File:  [gforth] / gforth / Attic / gforth.ds
Revision 1.12: download - view: text, annotated - select for diffs
Mon Mar 13 09:17:26 1995 UTC (24 years, 8 months ago) by anton
Branches: MAIN
CVS tags: HEAD
Added stuff for documenting Forth source and integrating it into the texi file
changed checks for DOMAINOS to checks for apollo (which is defined on apollos)
changed "-evaluate" (which did not work anyway) to "--evaluate"
added debugging.fs and assert.fs to startup.fs

    1: \input texinfo   @c -*-texinfo-*-
    2: @comment The source is gforth.ds, from which gforth.texi is generated
    3: @comment %**start of header (This is for running Texinfo on a region.)
    4: @setfilename gforth.info
    5: @settitle GNU Forth Manual
    6: @comment @setchapternewpage odd
    7: @comment %**end of header (This is for running Texinfo on a region.)
    8: 
    9: @ifinfo
   10: This file documents GNU Forth 0.0
   11: 
   12: Copyright @copyright{} 1994 GNU Forth Development Group
   13: 
   14:      Permission is granted to make and distribute verbatim copies of
   15:      this manual provided the copyright notice and this permission notice
   16:      are preserved on all copies.
   17:      
   18: @ignore
   19:      Permission is granted to process this file through TeX and print the
   20:      results, provided the printed document carries a copying permission
   21:      notice identical to this one except for the removal of this paragraph
   22:      (this paragraph not being relevant to the printed manual).
   23:      
   24: @end ignore
   25:      Permission is granted to copy and distribute modified versions of this
   26:      manual under the conditions for verbatim copying, provided also that the
   27:      sections entitled "Distribution" and "General Public License" are
   28:      included exactly as in the original, and provided that the entire
   29:      resulting derived work is distributed under the terms of a permission
   30:      notice identical to this one.
   31:      
   32:      Permission is granted to copy and distribute translations of this manual
   33:      into another language, under the above conditions for modified versions,
   34:      except that the sections entitled "Distribution" and "General Public
   35:      License" may be included in a translation approved by the author instead
   36:      of in the original English.
   37: @end ifinfo
   38: 
   39: @titlepage
   40: @sp 10
   41: @center @titlefont{GNU Forth Manual}
   42: @sp 2
   43: @center for version 0.0
   44: @sp 2
   45: @center Anton Ertl
   46: 
   47: @comment  The following two commands start the copyright page.
   48: @page
   49: @vskip 0pt plus 1filll
   50: Copyright @copyright{} 1994 GNU Forth Development Group
   51: 
   52: @comment !! Published by ... or You can get a copy of this manual ...
   53: 
   54:      Permission is granted to make and distribute verbatim copies of
   55:      this manual provided the copyright notice and this permission notice
   56:      are preserved on all copies.
   57:      
   58:      Permission is granted to copy and distribute modified versions of this
   59:      manual under the conditions for verbatim copying, provided also that the
   60:      sections entitled "Distribution" and "General Public License" are
   61:      included exactly as in the original, and provided that the entire
   62:      resulting derived work is distributed under the terms of a permission
   63:      notice identical to this one.
   64:      
   65:      Permission is granted to copy and distribute translations of this manual
   66:      into another language, under the above conditions for modified versions,
   67:      except that the sections entitled "Distribution" and "General Public
   68:      License" may be included in a translation approved by the author instead
   69:      of in the original English.
   70: @end titlepage
   71: 
   72: 
   73: @node Top, License, (dir), (dir)
   74: @ifinfo
   75: GNU Forth is a free implementation of ANS Forth available on many
   76: personal machines. This manual corresponds to version 0.0.
   77: @end ifinfo
   78: 
   79: @menu
   80: * License::                     
   81: * Goals::                       About the GNU Forth Project
   82: * Other Books::                 Things you might want to read
   83: * Invocation::                  Starting GNU Forth
   84: * Words::                       Forth words available in GNU Forth
   85: * ANS conformance::             Implementation-defined options etc.
   86: * Model::                       The abstract machine of GNU Forth
   87: * Emacs and GForth::            The GForth Mode
   88: * Internals::                   Implementation details
   89: * Bugs::                        How to report them
   90: * Pedigree::                    Ancestors of GNU Forth
   91: * Word Index::                  An item for each Forth word
   92: * Node Index::                  An item for each node
   93: @end menu
   94: 
   95: @node License, Goals, Top, Top
   96: @unnumbered License
   97: !! Insert GPL here
   98: 
   99: @iftex
  100: @unnumbered Preface
  101: This manual documents GNU Forth. The reader is expected to know
  102: Forth. This manual is primarily a reference manual. @xref{Other Books}
  103: for introductory material.
  104: @end iftex
  105: 
  106: @node    Goals, Other Books, License, Top
  107: @comment node-name,     next,           previous, up
  108: @chapter Goals of GNU Forth
  109: @cindex Goals
  110: The goal of the GNU Forth Project is to develop a standard model for
  111: ANSI Forth. This can be split into several subgoals:
  112: 
  113: @itemize @bullet
  114: @item
  115: GNU Forth should conform to the ANSI Forth standard.
  116: @item
  117: It should be a model, i.e. it should define all the
  118: implementation-dependent things.
  119: @item
  120: It should become standard, i.e. widely accepted and used. This goal
  121: is the most difficult one.
  122: @end itemize
  123: 
  124: To achieve these goals GNU Forth should be
  125: @itemize @bullet
  126: @item
  127: Similar to previous models (fig-Forth, F83)
  128: @item
  129: Powerful. It should provide for all the things that are considered
  130: necessary today and even some that are not yet considered necessary.
  131: @item
  132: Efficient. It should not get the reputation of being exceptionally
  133: slow.
  134: @item
  135: Free.
  136: @item
  137: Available on many machines/easy to port.
  138: @end itemize
  139: 
  140: Have we achieved these goals? GNU Forth conforms to the ANS Forth
  141: standard; it may be considered a model, but we have not yet documented
  142: which parts of the model are stable and which parts we are likely to
  143: change; it certainly has not yet become a de facto standard. It has some
  144: similarities and some differences to previous models; It has some
  145: powerful features, but not yet everything that we envisioned; on RISCs
  146: it is as fast as interpreters programmed in assembly, on
  147: register-starved machines it is not so fast, but still faster than any
  148: other C-based interpretive implementation; it is free and available on
  149: many machines.
  150: 
  151: @node Other Books, Invocation, Goals, Top
  152: @chapter Other books on ANS Forth
  153: 
  154: As the standard is relatively new, there are not many books out yet. It
  155: is not recommended to learn Forth by using GNU Forth and a book that is
  156: not written for ANS Forth, as you will not know your mistakes from the
  157: deviations of the book.
  158: 
  159: There is, of course, the standard, the definite reference if you want to
  160: write ANS Forth programs. It will be available in printed form from
  161: Global Engineering Documents !! somtime in spring or summer 1994. If you
  162: are lucky, you can still get dpANS6 (the draft that was approved as
  163: standard) by aftp from ftp.uu.net:/vendor/minerva/x3j14.
  164: 
  165: @cite{Forth: The new model} by Jack Woehr (!! Publisher) is an
  166: introductory book based on a draft version of the standard. It does not
  167: cover the whole standard. It also contains interesting background
  168: information (Jack Woehr was in the ANS Forth Technical Committe). It is
  169: not appropriate for complete newbies, but programmers experienced in
  170: other languages should find it ok.
  171: 
  172: @node Invocation, Words, Other Books, Top
  173: @chapter Invocation
  174: 
  175: You will usually just say @code{gforth}. In many other cases the default
  176: GNU Forth image will be invoked like this:
  177: 
  178: @example
  179: gforth [files] [-e forth-code]
  180: @end example
  181: 
  182: executing the contents of the files and the Forth code in the order they
  183: are given.
  184: 
  185: In general, the command line looks like this:
  186: 
  187: @example
  188: gforth [initialization options] [image-specific options]
  189: @end example
  190: 
  191: The initialization options must come before the rest of the command
  192: line. They are:
  193: 
  194: @table @code
  195: @item --image-file @var{file}
  196: Loads the Forth image @var{file} instead of the default
  197: @file{gforth.fi}.
  198: 
  199: @item --path @var{path}
  200: Uses @var{path} for searching the image file and Forth source code
  201: files instead of the default in the environment variable
  202: @code{GFORTHPATH} or the path specified at installation time (typically
  203: @file{/usr/local/lib/gforth:.}). A path is given as a @code{:}-separated
  204: list.
  205: 
  206: @item --dictionary-size @var{size}
  207: @item -m @var{size}
  208: Allocate @var{size} space for the Forth dictionary space instead of
  209: using the default specified in the image (typically 256K). The
  210: @var{size} specification consists of an integer and a unit (e.g.,
  211: @code{4M}). The unit can be one of @code{b} (bytes), @code{e} (element
  212: size, in this case Cells), @code{k} (kilobytes), and @code{M}
  213: (Megabytes). If no unit is specified, @code{e} is used.
  214: 
  215: @item --data-stack-size @var{size}
  216: @item -d @var{size}
  217: Allocate @var{size} space for the data stack instead of using the
  218: default specified in the image (typically 16K).
  219: 
  220: @item --return-stack-size @var{size}
  221: @item -r @var{size}
  222: Allocate @var{size} space for the return stack instead of using the
  223: default specified in the image (typically 16K).
  224: 
  225: @item --fp-stack-size @var{size}
  226: @item -f @var{size}
  227: Allocate @var{size} space for the floating point stack instead of
  228: using the default specified in the image (typically 16K). In this case
  229: the unit specifier @code{e} refers to floating point numbers.
  230: 
  231: @item --locals-stack-size @var{size}
  232: @item -l @var{size}
  233: Allocate @var{size} space for the locals stack instead of using the
  234: default specified in the image (typically 16K).
  235: 
  236: @end table
  237: 
  238: As explained above, the image-specific command-line arguments for the
  239: default image @file{gforth.fi} consist of a sequence of filenames and
  240: @code{-e @var{forth-code}} options that are interpreted in the seqence
  241: in which they are given. The @code{-e @var{forth-code}} or
  242: @code{--evaluate @var{forth-code}} option evaluates the forth
  243: code. This option takes only one argument; if you want to evaluate more
  244: Forth words, you have to quote them or use several @code{-e}s. To exit
  245: after processing the command line (instead of entering interactive mode)
  246: append @code{-e bye} to the command line.
  247: 
  248: Not yet implemented:
  249: On startup the system first executes the system initialization file
  250: (unless the option @code{--no-init-file} is given; note that the system
  251: resulting from using this option may not be ANS Forth conformant). Then
  252: the user initialization file @file{.gforth.fs} is executed, unless the
  253: option @code{--no-rc} is given; this file is first searched in @file{.},
  254: then in @file{~}, then in the normal path (see above).
