From 83c6612bf07228f70d25136c850f660721379632 Mon Sep 17 00:00:00 2001 From: rich Date: Sun, 7 Oct 2007 11:07:15 +0000 Subject: [PATCH] Restructure Makefile to add automated tests. Remove build-id section. Set up the data segment. COLD -> QUIT. Much rearrangement of the sections to improve overall memory usage. Don't align code. Simplify DUP. Simplify ?DUP. Comparison operations use set instead of jumps. Added C@C! and CMOVE. Removed _X, _Y, _Z. Make return_stack into a separate symbol for bottom of the stack, so it appears in disassembly. New symbol return_stack_top is R0. Added SYS_BRK. Simplify RDROP. WORD's and EMIT's buffers are now named. SNUMBER -> NUMBER, now far more powerful (can handle any number base and negative numbers correctly, and comes with a test suite). CREATE now no longer does an implicit WORD, but gets the name from the string at top of stack, allowing it to create anonymous words (see :NONAME). Added HIDE word. Removed INTERPRETER, added INTERPRET. Added PARSE ERROR with context in the INTERPRET function. Added SYSCALL0, 1, 2, 3. Added explicit initialisation of the data segment. Added UNLESS. Added C, WORDS now completely omits HIDDEN words. Generalise CFA> to work with any code/data pointer. Added execution tokens, EXECUTE, :NONAME, [']. Added exceptions, CATCH, THROW, ABORT. Added PRINT-STACK-TRACE. Added CSTRING. Removed STRNCMP. Added standard file and syscall words. Added UNUSED, MORECORE. Allow welcome message to be disabled (particularly during automated tests). Multiple documentation improvements. --- Makefile | 32 ++- jonesforth.S | 555 +++++++++++++++++++++++++++-------------- jonesforth.f | 520 ++++++++++++++++++++++++++++++-------- test_comparison.f | 73 ++++++ test_comparison.f.out | 70 ++++++ test_exception.f | 13 + test_exception.f.out | 6 + test_number.f | 11 + test_number.f.out | 5 + test_read_file.f | 24 ++ test_read_file.f.out | 7 + test_stack_trace.f | 9 + test_stack_trace.f.out | 6 + 13 files changed, 1040 insertions(+), 291 deletions(-) create mode 100644 test_comparison.f create mode 100644 test_comparison.f.out create mode 100644 test_exception.f create mode 100644 test_exception.f.out create mode 100644 test_number.f create mode 100644 test_number.f.out create mode 100644 test_read_file.f create mode 100644 test_read_file.f.out create mode 100644 test_stack_trace.f create mode 100644 test_stack_trace.f.out diff --git a/Makefile b/Makefile index 213eaf3..78e6c72 100644 --- a/Makefile +++ b/Makefile @@ -1,10 +1,34 @@ -# $Id: Makefile,v 1.5 2007-09-29 16:04:20 rich Exp $ +# $Id: Makefile,v 1.6 2007-10-07 11:07:15 rich Exp $ -all: - gcc -m32 -nostdlib -static -Wl,-Ttext,0 -o jonesforth jonesforth.S +SHELL := /bin/bash + +all: jonesforth + +jonesforth: jonesforth.S + gcc -m32 -nostdlib -static -Wl,-Ttext,0 -Wl,--build-id=none -o $@ $< run: - cat jonesforth.f - | ./jonesforth + cat jonesforth.f $(PROG) - | ./jonesforth + +clean: + rm -f jonesforth *~ core .test_* + +TESTS := $(patsubst %.f,%.test,$(wildcard test_*.f)) + +test check: $(TESTS) + +test_%.test: test_%.f jonesforth + @echo -n "$< ... " + @rm -f .$@ + @cat <(echo ': TEST-MODE ;') jonesforth.f $< <(echo 'TEST') | \ + ./jonesforth 2>&1 | \ + sed 's/DSP=[0-9]*//g' > .$@ + @diff -u .$@ $<.out + @rm -f .$@ + @echo "ok" + +.SUFFIXES: .f .test +.PHONY: test check remote: scp jonesforth.S jonesforth.f rjones@oirase:Desktop/ diff --git a/jonesforth.S b/jonesforth.S index c885150..081b537 100644 --- a/jonesforth.S +++ b/jonesforth.S @@ -1,11 +1,11 @@ /* A sometimes minimal FORTH compiler and tutorial for Linux / i386 systems. -*- asm -*- By Richard W.M. Jones http://annexia.org/forth This is PUBLIC DOMAIN (see public domain release statement below). - $Id: jonesforth.S,v 1.41 2007-09-29 23:11:27 rich Exp $ + $Id: jonesforth.S,v 1.42 2007-10-07 11:07:15 rich Exp $ - gcc -m32 -nostdlib -static -Wl,-Ttext,0 -o jonesforth jonesforth.S + gcc -m32 -nostdlib -static -Wl,-Ttext,0 -Wl,--build-id=none -o jonesforth jonesforth.S */ - .set JONES_VERSION,39 + .set JONES_VERSION,42 /* INTRODUCTION ---------------------------------------------------------------------- @@ -121,7 +121,7 @@ Again, to assemble this you will need gcc and gas (the GNU assembler). The commands to assemble and run the code (save this file as 'jonesforth.S') are: - gcc -m32 -nostdlib -static -Wl,-Ttext,0 -o jonesforth jonesforth.S + gcc -m32 -nostdlib -static -Wl,-Ttext,0 -Wl,--build-id=none -o jonesforth jonesforth.S cat jonesforth.f - | ./jonesforth If you want to run your own FORTH programs you can do: @@ -553,56 +553,27 @@ stack points -> | addr of DOUBLE | + 4 = +------------------+ This is what the set up code does. Does a tiny bit of house-keeping, sets up the separate return stack (NB: Linux gives us the ordinary parameter stack already), then - immediately jumps to a FORTH word called COLD. COLD stands for cold-start. In ISO - FORTH (but not in this FORTH), COLD can be called at any time to completely reset - the state of FORTH, and there is another word called WARM which does a partial reset. + immediately jumps to a FORTH word called QUIT. Despite its name, QUIT doesn't quit + anything. It resets some internal state and starts reading and interpreting commands. + (The reason it is called QUIT is because you can call QUIT from your own FORTH code + to "quit" your program and go back to interpreting). */ -/* ELF entry point. */ +/* Assembler entry point. */ .text .globl _start _start: cld - mov %esp,var_S0 // Store the initial data stack pointer. - mov $return_stack,%ebp // Initialise the return stack. + mov %esp,var_S0 // Save the initial data stack pointer in FORTH variable S0. + mov $return_stack_top,%ebp // Initialise the return stack. + call set_up_data_segment mov $cold_start,%esi // Initialise interpreter. NEXT // Run interpreter! .section .rodata cold_start: // High-level code without a codeword. - .int COLD - -/* - We also allocate some space for the return stack and some space to store user - definitions. These are static memory allocations using fixed-size buffers, but it - wouldn't be a great deal of work to make them dynamic. -*/ - - .bss -/* FORTH return stack. */ - .set RETURN_STACK_SIZE,8192 - .align 4096 - .space RETURN_STACK_SIZE -return_stack: // Initial top of return stack. - -/* The user definitions area: space for user-defined words and general memory allocations. */ - .set USER_DEFS_SIZE,65536 - .align 4096 -user_defs_start: - .space USER_DEFS_SIZE - -/* This is used as a temporary input buffer when reading from files or the terminal. */ - .set BUFFER_SIZE,4096 - .align 4096 -buffer: -_initbufftop: - .space BUFFER_SIZE -buffend: -currkey: - .int buffer -bufftop: - .int _initbufftop + .int QUIT /* BUILT-IN WORDS ---------------------------------------------------------------------- @@ -664,7 +635,7 @@ name_\label : .set link,name_\label .byte \flags+\namelen // flags + length byte .ascii "\name" // the name - .align 4 + .align 4 // padding to next 4 byte boundary .globl \label \label : .int DOCOL // codeword - the interpreter @@ -701,12 +672,12 @@ name_\label : .set link,name_\label .byte \flags+\namelen // flags + length byte .ascii "\name" // the name - .align 4 + .align 4 // padding to next 4 byte boundary .globl \label \label : .int code_\label // codeword .text - .align 4 + //.align 4 .globl code_\label code_\label : // assembler code follows .endm @@ -717,23 +688,22 @@ code_\label : // assembler code follows you can skip the details. */ - defcode "DUP",3,,DUP - pop %eax // duplicate top of stack - push %eax - push %eax - NEXT - defcode "DROP",4,,DROP pop %eax // drop top of stack NEXT defcode "SWAP",4,,SWAP - pop %eax // swap top of stack + pop %eax // swap top two elements on stack pop %ebx push %eax push %ebx NEXT + defcode "DUP",3,,DUP + mov (%esp),%eax // duplicate top of stack + push %eax + NEXT + defcode "OVER",4,,OVER