Eric Lee / smarc-fsl-linux-kernel

11 Apr, 2015

2 commits

  • 704f32d48   spi: bcm2835: enabling polling mode for transfers shorter than 30us ... Browse Code »

    In cases of short transfer times the CPU is spending lots of time
    in the interrupt handler and scheduler to reschedule the worker thread.

    Measurements show that we have times where it takes 29.32us to between
    the last clock change and the time that the worker-thread is running again
    returning from wait_for_completion_timeout().

    During this time the interrupt-handler is running calling complete()
    and then also the scheduler is rescheduling the worker thread.

    This time can vary depending on how much of the code is still in
    CPU-caches, when there is a burst of spi transfers the subsequent delays
    are in the order of 25us, so the value of 30us seems reasonable.

    With polling the whole transfer of 4 bytes at 10MHz finishes after 6.16us
    (CS down to up) with the real transfer (clock running) taking 3.56us.
    So the efficiency has much improved and is also freeing CPU cycles,
    reducing interrupts and context switches.

    Because of the above 30us seems to be a reasonable limit for polling.

    Signed-off-by: Martin Sperl
    Signed-off-by: Mark Brown

    Martin Sperl
     
  • a30a555d7   spi: bcm2835: transform native-cs to gpio-cs on first spi_setup ... Browse Code »

    Transforms the bcm-2835 native SPI-chip select to their gpio-cs equivalent.

    This allows for some support of some optimizations that are not
    possible due to HW-gliches on the CS line - especially filling
    the FIFO before enabling SPI interrupts (by writing to CS register)
    while the transfer is already in progress (See commit: e3a2be3030e2)

    This patch also works arround some issues in bcm2835-pinctrl which does not
    set the value when setting the GPIO as output - it just sets up output and
    (typically) leaves the GPIO as low. When a fix for this is merged then this
    gpio_set_value can get removed from bcm2835_spi_setup.

    Signed-off-by: Martin Sperl
    Signed-off-by: Mark Brown

    Martin Sperl
     

30 Mar, 2015

2 commits

27 Mar, 2015

1 commit

  • e34ff011c   spi: bcm2835: move to the transfer_one driver model ... Browse Code »

    This also allows for GPIO-CS to get used removing the limitation of
    2/3 SPI devises on the SPI bus.

    Fixes: spi-cs-high with native CS with multiple devices on the spi-bus
    resetting the chip selects to "normal" polarity after a finished
    transfer.

    No other functionality/improvements added.

    Tested with the following 4 devices on the spi-bus:
    * mcp2515 with native CS
    * mcp2515 with gpio CS
    * fb_st7735r with native CS
    (plus spi-cs-high via transistor inverting polarity)
    * enc28j60 with gpio-CS
    Tested-by: Martin Sperl

    Signed-off-by: Martin Sperl
    Signed-off-by: Mark Brown

    Martin Sperl
     

24 Mar, 2015

3 commits

21 Mar, 2015

1 commit

  • 342f948a1   spi: bcm2835: fix all checkpath --strict messages ... Browse Code »

    The following errors/warnings issued by checkpatch.pl --strict have been fixed:
    drivers/spi/spi-bcm2835.c:182: CHECK: Alignment should match open parenthesis
    drivers/spi/spi-bcm2835.c:191: CHECK: braces {} should be used on all arms of this statement
    drivers/spi/spi-bcm2835.c:234: CHECK: Alignment should match open parenthesis
    drivers/spi/spi-bcm2835.c:256: CHECK: Alignment should match open parenthesis
    drivers/spi/spi-bcm2835.c:271: CHECK: Alignment should match open parenthesis
    drivers/spi/spi-bcm2835.c:346: CHECK: Alignment should match open parenthesis
    total: 0 errors, 0 warnings, 6 checks, 403 lines checked

    In 2 locations the arguments had to get split/moved to the next line so that the
    line width stays below 80 chars.

