1 ;;;; This file is for compiler tests which have side effects (e.g.
2 ;;;; executing DEFUN) but which don't need any special side-effecting
3 ;;;; environmental stuff (e.g. DECLAIM of particular optimization
4 ;;;; settings). Similar tests which *do* expect special settings may
5 ;;;; be in files compiler-1.impure.lisp, compiler-2.impure.lisp, etc.
7 ;;;; This software is part of the SBCL system. See the README file for
10 ;;;; While most of SBCL is derived from the CMU CL system, the test
11 ;;;; files (like this one) were written from scratch after the fork
14 ;;;; This software is in the public domain and is provided with
15 ;;;; absolutely no warranty. See the COPYING and CREDITS files for
16 ;;;; more information.
18 (cl:in-package :cl-user)
20 (load "assertoid.lisp")
22 ;;; Old CMU CL code assumed that the names of "keyword" arguments are
23 ;;; necessarily self-evaluating symbols, but ANSI Common Lisp allows
24 ;;; them to be any symbols, not necessarily keywords, and thus not
25 ;;; necessarily self-evaluating. Make sure that this works.
26 (defun newfangled-cons (&key ((left-thing x)) ((right-thing y)))
28 (assert (equal (cons 1 2) (newfangled-cons 'right-thing 2 'left-thing 1)))
30 ;;; ANSI specifically says that duplicate keys are OK in lambda lists,
31 ;;; with no special exception for macro lambda lists. (As reported by
32 ;;; Pierre Mai on cmucl-imp 2001-03-30, Python didn't think so. The
33 ;;; rest of the thread had some entertainment value, at least for me
34 ;;; (WHN). The unbelievers were besmote and now even CMU CL will
35 ;;; conform to the spec in this regard. Who needs diplomacy when you
36 ;;; have brimstone?:-)
37 (defmacro ayup-duplicate-keys-are-ok-i-see-the-lite (&key k)
39 (assert (equal (ayup-duplicate-keys-are-ok-i-see-the-lite :k 112) 112))
40 (assert (equal (ayup-duplicate-keys-are-ok-i-see-the-lite :k 'x :k 'y) 'x))
42 ;;; As reported by Alexey Dejneka (sbcl-devel 2002-01-30), in
43 ;;; sbcl-0.7.1 plus his patch (i.e. essentially sbcl-0.7.1.2), the
44 ;;; compiler barfed on this, blowing up in FIND-IN-PHYSENV looking for
45 ;;; the LAMBDA-VAR named NUM. That was fixed in sbcl-0.7.1.3.
46 (defun parse-num (index)
53 (when (and (digs) (digs)) x))))
55 ;;; Bug 132: The compiler used to fail to compile INTEGER-valued CATCH
56 ;;; tags. This was fixed by Alexey Dejneka in sbcl-0.7.1.14. (INTEGER
57 ;;; catch tags are still a bad idea because EQ is used to compare
58 ;;; tags, and EQ comparison on INTEGERs is unportable; but now it's a
59 ;;; compiler warning instead of a failure to compile.)
61 (catch 0 (print 1331)))
63 ;;; Bug 150: In sbcl-0.7.1.15, compiling this code caused a failure in
64 ;;; SB-C::ADD-TEST-CONSTRAINTS:
65 ;;; The value NIL is not of type SB-C::CONTINUATION.
66 ;;; This bug was fixed by APD in sbcl-0.7.1.30.
67 (defun bug150-test1 ()
69 (flet ((wufn () (glorp table1 4.9)))
70 (gleep *uustk* #'wufn "#1" (list)))
71 (if (eql (lo foomax 3.2))
73 (error "not ~S" '(eql (lo foomax 3.2))))
75 ;;; A simpler test case for bug 150: The compiler died with the
76 ;;; same type error when trying to compile this.
