-format ELF64 executable
+;; vim: syntax=fasm
+
+;; At compile-time we load the module given by the environment variable
+;; OS_INCLUDE. All of the following these procedures should preserve the value
+;; of RSI and RSP. They may use other registers as they like.
+;;
+;; The module should provide the following:
+;;
+;; os_code_section
+;; Macro to start the text segment.
+;;
+;; os_data_section
+;; Macro to start the data segment.
+;;
+;; os_initialize
+;; Called at initialization.
+;;
+;; os_print_string
+;; Takes a string buffer in RCX and the length in RDX, and prints the string
+;; to the console.
+;;
+;; os_read_char
+;; Wait for the user to type a key, and then put the corresponding ASCII byte
+;; into RAX.
+;;
+;; os_terminate
+;; Shut down the system, returning the error code given in RAX.
+include '%OS_INCLUDE%'
;; The code in this macro is placed at the end of each Forth word. When we are
;; executing a definition, this code is what causes execution to resume at the
add rbp, 8
}
-segment readable executable
+;; The following macro generates the dictionary header. It updates the
+;; initial_latest_entry variable, which is used as the initial value of the
+;; latest_entry variable that is made available at runtime.
+;;
+;; The header contains a link to the previous entry, the length of the name of
+;; the word and the word itself as a string literal.
+;;
+;; This macro also defines a label LABEL_entry.
+initial_latest_entry = 0
+macro header label, name, immediate {
+ local .string_end
+
+label#_entry:
+ dq initial_latest_entry
+ if immediate eq
+ db 0
+ else
+ db 1
+ end if
+ db .string_end - ($ + 1)
+ db name
+ .string_end:
+label:
+
+initial_latest_entry = label#_entry
+}
+
+;; Define a Forth word that is implemented in assembly. See 'header' for details.
+macro forth_asm label, name, immediate {
+ header label, name, immediate
+ dq .start
+.start:
+}
+
+include "impl.asm" ; Misc. subroutines
+include "bootstrap.asm" ; Forth words encoded in Assembly
+
+os_code_section
main:
cld ; Clear direction flag so LODSQ does the right thing.
mov rbp, return_stack_top ; Initialize return stack
- mov rsi, program
- next
+ call os_initialize
+
+ mov rax, MAIN
+ jmp qword [rax]
program: dq MAIN
;; The codeword is the code that will be executed at the beginning of a forth
;; word. It needs to save the old RSI and update it to point to the next word to
;; execute.
-docol:
+header DOCOL, 'DOCOL'
pushr rsi ; Save old value of RSI on return stack; we will continue execution there after we are done executing this word
lea rsi, [rax + 8] ; RAX currently points to the address of the codeword, so we want to continue at RAX+8
next ; Execute word pointed to by RSI
;; This word is called at the end of a Forth definition. It just needs to
-;; restore the old value of RSI (saved by 'docol') and resume execution.
-EXIT_entry:
- dq 0
- db 4
- db 'EXIT'
-EXIT:
- dq .start
-.start:
+;; restore the old value of RSI (saved by 'DOCOL') and resume execution.
+forth_asm EXIT, 'EXIT'
popr rsi
next
;; LIT is a special word that reads the next "word pointer" and causes it to be
;; placed on the stack rather than executed.
-LIT_entry:
- dq EXIT_entry
- db 3
- db 'LIT'
-LIT:
- dq .start
-.start:
+forth_asm LIT, 'LIT'
lodsq
push rax
next
+;; When LITSTRING is encountered while executing a word, it instead reads a
+;; string from the definition of that word, and places that string on the stack
+;; as (buffer, length).
+forth_asm LITSTRING, 'LITSTRING'
+ lodsb
+ push rsi ; Buffer
+ movzx rax, al
+ push rax ; Length
+ add rsi, rax ; Skip over string before resuming execution
+ next
+
;; Given a string (a pointer following by a size), return the location of the
;; dictionary entry for that word. If no such word exists, return 0.