    Signed-off-by: Martin Sperl
    Signed-off-by: Mark Brown

    Martin Sperl
     

22 Dec, 2014

1 commit

20 Oct, 2014

1 commit

16 Feb, 2014

1 commit

  • 78e39523b   spi: Remove explictly set bus_num and num_chipselect to default setting ... Browse Code »

    The purpose of commit 1e8a52e18cfb
    "spi: By default setup spi_masters with 1 chipselect and dynamics bus number"
    is to avoid setting default value for bus_num and num_chipselect in spi master
    drivers. So let's remove the duplicate code.

    Signed-off-by: Axel Lin
    Acked-by: Uwe Kleine-König
    Acked-By: David Daney
    Acked-by: Stephen Warren
    Signed-off-by: Mark Brown

    Axel Lin
     

18 Dec, 2013

1 commit

28 Nov, 2013

1 commit

15 Nov, 2013

2 commits

26 Sep, 2013

1 commit

23 Aug, 2013

1 commit

  • 7bc003100   spi: bcm2835: Add spi_master_get() call to prevent use after free ... Browse Code »

    The call to spi_unregister_master results in device memory being freed, it must
    no longer be accessed afterwards. Thus call spi_master_get() to get an extra
    reference to the device and call spi_master_put() only after the last access to
    device data.

    Note, current code has an extra spi_master_put() call in bcm2835_spi_remove().
    Thus this patch just adds an spi_master_get() to balance the reference count.

    Signed-off-by: Axel Lin
    Signed-off-by: Mark Brown

    Axel Lin
     

05 Aug, 2013

1 commit

24 Jul, 2013

1 commit

13 May, 2013

1 commit

01 Apr, 2013

1 commit

13 Mar, 2013

1 commit

  • f8043872e   spi: add driver for BCM2835 ... Browse Code »

    The BCM2835 contains two forms of SPI master controller (one known
    simply as SPI0, and the other known as the "Universal SPI Master", in
    the auxilliary block) and one form of SPI slave controller. This patch
    adds support for the SPI0 controller.

    This driver is taken from Chris Boot's repository at
    git://github.com/bootc/linux.git rpi-linear
    as of commit 6de2905 "spi-bcm2708: fix printf with spurious %s".
    In the first SPI-related commit there, Chris wrote:

    Thanks to csoutreach / A Robinson for his driver which I used as an
    inspiration. You can find his version here:
    http://piface.openlx.org.uk/raspberry-pi-spi-kernel-driver-available-for

    Changes made during upstreaming:
    * Renamed bcm2708 to bcm2835 as per upstream naming for this SoC.
    * Removed support for brcm,realtime property.
    * Increased transfer timeout to 30 seconds.
    * Return IRQ_NONE from the IRQ handler if no interrupt was handled.
    * Disable TA (Transfer Active) and clear FIFOs on a transfer timeout.
    * Wrote device tree binding documentation.
    * Request unnamed clock rather than "sys_pclk"; the DT will provide the
    correct clock.
    * Assume that tfr->speed_hz and tfr->bits_per_word are always set in
    bcm2835_spi_start_transfer(), bcm2835_spi_transfer_one(), so no need
    to check spi->speed_hz or tft->bits_per_word.
    * Re-ordered probe() to remove the need for temporary variables.
    * Call clk_disable_unprepare() rather than just clk_unprepare() on probe()
    failure.
    * Don't use devm_request_irq(), to ensure that the IRQ doesn't fire after
    we've torn down the device, but not unhooked the IRQ.
    * Moved probe()'s call to clk_prepare_enable() so we can be sure the clock
    is enabled if the IRQ handler fires immediately.
    * Remove redundant checks from bcm2835_spi_check_transfer() and
    bcm2835_spi_setup().
    * Re-ordered IRQ handler to check for RXR before DONE. Added comments to
    ISR.
    * Removed empty prepare/unprepare implementations.
    * Removed use of devinit/devexit.
    * Added BCM2835_ prefix to defines.

    Signed-off-by: Chris Boot
    Signed-off-by: Stephen Warren
    Signed-off-by: Mark Brown

    Chris Boot