imx_set_termios() cleared RTS on every call, now fixed.
Reviewed-by: NSascha Hauer <s.hauer@pengutronix.de>
Tested-by: NSascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: NSergey Organov <sorganov@gmail.com>
Link: https://lore.kernel.org/r/1561558293-7683-5-git-send-email-sorganov@gmail.comSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Avoid repeating the same code for rs485 twice.

Make it obvious we clear CRTSCTS bit in termios->c_cflag whenever
sport->have_rtscts is false.

Make it obvious we clear UCR2_IRTS whenever CRTSCTS is set.
Reviewed-by: NSascha Hauer <s.hauer@pengutronix.de>
Tested-by: NSascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: NSergey Organov <sorganov@gmail.com>
Link: https://lore.kernel.org/r/1561558293-7683-4-git-send-email-sorganov@gmail.comSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Set common bits in a separate statement to make initialization
explicit and not repeat the common part.
Reviewed-by: NSascha Hauer <s.hauer@pengutronix.de>
Tested-by: NSascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: NSergey Organov <sorganov@gmail.com>
Reviewed-by: NUwe Kleine-Knig <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/1561558293-7683-3-git-send-email-sorganov@gmail.comSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

While the .flush_buffer() callback clears sci_port.tx_dma_len since
commit 1cf4a7ef ("serial: sh-sci: Fix race condition causing
garbage during shutdown"), it does not terminate a transmit DMA
operation that may be in progress.

Fix this by terminating any pending DMA operations, and resetting the
corresponding cookie.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NEugeniu Rosca <erosca@de.adit-jv.com>
Tested-by: NEugeniu Rosca <erosca@de.adit-jv.com>

Link: https://lore.kernel.org/r/20190624123540.20629-3-geert+renesas@glider.beSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When uart_flush_buffer() is called, the .flush_buffer() callback zeroes
the tx_dma_len field.  This may race with the work queue function
handling transmit DMA requests:

  1. If the buffer is flushed before the first DMA API call,
     dmaengine_prep_slave_single() may be called with a zero length,
     causing the DMA request to never complete, leading to messages
     like:

        rcar-dmac e7300000.dma-controller: Channel Address Error happen

     and, with debug enabled:

	sh-sci e6e88000.serial: sci_dma_tx_work_fn: ffff800639b55000: 0...0, cookie 126

     and DMA timeouts.

  2. If the buffer is flushed after the first DMA API call, but before
     the second, dma_sync_single_for_device() may be called with a zero
     length, causing the transmit data not to be flushed to RAM, and
     leading to stale data being output.

Fix this by:
  1. Letting sci_dma_tx_work_fn() return immediately if the transmit
     buffer is empty,
  2. Extending the critical section to cover all DMA preparational work,
     so tx_dma_len stays consistent for all of it,
  3. Using local copies of circ_buf.head and circ_buf.tail, to make sure
     they match the actual operation above.
Reported-by: NEugeniu Rosca <erosca@de.adit-jv.com>
Suggested-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NEugeniu Rosca <erosca@de.adit-jv.com>
Tested-by: NEugeniu Rosca <erosca@de.adit-jv.com>
Link: https://lore.kernel.org/r/20190624123540.20629-2-geert+renesas@glider.beSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Support for the Marvell MV64x60 line of bridge chips that contained
MPSC controllers has been removed and there are no other components
that have that controller so remove its driver.
Signed-off-by: NMark Greer <mgreer@animalcreek.com>
Link: https://lore.kernel.org/r/20190626160553.28518-1-mgreer@animalcreek.comSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Add the missing unlock before return from function serial8250_register_8250_port()
in the error handling case.
Signed-off-by: NWei Yongjun <weiyongjun1@huawei.com>
Link: https://lore.kernel.org/r/20190703055908.141294-1-weiyongjun1@huawei.comSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Adds a flush of RX and TX FIFOs, and fixes some errors:
- adds RX FIFO flush in startup fonction
- removes the useless transmitter enabling in startup fonction
  (e.g. receiver only, see Documentation/serial/driver)
- configures FIFO threshold before enabling it, rather than after
- flushes both TX and RX in set_termios function
Signed-off-by: NErwan Le Ray <erwan.leray@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Adds the support of RX FIFO threshold in order to improve the RX FIFO
management.
This is done by enabling fifo threshold interrupt, instead of relying
on rx empty/fifo not full irq. That basically generates one irq/char
currently. With this patch:
- RXCFG is set to half fifo size (e.g. 16/2 = 8 data for a 16 data depth
  FIFO)
- irq rate may be reduced by up to 1/RXCFG,  e.g. 1 over 8 with current
  RXCFG setting.
- Receiver timeout is used to gather chars when FIFO threshold isn't
  reached.
Signed-off-by: NErwan Le Ray <erwan.leray@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Adds the support of TX FIFO threshold in order to improve the TX FIFO
management:
- TX FIFO threshold irq enabling (instead of relying on tx empty / fifo
  not full irq that generates one irq/char)
- TXCFG is set to half fifo size (e.g. 16/2 = 8 data for a 16 data depth
  FIFO)
- irq rate may be reduced by up to 1/TXCFG,  e.g. 1 over 8 with current
  TXCFG setting.
Signed-off-by: NErwan Le Ray <erwan.leray@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Improves PIO transmission:
- Replaces the FIFO filling per character by a filling per blocks of
  characters, which provides better performances
- Replaces the active waiting loop by TX empty interrupt dynamic handling.
  TXE interrupt is now enabled when data has to be sent (ie when
  uart_circ is not empty), and inhibited when there is no more data to
  send (ie when uart_circ is empty).
Signed-off-by: NErwan Le Ray <erwan.leray@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Add support of RX timeout interrupts to limit the number of interrupts.
RX timeout is a number of bits (baud clock cycles) without
transmission seen in the receiver. One character  is used as an arbitrary
RX timeout value.
If parity is enabled, the number of bits has to include parity bit.
Signed-off-by: NGerald Baeza <gerald.baeza@st.com>
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NErwan Le Ray <erwan.leray@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This reverts commit 2e9fe539.

