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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762
|
;(include-book "m1")
;(certify-book "theorems-a-and-b" 1)
(in-package "M1")
(include-book "tmi-reductions")
(include-book "implementation")
; The tmi-reductions book contains a theorem establishing that the function
; TMI, as defined in the Boyer-Moore paper, "A Mechanical Proof of the Turing
; Completeness of Pure Lisp," is equivalent (modulo the representation of
; tapes) to an algorithm named tmi3 implementable on M1. The implementation
; book defines an M1 program (PI) and various schedules and proves that (PI)
; implements tmi3.
; For what it's worth: proof-1 required 160 defthms. This proof requires 138,
; mainly because defsys handles almost all of the implementation.lisp proofs.
; I also cleaned up the theorem A and B proofs a bit.
; In the Boyer-Moore paper, "A Mechanical Proof of the Turing Completeness of
; Pure Lisp" we prove two things: about a computational paradigm emulating a
; Turing machine.
; (a) If the Turing machine runs forever, then the emulator runs forever.
; This theorem is stated in its contrapositive form: if the emulator halts
; then the Turing machine halts.
; (b) If the Turing machine halts with tape tape, then the emulator halts with
; the same tape (modulo representation).
; Outline:
; Both of these theorems depend on a stronger Simulation Theorem that precisely
; relates running (PI) and TMI. So our first goal below is to prove the
; Simulation Theorem. Theorem B follows immediately from the Simulation
; Theorem. Theorem A, however, requires a little more work: we have to define
; a clock function that converts a schedule (under which the M1 emulator halts)
; to a clock (under which TMI halts). The definition of this conversion
; function depends crucially on the Monotonicity property of the M1 schedule
; function used in the Simulation Theorem: TMI doesn't halt, then the schedule
; for n+1 steps is greater than the schedule for n steps. Given Monotonicity,
; we can find an appropriate clock by searching upwards for ever greater
; clocks, knowing that the corresponding schedule gets longer and longer and
; that eventually it will exceed the length of the schedule that makes M1 halt.
; So our subgoals are: The Simulation Theorem, Monotonicity, Theorem A, and,
; finally, Theorem B.
; Convention on Clocks and Schedules
; Both tmi and run are ``non-terminating'' interpreters that have had
; artificial means imposed upon them to insure (abnormal) termination. For tmi
; the artificial means is a number that is decreased every time tmi recurs.
; When abnormal termination occurs, tmi returns nil; normal termination is
; indicated by returning the final tape, a cons. We call the number
; controlling tmi a ``clock.''
; For M1's run the artificial means is a list that is cdr'd every time run
; recurs. Abnormal termination is indicated by returning an M1 state that is
; not HALTed, by which we mean the next-inst in the returned state is something
; other than a HALT instruction. Normal termination is indicated by returning
; a state in which the next instruction is HALT. We call the list controlling
; run a ``thread schedule'' (in preparation for the eventual addition of
; threads to the model). But because M1's run isn't sensitive to the ``thread
; identifier'' listed in its schedule -- it always steps the only thread -- we
; can convert an M1 schedule to a ``clock,'' a number that determines how long
; the schedule is. The typical idiom for applications of run will be (run
; (repeat 'tick <clock>) s), where <clock> is a numeric expression.
; Our two theorems, A and B involve clocks.
; A: If run halts normally in so many clock ticks, then there exists a clock
; that makes tmi halt normally.
; B: If tmi halts normally in so many clock ticks, then there exists a clock
; that makes run halt normally and return the ``same'' tape.
; We thus talk about 4 clocks and we adopt the following naming conventions, by
; restating A and B in terms that explicitly identify the symbol we will use
; for each clock:
; A: If run halts normally in i clock ticks, then there exists a j
; that makes tmi halt normally.
; B: If tmi halts normally in n clock ticks, then there exists a k
; that makes run halt normally and return the ``same'' tape.
; Summarizing our clock naming conventions:
; i -- the number of steps run takes to halt in the hypothesis of theorem A
; j -- the number of steps tmi takes to halt in the conclusion of theorem A
; n -- the number of steps tmi takes to halt in the hypothesis of theorem B
; k -- the number of steps run takes to halt in the conclusion of theorem B
; Note that j is actually a function of i: given i, there exists a j.
; Note that k is actually a function of n: given n, there exists a k.
; So while i and n are variable symbols, j and k are function symbols in our
; quantifier-free setting.
