2 \ A sometimes minimal FORTH compiler and tutorial for Linux / i386 systems. -*- asm -*-
3 \ By Richard W.M. Jones <rich@annexia.org> http://annexia.org/forth
4 \ This is PUBLIC DOMAIN (see public domain release statement below).
5 \ $Id: jonesforth.f,v 1.14 2007-10-10 13:01:05 rich Exp $
7 \ The first part of this tutorial is in jonesforth.S. Get if from http://annexia.org/forth
9 \ PUBLIC DOMAIN ----------------------------------------------------------------------
11 \ I, the copyright holder of this work, hereby release it into the public domain. This applies worldwide.
13 \ In case this is not legally possible, I grant any entity the right to use this work for any purpose,
14 \ without any conditions, unless such conditions are required by law.
16 \ SETTING UP ----------------------------------------------------------------------
18 \ Let's get a few housekeeping things out of the way. Firstly because I need to draw lots of
19 \ ASCII-art diagrams to explain concepts, the best way to look at this is using a window which
20 \ uses a fixed width font and is at least this wide:
22 \<------------------------------------------------------------------------------------------------------------------------>
24 \ Secondly make sure TABS are set to 8 characters. The following should be a vertical
25 \ line. If not, sort out your tabs.
31 \ Thirdly I assume that your screen is at least 50 characters high.
33 \ START OF FORTH CODE ----------------------------------------------------------------------
35 \ We've now reached the stage where the FORTH system is running and self-hosting. All further
36 \ words can be written as FORTH itself, including words like IF, THEN, .", etc which in most
37 \ languages would be considered rather fundamental.
39 \ Some notes about the code:
41 \ I use indenting to show structure. The amount of whitespace has no meaning to FORTH however
42 \ except that you must use at least one whitespace character between words, and words themselves
43 \ cannot contain whitespace.
45 \ FORTH is case-sensitive. Use capslock!
47 \ The primitive word /MOD (DIVMOD) leaves both the quotient and the remainder on the stack. (On
48 \ i386, the idivl instruction gives both anyway). Now we can define the / and MOD in terms of /MOD
49 \ and a few other primitives.
53 \ Define some character constants
55 : BL 32 ; \ BL (BLank) is a standard FORTH word for space.
57 \ CR prints a carriage return
60 \ SPACE prints a space
63 \ The 2... versions of the standard operators work on pairs of stack entries. They're not used
64 \ very commonly so not really worth writing in assembler. Here is how they are defined in FORTH.
68 \ NEGATE leaves the negative of a number on the stack.
71 \ Standard words for booleans.
76 \ LITERAL takes whatever is on the stack and compiles LIT <foo>
79 , \ compile the literal itself (from the stack)
82 \ Now we can use [ and ] to insert literals which are calculated at compile time. (Recall that
83 \ [ and ] are the FORTH words which switch into and out of immediate mode.)
84 \ Within definitions, use [ ... ] LITERAL anywhere that '...' is a constant expression which you
85 \ would rather only compute once (at compile time, rather than calculating it each time your word runs).
87 [ \ go into immediate mode (temporarily)
88 CHAR : \ push the number 58 (ASCII code of colon) on the parameter stack
89 ] \ go back to compile mode
90 LITERAL \ compile LIT 58 as the definition of ':' word
93 \ A few more character constants defined the same way as above.
94 : ';' [ CHAR ; ] LITERAL ;
95 : '(' [ CHAR ( ] LITERAL ;
96 : ')' [ CHAR ) ] LITERAL ;
97 : '"' [ CHAR " ] LITERAL ;
98 : 'A' [ CHAR A ] LITERAL ;
99 : '0' [ CHAR 0 ] LITERAL ;
100 : '-' [ CHAR - ] LITERAL ;
101 : '.' [ CHAR . ] LITERAL ;
103 \ While compiling, '[COMPILE] word' compiles 'word' if it would otherwise be IMMEDIATE.
104 : [COMPILE] IMMEDIATE
105 WORD \ get the next word
106 FIND \ find it in the dictionary
107 >CFA \ get its codeword
111 \ RECURSE makes a recursive call to the current word that is being compiled.
113 \ Normally while a word is being compiled, it is marked HIDDEN so that references to the
114 \ same word within are calls to the previous definition of the word. However we still have
115 \ access to the word which we are currently compiling through the LATEST pointer so we
116 \ can use that to compile a recursive call.
118 LATEST @ \ LATEST points to the word being compiled at the moment
119 >CFA \ get the codeword
123 \ CONTROL STRUCTURES ----------------------------------------------------------------------
125 \ So far we have defined only very simple definitions. Before we can go further, we really need to
126 \ make some control structures, like IF ... THEN and loops. Luckily we can define arbitrary control
127 \ structures directly in FORTH.
129 \ Please note that the control structures as I have defined them here will only work inside compiled
130 \ words. If you try to type in expressions using IF, etc. in immediate mode, then they won't work.
131 \ Making these work in immediate mode is left as an exercise for the reader.
133 \ condition IF true-part THEN rest
134 \ -- compiles to: --> condition 0BRANCH OFFSET true-part rest
135 \ where OFFSET is the offset of 'rest'
136 \ condition IF true-part ELSE false-part THEN
137 \ -- compiles to: --> condition 0BRANCH OFFSET true-part BRANCH OFFSET2 false-part rest
138 \ where OFFSET if the offset of false-part and OFFSET2 is the offset of rest
140 \ IF is an IMMEDIATE word which compiles 0BRANCH followed by a dummy offset, and places
141 \ the address of the 0BRANCH on the stack. Later when we see THEN, we pop that address
142 \ off the stack, calculate the offset, and back-fill the offset.
144 ' 0BRANCH , \ compile 0BRANCH
145 HERE @ \ save location of the offset on the stack
146 0 , \ compile a dummy offset
151 HERE @ SWAP - \ calculate the offset from the address saved on the stack
152 SWAP ! \ store the offset in the back-filled location
156 ' BRANCH , \ definite branch to just over the false-part
157 HERE @ \ save location of the offset on the stack
158 0 , \ compile a dummy offset
159 SWAP \ now back-fill the original (IF) offset
160 DUP \ same as for THEN word above
165 \ BEGIN loop-part condition UNTIL
166 \ -- compiles to: --> loop-part condition 0BRANCH OFFSET
167 \ where OFFSET points back to the loop-part
168 \ This is like do { loop-part } while (condition) in the C language
170 HERE @ \ save location on the stack
174 ' 0BRANCH , \ compile 0BRANCH
175 HERE @ - \ calculate the offset from the address saved on the stack
176 , \ compile the offset here
179 \ BEGIN loop-part AGAIN
180 \ -- compiles to: --> loop-part BRANCH OFFSET
181 \ where OFFSET points back to the loop-part
182 \ In other words, an infinite loop which can only be returned from with EXIT
184 ' BRANCH , \ compile BRANCH
185 HERE @ - \ calculate the offset back
186 , \ compile the offset here
189 \ BEGIN condition WHILE loop-part REPEAT
190 \ -- compiles to: --> condition 0BRANCH OFFSET2 loop-part BRANCH OFFSET
191 \ where OFFSET points back to condition (the beginning) and OFFSET2 points to after the whole piece of code
192 \ So this is like a while (condition) { loop-part } loop in the C language
194 ' 0BRANCH , \ compile 0BRANCH
195 HERE @ \ save location of the offset2 on the stack
196 0 , \ compile a dummy offset2
200 ' BRANCH , \ compile BRANCH
201 SWAP \ get the original offset (from BEGIN)
202 HERE @ - , \ and compile it after BRANCH
204 HERE @ SWAP - \ calculate the offset2
205 SWAP ! \ and back-fill it in the original location
208 \ UNLESS is the same as IF but the test is reversed.
