From 0c1703f8fd84e8f75b0775e4225766582324eb32 Mon Sep 17 00:00:00 2001
From: uz <uz@b7a2c559-68d2-44c3-8de9-860c34a00d81>
Date: Tue, 3 Nov 2009 13:40:03 +0000
Subject: [PATCH] Added 16*16=32 multiplication routine.

git-svn-id: svn://svn.cc65.org/cc65/trunk@4434 b7a2c559-68d2-44c3-8de9-860c34a00d81
---
 libsrc/runtime/Makefile      |  1 +
 libsrc/runtime/mul16x16r32.s | 49 ++++++++++++++++++++++++++++++++++++
 libsrc/runtime/mul8x8r16.s   |  8 +++---
 3 files changed, 54 insertions(+), 4 deletions(-)
 create mode 100644 libsrc/runtime/mul16x16r32.s

diff --git a/libsrc/runtime/Makefile b/libsrc/runtime/Makefile
index 34052761c..b92863cab 100644
--- a/libsrc/runtime/Makefile
+++ b/libsrc/runtime/Makefile
@@ -152,6 +152,7 @@ OBJS = 	add.o		\
        	makebool.o	\
        	mod.o  		\
        	mul.o  		\
+        mul16x16r32.o   \
         mul8.o          \
         mul8x8r16.o     \
 	mulax3.o	\
diff --git a/libsrc/runtime/mul16x16r32.s b/libsrc/runtime/mul16x16r32.s
new file mode 100644
index 000000000..968903de5
--- /dev/null
+++ b/libsrc/runtime/mul16x16r32.s
@@ -0,0 +1,49 @@
+;
+; Ullrich von Bassewitz, 2010-11-03
+;
+; CC65 runtime: 16x16 => 32 multiplication
+;
+
+        .export         umul16x16r32
+       	.export	       	mul16x16r32 := umul16x16r32
+    	.importzp   	ptr1, ptr2, ptr3, ptr4, sreg
+
+
+;---------------------------------------------------------------------------
+; 16x16 => 32 multiplication routine.
+;
+;   lhs         rhs           result          result also in
+; -------------------------------------------------------------
+;   ptr1        ptr3           ax:sreg          ptr1:sreg
+;
+
+umul16x16r32:
+       	lda	#0
+       	ldx	ptr3+1	       	; Get high byte into register for speed
+       	sta    	sreg+1
+       	ldy    	#16   	       	; Number of bits
+
+        lsr     ptr1+1
+        ror     ptr1            ; Get first bit into carry
+@L0:    bcc     @L1
+
+      	clc
+      	adc	ptr3
+      	pha
+       	txa	      	       	; hi byte of left op
+      	adc	sreg+1
+      	sta	sreg+1
+      	pla
+
+@L1:    ror     sreg+1
+     	ror	a
+     	ror	ptr1+1
+     	ror	ptr1
+        dey
+        bne     @L0
+
+        sta     sreg            ; Save byte 3
+      	lda	ptr1	       	; Load the result
+      	ldx	ptr1+1
+      	rts	    		; Done
+
diff --git a/libsrc/runtime/mul8x8r16.s b/libsrc/runtime/mul8x8r16.s
index b1e679f69..a09618124 100644
--- a/libsrc/runtime/mul8x8r16.s
+++ b/libsrc/runtime/mul8x8r16.s
@@ -4,7 +4,8 @@
 ; CC65 runtime: 8x8 => 16 multiplication
 ;
 
-       	.export		mul8x8r16, umul8x8r16
+       	.export		umul8x8r16
+        .export         mul8x8r16 := umul8x8r16
     	.importzp	ptr1, ptr3
 
 
@@ -13,10 +14,9 @@
 ;
 ;   lhs         rhs           result          result also in
 ; -------------------------------------------------------------
-;   ptr1-lo     ptr3-lo         a/x             ptr1
+;   ptr1-lo     ptr3-lo         ax              ptr1
 ;
 
-mul8x8r16:
 umul8x8r16:
     	lda   	#0              ; Clear byte 1
        	ldy    	#8    	       	; Number of bits
@@ -28,7 +28,7 @@ umul8x8r16:
         ror     ptr1
         dey
         bne     @L0
-        tax         
+        tax
         stx     ptr1+1          ; Result in a/x and ptr1
     	lda	ptr1  	       	; Load the result
     	rts	   		; Done