/
parse-bytecode-5.1.1.rkt
699 lines (558 loc) · 21.3 KB
/
parse-bytecode-5.1.1.rkt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
#lang racket/base
;; Parsing Racket 5.1.1 bytecode structures into our own structures.
(require "typed-module-path.rkt"
"lam-entry-gensym.rkt"
"path-rewriter.rkt"
"../compiler/expression-structs.rkt"
"../compiler/lexical-structs.rkt"
"../parameters.rkt"
"../get-module-bytecode.rkt"
syntax/modresolve
compiler/zo-parse
racket/path
racket/match
racket/list)
(provide parse-bytecode
reset-lam-label-counter!/unit-testing)
;; current-module-path-index-resolver: (module-path-index (U Path #f) -> (U Symbol Path)) -> void
;; The module path index resolver figures out how to translate module path indices to module names.
(define current-module-path-index-resolver
(make-parameter
(lambda (mpi relative-to)
(cond
[(eq? mpi #f)
(current-module-path)]
[(self-module-path-index? mpi)
(current-module-path)]
[else
(resolve-module-path-index mpi relative-to)]))))
(define current-module-path-resolver
(make-parameter
(lambda (module-path relative-to)
(resolve-module-path module-path relative-to))))
(define (self-module-path-index? mpi)
(let-values ([(x y) (module-path-index-split mpi)])
(and (eq? x #f)
(eq? y #f))))
(define (explode-module-path-index mpi)
(let-values ([(x y) (module-path-index-split mpi)])
(cond
[(module-path-index? y)
(cons x (explode-module-path-index y))]
[else
(list x y)])))
;; seen-closures: (hashof symbol -> symbol)
;; As we're parsing, we watch for closure cycles. On any subsequent time where
;; we see a closure cycle, we break the cycle by generating an EmptyClosureReference.
;; The map is from the gen-id to the entry-point label of the lambda.
(define seen-closures (make-parameter (make-hasheq)))
;; Code is copied-and-pasted from compiler/decompile. Maps the primval ids to their respective
;; symbolic names.
(define primitive-table
;; Figure out number-to-id mapping for kernel functions in `primitive'
(let ([bindings
(let ([ns (make-base-empty-namespace)])
(parameterize ([current-namespace ns])
(namespace-require ''#%kernel)
(namespace-require ''#%unsafe)
(namespace-require ''#%flfxnum)
(for/list ([l (namespace-mapped-symbols)])
(cons l (with-handlers ([exn:fail? (lambda (x)
#f)])
(compile l))))))]
[table (make-hash)])
(for ([b (in-list bindings)])
(let ([v (and (cdr b)
(zo-parse (let ([out (open-output-bytes)])
(write (cdr b) out)
(close-output-port out)
(open-input-bytes (get-output-bytes out)))))])
(let ([n (match v
[(struct compilation-top (_ prefix (struct primval (n)))) n]
[else #f])])
(hash-set! table n (car b)))))
table))
;; parse-bytecode: (U Input-Port Path) -> Expression
;;
;; Given an input port, assumes the input is the byte representation of compiled-code.
;;
;; Given a path, assumes the path is for a module. It gets the module bytecode, and parses
;; that.
;;
;; TODO: this may be doing too much work. It doesn't quite feel like the right elements
;; are being manipulated here.
(define (parse-bytecode in)
(cond
[(input-port? in)
(parameterize ([seen-closures (make-hasheq)])
(let ([compilation-top (zo-parse in)])
(parse-top compilation-top)))]
[(compiled-expression? in)
(let ([op (open-output-bytes)])
(write in op)
(parse-bytecode (open-input-bytes (get-output-bytes op))))]
[(path? in)
(let*-values ([(normal-path) (normalize-path in)]
[(base file-path dir?) (split-path normal-path)])
(parameterize ([current-module-path normal-path]
[current-directory (cond [(path? base)
base]
[else
(error 'parse-bytecode)])])
(parse-bytecode
(open-input-bytes (get-module-bytecode normal-path)))))]
[else
(error 'parse-bytecode "Don't know how to parse from ~e" in)]))
(define (parse-top a-top)
(match a-top
[(struct compilation-top (max-let-depth prefix code))
(maybe-fix-module-name
(make-Top (parse-prefix prefix)
(parse-top-code code)))]))
;; maybe-fix-module-name: expression -> expression
;; When we're compiling a module directly from memory, it doesn't have a file path.