-FIND_entry:
- dq LIT_entry
- db 4
- db 'FIND'
-FIND:
- dq .start
-.start:
+forth_asm FIND, 'FIND'
mov [.rsi], rsi
- pop [.search_length]
- pop [.search_buffer]
- ;; RSI contains the entry we are currently looking at
+ pop [find.search_length]
+ pop [find.search_buffer]
mov rsi, [latest_entry] ; Start with the last added word
+ call find
+ push rsi
-.loop:
- movzx rcx, byte [rsi + 8] ; Length of word being looked at
- cmp rcx, [.search_length]
- jne .next ; If the words don't have the same length, we have the wrong word
-
- ;; Otherwise, we need to compare strings
- lea rdx, [rsi + 8 + 1] ; Location of character being compared in entry
- mov rdi, [.search_buffer] ; Location of character being compared in search buffer
-.compare_char:
- mov al, [rdx]
- mov ah, [rdi]
- cmp al, ah
- jne .next ; They don't match; try again
- inc rdx ; These characters match; look at the next ones
- inc rdi
- loop .compare_char
-
- jmp .found ; They match! We are done.
-
-.next:
- mov rsi, [rsi] ; Look at the previous entry
- cmp rsi, 0
- jnz .loop ; If there is no previous word, exit and return 0
-
-.found:
+ mov rsi, [.rsi]
+ next
push rsi
mov rsi, [.rsi]
next
+;; Given an entry in the dictionary, return a pointer to the codeword of that
+;; entry.
+forth_asm TCFA, '>CFA'
+ pop rax
+ add rax, 8 + 1 ; [rax] = length of name
+ movzx rbx, byte [rax]
+ inc rax
+ add rax, rbx ; [rax] = codeword
+ push rax
+ next
+
;; BRANCH is the fundamental mechanism for branching. BRANCH reads the next word
;; as a signed integer literal and jumps by that offset.
-BRANCH_entry:
- dq FIND_entry
- db 6
- db 'BRANCH'
-BRANCH:
- dq .start
-.start:
+forth_asm BRANCH, 'BRANCH'
add rsi, [rsi] ; [RSI], which is the next word, contains the offset; we add this to the instruction pointer.
next ; Then, we can just continue execution as normal
;; 0BRANCH is like BRANCH, but it jumps only if the top of the stack is zero.
-ZBRANCH:
- dq .start
-.start:
+forth_asm ZBRANCH, '0BRANCH'
;; Compare top of stack to see if we should branch
pop rax
cmp rax, 0
add rsi, 8 ; We need to skip over the next word, which contains the offset.
next
+;; Duplicate the top of the stack.
+forth_asm DUP_, 'DUP'
+ push qword [rsp]
+ next
+
+;; Execute the codeword at the given address.
+forth_asm EXEC, 'EXEC'
+ pop rax
+ jmp qword [rax]
+
;; Expects a character on the stack and prints it to standard output.
-EMIT:
- dq .start
-.start:
+forth_asm EMIT, 'EMIT'
pushr rsi
pushr rax
- mov rax, 1
- mov rdi, 1
- lea rsi, [rsp]
+
+ lea rcx, [rsp]
mov rdx, 1
- syscall
+ call os_print_string
+
add rsp, 8
popr rax
popr rsi
next
-;; Prints a newline to standard output.
-NEWLINE:
- dq docol
- dq LIT, $A
- dq EMIT
- dq EXIT
+;; Read a single character from the current input stream. Usually, this will wait
+;; for the user to press a key, and then return the corresponding character. When
+;; reading from a special buffer, it will instead return the next characater from
+;; that buffer.
+;;
+;; The ASCII character code is placed on the stack.
+forth_asm KEY, 'KEY'
+ call .impl
+ push rax
+ next
-;; Prints a space to standard output.
-SPACE_entry:
- dq BRANCH_entry
- db 5
- db 'SPACE'
-SPACE:
- dq docol
- dq LIT, ' '
- dq EMIT
- dq EXIT
+;; Result in RAX
+.impl:
+ ;; Are we reading from user input or from the input buffer?
+ cmp [input_buffer], 0
+ jne .from_buffer
-;; Read a word from standard input and push it onto the stack as a pointer and a
-;; size. The pointer is valid until the next call to READ_WORD.
-READ_WORD:
- dq .start
-.start:
- mov [.rsi], rsi
- mov [.rax], rax
+ ;; Reading user input
+ call os_read_char
+ ret
-.skip_whitespace:
- ;; Read characters into .char_buffer until one of them is not whitespace.
