3 Bugs can be reported on the help mailing list
4 sbcl-help@lists.sourceforge.net
5 or on the development mailing list
6 sbcl-devel@lists.sourceforge.net
8 Please include enough information in a bug report that someone reading
9 it can reproduce the problem, i.e. don't write
10 Subject: apparent bug in PRINT-OBJECT (or *PRINT-LENGTH*?)
11 PRINT-OBJECT doesn't seem to work with *PRINT-LENGTH*. Is this a bug?
13 Subject: apparent bug in PRINT-OBJECT (or *PRINT-LENGTH*?)
14 In sbcl-1.2.3 running under OpenBSD 4.5 on my Alpha box, when
15 I compile and load the file
16 (DEFSTRUCT (FOO (:PRINT-OBJECT (LAMBDA (X Y)
17 (LET ((*PRINT-LENGTH* 4))
20 then at the command line type
22 the program loops endlessly instead of printing the object.
27 There is also some information on bugs in the manual page and
28 in the TODO file. Eventually more such information may move here.
30 The gaps in the number sequence belong to old bug descriptions which
31 have gone away (typically because they were fixed, but sometimes for
32 other reasons, e.g. because they were moved elsewhere).
35 KNOWN BUGS OF NO SPECIAL CLASS:
38 DEFSTRUCT almost certainly should overwrite the old LAYOUT information
39 instead of just punting when a contradictory structure definition
40 is loaded. As it is, if you redefine DEFSTRUCTs in a way which
41 changes their layout, you probably have to rebuild your entire
42 program, even if you know or guess enough about the internals of
43 SBCL to wager that this (undefined in ANSI) operation would be safe.
46 ANSI specifies that a type mismatch in a structure slot
47 initialization value should not cause a warning.
49 This one might not be fixed for a while because while we're big
50 believers in ANSI compatibility and all, (1) there's no obvious
51 simple way to do it (short of disabling all warnings for type
52 mismatches everywhere), and (2) there's a good portable
53 workaround, and (3) by their own reasoning, it looks as though
54 ANSI may have gotten it wrong. ANSI justifies this specification
56 The restriction against issuing a warning for type mismatches
57 between a slot-initform and the corresponding slot's :TYPE
58 option is necessary because a slot-initform must be specified
59 in order to specify slot options; in some cases, no suitable
61 However, in SBCL (as in CMU CL or, for that matter, any compiler
62 which really understands Common Lisp types) a suitable default
63 does exist, in all cases, because the compiler understands the
64 concept of functions which never return (i.e. has return type NIL).
65 Thus, as a portable workaround, you can use a call to some
66 known-never-to-return function as the default. E.g.
68 (BAR (ERROR "missing :BAR argument")
69 :TYPE SOME-TYPE-TOO-HAIRY-TO-CONSTRUCT-AN-INSTANCE-OF))
71 (DECLAIM (FTYPE (FUNCTION () NIL) MISSING-ARG))
72 (DEFUN REQUIRED-ARG () ; workaround for SBCL non-ANSI slot init typing
73 (ERROR "missing required argument"))
75 (BAR (REQUIRED-ARG) :TYPE TRICKY-TYPE-OF-SOME-SORT)
76 (BLETCH (REQUIRED-ARG) :TYPE TRICKY-TYPE-OF-SOME-SORT)
77 (N-REFS-SO-FAR 0 :TYPE (INTEGER 0)))
78 Such code should compile without complaint and work correctly either
79 on SBCL or on any other completely compliant Common Lisp system.
82 bogus warnings about undefined functions for magic functions like
83 SB!C::%%DEFUN and SB!C::%DEFCONSTANT when cross-compiling files
84 like src/code/float.lisp. Fixing this will probably require
85 straightening out enough bootstrap consistency issues that
86 the cross-compiler can run with *TYPE-SYSTEM-INITIALIZED*.
87 Instead, the cross-compiler runs in a slightly flaky state
88 which is sane enough to compile SBCL itself, but which is
89 also unstable in several ways, including its inability
90 to really grok function declarations.
93 The "compiling top-level form:" output ought to be condensed.
94 Perhaps any number of such consecutive lines ought to turn into a
95 single "compiling top-level forms:" line.