  255: 
  256: @node Words, ANS conformance, Invocation, Top
  257: @chapter Forth Words
  258: 
  259: @menu
  260: * Notation::                    
  261: * Arithmetic::                  
  262: * Stack Manipulation::          
  263: * Memory access::               
  264: * Control Structures::          
  265: * Locals::                      
  266: * Defining Words::              
  267: * Wordlists::                   
  268: * Files::                       
  269: * Blocks::                      
  270: * Other I/O::                   
  271: * Programming Tools::           
  272: * Threading Words::             
  273: @end menu
  274: 
  275: @node Notation, Arithmetic, Words, Words
  276: @section Notation
  277: 
  278: The Forth words are described in this section in the glossary notation
  279: that has become a de-facto standard for Forth texts, i.e.
  280: 
  281: @format
  282: @var{word}     @var{Stack effect}   @var{wordset}   @var{pronunciation}
  283: @end format
  284: @var{Description}
  285: 
  286: @table @var
  287: @item word
  288: The name of the word. BTW, GNU Forth is case insensitive, so you can
  289: type the words in in lower case.
  290: 
  291: @item Stack effect
  292: The stack effect is written in the notation @code{@var{before} --
  293: @var{after}}, where @var{before} and @var{after} describe the top of
  294: stack entries before and after the execution of the word. The rest of
  295: the stack is not touched by the word. The top of stack is rightmost,
  296: i.e., a stack sequence is written as it is typed in. Note that GNU Forth
  297: uses a separate floating point stack, but a unified stack
  298: notation. Also, return stack effects are not shown in @var{stack
  299: effect}, but in @var{Description}. The name of a stack item describes
  300: the type and/or the function of the item. See below for a discussion of
  301: the types.
  302: 
  303: @item pronunciation
  304: How the word is pronounced
  305: 
  306: @item wordset
  307: The ANS Forth standard is divided into several wordsets. A standard
  308: system need not support all of them. So, the fewer wordsets your program
  309: uses the more portable it will be in theory. However, we suspect that
  310: most ANS Forth systems on personal machines will feature all
  311: wordsets. Words that are not defined in the ANS standard have
  312: @code{gforth} as wordset.
  313: 
  314: @item Description
  315: A description of the behaviour of the word.
  316: @end table
  317: 
  318: The type of a stack item is specified by the character(s) the name
  319: starts with:
  320: 
  321: @table @code
  322: @item f
  323: Bool, i.e. @code{false} or @code{true}.
  324: @item c
  325: Char
  326: @item w
  327: Cell, can contain an integer or an address
  328: @item n
  329: signed integer
  330: @item u
  331: unsigned integer
  332: @item d
  333: double sized signed integer
  334: @item ud
  335: double sized unsigned integer
  336: @item r
  337: Float
  338: @item a_
  339: Cell-aligned address
  340: @item c_
  341: Char-aligned address (note that a Char is two bytes in Windows NT)
  342: @item f_
  343: Float-aligned address
  344: @item df_
  345: Address aligned for IEEE double precision float
  346: @item sf_
  347: Address aligned for IEEE single precision float
  348: @item xt
  349: Execution token, same size as Cell
  350: @item wid
  351: Wordlist ID, same size as Cell
  352: @item f83name
  353: Pointer to a name structure
  354: @end table
  355: 
  356: @node Arithmetic, Stack Manipulation, Notation, Words
  357: @section Arithmetic
  358: Forth arithmetic is not checked, i.e., you will not hear about integer
  359: overflow on addition or multiplication, you may hear about division by
  360: zero if you are lucky. The operator is written after the operands, but
  361: the operands are still in the original order. I.e., the infix @code{2-1}
  362: corresponds to @code{2 1 -}. Forth offers a variety of division
  363: operators. If you perform division with potentially negative operands,
  364: you do not want to use @code{/} or @code{/mod} with its undefined
  365: behaviour, but rather @code{fm/mod} or @code{sm/mod} (probably the
  366: former, @pxref{Mixed precision}).
  367: 
  368: @menu
  369: * Single precision::            
  370: * Bitwise operations::          
  371: * Mixed precision::             operations with single and double-cell integers
  372: * Double precision::            Double-cell integer arithmetic
  373: * Floating Point::              
  374: @end menu
  375: 
  376: @node Single precision, Bitwise operations, Arithmetic, Arithmetic
  377: @subsection Single precision
  378: doc-+
  379: doc--
  380: doc-*
  381: doc-/
  382: doc-mod
  383: doc-/mod
  384: doc-negate
  385: doc-abs
  386: doc-min
  387: doc-max
  388: 
  389: @node Bitwise operations, Mixed precision, Single precision, Arithmetic
  390: @subsection Bitwise operations
  391: doc-and
  392: doc-or
  393: doc-xor
  394: doc-invert
  395: doc-2*
  396: doc-2/
  397: 
  398: @node Mixed precision, Double precision, Bitwise operations, Arithmetic
  399: @subsection Mixed precision
  400: doc-m+
  401: doc-*/
  402: doc-*/mod
  403: doc-m*
  404: doc-um*
  405: doc-m*/
  406: doc-um/mod
  407: doc-fm/mod
  408: doc-sm/rem
  409: 
  410: @node Double precision, Floating Point, Mixed precision, Arithmetic
  411: @subsection Double precision
  412: doc-d+
  413: doc-d-
  414: doc-dnegate
  415: doc-dabs
  416: doc-dmin
  417: doc-dmax
  418: 
  419: @node Floating Point,  , Double precision, Arithmetic
  420: @subsection Floating Point
  421: 
  422: Angles in floating point operations are given in radians (a full circle
  423: has 2 pi radians). Note, that gforth has a separate floating point
  424: stack, but we use the unified notation.
  425: 
  426: Floating point numbers have a number of unpleasant surprises for the
  427: unwary (e.g., floating point addition is not associative) and even a few
  428: for the wary. You should not use them unless you know what you are doing
  429: or you don't care that the results you get are totally bogus. If you
  430: want to learn about the problems of floating point numbers (and how to
  431: avoid them), you might start with @cite{David Goldberg, What Every
  432: Computer Scientist Should Know About Floating-Point Arithmetic, ACM
  433: Computing Surveys 23(1):5@minus{}48, March 1991}.
  434: 
  435: doc-f+
  436: doc-f-
  437: doc-f*
  438: doc-f/
  439: doc-fnegate
  440: doc-fabs
  441: doc-fmax
  442: doc-fmin
  443: doc-floor
  444: doc-fround
  445: doc-f**
  446: doc-fsqrt
  447: doc-fexp
  448: doc-fexpm1
  449: doc-fln
  450: doc-flnp1
  451: doc-flog
  452: doc-falog
  453: doc-fsin
  454: doc-fcos
  455: doc-fsincos
  456: doc-ftan
  457: doc-fasin
  458: doc-facos
  459: doc-fatan
  460: doc-fatan2
  461: doc-fsinh
  462: doc-fcosh
  463: doc-ftanh
  464: doc-fasinh
  465: doc-facosh
  466: doc-fatanh
  467: 
  468: @node Stack Manipulation, Memory access, Arithmetic, Words
  469: @section Stack Manipulation
  470: 
  471: gforth has a data stack (aka parameter stack) for characters, cells,
  472: addresses, and double cells, a floating point stack for floating point
  473: numbers, a return stack for storing the return addresses of colon
  474: definitions and other data, and a locals stack for storing local
  475: variables. Note that while every sane Forth has a separate floating
  476: point stack, this is not strictly required; an ANS Forth system could
  477: theoretically keep floating point numbers on the data stack. As an
  478: additional difficulty, you don't know how many cells a floating point
  479: number takes. It is reportedly possible to write words in a way that
  480: they work also for a unified stack model, but we do not recommend trying
  481: it. Instead, just say that your program has an environmental dependency
  482: on a separate FP stack.
  483: 
  484: Also, a Forth system is allowed to keep the local variables on the
  485: return stack. This is reasonable, as local variables usually eliminate
  486: the need to use the return stack explicitly. So, if you want to produce
  487: a standard complying program and if you are using local variables in a
  488: word, forget about return stack manipulations in that word (see the
  489: standard document for the exact rules).
  490: 
  491: @menu
  492: * Data stack::                  
  493: * Floating point stack::        
  494: * Return stack::                
  495: * Locals stack::                
  496: * Stack pointer manipulation::  
  497: @end menu
  498: 
  499: @node Data stack, Floating point stack, Stack Manipulation, Stack Manipulation
  500: @subsection Data stack
  501: doc-drop
  502: doc-nip
  503: doc-dup
  504: doc-over
  505: doc-tuck
  506: doc-swap
  507: doc-rot
  508: doc--rot
  509: doc-?dup
  510: doc-pick
  511: doc-roll
  512: doc-2drop
  513: doc-2nip
  514: doc-2dup
  515: doc-2over
  516: doc-2tuck
  517: doc-2swap
  518: doc-2rot
  519: 
  520: @node Floating point stack, Return stack, Data stack, Stack Manipulation
  521: @subsection Floating point stack
  522: doc-fdrop
  523: doc-fnip
  524: doc-fdup
  525: doc-fover
  526: doc-ftuck
  527: doc-fswap
  528: doc-frot
  529: 
  530: @node Return stack, Locals stack, Floating point stack, Stack Manipulation
  531: @subsection Return stack
  532: doc->r
  533: doc-r>
  534: doc-r@
  535: doc-rdrop
  536: doc-2>r
  537: doc-2r>
  538: doc-2r@
  539: doc-2rdrop
  540: 
  541: @node Locals stack, Stack pointer manipulation, Return stack, Stack Manipulation
  542: @subsection Locals stack
  543: 
  544: @node Stack pointer manipulation,  , Locals stack, Stack Manipulation
  545: @subsection Stack pointer manipulation
  546: doc-sp@
  547: doc-sp!
  548: doc-fp@
  549: doc-fp!
  550: doc-rp@
  551: doc-rp!
  552: doc-lp@
  553: doc-lp!
  554: 
  555: @node Memory access, Control Structures, Stack Manipulation, Words
  556: @section Memory access
  557: 
  558: @menu
  559: * Stack-Memory transfers::      
  560: * Address arithmetic::          
  561: * Memory block access::         
  562: @end menu
  563: 
  564: @node Stack-Memory transfers, Address arithmetic, Memory access, Memory access
  565: @subsection Stack-Memory transfers
  566: 
  567: doc-@
  568: doc-!
  569: doc-+!
  570: doc-c@
  571: doc-c!
  572: doc-2@
  573: doc-2!
  574: doc-f@
  575: doc-f!
  576: doc-sf@
  577: doc-sf!
  578: doc-df@
  579: doc-df!
