1
0
mirror of https://github.com/hsoft/collapseos.git synced 2024-11-23 22:38:06 +11:00
collapseos/recipes/rc2014
Virgil Dupras 865f4f9256 Move AT28 driver to blkfs
And adjust rc2014/eeprom recipe
2020-04-26 15:18:28 -04:00
..
eeprom Move AT28 driver to blkfs 2020-04-26 15:18:28 -04:00
ps2 emul/zasm: use libcfs 2019-12-31 15:07:39 -05:00
sdcard recipe/rc2014/selfhost: new recipe 2020-04-25 19:24:55 -04:00
selfhost recipe/rc2014/selfhost: new recipe 2020-04-25 19:24:55 -04:00
Makefile Move link.fs to blkfs 2020-04-26 14:37:54 -04:00
README.md Move AT28 driver to blkfs 2020-04-26 15:18:28 -04:00
xcomp.fs recipes/rc2014: include readln directly in stage 1 2020-04-26 14:52:55 -04:00

RC2014

The RC2014 is a nice and minimal z80 system that has the advantage of being available in an assembly kit. Assembling it yourself involves quite a bit of soldering due to the bus system. However, one very nice upside of that bus system is that each component is isolated and simple.

The machine used in this recipe is the "Classic" RC2014 with an 8k ROM module , 32k of RAM, a 7.3728Mhz clock and a serial I/O.

The ROM module being supplied in the assembly kit is an EPROM, not EEPROM, so you can't install Collapse OS on it. You'll have to supply your own.

There are many options around to boot arbitrary sources. What was used in this recipe was a AT28C64B EEPROM module. I chose it because it's compatible with the 8k ROM module which is very convenient. If you do the same, however, don't forget to set the A14 jumper to high because what is the A14 pin on the AT27 ROM module is the WE pin on the AT28! Setting the jumper high will keep is disabled.

This recipe is for installing a minimal Collapse OS system on the RC2014. There are other recipes related to the RC2014:

Recipe

The goal is to have the shell running and accessible through the Serial I/O.

You'll need specialized tools to write data to the AT28 EEPROM. There seems to be many devices around made to write in flash and EEPROM modules, but being in a "understand everything" mindset, I built my own. This is the device I use in this recipe.

Gathering parts

Configure your build

Modules used in this build are configured through the conf.fs file in this folder. There isn't much to configure, but it's there.

Build stage 1

Self-bootstrapping is in Forth's DNA, which is really nice, but it makes cross-compiling a bit tricky. It's usually much easier to bootstrap a Forth from itself than trying to compile it from a foreign host.

This makes us adopt a 2 stages strategy. A tiny core is built from a foreign host, and then we run that tiny core on the target machine and let it bootstrap itself, then write our full interpreter binary.

We could have this recipe automate that 2 stage build process all automatically, but that would rob you of all your fun, right? Instead, we'll run that 2nd stage on the RC2014 itself!

To build your stage 1, run make in this folder, this will yield stage1.bin. This will contain that tiny core and, appended to it, the Forth source code it needs to run to bootstrap itself. When it's finished bootstrapping, you will get a prompt to a full Forth interpreter.

Emulate

The Collapse OS project includes a RC2014 emulator suitable for this image. You can invoke it with make emul. See emul/hw/rc2014/README.md for details.

Write to the ROM

Plug your romwrite atmega328 to your computer and identify the tty bound to it. In my case (arduino uno), it's /dev/ttyACM0. Then:

screen /dev/ttyACM0 9600
CTRL-A + ":quit"
cat rom.bin | pv -L 10 > /dev/ttyACM0

See romwrite's README for details about these commands.

Note that this method is slow and clunky, but before long, you won't be using it anymore. Writing to an EEPROM is much easier and faster from a RC2014 running Collapse OS, so once you have that first Collapse OS ROM, you'll be much better equipped for further toying around (unless, of course, you already had tools to write to EEPROM. In which case, you'll be ignoring this section altogether).

Running

Put the AT28 in the ROM module, don't forget to set the A14 jumper high, then power the thing up. Connect the FTDI-to-TTL cable to the Serial I/O module and identify the tty bound to it (in my case, /dev/ttyUSB0). Then:

screen /dev/ttyUSB0 115200

Press the reset button on the RC2014 to have Forth begin its bootstrap process. Note that it has to build more than half of itself from source. It takes about 30 seconds to complete.

Once bootstrapping is done you should see the Collapse OS prompt. That's a full Forth interpreter. You can have fun right now.

However, that long boot time is kinda annoying. Moreover, that bootstrap code being in source form takes precious space from our 8K ROM. That brings us to building stage 2.