- mov rax, 0
- mov rdi, 0
- mov rsi, .char_buffer
- mov rdx, 1
- syscall
+.from_buffer:
+ ;; Reading from buffer
+ mov rax, [input_buffer]
+ movzx rax, byte [rax]
+
+ inc [input_buffer]
+ dec [input_buffer_length]
+ ret
- cmp [.char_buffer], ' '
+;; Read a word and push it onto the stack as a pointer and a size. The pointer
+;; is valid until the next call to READ_WORD.
+forth_asm READ_WORD, 'READ-WORD'
+ push rsi
+.skip_whitespace:
+ ;; Read characters until one of them is not whitespace.
+ call KEY.impl
+ ;; We consider newlines and spaces to be whitespace.
+ cmp al, ' '
je .skip_whitespace
- cmp [.char_buffer], $A
+ cmp al, $A
je .skip_whitespace
-.alpha:
;; We got a character that wasn't whitespace. Now read the actual word.
mov [.length], 0
.read_alpha:
- mov al, [.char_buffer]
movzx rbx, [.length]
mov rsi, .buffer
add rsi, rbx
mov [rsi], al
inc [.length]
- mov rax, 0
- mov rdi, 0
- mov rsi, .char_buffer
- mov rdx, 1
- syscall
+ call KEY.impl
- cmp [.char_buffer], ' '
+ cmp al, ' '
je .end
- cmp [.char_buffer], $A
+ cmp al, $A
jne .read_alpha
.end:
+ pop rsi
push .buffer
movzx rax, [.length]
push rax
- mov rsi, [.rsi]
- mov rax, [.rax]
-
next
;; Takes a string on the stack and replaces it with the decimal number that the
;; string represents.
-PARSE_NUMBER:
- dq .start
-.start:
- pop [.length] ; Length
- pop rdi ; String pointer
- mov r8, 0 ; Result
-
- ;; Add (10^(rcx-1) * parse_char(rdi[length - rcx])) to the accumulated value
- ;; for each rcx.
- mov rcx, [.length]
-.loop:
- ;; First, calcuate 10^(rcx - 1)
- mov rax, 1
-
- mov r9, rcx
- .exp_loop:
- dec r9
- jz .break
- mov rbx, 10
- mul rbx
- jmp .exp_loop
- .break:
-
- ;; Now, rax = 10^(rcx - 1).
-
- ;; We need to calulate the value of the character at rdi[length - rcx].
- mov rbx, rdi
- add rbx, [.length]
- sub rbx, rcx
- movzx rbx, byte [rbx]
- sub rbx, '0'
+forth_asm PARSE_NUMBER, 'PARSE-NUMBER'
+ pop rcx ; Length
+ pop rdi ; String pointer
- ;; Multiply this value by rax to get (10^(rcx-1) * parse_char(rdi[length - rcx])),
- ;; then add this to the result.
- mul rbx
-
- ;; Add that value to r8
- add r8, rax
-
- dec rcx
- jnz .loop
-
- push r8
+ push rsi
+ call parse_number
+ pop rsi
+ push rax ; Result
next
-READ_NUMBER:
- dq docol
- dq READ_WORD
- dq PARSE_NUMBER
- dq EXIT
-
;; Takes a string (in the form of a pointer and a length on the stack) and
;; prints it to standard output.
-TELL:
- dq .start
-.start:
- mov rbx, rsi
- mov rcx, rax
+forth_asm TELL, 'TELL'
+ pushr rax
+ pushr rsi
- mov rax, 1
- mov rdi, 1
- pop rdx ; Length
- pop rsi ; Buffer
- syscall
+ pop rdx ; Length
+ pop rcx ; Buffer
+ call os_print_string
- mov rax, rcx
- mov rsi, rbx
+ popr rsi
+ popr rax
next
;; Exit the program cleanly.