210 \ Note the use of [COMPILE]: Since IF is IMMEDIATE we don't want it to be executed while UNLESS
211 \ is compiling, but while UNLESS is running (which happens to be when whatever word using UNLESS is
212 \ being compiled -- whew!). So we use [COMPILE] to reverse the effect of marking IF as immediate.
213 \ This trick is generally used when we want to write our own control words without having to
214 \ implement them all in terms of the primitives 0BRANCH and BRANCH, but instead reusing simpler
215 \ control words like (in this instance) IF.
217 ' NOT , \ compile NOT (to reverse the test)
218 [COMPILE] IF \ continue by calling the normal IF
221 \ COMMENTS ----------------------------------------------------------------------
223 \ FORTH allows ( ... ) as comments within function definitions. This works by having an IMMEDIATE
224 \ word called ( which just drops input characters until it hits the corresponding ).
226 1 \ allowed nested parens by keeping track of depth
228 KEY \ read next character
229 DUP '(' = IF \ open paren?
230 DROP \ drop the open paren
233 ')' = IF \ close paren?
237 DUP 0= UNTIL \ continue until we reach matching close paren, depth 0
238 DROP \ drop the depth counter
242 From now on we can use ( ... ) for comments.
244 STACK NOTATION ----------------------------------------------------------------------
246 In FORTH style we can also use ( ... -- ... ) to show the effects that a word has on the
247 parameter stack. For example:
249 ( n -- ) means that the word consumes an integer (n) from the parameter stack.
250 ( b a -- c ) means that the word uses two integers (a and b, where a is at the top of stack)
251 and returns a single integer (c).
252 ( -- ) means the word has no effect on the stack
255 ( Some more complicated stack examples, showing the stack notation. )
256 : NIP ( x y -- y ) SWAP DROP ;
257 : TUCK ( x y -- y x y ) DUP ROT ;
258 : PICK ( x_u ... x_1 x_0 u -- x_u ... x_1 x_0 x_u )
259 1+ ( add one because of 'u' on the stack )
260 4 * ( multiply by the word size )
261 DSP@ + ( add to the stack pointer )
265 ( With the looping constructs, we can now write SPACES, which writes n spaces to stdout. )
268 DUP 0> ( while n > 0 )
270 SPACE ( print a space )
271 1- ( until we count down to 0 )
276 ( Standard words for manipulating BASE. )
277 : DECIMAL ( -- ) 10 BASE ! ;
278 : HEX ( -- ) 16 BASE ! ;
281 PRINTING NUMBERS ----------------------------------------------------------------------
283 The standard FORTH word . (DOT) is very important. It takes the number at the top
284 of the stack and prints it out. However first I'm going to implement some lower-level
287 U.R ( u width -- ) which prints an unsigned number, padded to a certain width
288 U. ( u -- ) which prints an unsigned number
289 .R ( n width -- ) which prints a signed number, padded to a certain width.
293 will print out these characters:
294 <space> <space> - 1 2 3
296 In other words, the number padded left to a certain number of characters.
298 The full number is printed even if it is wider than width, and this is what allows us to
299 define the ordinary functions U. and . (we just set width to zero knowing that the full
300 number will be printed anyway).
302 Another wrinkle of . and friends is that they obey the current base in the variable BASE.
303 BASE can be anything in the range 2 to 36.
305 While we're defining . &c we can also define .S which is a useful debugging tool. This
306 word prints the current stack (non-destructively) from top to bottom.
309 ( This is the underlying recursive definition of U. )
311 BASE @ /MOD ( width rem quot )
312 ?DUP IF ( if quotient <> 0 then )
313 RECURSE ( print the quotient )
316 ( print the remainder )
318 '0' ( decimal digits 0..9 )
320 10 - ( hex and beyond digits A..Z )
328 FORTH word .S prints the contents of the stack. It doesn't alter the stack.
329 Very useful for debugging.
332 DSP@ ( get current stack pointer )
336 DUP @ U. ( print the stack element )
343 ( This word returns the width (in characters) of an unsigned number in the current base )
344 : UWIDTH ( u -- width )
345 BASE @ / ( rem quot )
346 ?DUP IF ( if quotient <> 0 then )
347 RECURSE 1+ ( return 1+recursive call )
356 UWIDTH ( width u uwidth )
357 -ROT ( u uwidth width )
358 SWAP - ( u width-uwidth )
359 ( At this point if the requested width is narrower, we'll have a negative number on the stack.
360 Otherwise the number on the stack is the number of spaces to print. But SPACES won't print
361 a negative number of spaces anyway, so it's now safe to call SPACES ... )
363 ( ... and then call the underlying implementation of U. )
368 .R prints a signed number, padded to a certain width. We can't just print the sign
369 and call U.R because we want the sign to be next to the number ('-123' instead of '- 123').
375 1 ( save a flag to remember that it was negative | width n 1 )
384 SWAP ( flag width u )
385 DUP ( flag width u u )
386 UWIDTH ( flag width u uwidth )
387 -ROT ( flag u uwidth width )
388 SWAP - ( flag u width-uwidth )
393 IF ( was it negative? print the - character )
400 ( Finally we can define word . in terms of .R, with a trailing space. )
403 ( The real U., note the trailing space. )
406 ( ? fetches the integer at an address and prints it. )
407 : ? ( addr -- ) @ . ;
409 ( c a b WITHIN returns true if a <= c and c < b )
425 ( DEPTH returns the depth of the stack. )
428 4- ( adjust because S0 was on the stack when we pushed DSP )
432 ALIGNED takes an address and rounds it up (aligns it) to the next 4 byte boundary.
434 : ALIGNED ( addr -- addr )
435 3 + 3 INVERT AND ( (addr+3) & ~3 )
439 ALIGN aligns the HERE pointer, so the next word appended will be aligned properly.
441 : ALIGN HERE @ ALIGNED HERE ! ;
444 STRINGS ----------------------------------------------------------------------
446 S" string" is used in FORTH to define strings. It leaves the address of the string and
447 its length on the stack, (length at the top of stack). The space following S" is the normal
448 space between FORTH words and is not a part of the string.