  580: 
  581: @node Address arithmetic, Memory block access, Stack-Memory transfers, Memory access
  582: @subsection Address arithmetic
  583: 
  584: ANS Forth does not specify the sizes of the data types. Instead, it
  585: offers a number of words for computing sizes and doing address
  586: arithmetic. Basically, address arithmetic is performed in terms of
  587: address units (aus); on most systems the address unit is one byte. Note
  588: that a character may have more than one au, so @code{chars} is no noop
  589: (on systems where it is a noop, it compiles to nothing).
  590: 
  591: ANS Forth also defines words for aligning addresses for specific
  592: addresses. Many computers require that accesses to specific data types
  593: must only occur at specific addresses; e.g., that cells may only be
  594: accessed at addresses divisible by 4. Even if a machine allows unaligned
  595: accesses, it can usually perform aligned accesses faster. 
  596: 
  597: For the performance-concious: alignment operations are usually only
  598: necessary during the definition of a data structure, not during the
  599: (more frequent) accesses to it.
  600: 
  601: ANS Forth defines no words for character-aligning addresses. This is not
  602: an oversight, but reflects the fact that addresses that are not
  603: char-aligned have no use in the standard and therefore will not be
  604: created.
  605: 
  606: The standard guarantees that addresses returned by @code{CREATE}d words
  607: are cell-aligned; in addition, gforth guarantees that these addresses
  608: are aligned for all purposes.
  609: 
  610: Note that the standard defines a word @code{char}, which has nothing to
  611: do with address arithmetic.
  612: 
  613: doc-chars
  614: doc-char+
  615: doc-cells
  616: doc-cell+
  617: doc-align
  618: doc-aligned
  619: doc-floats
  620: doc-float+
  621: doc-falign
  622: doc-faligned
  623: doc-sfloats
  624: doc-sfloat+
  625: doc-sfalign
  626: doc-sfaligned
  627: doc-dfloats
  628: doc-dfloat+
  629: doc-dfalign
  630: doc-dfaligned
  631: doc-maxalign
  632: doc-maxaligned
  633: doc-cfalign
  634: doc-cfaligned
  635: doc-address-unit-bits
  636: 
  637: @node Memory block access,  , Address arithmetic, Memory access
  638: @subsection Memory block access
  639: 
  640: doc-move
  641: doc-erase
  642: 
  643: While the previous words work on address units, the rest works on
  644: characters.
  645: 
  646: doc-cmove
  647: doc-cmove>
  648: doc-fill
  649: doc-blank
  650: 
  651: @node Control Structures, Locals, Memory access, Words
  652: @section Control Structures
  653: 
  654: Control structures in Forth cannot be used in interpret state, only in
  655: compile state, i.e., in a colon definition. We do not like this
  656: limitation, but have not seen a satisfying way around it yet, although
  657: many schemes have been proposed.
  658: 
  659: @menu
  660: * Selection::                   
  661: * Simple Loops::                
  662: * Counted Loops::               
  663: * Arbitrary control structures::  
  664: * Calls and returns::           
  665: * Exception Handling::          
  666: @end menu
  667: 
  668: @node Selection, Simple Loops, Control Structures, Control Structures
  669: @subsection Selection
  670: 
  671: @example
  672: @var{flag}
  673: IF
  674:   @var{code}
  675: ENDIF
  676: @end example
  677: or
  678: @example
  679: @var{flag}
  680: IF
  681:   @var{code1}
  682: ELSE
  683:   @var{code2}
  684: ENDIF
  685: @end example
  686: 
  687: You can use @code{THEN} instead of @code{ENDIF}. Indeed, @code{THEN} is
  688: standard, and @code{ENDIF} is not, although it is quite popular. We
  689: recommend using @code{ENDIF}, because it is less confusing for people
  690: who also know other languages (and is not prone to reinforcing negative
  691: prejudices against Forth in these people). Adding @code{ENDIF} to a
  692: system that only supplies @code{THEN} is simple:
  693: @example
  694: : endif   POSTPONE then ; immediate
  695: @end example
  696: 
  697: [According to @cite{Webster's New Encyclopedic Dictionary}, @dfn{then
  698: (adv.)}  has the following meanings:
  699: @quotation
  700: ... 2b: following next after in order ... 3d: as a necessary consequence
  701: (if you were there, then you saw them).
  702: @end quotation
  703: Forth's @code{THEN} has the meaning 2b, whereas @code{THEN} in Pascal
  704: and many other programming languages has the meaning 3d.]
  705: 
  706: We also provide the words @code{?dup-if} and @code{?dup-0=-if}, so you
  707: can avoid using @code{?dup}.
  708: 
  709: @example
  710: @var{n}
  711: CASE
  712:   @var{n1} OF @var{code1} ENDOF
  713:   @var{n2} OF @var{code2} ENDOF
  714:   @dots{}
  715: ENDCASE
  716: @end example
  717: 
  718: Executes the first @var{codei}, where the @var{ni} is equal to
  719: @var{n}. A default case can be added by simply writing the code after
  720: the last @code{ENDOF}. It may use @var{n}, which is on top of the stack,
  721: but must not consume it.
  722: 
  723: @node Simple Loops, Counted Loops, Selection, Control Structures
  724: @subsection Simple Loops
  725: 
  726: @example
  727: BEGIN
  728:   @var{code1}
  729:   @var{flag}
  730: WHILE
  731:   @var{code2}
  732: REPEAT
  733: @end example
  734: 
  735: @var{code1} is executed and @var{flag} is computed. If it is true,
  736: @var{code2} is executed and the loop is restarted; If @var{flag} is false, execution continues after the @code{REPEAT}.
  737: 
  738: @example
  739: BEGIN
  740:   @var{code}
  741:   @var{flag}
  742: UNTIL
  743: @end example
  744: 
  745: @var{code} is executed. The loop is restarted if @code{flag} is false.
  746: 
  747: @example
  748: BEGIN
  749:   @var{code}
  750: AGAIN
  751: @end example
  752: 
  753: This is an endless loop.
  754: 
  755: @node Counted Loops, Arbitrary control structures, Simple Loops, Control Structures
  756: @subsection Counted Loops
  757: 
  758: The basic counted loop is:
  759: @example
  760: @var{limit} @var{start}
  761: ?DO
  762:   @var{body}
  763: LOOP
  764: @end example
  765: 
  766: This performs one iteration for every integer, starting from @var{start}
  767: and up to, but excluding @var{limit}. The counter, aka index, can be
  768: accessed with @code{i}. E.g., the loop
  769: @example
  770: 10 0 ?DO
  771:   i .
  772: LOOP
  773: @end example
  774: prints
  775: @example
  776: 0 1 2 3 4 5 6 7 8 9
  777: @end example
  778: The index of the innermost loop can be accessed with @code{i}, the index
  779: of the next loop with @code{j}, and the index of the third loop with
  780: @code{k}.
  781: 
  782: The loop control data are kept on the return stack, so there are some
  783: restrictions on mixing return stack accesses and counted loop
  784: words. E.g., if you put values on the return stack outside the loop, you
  785: cannot read them inside the loop. If you put values on the return stack
  786: within a loop, you have to remove them before the end of the loop and
  787: before accessing the index of the loop.
  788: 
  789: There are several variations on the counted loop:
  790: 
  791: @code{LEAVE} leaves the innermost counted loop immediately.
  792: 
  793: @code{LOOP} can be replaced with @code{@var{n} +LOOP}; this updates the
  794: index by @var{n} instead of by 1. The loop is terminated when the border
  795: between @var{limit-1} and @var{limit} is crossed. E.g.:
  796: 
  797: @code{4 0 ?DO  i .  2 +LOOP}   prints @code{0 2}
  798: 
  799: @code{4 1 ?DO  i .  2 +LOOP}   prints @code{1 3}
  800: 
  801: The behaviour of @code{@var{n} +LOOP} is peculiar when @var{n} is negative:
  802: 
  803: @code{-1 0 ?DO  i .  -1 +LOOP}  prints @code{0 -1}
  804: 
  805: @code{ 0 0 ?DO  i .  -1 +LOOP}  prints nothing
  806: 
  807: Therefore we recommend avoiding using @code{@var{n} +LOOP} with negative
  808: @var{n}. One alternative is @code{@var{n} S+LOOP}, where the negative
  809: case behaves symmetrical to the positive case:
  810: 
  811: @code{-2 0 ?DO  i .  -1 S+LOOP}  prints @code{0 -1}
  812: 
  813: @code{-1 0 ?DO  i .  -1 S+LOOP}  prints @code{0}
  814: 
  815: @code{ 0 0 ?DO  i .  -1 S+LOOP}  prints nothing
  816: 
  817: The loop is terminated when the border between @var{limit@minus{}sgn(n)} and
  818: @var{limit} is crossed. However, @code{S+LOOP} is not part of the ANS
  819: Forth standard.
  820: 
  821: @code{?DO} can be replaced by @code{DO}. @code{DO} enters the loop even
  822: when the start and the limit value are equal. We do not recommend using
  823: @code{DO}. It will just give you maintenance troubles.
  824: 
  825: @code{UNLOOP} is used to prepare for an abnormal loop exit, e.g., via
  826: @code{EXIT}. @code{UNLOOP} removes the loop control parameters from the
  827: return stack so @code{EXIT} can get to its return address.
  828: 
  829: Another counted loop is
  830: @example
  831: @var{n}
  832: FOR
  833:   @var{body}
  834: NEXT
  835: @end example
  836: This is the preferred loop of native code compiler writers who are too
  837: lazy to optimize @code{?DO} loops properly. In GNU Forth, this loop
  838: iterates @var{n+1} times; @code{i} produces values starting with @var{n}
  839: and ending with 0. Other Forth systems may behave differently, even if
  840: they support @code{FOR} loops.
  841: 
  842: @node Arbitrary control structures, Calls and returns, Counted Loops, Control Structures
  843: @subsection Arbitrary control structures
  844: 
  845: ANS Forth permits and supports using control structures in a non-nested
  846: way. Information about incomplete control structures is stored on the
  847: control-flow stack. This stack may be implemented on the Forth data
  848: stack, and this is what we have done in gforth.
  849: 
  850: An @i{orig} entry represents an unresolved forward branch, a @i{dest}
  851: entry represents a backward branch target. A few words are the basis for
  852: building any control structure possible (except control structures that
  853: need storage, like calls, coroutines, and backtracking).
  854: 
  855: doc-if
  856: doc-ahead
  857: doc-then
  858: doc-begin
  859: doc-until
  860: doc-again
  861: doc-cs-pick
  862: doc-cs-roll
  863: 
  864: On many systems control-flow stack items take one word, in gforth they
  865: currently take three (this may change in the future). Therefore it is a
  866: really good idea to manipulate the control flow stack with
  867: @code{cs-pick} and @code{cs-roll}, not with data stack manipulation
  868: words.
