--- trunk/xvidcore/src/image/ia64_asm/interpolate8x8_ia64.s	2002/07/11 00:15:59	290
+++ trunk/xvidcore/src/image/ia64_asm/interpolate8x8_ia64.s	2002/07/11 14:03:39	291
@@ -25,23 +25,24 @@
 	and r14 = -8,r33			// align src
 	mov r15 = r32			// get dest
 	mov r16 = r34			// stride
-	sub r17 = 1,r35			// 1-rounding
+//	sub r17 = 0,r0			// 1-rounding
+		
 	;;
 
 	add r18 = 8,r14			
-	mux1 r17 = r17, @brcst		// broadcast 1-rounding
+//	mux1 r17 = r17, @brcst		// broadcast 1-rounding
 
 	sub r24 = 64,r22			// lshift of src
 	add r26 = 8,r22			// rshift of src+1
 	sub r27 = 56,r22			// lshift of src+1
 
 	mov ar.lc = 7						// loopcounter
-	mov ar.ec = LL + SL +OL + AVL + AL + STL		// sum of latencies
+	mov ar.ec = LL + SL +OL + AVL  + STL		// sum of latencies
 	mov pr.rot = 1 << 16					// init pr regs for sw-pipeling
 
 	;;
-	.rotr ald1[LL+1],ald2[LL+1],shru1[SL+1],shl1[SL+1],shru2[SL+1],shl2[SL+1],or1[OL+1],or2[OL+1+AL],add1[AL+1],avg[AVL+1]
-	.rotp aldp[LL], sh1p[SL], or1p[OL], addp[AL], pavg1p[AVL],stp[STL]
+	.rotr ald1[LL+1],ald2[LL+1],shru1[SL+1],shl1[SL+1],shru2[SL+1],shl2[SL+1],or1[OL+1],or2[OL+1+AL],avg[AVL+1]
+	.rotp aldp[LL], sh1p[SL], or1p[OL], pavg1p[AVL],stp[STL]
 
 
 .Lloop_interpolate:
@@ -56,9 +57,9 @@
 	(or1p[0]) or or1[0] = shru1[SL],shl2[SL]		// merge things
 	(or1p[0]) or or2[0] = shru2[SL],shl1[SL]
 
-	(addp[0]) padd1.uus add1[0] = or1[OL],r17		// add 1-rounding
+//	(addp[0]) padd1.uus add1[0] = or1[OL],r17		// add 1-rounding
 
-	(pavg1p[0]) pavg1 avg[0] = add1[AL],or2[OL+AL]	// parallel average
+	(pavg1p[0]) pavg1 avg[0] = or1[OL],or2[OL]	// parallel average
 
 	 (stp[0]) st8 [r15] = avg[AVL]			// store results
 	 (stp[0]) add r15 = r15,r16
@@ -96,24 +97,24 @@
 	and r14 = -8,r33
 	mov r15 = r32
 	mov r16 = r34
-	sub r17 = 1,r35
+//	sub r17 = 0,r0
 	;;
 
 	add r18 = 8,r14
 	add r19 = r14,r16			// src + stride 
-	mux1 r17 = r17, @brcst
+//	mux1 r17 = r17, @brcst
 
 	sub r24 = 64,r22
 	;;	
 	add r26 = 8,r19			// src + stride + 8
 
 	mov ar.lc = 7
-	mov ar.ec = LL + SL +OL + AVL + AL + STL
+	mov ar.ec = LL + SL +OL + AVL + STL
 	mov pr.rot = 1 << 16
 
 	;;
-	.rotr ald1[LL+1],ald2[LL+1],ald3[LL+1],ald4[LL+1],shru1[SL+1],shl1[SL+1],shru2[SL+1],shl2[SL+1],or1[OL+1],or2[OL+1+AL],add1[AL+1],avg[AVL+1]
-	.rotp aldp[LL], sh1p[SL], or1p[OL], addp[AL], pavg1p[AVL],stp[STL]
+	.rotr ald1[LL+1],ald2[LL+1],ald3[LL+1],ald4[LL+1],shru1[SL+1],shl1[SL+1],shru2[SL+1],shl2[SL+1],or1[OL+1],or2[OL+1+AL],avg[AVL+1]
+	.rotp aldp[LL], sh1p[SL], or1p[OL], pavg1p[AVL],stp[STL]
 