77 (defun bug150-test2 ()
81 ;;; bug 147, fixed by APD 2002-04-28
83 ;;; This test case used to crash the compiler, e.g. with
84 ;;; failed AVER: "(= (LENGTH (BLOCK-SUCC CALL-BLOCK)) 1)"
85 (defun bug147 (string ind)
89 (typep (char string ind) '(member #\1)))
92 ;;; bug reported and fixed by Matthias Hoelzl sbcl-devel 2002-05-13
93 (defmacro foo-2002-05-13 () ''x)
94 (eval '(foo-2002-05-13))
95 (compile 'foo-2002-05-13)
96 (foo-2002-05-13) ; (The bug caused UNDEFINED-FUNCTION to be signalled here.)
98 ;;; floating point pain on the PPC.
100 ;;; This test case used to fail to compile on most powerpcs prior to
101 ;;; sbcl-0.7.4.2x, as floating point traps were being incorrectly
103 (defun floating-point-pain (x)
104 (declare (single-float x))
107 ;;; bug found and fixed ca. sbcl-0.7.5.12: The INTERSECTION-TYPE
108 ;;; here satisfies "is a subtype of ARRAY-TYPE", but can't be
109 ;;; accessed with ARRAY-TYPE accessors like
110 ;;; ARRAY-TYPE-SPECIALIZED-ELEMENT-TYPE, so ARRAY-related
111 ;;; DEFTRANSFORMs died with TYPE-ERROR at compile time when
112 ;;; compiling the DEFUN here.
113 (defun stupid-input-to-smart-array-deftransforms-0-7-5-12 (v)
114 (declare (type (and simple-vector fwd-type-ref) v))
117 ;;; Ca. sbcl-0.7.5.15 the compiler would fail an internal consistency
118 ;;; check on this code because it expected all calls to %INSTANCE-REF
119 ;;; to be transformed away, but its expectations were dashed by perverse
120 ;;; code containing app programmer errors like this.
121 (defstruct something-known-to-be-a-struct x y)
122 (multiple-value-bind (fun warnings-p failure-p)
126 (cond (t (a4 a2 a3))))
128 (declare (type (or simple-vector null) a5 a6))
129 (something-known-to-be-a-struct-x a5))
133 (cond ((and (funcall a12 a2)
138 (let ((a15 (a1 a2 a3)))
141 (values #'a17 #'a11))))
142 ;; Python sees the structure accessor on the known-not-to-be-a-struct
143 ;; A5 value and is very, very disappointed in you. (But it doesn't
144 ;; signal BUG any more.)
147 ;;; On the SPARC, there was an erroneous definition of some VOPs used
148 ;;; to compile LOGANDs, which would lead to compilation of the
149 ;;; following function giving rise to a compile-time error (bug
150 ;;; spotted and fixed by Raymond Toy for CMUCL)
151 (defun logand-sparc-bogons (a0 a1 a2 a3 a4 a5 a6 a7 a8 a9 a10)
152 (declare (type (unsigned-byte 32) a0)
153 (type (signed-byte 32) a1 a2 a3 a4 a5 a6 a7 a8 a9 a10)
154 ;; to ensure that the call is a candidate for
156 (optimize (speed 3) (safety 0) (compilation-speed 0) (debug 0)))
158 ;; the call that fails compilation
160 ;; a call to prevent the other arguments from being optimized away
161 (logand a1 a2 a3 a4 a5 a6 a7 a8 a9)))
163 ;;; bug 192, reported by Einar Floystad Dorum sbcl-devel 2002-08-14,
164 ;;; fixed in sbcl-0.7.6.26: Compiling this function in 0.7.6 caused
165 ;;; the compiler to try to constant-fold DATA-VECTOR-REF, which is OK,
166 ;;; except that there was no non-VOP definition of DATA-VECTOR-REF, so
178 (DECLARE (IGNORABLE S E))
188 (DECLARE (IGNORABLE S E))
189 (WHEN (EQL #\b E) (G910 (1+ I)))))))
196 (DECLARE (IGNORABLE S))
199 (G908 I))) "abcdefg" 0 (length "abcdefg")))
201 ;;; bugs #65, #70, and #109, closed by APD's patch sbcl-devel 2002-08-17