mov 4(%esp),%eax // get the second element of stack push %eax // and push it on top @@ -758,12 +728,11 @@ code_\label : // assembler code follows NEXT defcode "?DUP",4,,QDUP // duplicate top of stack if non-zero - pop %eax + movl (%esp),%eax test %eax,%eax jz 1f push %eax -1: push %eax - NEXT +1: NEXT defcode "1+",2,,INCR incl (%esp) // increment top of stack @@ -801,7 +770,7 @@ code_\label : // assembler code follows /* In this FORTH, only /MOD is primitive. Later we will define the / and MOD words in terms of the primitive /MOD. The design of the i386 assembly instruction idiv which - leaves both quotient and remainder makes this obvious choice. + leaves both quotient and remainder makes this the obvious choice. */ defcode "/MOD",4,,DIVMOD @@ -813,118 +782,115 @@ code_\label : // assembler code follows push %eax // push quotient NEXT +/* + Lots of comparison operations. + + ANS FORTH says that the comparison words should return all (binary) 1's for + TRUE and all 0's for FALSE. However this is a bit of a strange convention + so this FORTH breaks it and returns the more normal (for C programmers ...) + 1 meaning TRUE and 0 meaning FALSE. +*/ + defcode "=",1,,EQU // top two words are equal? pop %eax pop %ebx cmp %ebx,%eax - je 1f - pushl $0 - NEXT -1: pushl $1 + sete %al + movzbl %al,%eax + pushl %eax NEXT defcode "<>",2,,NEQU // top two words are not equal? pop %eax pop %ebx cmp %ebx,%eax - je 1f - pushl $1 - NEXT -1: pushl $0 + setne %al + movzbl %al,%eax + pushl %eax NEXT defcode "<",1,,LT pop %eax pop %ebx cmp %eax,%ebx - jl 1f - pushl $0 - NEXT -1: pushl $1 + setl %al + movzbl %al,%eax + pushl %eax NEXT defcode ">",1,,GT pop %eax pop %ebx cmp %eax,%ebx - jg 1f - pushl $0 - NEXT -1: pushl $1 + setg %al + movzbl %al,%eax + pushl %eax NEXT defcode "<=",2,,LE pop %eax pop %ebx cmp %eax,%ebx - jle 1f - pushl $0 - NEXT -1: pushl $1 + setle %al + movzbl %al,%eax + pushl %eax NEXT defcode ">=",2,,GE pop %eax pop %ebx cmp %eax,%ebx - jge 1f - pushl $0 - NEXT -1: pushl $1 + setge %al + movzbl %al,%eax + pushl %eax NEXT defcode "0=",2,,ZEQU // top of stack equals 0? pop %eax test %eax,%eax - jz 1f - pushl $0 - NEXT -1: pushl $1 + setz %al + movzbl %al,%eax + pushl %eax NEXT defcode "0<>",3,,ZNEQU // top of stack not 0? pop %eax test %eax,%eax - jnz 1f - pushl $0 - NEXT -1: pushl $1 + setnz %al + movzbl %al,%eax + pushl %eax NEXT defcode "0<",2,,ZLT // comparisons with 0 pop %eax test %eax,%eax - jl 1f - pushl $0 - NEXT -1: pushl $1 + setl %al + movzbl %al,%eax + pushl %eax NEXT defcode "0>",2,,ZGT pop %eax test %eax,%eax - jg 1f - pushl $0 - NEXT -1: pushl $1 + setg %al + movzbl %al,%eax + pushl %eax NEXT defcode "0<=",3,,ZLE pop %eax test %eax,%eax - jle 1f - pushl $0 - NEXT -1: pushl $1 + setle %al + movzbl %al,%eax + pushl %eax NEXT defcode "0>=",3,,ZGE pop %eax test %eax,%eax - jge 1f - pushl $0 - NEXT -1: pushl $1 + setge %al + movzbl %al,%eax + pushl %eax NEXT defcode "AND",3,,AND // bitwise AND @@ -942,7 +908,7 @@ code_\label : // assembler code follows xorl %eax,(%esp) NEXT - defcode "INVERT",6,,INVERT // this is the FORTH bitwise "NOT" function (cf. NEGATE) + defcode "INVERT",6,,INVERT // this is the FORTH bitwise "NOT" function (cf. NEGATE and NOT) notl (%esp) NEXT @@ -1011,8 +977,8 @@ code_\label : // assembler code follows +---------------------------+-------+-------+-------+-------+-------+ LIT is executed in the normal way, but what it does next is definitely not normal. It - looks at %esi (which now points to the literal 2), grabs it, pushes it on the stack, then - manipulates %esi in order to skip the literal as if it had never been there. + looks at %esi (which now points to the number 2), grabs it, pushes it on the stack, then + manipulates %esi in order to skip the number as if it had never been there. What's neat is that the whole grab/manipulate can be done using a single byte single i386 instruction, our old friend LODSL. Rather than me drawing more ASCII-art diagrams, @@ -1079,6 +1045,26 @@ code_\label : // assembler code follows push %eax // push value onto stack NEXT +/* C@C! is a useful byte copy primitive. */ + defcode "C@C!",4,,CCOPY + movl 4(%esp),%ebx // source address + movb (%ebx),%al // get source character + pop %edi // destination address + stosb // copy to destination + push %edi // increment destination address + incl 4(%esp) // increment source address + NEXT + +/* and CMOVE is a block copy operation. */ + defcode "CMOVE",5,,CMOVE + mov %esi,%edx // preserve %esi + pop %ecx // length + pop %edi // destination address + pop %esi // source address + rep movsb // copy source to destination + mov %edx,%esi // restore %esi + NEXT + /* BUILT-IN VARIABLES ---------------------------------------------------------------------- @@ -1110,19 +1096,13 @@ var_\name : STATE Is the interpreter executing code (0) or compiling a word (non-zero)? LATEST Points to the latest (most recently defined) word in the dictionary. HERE Points to the next free byte of memory. When compiling, compiled words go here. - _X These are three scratch variables, used by some standard dictionary words. - _Y - _Z S0 Stores the address of the top of the parameter stack. BASE The current base for printing and reading numbers. */ defvar "STATE",5,,STATE - defvar "HERE",4,,HERE,user_defs_start - defvar "LATEST",6,,LATEST,name_SYSCALL3 // SYSCALL3 must be last in built-in dictionary - defvar "_X",2,,TX - defvar "_Y",2,,TY - defvar "_Z",2,,TZ + defvar "HERE",4,,HERE + defvar "LATEST",6,,LATEST,name_SYSCALL0 // SYSCALL0 must be last in built-in dictionary defvar "S0",2,,SZ defvar "BASE",4,,BASE,10 @@ -1154,7 +1134,7 @@ var_\name : .endm defconst "VERSION",7,,VERSION,JONES_VERSION - defconst "R0",2,,RZ,return_stack + defconst "R0",2,,RZ,return_stack_top defconst "DOCOL",5,,__DOCOL,DOCOL defconst "F_IMMED",7,,__F_IMMED,F_IMMED defconst "F_HIDDEN",8,,__F_HIDDEN,F_HIDDEN @@ -1166,6 +1146,7 @@ var_\name : defconst "SYS_READ",8,,SYS_READ,__NR_read defconst "SYS_WRITE",9,,SYS_WRITE,__NR_write defconst "SYS_CREAT",9,,SYS_CREAT,__NR_creat + defconst "SYS_BRK",7,,SYS_BRK,__NR_brk defconst "O_RDONLY",8,,__O_RDONLY,0 defconst "O_WRONLY",8,,__O_WRONLY,1 @@ -1202,7 +1183,7 @@ var_\name : NEXT defcode "RDROP",5,,RDROP - lea 4(%ebp),%ebp // pop return stack and throw away + addl $4,%ebp // pop return stack and throw away NEXT /* @@ -1240,11 +1221,24 @@ var_\name : and compiling code, we might be reading words to execute, we might be asking for the user to type their name -- ultimately it all comes in through KEY. - The implementation of KEY uses an input buffer of a certain size (defined at the end of the - program). It calls the Linux read(2) system call to fill this buffer and tracks its position + The implementation of KEY uses an input buffer of a certain size (defined at the start of this + file). It calls the Linux read(2) system call to fill this buffer and tracks its position in the buffer using a couple of variables, and if it runs out of input buffer then it refills it automatically. The other thing that KEY does is if it detects that stdin has closed, it exits the program, which is why when you hit ^D the FORTH system cleanly exits. + + buffer bufftop + | | + V V + +-------------------------------+--------------------------------------+ + | INPUT READ FROM STDIN ....... | unused part of the buffer | + +-------------------------------+--------------------------------------+ + ^ + | + currkey (next character to read) + + <---------------------- BUFFER_SIZE (4096 bytes) ----------------------> + */ defcode "KEY",3,,KEY @@ -1254,18 +1248,18 @@ var_\name : _KEY: mov (currkey),%ebx cmp (bufftop),%ebx - jge 1f + jge 1f // exhausted the input buffer? xor %eax,%eax mov (%ebx),%al inc %ebx mov %ebx,(currkey) ret -1: // out of input; use read(2) to fetch more input from stdin +1: // Out of input; use read(2) to fetch more input from stdin. xor %ebx,%ebx // 1st param: stdin mov $buffer,%ecx // 2nd param: buffer mov %ecx,currkey - mov $buffend-buffer,%edx // 3rd param: max length + mov $BUFFER_SIZE,%edx // 3rd param: max length mov $__NR_read,%eax // syscall: read int $0x80 test %eax,%eax // If %eax <= 0, then exit. @@ -1274,11 +1268,18 @@ _KEY: mov %ecx,bufftop jmp _KEY -2: // error or out of input: exit +2: // Error or end of input: exit the program. xor %ebx,%ebx mov $__NR_exit,%eax // syscall: exit int $0x80 + .data + .align 4 +currkey: + .int buffer // Current place in input buffer (next character to read). +bufftop: + .int buffer // Last valid data in input buffer + 1. + /* By contrast, output is much simpler. The FORTH word EMIT writes out a single byte to stdout. This implementation just uses the write system call. No attempt is made to buffer output, but @@ -1293,8 +1294,8 @@ _EMIT: mov $1,%ebx // 1st param: stdout // write needs the address of the byte to write - mov %al,(2f) - mov $2f,%ecx // 2nd param: address + mov %al,emit_scratch + mov $emit_scratch,%ecx // 2nd param: address mov $1,%edx // 3rd param: nbytes = 1 @@ -1302,8 +1303,9 @@ _EMIT: int $0x80 ret - .bss -2: .space 1 // scratch used by EMIT + .data // NB: easier to fit in the .data section +emit_scratch: + .space 1 // scratch used by EMIT /* Back to input, WORD is a FORTH word which reads the next full word of input. @@ -1317,9 +1319,11 @@ _EMIT: a static C string). Also notice that WORD's internal buffer is just 32 bytes long and there is NO checking for overflow. 31 bytes happens to be the maximum length of a FORTH word that we support, and that is what WORD is used for: to read FORTH words when - we are compiling and executing code. The returned strings are not NUL-terminated, so - in some crazy-world you could define FORTH words containing ASCII NULs, although why - you'd want to is a bit beyond me. + we are compiling and executing code. The returned strings are not NUL-terminated. + + Start address+length is the normal way to represent strings in FORTH (not ending in an + ASCII NUL character as in C), and so FORTH strings can contain any character including NULs + and can be any length. WORD is not suitable for just reading strings (eg. user input) because of all the above peculiarities and limitations. @@ -1348,7 +1352,7 @@ _WORD: jbe 1b // if so, keep looking /* Search for the end of the word, storing chars as we go. */ - mov $5f,%edi // pointer to return buffer + mov $word_buffer,%edi // pointer to return buffer 2: stosb // add character to return buffer call _KEY // get next key, returned in %al @@ -1356,9 +1360,9 @@ _WORD: ja 2b // if not, keep looping /* Return the word (well, the static buffer) and length. */ - sub $5f,%edi + sub $word_buffer,%edi mov %edi,%ecx // return length of the word - mov $5f,%edi // return address of the word + mov $word_buffer,%edi // return address of the word ret /* Code to skip \ comments to end of the current line. */ @@ -1368,40 +1372,88 @@ _WORD: jne 3b jmp 1b - .bss + .data // NB: easier to fit in the .data section // A static buffer where WORD returns. Subsequent calls // overwrite this buffer. Maximum word length is 32 chars. -5: .space 32 +word_buffer: + .space 32 /* As well as reading in words we'll need to read in numbers and for that we are using a function - called SNUMBER. This parses a numeric string such as one returned by WORD and pushes the + called NUMBER. This parses a numeric string such as one returned by WORD and pushes the number on the parameter stack. - This function does absolutely no error checking, and in particular the length of the string - must be >= 1 bytes, and should contain only digits 0-9. If it doesn't you'll get random results. + The function uses the variable BASE as the base (radix) for conversion, so for example if + BASE is 2 then we expect a binary number. Normally BASE is 10. - This function is only used when reading literal numbers in code, and shouldn't really be used - in user code at all. + If the word starts with a '-' character then the returned value is negative. + + If the string can't be parsed as a number (or contains characters outside the current BASE) + then we need to return an error indication. So NUMBER actually returns two items on the stack. + At the top of stack we return the number of unconverted characters (ie. if 0 then all characters + were converted, so there is no error). Second from top of stack is the parsed number or a + partial value if there was an error. */ - defcode "SNUMBER",7,,SNUMBER - pop %edi - pop %ecx - call _SNUMBER - push %eax + defcode "NUMBER",6,,NUMBER + pop %ecx // length of string + pop %edi // start address of string + call _NUMBER + push %eax // parsed number + push %ecx // number of unparsed characters (0 = no error) NEXT -_SNUMBER: + +_NUMBER: xor %eax,%eax xor %ebx,%ebx -1: - imull $10,%eax // %eax *= 10 - movb (%edi),%bl + + test %ecx,%ecx // trying to parse a zero-length string is an error, but will return 0. + jz 5f + + movl var_BASE,%edx // get BASE (in %dl) + + // Check if first character is '-'. + movb (%edi),%bl // %bl = first character in string + inc %edi + push %eax // push 0 on stack + cmpb $'-',%bl // negative number? + jnz 2f + pop %eax + push %ebx // push <> 0 on stack, indicating negative + dec %ecx + jnz 1f + pop %ebx // error: string is only '-'. + movl $1,%ecx + ret + + // Loop reading digits. +1: imull %edx,%eax // %eax *= BASE + movb (%edi),%bl // %bl = next character in string inc %edi - subb $'0',%bl // ASCII -> digit + + // Convert 0-9, A-Z to a number 0-35. +2: subb $'0',%bl // < '0'? + jb 4f + cmp $10,%bl // <= '9'? + jb 3f + subb $17,%bl // < 'A'? (17 is 'A'-'0') + jb 4f + addb $10,%bl + +3: cmp %dl,%bl // >= BASE? + jge 4f + + // OK, so add it to %eax and loop. add %ebx,%eax dec %ecx jnz 1b - ret + + // Negate the result if first character was '-' (saved on the stack). +4: pop %ebx + test %ebx,%ebx + jz 5f + neg %eax + +5: ret /* DICTIONARY LOOK UPS ---------------------------------------------------------------------- @@ -1439,8 +1491,7 @@ _FIND: // Now we start searching backwards through the dictionary for this word. mov var_LATEST,%edx // LATEST points to name header of the latest word in the dictionary -1: - test %edx,%edx // NULL pointer? (end of the linked list) +1: test %edx,%edx // NULL pointer? (end of the linked list) je 4f // Compare the length expected and the length of the word. @@ -1466,8 +1517,7 @@ _FIND: mov %edx,%eax ret -2: - mov (%edx),%edx // Move back through the link field to the previous word +2: mov (%edx),%edx // Move back through the link field to the previous word jmp 1b // .. and loop. 4: // Not found. @@ -1491,6 +1541,7 @@ _FIND: +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+ | LINK | 6 | D | O | U | B | L | E | 0 | DOCOL | DUP | + | EXIT | +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+ + codeword Notes: @@ -1499,7 +1550,8 @@ _FIND: In this FORTH you cannot easily turn a codeword pointer back into a dictionary entry pointer, but that is not true in most FORTH implementations where they store a back pointer in the definition (with an obvious memory/complexity cost). The reason they do this is that it is useful to be - able to go backwards (codeword -> dictionary entry) in order to decompile FORTH definitions. + able to go backwards (codeword -> dictionary entry) in order to decompile FORTH definitions + quickly. What does CFA stand for? My best guess is "Code Field Address". */ @@ -1533,6 +1585,7 @@ _TCFA: +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+ | LINK | 6 | D | O | U | B | L | E | 0 | DOCOL | DUP | + | EXIT | +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+ + codeword (Note to those following the source of FIG-FORTH / ciforth: My >DFA definition is different from theirs, because they have an extra indirection). @@ -1570,7 +1623,7 @@ _TCFA: FORTH solves this rather