From f34f9329f1069b0dd0ea6947134f4b8474a181e7 Mon Sep 17 00:00:00 2001
From: Irmen de Jong <irmen@razorvine.net>
Date: Sat, 23 Jan 2021 19:49:53 +0100
Subject: [PATCH] fixed bug in memcopy

---
 compiler/res/prog8lib/c64/syslib.p8 | 26 ++++++++++-------
 docs/source/todo.rst                |  1 -
 examples/test.p8                    | 43 +++++++++++++++++------------
 3 files changed, 41 insertions(+), 29 deletions(-)

diff --git a/compiler/res/prog8lib/c64/syslib.p8 b/compiler/res/prog8lib/c64/syslib.p8
index fcc97c4be..9ed05e340 100644
--- a/compiler/res/prog8lib/c64/syslib.p8
+++ b/compiler/res/prog8lib/c64/syslib.p8
@@ -500,19 +500,23 @@ sys {
     asmsub memcopy(uword source @R0, uword target @R1, uword count @AY) clobbers(A,X,Y) {
         %asm {{
             ldx  cx16.r0
-            stx  P8ZP_SCRATCH_W1
+            stx  P8ZP_SCRATCH_W1        ; source in ZP
             ldx  cx16.r0+1
             stx  P8ZP_SCRATCH_W1+1
             ldx  cx16.r1
-            stx  P8ZP_SCRATCH_W2
+            stx  P8ZP_SCRATCH_W2        ; target in ZP
             ldx  cx16.r1+1
             stx  P8ZP_SCRATCH_W2+1
             cpy  #0
             bne  _longcopy
-            ; copy <= 255
-            tay
 
-_remainder
+            ; copy <= 255 bytes
+            tay
+            bne  _copyshort
+            rts     ; nothing to copy
+
+_copyshort
+            ; decrease source and target pointers so we can simply index by Y
             lda  P8ZP_SCRATCH_W1
             bne  +
             dec  P8ZP_SCRATCH_W1+1
@@ -521,18 +525,19 @@ _remainder
             bne  +
             dec  P8ZP_SCRATCH_W2+1
 +           dec  P8ZP_SCRATCH_W2
--           lda  (P8ZP_SCRATCH_W1), y
-            sta  (P8ZP_SCRATCH_W2), y
+
+-           lda  (P8ZP_SCRATCH_W1),y
+            sta  (P8ZP_SCRATCH_W2),y
             dey
             bne  -
             rts
 
 _longcopy
-            sta  P8ZP_SCRATCH_B1        ; lsb(count) = remainder
+            sta  P8ZP_SCRATCH_B1        ; lsb(count) = remainder in last page
             tya
             tax                         ; x = num pages (1+)
             ldy  #0
--           lda  (P8ZP_SCRATCH_W1),y    ; copy a page at a time
+-           lda  (P8ZP_SCRATCH_W1),y
             sta  (P8ZP_SCRATCH_W2),y
             iny
             bne  -
@@ -541,7 +546,8 @@ _longcopy
             dex
             bne  -
             ldy  P8ZP_SCRATCH_B1
-            jmp  _remainder
+            bne  _copyshort
+            rts
         }}
     }
 
diff --git a/docs/source/todo.rst b/docs/source/todo.rst
index 28551b145..fd1e0473f 100644
--- a/docs/source/todo.rst
+++ b/docs/source/todo.rst
@@ -2,7 +2,6 @@
 TODO
 ====
 
-- fix sys.memcopy()
 - allow  uwordpointer[index] syntax -> transform into @(uwordpointer+index)  allow index to be >255!
 - add offsetof() to get the byte offset of struct members
 - add any2(), all2(), max2(), min2(), reverse2(), sum2(), sort2() that take (array, startindex, length) arguments
diff --git a/examples/test.p8 b/examples/test.p8
index cac481529..601a5863d 100644
--- a/examples/test.p8
+++ b/examples/test.p8
@@ -8,26 +8,33 @@ main {
 
 
     sub start() {
-        ubyte[] sarray = [11,22,33]
-        ubyte[] tarray = [0,0,0]
+        ; TODO test memcopy
+        ; counts: 0, 1, 2, 254, 255, 256, 257, 512, 1000
 
-        uword target = &tarray
-        uword source = &sarray
-        ubyte bb
-        @(target) = @(source)
-        target++
-        source++
-        @(target) = @(source)
-        target++
-        source++
-        @(target) = @(source)
-        target++
-        source++
+        uword buffer=memory("buffer",1000)
+        uword ones=memory("ones",1000)
 
-        for bb in tarray {
-            txt.print_ub(bb)
-            txt.chrout('\n')
+        sys.memset(buffer, 1000, '.')
+        @(buffer) = '<'
+        @(buffer+255) = '>'
+        @(buffer+256) = '<'
+        @(buffer+511) = '>'
+        @(buffer+512) = '<'
+        @(buffer+767) = '>'
+        @(buffer+768) = '<'
+        @(buffer+999) = '!'
+        sys.memset(ones, 1000, '*')
+
+        txt.clear_screen()
+        txt.print("\n\n\n\n\n\n\n\n\n")
+
+        sys.memcopy(ones, buffer, 999)
+
+        uword scr = $0400
+        uword ix
+        for ix in 0 to 999 {
+            @(scr) = @(buffer+ix)
+            scr++
         }
     }
-
 }