450 This is tricky to define because it has to do different things depending on whether
451 we are compiling or in immediate mode. (Thus the word is marked IMMEDIATE so it can
452 detect this and do different things).
454 In compile mode we append
455 LITSTRING <string length> <string rounded up 4 bytes>
456 to the current word. The primitive LITSTRING does the right thing when the current
459 In immediate mode there isn't a particularly good place to put the string, but in this
460 case we put the string at HERE (but we _don't_ change HERE). This is meant as a temporary
461 location, likely to be overwritten soon after.
463 ( C, appends a byte to the current compiled word. )
465 HERE @ C! ( store the character in the compiled image )
466 1 HERE +! ( increment HERE pointer by 1 byte )
469 : S" IMMEDIATE ( -- addr len )
470 STATE @ IF ( compiling? )
471 ' LITSTRING , ( compile LITSTRING )
472 HERE @ ( save the address of the length word on the stack )
473 0 , ( dummy length - we don't know what it is yet )
475 KEY ( get next character of the string )
478 C, ( copy character )
480 DROP ( drop the double quote character at the end )
481 DUP ( get the saved address of the length word )
482 HERE @ SWAP - ( calculate the length )
483 4- ( subtract 4 (because we measured from the start of the length word) )
484 SWAP ! ( and back-fill the length location )
485 ALIGN ( round up to next multiple of 4 bytes for the remaining code )
486 ELSE ( immediate mode )
487 HERE @ ( get the start address of the temporary space )
492 OVER C! ( save next character )
493 1+ ( increment address )
495 DROP ( drop the final " character )
496 HERE @ - ( calculate the length )
497 HERE @ ( push the start address )
503 ." is the print string operator in FORTH. Example: ." Something to print"
504 The space after the operator is the ordinary space required between words and is not
505 a part of what is printed.
507 In immediate mode we just keep reading characters and printing them until we get to
508 the next double quote.
510 In compile mode we use S" to store the string, then add TELL afterwards:
511 LITSTRING <string length> <string rounded up to 4 bytes> TELL
513 It may be interesting to note the use of [COMPILE] to turn the call to the immediate
514 word S" into compilation of that word. It compiles it into the definition of .",
515 not into the definition of the word being compiled when this is running (complicated
518 : ." IMMEDIATE ( -- )
519 STATE @ IF ( compiling? )
520 [COMPILE] S" ( read the string, and compile LITSTRING, etc. )
521 ' TELL , ( compile the final TELL )
523 ( In immediate mode, just read characters and print them until we get
524 to the ending double quote. )
528 DROP ( drop the double quote character )
529 EXIT ( return from this function )
537 CONSTANTS AND VARIABLES ----------------------------------------------------------------------
539 In FORTH, global constants and variables are defined like this:
541 10 CONSTANT TEN when TEN is executed, it leaves the integer 10 on the stack
542 VARIABLE VAR when VAR is executed, it leaves the address of VAR on the stack
544 Constants can be read but not written, eg:
548 You can read a variable (in this example called VAR) by doing:
550 VAR @ leaves the value of VAR on the stack
551 VAR @ . CR prints the value of VAR
552 VAR ? CR same as above, since ? is the same as @ .
554 and update the variable by doing:
556 20 VAR ! sets VAR to 20
558 Note that variables are uninitialised (but see VALUE later on which provides initialised
559 variables with a slightly simpler syntax).
561 How can we define the words CONSTANT and VARIABLE?
563 The trick is to define a new word for the variable itself (eg. if the variable was called
564 'VAR' then we would define a new word called VAR). This is easy to do because we exposed
565 dictionary entry creation through the CREATE word (part of the definition of : above).
566 A call to WORD [TEN] CREATE (where [TEN] means that "TEN" is the next word in the input)
567 leaves the dictionary entry:
572 +---------+---+---+---+---+
573 | LINK | 3 | T | E | N |
574 +---------+---+---+---+---+
577 For CONSTANT we can continue by appending DOCOL (the codeword), then LIT followed by
578 the constant itself and then EXIT, forming a little word definition that returns the
581 +---------+---+---+---+---+------------+------------+------------+------------+
582 | LINK | 3 | T | E | N | DOCOL | LIT | 10 | EXIT |
583 +---------+---+---+---+---+------------+------------+------------+------------+
586 Notice that this word definition is exactly the same as you would have got if you had
589 Note for people reading the code below: DOCOL is a constant word which we defined in the
590 assembler part which returns the value of the assembler symbol of the same name.
593 WORD ( get the name (the name follows CONSTANT) )
594 CREATE ( make the dictionary entry )
595 DOCOL , ( append DOCOL (the codeword field of this word) )
596 ' LIT , ( append the codeword LIT )
597 , ( append the value on the top of the stack )
598 ' EXIT , ( append the codeword EXIT )
602 VARIABLE is a little bit harder because we need somewhere to put the variable. There is
603 nothing particularly special about the user memory (the area of memory pointed to by HERE
604 where we have previously just stored new word definitions). We can slice off bits of this
605 memory area to store anything we want, so one possible definition of VARIABLE might create
608 +--------------------------------------------------------------+
611 +---------+---------+---+---+---+---+------------+------------+---|--------+------------+
612 | <var> | LINK | 3 | V | A | R | DOCOL | LIT | <addr var> | EXIT |
613 +---------+---------+---+---+---+---+------------+------------+------------+------------+
616 where <var> is the place to store the variable, and <addr var> points back to it.
618 To make this more general let's define a couple of words which we can use to allocate
619 arbitrary memory from the user memory.
621 First ALLOT, where n ALLOT allocates n bytes of memory. (Note when calling this that
622 it's a very good idea to make sure that n is a multiple of 4, or at least that next time
623 a word is compiled that HERE has been left as a multiple of 4).
625 : ALLOT ( n -- addr )
626 HERE @ SWAP ( here n )
627 HERE +! ( adds n to HERE, after this the old value of HERE is still on the stack )
631 Second, CELLS. In FORTH the phrase 'n CELLS ALLOT' means allocate n integers of whatever size
632 is the natural size for integers on this machine architecture. On this 32 bit machine therefore
633 CELLS just multiplies the top of stack by 4.
635 : CELLS ( n -- n ) 4 * ;
638 So now we can define VARIABLE easily in much the same way as CONSTANT above. Refer to the
639 diagram above to see what the word that this creates will look like.