;; We rewrite the ModuleLocator to its given name.
(define (maybe-fix-module-name exp)
(match exp
[(struct Top (top-prefix
(struct Module ((and name (? symbol?))
(struct ModuleLocator ('self 'self))
module-prefix
module-requires
module-code))))
(make-Top top-prefix
(make-Module name
(make-ModuleLocator name name) (current-module-path)
module-prefix
module-requires
module-code))]
[else
exp]))
(define (parse-prefix a-prefix)
(match a-prefix
[(struct prefix (num-lifts toplevels stxs))
(make-Prefix
(append (map parse-prefix-toplevel toplevels)
(map (lambda (x) #f) stxs)
(if (empty? stxs) empty (list #f))
(build-list num-lifts (lambda (i) #f))))]))
;; parse-top-code: (U form Any -> Expression)
(define (parse-top-code code)
(cond
[(form? code)
(parse-form code)]
[else
(make-Constant code)]))
;; parse-prefix-toplevel: (U #f symbol global-bucket module-variable) -> (U False Symbol GlobalBucket ModuleVariable)
(define (parse-prefix-toplevel a-toplevel)
(cond
[(eq? a-toplevel #f)
#f]
[(symbol? a-toplevel)
a-toplevel]
[(global-bucket? a-toplevel)
(make-GlobalBucket (global-bucket-name a-toplevel))]
[(module-variable? a-toplevel)
(let ([resolver (current-module-path-index-resolver)])
(make-ModuleVariable (module-variable-sym a-toplevel)
(let ([resolved-path-name
(resolver (module-variable-modidx a-toplevel) (current-module-path))])
(wrap-module-name resolved-path-name))))]))
(define (wrap-module-name resolved-path-name)
(cond
[(symbol? resolved-path-name)
(make-ModuleLocator resolved-path-name resolved-path-name)]
[(path? resolved-path-name)
(let ([rewritten-path (rewrite-path resolved-path-name)])
(cond
[(symbol? rewritten-path)
(make-ModuleLocator (rewrite-path resolved-path-name)
(normalize-path resolved-path-name))]
[else
(error 'wrap-module-name "Unable to resolve module path ~s."
resolved-path-name)]))]))
;; parse-form: form -> (U Expression)
(define (parse-form a-form)
(cond
[(def-values? a-form)
(parse-def-values a-form)]
[(def-syntaxes? a-form)
(parse-def-syntaxes a-form)]
[(req? a-form)
(parse-req a-form)]
[(seq? a-form)
(parse-seq a-form)]
[(splice? a-form)
(parse-splice a-form)]
[(mod? a-form)
(parse-mod a-form)]
[(expr? a-form)
(parse-expr a-form)]
[else
(error 'parse-form "~s" a-form)]))
;; parse-def-values: def-values -> Expression
(define (parse-def-values form)
(match form
[(struct def-values (ids rhs))
(make-DefValues (map parse-toplevel ids)
(parse-expr-seq-constant rhs))]))
(define (parse-def-syntaxes form)
;; Currently, treat def-syntaxes as a no-op. The compiler will not produce
;; syntax transformers.