  869: 
  870: Some standard control structure words are built from these words:
  871: 
  872: doc-else
  873: doc-while
  874: doc-repeat
  875: 
  876: Counted loop words constitute a separate group of words:
  877: 
  878: doc-?do
  879: doc-do
  880: doc-for
  881: doc-loop
  882: doc-s+loop
  883: doc-+loop
  884: doc-next
  885: doc-leave
  886: doc-?leave
  887: doc-unloop
  888: doc-done
  889: 
  890: The standard does not allow using @code{cs-pick} and @code{cs-roll} on
  891: @i{do-sys}. Our system allows it, but it's your job to ensure that for
  892: every @code{?DO} etc. there is exactly one @code{UNLOOP} on any path
  893: through the definition (@code{LOOP} etc. compile an @code{UNLOOP} on the
  894: fall-through path). Also, you have to ensure that all @code{LEAVE}s are
  895: resolved (by using one of the loop-ending words or @code{DONE}).
  896: 
  897: Another group of control structure words are
  898: 
  899: doc-case
  900: doc-endcase
  901: doc-of
  902: doc-endof
  903: 
  904: @i{case-sys} and @i{of-sys} cannot be processed using @code{cs-pick} and
  905: @code{cs-roll}.
  906: 
  907: @subsubsection Programming Style
  908: 
  909: In order to ensure readability we recommend that you do not create
  910: arbitrary control structures directly, but define new control structure
  911: words for the control structure you want and use these words in your
  912: program.
  913: 
  914: E.g., instead of writing
  915: 
  916: @example
  917: begin
  918:   ...
  919: if [ 1 cs-roll ]
  920:   ...
  921: again then
  922: @end example
  923: 
  924: we recommend defining control structure words, e.g.,
  925: 
  926: @example
  927: : while ( dest -- orig dest )
  928:  POSTPONE if
  929:  1 cs-roll ; immediate
  930: 
  931: : repeat ( orig dest -- )
  932:  POSTPONE again
  933:  POSTPONE then ; immediate
  934: @end example
  935: 
  936: and then using these to create the control structure:
  937: 
  938: @example
  939: begin
  940:   ...
  941: while
  942:   ...
  943: repeat
  944: @end example
  945: 
  946: That's much easier to read, isn't it? Of course, @code{BEGIN} and
  947: @code{WHILE} are predefined, so in this example it would not be
  948: necessary to define them.
  949: 
  950: @node Calls and returns, Exception Handling, Arbitrary control structures, Control Structures
  951: @subsection Calls and returns
  952: 
  953: A definition can be called simply be writing the name of the
  954: definition. When the end of the definition is reached, it returns. An earlier return can be forced using
  955: 
  956: doc-exit
  957: 
  958: Don't forget to clean up the return stack and @code{UNLOOP} any
  959: outstanding @code{?DO}...@code{LOOP}s before @code{EXIT}ing. The
  960: primitive compiled by @code{EXIT} is
  961: 
  962: doc-;s
  963: 
  964: @node Exception Handling,  , Calls and returns, Control Structures
  965: @subsection Exception Handling
  966: 
  967: doc-catch
  968: doc-throw
  969: 
  970: @node Locals, Defining Words, Control Structures, Words
  971: @section Locals
  972: 
  973: Local variables can make Forth programming more enjoyable and Forth
  974: programs easier to read. Unfortunately, the locals of ANS Forth are
  975: laden with restrictions. Therefore, we provide not only the ANS Forth
  976: locals wordset, but also our own, more powerful locals wordset (we
  977: implemented the ANS Forth locals wordset through our locals wordset).
  978: 
  979: @menu
  980: * gforth locals::               
  981: * ANS Forth locals::            
  982: @end menu
  983: 
  984: @node gforth locals, ANS Forth locals, Locals, Locals
  985: @subsection gforth locals
  986: 
  987: Locals can be defined with
  988: 
  989: @example
  990: @{ local1 local2 ... -- comment @}
  991: @end example
  992: or
  993: @example
  994: @{ local1 local2 ... @}
  995: @end example
  996: 
  997: E.g.,
  998: @example
  999: : max @{ n1 n2 -- n3 @}
 1000:  n1 n2 > if
 1001:    n1
 1002:  else
 1003:    n2
 1004:  endif ;
 1005: @end example
 1006: 
 1007: The similarity of locals definitions with stack comments is intended. A
 1008: locals definition often replaces the stack comment of a word. The order
 1009: of the locals corresponds to the order in a stack comment and everything
 1010: after the @code{--} is really a comment.
 1011: 
 1012: This similarity has one disadvantage: It is too easy to confuse locals
 1013: declarations with stack comments, causing bugs and making them hard to
 1014: find. However, this problem can be avoided by appropriate coding
 1015: conventions: Do not use both notations in the same program. If you do,
 1016: they should be distinguished using additional means, e.g. by position.
 1017: 
 1018: The name of the local may be preceded by a type specifier, e.g.,
 1019: @code{F:} for a floating point value:
 1020: 
 1021: @example
 1022: : CX* @{ F: Ar F: Ai F: Br F: Bi -- Cr Ci @}
 1023: \ complex multiplication
 1024:  Ar Br f* Ai Bi f* f-
 1025:  Ar Bi f* Ai Br f* f+ ;
 1026: @end example
 1027: 
 1028: GNU Forth currently supports cells (@code{W:}, @code{W^}), doubles
 1029: (@code{D:}, @code{D^}), floats (@code{F:}, @code{F^}) and characters
 1030: (@code{C:}, @code{C^}) in two flavours: a value-flavoured local (defined
 1031: with @code{W:}, @code{D:} etc.) produces its value and can be changed
 1032: with @code{TO}. A variable-flavoured local (defined with @code{W^} etc.)
 1033: produces its address (which becomes invalid when the variable's scope is
 1034: left). E.g., the standard word @code{emit} can be defined in therms of
 1035: @code{type} like this:
 1036: 
 1037: @example
 1038: : emit @{ C^ char* -- @}
 1039:     char* 1 type ;
 1040: @end example
 1041: 
 1042: A local without type specifier is a @code{W:} local. Both flavours of
 1043: locals are initialized with values from the data or FP stack.
 1044: 
 1045: Currently there is no way to define locals with user-defined data
 1046: structures, but we are working on it.
 1047: 
 1048: GNU Forth allows defining locals everywhere in a colon definition. This
 1049: poses the following questions:
 1050: 
 1051: @menu
 1052: * Where are locals visible by name?::  
 1053: * How long do locals live? ::   
 1054: * Programming Style::           
 1055: * Implementation::              
 1056: @end menu
 1057: 
 1058: @node Where are locals visible by name?, How long do locals live?, gforth locals, gforth locals
 1059: @subsubsection Where are locals visible by name?
 1060: 
 1061: Basically, the answer is that locals are visible where you would expect
 1062: it in block-structured languages, and sometimes a little longer. If you
 1063: want to restrict the scope of a local, enclose its definition in
 1064: @code{SCOPE}...@code{ENDSCOPE}.
 1065: 
 1066: doc-scope
 1067: doc-endscope
 1068: 
 1069: These words behave like control structure words, so you can use them
 1070: with @code{CS-PICK} and @code{CS-ROLL} to restrict the scope in
 1071: arbitrary ways.
 1072: 
 1073: If you want a more exact answer to the visibility question, here's the
 1074: basic principle: A local is visible in all places that can only be
 1075: reached through the definition of the local@footnote{In compiler
 1076: construction terminology, all places dominated by the definition of the
 1077: local.}. In other words, it is not visible in places that can be reached
 1078: without going through the definition of the local. E.g., locals defined
 1079: in @code{IF}...@code{ENDIF} are visible until the @code{ENDIF}, locals
 1080: defined in @code{BEGIN}...@code{UNTIL} are visible after the
 1081: @code{UNTIL} (until, e.g., a subsequent @code{ENDSCOPE}).
 1082: 
 1083: The reasoning behind this solution is: We want to have the locals
 1084: visible as long as it is meaningful. The user can always make the
 1085: visibility shorter by using explicit scoping. In a place that can
 1086: only be reached through the definition of a local, the meaning of a
 1087: local name is clear. In other places it is not: How is the local
 1088: initialized at the control flow path that does not contain the
 1089: definition? Which local is meant, if the same name is defined twice in
 1090: two independent control flow paths?
 1091: 
 1092: This should be enough detail for nearly all users, so you can skip the
 1093: rest of this section. If you relly must know all the gory details and
 1094: options, read on.
 1095: 
 1096: In order to implement this rule, the compiler has to know which places
 1097: are unreachable. It knows this automatically after @code{AHEAD},
 1098: @code{AGAIN}, @code{EXIT} and @code{LEAVE}; in other cases (e.g., after
 1099: most @code{THROW}s), you can use the word @code{UNREACHABLE} to tell the
 1100: compiler that the control flow never reaches that place. If
 1101: @code{UNREACHABLE} is not used where it could, the only consequence is
 1102: that the visibility of some locals is more limited than the rule above
 1103: says. If @code{UNREACHABLE} is used where it should not (i.e., if you
 1104: lie to the compiler), buggy code will be produced.
 1105: 
 1106: Another problem with this rule is that at @code{BEGIN}, the compiler
 1107: does not know which locals will be visible on the incoming
 1108: back-edge. All problems discussed in the following are due to this
 1109: ignorance of the compiler (we discuss the problems using @code{BEGIN}
 1110: loops as examples; the discussion also applies to @code{?DO} and other
 1111: loops). Perhaps the most insidious example is:
 1112: @example
 1113: AHEAD
 1114: BEGIN
 1115:   x
 1116: [ 1 CS-ROLL ] THEN
 1117:   @{ x @}
 1118:   ...
 1119: UNTIL
 1120: @end example
 1121: 
 1122: This should be legal according to the visibility rule. The use of
 1123: @code{x} can only be reached through the definition; but that appears
 1124: textually below the use.
 1125: 
 1126: From this example it is clear that the visibility rules cannot be fully
 1127: implemented without major headaches. Our implementation treats common
 1128: cases as advertised and the exceptions are treated in a safe way: The
 1129: compiler makes a reasonable guess about the locals visible after a
 1130: @code{BEGIN}; if it is too pessimistic, the
 1131: user will get a spurious error about the local not being defined; if the
 1132: compiler is too optimistic, it will notice this later and issue a
 1133: warning. In the case above the compiler would complain about @code{x}
 1134: being undefined at its use. You can see from the obscure examples in
 1135: this section that it takes quite unusual control structures to get the
 1136: compiler into trouble, and even then it will often do fine.