elegantly and as you might expect in a very low-level way which allows you to change how the compiler works on your own code. - FORTH has an INTERPRETER function (a true interpreter this time, not DOCOL) which runs in a + FORTH has an INTERPRET function (a true interpreter this time, not DOCOL) which runs in a loop, reading words (using WORD), looking them up (using FIND), turning them into codeword pointers (using >CFA) and deciding what to do with them. @@ -1581,7 +1634,7 @@ _TCFA: The interesting stuff happens when STATE is non-zero -- compiling mode. In this mode the interpreter appends the codeword pointer to user memory (the HERE variable points to the next - free byte of user memory). + free byte of user memory -- see DATA SEGMENT section below). So you may be able to see how we could define : (COLON). The general plan is: @@ -1642,6 +1695,7 @@ _TCFA: IMMEDIATE then the interpreter runs it immediately _even if it's in compile mode_. This is how the word ; (SEMICOLON) works -- as a word flagged in the dictionary as IMMEDIATE. + And all it does is append the codeword for EXIT on to the current definition and switch back to immediate mode (set STATE back to 0). Shortly we'll see the actual definition of ; and we'll see that it's really a very simple definition, declared IMMEDIATE. @@ -1654,7 +1708,6 @@ _TCFA: len pad codeword ^ | HERE - STATE is set to 0. And that's it, job done, our new definition is compiled, and we're back in immediate mode @@ -1697,9 +1750,9 @@ _TCFA: defcode "CREATE",6,,CREATE - // Get the word. - call _WORD // Returns %ecx = length, %edi = pointer to word. - mov %edi,%ebx // %ebx = address of the word + // Get the name length and address. + pop %ecx // %ecx = length + pop %ebx // %ebx = address of name // Link pointer. movl var_HERE,%edi // %edi is the address of the header @@ -1727,7 +1780,7 @@ _TCFA: to use. The first is , (COMMA) which is a standard FORTH word which appends a 32 bit integer to the user - data area pointed to by HERE, and adds 4 to HERE. So the action of , (COMMA) is: + memory pointed to by HERE, and adds 4 to HERE. So the action of , (COMMA) is: previous value of HERE | @@ -1789,6 +1842,7 @@ _COMMA: */ defword ":",1,,COLON + .int WORD // Get the name of the new word .int CREATE // CREATE the dictionary entry / header .int LIT, DOCOL, COMMA // Append DOCOL (the codeword). .int LATEST, FETCH, HIDDEN // Make the word hidden (see below for definition). @@ -1845,6 +1899,8 @@ _COMMA: LATEST @ HIDDEN + 'HIDE word' toggles the flag on a named 'word'. + Setting this flag stops the word from being found by FIND, and so can be used to make 'private' words. For example, to break up a large word into smaller parts you might do: @@ -1852,9 +1908,9 @@ _COMMA: : SUB2 ... subword ... ; : SUB3 ... subword ... ; : MAIN ... defined in terms of SUB1, SUB2, SUB3 ... ; - WORD SUB1 FIND HIDDEN \ Hide SUB1 - WORD SUB2 FIND HIDDEN \ Hide SUB2 - WORD SUB3 FIND HIDDEN \ Hide SUB3 + HIDE SUB1 + HIDE SUB2 + HIDE SUB3 After this, only MAIN is 'exported' or seen by the rest of the program. */ @@ -1865,6 +1921,12 @@ _COMMA: xorb $F_HIDDEN,(%edi) // Toggle the HIDDEN bit. NEXT + defword "HIDE",4,,HIDE + .int WORD // Get the word (after HIDE). + .int FIND // Look up in the dictionary. + .int HIDDEN // Set F_HIDDEN flag. + .int EXIT // Return. + /* ' (TICK) is a standard FORTH word which returns the codeword pointer of the next word. @@ -1978,24 +2040,28 @@ _COMMA: NEXT /* - COLD START AND INTERPRETER ---------------------------------------------------------------------- + QUIT AND INTERPRET ---------------------------------------------------------------------- - COLD is the first FORTH function called, almost immediately after the FORTH system "boots". + QUIT is the first FORTH function called, almost immediately after the FORTH system "boots". + As explained before, QUIT doesn't "quit" anything. It does some initialisation (in particular + it clears the return stack) and it calls INTERPRET in a loop to interpret commands. The + reason it is called QUIT is because you can call it from your own FORTH words in order to + "quit" your program and start again at the user prompt. - INTERPRETER is the FORTH interpreter ("toploop", "toplevel" or "REPL" might be a more accurate + INTERPRET is the FORTH interpreter ("toploop", "toplevel" or "REPL" might be a more accurate description -- see: http://en.wikipedia.org/wiki/REPL). */ - // COLD must not return (ie. must not call EXIT). - defword "COLD",4,,COLD - .int INTERPRETER // call the interpreter loop (never returns) + // QUIT must not return (ie. must not call EXIT). + defword "QUIT",4,,QUIT + .int RZ,RSPSTORE // R0 RSP!, clear the return stack + .int INTERPRET // interpret the next word + .int BRANCH,-8 // and loop (indefinitely) -/* This interpreter is pretty simple, but remember that in FORTH you can always override - * it later with a more powerful one! +/* + This interpreter is pretty simple, but remember that in FORTH you can always override + it later with a more powerful one! */ - defword "INTERPRETER",11,,INTERPRETER - .int INTERPRET,RDROP,INTERPRETER - defcode "INTERPRET",9,,INTERPRET call _WORD // Returns %ecx = length, %edi = pointer to word. @@ -2020,7 +2086,9 @@ _COMMA: 1: // Not in the dictionary (not a word) so assume it's a literal number. incl interpret_is_lit - call _SNUMBER // Returns the parsed number in %eax + call _NUMBER // Returns the parsed number in %eax, %ecx > 0 if error + test %ecx,%ecx + jnz 6f mov %eax,%ebx mov $LIT,%eax // The word is LIT @@ -2044,14 +2112,44 @@ _COMMA: jnz 5f // Not a literal, execute it now. This never returns, but the codeword will - // eventually call NEXT which will reenter the loop in INTERPRETER. + // eventually call NEXT which will reenter the loop in QUIT. jmp *(%eax) 5: // Executing a literal, which means push it on the stack. push %ebx NEXT - .data +6: // Parse error (not a known word or a number in the current BASE). + // Print an error message followed by up to 40 characters of context. + mov $2,%ebx // 1st param: stderr + mov $errmsg,%ecx // 2nd param: error message + mov $errmsgend-errmsg,%edx // 3rd param: length of string + mov $__NR_write,%eax // write syscall + int $0x80 + + mov (currkey),%ecx // the error occurred just before currkey position + mov %ecx,%edx + sub $buffer,%edx // %edx = currkey - buffer (length in buffer before currkey) + cmp $40,%edx // if > 40, then print only 40 characters + jle 7f + mov $40,%edx +7: sub %edx,%ecx // %ecx = start of area to print, %edx = length + mov $__NR_write,%eax // write syscall + int $0x80 + + mov $errmsgnl,%ecx // newline + mov $1,%edx + mov $__NR_write,%eax // write syscall + int $0x80 + + NEXT + + .section .rodata +errmsg: .ascii "PARSE ERROR: " +errmsgend: +errmsgnl: .ascii "\n" + + .data // NB: easier to fit in the .data section .align 4 interpret_is_lit: .int 0 // Flag used to record if reading a literal @@ -2062,12 +2160,16 @@ interpret_is_lit: CHAR puts the ASCII code of the first character of the following word on the stack. For example CHAR A puts 65 on the stack. - SYSCALL3 makes a standard Linux system call. (See for a list of system call - numbers). This is the form to use when the function takes up to three parameters. + EXECUTE is used to run execution tokens. See the discussion of execution tokens in the + FORTH code for more details. + + SYSCALL0, SYSCALL1, SYSCALL2, SYSCALL3 make a standard Linux system call. (See + for a list of system call numbers). As their name suggests these forms take between 0 and 3 + syscall parameters, plus the system call number. - In this FORTH, SYSCALL3 must be the last word in the built-in (assembler) dictionary because we + In this FORTH, SYSCALL0 must be the last word in the built-in (assembler) dictionary because we initialise the LATEST variable to point to it. This means that if you want to extend