(make-Constant (void)))
(define (parse-req form)
(let ([resolver (current-module-path-resolver)])
(match form
[(struct req (reqs dummy))
(let ([require-statement (parse-req-reqs reqs)])
(match require-statement
[(list '#%require (and (? module-path?) path))
(let ([resolved-path ((current-module-path-resolver) path (current-module-path))])
(cond
[(symbol? resolved-path)
(make-Require (make-ModuleLocator resolved-path resolved-path))]
[(path? resolved-path)
(let ([rewritten-path (rewrite-path resolved-path)])
(cond
[(symbol? rewritten-path)
(make-Require (make-ModuleLocator rewritten-path
(normalize-path resolved-path)))]
[else
(printf "Internal error: I don't know how to handle the require for ~s" require-statement)
(error 'parse-req)]))]
[else
(printf "Internal error: I don't know how to handle the require for ~s" require-statement)
(error 'parse-req)]))]
[else
(printf "Internal error: I don't know how to handle the require for ~s" require-statement)
(error 'parse-req)]))])))
;; parse-req-reqs: (stx -> (listof ModuleLocator))
(define (parse-req-reqs reqs)
(match reqs
[(struct stx (encoded))
(unwrap-wrapped encoded)]))
(define (unwrap-wrapped encoded)
(cond [(wrapped? encoded)
(match encoded
[(struct wrapped (datum wraps certs))
(unwrap-wrapped datum)])]
[(pair? encoded)
(cons (unwrap-wrapped (car encoded))
(unwrap-wrapped (cdr encoded)))]
[(null? encoded)
null]
[else
encoded]))
;; parse-seq: seq -> Expression
(define (parse-seq form)
(match form
[(struct seq (forms))
(make-Seq (map parse-form-item forms))]))
;; parse-form-item: (U form Any) -> Expression
(define (parse-form-item item)
(cond
[(form? item)
(parse-form item)]
[else
(make-Constant item)]))
;; parse-splice: splice -> Expression
(define (parse-splice form)
(match form
[(struct splice (forms))
(make-Splice (map parse-splice-item forms))]))
;; parse-splice-item: (U form Any) -> Expression
(define (parse-splice-item item)
(cond
[(form? item)
(parse-form item)]
[else
(make-Constant item)]))
;; parse-mod: mod -> Expression
(define (parse-mod form)
(match form
[(struct mod (name srcname self-modidx prefix provides requires
body syntax-body unexported max-let-depth dummy lang-info
internal-context))
(let ([self-path
((current-module-path-index-resolver)
self-modidx
(current-module-path))])
(cond
[(symbol? self-path)
(make-Module name
(make-ModuleLocator self-path self-path)
(parse-prefix prefix)
(parse-mod-requires self-modidx requires)
(parse-mod-body body))]
[else
(let ([rewritten-path (rewrite-path self-path)])
(cond
[(symbol? rewritten-path)
(make-Module name
(make-ModuleLocator rewritten-path
(normalize-path self-path))
(parse-prefix prefix)
(parse-mod-requires self-modidx requires)
(parse-mod-body body))]
[else
(error 'parse-mod "Internal error: unable to resolve module path ~s" self-path)]))]))]))
;; parse-mod-requires: module-path-index (listof (pair (U Integer #f) (listof module-path-index))) -> (listof ModuleLocator)
(define (parse-mod-requires enclosing-module-path-index requires)
;; We only care about phase 0 --- the runtime.