 1137: 
 1138: If the @code{BEGIN} is reachable from above, the most optimistic guess
 1139: is that all locals visible before the @code{BEGIN} will also be
 1140: visible after the @code{BEGIN}. This guess is valid for all loops that
 1141: are entered only through the @code{BEGIN}, in particular, for normal
 1142: @code{BEGIN}...@code{WHILE}...@code{REPEAT} and
 1143: @code{BEGIN}...@code{UNTIL} loops and it is implemented in our
 1144: compiler. When the branch to the @code{BEGIN} is finally generated by
 1145: @code{AGAIN} or @code{UNTIL}, the compiler checks the guess and
 1146: warns the user if it was too optimisitic:
 1147: @example
 1148: IF
 1149:   @{ x @}
 1150: BEGIN
 1151:   \ x ? 
 1152: [ 1 cs-roll ] THEN
 1153:   ...
 1154: UNTIL
 1155: @end example
 1156: 
 1157: Here, @code{x} lives only until the @code{BEGIN}, but the compiler
 1158: optimistically assumes that it lives until the @code{THEN}. It notices
 1159: this difference when it compiles the @code{UNTIL} and issues a
 1160: warning. The user can avoid the warning, and make sure that @code{x}
 1161: is not used in the wrong area by using explicit scoping:
 1162: @example
 1163: IF
 1164:   SCOPE
 1165:   @{ x @}
 1166:   ENDSCOPE
 1167: BEGIN
 1168: [ 1 cs-roll ] THEN
 1169:   ...
 1170: UNTIL
 1171: @end example
 1172: 
 1173: Since the guess is optimistic, there will be no spurious error messages
 1174: about undefined locals.
 1175: 
 1176: If the @code{BEGIN} is not reachable from above (e.g., after
 1177: @code{AHEAD} or @code{EXIT}), the compiler cannot even make an
 1178: optimistic guess, as the locals visible after the @code{BEGIN} may be
 1179: defined later. Therefore, the compiler assumes that no locals are
 1180: visible after the @code{BEGIN}. However, the useer can use
 1181: @code{ASSUME-LIVE} to make the compiler assume that the same locals are
 1182: visible at the BEGIN as at the point where the item was created.
 1183: 
 1184: doc-assume-live
 1185: 
 1186: E.g.,
 1187: @example
 1188: @{ x @}
 1189: AHEAD
 1190: ASSUME-LIVE
 1191: BEGIN
 1192:   x
 1193: [ 1 CS-ROLL ] THEN
 1194:   ...
 1195: UNTIL
 1196: @end example
 1197: 
 1198: Other cases where the locals are defined before the @code{BEGIN} can be
 1199: handled by inserting an appropriate @code{CS-ROLL} before the
 1200: @code{ASSUME-LIVE} (and changing the control-flow stack manipulation
 1201: behind the @code{ASSUME-LIVE}).
 1202: 
 1203: Cases where locals are defined after the @code{BEGIN} (but should be
 1204: visible immediately after the @code{BEGIN}) can only be handled by
 1205: rearranging the loop. E.g., the ``most insidious'' example above can be
 1206: arranged into:
 1207: @example
 1208: BEGIN
 1209:   @{ x @}
 1210:   ... 0=
 1211: WHILE
 1212:   x
 1213: REPEAT
 1214: @end example
 1215: 
 1216: @node How long do locals live?, Programming Style, Where are locals visible by name?, gforth locals
 1217: @subsubsection How long do locals live?
 1218: 
 1219: The right answer for the lifetime question would be: A local lives at
 1220: least as long as it can be accessed. For a value-flavoured local this
 1221: means: until the end of its visibility. However, a variable-flavoured
 1222: local could be accessed through its address far beyond its visibility
 1223: scope. Ultimately, this would mean that such locals would have to be
 1224: garbage collected. Since this entails un-Forth-like implementation
 1225: complexities, I adopted the same cowardly solution as some other
 1226: languages (e.g., C): The local lives only as long as it is visible;
 1227: afterwards its address is invalid (and programs that access it
 1228: afterwards are erroneous).
 1229: 
 1230: @node Programming Style, Implementation, How long do locals live?, gforth locals
 1231: @subsubsection Programming Style
 1232: 
 1233: The freedom to define locals anywhere has the potential to change
 1234: programming styles dramatically. In particular, the need to use the
 1235: return stack for intermediate storage vanishes. Moreover, all stack
 1236: manipulations (except @code{PICK}s and @code{ROLL}s with run-time
 1237: determined arguments) can be eliminated: If the stack items are in the
 1238: wrong order, just write a locals definition for all of them; then
 1239: write the items in the order you want.
 1240: 
 1241: This seems a little far-fetched and eliminating stack manipulations is
 1242: unlikely to become a conscious programming objective. Still, the number
 1243: of stack manipulations will be reduced dramatically if local variables
 1244: are used liberally (e.g., compare @code{max} in @ref{gforth locals} with
 1245: a traditional implementation of @code{max}).
 1246: 
 1247: This shows one potential benefit of locals: making Forth programs more
 1248: readable. Of course, this benefit will only be realized if the
 1249: programmers continue to honour the principle of factoring instead of
 1250: using the added latitude to make the words longer.
 1251: 
 1252: Using @code{TO} can and should be avoided.  Without @code{TO},
 1253: every value-flavoured local has only a single assignment and many
 1254: advantages of functional languages apply to Forth. I.e., programs are
 1255: easier to analyse, to optimize and to read: It is clear from the
 1256: definition what the local stands for, it does not turn into something
 1257: different later.
 1258: 
 1259: E.g., a definition using @code{TO} might look like this:
 1260: @example
 1261: : strcmp @{ addr1 u1 addr2 u2 -- n @}
 1262:  u1 u2 min 0
 1263:  ?do
 1264:    addr1 c@ addr2 c@ - ?dup
 1265:    if
 1266:      unloop exit
 1267:    then
 1268:    addr1 char+ TO addr1
 1269:    addr2 char+ TO addr2
 1270:  loop
 1271:  u1 u2 - ;
 1272: @end example
 1273: Here, @code{TO} is used to update @code{addr1} and @code{addr2} at
 1274: every loop iteration. @code{strcmp} is a typical example of the
 1275: readability problems of using @code{TO}. When you start reading
 1276: @code{strcmp}, you think that @code{addr1} refers to the start of the
 1277: string. Only near the end of the loop you realize that it is something
 1278: else.
 1279: 
 1280: This can be avoided by defining two locals at the start of the loop that
 1281: are initialized with the right value for the current iteration.
 1282: @example
 1283: : strcmp @{ addr1 u1 addr2 u2 -- n @}
 1284:  addr1 addr2
 1285:  u1 u2 min 0 
 1286:  ?do @{ s1 s2 @}
 1287:    s1 c@ s2 c@ - ?dup 
 1288:    if
 1289:      unloop exit
 1290:    then
 1291:    s1 char+ s2 char+
 1292:  loop
 1293:  2drop
 1294:  u1 u2 - ;
 1295: @end example
 1296: Here it is clear from the start that @code{s1} has a different value
 1297: in every loop iteration.
 1298: 
 1299: @node Implementation,  , Programming Style, gforth locals
 1300: @subsubsection Implementation
 1301: 
 1302: GNU Forth uses an extra locals stack. The most compelling reason for
 1303: this is that the return stack is not float-aligned; using an extra stack
 1304: also eliminates the problems and restrictions of using the return stack
 1305: as locals stack. Like the other stacks, the locals stack grows toward
 1306: lower addresses. A few primitives allow an efficient implementation:
 1307: 
 1308: doc-@local#
 1309: doc-f@local#
 1310: doc-laddr#
 1311: doc-lp+!#
 1312: doc-lp!
 1313: doc->l
 1314: doc-f>l
 1315: 
 1316: In addition to these primitives, some specializations of these
 1317: primitives for commonly occurring inline arguments are provided for
 1318: efficiency reasons, e.g., @code{@@local0} as specialization of
 1319: @code{@@local#} for the inline argument 0. The following compiling words
 1320: compile the right specialized version, or the general version, as
 1321: appropriate:
 1322: 
 1323: doc-compile-@local
 1324: doc-compile-f@local
 1325: doc-compile-lp+!
 1326: 
 1327: Combinations of conditional branches and @code{lp+!#} like
 1328: @code{?branch-lp+!#} (the locals pointer is only changed if the branch
 1329: is taken) are provided for efficiency and correctness in loops.
 1330: 
 1331: A special area in the dictionary space is reserved for keeping the
 1332: local variable names. @code{@{} switches the dictionary pointer to this
 1333: area and @code{@}} switches it back and generates the locals
 1334: initializing code. @code{W:} etc.@ are normal defining words. This
 1335: special area is cleared at the start of every colon definition.
 1336: 
 1337: A special feature of GNU Forths dictionary is used to implement the
 1338: definition of locals without type specifiers: every wordlist (aka
 1339: vocabulary) has its own methods for searching
 1340: etc. (@pxref{Wordlists}). For the present purpose we defined a wordlist
 1341: with a special search method: When it is searched for a word, it
 1342: actually creates that word using @code{W:}. @code{@{} changes the search
 1343: order to first search the wordlist containing @code{@}}, @code{W:} etc.,
 1344: and then the wordlist for defining locals without type specifiers.
 1345: 
 1346: The lifetime rules support a stack discipline within a colon
 1347: definition: The lifetime of a local is either nested with other locals
 1348: lifetimes or it does not overlap them.
 1349: 
 1350: At @code{BEGIN}, @code{IF}, and @code{AHEAD} no code for locals stack
 1351: pointer manipulation is generated. Between control structure words
 1352: locals definitions can push locals onto the locals stack. @code{AGAIN}
 1353: is the simplest of the other three control flow words. It has to
 1354: restore the locals stack depth of the corresponding @code{BEGIN}
 1355: before branching. The code looks like this:
 1356: @format
 1357: @code{lp+!#} current-locals-size @minus{} dest-locals-size
 1358: @code{branch} <begin>
 1359: @end format
 1360: 
 1361: @code{UNTIL} is a little more complicated: If it branches back, it
 1362: must adjust the stack just like @code{AGAIN}. But if it falls through,
 1363: the locals stack must not be changed. The compiler generates the
 1364: following code:
 1365: @format
 1366: @code{?branch-lp+!#} <begin> current-locals-size @minus{} dest-locals-size
 1367: @end format
 1368: The locals stack pointer is only adjusted if the branch is taken.
 1369: 
 1370: @code{THEN} can produce somewhat inefficient code:
 1371: @format
 1372: @code{lp+!#} current-locals-size @minus{} orig-locals-size
 1373: <orig target>:
 1374: @code{lp+!#} orig-locals-size @minus{} new-locals-size
 1375: @end format
 1376: The second @code{lp+!#} adjusts the locals stack pointer from the
 1377: level at the @var{orig} point to the level after the @code{THEN}. The
 1378: first @code{lp+!#} adjusts the locals stack pointer from the current
 1379: level to the level at the orig point, so the complete effect is an
 1380: adjustment from the current level to the right level after the
 1381: @code{THEN}.
