annexia/jonesforth.s.txt uses LITSTRING, while jonesforth.s/f use LITS.

There's no real reason why this should be different from the original.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

The ARM Jonesforth is version 47, while the prelude
and x86 version under annexia were version 42.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

It's error-prone and doesn't quite help the readability, so
implicitly calculate the length within the macro.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

Now we can make machine code definitions in Forth. I've
decided to make all machine code emitters start with $
(such as $NEXT), although this may change in the future.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

putchar may buffer after all.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

Make ;CODE behave like the gforth variant, i.e. usable
inside a level 2 (word defining) word (i.e. after CREATE).

Before (see x86 jonesforth 47):

: FOO ;CODE

Now:

CODE FOO END-CODE

Or with CREATE:

: MKNOTHING WORD CREATE 0 , ;CODE END-CODE
MKNOTHING FOO

More examples... so now we could do something like:

defword "$DOCON",F_IMM,ASMDOCON
	.int LIT            @ r0 points to DFA
	ldr r12, [r0, #4]   @ read cell from DFA
	.int COMMA
	.int LIT
	PUSHDSP r12         @ push to stack
	.int COMMA
	.int EXIT

: MKCON
   WORD CREATE
   0 ,        ( push dummy codeword, rewritten by (;CODE))
   ,          ( actual constant )
;CODE
   $DOCON
END-CODE

5 MKCON CON5
CON5 . CR

CODE ... ENDCODE instead can be used to make `defcode'
definitions in Forth instead of jonesforth.s.

So now we could do something like:

@ push r0 to stack
defword "$<R0",F_IMM,ASMFROMR0
	.int LIT
	PUSHDSP r0
	.int COMMA
	.int EXIT

@ push r7 to stack
defword "$<R7",F_IMM,ASMFROMR7
	.int LIT
	PUSHDSP r7
	.int COMMA
	.int EXIT

@ pop stack to r0
defword "$>R0",F_IMM,ASMTOR0
	.int LIT
	POPDSP r0
	.int COMMA
	.int EXIT
@ pop stack to r7
defword "$>R7",F_IMM,ASMTOR7
	.int LIT
	POPDSP r7
	.int COMMA
 	.int EXIT

CODE SWAP $>R0 $>R7 $<R0 $<R7 END-CODE

HEX 1337 FOOF SWAP . . CR

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

DICT signature is changed.
   On success: ( -- dictionary_address )
   On failure ( -- word_addr word_len 0 )

Introduce DICT-CHECKED wrapper around DICT to log error about
unknown word and throws an ABORT.

Fix all words using DICT to instead use DICT-CHECKED.

As an unfortunate side effect, the exception handling code
had to travel upwards somewhat.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

[instead of segfaulting or printing nothing, of course]

Now:
   CODE FOO END-CODE
   SEE
...gives
   CODE FOO ( CODEWORD FOO+0 ) E49A0004 E5901000 E12FFF11 (END-CODE)

...where the hex stuff is the $NEXT macro.

CREATE ... ;CODE words:
   : MKNOTHING WORD CREATE 0 , 3 , ;CODE END-CODE
   MKNOTHING BAR
   SEE MKNOTHING
   SEE BAR
...gives
   : MKNOTHING WORD CREATE 0 , 3 , ;CODE ( MKNOTHING+9 ) E49A0004 E5901000 E12FFF11 (END-CODE)
   CODE BAR ( CODEWORD MKNOTHING+9 ) 3 (END-CODE)

...which shows you that BAR is a word created with MKNOTHING.

Native words:
   SEE 2SWAP
...gives
   CODE 2SWAP ( CODEWORD 85BC ) (END-CODE)

...since length isn't encoded defcode definitions, this is the best we can do.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

See my blog post for details.

http://osdevnotes.blogspot.com/2015/07/does-in-jonesforth.html

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

I really just want this to be a bunch of useful
stuff built around a generic jonesforth.s.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

Be more explicit about arch.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>

This is still a wip. Some basic things seem to work, but a lot more
is to be done.

Signed-off-by: Andrei Warkentin <andrey.warkentin@gmail.com>