(let ([resolver (current-module-path-index-resolver)])
(let loop ([requires requires])
(cond
[(empty? requires)
empty]
[(= (car (first requires))
0)
(map (lambda (m)
(let ([enclosing-path (resolver enclosing-module-path-index (current-module-path))])
(cond
[(symbol? enclosing-path)
(wrap-module-name (resolver m (current-module-path)))]
[(path? enclosing-path)
(wrap-module-name (resolver m enclosing-path))])))
(cdr (first requires)))]
[else
(loop (rest requires))]))))
;; parse-mod-body: (listof (or/c form? any/c)) -> Expression
(define (parse-mod-body body)
(let ([parse-item (lambda (item)
(cond
[(form? item)
(parse-form item)]
[else
(make-Constant item)]))])
(make-Splice (map parse-item body))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(define (parse-expr expr)
(cond
[(lam? expr)
(parse-lam expr (make-lam-label))]
[(closure? expr)
(parse-closure expr)]
[(case-lam? expr)
(parse-case-lam expr)]
[(let-one? expr)
(parse-let-one expr)]
[(let-void? expr)
(parse-let-void expr)]
[(install-value? expr)
(parse-install-value expr)]
[(let-rec? expr)
(parse-let-rec expr)]
[(boxenv? expr)
(parse-boxenv expr)]
[(localref? expr)
(parse-localref expr)]
[(toplevel? expr)
(parse-toplevel expr)]
[(topsyntax? expr)
(parse-topsyntax expr)]
[(application? expr)
(parse-application expr)]
[(branch? expr)
(parse-branch expr)]
[(with-cont-mark? expr)
(parse-with-cont-mark expr)]
[(beg0? expr)
(parse-beg0 expr)]
[(varref? expr)
(parse-varref expr)]
[(assign? expr)
(parse-assign expr)]
[(apply-values? expr)
(parse-apply-values expr)]
[(primval? expr)
(parse-primval expr)]))
(define (parse-lam expr entry-point-label)
(match expr
[(struct lam (name flags num-params param-types rest? closure-map closure-types max-let-depth body))
(let ([lam-name (extract-lam-name name)])
(make-Lam lam-name
num-params
rest?
(parse-expr-seq-constant body)
(vector->list closure-map)
entry-point-label))]))
;; parse-closure: closure -> (U Lam EmptyClosureReference)
;; Either parses as a regular lambda, or if we come across the same closure twice,
;; breaks the cycle by creating an EmptyClosureReference with the pre-existing lambda
;; entry point.
(define (parse-closure expr)
(match expr
[(struct closure (code gen-id))
(let ([seen (seen-closures)])
(cond
[(hash-has-key? seen gen-id)
(match code
[(struct lam (name flags num-params param-types rest? closure-map closure-types max-let-depth body))
(let ([lam-name (extract-lam-name name)])
(make-EmptyClosureReference lam-name
num-params
rest?
(hash-ref seen gen-id)))])]
[else
(let ([fresh-entry-point (make-lam-label)])
(hash-set! seen gen-id fresh-entry-point)
(parse-lam code fresh-entry-point))]))]))
;; extract-lam-name: (U Symbol Vector) -> (U Symbol LamPositionalName)
(define (extract-lam-name name)
(cond
[(symbol? name)
name]
[(vector? name)
(match name
[(vector (and (? symbol?) sym)
(and (? path?) source)
(and (? number?) line)
(and (? number?) column)
(and (? number?) offset)
(and (? number?) span)
_)
(let ([try-to-rewrite (rewrite-path source)])
(make-LamPositionalName sym
(if try-to-rewrite
(symbol->string try-to-rewrite)
(path->string source))
line
column
offset
span))]
[(vector (and (? symbol?) sym)
(and (? symbol?) source)
(and (? number?) line)
(and (? number?) column)
(and (? number?) offset)
(and (? number?) span)
_)
(make-LamPositionalName sym
(symbol->string source)
line
column
offset
span)]
[else
(string->symbol (format "~s" name))])]
[else
'unknown
;; The documentation says that the name must be a symbol or vector, but I'm seeing cases
;; where it returns the empty list when there's no information available.
]))
(define (parse-case-lam exp)
(match exp
[(struct case-lam (name clauses))
(let ([case-lam-label (make-lam-label)])
(make-CaseLam (extract-lam-name name)
(map (lambda (l)
(cond
[(closure? l)
(parse-closure l)]
[else
(parse-lam l (make-lam-label))]))
clauses)
case-lam-label))]))
(define (parse-let-one expr)
(match expr
[(struct let-one (rhs body flonum? unused?))