 1382: 
 1383: In a conventional Forth implementation a dest control-flow stack entry
 1384: is just the target address and an orig entry is just the address to be
 1385: patched. Our locals implementation adds a wordlist to every orig or dest
 1386: item. It is the list of locals visible (or assumed visible) at the point
 1387: described by the entry. Our implementation also adds a tag to identify
 1388: the kind of entry, in particular to differentiate between live and dead
 1389: (reachable and unreachable) orig entries.
 1390: 
 1391: A few unusual operations have to be performed on locals wordlists:
 1392: 
 1393: doc-common-list
 1394: doc-sub-list?
 1395: doc-list-size
 1396: 
 1397: Several features of our locals wordlist implementation make these
 1398: operations easy to implement: The locals wordlists are organised as
 1399: linked lists; the tails of these lists are shared, if the lists
 1400: contain some of the same locals; and the address of a name is greater
 1401: than the address of the names behind it in the list.
 1402: 
 1403: Another important implementation detail is the variable
 1404: @code{dead-code}. It is used by @code{BEGIN} and @code{THEN} to
 1405: determine if they can be reached directly or only through the branch
 1406: that they resolve. @code{dead-code} is set by @code{UNREACHABLE},
 1407: @code{AHEAD}, @code{EXIT} etc., and cleared at the start of a colon
 1408: definition, by @code{BEGIN} and usually by @code{THEN}.
 1409: 
 1410: Counted loops are similar to other loops in most respects, but
 1411: @code{LEAVE} requires special attention: It performs basically the same
 1412: service as @code{AHEAD}, but it does not create a control-flow stack
 1413: entry. Therefore the information has to be stored elsewhere;
 1414: traditionally, the information was stored in the target fields of the
 1415: branches created by the @code{LEAVE}s, by organizing these fields into a
 1416: linked list. Unfortunately, this clever trick does not provide enough
 1417: space for storing our extended control flow information. Therefore, we
 1418: introduce another stack, the leave stack. It contains the control-flow
 1419: stack entries for all unresolved @code{LEAVE}s.
 1420: 
 1421: Local names are kept until the end of the colon definition, even if
 1422: they are no longer visible in any control-flow path. In a few cases
 1423: this may lead to increased space needs for the locals name area, but
 1424: usually less than reclaiming this space would cost in code size.
 1425: 
 1426: 
 1427: @node ANS Forth locals,  , gforth locals, Locals
 1428: @subsection ANS Forth locals
 1429: 
 1430: The ANS Forth locals wordset does not define a syntax for locals, but
 1431: words that make it possible to define various syntaxes. One of the
 1432: possible syntaxes is a subset of the syntax we used in the gforth locals
 1433: wordset, i.e.:
 1434: 
 1435: @example
 1436: @{ local1 local2 ... -- comment @}
 1437: @end example
 1438: or
 1439: @example
 1440: @{ local1 local2 ... @}
 1441: @end example
 1442: 
 1443: The order of the locals corresponds to the order in a stack comment. The
 1444: restrictions are:
 1445: 
 1446: @itemize @bullet
 1447: @item
 1448: Locals can only be cell-sized values (no type specifers are allowed).
 1449: @item
 1450: Locals can be defined only outside control structures.
 1451: @item
 1452: Locals can interfere with explicit usage of the return stack. For the
 1453: exact (and long) rules, see the standard. If you don't use return stack
 1454: accessing words in a definition using locals, you will we all right. The
 1455: purpose of this rule is to make locals implementation on the return
 1456: stack easier.
 1457: @item
 1458: The whole definition must be in one line.
 1459: @end itemize
 1460: 
 1461: Locals defined in this way behave like @code{VALUE}s
 1462: (@xref{Values}). I.e., they are initialized from the stack. Using their
 1463: name produces their value. Their value can be changed using @code{TO}.
 1464: 
 1465: Since this syntax is supported by gforth directly, you need not do
 1466: anything to use it. If you want to port a program using this syntax to
 1467: another ANS Forth system, use @file{anslocal.fs} to implement the syntax
 1468: on the other system.
 1469: 
 1470: Note that a syntax shown in the standard, section A.13 looks
 1471: similar, but is quite different in having the order of locals
 1472: reversed. Beware!
 1473: 
 1474: The ANS Forth locals wordset itself consists of the following word
 1475: 
 1476: doc-(local)
 1477: 
 1478: The ANS Forth locals extension wordset defines a syntax, but it is so
 1479: awful that we strongly recommend not to use it. We have implemented this
 1480: syntax to make porting to gforth easy, but do not document it here. The
 1481: problem with this syntax is that the locals are defined in an order
 1482: reversed with respect to the standard stack comment notation, making
 1483: programs harder to read, and easier to misread and miswrite. The only
 1484: merit of this syntax is that it is easy to implement using the ANS Forth
 1485: locals wordset.
 1486: 
 1487: @node Defining Words, Wordlists, Locals, Words
 1488: @section Defining Words
 1489: 
 1490: @node Values,  , Defining Words, Defining Words
 1491: @subsection Values
 1492: 
 1493: @node Wordlists, Files, Defining Words, Words
 1494: @section Wordlists
 1495: 
 1496: @node Files, Blocks, Wordlists, Words
 1497: @section Files
 1498: 
 1499: @node Blocks, Other I/O, Files, Words
 1500: @section Blocks
 1501: 
 1502: @node Other I/O, Programming Tools, Blocks, Words
 1503: @section Other I/O
 1504: 
 1505: @node Programming Tools, Threading Words, Other I/O, Words
 1506: @section Programming Tools
 1507: 
 1508: @menu
 1509: * Debugging::                   Simple and quick.
 1510: * Assertions::                  Making your programs self-checking.
 1511: @end menu
 1512: 
 1513: @node Debugging, Assertions, Programming Tools, Programming Tools
 1514: @subsection Debugging
 1515: 
 1516: The simple debugging aids provided in @file{debugging.fs}
 1517: are meant to support a different style of debugging than the
 1518: tracing/stepping debuggers used in languages with long turn-around
 1519: times.
 1520: 
 1521: A much better (faster) way in fast-compilig languages is to add
 1522: printing code at well-selected places, let the program run, look at
 1523: the output, see where things went wrong, add more printing code, etc.,
 1524: until the bug is found.
 1525: 
 1526: The word @code{~~} is easy to insert. It just prints debugging
 1527: information (by default the source location and the stack contents). It
 1528: is also easy to remove (@kbd{C-x ~} in the Emacs Forth mode to
 1529: query-replace them with nothing). The deferred words
 1530: @code{printdebugdata} and @code{printdebugline} control the output of
 1531: @code{~~}. The default source location output format works well with
 1532: Emacs' compilation mode, so you can step through the program at the
 1533: source level using @kbd{C-x `} (the advantage over a stepping debugger
 1534: is that you can step in any direction and you know where the crash has
 1535: happened or where the strange data has occurred).
 1536: 
 1537: Note that the default actions clobber the contents of the pictured
 1538: numeric output string, so you should not use @code{~~}, e.g., between
 1539: @code{<#} and @code{#>}.
 1540: 
 1541: doc-~~
 1542: doc-printdebugdata
 1543: doc-printdebugline
 1544: 
 1545: @node Assertions,  , Debugging, Programming Tools
 1546: @subsection Assertions
 1547: 
 1548: It is a good idea to make your programs self-checking, in particular, if
 1549: you use an assumption (e.g., that a certain field of a data structure is
 1550: never zero) that may become wrong during maintenance. GForth supports
 1551: assertions for this purpose. They are used like this:
 1552: 
 1553: @example
 1554: assert( @var{flag} )
 1555: @end example
 1556: 
 1557: The code between @code{assert(} and @code{)} should compute a flag, that
 1558: should be true if everything is alright and false otherwise. It should
 1559: not change anything else on the stack. The overall stack effect of the
 1560: assertion is @code{( -- )}. E.g.
 1561: 
 1562: @example
 1563: assert( 1 1 + 2 = ) \ what we learn in school
 1564: assert( dup 0<> ) \ assert that the top of stack is not zero
 1565: assert( false ) \ this code should not be reached
 1566: @end example
 1567: 
 1568: The need for assertions is different at different times. During
 1569: debugging, we want more checking, in production we sometimes care more
 1570: for speed. Therefore, assertions can be turned off, i.e., the assertion
 1571: becomes a comment. Depending on the importance of an assertion and the
 1572: time it takes to check it, you may want to turn off some assertions and
 1573: keep others turned on. GForth provides several levels of assertions for
 1574: this purpose:
 1575: 
 1576: doc-assert0(
 1577: doc-assert1(
 1578: doc-assert2(
 1579: doc-assert3(
 1580: doc-assert(
 1581: doc-)
 1582: 
 1583: @code{Assert(} is the same as @code{assert1(}. The variable
 1584: @code{assert-level} specifies the highest assertions that are turned
 1585: on. I.e., at the default @code{assert-level} of one, @code{assert0(} and
 1586: @code{assert1(} assertions perform checking, while @code{assert2(} and
 1587: @code{assert3(} assertions are treated as comments.
 1588: 
 1589: Note that the @code{assert-level} is evaluated at compile-time, not at
 1590: run-time. I.e., you cannot turn assertions on or off at run-time, you
 1591: have to set the @code{assert-level} appropriately before compiling a
 1592: piece of code. You can compile several pieces of code at several
 1593: @code{assert-level}s (e.g., a trusted library at level 1 and newly
 1594: written code at level 3).
 1595: 
 1596: doc-assert-level
 1597: 
 1598: If an assertion fails, a message compatible with Emacs' compilation mode
 1599: is produced and the execution is aborted (currently with @code{ABORT"}.
 1600: If there is interest, we will introduce a special throw code. But if you
 1601: intend to @code{catch} a specific condition, using @code{throw} is
 1602: probably more appropriate than an assertion).
 1603: 
 1604: @node Threading Words,  , Programming Tools, Words
 1605: @section Threading Words
 1606: 
 1607: These words provide access to code addresses and other threading stuff
 1608: in gforth (and, possibly, other interpretive Forths). It more or less
 1609: abstracts away the differences between direct and indirect threading
 1610: (and, for direct threading, the machine dependences). However, at
 1611: present this wordset is still inclomplete. It is also pretty low-level;
 1612: some day it will hopefully be made unnecessary by an internals words set
 1613: that abstracts implementation details away completely.
 1614: 
 1615: doc->code-address
 1616: doc->does-code
 1617: doc-code-address!
 1618: doc-does-code!
 1619: doc-does-handler!