; The Simulation Theorem
(defun psi-clock (st tape pos tm w nnil n)
(clk+ 2
(main-clock nil st tape pos tm w nnil n)))
; Our goal is to express precisely the relationship between TMI and an m1 run of
; the M1 system (PSI). We take it in steps. First, we express the relationship
; between an m1 run of (PSI) and tmi3. Then we move to TMI terms by applying the
; functions that re-represent Turing machines and tapes.
(defthm tmi3-simulation
(implies (and (natp st)
(natp tape)
(natp pos)
(natp tm)
(natp w)
(equal nnil (nnil w))
(< st (expt 2 w)))
(equal
(m1 (make-state 0
'(0 0 0 0 0 0 0 0 0 0 0 0 0)
(push nnil
(push w
(push tm
(push pos (push tape (push st nil))))))
(psi))
(psi-clock st tape pos tm w nnil n))
(let ((s (make-state
*main*
'(0 0 0 0 0 0 0 0 0 0 0 0 0)
(push 2
(push nnil
(push w
(push tm
(push pos (push tape (push st nil)))))))
(psi))))
(if
(equal (mv-nth 0 (tmi3 st tape pos tm w n))
0)
(make-state
*tmi3-loop*
(main-final-locals nil st tape pos tm w nnil n s)
(main-final-stack nil st tape pos tm w nnil n s)
(psi))
(make-state
2
(update-nth*
0
'(0 0 0 0 0 0)
(main-final-locals nil st tape pos tm w nnil n s))
(push
(mv-nth 3 (tmi3 st tape pos tm w n))
(push
(mv-nth 2 (tmi3 st tape pos tm w n))
(push (mv-nth 1 (tmi3 st tape pos tm w n))
(push (mv-nth 0 (tmi3 st tape pos tm w n))
nil))))
(psi)))))))
(in-theory (disable psi-clock))
; I am not convinced the following theorem is sufficient for Theorem A. But let's go with it
; and see what happens.
(defthm integerp-mv-nth-convert-tape-to-tapen-pos
(implies (tapep tape)
(and (integerp (mv-nth 0 (convert-tape-to-tapen-pos tape)))
(integerp (mv-nth 1 (convert-tape-to-tapen-pos tape)))))
:hints (("Goal" :in-theory (enable acl2::mv-nth convert-tape-to-tapen-pos tapep)))
:rule-classes :rewrite) ;I'd settle for :tau-system if it were used in backchaining!
(defthm nonneg-mv-nth-convert-tape-to-tapen-pos
(implies (tapep tape)
(and (<= 0 (mv-nth 0 (convert-tape-to-tapen-pos tape)))
(<= 0 (mv-nth 1 (convert-tape-to-tapen-pos tape)))))
:hints (("Goal" :in-theory (enable acl2::mv-nth convert-tape-to-tapen-pos tapep)))
:rule-classes :linear)
(defthm positive-natp-ncode-rewrite-version
(implies (and (natp w)
(turing1-machinep tm w))
(and (integerp (ncode tm w))
(< 0 (ncode tm w)))))
(in-theory (disable tmi2-is-tmi1 tmi3-is-tmi2 renaming-map properties-of-instr))
; We disable the intermediate tmi theorems because they rewrite TMI3. We
; disable RENAMING-MAP otherwise (cdr (assoc st (renaming-map st tm))) becomes
; 0 and doesn't allow the assoc expression in m1-psi-is-tmi below to match the
; corresponding one in the theorem tmi3-is-tmi. We disable properties-of-instr
; because it (stupidly) FORCEs turing1-machinep when we need to stay at the
; turing-machinep level.
; We really ought to know these tau-like theorems:
(defthm tapep-new-tape
(implies (and (tapep tape)
(operationp op))
(tapep (new-tape op tape)))
:hints (("Goal" :in-theory (enable tapep new-tape))))
(defthm operationp-nth-2-instr
(implies (and (turing-machinep tm)
(instr st current-sym tm))
(operationp (nth 2 (instr st current-sym tm))))
:hints (("Goal" :in-theory (disable properties-of-instr)))) ; <--- forces turing1-machinep!
