summaryrefslogtreecommitdiff
path: root/readme.md
blob: 58c8dc782d05f580a4d99bfdddd2fd2b045c2c12 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
# jewelforth

it's a subroutine threaded forth for linux x86\_64.
mostly made for personal stuff.
it does not conform to any standards, i just implement what i want.
if you use it and it breaks, too bad

public domain

## dictionary
the dictionary follows a fairly standard format.

| field | size | forth name |
| :---- | :--- | :--------- |
| link to previous | 8 bytes | LFA (link field address) |
| flag | 1 byte | FFA (flag field address) |
| string length | 2 bytes | NFA (name field address) |
| string | variable length | still NFA |
| code | variable length | CFA (code field address) |

## the one reserved register

the working stack pointer is `r14`.
the other registers are used as general purpose registers;
`r11` in particular is the standard register used in
compiling calls.

## miscellaneous notes and stuff

### COMPILE,
this is an STC forth so when we compile a call we have to
write the bytes of a `call` in manually.
x86\_64 does not allow absolute jumps from an immediate address,
so a wonky but hopefully not too slow solution is to compile
`literal address → W` and then `call W`. It'll Be Fine?

- `mov r11, [cfa]` = `94 BB [CFA]`
- `call r11` = `41 FF D3`

### Some Links

- starting forth part 1: http://www.bradrodriguez.com/papers/moving1.htm

## silly little plans

### in-forth assembler
this would reap the most benefits from STC.
probably look at dusk's assemblers for how it should look like
or liek something liek dusk's lib/bm?
idk something to do stuff Fast if u need. would be fun