Re: [PATCH 00/11] staging: comedi: move comedi_fc functionality into core

[Ian Abbott <abbotti@mev.co.uk>]

On 15/09/14 19:21, Hartley Sweeten wrote:
> On Monday, September 15, 2014 5:46 AM, Ian Abbott wrote:
>> The "comedi_fc" module contains a few functions for use by several
>> low-level comedi drivers.  Since they are used by quite a few drivers,
>> it seems useful to migrate their functionality to the core "comedi"
>> module, renaming them in the process, while keeping the original
>> function names in "drivers/comedi_fc.h" available for now.  This set of
>> patches accomplishes that.
>>
>> The next stages after this will be to migrate the remaining inline
>> functions from "drivers/comedi_fc.h" (possibly making them externally
>> exported functions, although they are rather short), and then to update
>> the low-level drivers to call the functions with the new names,
>> eliminating the need to include "comedi_fc.h".  Then
>> "drivers/comedi_fc.h" can be removed as well.
>>
>> Note that PATCH 10 has a couple of checkpatch warnings due to
>> changing an existing block comment in non-standard block-comment style.
>>
>> PATCH 11 has a checkpatch warning to check whether MAINTAINERS needs
>> updating.  It doesn't.
>>
>> PATCH 11 might conflict with some other patch series that haven't been
>> committed yet.
>>
>> 01) staging: comedi: add comedi_bytes_per_scan()
>> 02) staging: comedi: comedi_fc: use comedi_bytes_per_scan()
>> 03) staging: comedi: add comedi_inc_scan_progress()
>> 04) staging: comedi: comedi_fc: use comedi_inc_scan_progress()
>> 05) staging: comedi: add comedi_handle_events()
>> 06) staging: comedi: comedi_fc: cfc_write_array_to_buffer() data is
>>      const
>> 07) staging: comedi: add comedi_write_array_to_buffer()
>> 08) staging: comedi: comedi_fc.h: use comedi_write_array_to_buffer()
>> 09) staging: comedi: add comedi_read_array_from_buffer()
>> 10) staging: comedi: migrate copyrights from "comedi_fc.c"
>> 11) staging: comedi: remove comedi_fc module
>>
>>   drivers/staging/comedi/Kconfig             |  31 -------
>>   drivers/staging/comedi/comedi_buf.c        |  66 +++++++++++++++
>>   drivers/staging/comedi/comedidev.h         |  11 +++
>>   drivers/staging/comedi/drivers.c           |  95 +++++++++++++++++++++
>>   drivers/staging/comedi/drivers/Makefile    |   1 -
>>   drivers/staging/comedi/drivers/comedi_fc.c | 132 -----------------------------
>>   drivers/staging/comedi/drivers/comedi_fc.h |  40 ++++++---
>>   7 files changed, 201 insertions(+), 175 deletions(-)
>
> Ian,
>
> Other than I think the comedi_fc.c functions should be migrated to
> comedi_buf.c instead of drivers.c I have not problems with this.

I think it's debatable. The only one of the three functions you 
mentioned that touches the buffer is comedi_inc_scan_progress(), and I'd 
rather keep that with comedi_bytes_per_scan().

> And, patch 11 does have a conflict with your "... use dma_alloc_coherent..."
> patches.

Yes, I'll have to fix up one or the other, depending which gets applied 
first (unless git can magically sort it out with a three way merge).

> But,
>
> Reviewed-by: H Hartley Sweeten <hsweeten@visionengravers.com>
>

Thanks!

-- 
-=( Ian Abbott @ MEV Ltd.    E-mail: <abbotti@mev.co.uk>        )=-
-=( Tel: +44 (0)161 477 1898   FAX: +44 (0)161 718 3587         )=-