642 1 CELLS ALLOT ( allocate 1 cell of memory, push the pointer to this memory )
643 WORD CREATE ( make the dictionary entry (the name follows VARIABLE) )
644 DOCOL , ( append DOCOL (the codeword field of this word) )
645 ' LIT , ( append the codeword LIT )
646 , ( append the pointer to the new memory )
647 ' EXIT , ( append the codeword EXIT )
651 VALUES ----------------------------------------------------------------------
653 VALUEs are like VARIABLEs but with a simpler syntax. You would generally use them when you
654 want a variable which is read often, and written infrequently.
656 20 VALUE VAL creates VAL with initial value 20
657 VAL pushes the value (20) directly on the stack
658 30 TO VAL updates VAL, setting it to 30
659 VAL pushes the value (30) directly on the stack
661 Notice that 'VAL' on its own doesn't return the address of the value, but the value itself,
662 making values simpler and more obvious to use than variables (no indirection through '@').
663 The price is a more complicated implementation, although despite the complexity there is no
664 performance penalty at runtime.
666 A naive implementation of 'TO' would be quite slow, involving a dictionary search each time.
667 But because this is FORTH we have complete control of the compiler so we can compile TO more
668 efficiently, turning:
672 and calculating <addr> (the address of the value) at compile time.
674 Now this is the clever bit. We'll compile our value like this:
676 +---------+---+---+---+---+------------+------------+------------+------------+
677 | LINK | 3 | V | A | L | DOCOL | LIT | <value> | EXIT |
678 +---------+---+---+---+---+------------+------------+------------+------------+
681 where <value> is the actual value itself. Note that when VAL executes, it will push the
682 value on the stack, which is what we want.
684 But what will TO use for the address <addr>? Why of course a pointer to that <value>:
686 code compiled - - - - --+------------+------------+------------+-- - - - -
687 by TO VAL | LIT | <addr> | ! |
688 - - - - --+------------+-----|------+------------+-- - - - -
691 +---------+---+---+---+---+------------+------------+------------+------------+
692 | LINK | 3 | V | A | L | DOCOL | LIT | <value> | EXIT |
693 +---------+---+---+---+---+------------+------------+------------+------------+
696 In other words, this is a kind of self-modifying code.
698 (Note to the people who want to modify this FORTH to add inlining: values defined this
699 way cannot be inlined).
702 WORD CREATE ( make the dictionary entry (the name follows VALUE) )
703 DOCOL , ( append DOCOL )
704 ' LIT , ( append the codeword LIT )
705 , ( append the initial value )
706 ' EXIT , ( append the codeword EXIT )
709 : TO IMMEDIATE ( n -- )
710 WORD ( get the name of the value )
711 FIND ( look it up in the dictionary )
712 >DFA ( get a pointer to the first data field (the 'LIT') )
713 4+ ( increment to point at the value )
714 STATE @ IF ( compiling? )
715 ' LIT , ( compile LIT )
716 , ( compile the address of the value )
718 ELSE ( immediate mode )
719 ! ( update it straightaway )
723 ( x +TO VAL adds x to VAL )
725 WORD ( get the name of the value )
726 FIND ( look it up in the dictionary )
727 >DFA ( get a pointer to the first data field (the 'LIT') )
728 4+ ( increment to point at the value )
729 STATE @ IF ( compiling? )
730 ' LIT , ( compile LIT )
731 , ( compile the address of the value )
732 ' +! , ( compile +! )
733 ELSE ( immediate mode )
734 +! ( update it straightaway )
739 PRINTING THE DICTIONARY ----------------------------------------------------------------------
741 ID. takes an address of a dictionary entry and prints the word's name.
743 For example: LATEST @ ID. would print the name of the last word that was defined.
746 4+ ( skip over the link pointer )
747 DUP C@ ( get the flags/length byte )
748 F_LENMASK AND ( mask out the flags - just want the length )
751 DUP 0> ( length > 0? )
753 SWAP 1+ ( addr len -- len addr+1 )
754 DUP C@ ( len addr -- len addr char | get the next character)
755 EMIT ( len addr char -- len addr | and print it)
756 SWAP 1- ( len addr -- addr len-1 | subtract one from length )
758 2DROP ( len addr -- )
762 'WORD word FIND ?HIDDEN' returns true if 'word' is flagged as hidden.
764 'WORD word FIND ?IMMEDIATE' returns true if 'word' is flagged as immediate.
767 4+ ( skip over the link pointer )
768 C@ ( get the flags/length byte )
769 F_HIDDEN AND ( mask the F_HIDDEN flag and return it (as a truth value) )
772 4+ ( skip over the link pointer )
773 C@ ( get the flags/length byte )
774 F_IMMED AND ( mask the F_IMMED flag and return it (as a truth value) )
778 WORDS prints all the words defined in the dictionary, starting with the word defined most recently.
779 However it doesn't print hidden words.
781 The implementation simply iterates backwards from LATEST using the link pointers.
784 LATEST @ ( start at LATEST dictionary entry )
786 ?DUP ( while link pointer is not null )
788 DUP ?HIDDEN NOT IF ( ignore hidden words )
789 DUP ID. ( but if not hidden, print the word )
792 @ ( dereference the link pointer - go to previous word )
798 FORGET ----------------------------------------------------------------------
800 So far we have only allocated words and memory. FORTH provides a rather primitive method
803 'FORGET word' deletes the definition of 'word' from the dictionary and everything defined
804 after it, including any variables and other memory allocated after.
806 The implementation is very simple - we look up the word (which returns the dictionary entry
807 address). Then we set HERE to point to that address, so in effect all future allocations
808 and definitions will overwrite memory starting at the word. We also need to set LATEST to
809 point to the previous word.
811 Note that you cannot FORGET built-in words (well, you can try but it will probably cause
814 XXX: Because we wrote VARIABLE to store the variable in memory allocated before the word,
815 in the current implementation VARIABLE FOO FORGET FOO will leak 1 cell of memory.