Reading LSR unconditionally but processing the error flags only if
UART_IIR_RDI bit was set before in IIR may lead to a loss of transmission
error information on UARTs where the transmission error flags are cleared
by a read of LSR. Information are lost in case an error is detected right
before the read of LSR while processing e.g. an UART_IIR_THRI interrupt.
Signed-off-by: NOliver Barta <o.barta89@gmail.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Fixes: 2e9fe539 ("serial: 8250: Don't service RX FIFO if interrupts are disabled")
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch permits the usage for GPIOs to control
the CTS/RTS/DTR/DSR/DCD/RI signals.

Changed by Stefan:
Only call mctrl_gpio_init(), if the device has no ACPI companion device
to not break existing ACPI based systems. Also only use the mctrl_gpio_
functions when "gpios" is available.

Use MSR / MCR <-> TIOCM wrapper functions.
Signed-off-by: NYegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: NStefan Roese <sr@denx.de>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Tested-by: NYegor Yefremov <yegorslists@googlemail.com>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Giulio Benetti <giulio.benetti@micronovasrl.com>
Cc: Yegor Yefremov <yegorslists@googlemail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch adds a check for the GPIOs property existence, before the
GPIO is requested. This fixes an issue seen when the 8250 mctrl_gpio
support is added (2nd patch in this patch series) on x86 platforms using
ACPI.

Here Mika's comments from 2016-08-09:

"
I noticed that with v4.8-rc1 serial console of some of our Broxton
systems does not work properly anymore. I'm able to see output but input
does not work.

I bisected it down to commit 4ef03d32
("tty/serial/8250: use mctrl_gpio helpers").

The reason why it fails is that in ACPI we do not have names for GPIOs
(except when _DSD is used) so we use the "idx" to index into _CRS GPIO
resources. Now mctrl_gpio_init_noauto() goes through a list of GPIOs
calling devm_gpiod_get_index_optional() passing "idx" of 0 for each. The
UART device in Broxton has following (simplified) ACPI description:

    Device (URT4)
    {
        ...
        Name (_CRS, ResourceTemplate () {
            GpioIo (Exclusive, PullDefault, 0x0000, 0x0000, IoRestrictionOutputOnly,
                    "\\_SB.GPO0", 0x00, ResourceConsumer)
            {
                0x003A
            }
            GpioIo (Exclusive, PullDefault, 0x0000, 0x0000, IoRestrictionOutputOnly,
                    "\\_SB.GPO0", 0x00, ResourceConsumer)
            {
                0x003D
            }
        })

In this case it finds the first GPIO (0x003A which happens to be RX pin
for that UART), turns it into GPIO which then breaks input for the UART
device. This also breaks systems with bluetooth connected to UART (those
typically have some GPIOs in their _CRS).

Any ideas how to fix this?

