serial: mvebu-uart: correctly calculate minimal possible baudrate

For default (x16) scheme which is currently used by mvebu-uart.c driver,
maximal divisor of UART base clock is 1023*16. Therefore there is limit for
minimal supported baudrate. This change calculate it correctly and prevents
setting invalid divisor 0 into hardware registers.

Signed-off-by: Pali Rohár <pali@kernel.org>
Fixes: 68a0db1d7da2 ("serial: mvebu-uart: add function to change baudrate")
Link: https://lore.kernel.org/r/20210624224909.6350-4-pali@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: mvebu-uart: do not allow changing baudrate when uartclk is not available

Testing mvuart->clk for non-error is not enough as mvuart->clk may contain
valid clk pointer but when clk_prepare_enable(mvuart->clk) failed then
port->uartclk is zero.

When mvuart->clk is not available then port->uartclk is zero too.

Parent clock rate port->uartclk is needed to calculate UART clock divisor
and without it is not possible to change baudrate.

So fix test condition when it is possible to change baudrate.

Signed-off-by: Pali Rohár <pali@kernel.org>
Fixes: 68a0db1d7da2 ("serial: mvebu-uart: add function to change baudrate")
Link: https://lore.kernel.org/r/20210624224909.6350-3-pali@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: mvebu-uart: fix calculation of clock divisor

The clock divisor should be rounded to the closest value.

Signed-off-by: Pali Rohár <pali@kernel.org>
Fixes: 68a0db1d7da2 ("serial: mvebu-uart: add function to change baudrate")
Cc: stable@vger.kernel.org # 0e4cf69ede87 ("serial: mvebu-uart: clarify the baud rate derivation")
Link: https://lore.kernel.org/r/20210624224909.6350-2-pali@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tty: make linux/tty_flip.h self-contained

If someone includes linux/tty_flip.h before linux/tty.h, they see
many compiler errors like:
 include/linux/tty_flip.h:23:30: error: invalid use of undefined type 'struct tty_port'
 include/linux/tty_flip.h:26:14: error: invalid use of undefined type 'struct tty_buffer'

tty_flip.h actually lexicographically sorts before tty.h. So if people
sort includes (as I tried in amiserial), the compilation suddenly
breaks.

Solve this by including linux/tty.h from linux/tty_flip.h, so that
everything is defined as needed.

Another alternative would be to uninline tty_insert_flip_char and just
insert forward declarations of tty_port and tty_buffer structs into
tty_flip.h as that inline is the only real user. But that would mean
slowing down the fast path without any good reason. (Provided the fix
is that easy and there were no real problems with this until now.)

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210625073511.4514-1-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: Prefer unsigned int to bare use of unsigned

Fix checkpatch warnings:
WARNING: Prefer 'unsigned int' to bare use of 'unsigned'

Signed-off-by: Jinchao Wang <wjc@cdjrlc.com>
Link: https://lore.kernel.org/r/20210624021207.58059-1-wjc@cdjrlc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: 8250: 8250_omap: Fix possible interrupt storm on K3 SoCs

On K3 family of SoCs (which includes AM654 SoC), it is observed that RX
TIMEOUT is signalled after RX FIFO has been drained, in which case a
dummy read of RX FIFO is required to clear RX TIMEOUT condition.
Otherwise, this would lead to an interrupt storm.

Fix this by introducing UART_RX_TIMEOUT_QUIRK flag and doing a dummy
read in IRQ handler when RX TIMEOUT is reported with no data in RX FIFO.

Fixes: be70874498f3 ("serial: 8250_omap: Add support for AM654 UART controller")
Reported-by: Jan Kiszka <jan.kiszka@siemens.com>
Tested-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Link: https://lore.kernel.org/r/20210622145704.11168-1-vigneshr@ti.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: qcom_geni_serial: use DT aliases according to DT bindings

Device tree bindings do not specify "hsuart" aliases, instead all serial
ports should use "serial" alias name as noted by Rob Herring [1].
Make qcom_geni_serial driver use "serial" alias and fallback to "hsuart"
if one is not found.

[1] https://lore.kernel.org/linux-arm-msm/6dd1f5cd-03c7-5945-9fa2-1c2698405110@linaro.org/

Cc: Rob Herring <robh+dt@kernel.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20210621211528.1607516-1-dmitry.baryshkov@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Revert "tty: serial: Add UART driver for Cortina-Access platform"

This reverts commit b61c8bf4694b5115766849378dcb8787ff54e65e.  It never
made it to a public mailing list and still needs some work based on the
review comments.  So revert it for now.

Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Link: https://lore.kernel.org/r/CAMuHMdXA9-ajoAza2JAW5879ECieMm1dbBbKHgJhDa7=3kWu3w@mail.gmail.com
Cc: Jason Li <jason.li@cortina-access.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tty: serial: Add UART driver for Cortina-Access platform

This driver supports Cortina Access UART IP integrated
in most all CAXXXX line of SoCs. Earlycom is also supported

Signed-off-by: Jason Li <jason.li@cortina-access.com>
Link: https://lore.kernel.org/r/20210615165750.31261-1-alex.nemirovsky@cortina-access.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

MAINTAINERS: add me back as mxser maintainer

I was given a few cards from MOXA (thanks a lot). Provided I can now
test changes, I cleaned up the driver a bit and can continue maintaining
it.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-71-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: Documentation, fix typos

Fix some typos in the mxser's documentation.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-70-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: Documentation, make the docs up-to-date

Remove all the obsolete information from the documentation of mxser
driver. Either it was about the out-of-tree driver, or it was superseded
by udev et al.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-69-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: Documentation, remove traces of callout device

Callout devices are ancient history, so remove its traces from mxser's
documentation.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-68-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: introduce mxser_16550A_or_MUST helper

This check is used in the code on many places. Either in positive or
negative form. So introduce a helper called mxser_16550A_or_MUST for
this purpose. And replace all the occurrences.

This simplifies many ifs in there.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-67-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: rename flags to old_speed in mxser_set_serial_info

Local 'flags' are used to store speed in mxser_set_serial_info, not
actual flags (by doing & ASYNC_SPD_MASK). So rename 'flags' to
'old_speed' (after what it is).

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-66-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: use port variable in mxser_set_serial_info

mxser_set_serial_info already defined a local 'port' of type struct
tty_port. So use it in the code everywhere.

This eliminates some text from the code there.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-65-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: access info->MCR under info->slock

info->MCR is protected by info->slock all over the code. Extend the lock
in mxser_tiocmget around the info->MCR fetch too.

It likely doesn't change anything, but it's always good to be consistent.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-64-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: fix typos around enhanced mode

Fix spelling of "enhanced" in macros and comments around them.

While "enchance" comes from the original Moxa's driver, I don't think it
was meant to be that. From the context, they obviously mean "enhanced".

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-63-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: decrypt FCR values

Currently, some of the values written to FCR are magic constants. But
they are composed of well-defined bits. Use these named macros in place
of the constants.

No changes in objdump -d.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-62-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove xmit_cnt < 0 tests

mxser_port::xmit_cnt is unsigned, it cannot be negative. Therefore
remove all the xmit_cnt < 0 tests.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-61-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: introduce and use start/stop_tx helpers

Starting and stopping TX is performed on many places in the driver. It
is repeated outb and info->IER modification.  Create few helpers
(__mxser_start_tx, mxser_start_tx, __mxser_stop_tx) and use them in the
code instead of open coding.

This should be a base for an uart driver for this device in the future.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-60-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: extract and dedup CTS handling

CTS is handled on two locations in mxser. Extract the code into
separation function mxser_handle_cts. And call it from the two
locations.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-59-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: don't start TX from tty_operations::put_char

TX is not supposed to be started from tty_operations::put_char. Line
disciplines ought to call tty_operations::put_char several times and
then do a single tty_operations::flush_chars to start TX.

Fix this in mxser by removing TX startup from mxser_put_char.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-58-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: cleanup mxser_dtr_rts

mxser_dtr_rts now does all inb, outb, & or | in a single statement.  The
code is hard to follow, so make it more readable by doing one-line =
one-statement.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-57-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: cleanup mxser_rs_break

mxser_rs_break now does all inb, outb, & or | in a single statement. The
code is hard to follow, so make it more readable by doing one-line =
one-statement.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-56-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove pointless ioaddr checks

struct mxser_port::ioaddr can never be NULL once set in ->probe. Hence
the checks for it are pointless. Remove them.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-55-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: alloc only needed # of ports

Many boards have only 2 ports. But we always allocate 8 ports (struct
mxser_port) in each struct mxser_board.

Switch this to an empty/variable sized array ([]) and allocate the
needed number of ports using struct_size helper in ->probe.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-54-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: alloc struct mxser_board dynamically

There is no need to preallocate an array of four struct mxser_board's.
In most cases a single board or two will be present in a machine. So
allocate struct mxser_board as needed in ->probe.

This makes mxser_boards a bit array. There we store which indexes are
free (unallocated).