818 WORD FIND ( find the word, gets the dictionary entry address )
819 DUP @ LATEST ! ( set LATEST to point to the previous word )
820 HERE ! ( and store HERE with the dictionary address )
824 DUMP ----------------------------------------------------------------------
826 DUMP is used to dump out the contents of memory, in the 'traditional' hexdump format.
828 Notice that the parameters to DUMP (address, length) are compatible with string words
831 : DUMP ( addr len -- )
832 BASE @ ROT ( save the current BASE at the bottom of the stack )
833 HEX ( and switch to hexadecimal mode )
836 ?DUP ( while len > 0 )
838 OVER 8 U.R ( print the address )
841 ( print up to 16 words on this line )
842 2DUP ( addr len addr len )
843 1- 15 AND 1+ ( addr len addr linelen )
845 ?DUP ( while linelen > 0 )
847 SWAP ( addr len linelen addr )
848 DUP C@ ( addr len linelen addr byte )
849 2 .R SPACE ( print the byte )
850 1+ SWAP 1- ( addr len linelen addr -- addr len addr+1 linelen-1 )
854 ( print the ASCII equivalents )
855 2DUP 1- 15 AND 1+ ( addr len addr linelen )
857 ?DUP ( while linelen > 0)
859 SWAP ( addr len linelen addr )
860 DUP C@ ( addr len linelen addr byte )
861 DUP 32 128 WITHIN IF ( 32 <= c < 128? )
866 1+ SWAP 1- ( addr len linelen addr -- addr len addr+1 linelen-1 )
871 DUP 1- 15 AND 1+ ( addr len linelen )
872 DUP ( addr len linelen linelen )
873 ROT ( addr linelen len linelen )
874 - ( addr linelen len-linelen )
875 ROT ( len-linelen addr linelen )
876 + ( len-linelen addr+linelen )
877 SWAP ( addr-linelen len-linelen )
880 DROP ( restore stack )
881 BASE ! ( restore saved BASE )
885 CASE ----------------------------------------------------------------------
887 CASE...ENDCASE is how we do switch statements in FORTH. There is no generally
888 agreed syntax for this, so I've gone for the syntax mandated by the ISO standard
891 ( some value on the stack )
899 The CASE statement tests the value on the stack by comparing it for equality with
900 test1, test2, ..., testn and executes the matching piece of code within OF ... ENDOF.
901 If none of the test values match then the default case is executed. Inside the ... of
902 the default case, the value is still at the top of stack (it is implicitly DROP-ed
903 by ENDCASE). When ENDOF is executed it jumps after ENDCASE (ie. there is no "fall-through"
904 and no need for a break statement like in C).
906 The default case may be omitted. In fact the tests may also be omitted so that you
907 just have a default case, although this is probably not very useful.
909 An example (assuming that 'q', etc. are words which push the ASCII value of the letter
915 'q' OF 1 TO QUIT ENDOF
916 's' OF 1 TO SLEEP ENDOF
918 ." Sorry, I didn't understand key <" DUP EMIT ." >, try again." CR
921 (In some versions of FORTH, more advanced tests are supported, such as ranges, etc.
922 Other versions of FORTH need you to write OTHERWISE to indicate the default case.
923 As I said above, this FORTH tries to follow the ANS FORTH standard).
925 The implementation of CASE...ENDCASE is somewhat non-trivial. I'm following the
926 implementations from here:
927 http://www.uni-giessen.de/faq/archiv/forthfaq.case_endcase/msg00000.html
929 The general plan is to compile the code as a series of IF statements:
931 CASE (push 0 on the immediate-mode parameter stack)
932 test1 OF ... ENDOF test1 OVER = IF DROP ... ELSE
933 test2 OF ... ENDOF test2 OVER = IF DROP ... ELSE
934 testn OF ... ENDOF testn OVER = IF DROP ... ELSE
935 ... ( default case ) ...
936 ENDCASE DROP THEN [THEN [THEN ...]]
938 The CASE statement pushes 0 on the immediate-mode parameter stack, and that number
939 is used to count how many THEN statements we need when we get to ENDCASE so that each
940 IF has a matching THEN. The counting is done implicitly. If you recall from the
941 implementation above of IF, each IF pushes a code address on the immediate-mode stack,
942 and these addresses are non-zero, so by the time we get to ENDCASE the stack contains
943 some number of non-zeroes, followed by a zero. The number of non-zeroes is how many
944 times IF has been called, so how many times we need to match it with THEN.
946 This code uses [COMPILE] so that we compile calls to IF, ELSE, THEN instead of
947 actually calling them while we're compiling the words below.
949 As is the case with all of our control structures, they only work within word
950 definitions, not in immediate mode.
953 0 ( push 0 to mark the bottom of the stack )
957 ' OVER , ( compile OVER )
959 [COMPILE] IF ( compile IF )
960 ' DROP , ( compile DROP )
964 [COMPILE] ELSE ( ENDOF is the same as ELSE )
968 ' DROP , ( compile DROP )
970 ( keep compiling THEN until we get to our zero marker )
979 DECOMPILER ----------------------------------------------------------------------
981 CFA> is the opposite of >CFA. It takes a codeword and tries to find the matching
982 dictionary definition. (In truth, it works with any pointer into a word, not just
983 the codeword pointer, and this is needed to do stack traces).
985 In this FORTH this is not so easy. In fact we have to search through the dictionary
986 because we don't have a convenient back-pointer (as is often the case in other versions
987 of FORTH). Because of this search, CFA> should not be used when performance is critical,
988 so it is only used for debugging tools such as the decompiler and printing stack
991 This word returns 0 if it doesn't find a match.
994 LATEST @ ( start at LATEST dictionary entry )
996 ?DUP ( while link pointer is not null )
998 2DUP SWAP ( cfa curr curr cfa )
999 < IF ( current dictionary entry < cfa? )
1000 NIP ( leave curr dictionary entry on the stack )
1003 @ ( follow link pointer back )
1005 DROP ( restore stack )
1006 0 ( sorry, nothing found )
1010 SEE decompiles a FORTH word.
1012 We search for the dictionary entry of the word, then search again for the next
1013 word (effectively, the end of the compiled word). This results in two pointers:
1015 +---------+---+---+---+---+------------+------------+------------+------------+
1016 | LINK | 3 | T | E | N | DOCOL | LIT | 10 | EXIT |
1017 +---------+---+---+---+---+------------+------------+------------+------------+
1020 Start of word End of word
1022 With this information we can have a go at decompiling the word. We need to
1023 recognise "meta-words" like LIT, LITSTRING, BRANCH, etc. and treat those separately.
