
intention of the timer much clearer, the driver easier to understand and
simplifies assumptions about the timer. The latter avoids race
conditions if these assumptions were not met or confused.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Link: https://lore.kernel.org/r/20240506114016.30498-9-wsa+renesas@sang-engineering.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When using a high speed clock with a low baud rate, the 4x prescaler is
automatically selected if required. In that case, sc16is7xx_set_baud()
properly configures the chip registers, but returns an incorrect baud
rate by not taking into account the prescaler value. This incorrect baud
rate is then fed to uart_update_timeout().
For example, with an input clock of 80MHz, and a selected baud rate of 50,
sc16is7xx_set_baud() will return 200 instead of 50.
Fix this by first changing the prescaler variable to hold the selected
prescaler value instead of the MCR bitfield. Then properly take into
account the selected prescaler value in the return value computation.
Also add better documentation about the divisor value computation.
Fixes: dfeae619d7 ("serial: sc16is7xx")
Cc: stable@vger.kernel.org
Signed-off-by: Hugo Villeneuve <hvilleneuve@dimonoff.com>
Reviewed-by: Jiri Slaby <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20240430200431.4102923-1-hugo@hugovil.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a scenario when resuming from some power saving states
with no_console_suspend where console output can be generated
before the 8250_bcm7271 driver gets the opportunity to restore
the baud_mux_clk frequency. Since the baud_mux_clk is at its
default frequency at this time the output can be garbled until
the driver gets the opportunity to resume.
Since this is only an issue with console use of the serial port
during that window and the console isn't likely to use baud
rates that require alternate baud_mux_clk frequencies, allow the
driver to select the default_mux_rate if it is accurate enough.
Fixes: 41a469482d ("serial: 8250: Add new 8250-core based Broadcom STB driver")
Cc: stable@vger.kernel.org
Signed-off-by: Doug Berger <opendmb@gmail.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Tested-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20240424222559.1844045-1-opendmb@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current implementation uses either gsm0_receive() or gsm1_receive()
depending on whether the user configured the mux in basic or advanced
option mode. Both functions share some state values over the same logical
elements of the frame. However, both frame types differ in their nature.
gsm0_receive() uses non-transparency framing, whereas gsm1_receive() uses
transparency mechanism. Switching between both modes leaves the receive
function in an undefined state when done during frame reception.
Fix this by splitting both states. Add gsm0_receive_state_check_and_fix()
and gsm1_receive_state_check_and_fix() to ensure that gsm->state is reset
after a change of gsm->receive.
Note that gsm->state is only accessed in:
- gsm0_receive()
- gsm1_receive()
- gsm_error()
Fixes: e1eaea46bb ("tty: n_gsm line discipline")
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Starke <daniel.starke@siemens.com>
Link: https://lore.kernel.org/r/20240424054842.7741-2-daniel.starke@siemens.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Assuming the following:
- side A configures the n_gsm in basic option mode
- side B sends the header of a basic option mode frame with data length 1
- side A switches to advanced option mode
- side B sends 2 data bytes which exceeds gsm->len
Reason: gsm->len is not used in advanced option mode.
- side A switches to basic option mode
- side B keeps sending until gsm0_receive() writes past gsm->buf
Reason: Neither gsm->state nor gsm->len have been reset after
reconfiguration.
Fix this by changing gsm->count to gsm->len comparison from equal to less
than. Also add upper limit checks against the constant MAX_MRU in
gsm0_receive() and gsm1_receive() to harden against memory corruption of
gsm->len and gsm->mru.
All other checks remain as we still need to limit the data according to the
user configuration and actual payload size.
Reported-by: j51569436@gmail.com
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218708
Tested-by: j51569436@gmail.com
Fixes: e1eaea46bb ("tty: n_gsm line discipline")
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Starke <daniel.starke@siemens.com>
Link: https://lore.kernel.org/r/20240424054842.7741-1-daniel.starke@siemens.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case the UART port is used as a console, no_console_suspend is
available in bootargs and UART port is part of a software-controlled power
domain we need to call device_set_awake_path(). This lets the power
domain core code know that this domain should not be powered off
during system suspend. Otherwise, the UART port power domain is turned off,
nothing is printed while suspending and the suspend/resume process is
blocked. This was detected on the Renesas RZ/G3S SoC while adding support
for power domains.
Based on code investigation (on v6.9-rc5), this issue is present on other
SoCs (e.g., Renesas R-Mobile A1 [1], IMX8QXP [2]) and different SoCs have
particular implementation to handle it. Due to this the patch added the
call of device_set_awake_path() in uart_suspend_port() instead of having
it in the platform specific UART driver.
[1] drivers/pmdomain/renesas/rmobile-sysc.c:116
[2] drivers/pmdomain/imx/scu-pd.c:357
Suggested-by: Ulf Hansson <ulf.hansson@linaro.org>
Suggested-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Tony Lindgren <tony.lindgren@linux.intel.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Link: https://lore.kernel.org/r/20240430095930.2806067-1-claudiu.beznea.uj@bp.renesas.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is one of the hackiest Alpha machines, and the only one without
PCI support. Removing this allows cleaning up code in eise and tty
drivers in addition to the architecture code.
Acked-by: Paul E. McKenney <paulmck@kernel.org>
Acked-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
We want the tty fixes in here as well, and it resolves a merge conflict
in:
drivers/tty/serial/serial_core.c
as well.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is a large patch but is only additions. All changes and removals
are made in previous patches in this series.
- Add CTI board_init and port setup functions for each UART type
- Add CTI_EXAR_DEVICE() and CTI_PCI_DEVICE() macros
- Add support for reading a word from the Exar EEPROM.
- Add support for configuring and setting a single MPIO
- Add various helper functions for CTI boards.
- Add osc_freq to struct exar8250
Signed-off-by: Parker Newman <pnewman@connecttech.com>
Link: https://lore.kernel.org/r/ae4a66e7342b686cb8d4b15317585dfb37222cf4.1713382717.git.pnewman@connecttech.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