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-53-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: make mxser_board::idx really an index

mxser_board::idx is named as index, but actually contain a base. So
assign index there. And for base, introduce a local variable.

This is needed for the next patch (use index as a bit index).

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-52-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: rework and simplify mxser_open

In mxser_open, tty->port is already set thanks to
tty_port_register_device in ->probe. So we can simply use container_of
to deduce the struct mxser_port's pointer. Thus avoiding divisions and
complex computations.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-51-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: cleanup mxser_process_txrx_fifo

Rename process_txrx_fifo to mxser_process_txrx_fifo and:
* remove useless parentheses
* return from the 'if's true branch and process the rest in normal code
  flow (shift the code one level left)

All this to make the code more readable.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-50-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove mxser_cardinfo

Now that mxser_cards array (of struct mxser_cardinfo) contains only port
count and a quirk, we can encode these two directly into
pci_device_id::driver_data. And we can remove this struct completely.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-49-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove info message from probe

There is a dev_info in ->probe which prints device's name, bus and slot
numbers. All these can be fetched from the lspci output. So remove this
useless print.

This eliminates the whole mxser_cardinfo::name field.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-48-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: move request irq to probe and switch to managed

Move request_irq from mxser_initbrd to mxser_probe so that we can switch
it to managed request. It simplifies the cleanup code.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-47-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: pci, switch to managed resources

Switch to managed resources for PCI using pcim_enable_device.
Regions and device disabling is now taken care of by the core and we
need not unwind. Neither in ->probe, nor in ->remove. That simplifies
the code.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-46-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: inline mxser_board_remove into mxser_remove

The only user of mxser_board_remove is mxser_remove. Move there those
few lines.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-45-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: move board init into mxser_initbrd

What's done in ->probe for board is actually its initialization. So move
it to mxser_initbrd where it belongs.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-44-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: clean up the rest of MUST helpers

Clean up the rest of MUST helpers. They all can use the newly added
mxser_must_set_EFR. And passing a bool instead of having two functions
(_enable+_disable) simplifies the code a lot too.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-43-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: introduce mxser_must_select_bank and use it

Again, selecting a bank in the EFR register is duplicated many times.
Extract it to a separate function (mxser_must_select_bank) and use it on
all the places.

This cleans up most of the helpers. And mxser_get_must_hardware_id now
returns the HW ID directly, not via a by-reference parameter.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-42-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: dedup mxser_must_set_enhance_mode

There are multiple copies of setting the EFR register. Separate it to a
new universal function called __mxser_must_set_EFR. And use it in the
code by introducing mxser_must_set_enhance_mode instead of
mxser_disable_must_enchance_mode and mxser_enable_must_enchance_mode.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-41-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: simplify mxser_ioctl_op_mode

* ModeMask local array is just inverted OP_MODE_MASK shifted by shiftbit.
  Drop this array and use the value directly.
* return from the if's true branch and drop 'else' branch by moving the
  code completely outside the if.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-40-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: extract mxser_ioctl_op_mode

This ioctl code sets RS232, RS422 or RS485 mode. Move the code to a
separate function (mxser_ioctl_op_mode) with a note that these
mxser-special ioctls should be likely replaced eventually by TIOCGRS485
and TIOCSRS485.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-39-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: cleanup mxser_change_speed

* use UART macros instead of magic constants
* let the default case reuse already existing code (CS5 case)

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-38-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop mxser_port::custom_divisor

It's an once-set constant, so define a macro for the constant
(MXSER_CUSTOM_DIVISOR) and use it in the code instead.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-37-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop mxser_port::baud_base

It's an once-set constant, so define a macro for the constant
(MXSER_BAUD_BASE) and use it in the code instead.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-36-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove unused mxser_port::stop_rx

It's been a long time since mxser_port::stop_rx was abandoned. Remove it.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-35-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop unused mxser_port::normal_termios

After the previous patches, mxser_port::normal_termios is unused, so
time to wave bye-bye.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-34-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: don't allocate MXSER_PORTS + 1

The MXSER_PORTS's tty_device is never registered and neither its
tty_port instance exists. Hence, it's quite pointless to allocate it.

It used to be used for global information fetches via ioctls. We have
just removed these as non-existent tty_device+tty_port was exactly the
reason to remove the ioctls.