203 ;;; This was "YA code deletion bug" whose symptom was the failure of
205 ;;; (EQ (C::LAMBDA-TAIL-SET C::CALLER)
206 ;;; (C::LAMBDA-TAIL-SET (C::LAMBDA-HOME C::CALLEE)))
208 (defun bug65-1 (termx termy) ; from Carl Witty on submit bugs list, debian.org
210 ((alpha-equal-bound-term-lists (listx listy)
211 (or (and (null listx) (null listy))
213 (let ((bindings-x (bindings-of-bound-term (car listx)))
214 (bindings-y (bindings-of-bound-term (car listy))))
215 (if (and (null bindings-x) (null bindings-y))
216 (alpha-equal-terms (term-of-bound-term (car listx))
217 (term-of-bound-term (car listy)))
218 (and (= (length bindings-x) (length bindings-y))
220 (enter-binding-pairs (bindings-of-bound-term (car listx))
221 (bindings-of-bound-term (car listy)))
222 (alpha-equal-terms (term-of-bound-term (car listx))
223 (term-of-bound-term (car listy)))
224 (exit-binding-pairs (bindings-of-bound-term (car listx))
225 (bindings-of-bound-term (car listy)))))))
226 (alpha-equal-bound-term-lists (cdr listx) (cdr listy)))))
228 (alpha-equal-terms (termx termy)
229 (if (and (variable-p termx)
231 (equal-bindings (id-of-variable-term termx)
232 (id-of-variable-term termy))
233 (and (equal-operators-p (operator-of-term termx) (operator-of-term termy))
234 (alpha-equal-bound-term-lists (bound-terms-of-term termx)
235 (bound-terms-of-term termy))))))
239 (with-variable-invocation (alpha-equal-terms termx termy))))))
240 (defun bug65-2 () ; from Bob Rogers cmucl-imp 1999-07-28
241 ;; Given an FSSP alignment file named by the argument . . .
242 (labels ((get-fssp-char ()
246 ;; Stub body, enough to tickle the bug.
247 (list (read-fssp-char)
249 (defun bug70 ; from David Young cmucl-help 30 Nov 2000
250 (item sequence &key (test #'eql))
251 (labels ((find-item (obj seq test &optional (val nil))
252 (let ((item (first seq)))
255 ((funcall test obj item)
261 (nconc val `(,item))))))))
262 (find-item item sequence test)))
263 (defun bug109 () ; originally from CMU CL bugs collection, reported as
264 ; SBCL bug by MNA 2001-06-25
269 ;; Uncomment and it works
273 ;;; bug 192a, fixed by APD "more strict type checking" patch
274 ;;; (sbcl-devel 2002-08-07)
276 (declare (optimize (speed 0) (safety 3)))
277 ;; Even with bug 192a, this declaration was checked as an assertion.
281 ;; Because of bug 192a, this declaration was trusted without checking.
282 (declare (single-float x))
284 (assert (null (ignore-errors (bug192a nil))))
285 (multiple-value-bind (result error) (ignore-errors (bug192a 100))
286 (assert (null result))
287 (assert (equal (type-error-expected-type error) 'single-float)))
289 ;;; bug 194, fixed in part by APD "more strict type checking" patch
290 ;;; (sbcl-devel 2002-08-07)
292 #+nil ; FIXME: still broken in 0.7.7.19 (after patch)
293 (multiple-value-bind (result error)
294 (ignore-errors (multiple-value-prog1 (progn (the real '(1 2 3)))))
295 (assert (null result))
296 (assert (typep error 'type-error)))
297 #+nil ; FIXME: still broken in 0.7.7.19 (after patch)
298 (multiple-value-bind (result error)
299 (ignore-errors (the real '(1 2 3)))
300 (assert (null result))
301 (assert (typep error 'type-error))))
303 ;;; BUG 48a. and b. (symbol-macrolet handling), fixed by Eric Marsden
304 ;;; and Raymond Toy for CMUCL, fix ported for sbcl-0.7.6.18.