 1620: doc-/does-handler
 1621: 
 1622: @node ANS conformance, Model, Words, Top
 1623: @chapter ANS conformance
 1624: 
 1625: @node Model, Emacs and GForth, ANS conformance, Top
 1626: @chapter Model
 1627: 
 1628: @node Emacs and GForth, Internals, Model, Top
 1629: @chapter Emacs and GForth
 1630: 
 1631: GForth comes with @file{gforth.el}, an improved version of
 1632: @file{forth.el} by Goran Rydqvist (icluded in the TILE package). The
 1633: improvements are a better (but still not perfect) handling of
 1634: indentation. I have also added comment paragraph filling (@kbd{M-q}),
 1635: commenting (@kbd{C-x \}) and uncommenting (@kbd{C-u C-x \}) regions and
 1636: removing debugging tracers (@kbd{C-x ~}, @pxref{Debugging}). I left the
 1637: stuff I do not use alone, even though some of it only makes sense for
 1638: TILE. To get a description of these features, enter Forth mode and type
 1639: @kbd{C-h m}.
 1640: 
 1641: In addition, GForth supports Emacs quite well: The source code locations
 1642: given in error messages, debugging output (from @code{~~}) and failed
 1643: assertion messages are in the right format for Emacs' compilation mode
 1644: (@pxref{Compilation, , Running Compilations under Emacs, emacs, Emacs
 1645: Manual}) so the source location corresponding to an error or other
 1646: message is only a few keystrokes away (@kbd{C-x `} for the next error,
 1647: @kbd{C-c C-c} for the error under the cursor).
 1648: 
 1649: Also, if you @code{include} @file{etags.fs}, a new @file{TAGS} file
 1650: (@pxref{Tags, , Tags Tables, emacs, Emacs Manual}) will be produced that
 1651: contains the definitions of all words defined afterwards. You can then
 1652: find the source for a word using @kbd{M-.}. Note that emacs can use
 1653: several tags files at the same time (e.g., one for the gforth sources
 1654: and one for your program).
 1655: 
 1656: To get all these benefits, add the following lines to your @file{.emacs}
 1657: file:
 1658: 
 1659: @example
 1660: (autoload 'forth-mode "gforth.el")
 1661: (setq auto-mode-alist (cons '("\\.fs\\'" . forth-mode) auto-mode-alist))
 1662: @end example
 1663: 
 1664: @node Internals, Bugs, Emacs and GForth, Top
 1665: @chapter Internals
 1666: 
 1667: Reading this section is not necessary for programming with gforth. It
 1668: should be helpful for finding your way in the gforth sources.
 1669: 
 1670: @menu
 1671: * Portability::                 
 1672: * Threading::                   
 1673: * Primitives::                  
 1674: * System Architecture::         
 1675: @end menu
 1676: 
 1677: @node Portability, Threading, Internals, Internals
 1678: @section Portability
 1679: 
 1680: One of the main goals of the effort is availability across a wide range
 1681: of personal machines. fig-Forth, and, to a lesser extent, F83, achieved
 1682: this goal by manually coding the engine in assembly language for several
 1683: then-popular processors. This approach is very labor-intensive and the
 1684: results are short-lived due to progress in computer architecture.
 1685: 
 1686: Others have avoided this problem by coding in C, e.g., Mitch Bradley
 1687: (cforth), Mikael Patel (TILE) and Dirk Zoller (pfe). This approach is
 1688: particularly popular for UNIX-based Forths due to the large variety of
 1689: architectures of UNIX machines. Unfortunately an implementation in C
 1690: does not mix well with the goals of efficiency and with using
 1691: traditional techniques: Indirect or direct threading cannot be expressed
 1692: in C, and switch threading, the fastest technique available in C, is
 1693: significantly slower. Another problem with C is that it's very
 1694: cumbersome to express double integer arithmetic.
 1695: 
 1696: Fortunately, there is a portable language that does not have these
 1697: limitations: GNU C, the version of C processed by the GNU C compiler
 1698: (@pxref{C Extensions, , Extensions to the C Language Family, gcc.info,
 1699: GNU C Manual}). Its labels as values feature (@pxref{Labels as Values, ,
 1700: Labels as Values, gcc.info, GNU C Manual}) makes direct and indirect
 1701: threading possible, its @code{long long} type (@pxref{Long Long, ,
 1702: Double-Word Integers, gcc.info, GNU C Manual}) corresponds to Forths
 1703: double numbers. GNU C is available for free on all important (and many
 1704: unimportant) UNIX machines, VMS, 80386s running MS-DOS, the Amiga, and
 1705: the Atari ST, so a Forth written in GNU C can run on all these
 1706: machines@footnote{Due to Apple's look-and-feel lawsuit it is not
 1707: available on the Mac (@pxref{Boycott, , Protect Your Freedom---Fight
 1708: ``Look And Feel'', gcc.info, GNU C Manual}).}.
 1709: 
 1710: Writing in a portable language has the reputation of producing code that
 1711: is slower than assembly. For our Forth engine we repeatedly looked at
 1712: the code produced by the compiler and eliminated most compiler-induced
 1713: inefficiencies by appropriate changes in the source-code.
 1714: 
 1715: However, register allocation cannot be portably influenced by the
 1716: programmer, leading to some inefficiencies on register-starved
 1717: machines. We use explicit register declarations (@pxref{Explicit Reg
 1718: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) to
 1719: improve the speed on some machines. They are turned on by using the
 1720: @code{gcc} switch @code{-DFORCE_REG}. Unfortunately, this feature not
 1721: only depends on the machine, but also on the compiler version: On some
 1722: machines some compiler versions produce incorrect code when certain
 1723: explicit register declarations are used. So by default
 1724: @code{-DFORCE_REG} is not used.
 1725: 
 1726: @node Threading, Primitives, Portability, Internals
 1727: @section Threading
 1728: 
 1729: GNU C's labels as values extension (available since @code{gcc-2.0},
 1730: @pxref{Labels as Values, , Labels as Values, gcc.info, GNU C Manual})
 1731: makes it possible to take the address of @var{label} by writing
 1732: @code{&&@var{label}}.  This address can then be used in a statement like
 1733: @code{goto *@var{address}}. I.e., @code{goto *&&x} is the same as
 1734: @code{goto x}.
 1735: 
 1736: With this feature an indirect threaded NEXT looks like:
 1737: @example
 1738: cfa = *ip++;
 1739: ca = *cfa;
 1740: goto *ca;
 1741: @end example
 1742: For those unfamiliar with the names: @code{ip} is the Forth instruction
 1743: pointer; the @code{cfa} (code-field address) corresponds to ANS Forths
 1744: execution token and points to the code field of the next word to be
 1745: executed; The @code{ca} (code address) fetched from there points to some
 1746: executable code, e.g., a primitive or the colon definition handler
 1747: @code{docol}.
 1748: 
 1749: Direct threading is even simpler:
 1750: @example
 1751: ca = *ip++;
 1752: goto *ca;
 1753: @end example
 1754: 
 1755: Of course we have packaged the whole thing neatly in macros called
 1756: @code{NEXT} and @code{NEXT1} (the part of NEXT after fetching the cfa).
 1757: 
 1758: @menu
 1759: * Scheduling::                  
 1760: * Direct or Indirect Threaded?::  
 1761: * DOES>::                       
 1762: @end menu
 1763: 
 1764: @node Scheduling, Direct or Indirect Threaded?, Threading, Threading
 1765: @subsection Scheduling
 1766: 
 1767: There is a little complication: Pipelined and superscalar processors,
 1768: i.e., RISC and some modern CISC machines can process independent
 1769: instructions while waiting for the results of an instruction. The
 1770: compiler usually reorders (schedules) the instructions in a way that
 1771: achieves good usage of these delay slots. However, on our first tries
 1772: the compiler did not do well on scheduling primitives. E.g., for
 1773: @code{+} implemented as
 1774: @example
 1775: n=sp[0]+sp[1];
 1776: sp++;
 1777: sp[0]=n;
 1778: NEXT;
 1779: @end example
 1780: the NEXT comes strictly after the other code, i.e., there is nearly no
 1781: scheduling. After a little thought the problem becomes clear: The
 1782: compiler cannot know that sp and ip point to different addresses (and
 1783: the version of @code{gcc} we used would not know it even if it was
 1784: possible), so it could not move the load of the cfa above the store to
 1785: the TOS. Indeed the pointers could be the same, if code on or very near
 1786: the top of stack were executed. In the interest of speed we chose to
 1787: forbid this probably unused ``feature'' and helped the compiler in
 1788: scheduling: NEXT is divided into the loading part (@code{NEXT_P1}) and
 1789: the goto part (@code{NEXT_P2}). @code{+} now looks like:
 1790: @example
 1791: n=sp[0]+sp[1];
 1792: sp++;
 1793: NEXT_P1;
 1794: sp[0]=n;
 1795: NEXT_P2;
 1796: @end example
 1797: This can be scheduled optimally by the compiler.
 1798: 
 1799: This division can be turned off with the switch @code{-DCISC_NEXT}. This
 1800: switch is on by default on machines that do not profit from scheduling
 1801: (e.g., the 80386), in order to preserve registers.
 1802: 
 1803: @node Direct or Indirect Threaded?, DOES>, Scheduling, Threading
 1804: @subsection Direct or Indirect Threaded?
 1805: 
 1806: Both! After packaging the nasty details in macro definitions we
 1807: realized that we could switch between direct and indirect threading by
 1808: simply setting a compilation flag (@code{-DDIRECT_THREADED}) and
 1809: defining a few machine-specific macros for the direct-threading case.
 1810: On the Forth level we also offer access words that hide the
 1811: differences between the threading methods (@pxref{Threading Words}).
 1812: 
 1813: Indirect threading is implemented completely
 1814: machine-independently. Direct threading needs routines for creating
 1815: jumps to the executable code (e.g. to docol or dodoes). These routines
 1816: are inherently machine-dependent, but they do not amount to many source
 1817: lines. I.e., even porting direct threading to a new machine is a small
 1818: effort.
 1819: 
 1820: @node DOES>,  , Direct or Indirect Threaded?, Threading
 1821: @subsection DOES>
 1822: One of the most complex parts of a Forth engine is @code{dodoes}, i.e.,
 1823: the chunk of code executed by every word defined by a
 1824: @code{CREATE}...@code{DOES>} pair. The main problem here is: How to find
 1825: the Forth code to be executed, i.e. the code after the @code{DOES>} (the
 1826: DOES-code)? There are two solutions:
 1827: 
 1828: In fig-Forth the code field points directly to the dodoes and the
 1829: DOES-code address is stored in the cell after the code address
 1830: (i.e. at cfa cell+). It may seem that this solution is illegal in the
 1831: Forth-79 and all later standards, because in fig-Forth this address
 1832: lies in the body (which is illegal in these standards). However, by
 1833: making the code field larger for all words this solution becomes legal
 1834: again. We use this approach for the indirect threaded version. Leaving
 1835: a cell unused in most words is a bit wasteful, but on the machines we
 1836: are targetting this is hardly a problem. The other reason for having a
 1837: code field size of two cells is to avoid having different image files
 1838: for direct and indirect threaded systems (@pxref{System Architecture}).
