Programming notes.org 5.2 KB
## Copyright (C) 2016 Jeremiah Orians ## This file is part of stage0. ## ## stage0 is free software: you can redistribute it and/or modify ## it under the terms of the GNU General Public License as published by ## the Free Software Foundation, either version 3 of the License, or ## (at your option) any later version. ## ## stage0 is distributed in the hope that it will be useful, ## but WITHOUT ANY WARRANTY; without even the implied warranty of ## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ## GNU General Public License for more details. ## ## You should have received a copy of the GNU General Public License ## along with stage0. If not, see <http://www.gnu.org/licenses/>.
High level programming information that is likely useful
M0
What is it?
M0 is the most minimal macro-assembler possible to self bootstrap on the Knight ISA specification. In fact it is so minimal that a definition file or individual definitions need to be prepended to source files for it to work. And example of a definitions is the following: DEFINE ADD 05000
What are its assumptions?
- all numbers outside of quotes are to be converted to 4 hex nybbles.
- all strings are to be converted into hex form and start with a " and end with a ", no nesting
- all definitions need to be applied if possible.
- definitions can not be recursive or nested
- all other details about the resulting binary will be handled by a hex2 assembler or better.
- Line comments start with # or ; and continue to the end of the line and are not output
How to use it
Simply create a definition, such as DEFINE ADD 05000 and then use ADD where you need 05000 That is it.
Does it provide any error checking?
No
Seriously?
M0 assumes you write perfect code and never make a single mistake. If you want an macro assembler that helps to catch dumb mistakes, use something else.
Can I use M0 without stage0 vm?
Yes, that is why https://github.com/oriansj/mescc-tools exists which will also provide some basic error checking (duplicate definitions, missing labels, duplicate labels, etc)
Hex2
What is it?
Hex2 is a standard set of rules that are architecture dependent which when combined with an implementation that support it functions like an advanced hex assembler. The goal behind hex2 is that any programmer should be able to implement a compatable implemention within a couple hours in C, or a couple of days in assembly or a couple of weeks if they hand encode it in hex itself. Its design is optimal for instruction sets that are nybble aligned or at worst verbose for instruction sets that are byte aligned.
What are its assumptions?
The most important assumption is that ALL labels are globally unique and should have only 1 instance.
Hex2 only assumes 4 types of inputs:
- Hex nybbles to be assembled
- Line comments beginning with # or ; to be ignored and dropped
- Labels that begin with :
- Pointers to labels that start with @,$ or &
For pointers starting with @, it is assumed that the user wants the 16bit relative displacement required by the preceeding instruction to the label; this is primarily used for jump, calli address immediates and PC relative loads and stores. For pointers starting with $, it is assumed that the user wants the 16bit absolute address of the label; this is primarily used for pointer comparisons involving hand encoded objects. For pointers starting with &, it is assumed that the user wants the 32bit absolute address of the label; this is primarily for hand encoding objects or instructions that want 32bit absolute addresses
Should one wish to manually calculate the displacement field of an instruction, simply subtract the address of the byte after the immediate from the address of the target.
More advanced?
It is possible for a hex2 assembler to support more than just the minimal set as defined above and is frequently the case on complicated instruction sets. However, to be considered hex2 compliant a program should not depend on those additions. However, a hex assembler which supports those extensions can be considered fully compliant as long as hex2 compliant programs can be compiled correctly.
How to use it
In theory, you should never have to use hex2 directly as it is a rather tedious language to program in. Its primary design is to allow extremely low level and auditable bootstrap programs. But should you wish to do so, it'll probably be advisable for it to look something like this: :Store_String_0 05049045 # STOREX8 R0 R4 R5 ; Store the Byte 42100100 # FGETC ; Get next Byte 0F550001 # ADDUI R5 R5 1 ; Prep for next loop C306 @Store_String_0 # CMPJUMPI.NE R0 R6 @Store_String_0 ; Loop if matching not found
With very clear label names, a mixture of low and high level comments. It does become possible to write fixable code. But lets be honest, the only time you need to write hex2 is: When bootstrapping something new/ugly without M0 OR All of stage0 was lost due to a category 5 disaster.
Should you use it?
Only if engaged in low level bootstrapping with the desire to defend against the trusting trust attack in an auditable fashion. Otherwise, just use gcc or llvm and save yourself alot of responsibility.