We cannot just drop the _CRS index lookup fallback because that would
break many existing machines out there so maybe we can limit this to
only DT enabled machines. Or alternatively probe if the property first
exists before trying to acquire the GPIOs (using
device_property_present()).
"

This patch implements the fix suggested by Mika in his statement above.
Signed-off-by: NStefan Roese <sr@denx.de>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: NYegor Yefremov <yegorslists@googlemail.com>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Yegor Yefremov <yegorslists@googlemail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Giulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Fixes: 6bf4e42f ("serial: 8250: Add support for higher baud rates to Pericom chips")
Signed-off-by: Nkbuild test robot <lkp@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

A bug was introduced by commit b3b57646 ("tty: serial_core: convert
uart_open to use tty_port_open"). It caused a constant warning printed
into the system log regarding the tty and port counter mismatch:

[   21.644197] ttyS ttySx: tty_port_close_start: tty->count = 1 port count = 2

in case if session hangup was detected so the warning is printed starting
from the second open-close iteration.

Particularly the problem was discovered in situation when there is a
serial tty device without hardware back-end being setup. It is considered
by the tty-serial subsystems as a hardware problem with session hang up.
In this case uart_startup() will return a positive value with TTY_IO_ERROR
flag set in corresponding tty_struct instance. The same value will get
passed to be returned from the activate() callback and then being returned
from tty_port_open(). But since in this case tty_port_block_til_ready()
isn't called the TTY_PORT_ACTIVE flag isn't set (while the method had been
called before tty_port_open conversion was introduced and the rest of the
subsystem code expected the bit being set in this case), which prevents the
uart_hangup() method to perform any cleanups including the tty port
counter setting to zero. So the next attempt to open/close the tty device
will discover the counters mismatch.

In order to fix the problem we need to manually set the TTY_PORT_ACTIVE
flag in case if uart_startup() returned a positive value. In this case
the hang up procedure will perform a full set of cleanup actions including
the port ref-counter resetting.

Fixes: b3b57646 "tty: serial_core: convert uart_open to use tty_port_open"
Signed-off-by: NSerge Semin <fancer.lancer@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch adds wrapper functions to convert MSR <-> TIOCM and also
MCR <-> TIOCM. These functions are used now in serial8250_do_set_mctrl()
and serial8250_do_get_mctrl().
Signed-off-by: NStefan Roese <sr@denx.de>
Suggested-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Yegor Yefremov <yegorslists@googlemail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Giulio Benetti <giulio.benetti@micronovasrl.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor out similar code pieces that set or clear UART_IER_THRI bit to
serial8250_{set,clear}_THRI() helpers.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

There is no need to add an out-of-memory error message inside
the driver because the core MM code will take care of it.

Remove the unneeded OOM error message.
Signed-off-by: NFabio Estevam <festevam@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

There is no reason to call return at the end of function which should
return void.

The patch is also remove one checkpatch warning:
WARNING: void function return statements are not generally useful
+	return;
+}

Fixes: 6ee04c6c ("tty: xuartps: Add polled mode support for xuartps")
Signed-off-by: NNava kishore Manne <nava.manne@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch fixes this checkpatch warning:
WARNING: macros should not use a trailing semicolon
+#define to_cdns_uart(_nb) container_of(_nb, struct cdns_uart, \
+		clk_rate_change_nb);

Fixes: d9bb3fb1 ("tty: xuartps: Rebrand driver as Cadence UART")
Signed-off-by: NNava kishore Manne <nava.manne@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Trivial patch which fixes one checkpatch warning:
WARNING: line over 80 characters
+		       !(readl(port->membase + CDNS_UART_SR)
			& CDNS_UART_SR_TXFULL)) {

Fixes: c8dbdc84 ("serial: xuartps: Rewrite the interrupt handling logic")
Signed-off-by: NNava kishore Manne <nava.manne@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Trivial patch which fixes this checkpatch warning:
WARNING: Avoid multiple line dereference - prefer 'port->state->xmit.tail'
+				port->state->xmit.buf[port->state->xmit.
+				tail], port->membase + CDNS_UART_FIFO);

Fixes: c8dbdc84 ("serial: xuartps: Rewrite the interrupt handling logic")
Signed-off-by: NNava kishore Manne <nava.manne@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Octal permission is preffered compare to symbolic one.

This patch fixes checkpatch warnings:
Symbolic permissions 'S_IRUGO' are not preferred. Consider using octal
permissions '0444'.

Fixes: 85baf542 ("tty: xuartps: support 64 byte FIFO size")
Signed-off-by: NNava kishore Manne <nava.manne@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Let kernel to find out major number dynamically for the first device and
then reuse it for other instances.
This fixes the issue that each uart is registered with a
different major number.

After the patch:
crw-------    1 root     root      253,   0 Jun 10 08:31 /dev/ttyPS0
crw--w----    1 root     root      253,   1 Jan  1  1970 /dev/ttyPS1

Fixes: 024ca329 ("serial: uartps: Register own uart console and driver structures")
Signed-off-by: NShubhrajyoti Datta <shubhrajyoti.datta@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The Pericom chips can achieve additional baud rates by programming the
sample clock register. The baud rates can be described as
921600 * 16 / (16 - scr) for scr values 5 to 15. The divisor is set to 1
for these baud rates.