98 The way that the compiler munges types with arguments together
99 with types with no arguments (in e.g. TYPE-EXPAND) leads to
100 weirdness visible to the user:
101 (DEFTYPE FOO () 'FIXNUM)
103 (TYPEP 11 '(FOO)) => T, which seems weird
104 (TYPEP 11 'FIXNUM) => T
105 (TYPEP 11 '(FIXNUM)) signals an error, as it should
106 The situation is complicated by the presence of Common Lisp types
107 like UNSIGNED-BYTE (which can either be used in list form or alone)
108 so I'm not 100% sure that the behavior above is actually illegal.
109 But I'm 90+% sure, and the following related behavior,
111 treating the bare symbol AND as equivalent to '(AND), is specifically
112 forbidden (by the ANSI specification of the AND type).
115 It would be nice if the
117 (during macroexpansion)
118 said what macroexpansion was at fault, e.g.
120 (during macroexpansion of IN-PACKAGE,
121 during macroexpansion of DEFFOO)
124 (SUBTYPEP '(FUNCTION (T BOOLEAN) NIL)
125 '(FUNCTION (FIXNUM FIXNUM) NIL)) => T, T
126 (Also, when this is fixed, we can enable the code in PROCLAIM which
127 checks for incompatible FTYPE redeclarations.)
130 (I *think* this is a bug. It certainly seems like strange behavior. But
131 the ANSI spec is scary, dark, and deep.. -- WHN)
132 (FORMAT NIL "~,1G" 1.4) => "1. "
133 (FORMAT NIL "~3,1G" 1.4) => "1. "
136 from Marco Antoniotti on cmucl-imp mailing list 1 Mar 2000:
138 (setf (find-class 'ccc1) (find-class 'ccc))
139 (defmethod zut ((c ccc1)) 123)
140 DTC's recommended workaround from the mailing list 3 Mar 2000:
141 (setf (pcl::find-class 'ccc1) (pcl::find-class 'ccc))
144 The ANSI spec, in section "22.3.5.2 Tilde Less-Than-Sign: Logical Block",
145 says that an error is signalled if ~W, ~_, ~<...~:>, ~I, or ~:T is used
146 inside "~<..~>" (without the colon modifier on the closing syntax).
147 However, SBCL doesn't do this:
148 * (FORMAT T "~<munge~wegnum~>" 12)
153 Sometimes (SB-EXT:QUIT) fails with
154 Argh! maximum interrupt nesting depth (4096) exceeded, exiting
155 Process inferior-lisp exited abnormally with code 1
156 I haven't noticed a repeatable case of this yet.
159 In some cases the compiler believes type declarations on array
160 elements without checking them, e.g.
161 (DECLAIM (OPTIMIZE (SAFETY 3) (SPEED 1) (SPACE 1)))
164 (DECLARE (TYPE (SIMPLE-ARRAY CONS 1) X))
165 (WHEN (CONSP (AREF X 0))
167 (BAR (VECTOR (MAKE-FOO :A 11 :B 12)))
170 in SBCL 0.6.5 (and also in CMU CL 18b). This does not happen for
171 all cases, e.g. the type assumption *is* checked if the array
172 elements are declared to be of some structure type instead of CONS.
175 The printer doesn't report closures very well. This is true in
179 #<Closure Over Function "DEFUN STRUCTURE-SLOT-ACCESSOR" {134D1A1}>
180 It would be nice to make closures have a settable name slot,
181 and make things like DEFSTRUCT and FLET, which create closures,
182 set helpful values into this slot.
185 And as long as we're wishing, it would be awfully nice if INSPECT could
186 also report on closures, telling about the values of the bound variables.
189 The compiler assumes that any time a function of declared FTYPE
190 doesn't signal an error, its arguments were of the declared type.