;; fixme: use flonum? and unused? to generate better code.
(make-Let1 (parse-expr-seq-constant rhs)
(parse-expr-seq-constant body))]))
;; parse-expr-seq-constant: (U expr seq Any) -> Expression
(define (parse-expr-seq-constant x)
(cond
[(expr? x) (parse-expr x)]
[(seq? x) (parse-seq x)]
[else (make-Constant x)]))
(define (parse-let-void expr)
(match expr
[(struct let-void (count boxes? body))
(make-LetVoid count (parse-expr-seq-constant body) boxes?)]))
(define (parse-install-value expr)
(match expr
[(struct install-value (count pos boxes? rhs body))
(make-Seq (list (make-InstallValue count pos (parse-expr-seq-constant rhs) boxes?)
(parse-expr-seq-constant body)))]))
(define (parse-let-rec expr)
(match expr
[(struct let-rec (procs body))
(make-LetRec (map (lambda (p) (parse-lam p (make-lam-label)))
procs)
(parse-expr-seq-constant body))]))
(define (parse-boxenv expr)
(match expr
[(struct boxenv (pos body))
(make-BoxEnv pos (parse-expr-seq-constant body))]))
(define (parse-localref expr)
(match expr
[(struct localref (unbox? pos clear? other-clears? flonum?))
;; FIXME: we should use clear? at the very least: as I understand it,
;; this is here to maintain safe-for-space behavior.
;; We should also make use of flonum information to generate better code.
(make-LocalRef pos unbox?)]))
(define (parse-toplevel expr)
(match expr
;; FIXME: we should also keep track of const? and ready? to produce better code, and to
;; do the required runtime checks when necessary (const?=#f, ready?=#f)
[(struct toplevel (depth pos const? ready?))
(make-ToplevelRef depth pos)]))
(define (parse-topsyntax expr)
;; We should not get into this because we're only parsing the runtime part of
;; the bytecode. Treated as a no-op.
(make-Constant (void)))
(define (parse-application expr)
(match expr
[(struct application (rator rands))
(make-App (parse-application-rator rator)
(map parse-application-rand rands))]))
(define (parse-application-rator rator)
(cond
[(expr? rator)
(parse-expr rator)]
[(seq? rator)
(parse-seq rator)]
[else
(make-Constant rator)]))
(define (parse-application-rand rand)
(cond
[(expr? rand)
(parse-expr rand)]
[(seq? rand)
(parse-seq rand)]
[else
(make-Constant rand)]))
(define (parse-branch expr)
(match expr
[(struct branch (test then else))
(make-Branch (parse-expr-seq-constant test)
(parse-expr-seq-constant then)
(parse-expr-seq-constant else))]))
(define (parse-with-cont-mark expr)
(match expr
[(struct with-cont-mark (key val body))
(make-WithContMark (parse-expr-seq-constant key)
(parse-expr-seq-constant val)
(parse-expr-seq-constant body))]))
(define (parse-beg0 expr)
(match expr
[(struct beg0 (seq))
(make-Begin0 (map parse-expr-seq-constant seq))]))
(define (parse-varref expr)
(match expr
[(struct varref (toplevel))
(make-VariableReference (parse-toplevel toplevel))]))
(define (parse-assign expr)
(match expr
[(struct assign ((struct toplevel (depth pos const? ready?)) rhs undef-ok?))
(make-ToplevelSet depth pos (parse-expr-seq-constant rhs))]))
(define (parse-apply-values expr)
(match expr
[(struct apply-values (proc args-expr))
(make-ApplyValues (parse-expr-seq-constant proc)
(parse-expr-seq-constant args-expr))]))
(define (parse-primval expr)
(match expr
[(struct primval (id))
(let ([name (hash-ref primitive-table id)])
(make-PrimitiveKernelValue name))]))