(defthm symbolp-nth-3-instr
(implies (and (turing-machinep tm)
(instr st current-sym tm))
(symbolp (nth 3 (instr st current-sym tm)))))
(defthm tapep-tmi
(implies (and (symbolp st)
(tapep tape)
(turing-machinep tm)
(tmi st tape tm n))
(tapep (tmi st tape tm n))))
#|
(defthm run-repeat
(equal (run (repeat 'tick (len sched)) s)
(run sched s))
:hints (("Goal" :in-theory (enable run))))
|#
(defun find-k (st tape tm n)
(let* ((map (renaming-map st tm))
(st-prime (cdr (assoc st map)))
(tape-prime (mv-nth 0 (mv-list 2 (convert-tape-to-tapen-pos tape))))
(pos-prime (mv-nth 1 (mv-list 2 (convert-tape-to-tapen-pos tape))))
(w (max-width tm map))
(tm-prime (ncode (tm-to-tm1 tm map) w)))
(psi-clock st-prime tape-prime pos-prime tm-prime w (nnil w) n)))
(defthm simulation
(implies (and (symbolp st)
(tapep tape)
(turing-machinep tm))
(let* ((map (renaming-map st tm))
(st-prime (cdr (assoc st map)))
(tape-prime (mv-nth 0 (convert-tape-to-tapen-pos tape)))
(pos-prime (mv-nth 1 (convert-tape-to-tapen-pos tape)))
(w (max-width1 (tm-to-tm1 tm map)))
(nnil (nnil w))
(tm-prime (ncode (tm-to-tm1 tm map) w))
(s-final
(m1 (make-state 0
'(0 0 0 0 0 0 0 0 0 0 0 0 0)
(push nnil
(push w
(push tm-prime
(push pos-prime
(push tape-prime
(push st-prime nil))))))
(psi))
(find-k st tape tm n))))
(and (iff (equal (next-inst s-final) '(HALT))
(tmi st tape tm n))
(implies (tmi st tape tm n)
(equal (decode-tape-and-pos
(top (pop (stack s-final)))
(top (stack s-final)))
(tmi st tape tm n))))))
:hints (("Goal" :do-not-induct t)))
(in-theory (disable find-k next-inst))
(defun ncode-st (st map)
(cdr (assoc st map)))
(defun ncode-tm (tm map w)
(ncode (tm-to-tm1 tm map) w))
(defun ncode-tape (tape)
(mv-let (tapen pos)
(convert-tape-to-tapen-pos tape)
(declare (ignore pos))
tapen))
(defun ncode-pos (tape)
(mv-let (tapen pos)
(convert-tape-to-tapen-pos tape)
(declare (ignore tapen))
pos))
(defmacro with-conventions (term)
`(let* ((map (renaming-map st tm))
(w (max-width tm map))
(nnil (nnil w))
(st^prime (ncode-st st map))
(tm^prime (ncode-tm tm map w))
(tape^prime (ncode-tape tape))
(pos^prime (ncode-pos tape))
(s_0 (make-state 0
'(0 0 0 0 0 0 0 0 0 0 0 0 0)
(push* nnil w tm^prime pos^prime tape^prime st^prime nil)
(Psi))))
(implies (and (symbolp st) (tapep tape) (turing-machinep tm))
,term)))
(defthm theorem-b
(with-conventions
(implies (and (natp n)
(tmi st tape tm n))
(let ((s_f (M1 s_0 (find-k st tape tm n))))
(and (haltedp s_f)
(equal (decode-tape-and-pos
(top (pop (stack s_f)))
(top (stack s_f)))
(tmi st tape tm n)))))))
; Now we turn to Theorem A. Recall that here we have a clock i at which m1
; halts normally and wish to exhibit a j at which tmi halts normally. We find
; j by searching up from 0. For each value of j, see if tmi halts normally.
; If not, increment j by 1 and repeat. Why does this search terminate?
; Consider (find-k ... j), i.e., the clock for m1 corresponding to the current j.
; We know by the Simulation theorem tmi halts (or not) at j iff m1 halts at (find-k
; ... j). So if tmi is not halted at j, then m1 is not halted at (find-k ...j).
; But eventually, (find-k ...j) exceeds i, the time at which m1 does halt. Why do
; we know (find-k ...j) will eventually exceed i? Because k is monotonic: as j
; increases, k increases, provided tmi has not halted. Note that we are also
; relying on the fact that once m1 halts normally it stays halted, i.e., if
; m1 is halted at i then it is halted at k if i <= k. So we can bound the
; search for j by looking until i <= (find-k ...j).
;
; So we prove Monotonicity next, then prove that m1 stays halted, then define
; the search mechanism for j, and then prove theorem a.