1026 WORD FIND ( find the dictionary entry to decompile )
1028 ( Now we search again, looking for the next word in the dictionary. This gives us
1029 the length of the word that we will be decompiling. (Well, mostly it does). )
1030 HERE @ ( address of the end of the last compiled word )
1031 LATEST @ ( word last curr )
1033 2 PICK ( word last curr word )
1034 OVER ( word last curr word curr )
1035 <> ( word last curr word<>curr? )
1036 WHILE ( word last curr )
1038 DUP @ ( word curr prev (which becomes: word last curr) )
1041 DROP ( at this point, the stack is: start-of-word end-of-word )
1042 SWAP ( end-of-word start-of-word )
1044 ( begin the definition with : NAME [IMMEDIATE] )
1045 ':' EMIT SPACE DUP ID. SPACE
1046 DUP ?IMMEDIATE IF ." IMMEDIATE " THEN
1048 >DFA ( get the data address, ie. points after DOCOL | end-of-word start-of-data )
1050 ( now we start decompiling until we hit the end of the word )
1054 DUP @ ( end start codeword )
1057 ' LIT OF ( is it LIT ? )
1058 4 + DUP @ ( get next word which is the integer constant )
1061 ' LITSTRING OF ( is it LITSTRING ? )
1062 [ CHAR S ] LITERAL EMIT '"' EMIT SPACE ( print S"<space> )
1063 4 + DUP @ ( get the length word )
1064 SWAP 4 + SWAP ( end start+4 length )
1065 2DUP TELL ( print the string )
1066 '"' EMIT SPACE ( finish the string with a final quote )
1067 + ALIGNED ( end start+4+len, aligned )
1068 4 - ( because we're about to add 4 below )
1070 ' 0BRANCH OF ( is it 0BRANCH ? )
1072 4 + DUP @ ( print the offset )
1076 ' BRANCH OF ( is it BRANCH ? )
1078 4 + DUP @ ( print the offset )
1082 ' ' OF ( is it ' (TICK) ? )
1083 [ CHAR ' ] LITERAL EMIT SPACE
1084 4 + DUP @ ( get the next codeword )
1085 CFA> ( and force it to be printed as a dictionary entry )
1088 ' EXIT OF ( is it EXIT? )
1089 ( We expect the last word to be EXIT, and if it is then we don't print it
1090 because EXIT is normally implied by ;. EXIT can also appear in the middle
1091 of words, and then it needs to be printed. )
1092 2DUP ( end start end start )
1093 4 + ( end start end start+4 )
1094 <> IF ( end start | we're not at the end )
1099 DUP ( in the default case we always need to DUP before using )
1100 CFA> ( look up the codeword to get the dictionary entry )
1101 ID. SPACE ( and print it )
1109 2DROP ( restore stack )
1113 EXECUTION TOKENS ----------------------------------------------------------------------
1115 Standard FORTH defines a concept called an 'execution token' (or 'xt') which is very
1116 similar to a function pointer in C. We map the execution token to a codeword address.
1118 execution token of DOUBLE is the address of this codeword
1121 +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+
1122 | LINK | 6 | D | O | U | B | L | E | 0 | DOCOL | DUP | + | EXIT |
1123 +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+
1126 There is one assembler primitive for execution tokens, EXECUTE ( xt -- ), which runs them.
1128 You can make an execution token for an existing word the long way using >CFA,
1129 ie: WORD [foo] FIND >CFA will push the xt for foo onto the stack where foo is the
1130 next word in input. So a very slow way to run DOUBLE might be:
1133 : SLOW WORD FIND >CFA EXECUTE ;
1134 5 SLOW DOUBLE . CR \ prints 10
1136 We also offer a simpler and faster way to get the execution token of any word FOO:
1140 (Exercises for readers: (1) What is the difference between ['] FOO and ' FOO?
1141 (2) What is the relationship between ', ['] and LIT?)
1143 More useful is to define anonymous words and/or to assign xt's to variables.
1145 To define an anonymous word (and push its xt on the stack) use :NONAME ... ; as in this
1148 :NONAME ." anon word was called" CR ; \ pushes xt on the stack
1149 DUP EXECUTE EXECUTE \ executes the anon word twice
1151 Stack parameters work as expected:
1153 :NONAME ." called with parameter " . CR ;
1155 10 SWAP EXECUTE \ prints 'called with parameter 10'
1156 20 SWAP EXECUTE \ prints 'called with parameter 20'
1158 Notice that the above code has a memory leak: the anonymous word is still compiled
1159 into the data segment, so even if you lose track of the xt, the word continues to
1160 occupy memory. A good way to keep track of the xt and thus avoid the memory leak is
1161 to assign it to a CONSTANT, VARIABLE or VALUE:
1164 :NONAME ." anon word was called" CR ; TO ANON
1168 Another use of :NONAME is to create an array of functions which can be called quickly
1169 (think: fast switch statement). This example is adapted from the ANS FORTH standard:
1171 10 CELLS ALLOT CONSTANT CMD-TABLE
1172 : SET-CMD CELLS CMD-TABLE + ! ;
1173 : CALL-CMD CELLS CMD-TABLE + @ EXECUTE ;
1175 :NONAME ." alternate 0 was called" CR ; 0 SET-CMD
1176 :NONAME ." alternate 1 was called" CR ; 1 SET-CMD
1178 :NONAME ." alternate 9 was called" CR ; 9 SET-CMD
1185 0 0 CREATE ( create a word with no name - we need a dictionary header because ; expects it )
1186 HERE @ ( current HERE value is the address of the codeword, ie. the xt )
1187 DOCOL , ( compile DOCOL (the codeword) )
1188 ] ( go into compile mode )
1192 ' LIT , ( compile LIT )
1196 EXCEPTIONS ----------------------------------------------------------------------
1198 Amazingly enough, exceptions can be implemented directly in FORTH, in fact rather easily.
1200 The general usage is as follows:
1202 : FOO ( n -- ) THROW ;
1205 25 ['] FOO CATCH \ execute 25 FOO, catching any exception
1207 ." called FOO and it threw exception number: "
1209 DROP \ we have to drop the argument of FOO (25)
1212 \ prints: called FOO and it threw exception number: 25
1214 CATCH runs an execution token and detects whether it throws any exception or not. The
1215 stack signature of CATCH is rather complicated:
1217 ( a_n-1 ... a_1 a_0 xt -- r_m-1 ... r_1 r_0 0 ) if xt did NOT throw an exception
1218 ( a_n-1 ... a_1 a_0 xt -- ?_n-1 ... ?_1 ?_0 e ) if xt DID throw exception 'e'
1220 where a_i and r_i are the (arbitrary number of) argument and return stack contents
1221 before and after xt is EXECUTEd. Notice in particular the case where an exception
1222 is thrown, the stack pointer is restored so that there are n of _something_ on the
1223 stack in the positions where the arguments a_i used to be. We don't really guarantee
1224 what is on the stack -- perhaps the original arguments, and perhaps other nonsense --
1225 it largely depends on the implementation of the word that was executed.
1227 THROW, ABORT and a few others throw exceptions.
1229 Exception numbers are non-zero integers. By convention the positive numbers can be used
1230 for app-specific exceptions and the negative numbers have certain meanings defined in
1231 the ANS FORTH standard. (For example, -1 is the exception thrown by ABORT).
1233 0 THROW does nothing. This is the stack signature of THROW:
1236 ( * e -- ?_n-1 ... ?_1 ?_0 e ) the stack is restored to the state from the corresponding CATCH
1238 The implementation hangs on the definitions of CATCH and THROW and the state shared
1241 Up to this point, the return stack has consisted merely of a list of return addresses,
1242 with the top of the return stack being the return address where we will resume executing
1243 when the current word EXITs. However CATCH will push a more complicated 'exception stack
1244 frame' on the return stack. The exception stack frame records some things about the
1245 state of execution at the time that CATCH was called.
