--- xvid.c	2004/04/05 20:36:36	1.50
+++ xvid.c	2004/08/10 21:58:55	1.55
@@ -19,7 +19,7 @@
  *  along with this program ; if not, write to the Free Software
  *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  *
- * $Id: xvid.c,v 1.50 2004/04/05 20:36:36 edgomez Exp $
+ * $Id: xvid.c,v 1.55 2004/08/10 21:58:55 edgomez Exp $
  *
  ****************************************************************************/
 
@@ -222,6 +222,11 @@
 	interpolate8x8_halfpel_v  = interpolate8x8_halfpel_v_c;
 	interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_c;
 
+	interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_c;
+	interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_c;
+	interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_c;
+	interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_c;
+
 	interpolate16x16_lowpass_h = interpolate16x16_lowpass_h_c;
 	interpolate16x16_lowpass_v = interpolate16x16_lowpass_v_c;
 	interpolate16x16_lowpass_hv = interpolate16x16_lowpass_hv_c;
@@ -302,6 +307,7 @@
 	dev16      = dev16_c;
 	sad16v	   = sad16v_c;
 	sse8_16bit = sse8_16bit_c;
+	sse8_8bit  = sse8_8bit_c;
 
 #if defined(ARCH_IS_IA32)
 
@@ -356,6 +362,11 @@
 		interpolate8x8_halfpel_v  = interpolate8x8_halfpel_v_mmx;
 		interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_mmx;
 
+		interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_mmx;
+		interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_mmx;
+		interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_mmx;
+		interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_mmx;
+
 		interpolate8x8_6tap_lowpass_h = interpolate8x8_6tap_lowpass_h_mmx;
 		interpolate8x8_6tap_lowpass_v = interpolate8x8_6tap_lowpass_v_mmx;
 
@@ -389,14 +400,15 @@
 		yv12_to_uyvyi = yv12_to_uyvyi_mmx;
 
 		/* Motion estimation related functions */
-		calc_cbp = calc_cbp_mmx;
-		sad16    = sad16_mmx;
-		sad8     = sad8_mmx;
-		sad16bi = sad16bi_mmx;
-		sad8bi  = sad8bi_mmx;
-		dev16    = dev16_mmx;
-		sad16v	 = sad16v_mmx;
+		calc_cbp   = calc_cbp_mmx;
+		sad16      = sad16_mmx;
+		sad8       = sad8_mmx;
+		sad16bi    = sad16bi_mmx;
+		sad8bi     = sad8bi_mmx;
+		dev16      = dev16_mmx;
+		sad16v	   = sad16v_mmx;
 		sse8_16bit = sse8_16bit_mmx;
+		sse8_8bit  = sse8_8bit_mmx;
 	}
 
 	/* these 3dnow functions are faster than mmx, but slower than xmm. */
@@ -423,6 +435,11 @@
 		interpolate8x8_halfpel_h  = interpolate8x8_halfpel_h_xmm;
 		interpolate8x8_halfpel_v  = interpolate8x8_halfpel_v_xmm;
 		interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_xmm;
+		
+		interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_xmm;
+		interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_xmm;
+		interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_xmm;
+		interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_xmm;
 
 		/* reduced resolution */
 		copy_upsampled_8x8_16to8 = xvid_Copy_Upsampled_8x8_16To8_xmm;
@@ -462,38 +479,43 @@
 
 	if ((cpu_flags & XVID_CPU_3DNOWEXT)) {
 
-		/* Inverse DCT */
-		idct =  idct_3dne;
-
 		/* Buffer transfer */
 		transfer_8to16copy =  transfer_8to16copy_3dne;
 		transfer_16to8copy = transfer_16to8copy_3dne;
 		transfer_8to16sub =  transfer_8to16sub_3dne;
 		transfer_8to16subro =  transfer_8to16subro_3dne;
-		transfer_8to16sub2 =  transfer_8to16sub2_3dne;
 		transfer_16to8add = transfer_16to8add_3dne;
 		transfer8x8_copy = transfer8x8_copy_3dne;
 
-		/* Quantization */
-		quant_h263_intra = quant_h263_intra_3dne;
-		quant_h263_inter = quant_h263_inter_3dne;
-		dequant_mpeg_intra = dequant_mpeg_intra_3dne;
-		dequant_mpeg_inter = dequant_mpeg_inter_3dne;
-		dequant_h263_intra = dequant_h263_intra_3dne;
-		dequant_h263_inter = dequant_h263_inter_3dne;
-
-		/* ME functions */
-		calc_cbp = calc_cbp_3dne;
-		sad16 = sad16_3dne;
-		sad8 = sad8_3dne;
-		sad16bi = sad16bi_3dne;
-		sad8bi = sad8bi_3dne;
-		dev16 = dev16_3dne;
-
-		/* Interpolation */
-		interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_3dne;
-		interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_3dne;
-		interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_3dne;
+		if ((cpu_flags & XVID_CPU_MMXEXT)) {
+			/* Inverse DCT */
+			idct =  idct_3dne;
+
+			/* Buffer transfer */
+			transfer_8to16sub2 =  transfer_8to16sub2_3dne;
+
+			/* Interpolation */
+			interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_3dne;
+			interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_3dne;
+			interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_3dne;
+
+			/* Quantization */
+			quant_h263_intra = quant_h263_intra_3dne;		/* cmov only */
+			quant_h263_inter = quant_h263_inter_3dne;
+			dequant_mpeg_intra = dequant_mpeg_intra_3dne;	/* cmov only */
+			dequant_mpeg_inter = dequant_mpeg_inter_3dne;
+			dequant_h263_intra = dequant_h263_intra_3dne;
+			dequant_h263_inter = dequant_h263_inter_3dne;
+
+			/* ME functions */
+			calc_cbp = calc_cbp_3dne;
+
+			sad16 = sad16_3dne;
+			sad8 = sad8_3dne;
+			sad16bi = sad16bi_3dne;
+			sad8bi = sad8bi_3dne;
+			dev16 = dev16_3dne;
+		}
 	}
 
 	if ((cpu_flags & XVID_CPU_SSE2)) {
@@ -513,6 +535,9 @@
 		/* DCT operators
 		 * no iDCT because it's not "Walken matching" */
 		fdct = fdct_sse2_skal;
+
+		/* postprocessing */
+		image_brightness = image_brightness_sse2;
 	}
 #endif /* ARCH_IS_IA32 */
 
@@ -609,7 +634,7 @@
 		return XVID_ERR_VERSION;
 
 	info->actual_version = XVID_VERSION;
-	info->build = "xvid-1.0.0";
+	info->build = "xvid-1.1-cvshead";
 	info->cpu_flags = detect_cpu_flags();
 
 #if defined(_SMP) && defined(WIN32)