Adds new quirk for Pericom chips other than the four port chips to use
the
Signed-off-by: NJay Dolan <jay.dolan@accesio.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

imx_uart_set_termios() called imx_uart_rts_active(), or
imx_uart_rts_inactive() before taking port->port.lock.

As a consequence, sport->port.mctrl that these functions modify
could have been changed without holding port->port.lock.

Moved locking of port->port.lock above the calls to fix the issue.
Signed-off-by: NSergey Organov <sorganov@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This reverts commit c70669ec as it
breaks the build.
Reported-by: NStephen Rothwell <sfr@canb.auug.org.au>
Reported-by: Nkbuild test robot <lkp@intel.com>
Cc: Bich Hemon <bich.hemon@st.com>
Cc: Erwan Le Ray <erwan.leray@st.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The function msm_wait_for_xmitr can be taken with interrupts
disabled. In order to avoid a potential system lockup - demonstrated
under stress testing conditions on SoC QCS404/5 - make sure we wait
for a bounded amount of time.

Tested on SoC QCS404.
Signed-off-by: NJorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Interrupt handler checked THRE bit (transmitter holding register
empty) in LSR to detect if TX fifo is empty.
In case when there is only receive interrupts the TX handling
got called because THRE bit in LSR is set when there is no
transmission (FIFO empty). TX handling caused TX stop, which in
RS-485 half-duplex mode actually resets receiver FIFO. This is not
desired during reception because of possible data loss.

The fix is to check if THRI is set in IER in addition of the TX
fifo status. THRI in IER is set when TX is started and cleared
when TX is stopped.
This ensures that TX handling is only called when there is really
transmission on going and an interrupt for THRE and not when there
are only RX interrupts.
Signed-off-by: NKimmo Rautkoski <ext-kimmo.rautkoski@vaisala.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Remove spin_lock_irqsave in stm32_config_rs485, it cause recursive locking.
Already locked in uart_set_rs485_config.

Fixes: 1bcda09d ("serial: stm32: add support for RS485 hardware control mode")
Signed-off-by: NBorut Seljak <borut.seljak@t-2.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Deferred probe is an expected return value for clk_get() on many
platforms. The driver deals with it properly, so there's no need
to output a warning that may potentially confuse users.
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

dev_info() is more appropriate for printing messages inside drivers, so
switch to dev_info().
Signed-off-by: NFabio Estevam <festevam@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

dev_info() is more appropriate for printing messages inside drivers, so
switch to dev_info().
Signed-off-by: NFabio Estevam <festevam@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

In order to support Rx in-band wakeup, we need to enable irq wake on an
edge sensitive interrupt of Rx pin before suspend and disable it when
resuming.

This interrupt is used only as wake source to resume the system when
suspended. Note that the sent character will be lost as the controller is
actually suspended.

We use this to support wakeup on bluetooth. Bluetooth will repeatedly send
0xFD to wakeup host. Once host detects Rx falling, an interrupt is
triggered, and the system leaves sleep state. Then, the bluetooth driver
will send 0xFC to bluetooth and bluetooth can start to send normal HCI
packets.
Signed-off-by: NClaire Chang <tientzu@chromium.org>
Reviewed-by: NNicolas Boichat <drinkcat@chromium.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

In case the cable is not connected then the target gets into
an infinite wait for tx empty.
Add a timeout to the tx empty wait.
Reported-by: NJean-Francois Dagenais <jeff.dagenais@gmail.com>
Signed-off-by: NShubhrajyoti Datta <shubhrajyoti.datta@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently we are doing a read of the status register.
Move the spinlock after that as the reads need not be spinlock
protected. This patch prevents relaxing the cpu with spinlock held.
Signed-off-by: NShubhrajyoti Datta <shubhrajyoti.datta@xilinx.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Modify uart rx and complete for DMA
Signed-off-by: NLong Cheng <long.cheng@mediatek.com>
Reviewed-by: NVinod Koul <vkoul@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When modprobe/rmmod/modprobe module, if platform_driver_register() fails,
the kernel complained,

  proc_dir_entry 'driver/digicolor-usart' already registered
  WARNING: CPU: 1 PID: 5636 at fs/proc/generic.c:360 proc_register+0x19d/0x270

Fix this by adding uart_unregister_driver() when platform_driver_register() fails.
Reported-by: NHulk Robot <hulkci@huawei.com>
Signed-off-by: NKefeng Wang <wangkefeng.wang@huawei.com>
Acked-by: NBaruch Siach <baruch@tkos.co.il>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>