the assembler - part, you must put new words before SYSCALL3, or else change how LATEST is initialised. + part, you must put new words before SYSCALL0, or else change how LATEST is initialised. */ defcode "CHAR",4,,CHAR @@ -2077,6 +2179,11 @@ interpret_is_lit: push %eax // Push it onto the stack. NEXT + defcode "EXECUTE",7,,EXECUTE + pop %eax // Get xt into %eax + jmp *(%eax) // and jump to it. + // After xt runs its NEXT will continue executing the current word. + defcode "SYSCALL3",8,,SYSCALL3 pop %eax // System call number (see ) pop %ebx // First parameter. @@ -2086,6 +2193,83 @@ interpret_is_lit: push %eax // Result (negative for -errno) NEXT + defcode "SYSCALL2",8,,SYSCALL2 + pop %eax // System call number (see ) + pop %ebx // First parameter. + pop %ecx // Second parameter + int $0x80 + push %eax // Result (negative for -errno) + NEXT + + defcode "SYSCALL1",8,,SYSCALL1 + pop %eax // System call number (see ) + pop %ebx // First parameter. + int $0x80 + push %eax // Result (negative for -errno) + NEXT + + defcode "SYSCALL0",8,,SYSCALL0 + pop %eax // System call number (see ) + int $0x80 + push %eax // Result (negative for -errno) + NEXT + +/* + DATA SEGMENT ---------------------------------------------------------------------- + + Here we set up the Linux data segment, used for user definitions and variously known as just + the 'data segment', 'user memory' or 'user definitions area'. It is an area of memory which + grows upwards and stores both newly-defined FORTH words and global variables of various + sorts. + + It is completely analogous to the C heap, except there is no generalised 'malloc' and 'free' + (but as with everything in FORTH, writing such functions would just be a Simple Matter + Of Programming). Instead in normal use the data segment just grows upwards as new FORTH + words are defined/appended to it. + + There are various "features" of the GNU toolchain which make setting up the data segment + more complicated than it really needs to be. One is the GNU linker which inserts a random + "build ID" segment. Another is Address Space Randomization which means we can't tell + where the kernel will choose to place the data segment (or the stack for that matter). + + Therefore writing this set_up_data_segment assembler routine is a little more complicated + than it really needs to be. We ask the Linux kernel where it thinks the data segment starts + using the brk(2) system call, then ask it to reserve some initial space (also using brk(2)). + + You don't need to worry about this code. +*/ + .text + .set INITIAL_DATA_SEGMENT_SIZE,65536 +set_up_data_segment: + xor %ebx,%ebx // Call brk(0) + movl $__NR_brk,%eax + int $0x80 + movl %eax,var_HERE // Initialise HERE to point at beginning of data segment. + addl $INITIAL_DATA_SEGMENT_SIZE,%eax // Reserve nn bytes of memory for initial data segment. + movl %eax,%ebx // Call brk(HERE+INITIAL_DATA_SEGMENT_SIZE) + movl $__NR_brk,%eax + int $0x80 + ret + +/* + We allocate static buffers for the return static and input buffer (used when + reading in files and text that the user types in). +*/ + .set RETURN_STACK_SIZE,8192 + .set BUFFER_SIZE,4096 + + .bss +/* FORTH return stack. */ + .align 4096 +return_stack: + .space RETURN_STACK_SIZE +return_stack_top: // Initial top of return stack. + +/* This is used as a temporary input buffer when reading from files or the terminal. */ + .align 4096 +buffer: + .space BUFFER_SIZE + /* START OF FORTH CODE ---------------------------------------------------------------------- @@ -2094,7 +2278,8 @@ interpret_is_lit: languages would be considered rather fundamental. I used to append this here in the assembly file, but I got sick of fighting against gas's - stupid (lack of) multiline string syntax. So now that is in a separate file called jonesforth.f + crack-smoking (lack of) multiline string syntax. So now that is in a separate file called + jonesforth.f If you don't already have that file, download it from http://annexia.org/forth in order to continue the tutorial. diff --git a/jonesforth.f b/jonesforth.f index f3eebe3..025d9b0 100644 --- a/jonesforth.f +++ b/jonesforth.f @@ -2,7 +2,7 @@ \ A sometimes minimal FORTH compiler and tutorial for Linux / i386 systems. -*- asm -*- \ By Richard W.M. Jones http://annexia.org/forth \ This is PUBLIC DOMAIN (see public domain release statement below). -\ $Id: jonesforth.f,v 1.12 2007-09-30 14:37:00 rich Exp $ +\ $Id: jonesforth.f,v 1.13 2007-10-07 11:07:15 rich Exp $ \ \ The first part of this tutorial is in jonesforth.S. Get if from http://annexia.org/forth \ @@ -60,11 +60,6 @@ \ SPACE prints a space : SPACE BL EMIT ; -\ DUP, DROP are defined in assembly for speed, but this is how you might define them -\ in FORTH. Notice use of the scratch variables _X and _Y. -\ : DUP _X ! _X @ _X @ ; -\ : DROP _X ! ; - \ The 2... versions of the standard operators work on pairs of stack entries. They're not used \ very commonly so not really worth writing in assembler. Here is how they are defined in FORTH. : 2DUP OVER OVER ; @@ -214,6 +209,19 @@ SWAP ! \ and back-fill it in the original location ; +\ UNLESS is the same as IF but the test is reversed. +\ +\ Note the use of [COMPILE]: Since IF is IMMEDIATE we don't want it to be executed while UNLESS +\ is compiling, but while UNLESS is running (which happens to be when whatever word using UNLESS is +\ being compiled -- whew!). So we use [COMPILE] to reverse the effect of marking IF as immediate. +\ This trick is generally used when we want to write our own control words without having to +\ implement them all in terms of the primitives 0BRANCH and BRANCH, but instead reusing simpler +\ control words like (in this instance) IF. +: UNLESS IMMEDIATE + ' NOT , \ compile NOT (to reverse the test) + [COMPILE] IF \ continue by calling the normal IF +; + \ COMMENTS ---------------------------------------------------------------------- \ \ FORTH allows ( ... ) as comments within function definitions. This works by having an IMMEDIATE @@ -456,6 +464,12 @@ case we put the string at HERE (but we _don't_ change HERE). This is meant as a temporary location, likely to be overwritten soon after. ) +( C, appends a byte to the current compiled word. ) +: C, + HERE @ C! ( store the character in the compiled image ) + 1 HERE +! ( increment HERE pointer by 1 byte ) +; + : S" IMMEDIATE ( -- addr len ) STATE @ IF ( compiling? ) ' LITSTRING , ( compile LITSTRING ) @@ -465,8 +479,7 @@ KEY ( get next character of the string ) DUP '"' <> WHILE - HERE @ C! ( store the character in the compiled image ) - 1 HERE +! ( increment HERE pointer by 1 byte ) + C, ( copy character ) REPEAT DROP ( drop the double quote character at the end ) DUP ( get the saved address of the length word ) @@ -554,7 +567,8 @@ The trick is to define a new word for the variable itself (eg. if the variable was called 'VAR' then we would define a new word called VAR). This is easy to do because we exposed dictionary entry creation through the CREATE word (part of the definition of : above). - A call to CREATE TEN leaves the dictionary entry: + A call to WORD [TEN] CREATE (where [TEN] means that "TEN" is the next word in the input) + leaves the dictionary entry: +--- HERE | @@ -580,7 +594,8 @@ assembler part which returns the value of the assembler symbol of the same name. ) : CONSTANT - CREATE ( make the dictionary entry (the name follows CONSTANT) ) + WORD ( get the name (the name follows CONSTANT) ) + CREATE ( make the dictionary entry ) DOCOL , ( append DOCOL (the codeword field of this word) ) ' LIT , ( append the codeword LIT ) , ( append the value on the top of the stack ) @@ -589,10 +604,10 @@ ( VARIABLE is a little bit harder because we need somewhere to put the variable. There is - nothing particularly special about the 'user definitions area' (the area of memory pointed - to by HERE where we have previously