Re: iio_simple_dummy_buffer_dma

[Lars-Peter Clausen <lars@metafoo.de>]

On 03/30/2016 08:08 AM, John Syne wrote:
> Hi Lars,
>
> I did manage to get the dummy example with iio_simple_dummy_buffer_dma working. It turns out I had an issue in my dummy Kconfig when I added back the iio_simple_dummy_buffer_dma support. Now I can run iiod and connect iioScope and see sine, square, etc waveforms. I can also use generic_buffer and that works also. 
Great!
>
> Now I’m trying to figure out how to add my own hardware using part 2 of iio_simple_dummy_buffer_dma as a template. I was wondering when you expect to release the “overhauled version”.
Soon... Sorry I don't have a estimated date, at the moment this is not
top priority. I'll get to it as soon as I'm not completely 
>
> Has anyone used the IIO-DMA framework in a real driver? In your IIO-DMA patch notes, you said "Andrey Yurovsky has also successfully added DMA buffer support to the existing AT91 ADC”, but I have looked at this driver and I don’t see any IIO-DMA framework added. 
The iio-buffer-dmaengine driver is in a sense a real driver, but uses
the DMAengine framework to execute the DMA transfers. But you can easily
replace the calls to the DMAengine framework with direct access to the
DMA hardware. The AT91 support never made it upstream and Andrey's
e-mail bounces these days, no idea what happened to the code.
>
> Also, have you made any progress on adding the high speed interface via mmap or vmsplice. 
Nope, not yet. Again, I can't say when I'll get to this.
>
> Thank you and Jonathan for all the work you guys do on IIO. The IIO framework is a really helpful for what I need, especially with the DMA support. 
>
>
> Regards,
> John
>
>
>> On Feb 22, 2016, at 3:44 AM, Lars-Peter Clausen <lars@metafoo.de> wrote:
>>
>> On 02/22/2016 05:51 AM, John Syne wrote:
>>> BTW, I seem to remember that Lars removed iio_simple_dummy_buffer_dma.c in his V3 patch so this won’t be in the IIO repo, but I don’t know how to test the IIO DMA framework without this file.
>>>
>> Hi,
>>
>> The patch was dropped for now to avoid conflicts with other ongoing work on
>> the dummy driver. The patch will come back in a slightly overhauled version
>> soonish.