[PATCH v2 0/6] continuous regeneration of finite samples

[Ian Abbott]

On 27/01/16 21:28, Spencer E. Olson wrote:
> These patches implement a proposal to very slightly extend the interpretation of
> cmd->stop_arg when cmd->stop_src == TRIG_NONE.  Normally, for output subdevices,
> when the user specifies cmd->stop_src == TRIG_NONE, the device should
> continuously output data that must also be continually fed by the user via
> either a mmap/comedi_mark_buffer_written or an write(...) operation.  The intent
> of this proposal is to also allow a user to have a specific buffer repeated
> as-is indefinitely.  This is accomplished by the user specifying the length of
> the fixed buffer via cmd->stop_arg > 0.  It should be noted that cmd->stop_arg
> in this case must also be <= the size of the available dma buffer.  The prior
> behavior is maintained when cmd->stop_arg == 0, which is also enforced by the
> *_cmdtest functions for many, if not all output devices.  All devices which do
> not implement this new behavior should simply continue to enforce cmd->stop_arg
> == 0 in *_cmdtest functions.
>
> These patches implement the extended interpretation of cmd->stop_arg >0 for when
> cmd->stop_src==TRIG_NONE for National Instruments m/e-series devices only that
> are based on the ni_mio_common driver.  Currently, these changes only apply to
> analog output and cdio output devices.
>
> This patches also implement suggestions made by Ian Abbot, namely:
> (1) all patches were run through checkpatch.pl and fixed accordingly to comply
> with coding standards,
> (2) all block comments were changed to comply with comments beginning on the
> line after the /* delimiter,
> (3) dev_warn use in drivers/mite.c as used during debugging was changed to
> dev_dbg,
> and
> (4) mite.c logic statements were made simpler and more robust following Ian's
> suggestions.
>
> ** ** v2 changes ** **
>
> (5) Prior PATCH 2/3 (staging: comedi: ni_mio_comon: adds finite regeneration to
> AO output) was split into 4 patches (current PATCHES [2-5]/6) that keep the
> changes a bit more atomic.
>
> Spencer E. Olson (6):
>    staging: comedi: mite: enable continuous regeneration of finite
>      samples
>    staging: comedi: ni_mio_common:  Cleans up/clarifies ni_ao_reset
>    staging: comedi: ni_mio_common:  make more bits in ao_cmd1 reg be
>      stateful
>    staging: comedi: ni_mio_common:  Cleans up/clarifies ni_ao_cmd
>    staging: comedi: ni_mio_common:  adds finite regeneration to AO output
>    staging: comedi: ni_mio_common: add finite regeneration to dio output
>
>   drivers/staging/comedi/drivers/mite.c          |  72 +++-
>   drivers/staging/comedi/drivers/mite.h          |   3 +
>   drivers/staging/comedi/drivers/ni_mio_common.c | 573 ++++++++++++++++++-------
>   3 files changed, 479 insertions(+), 169 deletions(-)
>

Thanks!

Reviewed-by: Ian Abbott <abbotti at mev.co.uk>

-- 
-=( Ian Abbott @ MEV Ltd.    E-mail: <abbotti at mev.co.uk> )=-
-=(                          Web: http://www.mev.co.uk/  )=-