1247 When called, THROW walks up the return stack (the process is called 'unwinding') until
1248 it finds the exception stack frame. It then uses the data in the exception stack frame
1249 to restore the state allowing execution to continue after the matching CATCH. (If it
1250 unwinds the stack and doesn't find the exception stack frame then it prints a message
1251 and drops back to the prompt, which is also normal behaviour for so-called 'uncaught
1254 This is what the exception stack frame looks like. (As is conventional, the return stack
1255 is shown growing downwards from higher to lower memory addresses).
1257 +------------------------------+
1258 | return address from CATCH | Notice this is already on the
1259 | | return stack when CATCH is called.
1260 +------------------------------+
1261 | original parameter stack |
1263 +------------------------------+ ^
1264 | exception stack marker | |
1265 | (EXCEPTION-MARKER) | | Direction of stack
1266 +------------------------------+ | unwinding by THROW.
1270 The EXCEPTION-MARKER marks the entry as being an exception stack frame rather than an
1271 ordinary return address, and it is this which THROW "notices" as it is unwinding the
1272 stack. (If you want to implement more advanced exceptions such as TRY...WITH then
1273 you'll need to use a different value of marker if you want the old and new exception stack
1274 frame layouts to coexist).
1276 What happens if the executed word doesn't throw an exception? It will eventually
1277 return and call EXCEPTION-MARKER, so EXCEPTION-MARKER had better do something sensible
1278 without us needing to modify EXIT. This nicely gives us a suitable definition of
1279 EXCEPTION-MARKER, namely a function that just drops the stack frame and itself
1280 returns (thus "returning" from the original CATCH).
1282 One thing to take from this is that exceptions are a relatively lightweight mechanism
1287 RDROP ( drop the original parameter stack pointer )
1288 0 ( there was no exception, this is the normal return path )
1291 : CATCH ( xt -- exn? )
1292 DSP@ 4+ >R ( save parameter stack pointer (+4 because of xt) on the return stack )
1293 ' EXCEPTION-MARKER 4+ ( push the address of the RDROP inside EXCEPTION-MARKER ... )
1294 >R ( ... on to the return stack so it acts like a return address )
1295 EXECUTE ( execute the nested function )
1299 ?DUP IF ( only act if the exception code <> 0 )
1300 RSP@ ( get return stack pointer )
1302 DUP R0 4- < ( RSP < R0 )
1304 DUP @ ( get the return stack entry )
1305 ' EXCEPTION-MARKER 4+ = IF ( found the EXCEPTION-MARKER on the return stack )
1306 4+ ( skip the EXCEPTION-MARKER on the return stack )
1307 RSP! ( restore the return stack pointer )
1309 ( Restore the parameter stack. )
1310 DUP DUP DUP ( reserve some working space so the stack for this word
1311 doesn't coincide with the part of the stack being restored )
1312 R> ( get the saved parameter stack pointer | n dsp )
1313 4- ( reserve space on the stack to store n )
1314 SWAP OVER ( dsp n dsp )
1315 ! ( write n on the stack )
1316 DSP! EXIT ( restore the parameter stack pointer, immediately exit )
1321 ( No matching catch - print a message and restart the INTERPRETer. )
1340 ( Print a stack trace by walking up the return stack. )
1342 RSP@ ( start at caller of this function )
1344 DUP R0 4- < ( RSP < R0 )
1346 DUP @ ( get the return stack entry )
1348 ' EXCEPTION-MARKER 4+ OF ( is it the exception stack frame? )
1350 4+ DUP @ U. ( print saved stack pointer )
1355 CFA> ( look up the codeword to get the dictionary entry )
1356 ?DUP IF ( and print it )
1357 2DUP ( dea addr dea )
1358 ID. ( print word from dictionary entry )
1359 [ CHAR + ] LITERAL EMIT
1360 SWAP >DFA 4+ - . ( print offset )
1363 4+ ( move up the stack )
1370 C STRINGS ----------------------------------------------------------------------
1372 FORTH strings are represented by a start address and length kept on the stack or in memory.
1374 Most FORTHs don't handle C strings, but we need them in order to access the process arguments
1375 and environment left on the stack by the Linux kernel, and to make some system calls.
1377 Operation Input Output FORTH word Notes
1378 ----------------------------------------------------------------------
1380 Create FORTH string addr len S" ..."
1382 Create C string c-addr Z" ..."
1384 C -> FORTH c-addr addr len DUP STRLEN
1386 FORTH -> C addr len c-addr CSTRING Allocated in a temporary buffer, so
1387 should be consumed / copied immediately.
1388 FORTH string should not contain NULs.
1390 For example, DUP STRLEN TELL prints a C string.
1394 Z" .." is like S" ..." except that the string is terminated by an ASCII NUL character.
1396 To make it more like a C string, at runtime Z" just leaves the address of the string
1397 on the stack (not address & length as with S"). To implement this we need to add the
1398 extra NUL to the string and also a DROP instruction afterwards. Apart from that the
1399 implementation just a modified S".
1402 STATE @ IF ( compiling? )
1403 ' LITSTRING , ( compile LITSTRING )
1404 HERE @ ( save the address of the length word on the stack )
1405 0 , ( dummy length - we don't know what it is yet )
1407 KEY ( get next character of the string )
1410 HERE @ C! ( store the character in the compiled image )
1411 1 HERE +! ( increment HERE pointer by 1 byte )
1413 0 HERE @ C! ( add the ASCII NUL byte )
1415 DROP ( drop the double quote character at the end )
1416 DUP ( get the saved address of the length word )
1417 HERE @ SWAP - ( calculate the length )
1418 4- ( subtract 4 (because we measured from the start of the length word) )
1419 SWAP ! ( and back-fill the length location )
1420 ALIGN ( round up to next multiple of 4 bytes for the remaining code )
1421 ' DROP , ( compile DROP (to drop the length) )
1422 ELSE ( immediate mode )
1423 HERE @ ( get the start address of the temporary space )
1428 OVER C! ( save next character )
1429 1+ ( increment address )
1431 DROP ( drop the final " character )
1432 0 SWAP C! ( store final ASCII NUL )
1433 HERE @ ( push the start address )
1437 : STRLEN ( str -- len )
1438 DUP ( save start address )
1440 DUP C@ 0<> ( zero byte found? )
1445 SWAP - ( calculate the length )
1448 : CSTRING ( addr len -- c-addr )
1449 SWAP OVER ( len saddr len )
1450 HERE @ SWAP ( len saddr daddr len )
1453 HERE @ + ( daddr+len )
1454 0 SWAP C! ( store terminating NUL char )
1456 HERE @ ( push start address )
1460 THE ENVIRONMENT ----------------------------------------------------------------------
1462 Linux makes the process arguments and environment available to us on the stack.
1464 The top of stack pointer is saved by the early assembler code when we start up in the FORTH
1465 variable S0, and starting at this pointer we can read out the command line arguments and the
1468 Starting at S0, S0 itself points to argc (the number of command line arguments).
1470 S0+4 points to argv[0], S0+8 points to argv[1] etc up to argv[argc-1].
1472 argv[argc] is a NULL pointer.