just stored new word definitions). We can slice off - bits of this memory area to store anything we want, so one possible definition of - VARIABLE might create this: + nothing particularly special about the user memory (the area of memory pointed to by HERE + where we have previously just stored new word definitions). We can slice off bits of this + memory area to store anything we want, so one possible definition of VARIABLE might create + this: +--------------------------------------------------------------+ | | @@ -605,7 +620,7 @@ where is the place to store the variable, and points back to it. To make this more general let's define a couple of words which we can use to allocate - arbitrary memory from the user definitions area. + arbitrary memory from the user memory. First ALLOT, where n ALLOT allocates n bytes of memory. (Note when calling this that it's a very good idea to make sure that n is a multiple of 4, or at least that next time @@ -629,7 +644,7 @@ ) : VARIABLE 1 CELLS ALLOT ( allocate 1 cell of memory, push the pointer to this memory ) - CREATE ( make the dictionary entry (the name follows VARIABLE) ) + WORD CREATE ( make the dictionary entry (the name follows VARIABLE) ) DOCOL , ( append DOCOL (the codeword field of this word) ) ' LIT , ( append the codeword LIT ) , ( append the pointer to the new memory ) @@ -687,7 +702,7 @@ way cannot be inlined). ) : VALUE ( n -- ) - CREATE ( make the dictionary entry (the name follows VALUE) ) + WORD CREATE ( make the dictionary entry (the name follows VALUE) ) DOCOL , ( append DOCOL ) ' LIT , ( append the codeword LIT ) , ( append the initial value ) @@ -775,8 +790,8 @@ WHILE DUP ?HIDDEN NOT IF ( ignore hidden words ) DUP ID. ( but if not hidden, print the word ) + SPACE THEN - SPACE @ ( dereference the link pointer - go to previous word ) REPEAT CR @@ -967,11 +982,14 @@ DECOMPILER ---------------------------------------------------------------------- CFA> is the opposite of >CFA. It takes a codeword and tries to find the matching - dictionary definition. + dictionary definition. (In truth, it works with any pointer into a word, not just + the codeword pointer, and this is needed to do stack traces). In this FORTH this is not so easy. In fact we have to search through the dictionary because we don't have a convenient back-pointer (as is often the case in other versions - of FORTH). + of FORTH). Because of this search, CFA> should not be used when performance is critical, + so it is only used for debugging tools such as the decompiler and printing stack + traces. This word returns 0 if it doesn't find a match. ) @@ -980,11 +998,10 @@ BEGIN ?DUP ( while link pointer is not null ) WHILE - DUP >CFA ( cfa curr curr-cfa ) - 2 PICK ( cfa curr curr-cfa cfa ) - = IF ( found a match? ) + 2DUP SWAP ( cfa curr curr cfa ) + < IF ( current dictionary entry < cfa? ) NIP ( leave curr dictionary entry on the stack ) - EXIT ( and return from the function ) + EXIT THEN @ ( follow link pointer back ) REPEAT @@ -1095,17 +1112,285 @@ 2DROP ( restore stack ) ; +( + EXECUTION TOKENS ---------------------------------------------------------------------- + + Standard FORTH defines a concept called an 'execution token' (or 'xt') which is very + similar to a function pointer in C. We map the execution token to a codeword address. + + execution token of DOUBLE is the address of this codeword + | + V + +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+ + | LINK | 6 | D | O | U | B | L | E | 0 | DOCOL | DUP | + | EXIT | + +---------+---+---+---+---+---+---+---+---+------------+------------+------------+------------+ + len pad codeword ^ + + There is one assembler primitive for execution tokens, EXECUTE ( xt -- ), which runs them. + + You can make an execution token for an existing word the long way using >CFA, + ie: WORD [foo] FIND >CFA will push the xt for foo onto the stack where foo is the + next word in input. So a very slow way to run DOUBLE might be: + + : DOUBLE DUP + ; + : SLOW WORD FIND >CFA EXECUTE ; + 5 SLOW DOUBLE . CR \ prints 10 + + We also offer a simpler and faster way to get the execution token of any word FOO: + + ['] FOO + + (Exercises for readers: (1) What is the difference between ['] FOO and ' FOO? + (2) What is the relationship between ', ['] and LIT?) + + More useful is to define anonymous words and/or to assign xt's to variables. + + To define an anonymous word (and push its xt on the stack) use :NONAME ... ; as in this + example: + + :NONAME ." anon word was called" CR ; \ pushes xt on the stack + DUP EXECUTE EXECUTE \ executes the anon word twice + + Stack parameters work as expected: + + :NONAME ." called with parameter " . CR ; + DUP + 10 SWAP EXECUTE \ prints 'called with parameter 10' + 20 SWAP EXECUTE \ prints 'called with parameter 20' + + Notice that the above code has a memory leak: the anonymous word is still compiled + into the data segment, so even if you lose track of the xt, the word continues to + occupy memory. A good way to keep track of the xt and thus avoid the memory leak is + to assign it to a CONSTANT, VARIABLE or VALUE: + + 0 VALUE ANON + :NONAME ." anon word was called" CR ; TO ANON + ANON EXECUTE + ANON EXECUTE + + Another use of :NONAME is to create an array of functions which can be called quickly + (think: fast switch statement). This example is adapted from the ANS FORTH standard: + + 10 CELLS ALLOT CONSTANT CMD-TABLE + : SET-CMD CELLS CMD-TABLE + ! ; + : CALL-CMD CELLS CMD-TABLE + @ EXECUTE ; + + :NONAME ." alternate 0 was called" CR ; 0 SET-CMD + :NONAME ." alternate 1 was called" CR ; 1 SET-CMD + \ etc... + :NONAME ." alternate 9 was called" CR ; 9 SET-CMD + + 0 CALL-CMD + 1 CALL-CMD +) + +: :NONAME + 0 0 CREATE ( create a word with no name - we need a dictionary header because ; expects it ) + HERE @ ( current HERE value is the address of the codeword, ie. the xt ) + DOCOL , ( compile DOCOL (the codeword) ) + ] ( go into compile mode ) +; + +: ['] IMMEDIATE + ' LIT , ( compile LIT ) +; + +( + EXCEPTIONS ---------------------------------------------------------------------- + + Amazingly enough, exceptions can be implemented directly in FORTH, in fact rather easily. + + The general usage is as follows: + + : FOO ( n -- ) THROW ; + + : TEST-EXCEPTIONS + 25 ['] FOO CATCH \ execute 25 FOO, catching any exception + ?DUP IF + ." called FOO and it threw exception number: " + . CR + DROP \ we have to drop the argument of FOO (25) + THEN + ; + \ prints: called FOO and it threw exception number: 25 + + CATCH runs an execution token and detects whether it throws any exception or not. The + stack signature of CATCH is rather complicated: + + ( a_n-1 ... a_1 a_0 xt -- r_m-1 ... r_1 r_0 0 ) if xt did NOT throw an exception + ( a_n-1 ... a_1 a_0 xt -- ?_n-1 ... ?_1 ?_0 e ) if xt DID throw exception 'e' + + where a_i and r_i are the (arbitrary number of) argument and return stack contents + before and after xt is EXECUTEd. Notice in particular the case where an exception + is thrown, the stack pointer is restored so that there are n of _something_ on the + stack in the positions where the arguments a_i used to be. We don't really guarantee + what is on the stack -- perhaps the original arguments, and perhaps other nonsense -- + it largely depends on the implementation of the word that was executed. + + THROW, ABORT and a few others throw exceptions. + + Exception numbers are non-zero integers. By convention the positive numbers can be used + for app-specific exceptions and the negative numbers have certain meanings defined in + the ANS FORTH standard. (For example, -1 is the exception thrown by ABORT). + + 0 THROW does nothing. This is the stack signature of THROW: + + ( 0 -- ) + ( * e -- ?_n-1 ... ?_1 ?