module/language/assembly/decompile-bytecode.scm

   1 ;;; Guile VM code converters
   2
   3 ;; Copyright (C) 2001, 2009 Free Software Foundation, Inc.
   4
   5 ;;;; This library is free software; you can redistribute it and/or
   6 ;;;; modify it under the terms of the GNU Lesser General Public
   7 ;;;; License as published by the Free Software Foundation; either
   8 ;;;; version 3 of the License, or (at your option) any later version.
   9 ;;;;
  10 ;;;; This library is distributed in the hope that it will be useful,
  11 ;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 ;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13 ;;;; Lesser General Public License for more details.
  14 ;;;;
  15 ;;;; You should have received a copy of the GNU Lesser General Public
  16 ;;;; License along with this library; if not, write to the Free Software
  17 ;;;; Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  18
  19 ;;; Code:
  20
  21 (define-module (language assembly decompile-bytecode)
  22   #:use-module (system vm instruction)
  23   #:use-module (system base pmatch)
  24   #:use-module (srfi srfi-4)
  25   #:use-module (rnrs bytevector)
  26   #:use-module (language assembly)
  27   #:export (decompile-bytecode))
  28
  29 (define (decompile-bytecode x env opts)
  30   (let ((i 0) (size (u8vector-length x)))
  31     (define (pop)
  32       (let ((b (cond ((< i size) (u8vector-ref x i))
  33                      ((= i size) #f)
  34                      (else (error "tried to decode too many bytes")))))
  35         (if b (set! i (1+ i)))
  36         b))
  37     (let ((ret (decode-load-program pop)))
  38       (if (= i size)
  39           (values ret env)
  40           (error "bad bytecode: only decoded ~a out of ~a bytes" i size)))))
  41
  42 (define (br-instruction? x)
  43   (memq x '(br br-if br-if-not br-if-eq br-if-not-eq br-if-null br-if-not-null)))
  44
  45 (define (bytes->s16 a b)
  46   (let ((x (+ (ash a 8) b)))
  47     (if (zero? (logand (ash 1 15) x))
  48         x
  49         (- x (ash 1 16)))))
  50
  51 ;; FIXME: this is a little-endian disassembly!!!
  52 (define (decode-load-program pop)
  53   (let* ((nargs (pop)) (nrest (pop)) (nlocs0 (pop)) (nlocs1 (pop))
  54          (nlocs (+ nlocs0 (ash nlocs1 8)))
  55          (a (pop)) (b (pop)) (c (pop)) (d (pop))
  56          (e (pop)) (f (pop)) (g (pop)) (h (pop))
  57          (len (+ a (ash b 8) (ash c 16) (ash d 24)))
  58          (metalen (+ e (ash f 8) (ash g 16) (ash h 24)))
  59          (totlen (+ len metalen))
  60          (pad0 (pop)) (pad1 (pop)) (pad2 (pop)) (pad3 (pop))
  61          (labels '())
  62          (i 0))
  63     (define (ensure-label rel1 rel2)
  64       (let ((where (+ i (bytes->s16 rel1 rel2))))
  65         (or (assv-ref labels where)
  66             (begin
  67               (let ((l (gensym ":L")))
  68                 (set! labels (acons where l labels))
  69                 l)))))
  70     (define (sub-pop) ;; ...records. ha. ha.
  71       (let ((b (cond ((< i len) (pop))
  72                      ((= i len) #f)
  73                      (else (error "tried to decode too many bytes")))))
  74         (if b (set! i (1+ i)))
  75         b))
  76     (let lp ((out '()))
  77       (cond ((> i len)
  78              (error "error decoding program -- read too many bytes" out))
  79             ((= i len)
  80              `(load-program ,nargs ,nrest ,nlocs
  81                             ,(map (lambda (x) (cons (cdr x) (car x)))
  82                                   (reverse labels))
  83                             ,len
  84                             ,(if (zero? metalen) #f (decode-load-program pop))
  85                             ,@(reverse! out)))
  86             (else
  87              (let ((exp (decode-bytecode sub-pop)))
  88                (pmatch exp
  89                  ((,br ,rel1 ,rel2) (guard (br-instruction? br))
  90                   (lp (cons `(,br ,(ensure-label rel1 rel2)) out)))
  91                  ((mv-call ,n ,rel1 ,rel2)
  92                   (lp (cons `(mv-call ,n ,(ensure-label rel1 rel2)) out)))
  93                  (else
  94                   (lp (cons exp out))))))))))
  95
  96 (define (decode-bytecode pop)
  97   (and=> (pop)
  98          (lambda (opcode)
  99            (let ((inst (opcode->instruction opcode)))
 100              (cond
 101               ((eq? inst 'load-program)
 102                (decode-load-program pop))
 103               ((< (instruction-length inst) 0)
 104                (let* ((make-sequence
 105                        (if (eq? inst 'load-array)
 106                            make-bytevector
 107                            make-string))
 108                       (sequence-set!
 109                        (if (eq? inst 'load-array)
 110                            bytevector-u8-set!
 111                            (lambda (str pos value)
 112                              (string-set! str pos (integer->char value)))))
 113
 114                       (len (let* ((a (pop)) (b (pop)) (c (pop)))
 115                              (+ (ash a 16) (ash b 8) c)))
 116                       (seq (make-sequence len)))
 117                  (let lp ((i 0))
 118                    (if (= i len)
 119                        `(,inst ,seq)
 120                        (begin
 121                          (sequence-set! seq i (pop))
 122                          (lp (1+ i)))))))
 123               (else
 124                ;; fixed length
 125                (let lp ((n (instruction-length inst)) (out (list inst)))
 126                  (if (zero? n)
 127                      (reverse! out)
 128                      (lp (1- n) (cons (pop) out))))))))))