-TERMINATE:
- dq .start
-.start:
- mov rax, $3C
- mov rdi, 0
- syscall
-
-PUSH_HELLO_CHARS:
- dq docol
- dq LIT, $A
- dq LIT, 'o'
- dq LIT, 'l'
- dq LIT, 'l'
- dq LIT, 'e'
- dq LIT, 'H'
- dq EXIT
+forth_asm TERMINATE, 'TERMINATE'
+ mov rax, 0
+ call os_terminate
-PUSH_YOU_TYPED:
- dq .start
-.start:
- push you_typed_string
- push you_typed_string.length
- next
-
-HELLO_entry:
- dq SPACE_entry
- db 5
- db 'HELLO'
-HELLO:
- dq docol
- dq LIT, 'H', EMIT
- dq LIT, 'e', EMIT
- dq LIT, 'l', EMIT
- dq LIT, 'l', EMIT
- dq LIT, 'o', EMIT
- dq LIT, '!', EMIT
- dq NEWLINE
- dq EXIT
+;; Duplicate a pair of elements.
+forth_asm PAIRDUP, '2DUP'
+ pop rbx
+ pop rax
+ push rax
+ push rbx
+ push rax
+ push rbx
+ next
+
+;; Swap the top two elements on the stack.
+forth_asm SWAP, 'SWAP'
+ pop rax
+ pop rbx
+ push rax
+ push rbx
+ next
+
+;; Remove the top element from the stack.
+forth_asm DROP, 'DROP'
+ add rsp, 8
+ next
+
+forth_asm NOT_, 'NOT'
+ pop rax
+ cmp rax, 0
+ jz .false
+.true:
+ push 0
+ next
+.false:
+ push 1
+ next
;; .U prints the value on the stack as an unsigned integer in hexadecimal.
-DOTU_entry:
- dq HELLO_entry
- db 2
- db '.U'
-DOTU:
- dq .start
-.start:
+forth_asm DOTU, '.U'
mov [.length], 0
mov [.printed_length], 1
pop rax ; RAX = value to print
loop .flip
;; Print the buffer
- mov rax, 1
- mov rdi, 1
- mov rsi, .buffer
+ mov rcx, .buffer
mov rdx, [.printed_length]
- syscall
+ call os_print_string
;; Restore RSI and continue execution
pop rsi
next
-MAIN:
- dq docol
- dq HELLO
- dq LIT, SPACE_entry, DOTU, NEWLINE
- dq LIT, HELLO_entry, DOTU, NEWLINE
- dq LIT, DOTU_entry, DOTU, NEWLINE
- dq LIT, SPACE_string, LIT, SPACE_string.length, TELL, SPACE
- dq LIT, SPACE_string, LIT, SPACE_string.length, FIND, DOTU, NEWLINE
- dq LIT, HELLO_string, LIT, HELLO_string.length, TELL, SPACE
- dq LIT, HELLO_string, LIT, HELLO_string.length, FIND, DOTU, NEWLINE
- dq LIT, DOTU_string, LIT, DOTU_string.length, TELL, SPACE
- dq LIT, DOTU_string, LIT, DOTU_string.length, FIND, DOTU, NEWLINE
- dq LIT, HELLA_string, LIT, HELLA_string.length, TELL, SPACE
- dq LIT, HELLA_string, LIT, HELLA_string.length, FIND, DOTU, NEWLINE
+;; Takes a value and an address, and stores the value at the given address.
+forth_asm PUT, '!'
+ pop rbx ; Address
+ pop rax ; Value
+ mov [rbx], rax
+ next
+
+;; Takes an address and returns the value at the given address.
+forth_asm GET, '@'
+ pop rax
+ mov rax, [rax]
+ push rax
+ next
+
+forth_asm PUT_BYTE, 'C!'
+ pop rbx
+ pop rax ; Value
+ mov [rbx], al
+ next
+
+forth_asm GET_BYTE, 'C@'
+ pop rax
+ movzx rax, byte [rax]
+ push rax
+ next
+
+;; Add two integers on the stack.
+forth_asm PLUS, '+'
+ pop rax
+ pop rbx
+ add rax, rbx
+ push rax
+ next
+
+;; Calculate difference between two integers on the stack. The second number is
+;; subtracted from the first.
+forth_asm MINUS, '-'
+ pop rax
+ pop rbx
+ sub rbx, rax
+ push rbx
+ next
+
+;; Given two integers a and b on the stack, pushes the quotient and remainder of
+;; division of a by b.