#|
(defthm len-repeat
(equal (len (repeat x n))
(nfix n)))
|#
(in-theory (enable binary-clk+))
(defthm k-non-0
(implies (and (natp st)
(natp tape)
(natp pos)
(natp tm)
(natp w)
(not (zp n)))
(< 0 (find-k st tape tm n)))
:hints (("Goal" :in-theory (enable find-k psi-clock)))
:rule-classes :linear)
; The first problem with monotonicity is that at the level of k it is about
; main-clock, main-loop-clock, tmi3-clock, and tmi3-loop-clock. But all the
; action is happening at tmi3-loop-clock, where the computation actually hangs.
; So I first prove that tmi3-loop-clock is monotonic and then raise that result
; up through the others.
(defun tmi3-trace (st tape pos tm w n)
(declare (xargs :measure (acl2-count n)))
(cond ((zp n)
nil)
((equal (ninstr st (current-symn tape pos) tm w) -1)
(list (list t st tape pos)))
(t
(cons (list nil st tape pos)
(mv-let (new-tape new-pos)
(new-tape2 (nop (ninstr st (current-symn tape pos) tm w) w) tape pos)
(tmi3-trace (nst-out (ninstr st (current-symn tape pos) tm w) w)
new-tape
new-pos
tm
w
(- n 1)))))))
(defun k-halt (st tape pos tm w)
(CLK+ 5
(CLK+ (CURRENT-SYMN-CLOCK '(1 0 0 14)
TAPE POS)
(CLK+ 5
(CLK+ (NINSTR1-CLOCK '(0 0 14)
ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL w))
7)))))
(defun k-step (st tape pos tm w)
(CLK+
5
(CLK+
(CURRENT-SYMN-CLOCK '(1 0 0 14)
TAPE POS)
(CLK+
5
(CLK+
(NINSTR1-CLOCK '(0 0 14)
ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL W))
(CLK+
8
(CLK+
(CURRENT-SYMN-CLOCK '(1 0 0 2 14)
TAPE POS)
(CLK+
5
(CLK+
(NINSTR1-CLOCK '(0 0 2 14)
ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL W))
(CLK+
3
(CLK+
(NST-OUT-CLOCK
'(0 2 14)
(!NINSTR1 ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL W))
W)
(CLK+
5
(CLK+
(CURRENT-SYMN-CLOCK '(1 0 0 1 2 14)
TAPE POS)
(CLK+
5
(CLK+
(NINSTR1-CLOCK '(0 0 1 2 14)
ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL W))
(CLK+
3
(CLK+
(NOP-CLOCK
'(0 1 2 14)
(!NINSTR1 ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL W))
W)
(CLK+
4
(CLK+
(NEW-TAPE2-CLOCK
'(1 2 14)
(!NOP
(!NINSTR1 ST (!CURRENT-SYMN TAPE POS)
TM W (NNIL W))
W)
TAPE POS)
10)))))))))))))))))))
(defun k* (trace tm w)
(if (endp trace)
0
(if (car (car trace))
(k-halt (nth 1 (car trace))
(nth 2 (car trace))
(nth 3 (car trace))
tm w)
(+ (k-step (nth 1 (car trace))
(nth 2 (car trace))
(nth 3 (car trace))
tm w)
(k* (cdr trace) tm w)))))
(defthm tmi3-loop-clock-is-k*
(implies (and (natp st)
(natp tape)
(natp pos)
(natp tm)
(natp w)
(< st (expt 2 w)))
(equal (tmi3-loop-clock st tape pos tm w (nnil w) n)
(k* (tmi3-trace st tape pos tm w n) tm w)))
:hints (("Goal" :in-theory (enable tmi3-loop-clock))
("Subgoal *1/10'" :expand (TMI3-LOOP-CLOCK ST TAPE POS TM W (NNIL W) N))))
(defthm positive-k-halt
(and (integerp (k-halt st tape pos tm w))
(< 0 (k-halt st tape pos tm w)))
:rule-classes :type-prescription)
(defthm positive-k-step
(and (integerp (k-step st tape pos tm w))
(< 0 (k-step st tape pos tm w)))
:rule-classes :type-prescription)
(in-theory (disable k-halt k-step))
(defthm positive-k*
(implies (consp x)
(< 0 (k* x tm w)))
:rule-classes :linear)
; -----------------------------------------------------------------
