Add an alternative avx2 s32_to_f32d implementation that doesn't use the
gather function for when gather is slow.
Don't overwrite the orinal cpu_flags but store the selected flags in a
new variable. Use this to debug the selected function cpu flags.
Build libraries with defines from previous libraries so that we can
reuse functions from them.
We can then remove the SSE2 | SLOW_GATHER function selection from the
list. We will now select avx2 and it will then switch implementations
based on the CPU flags.
Make the params introspectable, use the names to configure them in
audioconvert.
Only use precomputed filters when using the defaults.
Tweak the kaiser filter alpha and n_taps based on values calculated from
stopband attenuation and transition bandwidth.
Add a resampler option to prefill the resampler with 0. This then
results in the resampler always outputing and consuming the same
amount of data instead of a short buffer in the beginning.
Also emit port info the the merger monitor ports.
Proxy params and buffers to the right monitor ports.
fmtconvert does not always have dsp ports
Increase resampler output buffer sizes so that we can up and downsample.
Fix little off by one in native resampler.
Fix passthrough in speex resampler.
Add a sinc based resampler that, unlike speex, avoids memcpy and
works directly on the source data. It also allows for ssse3
optimizations and aligned loads. It will later switch to table
interpolation when doing variable rate.
