linux-stable.git/drivers/iio, branch v4.17.4

iio: sca3000: Fix an error handling path in 'sca3000_probe()'

2018-07-03T09:27:00+00:00

commit 4a5b45383ca371e123ba103d34d4b3b87616245c upstream.

Use 'devm_iio_kfifo_allocate()' instead of 'iio_kfifo_allocate()' in order
to simplify code and avoid a memory leak in an error path in
'sca3000_probe()'. A call to 'sca3000_unconfigure_ring()' was missing.

Sent via the next merge window as unimportant bug and there are
other patches dependent on it.

Signed-off-by: Christophe JAILLET 
Cc: 
Signed-off-by: Jonathan Cameron 
Signed-off-by: Greg Kroah-Hartman

iio: adc: ad7791: remove sample freq sysfs attributes

2018-07-03T09:27:00+00:00

commit 7eb6b35d93c356f1afebbfb808bc296d6351e708 upstream.

In the current state, these attributes are broken, because they are
registered already, and the kernel throws a warning.
The first registration happens via the `IIO_CHAN_INFO_SAMP_FREQ` flag from
the `ad_sigma_delta` driver.

In this commit these attrs are removed, and in the following the
IIO_CHAN_INFO_SAMP_FREQ behavior will be implemented, which replaces these
hooks.

This is done to make things a bit easier to review as there is a bit of
overlap in the patch if it's done all at once.

Fixes: a13e831fcaa7 ("staging: iio: ad7192: implement IIO_CHAN_INFO_SAMP_FREQ")

Signed-off-by: Alexandru Ardelean 
Cc: 
Signed-off-by: Jonathan Cameron 
Signed-off-by: Greg Kroah-Hartman

Merge tag 'iio-fixes-for-4.17a' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-linus

2018-05-09T15:08:29+00:00

Jonathan writes:

First round of IIO fixes for the 4.17 cycle.

* core
  - fix up some issues with overflow etc around wrong types
    for some fo the kfifo handling functions.  Seems unlikely
    this would be triggered in reality but the fixes are simple
    so let's tidy them up.  Second patch deals with checking
    the userspace value passed for length for potential overflow.
* ad7793
  - Catch up with changes to the ad_sigma_delta core and use
    read_raw / write_raw iwth IIO_CHAN_INFO_SAMP_FEW to handle
    sampling frequency control.
* at91-sama5d2
  - Channel config for differential channels was completely broken.
  - Missing Kconfig dependency for buffer support.
* hid-sensor
  - Fix an issue with powering up after resume due to wrong reference
    counting.
* stm32-dfsdm
  - Fix an issue with second writes of the oversampling settings
    failing.
  - Fix an issue with the sample rate being set to half of requested
    value when particular clock source is used.

iio: adc: select buffer for at91-sama5d2_adc

2018-04-21T14:57:10+00:00

We need to select the buffer code, otherwise we get build errors
with undefined functions on the trigger and buffer,
if we select just IIO and then AT91_SAMA5D2_ADC from menuconfig

This adds a Kconfig 'select' statement like other ADC
drivers have it already.

Fixes: 5e1a1da0f8c9 ("iio: adc: at91-sama5d2_adc: add hw trigger and buffer support")
Signed-off-by: Eugen Hristev 
Cc: 
Signed-off-by: Jonathan Cameron

iio: hid-sensor-trigger: Fix sometimes not powering up the sensor after resume

2018-04-21T14:52:35+00:00

hid_sensor_set_power_work() powers the sensors back up after a resume
based on the user_requested_state atomic_t.

But hid_sensor_power_state() treats this as a boolean flag, leading to
the following problematic scenario:

1) Some app starts using the iio-sensor in buffered / triggered mode,
   hid_sensor_data_rdy_trigger_set_state(true) gets called, setting
   user_requested_state to 1.
2) Something directly accesses a _raw value through sysfs, leading
   to a call to hid_sensor_power_state(true) followed by
   hid_sensor_power_state(false) call, this sets user_requested_state
   to 1 followed by setting it to 0.
3) Suspend/resume the machine, hid_sensor_set_power_work() now does
   NOT power the sensor back up because user_requested_state (wrongly)
   is 0. Which stops the app using the sensor in buffered mode from
   receiving any new values.

This commit changes user_requested_state to a counter tracking how many
times hid_sensor_power_state(true) was called instead, fixing this.

Cc: Bastien Nocera 
Cc: Srinivas Pandruvada 
Signed-off-by: Hans de Goede 
Acked-by: Srinivas Pandruvada 
Cc: 
Signed-off-by: Jonathan Cameron

iio: adc: at91-sama5d2_adc: fix channel configuration for differential channels

2018-04-15T19:21:33+00:00

When iterating through the channels, the index in the array is not the
scan index. Added an xlate function to translate to the proper index.
The result of the bug is that the channel array is indexed with a wrong index,
thus instead of the proper channel, we access invalid memory, which may
lead to invalid results and/or corruption.
This will be used also for devicetree channel xlate.

