shithub: opus

Info  •  Files  •  Log  •  Branches

Replacing some 32x16 multiplies with 16x16 when both operands are 16 bits

--- a/silk/NLSF_del_dec_quant.c

+++ b/silk/NLSF_del_dec_quant.c

@@ -46,8 +46,9 @@

 )

 {

     opus_int         i, j, nStates, ind_tmp, ind_min_max, ind_max_min, in_Q10, res_Q10;

-    opus_int         pred_Q10, diff_Q10, out0_Q10, out1_Q10, rate0_Q5, rate1_Q5;

-    opus_int32       RD_tmp_Q25, min_Q25, min_max_Q25, max_min_Q25, pred_coef_Q16;

+    opus_int         pred_Q10, diff_Q10, rate0_Q5, rate1_Q5;

+    opus_int16       out0_Q10, out1_Q10;

+    opus_int32       RD_tmp_Q25, min_Q25, min_max_Q25, max_min_Q25;

     opus_int         ind_sort[         NLSF_QUANT_DEL_DEC_STATES ];

     opus_int8        ind[              NLSF_QUANT_DEL_DEC_STATES ][ MAX_LPC_ORDER ];

     opus_int16       prev_out_Q10[ 2 * NLSF_QUANT_DEL_DEC_STATES ];

@@ -74,8 +75,8 @@

             out0_Q10 = silk_ADD16( out0_Q10, SILK_FIX_CONST( NLSF_QUANT_LEVEL_ADJ, 10 ) );

             out1_Q10 = silk_ADD16( out1_Q10, SILK_FIX_CONST( NLSF_QUANT_LEVEL_ADJ, 10 ) );

         }

-        out0_Q10_table[ i + NLSF_QUANT_MAX_AMPLITUDE_EXT ] = silk_SMULWB( (opus_int32)out0_Q10, quant_step_size_Q16 );

-        out1_Q10_table[ i + NLSF_QUANT_MAX_AMPLITUDE_EXT ] = silk_SMULWB( (opus_int32)out1_Q10, quant_step_size_Q16 );

+        out0_Q10_table[ i + NLSF_QUANT_MAX_AMPLITUDE_EXT ] = silk_RSHIFT( silk_SMULBB( out0_Q10, quant_step_size_Q16 ), 16 );

+        out1_Q10_table[ i + NLSF_QUANT_MAX_AMPLITUDE_EXT ] = silk_RSHIFT( silk_SMULBB( out1_Q10, quant_step_size_Q16 ), 16 );

     }

     silk_assert( (NLSF_QUANT_DEL_DEC_STATES & (NLSF_QUANT_DEL_DEC_STATES-1)) == 0 );     /* must be power of two */

@@ -85,12 +86,11 @@

     prev_out_Q10[ 0 ] = 0;

     for( i = order - 1; ; i-- ) {

         rates_Q5 = &ec_rates_Q5[ ec_ix[ i ] ];

-        pred_coef_Q16 = silk_LSHIFT( (opus_int32)pred_coef_Q8[ i ], 8 );

         in_Q10 = x_Q10[ i ];

         for( j = 0; j < nStates; j++ ) {

-            pred_Q10 = silk_SMULWB( pred_coef_Q16, prev_out_Q10[ j ] );

+            pred_Q10 = silk_RSHIFT( silk_SMULBB( (opus_int16)pred_coef_Q8[ i ], prev_out_Q10[ j ] ), 8 );

             res_Q10  = silk_SUB16( in_Q10, pred_Q10 );

-            ind_tmp  = silk_SMULWB( (opus_int32)inv_quant_step_size_Q6, res_Q10 );

+            ind_tmp  = silk_RSHIFT( silk_SMULBB( inv_quant_step_size_Q6, res_Q10 ), 16 );

             ind_tmp  = silk_LIMIT( ind_tmp, -NLSF_QUANT_MAX_AMPLITUDE_EXT, NLSF_QUANT_MAX_AMPLITUDE_EXT-1 );

             ind[ j ][ i ] = (opus_int8)ind_tmp;