1474 After that the stack contains environment variables, a set of pointers to strings of the
1475 form NAME=VALUE and on until we get to another NULL pointer.
1477 The first word that we define, ARGC, pushes the number of command line arguments (note that
1478 as with C argc, this includes the name of the command).
1485 n ARGV gets the nth command line argument.
1487 For example to print the command name you would do:
1490 : ARGV ( n -- str u )
1491 1+ CELLS S0 @ + ( get the address of argv[n] entry )
1492 @ ( get the address of the string )
1493 DUP STRLEN ( and get its length / turn it into a FORTH string )
1497 ENVIRON returns the address of the first environment string. The list of strings ends
1498 with a NULL pointer.
1500 For example to print the first string in the environment you could do:
1501 ENVIRON @ DUP STRLEN TELL
1503 : ENVIRON ( -- addr )
1504 ARGC ( number of command line parameters on the stack to skip )
1505 2 + ( skip command line count and NULL pointer after the command line args )
1506 CELLS ( convert to an offset )
1507 S0 @ + ( add to base stack address )
1511 SYSTEM CALLS AND FILES ----------------------------------------------------------------------
1513 Miscellaneous words related to system calls, and standard access to files.
1516 ( BYE exits by calling the Linux exit(2) syscall. )
1518 0 ( return code (0) )
1519 SYS_EXIT ( system call number )
1524 UNUSED returns the number of cells remaining in the user memory (data segment).
1526 For our implementation we will use Linux brk(2) system call to find out the end
1527 of the data segment and subtract HERE from it.
1529 : GET-BRK ( -- brkpoint )
1530 0 SYS_BRK SYSCALL1 ( call brk(0) )
1534 GET-BRK ( get end of data segment according to the kernel )
1535 HERE @ ( get current position in data segment )
1537 4 / ( returns number of cells )
1541 MORECORE increases the data segment by the specified number of (4 byte) cells.
1543 NB. The number of cells requested should normally be a multiple of 1024. The
1544 reason is that Linux can't extend the data segment by less than a single page
1545 (4096 bytes or 1024 cells).
1547 This FORTH doesn't automatically increase the size of the data segment "on demand"
1548 (ie. when , (COMMA), ALLOT, CREATE, and so on are used). Instead the programmer
1549 needs to be aware of how much space a large allocation will take, check UNUSED, and
1550 call MORECORE if necessary. A simple programming exercise is to change the
1551 implementation of the data segment so that MORECORE is called automatically if
1552 the program needs more memory.
1554 : BRK ( brkpoint -- )
1558 : MORECORE ( cells -- )
1563 Standard FORTH provides some simple file access primitives which we model on
1564 top of Linux syscalls.
1566 The main complication is converting FORTH strings (address & length) into C
1567 strings for the Linux kernel.
1569 Notice there is no buffering in this implementation.
1572 : R/O ( -- fam ) O_RDONLY ;
1573 : R/W ( -- fam ) O_RDWR ;
1575 : OPEN-FILE ( addr u fam -- fd 0 (if successful) | c-addr u fam -- fd errno (if there was an error) )
1577 CSTRING ( fam cstring )
1578 SYS_OPEN SYSCALL2 ( open (filename, flags) )
1580 DUP 0< IF ( errno? )
1587 : CREATE-FILE ( addr u fam -- fd 0 (if successful) | c-addr u fam -- fd errno (if there was an error) )
1591 CSTRING ( fam cstring )
1592 420 ROT ( 0644 fam cstring )
1593 SYS_OPEN SYSCALL3 ( open (filename, flags|O_TRUNC|O_CREAT, 0644) )
1595 DUP 0< IF ( errno? )
1602 : CLOSE-FILE ( fd -- 0 (if successful) | fd -- errno (if there was an error) )
1607 : READ-FILE ( addr u fd -- u2 0 (if successful) | addr u fd -- 0 0 (if EOF) | addr u fd -- u2 errno (if error) )
1608 ROT SWAP -ROT ( u addr fd )
1612 DUP 0< IF ( errno? )
1620 PERROR prints a message for an errno, similar to C's perror(3) but we don't have the extensive
1621 list of strerror strings available, so all we can do is print the errno.
1623 : PERROR ( errno addr u -- )
1631 ASSEMBLER CODE ----------------------------------------------------------------------
1633 This is just the outline of a simple assembler, allowing you to write FORTH primitives
1634 in assembly language.
1636 Assembly primitives begin ': NAME' in the normal way, but are ended with ;CODE. ;CODE
1637 updates the header so that the codeword isn't DOCOL, but points instead to the assembled
1638 code (in the DFA part of the word).
1640 We provide a convenience macro NEXT (you guessed the rest).
1642 The rest consists of some immediate words which expand into machine code appended to the
1643 definition of the word. Only a very tiny part of the i386 assembly space is covered, just
1644 enough to write a few assembler primitives below.
1648 ALIGN ( machine code is assembled in bytes so isn't necessarily aligned at the end )
1650 HIDDEN ( unhide the word )
1651 DUP >DFA SWAP >CFA ! ( change the codeword to point to the data area )
1652 [COMPILE] [ ( go back to immediate mode )
1657 ( Equivalent to the NEXT macro )
1658 : NEXT IMMEDIATE AD C, FF C, 20 C, ;
1660 ( The i386 registers )
1670 ( i386 stack instructions )
1671 : PUSH IMMEDIATE 50 + C, ;
1672 : POP IMMEDIATE 58 + C, ;
1674 ( RDTSC instruction )
1675 : RDTSC IMMEDIATE 0F C, 31 C, ;
1680 RDTSC is an assembler primitive which reads the Pentium timestamp counter (a very fine-
1681 grained counter which counts processor clock cycles). Because the TSC is 64 bits wide
1682 we have to push it onto the stack in two slots.
1684 : RDTSC ( -- lsb msb )
1685 RDTSC ( writes the result in %edx:%eax )
1686 EAX PUSH ( push lsb )
1687 EDX PUSH ( push msb )
1692 NOTES ----------------------------------------------------------------------
1694 DOES> isn't possible to implement with this FORTH because we don't have a separate
1699 WELCOME MESSAGE ----------------------------------------------------------------------
1701 Print the version and OK prompt.
1705 S" TEST-MODE" FIND NOT IF
1706 ." JONESFORTH VERSION " VERSION . CR
1707 UNUSED . ." CELLS REMAINING" CR