It would be peculiar to add this "virtual" port as we require a backing
tty_port since commit ecbbfd44a08f (TTY: move tty buffers to tty_port)
in 2012.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-33-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove cnt from mxser_receive_chars

After the previous ioctls removal, cnt is needed only in
mxser_receive_chars_old now. So remove it from mxser_receive_chars and
mxser_receive_chars_new and account only in mxser_receive_chars_old.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-32-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_HighSpeedOn ioctl

1) The semantics are defined exactly nowhere.
2) The ioctl definition was never exposed to userspace, so noone should
   actually be reliably using this.
3) It's not even mentioned in any of the mxser utilities.

Provided the above, remove this mxser-special ioctl.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-31-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_ASPP_OQUEUE ioctl

1) The semantics are defined exactly nowhere.
2) The ioctl definition was never exposed to userspace, so noone should
   actually be reliably using this.
3) It's not even mentioned in any of the mxser utilities.

Provided the above, remove this mxser-special ioctl.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-30-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_GETMSTATUS ioctl

1) MOXA_GETMSTATUS is broken at least since commit ecbbfd44a08f (TTY:
   move tty buffers to tty_port) in 2012.
   Even though it's mentioned in the mon_pa moxa utility, obviously
   noone ever tried that in that timeframe.
2) The ioctl definition was never exposed to userspace, so noone
   should actually be reliably using this.

Provided the above, remove this mxser-special ioctl.

And remove the whole concept of mxser_ioctl_special as this was the last
ioctl in there.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-29-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_GETDATACOUNT ioctl

1) MOXA_GETDATACOUNT is broken at least since commit ecbbfd44a08f (TTY:
   move tty buffers to tty_port) in 2012.
   Even though it's mentioned in the mon_p moxa utility, obviously noone
   ever tried that in that timeframe.
2) The ioctl definition was never exposed to userspace, so noone should
   actually be reliably using this.

Provided the above, remove this mxser-special ioctl.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-28-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_CHKPORTENABLE ioctl

1) MOXA_CHKPORTENABLE is broken at least since commit ecbbfd44a08f (TTY:
   move tty buffers to tty_port) in 2012.
2) The ioctl definition was never exposed to userspace, so noone should
   actually be reliably using this.
3) It's not even mentioned in any of the mxser utilities.

Provided the above, remove this mxser-special ioctl.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-27-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_ASPP_LSTATUS ioctl

1) The semantics are defined exactly nowhere.
2) The ioctl definition was never exposed to userspace, so noone should
   actually be reliably using this.
3) It's not even mentioned in any of the mxser utilities.

Provided the above, remove this mxser-special ioctl.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-26-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_ASPP_MON and friends

1) MOXA_ASPP_MON_EXT is broken at least since commit ecbbfd44a08f (TTY:
   move tty buffers to tty_port) in 2012.
2) The ioctl definitions were never exposed to userspace, so noone should
   actually be reliably using them.
3) They're not even mentioned in any of the mxser utilities.

So remove all these mxser-special ioctls: MOXA_ASPP_MON,
MOXA_ASPP_MON_EXT, and MOXA_SDS_RSTICOUNTER.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-25-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_SET_BAUD_METHOD ioctl

1) The semantics are defined exactly nowhere.
2) The ioctl definition was never exposed to userspace, so noone should
   actually be reliably using this.
3) It's not even mentioned in any of the mxser utilities.

Provided the above, remove this mxser-special ioctl.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-24-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove MOXA_GET_MAJOR deprecated ioctl

MOXA_GET_MAJOR was deprecated in commit 41aee9a121fd (Char: mxser, ioctl
cleanup) in 2008. Remove this ioctl finally.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-23-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop unused MOXA_DIAGNOSE macro

It's unused.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-22-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop UART_MCR_AFE and UART_LSR_SPECIAL defines

* UART_MCR_AFE is a redefinition of serial_reg.h's one.
* UART_LSR_SPECIAL is a redefinition of serial_reg.h's
  UART_LSR_BRK_ERROR_BITS.

So remove both and replace the latter uses by already defined
UART_LSR_BRK_ERROR_BITS.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-21-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: make xmit ring buffer variables unsigned

Make the xmit ring buffer pointers and counter (xmit_head, xmit_tail,
xmit_cnt) unsigned. Now, the assumptions for the compiler are clear --
they cannot be negative.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-20-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: correct types for uart variables

In struct mxser_port, all the UART registers and masks, and the xon/xoff
character are stored into int. Let's re-type all these as UART registers
are 8bit (u8) and xon/xoff character is an unsigned char (cc_t in
ktermios).

