Made bitblt work in 32bpp mode

src/graphics.asm

@@ -17,8 +17,8 @@ include errno
 ;    101 - reserved
 ;    110 - reserved
 ;    111 - SRCALPHA (copies source over destination if: source != 0 (for 8bpp), source.alpha != 0 (for 32bpp))
-; Stack argument 1: Width (in bytes)
-; Stack argument 2: Height (in bytes)
+; Stack argument 1: Width
+; Stack argument 2: Height
 ; Result: None
 ; Clobbers: r1, r2, r3, r4, r5, r6, r7
 ; Errors: BUFFER_DEPTH_MISMATCH - if the buffers have different bit depths
@@ -40,22 +40,29 @@ pub bitblit:
         jmp r13
     bitblit_depth_good:
     push r8
-    add sp, sp, 12
-    load_32 r8, [sp] ; r8 = height
-    add sp, sp, 4
-    load_32 r9, [sp] ; r9 = width
-
-    sub sp, sp, 16
     push r10
     push r11
     push r12
     push r13
+	push r7
+	mov r7, r8
+	lsr r7, r7, 4
 
     add r10, r1, 4 ; r10 = pointer to destination stride
     load_16 r10, [r10] ; r10 = destination stride
-
+	lsl r10, r10, r7
+	
     add r11, r2, 4 ; r11 = pointer to source stride
     load_16 r11, [r11] ; r11 = source stride
+	lsl r11, r11, r7
+	
+	add sp, sp, 32
+    load_32 r8, [sp] ; r8 = height
+    add sp, sp, 4
+    load_32 r9, [sp] ; r9 = width
+    lsl r9, r9, r7
+    sub sp, sp, 36
+
 
     ; Calculating destination index
     push r10
@@ -74,6 +81,7 @@ pub bitblit:
     add r4, r4, r12 ; Now r4 is dest_y * dest_stride
     load_32 r1, [r1] ; r1 = destination data pointer
     add r1, r1, r4 ; Now r1 is dest_pointer + dest_y * dest_stride
+    lsl r3, r3, r7
     add r1, r1, r3 ; Now r1 is dest_pointer + dest_y * dest_stride + dest_x
 	pop r10
 
@@ -94,8 +102,11 @@ pub bitblit:
     add r6, r6, r12 ; Now r6 is src_y * src_stride
     load_32 r2, [r2] ; r2 = source data pointer
     add r2, r2, r6 ; Now r2 is src_pointer + src_y * src_stride
+    lsl r5, r5, r7
     add r2, r2, r5 ; Now r2 is src_pointer + src_y * src_stride + src_x
 	pop r11
+	
+	pop r7
 
     ; r3, r4, r5, r6 are now free
 
@@ -133,8 +144,7 @@ pub bitblit:
             push r2
 	        bitblit_copy_line_loop: ; Main loop that copies 4 bytes at a time
 	        	add r3, r3, 4
-	            cmp r9, r3
-	            je bitblit_copy_line_finish
+	        	cmp r9, r3
 	            jb bitblit_copy_line_end
 	            load_32 r6, [r5] ; r6 = [src]
 	            load_32 r13, [r4] ; r13 = [dest]
@@ -201,6 +211,8 @@ pub bitblit:
                     mov r6, r13 ; If it was zero, we take [dest], i.e. don't overwrite
 	            bitblit_save_bits:
 	                store_32 [r4], r6
+	                cmp r9, r3
+	            	je bitblit_copy_line_finish
 	                add r4, r4, 4
 	                add r5, r5, 4
 	                jmp bitblit_copy_line_loop
@@ -253,4 +265,4 @@ pub bitblit:
         pop r8
         add sp, sp, 8
         mov flags, errno.OK
-        jmp r13
+        jmp r13