[non-BE] Resolve "Wrong decimation index calculation in ivas_lfe_tdplc_fx"

• Related issues: #2630 (closed)
• Requested reviewers: @malenovskyvl

Reason why this change is needed

• In ivas_lfe_tdplc(), the float refrence calculation is as follows:

rec_frame_us[i] = rec_frame_us[( i * 48000 ) / output_Fs];

But BASOP:

idx = BASOP_Util_Divide3232_Scale( Mpy_32_32( L_shl( i, 16 ), L_shl( 44100, 15 ) ), output_Fs, &temp_q );
idx = shr( idx, add( sub( 15, temp_q ), 1 ) );

This computes idx = i * 22050 / output_Fs instead of i * 48000 / output_Fs.

Description of the change

• decimation index uses 48000 (was 44100) and the extra >>1 is dropped.

Affected operating points

• non-BE