add skill and some adjustments

attempt_1/duffle.amd64.win32.h

@@ -182,6 +182,24 @@ IA_ U8 align_pow2(U8 x, U8 b) {
 	return ((x + b - 1) & (~(b - 1)));
 }
 
+#if 0
+#pragma clang optimize off
+// TODO(Ed): Replace these later (only matters if CRT is not used)
+void* memset(void* dest, int c, U8 count) {
+    U1* bytes = (U1*)dest;
+    while (count--) *bytes++ = (U1)c;
+    return dest;
+}
+
+void* memcpy(void* dest, const void* src, U8 count) {
+    U1* d = (U1*)dest;
+    const U1* s = (const U1*)src;
+    while (count--) *d++ = *s++;
+    return dest;
+}
+#pragma clang optimize on
+#endif
+
 IA_ U8 mem_copy            (U8 dest, U8 src,   U8 len) { return (U8)(__builtin_memcpy ((void*)dest, (void const*)src,   len)); }
 IA_ U8 mem_copy_overlapping(U8 dest, U8 src,   U8 len) { return (U8)(__builtin_memmove((void*)dest, (void const*)src,   len)); }
 IA_ U8 mem_fill            (U8 dest, U8 value, U8 len) { return (U8)(__builtin_memset ((void*)dest, (int)        value, len)); }