+forth_asm TIMESMOD, '/MOD'
+ pop rbx ; b
+ pop rax ; a
+ mov rdx, 0
+ div rbx
+ push rax ; a / b
+ push rdx ; a % b
+ next
+
+;; Read input until next " character is found. Push a string containing the
+;; input on the stack as (buffer length). Note that the buffer is only valid
+;; until the next call to S" and that no more than 255 characters can be read.
+forth_asm READ_STRING, 'S"'
+ ;; If the input buffer is set, we should read from there instead.
+ cmp [input_buffer], 0
+ jne read_string_buffer
+
+ push rsi
+
+ mov [.length], 0
+
+.read_char:
+ call os_read_char
+ cmp al, '"'
+ je .done
+
+ mov rdx, .buffer
+ add rdx, [.length]
+ mov [rdx], al
+ inc [.length]
+ jmp .read_char
+
+.done:
+ pop rsi
+
+ push .buffer
+ push [.length]
+
+ next
+
+read_string_buffer:
+ push rsi
+
+ ;; We borrow READ_STRING's buffer. They won't mind.
+ mov [READ_STRING.length], 0
+
+.read_char:
+ mov rbx, [input_buffer]
+ mov al, [rbx]
+ cmp al, '"'
+ je .done
+
+ mov rdx, READ_STRING.buffer
+ add rdx, [READ_STRING.length]
+ mov [rdx], al
+ inc [READ_STRING.length]
+
+ inc [input_buffer]
+ dec [input_buffer_length]
+
+ jmp .read_char
+
+.done:
+ pop rsi
+
+ ;; Skip closing "
+ inc [input_buffer]
+ dec [input_buffer_length]
+
+ push READ_STRING.buffer
+ push [READ_STRING.length]
+
+ next
+
+;; CREATE inserts a new header in the dictionary, and updates LATEST so that it
+;; points to the header. To compile a word, the user can then call ',' to
+;; continue to append data after the header.
+;;
+;; It takes the name of the word as a string (address length) on the stack.
+forth_asm CREATE, 'CREATE'
+ pop rcx ; Word string length
+ pop rdx ; Word string pointer
+
+ mov rdi, [here] ; rdi = Address at which to insert this entry
+ mov rax, [latest_entry] ; rax = Address of the previous entry
+ mov [rdi], rax ; Insert link to previous entry
+ mov [latest_entry], rdi ; Update LATEST to point to this word
+
+ add rdi, 8
+ mov [rdi], byte 0 ; Insert immediate flag
+
+ add rdi, 1
+ mov [rdi], byte cl ; Insert length
+
+ ;; Insert word string
+ add rdi, 1
+
+ push rsi
+ mov rsi, rdx ; rsi = Word string pointer
+ rep movsb
+ pop rsi
+
+ ;; Update HERE
+ mov [here], rdi
+
+ next
+
+forth_asm TICK, "'"
+ lodsq
+ push rax
+ next
+
+forth_asm ROT, 'ROT'
+ pop rax
+ pop rbx
+ pop rdx
+ push rax
+ push rdx
+ push rbx
+ next
+
+forth_asm PICK, 'PICK'
+ pop rax
+ lea rax, [rsp + 8 * rax]
+ mov rax, [rax]
+ push rax
+ next
+
+forth_asm EQL, '='
+ pop rax
+ pop rbx
+ cmp rax, rbx
+ je .eq
+.noteq:
+ push 0
+ next
+.eq:
+ push 1
+ next
+
+forth MAIN, 'MAIN'
+ dq SYSCODE
+ dq INTERPRET_STRING
+ dq INTERPRET
+ dq BRANCH, -8 * 2
dq TERMINATE
-segment readable writable
+;; EFI:
+
+forth EFI_SYSTEM_TABLE_CONSTANT, 'SystemTable'
+ dq LIT, system_table, GET
+ dq EXIT
+
+forth_asm EFICALL2, 'EFICALL2'
+ pop rax ; function pointer
+ pop rdx ; 2nd argument
+ pop rcx ; 1st argument
+
+ sub rsp, 32
+ call rax
+ add rsp, 32
+
+ next
+
+forth_asm EFICALL3, 'EFICALL3'
+ pop rax ; function pointer
+ pop r8 ; 3rd argument
