added assembly docs

docs/assembly.md

@@ -0,0 +1,23 @@
+# Assembly
+
+Assembly tokenization should be workiung completely. It knows what instructions the CPU implementation supports and parses based on them.
+
+## Pseudo-ops
+The following pseudo-ops are implemented as of yet:
+* `.space <len>` reverse <len> bytes of zero
+* `.ascii 'text'` put text into memory
+* `.asciiz 'text'` put text into memory (null terminated)
+* `.sextion .<name>` same as `.<name>` see sections:
+
+## Sections: 
+Currently only these three sections are supported: 
+* `data` read-write data (non-executable)
+* `rodata` read-only data (non-executable)
+* `.text` executable data (read-only)
+
+## Allocating stack
+another pseudo-op is recognized: `.stack <len>`. This marks the executable as requesting at least `<len>` bytes of stack. 
+If the loader repsects this wish, the sp is initialized pointing to the end of the stack.
+
+ 
+