Added division routine. Removed C callable entry points from mul routines,

they're now in common.


git-svn-id: svn://svn.cc65.org/cc65/trunk@4441 b7a2c559-68d2-44c3-8de9-860c34a00d81

libsrc/runtime/Makefile

@@ -218,6 +218,7 @@ OBJS = 	add.o		\
        	ugt.o		\
        	ule.o		\
        	ult.o		\
+        udiv32by16r16.o \
        	umod.o		\
         umul16x16r32.o  \
        	xor.o           \

libsrc/runtime/imul16x16r32.s

@@ -4,8 +4,8 @@
 ; CC65 runtime: 16x16 => 32 signed multiplication
 ;
 
-        .export         imul16x16r32, _cc65_imul16x16r32
-        .import         popax, negax, umul16x16r32m, negeax
+        .export         imul16x16r32
+        .import         negax, umul16x16r32m, negeax
     	.importzp     	ptr1, ptr3, tmp1
 
 
@@ -19,12 +19,6 @@
 ; There is probably a faster way to do this.
 ;
 
-
-_cc65_imul16x16r32:
-        sta     ptr1
-        stx     ptr1+1
-        jsr     popax
-
 imul16x16r32:
         stx     tmp1
         cpx     #0

libsrc/runtime/udiv32by16r16.s

@@ -0,0 +1,55 @@
+;
+; Ullrich von Bassewitz, 2009-11-04
+;
+; CC65 runtime: 32by16 => 16 unsigned division
+;
+
+       	.export	       	udiv32by16r16, udiv32by16r16m
+ 	.importzp   	ptr1, ptr2, ptr3, sreg
+
+
+;---------------------------------------------------------------------------
+; 32by16 division. Divide ptr1:ptr2 by ptr3. Result is in ptr1, remainder
+; in sreg.
+;
+;   lhs         rhs           result      result also in    remainder
+; -----------------------------------------------------------------------
+;   ptr1:ptr2   ptr3          ax          ptr1              sreg
+;
+
+
+udiv32by16r16:
+        sta     ptr3
+        stx     ptr3+1
+udiv32by16r16m:
+        lda    	#0
+ 	sta	sreg+1
+ 	ldy	#32
+
+L0:    	asl     ptr1
+        rol     ptr1+1
+        rol     ptr2
+        rol     ptr2+1
+        rol     a
+        rol     sreg+1
+
+ 	pha
+ 	cmp	ptr3
+ 	lda	sreg+1
+ 	sbc	ptr3+1
+ 	bcc	L1
+
+ 	sta	sreg+1
+ 	pla
+ 	sbc	ptr3
+ 	pha
+ 	inc	ptr1
+
+L1:	pla
+ 	dey
+ 	bne	L0
+ 	sta	sreg
+        lda     ptr1
+        ldx     ptr1+1
+ 	rts
+

libsrc/runtime/umul16x16r32.s

@@ -4,8 +4,7 @@
 ; CC65 runtime: 16x16 => 32 unsigned multiplication
 ;
 
-        .export         _cc65_umul16x16r32, umul16x16r32, umul16x16r32m
-        .import         popax
+        .export         umul16x16r32, umul16x16r32m
     	.importzp      	ptr1, ptr3, sreg
 
 
@@ -17,11 +16,6 @@
 ;   ptr1        ax            ax:sreg          ptr1:sreg
 ;
 
-_cc65_umul16x16r32:
-        sta     ptr1
-        stx     ptr1+1
-        jsr     popax
-
 umul16x16r32:
         sta     ptr3
         stx     ptr3+1