[non-BE] Bug in calculating parcor coefficient in Calc_rc0_h()

This is a follow up of !2851 (comment 111231).

There seems to be missing a scaling of the first LP filter response coefficient h[0] in function Calc_rc0_h. See the comment at https://forge.3gpp.org/rep/sa4/audio/ivas-basop/-/blob/main/lib_com/swb_tbe_com_fx.c#L660:

    /* computation of the autocorrelation function acf */
    IF( element_mode != EVS_MONO && enc_dec == ENC )
    {
        L_acc = L_mult( h[0], h[0] ); // TBV: shouldn't it be "shr( h[0], 2 )" ?
        FOR( i = 1; i < LONG_H_ST; i++ )
        {
            tmp2 = shr( h[i], 2 );
            L_acc = L_mac( L_acc, tmp2, tmp2 );
        }
        sh_acf = norm_l( L_acc );
        L_acc = L_shl( L_acc, sh_acf );
        acf0 = extract_h( L_acc );