(defthm trace-extension
(implies (and (natp n)
(equal (mv-nth 0 (tmi3 st tape pos tm w n)) 0))
(equal (tmi3-trace st tape pos tm w (+ 1 n))
(append (tmi3-trace st tape pos tm w n)
(mv-let (flg st1 tape1 pos1)
(tmi3 st tape pos tm w n)
(declare (ignore flg))
(list (list (EQUAL (NINSTR ST1 (CURRENT-SYMN tape1 pos1) TM W)
-1)
st1 tape1 pos1)))))))
(defun tracep (x)
(cond ((endp x) t)
((car (car x)) (endp (cdr x)))
(t (tracep (cdr x)))))
(defthm tracep-tmi3-trace
(tracep (tmi3-trace st tape pos tm w n)))
(defthm k*-append
(implies (tracep (append a b))
(equal (k* (append a b) tm w)
(+ (k* a tm w) (k* b tm w)))))
(defthm tracep-append
(equal (tracep (append a b))
(if (tracep a)
(if (car (car (last a)))
(endp b)
(tracep b))
nil)))
(defthm tmi3-v-last-tmi3-trace
(implies (and (integerp n)
(<= 0 n)
(force (equal 0 (mv-nth 0 (tmi3 st tape pos tm w n)))))
(not (car (car (last (tmi3-trace st tape pos tm w n))))))
:hints (("Goal" :in-theory (enable tmi3))))
(defthm k*-tmi3-trace-monotonic
(implies (and (natp n)
(equal (mv-nth 0 (tmi3 st tape pos tm w n))
0))
(< (k* (tmi3-trace st tape pos tm w n) tm w)
(k* (tmi3-trace st tape pos tm w (+ n 1)) tm w)))
:rule-classes :linear)
(defthm final-tmi3-state-is-proper
(implies (and (natp st)
(natp tape)
(natp pos)
(natp tm)
(natp w)
(< st (expt 2 w)))
(and (natp (mv-nth 1 (tmi3 st tape pos tm w n)))
(< (mv-nth 1 (tmi3 st tape pos tm w n)) (expt 2 w))))
:rule-classes
((:linear :corollary
(implies (and (natp st)
(natp tape)
(natp pos)
(natp tm)
(natp w)
(< st (expt 2 w)))
(and (<= 0 (mv-nth 1 (tmi3 st tape pos tm w n)))
(< (mv-nth 1 (tmi3 st tape pos tm w n)) (expt 2 w)))))
(:rewrite
:corollary
(implies (and (natp st)
(natp tape)
(natp pos)
(natp tm)
(natp w)
(< st (expt 2 w)))
(integerp (mv-nth 1 (tmi3 st tape pos tm w n)))))))
(defthm acl2-numberp-cdr-assoc-equal-st-renaming-map-st
(acl2-numberp (CDR (ASSOC-EQUAL ST (RENAMING-MAP ST TM))))
:hints (("Goal" :in-theory (enable renaming-map))))
(defthm find-k-monotonic
(implies (and (symbolp st)
(tapep tape)
(turing-machinep tm)
(natp n)
(force (not (tmi st tape tm n))))
(< (find-k st tape tm n)
(find-k st tape tm (+ 1 n))))
:hints (("Goal" :do-not-induct t
:in-theory (e/d (find-k psi-clock main-clock main-loop-clock tmi3-clock)
(turing1-4-tuple)
)))
:rule-classes :linear)
(defthm 0<-find-k
(< 0 (find-k st tape tm n))
:hints (("Goal" :in-theory (enable find-k psi-clock))))
(defun find-j1 (st tape tm i j)
(declare (xargs :measure (nfix (- (+ 1 (nfix i))
(find-k st tape tm j)))
:otf-flg t))
(if (and (symbolp st)
(tapep tape)
(turing-machinep tm)
(natp i)
(natp j))
(if (equal (tmi st tape tm j) nil)
(if (<= i (find-k st tape tm j))
j
(find-j1 st tape tm i (+ 1 j)))
j)
0))
(defun find-j (st tape tm i)
(find-j1 st tape tm i 0))
; The crucial property of find-j1 is that it either returns a j
; that makes tmi halt or else an j whose k is greater than i.
(defthm crucial-property-of-find-j1
(implies (and (symbolp st)
(tapep tape)
(turing-machinep tm)
(natp i)
(natp j))
(or (tmi st tape tm (find-j1 st tape tm i j))
(<= i (find-k st tape tm
(find-j1 st tape tm i j)))))
:rule-classes nil)
; Now we work on the M1 Stays Halted theorem.