This save some bytes in memory, but more importantly the change makes it
what it really is.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-19-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove else from LSR bits checks

UART_LSR_SPECIAL is composed of UART_LSR_BI, UART_LSR_PE, UART_LSR_FE,
UART_LSR_OE. So status cannot be anything else. Remove the unused else
branch.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-18-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: extract mxser_receive_chars_old

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-17-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: extract mxser_receive_chars_new

Extract the new handling of receive chars (using the MUST chip
enhancements) into a separate function. This eliminates the need for the
intr_old label and makes the code of mxser_receive_chars more compact.
In the next step, we will extract the old handling too.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-16-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: simplify mxser_interrupt and drop mxser_board::vector_mask

mxser_board::vector_mask is just a bitfield with bits set for all
available ports.  We can obtain this value simply by
"BIT(brd->info->nports) - 1" directly in the ISR. So remove vector_mask
and simplify the code a bit.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-15-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: extract port ISR

The ISR is terrible mix of letters. Let's extract the proper per-port
handling to a separate function called mxser_port_isr. This way, we can
actually see what both mxser_interrupt and mxser_port_isr do now.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-14-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: cleanup LSR handling in mxser_receive_chars

mxser_interrupt is a bit confusing as it passes a pointer to LSR
variable ("status") to mxser_receive_chars. That function can indeed
update the LCR variable. Make this clear by not passing a variable by
reference, instead, return updated status. And change its type
accordingly -- to u8.

Note that the next patches will rework ISR so that even the definition
of status will be u8 soon too.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-13-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove nonsense from ISR

request_irq was passed a board. There is no need to check dev_id and/or
find the appropriate board in the ISR. Neither check if board is NULL.
Remove this nonsense.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-12-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: move max_baud from port to board

mxser_port::max_baud is actually a board property, not port's. So move
it to mxser_board.

While at it, drop the printk informing about max baud rate during probe.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-11-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop constant board::uart_type

board::uart_type is always set to PORT_16550A. So, use this constant in
the code instead. And drop the mxser_board member.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-10-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: introduce enum mxser_must_hwid

Provided the MUST HW ID is an enumeration, define one (enum
mxser_must_hwid) and use it in the code.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-9-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: rename mxser_board::chip_flag to must_hwid

It is named IsMoxaMustChipFlag in the original Moxa's driver. Over the
time, it was reduced to chip_flag which doesn't make much sense. So dub
it again by what it is: must_hwid (mxser_must_get_hwid assigns to it).

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-8-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: rename CheckIsMoxaMust to mxser_get_must_hwid

CheckIsMoxaMust actually finds out what the card is. So rename it to
mxser_must_get_hwid, so that the name conforms to the semantics of the
function.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-7-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: cleanup Gpci_uart_info struct

The Gpci_uart_info structure contains up to tripled information and
chose weird types for the members. Deduplicate the members and choose
proper types for the data -- everything can be u8 except baud rate.
There, we use speed_t, which is used in this context.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-6-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: integrate mxser.h into .c

As mxser.h serves only mxser.c, integrate the definitions into .c and
drop .h. There are some "semi-public" ioctl definitions, but they are
mostly go away in the next patches. If something, they would need to
live in an uapi header anyway.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-5-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: remove info printout from init

There is no reason/desire to print module information upon module load.
Drop this printk (and a version define).

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-4-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: renumber mxser_cards

After the ISA cards removal in the previous patch, the mxser_cards array
has holes in it. So renumber the array while updating PCI device's
driver_data.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-3-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxser: drop ISA support

While mxser PCI cards are still around and produced (Moxa provided me
with two recently), ISA cards are obsolete for a long time. I haven't
seen anyone using the cards and the ISA code paths are barely tested.

Hence, remove ISA support from mxser driver so that we can clean the
driver up.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210618061516.662-2-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial_cs: Add Option International GSM-Ready 56K/ISDN modem

Add support for Option International GSM-Ready 56K/ISDN PCMCIA modem
card.

Signed-off-by: Ondrej Zary <linux@zary.sk>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210611201940.23898-2-linux@zary.sk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial_cs: remove wrong GLOBETROTTER.cis entry

The GLOBETROTTER.cis entry in serial_cs matches more devices than
intended and breaks them. Remove it.

