resampler: Precompute some common filter coefficients

While this is quite fast on x86 (order of a few microseconds), the
computation can take a few milliseconds on ARM (measured at 1.9ms (32000
-> 48000) and 3.3ms (32000 -> 44100) on a Cortex A53).

Let's precompute some common rates so that we can avoid this overhead on
each stream (or any other audioconvert) instantiation. The approach
taken here is to write a little program to create the resampler
instance, and run that on the host at compile-time to generate some
common rate conversions.

meson.build

@@ -50,6 +50,7 @@ pipewire_confdatadir = pipewire_datadir / 'pipewire'
 modules_install_dir = pipewire_libdir / pipewire_name
 
 cc = meson.get_compiler('c')
+cc_native = meson.get_compiler('c', native: true)
 
 if cc.has_header('features.h') and cc.get_define('__GLIBC__', prefix: '#include <features.h>') != ''
   # glibc ld.so interprets ${LIB} in a library loading path with an
@@ -112,6 +113,8 @@ cc_flags = common_flags + [
 ]
 add_project_arguments(cc.get_supported_arguments(cc_flags), language: 'c')
 
+cc_flags_native = cc_native.get_supported_arguments(cc_flags)
+
 have_cpp = add_languages('cpp', native: false, required : false)
 
 if have_cpp
@@ -276,6 +279,7 @@ configure_file(input : 'Makefile.in',
 
 # Find dependencies
 mathlib = cc.find_library('m', required : false)
+mathlib_native = cc_native.find_library('m', required : false)
 rt_lib = cc.find_library('rt', required : false) # clock_gettime
 dl_lib = cc.find_library('dl', required : false)
 pthread_lib = dependency('threads')