305 (multiple-value-bind (function warnings-p failure-p)
306 (compile nil '(lambda () (symbol-macrolet ((t nil)) t)))
308 (assert (raises-error? (funcall function) program-error)))
309 (multiple-value-bind (function warnings-p failure-p)
312 (symbol-macrolet ((*standard-input* nil))
315 (assert (raises-error? (funcall function) program-error)))
317 BUG 48c, not yet fixed:
318 (multiple-value-bind (function warnings-p failure-p)
319 (compile nil '(lambda () (symbol-macrolet ((s nil)) (declare (special s)) s)))
321 (assert (raises-error? (funcall function) program-error)))
324 ;;; bug 120a: Turned out to be constraining code looking like (if foo
325 ;;; <X> <X>) where <X> was optimized by the compiler to be the exact
326 ;;; same block in both cases, but not turned into (PROGN FOO <X>).
327 ;;; Fixed by APD in sbcl-0.7.7.2, who provided this test:
328 (declaim (inline dont-constrain-if-too-much))
329 (defun dont-constrain-if-too-much (frame up-frame)
330 (declare (optimize (speed 3) (safety 1) (debug 1)))
331 (if (or (not frame) t)
334 (defun dont-constrain-if-too-much-aux (x y)
335 (declare (optimize (speed 3) (safety 1) (debug 1)))
336 (if x t (if y t (dont-constrain-if-too-much x y))))
338 (assert (null (dont-constrain-if-too-much-aux nil nil)))
340 ;;; TYPE-ERROR confusion ca. sbcl-0.7.7.24, reported and fixed by
341 ;;; APD sbcl-devel 2002-09-14
342 (defun exercise-0-7-7-24-bug (x)
343 (declare (integer x))
345 (setf y (the single-float (if (> x 0) x 3f0)))
347 (multiple-value-bind (v e) (ignore-errors (exercise-0-7-7-24-bug 4))
349 (assert (typep e 'type-error)))
350 (assert (equal (exercise-0-7-7-24-bug -4) '(3f0 3f0)))
352 ;;; non-intersecting type declarations were DWIMing in a confusing
353 ;;; fashion until sbcl-0.7.7.28, when APD reported and fixed the
355 (defun non-intersecting-the (x)
357 (setf y (the single-float (the integer x)))
360 (raises-error? (foo 3) type-error)
361 (raises-error? (foo 3f0) type-error)
363 ;;;; tests not in the problem domain, but of the consistency of the
364 ;;;; compiler machinery itself
368 ;;; Hunt for wrong-looking things in fundamental compiler definitions,
369 ;;; and gripe about them.
371 ;;; FIXME: It should be possible to (1) repair the things that this
372 ;;; code gripes about, and then (2) make the code signal errors
373 ;;; instead of just printing complaints to standard output, in order
374 ;;; to prevent the code from later falling back into disrepair.
375 (defun grovel-results (function)
376 (dolist (template (fun-info-templates (info :function :info function)))
377 (when (template-more-results-type template)
378 (format t "~&Template ~A has :MORE results, and translates ~A.~%"
379 (template-name template)
382 (when (eq (template-result-types template) :conditional)
385 (let ((types (template-result-types template))
386 (result-type (fun-type-returns (info :function :type function))))
388 ((values-type-p result-type)
389 (do ((ltypes (append (args-type-required result-type)
390 (args-type-optional result-type))
392 (types types (rest types)))
395 (format t "~&More types than ltypes in ~A, translating ~A.~%"
396 (template-name template)
400 (unless (null ltypes)
401 (format t "~&More ltypes than types in ~A, translating ~A.~%"
402 (template-name template)
405 ((eq result-type (specifier-type nil))
407 (format t "~&Template ~A returns values for function ~A with RESULT-TYPE NIL.~%"
408 (template-name template)
411 ((/= (length types) 1)
412 (format t "~&Template ~A isn't returning 1 value for ~A.~%"
413 (template-name template)
417 (defun identify-suspect-vops (&optional (env (first
418 (last *info-environment*))))
419 (do-info (env :class class :type type :name name :value value)
420 (when (and (eq class :function) (eq type :type))
421 ;; OK, so we have an entry in the INFO database. Now, if ...
422 (let* ((info (info :function :info name))
423 (templates (and info (fun-info-templates info))))
425 ;; ... it has translators
426 (grovel-results name))))))
427 (identify-suspect-vops)
430 (quit :unix-status 104)