--- researchv9/jerq/src/lib/j/bitblt.c	2018/04/24 17:21:59	1.1
+++ researchv9/jerq/src/lib/j/bitblt.c	2018/04/24 17:32:48	1.1.1.2
@@ -18,7 +18,9 @@
 #undef	sw
 #define DX1	m
 
+#undef bitblt
 bitblt(sm,r,dm,p,fc)
+#define bitblt Sbitblt
 Bitmap *sm,*dm;
 Rectangle r;
 Point p;
@@ -34,13 +36,19 @@ int fc;
 	int dummy1, dummy2;		/* for future use, CANNOT CHANGE */
 					/* NUMBER OF LOCALS DUE TO %fp 	*/
 					/* TRICKS			*/
+
+
 	/* clip to the source Bitmap */
-	if(r.origin.x < sm->rect.origin.x)
+	if(r.origin.x < sm->rect.origin.x){
+		p.x+=sm->rect.origin.x-r.origin.x;
 		r.origin.x=sm->rect.origin.x;
+	}
 	if(r.corner.x > sm->rect.corner.x)
 		r.corner.x=sm->rect.corner.x;
-	if(r.origin.y < sm->rect.origin.y)
+	if(r.origin.y < sm->rect.origin.y){
+		p.y+=sm->rect.origin.y-r.origin.y;
 		r.origin.y=sm->rect.origin.y;
+	}
 	if(r.corner.y > sm->rect.corner.y)
 		r.corner.y=sm->rect.corner.y;
 	/*
@@ -73,7 +81,12 @@ int fc;
 	mask1 = ~ntmask1;
 	mask2 = topbits[((p.x+DX1) & 31) + 1];
 	ntmask2 = ~mask2;
-	w = ((p.x+DX1) >> 5) - (p.x >> 5);	/* inner loop+1: sub 1 later*/
+	/*
+		the following code is wrong.
+		the moral is, do it right
+	*/
+	/*w = ((p.x+DX1) >> 5) - (p.x >> 5);	/* inner loop+1: sub 1 later*/
+	w = (((p.x+DX1)&~31)/32) - ((p.x&~31)/32);	/* inner loop+1: sub 1 later*/
 	sw = ((sm->width - w) << 2);	/* sleazy hack to avoid shift */
 	dw = ((dm->width - w) << 2);	/* in outer, inner loops */
 	if (sm == dm) {		/* may have to mess with loop order */
@@ -434,54 +447,6 @@ int fc;
 		} while (--h > 0);
 		break;
 	case F_STORE | NOSHIFT:
-		if (w > 25)
-			goto widestore;
-		i = 25 - w;
-		m = (i << 2) + (i << 1);
-		
-		asm(" LLSW3 &0x2,0x10(%fp),%r0");	/* bytewidth in r0*/
-		asm(" ADDW2 &4,%r0");	/* r0 +=4 for loop offsets */
-		asm(" MOVAW B_FS_N,%r1");
-		asm(" ADDW2 %r4,%r1");	/* r1 += &label, or jump addr*/
-		m = h;		/* m is free reg, use it as height */
-		asm(" MOVW 0x20(%fp),%r2");	/* r2 = mask1 */
-		asm(" PUSHW %ap");	/* save fp and ap */
-		asm(" MOVW 0x24(%fp),%ap");	/* ap = mask2 */
-		asm(" PUSHW %fp");
-		asm(" MOVW %r1,%fp");	/* put jump index into fp */
-
-		do{
-
-		asm(" XORW3 0(%r7),0(%r8),%r1");
-		asm(" ANDW2 %r2,%r1");
-		asm(" XORW2 %r1,0(%r7)");
-
-		asm(" JMP     0(%fp)");
-		asm("B_FS_N: ");		/* label */
-		asm(" MOVW    0x64(%r8),0x64(%r7)");
-		asm(" MOVW    0x60(%r8),0x60(%r7)"); asm(" MOVW    0x5c(%r8),0x5c(%r7)");
-		asm(" MOVW    0x58(%r8),0x58(%r7)"); asm(" MOVW    0x54(%r8),0x54(%r7)");
-		asm(" MOVW    0x50(%r8),0x50(%r7)"); asm(" MOVW    0x4c(%r8),0x4c(%r7)"); asm(" MOVW    0x48(%r8),0x48(%r7)"); asm(" MOVW    0x44(%r8),0x44(%r7)");
-		asm(" MOVW    0x40(%r8),0x40(%r7)"); asm(" MOVW    0x3c(%r8),0x3c(%r7)"); asm(" MOVW    0x38(%r8),0x38(%r7)"); asm(" MOVW    0x34(%r8),0x34(%r7)");
-		asm(" MOVW    0x30(%r8),0x30(%r7)"); asm(" MOVW    0x2c(%r8),0x2c(%r7)"); asm(" MOVW    0x28(%r8),0x28(%r7)"); asm(" MOVW    0x24(%r8),0x24(%r7)");
-		asm(" MOVW    0x20(%r8),0x20(%r7)"); asm(" MOVW    0x1c(%r8),0x1c(%r7)"); asm(" MOVW    0x18(%r8),0x18(%r7)");
-		asm(" MOVW    0x14(%r8),0x14(%r7)");
-		asm(" MOVW    0x10(%r8),0x10(%r7)");
-		asm(" MOVW    0xc(%r8),0xc(%r7)");
-		asm(" MOVW    0x8(%r8),0x8(%r7)");
-		asm(" MOVW    0x4(%r8),0x4(%r7)");
-		asm(" ADDW2   %r0,%r8");
-		asm(" ADDW2   %r0,%r7");
-		asm(" XORW3 0(%r7),0(%r8),%r1");
-		asm(" ANDW2 %ap,%r1");	/* ap has mask2 */
-		asm(" XORW2 %r1,0(%r7)");
-		asm(" ADDW2   %r6,%r8");
-		asm(" ADDW2   %r5,%r7");
-	} while (--m != 0);
-		asm(" POPW %fp");
-		asm(" POPW %ap");
-	break;
-	/* ABOVE ALMOST EQUAL TO : */
 widestore:
 		b = w >> 2;
 		w = w & 3;