improve WMOPS performance of ivas_dirac_dec_output_synthesis_cov_param_mc_synthesise_slot_fx()

Basic info

Said function is calling a generalized Matrix Multiplication function, which requires the output to be shifted back and forth to match the representation.

By inlining the matrix multiplication, the shifting becomes unnecessary and the WMOPS performance can be increased.

• Float reference:
  • Encoder (float):
  • Decoder (float):
• Fixed point:
  • Encoder (fixed):
  • Decoder (fixed):

Bug description

Ways to reproduce

Box folder: ...\Box_EXTERNAL_IVAS_BASOP_VERIFICATION\issues\issue-xxxx