--- /dev/null
+config FAAD2_ALWAYS_USE_FLOATING_POINT
+ depends on PACKAGE_libfaad2
+ bool "Build FAAD2 to use floating-point math even if using soft floats."
+ default n
+ help
+ By default, libfaad2 will be built to use fixed point rather
+ than floating point math if emulated floating point operations
+ are being used. (See the CONFIG_SOFT_FLOAT option.) Set this
+ option if you would like to use floating point math regardless
+ of whether floating point emulation is in use.
+
+ For the best real-time decoding ability, leave this option off.