--- qpel.h	2004/08/10 21:58:55	1.3
+++ qpel.h	2005/10/26 12:38:34	1.7
@@ -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: qpel.h,v 1.3 2004/08/10 21:58:55 edgomez Exp $
+ * $Id: qpel.h,v 1.7 2005/10/26 12:38:34 Skal Exp $
  *
  ****************************************************************************/
 
@@ -78,6 +78,16 @@
 extern XVID_QP_FUNCS xvid_QP_Add_Funcs_mmx;
 #endif
 
+#ifdef ARCH_IS_PPC
+extern XVID_QP_FUNCS xvid_QP_Funcs_Altivec_C;
+extern XVID_QP_FUNCS xvid_QP_Add_Funcs_Altivec_C;
+#endif
+
+#ifdef ARCH_IS_X86_64
+extern XVID_QP_FUNCS xvid_QP_Funcs_x86_64;
+extern XVID_QP_FUNCS xvid_QP_Add_Funcs_x86_64;
+#endif
+
 extern XVID_QP_FUNCS *xvid_QP_Funcs;      /* <- main pointer for enc/dec structure */
 extern XVID_QP_FUNCS *xvid_QP_Add_Funcs;  /* <- main pointer for enc/dec structure */
 
@@ -120,22 +130,17 @@
 
 	int32_t x_int, y_int;
 
-	const int32_t xRef = x*4 + dx;
-	const int32_t yRef = y*4 + dy;
+	const int32_t xRef = (int)x*4 + dx;
+	const int32_t yRef = (int)y*4 + dy;
 
 	Ops = xvid_QP_Funcs;
 	quads = (dx&3) | ((dy&3)<<2);
 
-	x_int = xRef/4;
-	if (xRef < 0 && xRef % 4)
-		x_int--;
-
-	y_int	 = yRef/4;
-	if (yRef < 0 && yRef % 4)
-		y_int--;
+	x_int = xRef >> 2;
+	y_int = yRef >> 2;
 
 	dst = cur + y * stride + x;
-	src = refn + y_int * stride + x_int;
+	src = refn + y_int * (int)stride + x_int;
 
 	tmp = refh; /* we need at least a 16 x stride scratch block */
 
@@ -223,23 +228,18 @@
 
 	int32_t x_int, y_int;
 
-	const int32_t xRef = x*4 + dx;
-	const int32_t yRef = y*4 + dy;
+	const int32_t xRef = (int)x*4 + dx;
+	const int32_t yRef = (int)y*4 + dy;
 
 	Ops = xvid_QP_Add_Funcs;
 	Ops_Copy = xvid_QP_Funcs;
 	quads = (dx&3) | ((dy&3)<<2);
 
-	x_int = xRef/4;
-	if (xRef < 0 && xRef % 4)
-		x_int--;
-
-	y_int	 = yRef/4;
-	if (yRef < 0 && yRef % 4)
-		y_int--;
+	x_int = xRef >> 2;
+	y_int = yRef >> 2;
 
 	dst = cur + y * stride + x;
-	src = refn + y_int * stride + x_int;
+	src = refn + y_int * (int)stride + x_int;
 
 	tmp = refh; /* we need at least a 16 x stride scratch block */
 
@@ -328,22 +328,17 @@
 
 	int32_t x_int, y_int;
 
-	const int32_t xRef = x*4 + dx;
-	const int32_t yRef = y*4 + dy;
+	const int32_t xRef = (int)x*4 + dx;
+	const int32_t yRef = (int)y*4 + dy;
 
 	Ops = xvid_QP_Funcs;
 	quads = (dx&3) | ((dy&3)<<2);
 
-	x_int = xRef/4;
-	if (xRef < 0 && xRef % 4)
-		x_int--;
-
-	y_int	 = yRef/4;
-	if (yRef < 0 && yRef % 4)
-		y_int--;
+	x_int = xRef >> 2;
+	y_int = yRef >> 2;
 
 	dst = cur + y * stride + x;
-	src = refn + y_int * stride + x_int;
+	src = refn + y_int * (int)stride + x_int;
 
 	tmp = refh; /* we need at least a 16 x stride scratch block */
 
@@ -428,22 +423,17 @@
 
 	int32_t x_int, y_int;
 
-	const int32_t xRef = x*4 + dx;
-	const int32_t yRef = y*4 + dy;
+	const int32_t xRef = (int)x*4 + dx;
+	const int32_t yRef = (int)y*4 + dy;
 
 	Ops = xvid_QP_Funcs;
 	quads = (dx&3) | ((dy&3)<<2);
 
-	x_int = xRef/4;
-	if (xRef < 0 && xRef % 4)
-		x_int--;
-
-	y_int	 = yRef/4;
-	if (yRef < 0 && yRef % 4)
-		y_int--;
+	x_int = xRef >> 2;
+	y_int = yRef >> 2;
 
 	dst = cur + y * stride + x;
-	src = refn + y_int * stride + x_int;
+	src = refn + y_int * (int)stride + x_int;
 
 	tmp = refh; /* we need at least a 16 x stride scratch block */
 
@@ -528,23 +518,18 @@
 
 	int32_t x_int, y_int;
 
-	const int32_t xRef = x*4 + dx;
-	const int32_t yRef = y*4 + dy;
+	const int32_t xRef = (int)x*4 + dx;
+	const int32_t yRef = (int)y*4 + dy;
 
 	Ops = xvid_QP_Add_Funcs;
 	Ops_Copy = xvid_QP_Funcs;
 	quads = (dx&3) | ((dy&3)<<2);
 
-	x_int = xRef/4;
-	if (xRef < 0 && xRef % 4)
-		x_int--;
-
-	y_int	 = yRef/4;
-	if (yRef < 0 && yRef % 4)
-		y_int--;
+	x_int = xRef >> 2;
+	y_int = yRef >> 2;
 
 	dst = cur + y * stride + x;
-	src = refn + y_int * stride + x_int;
+	src = refn + y_int * (int)stride + x_int;
 
 	tmp = refh; /* we need at least a 16 x stride scratch block */