+ pop rdx ; 2nd argument
+ pop rcx ; 1st argument
+
+ sub rsp, 32
+ call rax
+ add rsp, 32
+
+ push rax
+
+ next
+
+forth_asm EFICALL10, 'EFICALL10'
+ pop rax ; function pointer
+
+ mov rcx, [rsp + 8 * 9]
+ mov rdx, [rsp + 8 * 8]
+ mov r8, [rsp + 8 * 7]
+ mov r9, [rsp + 8 * 6]
+
+ ;; Reverse order of stack arguments
+ mov r10, [rsp + 8 * 5]
+ mov r11, [rsp + 8 * 0]
+ mov [rsp + 8 * 5], r11
+ mov [rsp + 8 * 0], r10
-latest_entry dq DOTU_entry
+ mov r10, [rsp + 8 * 4]
+ mov r11, [rsp + 8 * 1]
+ mov [rsp + 8 * 4], r11
+ mov [rsp + 8 * 1], r10
-SPACE_string db 'SPACE'
-.length = $ - SPACE_string
-HELLO_string db 'HELLO'
-.length = $ - HELLO_string
-DOTU_string db '.U'
-.length = $ - DOTU_string
-HELLA_string db 'HELLA'
-.length = $ - HELLA_string
+ mov r10, [rsp + 8 * 3]
+ mov r11, [rsp + 8 * 2]
+ mov [rsp + 8 * 3], r11
+ mov [rsp + 8 * 2], r10
+ sub rsp, 32
+ call rax
+ add rsp, 32 + 8 * 10
-you_typed_string db 'You typed: '
-.length = $ - you_typed_string
+ push rax
+
+ next
+
+;; Built-in variables:
+
+forth STATE, 'STATE'
+ dq LIT, var_STATE
+ dq EXIT
+
+forth LATEST, 'LATEST'
+ dq LIT, latest_entry
+ dq EXIT
+
+forth HERE, 'HERE'
+ dq LIT, here
+ dq EXIT
+
+forth SYSCODE, 'SYSCODE'
+ dq LIT, sysf
+ dq LIT, sysf.len
+ dq EXIT
+
+forth INPUT_BUFFER, 'INPUT-BUFFER'
+ dq LIT, input_buffer
+ dq EXIT
+
+forth INPUT_LENGTH, 'INPUT-LENGTH'
+ dq LIT, input_buffer_length
+ dq EXIT
+
+os_data_section
+
+;; The LATEST variable holds a pointer to the word that was last added to the
+;; dictionary. This pointer is updated as new words are added, and its value is
+;; used by FIND to look up words.
+latest_entry dq initial_latest_entry
+
+;; The STATE variable is 0 when the interpreter is executing, and non-zero when
+;; it is compiling.
+var_STATE dq 0
+
+;; The interpreter can read either from standard input or from a buffer. When
+;; input-buffer is set (non-null), words like READ-WORD and S" will use this
+;; buffer instead of reading user input.
+input_buffer dq 0
+input_buffer_length dq 0
-FIND.search_length dq ?
-FIND.search_buffer dq ?
FIND.rsi dq ?
READ_WORD.rsi dq ?
-READ_WORD.rax dq ?
-READ_WORD.max_size = $FF
-READ_WORD.buffer rb READ_WORD.max_size
-READ_WORD.length db ?
-READ_WORD.char_buffer db ?
+READ_WORD.rbp dq ?
+
+READ_STRING.char_buffer db ?
+READ_STRING.buffer rb $FF
+READ_STRING.length dq ?
DOTU.chars db '0123456789ABCDEF'
DOTU.buffer rq 16 ; 64-bit number has no more than 16 digits in hex
DOTU.length dq ?
DOTU.printed_length dq ?
-PARSE_NUMBER.length dq ?
+KEY.buffer dq ?
+
+READ_WORD.buffer rb $FF
+READ_WORD.length db ?
+
+;; Reserve space for compiled words, accessed through HERE.
+here dq here_top
+here_top rq $4000
;; Return stack
rq $2000
return_stack_top:
+
+;; We store some Forth code in sys.f that defined common words that the user
+;; would expect to have available at startup. To execute these words, we just
+;; include the file directly in the binary, and then interpret it at startup.
+sysf:
+file 'sys.f'
+file 'uefi.f'
+sysf.len = $ - sysf
+