6502/COMMON
1
0
Fork 0
mirror of https://github.com/Russell-S-Harper/COMMON.git synced 2024-11-24 08:30:57 +00:00
Code Issues Projects Releases Wiki Activity
Advances some ideas using Steve Wozniak's 6502 SWEET16 interpreted byte-code language as inspiration.
34 Commits 2 Branches 0 Tags 212 KiB
C 59.8%
Assembly 27.9%
Yacc 5.2%
C++ 4.5%
Makefile 1.6%
Other 1%
8bbb5016cd
Go to file
Russell-S-Harper 8bbb5016cd Was using up too many S registers, so moved internal functions to use new registers I0 - I7, added IEEE rounding to MUL and DIV, refactored to reduce size, and trimmed trailing spaces.
2018-08-21 09:14:16 -04:00
common Was using up too many S registers, so moved internal functions to use new registers I0 - I7, added IEEE rounding to MUL and DIV, refactored to reduce size, and trimmed trailing spaces. 2018-08-21 09:14:16 -04:00
emulator Was using up too many S registers, so moved internal functions to use new registers I0 - I7, added IEEE rounding to MUL and DIV, refactored to reduce size, and trimmed trailing spaces. 2018-08-21 09:14:16 -04:00
xa-pre-process Was using up too many S registers, so moved internal functions to use new registers I0 - I7, added IEEE rounding to MUL and DIV, refactored to reduce size, and trimmed trailing spaces. 2018-08-21 09:14:16 -04:00
LICENSE Initial commit 2018-08-06 21:55:07 -04:00
Makefile Renamed common-post-process to xa-pre-process. 2018-08-07 07:03:21 -04:00
README.md Revising README. 2018-08-15 08:02:31 -04:00

README.md

COMMON

Advances some ideas using Steve Wozniak’s 6502 SWEET16 interpreted byte-code language as inspiration. While the goal of SWEET16 was brevity, the goal of COMMON is functionality. The intent is to make a platform suitable for many commercial, scientific, and engineering applications.

For example:

  • native type is equivalent to fixed-point decimal ±######.###
  • easier support for banked memory
  • easier support for higher language compilers
  • arithmetic operations add, subtract, multiply, divide, and modulus
  • inherent overflow/underflow detection
  • all control branching is 16-bit relative, for easier relocatable code
  • support for custom system/user functions, akin to INT in x86

Why 6502 and not, for example, x86?

  • 6502 assembler is very easy and has a large archive of existing functions
  • existing 6502 SWEET16 already has the “hard work” done
  • interesting to see it run in newer versions of 6502 processors
  • how do you think Bender does what he does? (or the Terminator!)

In progress:

  • add all the instructions (see common/common.h for the list)
  • a simple unit test suite to ensure each instruction is correct

The meat of the project:

  • common/common.h: details of instructions
  • common/common.asm: assembler code for the instructions
  • common/macros.h: macros used to define the interpreted byte-code
  • common/page6.src: sample source file using the macros

Auxiliary:

  • emulator/*: 6502 emulator (borrowed Mike Chambers’ Fake6502 CPU emulator v1.1 ©2011)
  • xa-pre-process/*: utility xapp to convert 32-bit fixed decimal quantities so that xa can use them

Right now, for testing purposes, the code builds everything into one file system.obj and runs the code in the last block loaded, in this case, the code corresponding to page6.src. Eventually will support decoupling of system and application files. Application files will be inherently relocatable.

To build and run:

make all
make run

The makefiles use re2c, flex, bison, gcc, cpp, and xa. Will eventually provide a ./configure.