Closed zuckschwerdt closed 5 years ago
This matches the TX streamlining in #6 for RX.
You are a real pro here. I'm giving you commit access
I got three more patches lined up: TX is MSB first, not LSB like RX; a fast copy optimization if hardware buffers align to user buffers (they do on the default using the Pluto or SoapyRemote); and then DMA on top of that, getting rid of another buffer ;) I'll PR all three in the next days for review and for documentation to the community. But feel free to let those sit, I'll then just merge after a week or so.
the current code copies the RX data to an intermediate buffer just to copy that buffer to the output user buffer again. This change removes the intermediate copy step and folds the two tight copy loops for
iio_channel_convert()
and sample format conversion into one.