Double precision accelerator

[ikostosk]

Is there any library or source code example of using the double-precision accelerator instructions (as described here: https://archive.eettaiwan.com/www.eetta ... S=DOWNLOAD) which as I understand are included in ESP32 cores ?

I am trying to run a code (IIR filters) where single-precision just barely doesn't work and the double-precision software emulation is just barely too slow. All I need is double precision addition and multiplication in some critical part of the code. According to the application note above, using these instruction has a promise of at least 2x acceleration, which would be ideal for the moment...