1
0
mirror of https://github.com/hsoft/collapseos.git synced 2024-12-24 14:18:06 +11:00

Add bootstrap guide

This commit is contained in:
Virgil Dupras 2020-05-21 15:25:12 -04:00
parent b5683f447b
commit 0939241db1
10 changed files with 126 additions and 2 deletions

View File

@ -6,7 +6,7 @@ MASTER INDEX
150 Extra words
200 Z80 assembler 260 Cross compilation
280 Z80 boot code 350 Core words
410 PS/2 keyboard subsystem
410 PS/2 keyboard subsystem 420 Bootstrap guide
490 TRS-80 Recipe 520 Fonts
550 TI-84+ Recipe 580 RC2014 Recipe
620 Sega Master System Recipe

View File

@ -13,4 +13,4 @@ Contents
4 Number literals 6 Compilation vs meta-comp.
8 Interpreter I/O 11 Signed-ness
14 Addressed devices 17 DOES>
18 Disk blocks
18 Disk blocks 21 How blocks are organized

13
blk/021 Normal file
View File

@ -0,0 +1,13 @@
How blocks are organized
Organization of contiguous blocks is an ongoing challenge and
Collapse OS' blocks are never as tidy as they should, but we
try to strive towards a few goals:
1. Block 0 contains documentation discovery core keys to the
uninitiated.
2. First section (up to B100) is usage documentation.
3. B100-B200 are for runtime usage utilities
4. B200-B500 are for bootstrapping
5. The rest is for recipes.
6. I'm not sure yet how I'll organize multiple arches.

16
blk/420 Normal file
View File

@ -0,0 +1,16 @@
Bootstrap guide
You want to deploy Collapse OS on a new system? Start here.
What is Collapse OS? It is a binary placed either in ROM on
in RAM by a bootloader. That binary, when executed, initializes
itself to a Forth interpreter. In most cases, that Forth
interpreter will have some access to a mass storage device,
which allows it to access Collapse OS' disk blocks and come
to this block to bootstrap itself some more.
This binary can be separated in 5 distinct layers:
1. Boot code (B280)
2. Boot words (B305)
3. Core words (low) (B350)
4. Drivers (cont.)

16
blk/421 Normal file
View File

@ -0,0 +1,16 @@
5. Core words (high)
Boot code (B280)
This part contains core routines that underpins Forth fundamen-
tal structures: dict navigation and search, PSP/RSP bounds
checks, word types (atom, native, literals, "does type"), etc.
It also of course does core initialization: set RSP/PSP, HERE
CURRENT, then find BOOT and call it (see B89).
It also contains what we call the "stable ABI" in its first
0x100 bytes. The beginning og the dict is intertwined in this
layer because EXIT, (br), (?br) and (loop) are part of the
stable ABI.
(cont.)

16
blk/422 Normal file
View File

@ -0,0 +1,16 @@
Boot words (B305)
Then come the implementation of core Forth words in native
assembly. Performance is not Collapse OS' primary design goal,
so we try to keep this section to a minimum: we much prefer
to implement our words in Forth.
However, some words are in this section for performance
reasons. Sometimes, the gain is too great to pass up.
Core words (low) (B350)
Then comes the part where we begin defining words in Forth.
Core words are designed to be cross-compiled (B260), from a
full Forth interpreter. This means that it has access to more
than boot words. This somes with tricky limitations. (cont.)

16
blk/423 Normal file
View File

@ -0,0 +1,16 @@
See B260 for details.
Drivers
Up until now, we haven't implemented EMIT or KEY yet: those
words are defined in the "high" part of core words because we
generally need machine-specific drivers to implement (emit) and
(key).
Well, now is their time to shine. We split core in two
precisely to fit drivers in there. This way. they have access
to a pretty good vocabulary and they're also give the oppor-
tunity to provide (emit) and (key).
(cont.)

16
blk/424 Normal file
View File

@ -0,0 +1,16 @@
Core words (high) (B350)
Then come EMIT, KEY and everything that depend on it, until
we have a full Forth interpreter. At the very end, we define
tricky IMMEDIATEs that, if defined earlier, would break cross
compilation.
We end that with a hook words which is also where CURRENT will
be on boot.
So that's the anatomy of a Collapse OS binary. How do you build
one? If your machine is already covered by a recipe, you're in
luck: follow instructions.
If you're deploying to a new machine, you'll have to write a
new xcomp (cross compilation) unit. Let's look at its (cont.)

16
blk/425 Normal file
View File

@ -0,0 +1,16 @@
anatomy. First, we have constants. Some of them are device-
specific, but some of them are always there. RAMSTART is the
address at which the RAM starts on the system. System variables
will go there and HERE will go after it.
RS_ADDR is where RSP starts and PS_ADDR is where PSP starts.
RSP and PSP are designed to be contiguous. RSP goes up and PSP
goes down. If they meet, we know we have a stack overflow.
Then, we load the assembler and cross compilation unit, which
will be needed for the task ahead.
Then, it's a matter of adding layer after layer. For most
system, all those layers except the drivers will be added the
same way. Drivers are a bit tricker and machine specific. I
can't help you there, you'll have to use your wits. (cont.)

15
blk/426 Normal file
View File

@ -0,0 +1,15 @@
After we've loaded the high part of the core words, we're at
the "wrapping up" part. We add what we call a "hook word" (an
empty word with a single letter name) which doesn't cost us
much and can be very useful if we need to augment the binary
with more words, and at that point we have our future boot
CURRENT, which PC yields. That is why we write it to the
LATEST field of the stable ABI: This value will be used at
boot.
After the last word of the dictionary comes the "source init"
part. The boot sequence is designed to interpret whatever comes
after LATEST as Forth source, and this, until it reads ASCII
EOT character (4). This is generally used for driver init.
Good luck!