Fixes: 5e1a1da0f ("iio: adc: at91-sama5d2_adc: add hw trigger and buffer support")
Fixes: 073c66201 ("iio: adc: at91-sama5d2_adc: add support for DMA")
Signed-off-by: Eugen Hristev 
Cc: 
Signed-off-by: Jonathan Cameron

Merge tag 'gpio-v4.17-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio

2018-04-05T16:51:41+00:00

Pull GPIO updates from Linus Walleij:
 "This is the bulk of GPIO changes for the v4.17 kernel cycle:

  New drivers:

   - Nintendo Wii GameCube GPIO, known as "Hollywood"

   - Raspberry Pi mailbox service GPIO expander

   - Spreadtrum main SC9860 SoC and IEC GPIO controllers.

  Improvements:

   - Implemented .get_multiple() callback for most of the
     high-performance industrial GPIO cards for the ISA bus.

   - ISA GPIO drivers now select the ISA_BUS_API instead of depending on
     it. This is merged with the same pattern for all the ISA drivers
     and some other Kconfig cleanups related to this.

  Cleanup:

   - Delete the TZ1090 GPIO drivers following the deletion of this SoC
     from the ARM tree.

   - Move the documentation over to driver-api to conform with the rest
     of the kernel documentation build.

   - Continue to make the GPIO drivers include only
      and not the too broad  that we
     want to get rid of.

   - Managed to remove VLA allocation from two drivers pending more
     fixes in this area for the next merge window.

   - Misc janitorial fixes"

* tag 'gpio-v4.17-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio: (77 commits)
  gpio: Add Spreadtrum PMIC EIC driver support
  gpio: Add Spreadtrum EIC driver support
  dt-bindings: gpio: Add Spreadtrum EIC controller documentation
  gpio: ath79: Fix potential NULL dereference in ath79_gpio_probe()
  pinctrl: qcom: Don't allow protected pins to be requested
  gpiolib: Support 'gpio-reserved-ranges' property
  gpiolib: Change bitmap allocation to kmalloc_array
  gpiolib: Extract mask allocation into subroutine
  dt-bindings: gpio: Add a gpio-reserved-ranges property
  gpio: mockup: fix a potential crash when creating debugfs entries
  gpio: pca953x: add compatibility for pcal6524 and pcal9555a
  gpio: dwapb: Add support for a bus clock
  gpio: Remove VLA from xra1403 driver
  gpio: Remove VLA from MAX3191X driver
  gpio: ws16c48: Implement get_multiple callback
  gpio: gpio-mm: Implement get_multiple callback
  gpio: 104-idi-48: Implement get_multiple callback
  gpio: 104-dio-48e: Implement get_multiple callback
  gpio: pcie-idio-24: Implement get_multiple/set_multiple callbacks
  gpio: pci-idio-16: Implement get_multiple callback
  ...

iio:kfifo_buf: check for uint overflow

2018-03-30T10:16:12+00:00

Currently, the following causes a kernel OOPS in memcpy:

echo 1073741825 > buffer/length
echo 1 > buffer/enable

Note that using 1073741824 instead of 1073741825 causes "write error:
Cannot allocate memory" but no OOPS.

This is because 1073741824 == 2^30 and 1073741825 == 2^30+1. Since kfifo
rounds up to the nearest power of 2, it will actually call kmalloc with
roundup_pow_of_two(length) * bytes_per_datum.

Using length == 1073741825 and bytes_per_datum == 2, we get:

kmalloc(roundup_pow_of_two(1073741825) * 2
or kmalloc(2147483648 * 2)
or kmalloc(4294967296)
or kmalloc(UINT_MAX + 1)

so this overflows to 0, causing kmalloc to return ZERO_SIZE_PTR and
subsequent memcpy to fail once the device is enabled.

Fix this by checking for overflow prior to allocating a kfifo. With this
check added, the above code returns -EINVAL when enabling the buffer,
rather than causing an OOPS.

Signed-off-by: Martin Kelly 
cc: 
Signed-off-by: Jonathan Cameron

iio:buffer: make length types match kfifo types

2018-03-30T10:14:32+00:00

Currently, we use int for buffer length and bytes_per_datum. However,
kfifo uses unsigned int for length and size_t for element size. We need
to make sure these matches or we will have bugs related to overflow (in
the range between INT_MAX and UINT_MAX for length, for example).

In addition, set_bytes_per_datum uses size_t while bytes_per_datum is an
int, which would cause bugs for large values of bytes_per_datum.

Change buffer length to use unsigned int and bytes_per_datum to use
size_t.

Signed-off-by: Martin Kelly 
Cc: 
Signed-off-by: Jonathan Cameron

Merge 4.16-rc7 into staging-next

2018-03-28T11:33:37+00:00

We want the IIO and staging driver fixes in here as well.

Signed-off-by: Greg Kroah-Hartman