 1839: 
 1840: The other approach is that the code field points or jumps to the cell
 1841: after @code{DOES}. In this variant there is a jump to @code{dodoes} at
 1842: this address. @code{dodoes} can then get the DOES-code address by
 1843: computing the code address, i.e., the address of the jump to dodoes,
 1844: and add the length of that jump field. A variant of this is to have a
 1845: call to @code{dodoes} after the @code{DOES>}; then the return address
 1846: (which can be found in the return register on RISCs) is the DOES-code
 1847: address. Since the two cells available in the code field are usually
 1848: used up by the jump to the code address in direct threading, we use
 1849: this approach for direct threading. We did not want to add another
 1850: cell to the code field.
 1851: 
 1852: @node Primitives, System Architecture, Threading, Internals
 1853: @section Primitives
 1854: 
 1855: @menu
 1856: * Automatic Generation::        
 1857: * TOS Optimization::            
 1858: * Produced code::               
 1859: @end menu
 1860: 
 1861: @node Automatic Generation, TOS Optimization, Primitives, Primitives
 1862: @subsection Automatic Generation
 1863: 
 1864: Since the primitives are implemented in a portable language, there is no
 1865: longer any need to minimize the number of primitives. On the contrary,
 1866: having many primitives is an advantage: speed. In order to reduce the
 1867: number of errors in primitives and to make programming them easier, we
 1868: provide a tool, the primitive generator (@file{prims2x.fs}), that
 1869: automatically generates most (and sometimes all) of the C code for a
 1870: primitive from the stack effect notation.  The source for a primitive
 1871: has the following form:
 1872: 
 1873: @format
 1874: @var{Forth-name}	@var{stack-effect}	@var{category}	[@var{pronounc.}]
 1875: [@code{""}@var{glossary entry}@code{""}]
 1876: @var{C code}
 1877: [@code{:}
 1878: @var{Forth code}]
 1879: @end format
 1880: 
 1881: The items in brackets are optional. The category and glossary fields
 1882: are there for generating the documentation, the Forth code is there
 1883: for manual implementations on machines without GNU C. E.g., the source
 1884: for the primitive @code{+} is:
 1885: @example
 1886: +    n1 n2 -- n    core    plus
 1887: n = n1+n2;
 1888: @end example
 1889: 
 1890: This looks like a specification, but in fact @code{n = n1+n2} is C
 1891: code. Our primitive generation tool extracts a lot of information from
 1892: the stack effect notations@footnote{We use a one-stack notation, even
 1893: though we have separate data and floating-point stacks; The separate
 1894: notation can be generated easily from the unified notation.}: The number
 1895: of items popped from and pushed on the stack, their type, and by what
 1896: name they are referred to in the C code. It then generates a C code
 1897: prelude and postlude for each primitive. The final C code for @code{+}
 1898: looks like this:
 1899: 
 1900: @example
 1901: I_plus:	/* + ( n1 n2 -- n ) */  /* label, stack effect */
 1902: /*  */                          /* documentation */
 1903: @{
 1904: DEF_CA                          /* definition of variable ca (indirect threading) */
 1905: Cell n1;                        /* definitions of variables */
 1906: Cell n2;
 1907: Cell n;
 1908: n1 = (Cell) sp[1];              /* input */
 1909: n2 = (Cell) TOS;
 1910: sp += 1;                        /* stack adjustment */
 1911: NAME("+")                       /* debugging output (with -DDEBUG) */
 1912: @{
 1913: n = n1+n2;                      /* C code taken from the source */
 1914: @}
 1915: NEXT_P1;                        /* NEXT part 1 */
 1916: TOS = (Cell)n;                  /* output */
 1917: NEXT_P2;                        /* NEXT part 2 */
 1918: @}
 1919: @end example
 1920: 
 1921: This looks long and inefficient, but the GNU C compiler optimizes quite
 1922: well and produces optimal code for @code{+} on, e.g., the R3000 and the
 1923: HP RISC machines: Defining the @code{n}s does not produce any code, and
 1924: using them as intermediate storage also adds no cost.
 1925: 
 1926: There are also other optimizations, that are not illustrated by this
 1927: example: Assignments between simple variables are usually for free (copy
 1928: propagation). If one of the stack items is not used by the primitive
 1929: (e.g.  in @code{drop}), the compiler eliminates the load from the stack
 1930: (dead code elimination). On the other hand, there are some things that
 1931: the compiler does not do, therefore they are performed by
 1932: @file{prims2x.fs}: The compiler does not optimize code away that stores
 1933: a stack item to the place where it just came from (e.g., @code{over}).
 1934: 
 1935: While programming a primitive is usually easy, there are a few cases
 1936: where the programmer has to take the actions of the generator into
 1937: account, most notably @code{?dup}, but also words that do not (always)
 1938: fall through to NEXT.
 1939: 
 1940: @node TOS Optimization, Produced code, Automatic Generation, Primitives
 1941: @subsection TOS Optimization
 1942: 
 1943: An important optimization for stack machine emulators, e.g., Forth
 1944: engines, is keeping  one or more of the top stack items in
 1945: registers.  If a word has the stack effect @var{in1}...@var{inx} @code{--}
 1946: @var{out1}...@var{outy}, keeping the top @var{n} items in registers
 1947: @itemize
 1948: @item
 1949: is better than keeping @var{n-1} items, if @var{x>=n} and @var{y>=n},
 1950: due to fewer loads from and stores to the stack.
 1951: @item is slower than keeping @var{n-1} items, if @var{x<>y} and @var{x<n} and
 1952: @var{y<n}, due to additional moves between registers.
 1953: @end itemize
 1954: 
 1955: In particular, keeping one item in a register is never a disadvantage,
 1956: if there are enough registers. Keeping two items in registers is a
 1957: disadvantage for frequent words like @code{?branch}, constants,
 1958: variables, literals and @code{i}. Therefore our generator only produces
 1959: code that keeps zero or one items in registers. The generated C code
 1960: covers both cases; the selection between these alternatives is made at
 1961: C-compile time using the switch @code{-DUSE_TOS}. @code{TOS} in the C
 1962: code for @code{+} is just a simple variable name in the one-item case,
 1963: otherwise it is a macro that expands into @code{sp[0]}. Note that the
 1964: GNU C compiler tries to keep simple variables like @code{TOS} in
 1965: registers, and it usually succeeds, if there are enough registers.
 1966: 
 1967: The primitive generator performs the TOS optimization for the
 1968: floating-point stack, too (@code{-DUSE_FTOS}). For floating-point
 1969: operations the benefit of this optimization is even larger:
 1970: floating-point operations take quite long on most processors, but can be
 1971: performed in parallel with other operations as long as their results are
 1972: not used. If the FP-TOS is kept in a register, this works. If
 1973: it is kept on the stack, i.e., in memory, the store into memory has to
 1974: wait for the result of the floating-point operation, lengthening the
 1975: execution time of the primitive considerably.
 1976: 
 1977: The TOS optimization makes the automatic generation of primitives a
 1978: bit more complicated. Just replacing all occurrences of @code{sp[0]} by
 1979: @code{TOS} is not sufficient. There are some special cases to
 1980: consider:
 1981: @itemize
 1982: @item In the case of @code{dup ( w -- w w )} the generator must not
 1983: eliminate the store to the original location of the item on the stack,
 1984: if the TOS optimization is turned on.
 1985: @item Primitives with stack effects of the form @code{--}
 1986: @var{out1}...@var{outy} must store the TOS to the stack at the start.
 1987: Likewise, primitives with the stack effect @var{in1}...@var{inx} @code{--}
 1988: must load the TOS from the stack at the end. But for the null stack
 1989: effect @code{--} no stores or loads should be generated.
 1990: @end itemize
 1991: 
 1992: @node Produced code,  , TOS Optimization, Primitives
 1993: @subsection Produced code
 1994: 
 1995: To see what assembly code is produced for the primitives on your machine
 1996: with your compiler and your flag settings, type @code{make engine.s} and
 1997: look at the resulting file @file{engine.s}.
 1998: 
 1999: @node System Architecture,  , Primitives, Internals
 2000: @section System Architecture
 2001: 
 2002: Our Forth system consists not only of primitives, but also of
 2003: definitions written in Forth. Since the Forth compiler itself belongs
 2004: to those definitions, it is not possible to start the system with the
 2005: primitives and the Forth source alone. Therefore we provide the Forth
 2006: code as an image file in nearly executable form. At the start of the
 2007: system a C routine loads the image file into memory, sets up the
 2008: memory (stacks etc.) according to information in the image file, and
 2009: starts executing Forth code.
 2010: 
 2011: The image file format is a compromise between the goals of making it
 2012: easy to generate image files and making them portable. The easiest way
 2013: to generate an image file is to just generate a memory dump. However,
 2014: this kind of image file cannot be used on a different machine, or on
 2015: the next version of the engine on the same machine, it even might not
 2016: work with the same engine compiled by a different version of the C
 2017: compiler. We would like to have as few versions of the image file as
 2018: possible, because we do not want to distribute many versions of the
 2019: same image file, and to make it easy for the users to use their image
 2020: files on many machines. We currently need to create a different image
 2021: file for machines with different cell sizes and different byte order
 2022: (little- or big-endian)@footnote{We consider adding information to the
 2023: image file that enables the loader to change the byte order.}.
 2024: 
 2025: Forth code that is going to end up in a portable image file has to
 2026: comply to some restrictions: addresses have to be stored in memory with
 2027: special words (@code{A!}, @code{A,}, etc.) in order to make the code
 2028: relocatable. Cells, floats, etc., have to be stored at the natural
 2029: alignment boundaries@footnote{E.g., store floats (8 bytes) at an address
 2030: dividable by~8. This happens automatically in our system when you use
 2031: the ANS Forth alignment words.}, in order to avoid alignment faults on
 2032: machines with stricter alignment. The image file is produced by a
 2033: metacompiler (@file{cross.fs}).
 2034: 
 2035: So, unlike the image file of Mitch Bradleys @code{cforth}, our image
 2036: file is not directly executable, but has to undergo some manipulations
 2037: during loading. Address relocation is performed at image load-time, not
 2038: at run-time. The loader also has to replace tokens standing for
 2039: primitive calls with the appropriate code-field addresses (or code
 2040: addresses in the case of direct threading).
 2041: 
 2042: @node Bugs, Pedigree, Internals, Top
 2043: @chapter Bugs
 2044: 
 2045: @node Pedigree, Word Index, Bugs, Top
 2046: @chapter Pedigree
 2047: 
 2048: @node Word Index, Node Index, Pedigree, Top
 2049: @chapter Word Index
 2050: 
 2051: @node Node Index,  , Word Index, Top
 2052: @chapter Node Index
 2053: 
 2054: @contents
 2055: @bye
 2056: 

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>