_0 e ) the stack is restored to the state from the corresponding CATCH + + The implementation hangs on the definitions of CATCH and THROW and the state shared + between them. + + Up to this point, the return stack has consisted merely of a list of return addresses, + with the top of the return stack being the return address where we will resume executing + when the current word EXITs. However CATCH will push a more complicated 'exception stack + frame' on the return stack. The exception stack frame records some things about the + state of execution at the time that CATCH was called. + + When called, THROW walks up the return stack (the process is called 'unwinding') until + it finds the exception stack frame. It then uses the data in the exception stack frame + to restore the state allowing execution to continue after the matching CATCH. (If it + unwinds the stack and doesn't find the exception stack frame then it prints a message + and drops back to the prompt, which is also normal behaviour for so-called 'uncaught + exceptions'). + + This is what the exception stack frame looks like. (As is conventional, the return stack + is shown growing downwards from higher to lower memory addresses). + + +------------------------------+ + | return address from CATCH | Notice this is already on the + | | return stack when CATCH is called. + +------------------------------+ + | original parameter stack | + | pointer | + +------------------------------+ ^ + | exception stack marker | | + | (EXCEPTION-MARKER) | | Direction of stack + +------------------------------+ | unwinding by THROW. + | + | + + The EXCEPTION-MARKER marks the entry as being an exception stack frame rather than an + ordinary return address, and it is this which THROW "notices" as it is unwinding the + stack. (If you want to implement more advanced exceptions such as TRY...WITH then + you'll need to use a different value of marker if you want the old and new exception stack + frame layouts to coexist). + + What happens if the executed word doesn't throw an exception? It will eventually + return and call EXCEPTION-MARKER, so EXCEPTION-MARKER had better do something sensible + without us needing to modify EXIT. This nicely gives us a suitable definition of + EXCEPTION-MARKER, namely a function that just drops the stack frame and itself + returns (thus "returning" from the original CATCH). + + One thing to take from this is that exceptions are a relatively lightweight mechanism + in FORTH. +) + +: EXCEPTION-MARKER + RDROP ( drop the original parameter stack pointer ) + 0 ( there was no exception, this is the normal return path ) +; + +: CATCH ( xt -- exn? ) + DSP@ 4+ >R ( save parameter stack pointer (+4 because of xt) on the return stack ) + ' EXCEPTION-MARKER 4+ ( push the address of the RDROP inside EXCEPTION-MARKER ... ) + >R ( ... on to the return stack so it acts like a return address ) + EXECUTE ( execute the nested function ) +; + +: THROW ( n -- ) + ?DUP IF ( only act if the exception code <> 0 ) + RSP@ ( get return stack pointer ) + BEGIN + DUP R0 4- < ( RSP < R0 ) + WHILE + DUP @ ( get the return stack entry ) + ' EXCEPTION-MARKER 4+ = IF ( found the EXCEPTION-MARKER on the return stack ) + 4+ ( skip the EXCEPTION-MARKER on the return stack ) + RSP! ( restore the return stack pointer ) + + ( Restore the parameter stack. ) + DUP DUP DUP ( reserve some working space so the stack for this word + doesn't coincide with the part of the stack being restored ) + R> ( get the saved parameter stack pointer | n dsp ) + 4- ( reserve space on the stack to store n ) + SWAP OVER ( dsp n dsp ) + ! ( write n on the stack ) + DSP! EXIT ( restore the parameter stack pointer, immediately exit ) + THEN + 4+ + REPEAT + + ( No matching catch - print a message and restart the INTERPRETer. ) + DROP + + CASE + 0 1- OF ( ABORT ) + ." ABORTED" CR + ENDOF + ( default case ) + ." UNCAUGHT THROW " + DUP . CR + ENDCASE + QUIT + THEN +; + +: ABORT ( -- ) + 0 1- THROW +; + +( Print a stack trace by walking up the return stack. ) +: PRINT-STACK-TRACE + RSP@ ( start at caller of this function ) + BEGIN + DUP R0 4- < ( RSP < R0 ) + WHILE + DUP @ ( get the return stack entry ) + CASE + ' EXCEPTION-MARKER 4+ OF ( is it the exception stack frame? ) + ." CATCH ( DSP=" + 4+ DUP @ U. ( print saved stack pointer ) + ." ) " + ENDOF + ( default case ) + DUP + CFA> ( look up the codeword to get the dictionary entry ) + ?DUP IF ( and print it ) + 2DUP ( dea addr dea ) + ID. ( print word from dictionary entry ) + [ CHAR + ] LITERAL EMIT + SWAP >DFA 4+ - . ( print offset ) + THEN + ENDCASE + 4+ ( move up the stack ) + REPEAT + DROP + CR +; + ( C STRINGS ---------------------------------------------------------------------- FORTH strings are represented by a start address and length kept on the stack or in memory. Most FORTHs don't handle C strings, but we need them in order to access the process arguments - and environment left on the stack by the Linux kernel. + and environment left on the stack by the Linux kernel, and to make some system calls. + + Operation Input Output FORTH word Notes + ---------------------------------------------------------------------- + + Create FORTH string addr len S" ..." + + Create C string c-addr Z" ..." + + C -> FORTH c-addr addr len DUP STRLEN - The main function we need is STRLEN which works out the length of a C string. DUP STRLEN is - a common idiom which 'converts' a C string into a FORTH string. (For example, DUP STRLEN TELL - prints a C string). + FORTH -> C addr len c-addr CSTRING Allocated in a temporary buffer, so + should be consumed / copied immediately. + FORTH string should not contain NULs. + + For example, DUP STRLEN TELL prints a C string. ) ( @@ -1152,7 +1437,6 @@ THEN ; -( STRLEN returns the length of a C string ) : STRLEN ( str -- len ) DUP ( save start address ) BEGIN @@ -1164,42 +1448,15 @@ SWAP - ( calculate the length ) ; -( - STRNCMP compares two strings up to a length. As with C's strncmp it returns 0 if they - are equal, or a number > 0 or < 0 indicating their order. -) -: STRNCMP ( str1 str2 len -- eq? ) - BEGIN - ?DUP - WHILE - ROT ( len str1 str2 ) - DUP C@ ( len str1 str2 char2 ) - 2 PICK C@ ( len str1 str2 char2 char1 ) - OVER ( len str1 str2 char2 char1 char2 ) - - ( len str1 str2 char2 char1-char2 ) - - ?DUP IF ( strings not the same at this position? ) - NIP ( len str1 str2 diff ) - ROT ( len diff str1 str2 ) - DROP DROP ( len diff ) - NIP ( diff ) - EXIT - THEN +: CSTRING ( addr len -- c-addr ) + SWAP OVER ( len saddr len ) + HERE @ SWAP ( len saddr daddr len ) + CMOVE ( len ) - 0= IF ( characters are equal, but is this the end of the C string? ) - DROP DROP DROP - 0 - EXIT - THEN + HERE @ + ( daddr+len ) + 0 SWAP C! ( store terminating NUL char ) - 1+ ( len str1 str2+1 ) - ROT ( str2+1 len str1 ) - 1+ ROT ( str1+1 str2+1 len ) - 1- ( str1+1 str2+1 len-1 ) - REPEAT - - 2DROP ( restore stack ) - 0 ( equal ) + HERE @ ( push start address ) ; ( @@ -1254,73 +1511,124 @@ ; ( - SYSTEM CALLS ---------------------------------------------------------------------- + SYSTEM CALLS AND FILES ---------------------------------------------------------------------- - Some wrappers around Linux system calls + Miscellaneous words related to system calls, and standard access to files. ) ( BYE exits by calling the Linux exit(2) syscall. ) : BYE ( -- ) - 0 - 0 0 ( return code (0) ) SYS_EXIT ( system call number ) - SYSCALL3 + SYSCALL1 ; ( - OPEN, CREAT and CLOSE are just like the Linux syscalls open(2), creat(2) and close(2). + UNUSED returns the number of cells remaining in the user memory (data segment). - Notice that they take C strings and may return error codes (-errno). + For our implementation we will use Linux brk(2) system call to find out the end + of the data segment and subtract HERE from it. ) -: OPEN ( mode flags c-pathname -- ret ) - SYS_OPEN - SYSCALL3 +: GET-BRK ( -- brkpoint ) + 0 SYS_BRK SYSCALL1 ( call brk(0) ) ; -: CREAT ( mode c-pathname -- ret ) - 0 ROT - SYS_CREAT - SYSCALL3 +: UNUSED ( -- n ) + GET-BRK ( get end of data segment according to the kernel ) + HERE @ ( get current position in data segment ) + - + 4 / ( returns number of cells ) ; -: CLOSE ( fd -- ret ) - 0 SWAP 0 ROT - SYS_CLOSE - SYSCALL3 +( + MORECORE increases the data segment by the specified number of (4 byte) cells. + + NB. The number of cells requested should normally be a multiple of 1024. The + reason is that Linux can't extend the data segment by less than a single page + (4096 bytes or 1024 cells). + + This FORTH doesn't automatically increase the size of the data segment "on demand" + (ie. when , (COMMA), ALLOT, CREATE, and so on are used). Instead the programmer + needs to be aware of how much space a large allocation will take, check UNUSED, and + call MORECORE if necessary. A simple programming exercise is to change the + implementation of the data segment so that MORECORE is called automatically if + the program needs more memory. +) +: BRK ( brkpoint -- ) + SYS_BRK SYSCALL1 ; -( READ and WRITE system calls. ) -: READ ( len buffer fd -- ret ) - SYS_READ - SYSCALL3 -; - -: WRITE ( len buffer fd -- ret ) - SYS_WRITE - SYSCALL3 -; +: MORECORE ( cells -- ) + CELLS GET-BRK + BRK +; ( - ANS FORTH ---------------------------------------------------------------------- + Standard FORTH provides some simple file access primitives which we model on + top of Linux syscalls. - From this point we're trying to fill in the missing parts of the ISO standard, commonly - referred to as ANS FORTH. + The main complication is converting FORTH strings (address & length) into C + strings for the Linux kernel. - http://www.taygeta.com/forth/dpans.html - http://www.taygeta.com/forth/dpansf.htm (list of words) + Notice there is no buffering in this implementation. ) -( C, writes a byte at the HERE pointer. ) -: C, HERE @ C! 1 HERE +! ; - - +: R/O ( -- fam ) O_RDONLY ; +: R/W ( -- fam ) O_RDWR ; +: OPEN-FILE ( addr u fam -- fd 0 (if successful) | c-addr u fam -- fd errno (if there was an error) ) + ROT ( fam addr u ) + CSTRING ( fam cstring ) + SYS_OPEN SYSCALL2 ( open (filename, flags) ) + DUP ( fd fd ) + DUP 0< IF ( errno? ) + NEGATE ( fd errno ) + ELSE + DROP 0 ( fd 0 ) + THEN +; +: CREATE-FILE ( addr u fam -- fd 0 (if successful) | c-addr u fam -- fd errno (if there was an error) ) + O_CREAT OR + O_TRUNC OR + ROT ( fam addr u ) + CSTRING ( fam cstring ) + 420 ROT ( 0644 fam cstring ) + SYS_OPEN SYSCALL3 ( open (filename, flags|O_TRUNC|O_CREAT, 0644) ) + DUP ( fd fd ) + DUP 0< IF ( errno? ) + NEGATE ( fd errno ) + ELSE + DROP 0 ( fd 0 ) + THEN +; +: CLOSE-FILE ( fd -- 0 (if successful) | fd -- errno (if there was an error) ) + SYS_CLOSE SYSCALL1 + NEGATE +; +: READ-FILE ( addr u fd -- u2 0 (if successful) | addr u fd -- 0 0 (if EOF) | addr u fd -- u2 errno (if error) ) + ROT SWAP -ROT ( u addr fd ) + SYS_READ SYSCALL3 + DUP ( u2 u2 ) + DUP 0< IF ( errno? ) + NEGATE ( u2 errno ) + ELSE + DROP 0 ( u2 0 ) + THEN +; +( + PERROR prints a message for an errno, similar to C's perror(3) but we don't have the extensive + list of strerror strings available, so all we can do is print the errno. +) +: PERROR ( errno addr u -- ) + TELL + ':' EMIT SPACE + ." ERRNO=" + . CR +; ( NOTES ---------------------------------------------------------------------- @@ -1335,5 +1643,13 @@ Print the version and OK prompt. ) -." JONESFORTH VERSION " VERSION . CR -." OK " +: WELCOME + S" TEST-MODE" FIND NOT IF + ." JONESFORTH VERSION " VERSION . CR + UNUSED . ." CELLS REMAINING" CR + ." OK " + THEN +; + +WELCOME +HIDE WELCOME diff --git a/test_comparison.f b/test_comparison.f new file mode 100644 index 0000000..f4908fb --- /dev/null +++ b/test_comparison.f @@ -0,0 +1,73 @@ +( -*- text -*- ) + +: TEST + 1 0 < . CR + 0 1 < . CR + 1 -1 < . CR + -1 1 < . CR + -1 0 < . CR + 0 -1 < . CR CR + + 1 0 > . CR + 0 1 > . CR + 1 -1 > . CR + -1 1 > . CR + -1 0 > . CR + 0 -1 > . CR CR + + 1 1 <= . CR + 0 0 <= . CR + -1 -1 <= . CR + 1 0 <= . CR + 0 1 <= . CR + 1 -1 <= . CR + -1 1 <= . CR + -1 0 <= . CR + 0 -1 <= . CR CR + + 1 1 >= . CR + 0 0 >= . CR + -1 -1 >= . CR + 1 0 >= . CR + 0 1 >= . CR + 1 -1 >= . CR + -1 1 >= . CR + -1 0 >= . CR + 0 -1 >= . CR CR + + 1 1 = . CR + 1 0 = . CR + 0 0 = . CR + 1 -1 = . CR + -1 -1 = . CR CR + + 1 1 <> . CR + 1 0 <> . CR + 0 0 <> . CR + 1 -1 <> . CR + -1 -1 <> . CR CR + + 1 0= . CR + 0 0= . CR + -1 0= . CR CR + + 1 0<> . CR + 0 0<> . CR + -1 0<> . CR CR + + 1 0< . CR + 0 0< . CR + -1 0< . CR CR + + 1 0> . CR + 0 0> . CR + -1 0> . CR CR + + 1 0<= . CR + 0 0<= . CR + -1 0<= . CR CR + + 1 0>= . CR + 0 0>= . CR + -1 0>= . CR CR +; diff --git a/test_comparison.f.out b/test_comparison.f.out new file mode 100644 index 0000000..40d6034 --- /dev/null +++ b/test_comparison.f.out @@ -0,0 +1,70 @@ +0 +1 +0 +1 +1 +0 + +1 +0 +1 +0 +0 +1 + +1 +1 +1 +0 +1 +0 +1 +1 +0 + +1 +1 +1 +1 +0 +1 +0 +0 +1 + +1 +0 +1 +0 +1 + +0 +1 +0 +1 +0 + +0 +1 +0 + +1 +0 +1 + +0 +0 +1 + +1 +0 +0 + +0 +1 +1 + +1 +1 +0 + diff --git a/test_exception.f b/test_exception.f new file mode 100644 index 0000000..e8f57d5 --- /dev/null +++ b/test_exception.f @@ -0,0 +1,13 @@ +( -*- text -*- ) + +: TEST4 PRINT-STACK-TRACE THROW ; + +: TEST3 0 TEST4 26 TEST4 ; + +: TEST2 + ['] TEST3 CATCH + ?DUP IF ." TEST3 threw exception " . CR THEN + TEST3 +; + +: TEST TEST2 ; diff --git a/test_exception.f.out b/test_exception.f.out new file mode 100644 index 0000000..a62679f --- /dev/null +++ b/test_exception.f.out @@ -0,0 +1,6 @@ +TEST4+0 TEST3+8 CATCH+28 CATCH ( ) TEST2+8 TEST+0 +TEST4+0 TEST3+20 CATCH+28 CATCH ( ) TEST2+8 TEST+0 +TEST3 threw exception 26 +TEST4+0 TEST3+8 TEST2+68 TEST+0 +TEST4+0 TEST3+20 TEST2+68 TEST+0 +UNCAUGHT THROW 26 diff --git a/test_number.f b/test_number.f new file mode 100644 index 0000000..846347f --- /dev/null +++ b/test_number.f @@ -0,0 +1,11 @@ +( -*- text -*- ) + +: TEST + 123 . CR + [ HEX -7F ] LITERAL DECIMAL . CR + [ HEX 7FF77FF7 ] LITERAL HEX . CR + [ HEX -7FF77FF7 ] LITERAL 2 BASE ! . CR + [ 2 BASE ! 1111111111101110111111111110111 ] LITERAL HEX . CR +; + +DECIMAL ( restore immediate-mode base ) diff --git a/test_number.f.out b/test_number.f.out new file mode 100644 index 0000000..734c750 --- /dev/null +++ b/test_number.f.out @@ -0,0 +1,5 @@ +123 +-127 +7FF77FF7 +-1111111111101110111111111110111 +7FF77FF7 diff --git a/test_read_file.f b/test_read_file.f new file mode 100644 index 0000000..f3a6f45 --- /dev/null +++ b/test_read_file.f @@ -0,0 +1,24 @@ +( -*- text -*- + Test READ-FILE. + $Id: test_read_file.f,v 1.1 2007-10-07 11:07:15 rich Exp $ +) + +0 VALUE FD +100 CELLS ALLOT CONSTANT BUFFER + +: TEST + S" /etc/fstab" R/O OPEN-FILE + ?DUP IF S" /etc/fstab" PERROR QUIT THEN + + TO FD + + BEGIN + BUFFER 100 CELLS FD READ-FILE + ?DUP IF S" READ-FILE" PERROR QUIT THEN + DUP + BUFFER SWAP TELL + 0= UNTIL + + FD CLOSE-FILE + ?DUP IF S" CLOSE-FILE" PERROR QUIT THEN +; diff --git a/test_read_file.f.out b/test_read_file.f.out new file mode 100644 index 0000000..1b95f7b --- /dev/null +++ b/test_read_file.f.out @@ -0,0 +1,7 @@ +/dev/VolGroup00/LogVol00 / ext3 defaults,noatime 1 1 +LABEL=/boot /boot ext3 defaults 1 2 +tmpfs /dev/shm tmpfs defaults 0 0 +devpts /dev/pts devpts gid=5,mode=620 0 0 +sysfs /sys sysfs defaults 0 0 +proc /proc proc defaults 0 0 +/dev/VolGroup00/LogVol01 swap swap defaults 0 0 diff --git a/test_stack_trace.f b/test_stack_trace.f new file mode 100644 index 0000000..e72f0fa --- /dev/null +++ b/test_stack_trace.f @@ -0,0 +1,9 @@ +( -*- text -*- ) + +: TEST4 PRINT-STACK-TRACE ; + +: TEST3 TEST4 1 2 + . CR TEST4 ; + +: TEST2 TEST3 TEST3 ; + +: TEST TEST2 ; diff --git a/test_stack_trace.f.out b/test_stack_trace.f.out new file mode 100644 index 0000000..3595aa2 --- /dev/null +++ b/test_stack_trace.f.out @@ -0,0 +1,6 @@ +TEST4+0 TEST3+0 TEST2+0 TEST+0 +3 +TEST4+0 TEST3+32 TEST2+0 TEST+0 +TEST4+0 TEST3+0 TEST2+4 TEST+0 +3 +TEST4+0 TEST3+32 TEST2+4 TEST+0 -- 2.39.2