Example: # pccardctl info
PRODID_1="Option International
"
PRODID_2="GSM-Ready 56K/ISDN
"
PRODID_3="021
"
PRODID_4="A
"
MANFID=0013,0000
FUNCID=0

result:
pcmcia 0.0: Direct firmware load for cis/GLOBETROTTER.cis failed with error -2

The GLOBETROTTER.cis is nowhere to be found. There's GLOBETROTTER.cis.ihex at
https://netdev.vger.kernel.narkive.com/h4inqdxM/patch-axnet-cs-fix-phy-id-detection-for-bogus-asix-chip#post41
It's from completely diffetent card:
vers_1 4.1, "Option International", "GSM/GPRS GlobeTrotter", "001", "A"

Signed-off-by: Ondrej Zary <linux@zary.sk>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210611201940.23898-1-linux@zary.sk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

MIPS: Malta: Enable magic multipliers for Super I/O UARTs

The SMSC FDC37M817 Super I/O chip has a configuration feature that lets
it support special UART divisor values of 32770 and 32769 for bit rates
of 230400 and 460800 bits per second respectively.  Our 8250 driver core
provides support for these special divisors via the UPF_MAGIC_MULTIPLIER
flag, and YAMON firmware unconditionally configures the Super I/O chip
with these divisors enabled as well, so all we need to do in platform
setup for these bit rates to work is to set the flag.

Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2105182249380.3032@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: 8250: Handle custom baud rates in UPF_MAGIC_MULTIPLIER range

Handle custom baud rates requested in the UPF_MAGIC_MULTIPLIER range
with BOTHER.  Currently matching is exact, that is if a baud rate that
is not either of clk/4 or clk/8 is requested, then we fall through to
the default case, which will just divide the clock rate by 16 times the
rate requested, round it to closest integer, and possibly yield even
worse results then if clamping to the extra baud rates was chosen.

So for example if we have the usual base rate of 115200 and request a
rate of 230399, then the fall-through divisor calculation will yield 1,
and consequently the baud rate of 115200 will be programmed even though
obviously the magic rate of 230400 would be more appropriate.

Make the selection of the magic rates range-qualified then and use clk/4
for rates from clk/6 up (assuming `serial8250_get_baud_rate' has already
rejected any rates too far beyond clk/4), and otherwise use clk/8 for
rates from clk/12 up, and finally fall through to the standard divisor
calculation.  Explicitly void using the undocumented rate of 153600bps
and stick to documented divisor values only.

Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2105200232090.29169@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: 8250: Actually allow UPF_MAGIC_MULTIPLIER baud rates

Support for magic baud rate divisors of 32770 and 32769 used with SMSC
Super I/O chips for extra baud rates of 230400 and 460800 respectively
where base rate is 115200[1] has been added around Linux 2.5.64, which
predates our repo history, but the origin could be identified as commit
2a717aad772f ("Merge with Linux 2.5.64.") with the old MIPS/Linux repo
also at: <git://git.kernel.org/pub/scm/linux/kernel/git/ralf/linux.git>.

Code that is now in `serial8250_do_get_divisor' was added back then to
`serial8250_get_divisor', but that code would only ever trigger if one
of the higher baud rates was actually requested, and that cannot ever
happen, because the earlier call to `serial8250_get_baud_rate' never
returns them.  This is because it calls `uart_get_baud_rate' with the
maximum requested being the base rate, that is clk/16 or 115200 for SMSC
chips at their nominal clock rate.

Fix it then and allow UPF_MAGIC_MULTIPLIER baud rates to be selected, by
requesting the maximum baud rate of clk/4 rather than clk/16 if the flag
has been set.  Also correct the minimum baud rate, observing that these
ports only support actual (non-magic) divisors of up to 32767 only.

References:

[1] "FDC37M81x, PC98/99 Compliant Enhanced Super I/O Controller with
    Keyboard/Mouse Wake-Up", Standard Microsystems Corporation, Rev.
    03/27/2000, Table 31 - "Baud Rates", p. 77

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2105190412280.29169@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: 8250: Document SMSC Super I/O UART peculiarities

Contrary to what SMSC documentation says and unlike NS16C550A UARTs the
SMSC Super I/O IC claims compatibility with the SMSC UART implementation
does not support dividing the internal PLL clock by any divisor from 1
to 65535[1], with the exception of two magic divisors of 32769 and 32770
used respectively to select the high-speed data rates of 460800bps and
230400bps[2] if enabled on a port-by-port basis in with the Serial Port
Mode Register in the Device Configuration Space[3][4].

Instead empirical evidence indicates that the divisor, as stored in the
DLL and DLM register pair, has the range of 1 to 32767 only, and bit 7
of the DLM register (bit 15 of the divisor) effectively serves as a
selection bit for the prescaler from the base frequency of 7.3728MHz,
either 4 if the bit is 0, or 1 if the bit is 1 and high-speed operation
has been enabled with the Serial Port Mode Register.

So if high-speed operation has not been enabled, then say the values of
1 and 32769 (0x8001) written into the combined DLL and DLM register pair
both select the divisor of 1 and the baud rate of 115200bps.

