Re: [PATCH 0/4 V6] staging:iio: Add support for multiple buffers (testing required!)

[Jonathan Cameron <jic23@kernel.org>]

I've let this sit for nearly long enough I think.  If no one shouts
in the next few days (or asks for more time to take a look). I'll add
patches 1-3 of this the togreg branch of iio.git and send a pull request
for it to Greg.  The input client driver can wait for a bit.

I'm really really keen to get this lot in place in time to have a few
weeks of testing in linux-next before the merge window opens.

(it's over a year old now as Peter noticed, and I'm getting serriously
bored of it ;)

> Same as V5 but with the changes from Lars-Peter's review to close some
> memory leaks and catch up with some changes in the subsystem that had
> passed this by.  It should also now be possible to disable the
> timestamp! (oops).
> 
> Thanks to Lars for his work reviewing my patches and also for his
> excellent work reviewing in general! Without him the whole subsystem
> would be moving a lot slower.
> 
> 
> Again, more testing would be good (thanks for those who already
> have).
> 
> In quick summary this is needed to do interrupt driven in kernel
> clients as it adds the option to push to multiple 'buffer' interfaces.
> Note that some of these may not actually do any buffering but rather
> push directly into another subsytem (the example is input).
> 
> I would like to merge at least the first 3 patches early in this cycle.
> What I need is some testing across a reasonable range of devices.  Review
> is welcome of course as well!
> 
> The complexity in this set is entirely in the changes to industrialio-buffer.c
> concerning how we go about adding and removing client buffers.  To do this
> we need to work out which scan mask is appropriate given the requests of
> all the given buffers (and whether we can satisfy it).  I've been testing
> with the max1363 driver which probably has the most complex set of available
> scan masks.  Thus when a new buffer is added or one is removed the whole
> buffer set are torn down and built up again.  This may be undesirable for
> some applications but should occur only rarely.
> 
> The example driver is for bridging to the input subsystem and still needs
> some work, but that can occur later in this merge cycle whereas the changes
> to the core really want to go in asap.
> 
> Changes since v3 
> 
> Rebased and added other drivers that have occured whilst this patch series
> has been languishing in my local tree.
> 
> Changes since V2.
> 
> Improved handling of error conditions in the buffer insertion and
> removal code.  (suggested by Lars-Peter).  I want to take another
> look at these, but didn't want to keep people from testing this
> in the meantime.
> 
> Also a number of minor fixes that came up during testing.
> 
> Thanks,
> 
> Jonathan
> 
> 
> Jonathan Cameron (4):
>   staging:iio: Add support for multiple buffers
>   staging:iio:in kernel users: Add a data field for channel specific
>     info.
>   staging:iio: add a callback buffer for in kernel push interface
>   staging:iio: Proof of concept input driver.
> 
>  drivers/iio/Kconfig                             |   6 +
>  drivers/iio/Makefile                            |   1 +
>  drivers/iio/accel/hid-sensor-accel-3d.c         |  15 +-
>  drivers/iio/adc/ad7266.c                        |   3 +-
>  drivers/iio/adc/ad7476.c                        |   2 +-
>  drivers/iio/adc/ad_sigma_delta.c                |   2 +-
>  drivers/iio/adc/at91_adc.c                      |   3 +-
>  drivers/iio/buffer_cb.c                         | 113 +++++++
>  drivers/iio/gyro/hid-sensor-gyro-3d.c           |  15 +-
>  drivers/iio/industrialio-buffer.c               | 380 ++++++++++++++++--------
>  drivers/iio/industrialio-core.c                 |   1 +
>  drivers/iio/inkern.c                            |   1 +
>  drivers/iio/light/adjd_s311.c                   |   3 +-
>  drivers/iio/light/hid-sensor-als.c              |  15 +-
>  drivers/iio/magnetometer/hid-sensor-magn-3d.c   |  15 +-
>  drivers/staging/iio/Kconfig                     |  11 +
>  drivers/staging/iio/Makefile                    |   1 +
>  drivers/staging/iio/accel/adis16201_ring.c      |   2 +-
>  drivers/staging/iio/accel/adis16203_ring.c      |   2 +-
>  drivers/staging/iio/accel/adis16204_ring.c      |   2 +-
>  drivers/staging/iio/accel/adis16209_ring.c      |   2 +-
>  drivers/staging/iio/accel/adis16240_ring.c      |   2 +-
>  drivers/staging/iio/accel/lis3l02dq_ring.c      |   2 +-
>  drivers/staging/iio/adc/ad7298_ring.c           |   2 +-
>  drivers/staging/iio/adc/ad7606_ring.c           |   2 +-
>  drivers/staging/iio/adc/ad7887_ring.c           |   2 +-
>  drivers/staging/iio/adc/ad799x_ring.c           |   2 +-
>  drivers/staging/iio/adc/max1363_ring.c          |   2 +-
>  drivers/staging/iio/gyro/adis16260_ring.c       |   2 +-
>  drivers/staging/iio/iio_input.c                 | 225 ++++++++++++++
>  drivers/staging/iio/iio_input.h                 |  23 ++
>  drivers/staging/iio/iio_simple_dummy_buffer.c   |   5 +-
>  drivers/staging/iio/impedance-analyzer/ad5933.c |   4 +-
>  drivers/staging/iio/imu/adis16400_ring.c        |   5 +-
>  drivers/staging/iio/meter/ade7758_ring.c        |   2 +-
>  include/linux/iio/buffer.h                      |  24 +-
>  include/linux/iio/consumer.h                    |  48 +++
>  include/linux/iio/iio.h                         |   2 +
>  include/linux/iio/machine.h                     |   2 +
>  39 files changed, 743 insertions(+), 208 deletions(-)
>  create mode 100644 drivers/iio/buffer_cb.c
>  create mode 100644 drivers/staging/iio/iio_input.c
>  create mode 100644 drivers/staging/iio/iio_input.h
>