(defthm program-step
(equal (program (step s))
(program s))
:hints (("Goal" :in-theory (enable step))))
(defthm program-m1
(equal (program (m1 s a))
(program s))
:hints (("Goal" :in-theory (enable m1))))
(defthm m1-stays-halted
(implies (equal (next-inst s) '(HALT))
(equal (m1 s clock) s))
:hints (("Goal" :induct (m1 s clock)
:in-theory (enable m1 step))))
;(defun run-repeat-hint (k s)
; (if (zp k)
; s
; (run-repeat-hint (- k 1) (step s))))
(defthm m1-stays-halted-clk+-version
(implies (equal (next-inst (m1 s a)) '(HALT))
(equal (m1 s (clk+ a b))
(m1 s a)))
:hints (("Goal" :in-theory (disable binary-clk+))))
#|
(defthm ap-repeat
(implies (and (natp i)
(natp j))
(equal (clk+ (repeat 'tick i)
(repeat 'tick j))
(repeat 'tick (+ i j)))))
|#
(defthm m1-stays-halted-repeat-version
(implies (and (<= i k)
(natp i)
(natp k)
(equal (next-inst (m1 s i)) '(HALT)))
(equal (m1 s k)
(m1 s i)))
:hints (("Goal" :use (:instance m1-stays-halted-clk+-version
(a i)
(b (- k i)))
:do-not-induct t
:in-theory (disable m1-clk+ m1-stays-halted-clk+-version)))
:rule-classes nil)
(in-theory (disable find-j1))
(defthm theorem-a
(with-conventions
(implies (natp i)
(let ((s_f (m1 s_0 i)))
(implies
(haltedp s_f)
(tmi st tape tm (find-j st tape tm i))))))
; Proof: Let s-init be the initial M1 state above. (m1
; s-init i) terminates. find-j1 marches up from 0 looking for a j that makes tmi
; halt. If it finds it we're done. Otherwise, it eventually finds a j such
; tmi doesn't halt but (find-k .... j) exceeds i. That's a contradiction because if
; tmi doesn't halt at j then M1 doesn't halt at (find-k ... j), by the Simulation
; theorem and the fact that M1 stays halted. Q.E.D.
:hints (("Goal" :do-not-induct t
:use ((:instance crucial-property-of-find-j1 (j 0))
(:instance m1-stays-halted-repeat-version
(i i)
(k (find-k st tape tm (find-j1 st tape tm i 0)))
(s (MAKE-STATE
0 '(0 0 0 0 0 0 0 0 0 0 0 0 0)
(PUSH
(NNIL (MAX-WIDTH TM (RENAMING-MAP ST TM)))
(PUSH
(MAX-WIDTH TM (RENAMING-MAP ST TM))
(PUSH
(NCODE (TM-TO-TM1 TM (RENAMING-MAP ST TM))
(MAX-WIDTH TM (RENAMING-MAP ST TM)))
(PUSH (ACL2::MV-NTH 1 (CONVERT-TAPE-TO-TAPEN-POS TAPE))
(PUSH (ACL2::MV-NTH 0 (CONVERT-TAPE-TO-TAPEN-POS TAPE))
(PUSH (CDR (ACL2::ASSOC-EQUAL ST (RENAMING-MAP ST TM)))
NIL))))))
(PSI))))))))
; Revision
(defun down (st tape tm)
; Down is a computable function.
(let* ((map (renaming-map st tm))
(st-prime (cdr (assoc st map)))
(tape-prime (mv-nth 0 (mv-list 2 (convert-tape-to-tapen-pos tape))))
(pos-prime (mv-nth 1 (mv-list 2 (convert-tape-to-tapen-pos tape))))
(w (max-width tm map))
(nnil (nnil w))
(tm-prime (ncode (tm-to-tm1 tm map) w)))
(make-state 0
'(0 0 0 0 0 0 0 0 0 0 0 0 0)
(push nnil
(push w
(push tm-prime
(push pos-prime
(push tape-prime
(push st-prime nil))))))
(psi))))
(defun up (s)
(DECODE-TAPE-AND-POS (TOP (POP (STACK s)))
(TOP (STACK s))))
(defthm a
(implies
(and (symbolp st)
(tapep tape)
(turing-machinep tm)
(natp i))
(let ((s_f (m1 (down st tape tm) i)))
(implies (haltedp s_f)
(tmi st tape tm (find-j st tape tm i)))))
:hints (("Goal" :use theorem-a :in-theory (disable theorem-a) :do-not-induct t))
:rule-classes nil)
(defthm b
(implies (and (symbolp st)
(tapep tape)
(turing-machinep tm)
(tmi st tape tm n))
(let ((s_f (m1 (down st tape tm)
(find-k st tape tm n))))
(and (haltedp s_f)
(equal (up s_f)
(tmi st tape tm n)))))
:rule-classes nil)
|