[1] "FDC37M81x, PC98/99 Compliant Enhanced Super I/O Controller with
    Keyboard/Mouse Wake-Up", Standard Microsystems Corporation, Rev.
    03/27/2000, Section "Programmable Baud Rate Generator (and Divisor
    Latches DLH, DLL)", p. 75

[2] same, Table 31 - "Baud Rates", p. 77

[3] same, Table 60 - "Serial Port 1, Logical Device 4 [Logical Device
    Number = 0x04]", p. 153

[4] same, Table 61 - "Serial Port 2, Logical Device 5 [Logical Device
    Number = 0x05]", p. 153

Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2106092330530.5469@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

dt-bindings: serial: Move omap-serial.txt to YAML schema

Convert serial-omap.txt to YAML schema for better checks and documentation.

Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Link: https://lore.kernel.org/r/20210610182227.2480-1-vigneshr@ti.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: 8250: Correct the clock for OxSemi PCIe devices

Oxford Semiconductor PCIe (Tornado) serial port devices are driven by a
fixed 62.5MHz clock input derived from the 100MHz PCI Express clock.

In the enhanced (650) mode, which we select in `autoconfig_has_efr' by
setting the ECB bit in the EFR register, and in the absence of clock
reconfiguration, which we currently don't do, the clock rate is divided
only by the oversampling rate of 16 as it is supplied to the baud rate
generator, yielding the baud base of 3906250.  This comes from the reset
values of the TCR and MCR[7] registers which are both zero[1][2][3][4],
choosing the oversampling rate of 16 and the normal (divide by 1) baud
rate generator prescaler respectively.  This is the rate that is divided
by the value held in the divisor latch to determine the baud rate used.

Replace the incorrect baud base of 4000000 with the right value of
3906250 then.

References:

[1] "OXPCIe200 PCI Express Multi-Port Bridge", Oxford Semiconductor,
    Inc., DS-0045, 10 Nov 2008, Section "Reset Configuration", p. 72

[2] "OXPCIe952 PCI Express Bridge to Dual Serial & Parallel Port",
    Oxford Semiconductor, Inc., DS-0046, Mar 06 08, Section "Reset
    Configuration", p. 27

[3] "OXPCIe954 PCI Express Bridge to Quad Serial Port", Oxford
    Semiconductor, Inc., DS-0047, Feb 08, Section "Reset Configuration",
    p. 28

[4] "OXPCIe958 PCI Express Bridge to Octal Serial Port", Oxford
    Semiconductor, Inc., DS-0048, Feb 08, Section "Reset Configuration",
    p. 28

Fixes: 7106b4e333bae ("8250: Oxford Semiconductor Devices")
Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2106100203510.5469@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: 8250: Dissociate 4MHz Titan ports from Oxford ports

Oxford Semiconductor PCIe (Tornado) serial port devices have their baud
base set incorrectly, however their `pciserial_board' entries have been
reused for Titan serial port devices.  Define own entries for the latter
devices then, carrying over the settings, so that Oxford entries can be
fixed.

Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2106100142310.5469@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tty: make use of tty_get_{char,frame}_size

In the previous patch, we introduced tty_get_char_size() and
tty_get_frame_size() for computing character and frame sizes,
respectively. Here, we make use of them in various tty drivers where
applicable.

The stats look nice: 12 insertions, 169 deletions.

Cc: Arnd Bergmann <arnd@arndb.de>
Cc: David Lin <dtwlin@gmail.com>
Cc: Johan Hovold <johan@kernel.org>
Cc: Alex Elder <elder@kernel.org>
Cc: Shawn Guo <shawnguo@kernel.org>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Andy Gross <agross@kernel.org>
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Cc: Alexandre Torgue <alexandre.torgue@foss.st.com>
Cc: Oliver Neukum <oneukum@suse.com>
Acked-by: Alex Elder <elder@kernel.org>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210610090247.2593-4-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cypress_m8: switch data_bits to real character bits

Make data_bits what it really is. Assign proper bit counts to data_bits
instead of magic 0..3. There are two reasons:
1) it's clear what we store there, and
2) it will make the transition to tty_tty_get_char_size() in the next
   patch easier.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210610090247.2593-3-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mxs-auart: redefine AUART_LINECTRL_WLEN to accept bits count

It's a bit illogical to take magic constants in AUART_LINECTRL_WLEN.
Like in auart_console_get_options(), 2 means 7 bits. Switch
AUART_LINECTRL_WLEN to accept bit length, i.e. let it subtract 5 from
the parameter before doing the logic.