Building stage 2

You're about to learn a lot about this platform and its self-bootstrapping nature, but its a bumpy ride. Grab something. Why not a beer?

Our stage 1 prompt is the result of Forth's inner core interpreting the source code of the Full Forth, which was appended to the binary inner core in ROM. This results in a compiled dictionary, in RAM, at address 0x8000+system RAM.

Wouldn't it be great if we could save that compiled binary in ROM and save the system the trouble of recompiling itself on boot?

Unfortunately, this compiled dictionary isn't usable as-is. Offsets compiled in there are compiled based on a 0x8000-or-so base offset. What we need is a 0xa00-or-so base offset, that is, something suitable to be appended to the boot binary, in ROM, in binary form.

Fortunately, inside the compiled source is the contents of the Linker (B120) which will allow us to relink our compiled dictionary so that in can be relocated in ROM, next to our boot binary. I won't go into relinking details. Look at the source. For now, let's just use it:

RLCORE

That command will take the dict from ' H@ up to CURRENT, copy it in free memory and then relocate it. It will print 3 addresses during its processing.

The first address is the top copied address. The process didn't touch memory above this point. The second address is the wordref of the last copied entry. The 3rd is the bottom address of the copied dict. When that last address is printed, the processing is over (because we don't have a > prompt, we don't have any other indicator that the process is over).

Assembling the stage 2 binary

At that point, we have a fully relocated binary in memory. Depending on our situations, the next steps differ.

  • If we're on a RC2014 that has writing capabilities to permanent storage, we'll want to assemble that binary directly on the RC2014 and write it to permanent storage.
  • If we're on a RC2014 that doesn't have those capabilities, we'll want to dump memory on our modern environment using /tools/memdump and then assemble that binary there.
  • If we're in the emulator, we'll want to dump our memory using CTRL+E and then assemble our stage 2 binary from that dump.

In these instructions, we assume an emulated environment. I'll use actual offsets of an actual assembling session, but these of course are only examples. It is very likely that these will not be the same offsets for you.

So you've pressed CTRL+E and you have a memdump file. Open it with a hex editor (I like hexedit) to have a look around and to decide what we'll extract from that memdump. RLCORE already gave you important offsets (in my case, 9a3c, 99f6 and 8d60), but although the beginning of will always be the same (8d60), the end offset depends on the situation.

If you look at data between 99f6 and 9a3c, you'll see that this data is not 100% dictionary entry material. Some of it is buffer data allocated at initialization. To locate the end of a word, look for 0042, the address for EXIT. In my case, it's at 9a1a and it's the end of the INIT word.

Moreover, the INIT routine that is in there is not quite what we want, because it doesn't contain the HERE adjustment that we find in pre.fs. We'll want to exclude it from our binary, so let's go a bit further, at 99cf, ending at 99de.

So, the end of our compiled dict is actually 99de. Alright, let's extract it:

dd if=memdump bs=1 skip=36192 count=3198 > dict.bin

36192 is 8d60 and 3198 is 99de-8d60. This needs to be prepended by the boot binary. We already have stage1.bin, but this binary contains bootstrap source code we don't need any more. To strip it, we'll need to dd it out to LATEST, in my case 098b:

dd if=stage1.bin bs=1 count=2443 > s1pre.bin

Now we can combine our binaries:

cat s1pre.bin dict.bin > stage2.bin

Is it ready to run yet? no. There are 3 adjustments we need to manually make using our hex editor.

  1. We need to link H@ to the hook word of the boot binary. In my case, it's a matter of writing 02 at 08ec and 00 at 08ed, H@'s prev field.
  2. We need to end our binary with a hook word. It can have a zero-length name and the prev field needs to properly point to the previous wordref. In my case, that was RLCORE at offset 1559 for a stage2.bin size of 1568, which means that I appended 0F 00 00 at the end of the file.
  3. Finally, we need to adjust LATEST which is at offset 08. This needs to point to the last wordref of the file, which is equal to the length of stage2.bin because we've just added a hook word. This means that we write 6B at offset 08 and 15 at offset 09.

Now are we ready yet? ALMOST! There's one last thing we need to do: add runtime source. In our case, because we have a compiled dict, the only source we need to include is initialization code. We've stripped it from our stage1 earlier, we need to re-add it.

Look at xcomp.fs. You see that ," bla bla bla" line? That's initialization code. Copy it to a file like run.fs (without the ,") and build your final binary:

cat stage2.bin run.fs > stage2r.bin

That's it! our binary is ready to run!

../../emul/hw/rc2014/classic stage2r.bin

And there you have it, a stage2 binary that you've assembled yourself.