 
 .Lloop_interpolate2:
@@ -130,9 +131,9 @@
 	(or1p[0]) or or1[0] = shru1[SL],shl1[SL]
 	(or1p[0]) or or2[0] = shru2[SL],shl2[SL]
 
-	(addp[0]) padd1.uus add1[0] = or1[OL],r17
+//	(addp[0]) padd1.uus add1[0] = or1[OL],r17
 
-	(pavg1p[0]) pavg1 avg[0] = add1[AL],or2[OL+AL]
+	(pavg1p[0]) pavg1 avg[0] = or1[OL],or2[OL]
 
 	 (stp[0]) st8 [r15] = avg[AVL]
 	 (stp[0]) add r15 = r15,r16
@@ -170,12 +171,12 @@
 	and r14 = -8,r33
 	mov r15 = r32
 	mov r16 = r34
-	sub r17 = 1,r35
+//	sub r17 = 0,r0
 	;;
 
 	add r18 = 8,r14
 	add r19 = r14,r16
-	mux1 r17 = r17, @brcst
+//	mux1 r17 = r17, @brcst
 
 	add r27 = 8,r22
 	sub r28 = 56,r22
@@ -184,12 +185,12 @@
 	add r26 = 8,r19
 
 	mov ar.lc = 7
-	mov ar.ec = LL + SL +OL + 2*AVL + AL + STL
+	mov ar.ec = LL + SL +OL + 2*AVL  + STL
 	mov pr.rot = 1 << 16
 
 	;;
-	.rotr ald1[LL+1],ald2[LL+1],ald3[LL+1],ald4[LL+1],shru1[SL+1],shl1[SL+1],shru2[SL+1],shl2[SL+1],shl3[SL+1],shru3[SL+1],shl4[SL+1],shru4[SL+1],or1[OL+1],or2[OL+1+AL],or3[OL+AL+1],or4[OL+AL+1],add1[AL+1],avg[AVL+1],avg1[AVL+1],avg2[AVL+1]
-	.rotp aldp[LL], sh1p[SL], or1p[OL], addp[AL],pavg1p[AVL],pavg2p[AVL],stp[STL]
+	.rotr ald1[LL+1],ald2[LL+1],ald3[LL+1],ald4[LL+1],shru1[SL+1],shl1[SL+1],shru2[SL+1],shl2[SL+1],shl3[SL+1],shru3[SL+1],shl4[SL+1],shru4[SL+1],or1[OL+1],or2[OL+1+AL],or3[OL+AL+1],or4[OL+AL+1],avg[AVL+1],avg1[AVL+1],avg2[AVL+1]
+	.rotp aldp[LL], sh1p[SL], or1p[OL],pavg1p[AVL],pavg2p[AVL],stp[STL]
 
 
 .Lloop_interpolate3:
@@ -213,10 +214,10 @@
 	(or1p[0]) or or3[0] = shru3[SL],shl3[SL]
 	(or1p[0]) or or4[0] = shru4[SL],shl4[SL]
 
-	(addp[0]) padd1.uus add1[0] = or1[OL],r17
+//	(addp[0]) padd1.uus add1[0] = or1[OL],r17
 
-	(pavg1p[0]) pavg1 avg[0] = add1[AL],or2[OL+AL]
-	(pavg1p[0]) pavg1 avg1[0] = or3[OL+AL],or4[OL+AL]
+	(pavg1p[0]) pavg1 avg[0] = or1[OL],or2[OL]
+	(pavg1p[0]) pavg1 avg1[0] = or3[OL],or4[OL]
 	
 	(pavg2p[0]) pavg1 avg2[0] = avg[AVL],avg1[AVL]