This will ease conversion from CSIZE to bits in mxs_auart_settermios()
in the next patch.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210610090247.2593-2-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tty: make tty_get_{char,frame}_size available

Many tty drivers contain code to compute bits count depending on termios
cflags. So extract this code from serial core to two separate tty helper
functions:
* tty_get_char_size -- only size of a character, without flags,
* tty_get_frame_size -- complete size of a frame including flags.

In the next patch, calls to these new functions replace many copies of
this code.

Note that we accept only cflag as a parameter. That's because some
callers like pch_uart_startup or sunsab_console_setup don't have at hand
termios which we could pass around.

Cc: Joe Perches <joe@perches.com>
Cc: Johan Hovold <johan@kernel.org>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210610090247.2593-1-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ipwireless: remove unused ipw_tty::closing

It's only set, but never read because commit 01261cb94318 ("tty:
ipwireless: Remove tty->closing abort from ipw_open()") removed last use
of tty::closing.

Cc: Jiri Kosina <jikos@kernel.org>
Cc: David Sterba <dsterba@suse.com>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20210610090307.2689-1-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: stm32: defer probe for dma devices

Defer device probe operation when a DMA channel request probe deferral.

With this change both DMA channels are acquired before DMA channels are
configured. Once no probe deferral is expected, DMAs are
configured and any failure in their configuration will make the driver
to fallback to interrupt mode as prior this change.

Signed-off-by: Erwan Le Ray <erwan.leray@foss.st.com>
Signed-off-by: Etienne Carriere <etienne.carriere@foss.st.com>
Link: https://lore.kernel.org/r/20210610100020.2318-1-erwan.leray@foss.st.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: tegra-tcu: Reorder channel initialization

Request the RX mailbox only after initializing the UART data
structures. Otherwise it can rarely happen that the receive
callback is called before the UART is ready.

Fixes: 2d908b38d409 ("serial: Add Tegra Combined UART driver")
Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Link: https://lore.kernel.org/r/20210609155655.3567545-1-mperttunen@nvidia.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

serial: sh-sci: Stop dmaengine transfer in sci_stop_tx()

Stop dmaengine transfer in sci_stop_tx(). Otherwise, the following
message is possible output when system enters suspend and while
transferring data, because clearing TIE bit in SCSCR is not able to
stop any dmaengine transfer.

    sh-sci e6550000.serial: ttySC1: Unable to drain transmitter

Note that this driver has already used some #ifdef in the .c file
so that this patch also uses #ifdef to fix the issue. Otherwise,
build errors happens if the CONFIG_SERIAL_SH_SCI_DMA is disabled.

Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.")
Cc: <stable@vger.kernel.org> # v4.9+
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Link: https://lore.kernel.org/r/20210610110806.277932-1-yoshihiro.shimoda.uh@renesas.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Merge tag 'v5.13-rc6' into tty-next

We want the tty fixes in here as well.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Linux 5.13-rc6

Merge tag 'perf-tools-fixes-for-v5.13-2021-06-13' of git://git./linux/kernel/git/acme/linux

Pull perf tools fixes from Arnaldo Carvalho de Melo:

 - Correct buffer copying when peeking events

 - Sync cpufeatures/disabled-features.h header with the kernel sources

* tag 'perf-tools-fixes-for-v5.13-2021-06-13' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
  tools headers cpufeatures: Sync with the kernel sources
  perf session: Correct buffer copying when peeking events

Merge tag 'nfs-for-5.13-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client bugfixes from Trond Myklebust:
 "Highlights include:

  Stable fixes:

   - Fix use-after-free in nfs4_init_client()

  Bugfixes:

   - Fix deadlock between nfs4_evict_inode() and nfs4_opendata_get_inode()

   - Fix second deadlock in nfs4_evict_inode()

   - nfs4_proc_set_acl should not change the value of NFS_CAP_UIDGID_NOMAP

   - Fix setting of the NFS_CAP_SECURITY_LABEL capability"

* tag 'nfs-for-5.13-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  NFSv4: Fix second deadlock in nfs4_evict_inode()
  NFSv4: Fix deadlock between nfs4_evict_inode() and nfs4_opendata_get_inode()
  NFS: FMODE_READ and friends are C macros, not enum types
  NFS: Fix a potential NULL dereference in nfs_get_client()
  NFS: Fix use-after-free in nfs4_init_client()
  NFS: Ensure the NFS_CAP_SECURITY_LABEL capability is set when appropriate
  NFSv4: nfs4_proc_set_acl needs to restore NFS_CAP_UIDGID_NOMAP on error.