191 E.g. compiling and loading
192 (DECLAIM (OPTIMIZE (SAFETY 3)))
193 (DEFUN FACTORIAL (X) (GAMMA (1+ X)))
195 (DECLAIM (FTYPE (FUNCTION (UNSIGNED-BYTE)) FACTORIAL))
197 (COND ((> (FACTORIAL X) 1.0E6)
198 (FORMAT T "too big~%"))
200 (FORMAT T "exactly ~S~%" (FACTORIAL X)))
202 (FORMAT T "approximately ~S~%" (FACTORIAL X)))))
205 will cause the INTEGERP case to be selected, giving bogus output a la
207 This violates the "declarations are assertions" principle.
208 According to the ANSI spec, in the section "System Class FUNCTION",
209 this is a case of "lying to the compiler", but the lying is done
210 by the code which calls FACTORIAL with non-UNSIGNED-BYTE arguments,
211 not by the unexpectedly general definition of FACTORIAL. In any case,
212 "declarations are assertions" means that lying to the compiler should
213 cause an error to be signalled, and should not cause a bogus
214 result to be returned. Thus, the compiler should not assume
215 that arbitrary functions check their argument types. (It might
216 make sense to add another flag (CHECKED?) to DEFKNOWN to
217 identify functions which *do* check their argument types.)
218 (Also, verify that the compiler handles declared function
219 return types as assertions.)
222 DEFMETHOD doesn't check the syntax of &REST argument lists properly,
223 accepting &REST even when it's not followed by an argument name:
224 (DEFMETHOD FOO ((X T) &REST) NIL)
227 TYPEP of VALUES types is sometimes implemented very inefficiently, e.g. in
228 (DEFTYPE INDEXOID () '(INTEGER 0 1000))
230 (DECLARE (TYPE INDEXOID X))
231 (THE (VALUES INDEXOID)
233 where the implementation of the type check in function FOO
234 includes a full call to %TYPEP. There are also some fundamental problems
235 with the interpretation of VALUES types (inherited from CMU CL, and
236 from the ANSI CL standard) as discussed on the cmucl-imp@cons.org
237 mailing list, e.g. in Robert Maclachlan's post of 21 Jun 2000.
240 The definitions of SIGCONTEXT-FLOAT-REGISTER and
241 %SET-SIGCONTEXT-FLOAT-REGISTER in x86-vm.lisp say they're not
242 supported on FreeBSD because the floating point state is not saved,
243 but at least as of FreeBSD 4.0, the floating point state *is* saved,
244 so they could be supported after all. Very likely
245 SIGCONTEXT-FLOATING-POINT-MODES could now be supported, too.
248 (as discussed by Douglas Crosher on the cmucl-imp mailing list ca.
249 Aug. 10, 2000): CMUCL currently interprets 'member as '(member); same
250 issue with 'union, 'and, 'or etc. So even though according to the
251 ANSI spec, bare 'MEMBER, 'AND, and 'OR are not legal types, CMUCL
252 (and now SBCL) interpret them as legal types.
255 ANSI specifies DEFINE-SYMBOL-MACRO, but it's not defined in SBCL.
256 CMU CL added it ca. Aug 13, 2000, after some discussion on the mailing
257 list, and it is probably possible to use substantially the same
258 patches to add it to SBCL.
261 a slew of floating-point-related errors reported by Peter Van Eynde
263 a: (fixed in sbcl-0.6.11.25)
264 b: SBCL's value for LEAST-POSITIVE-SHORT-FLOAT is bogus, and
265 should probably be 1.4012985e-45. In SBCL,
266 (/ LEAST-POSITIVE-SHORT-FLOAT 2) returns a number smaller
267 than LEAST-POSITIVE-SHORT-FLOAT. Similar problems
268 exist for LEAST-NEGATIVE-SHORT-FLOAT, LEAST-POSITIVE-LONG-FLOAT,
269 and LEAST-NEGATIVE-LONG-FLOAT.
270 c: Many expressions generate floating infinity:
275 PVE's regression tests want them to raise errors. SBCL
276 generates the infinities instead, which may or may not be
278 d: (in section12.erg) various forms a la
279 (FLOAT 1 DOUBLE-FLOAT-EPSILON)
280 don't give the right behavior.
283 type safety errors reported by Peter Van Eynde July 25, 2000:
284 a: (COERCE (QUOTE (A B C)) (QUOTE (VECTOR * 4)))
286 In general lengths of array type specifications aren't
287 checked by COERCE, so it fails when the spec is
288 (VECTOR 4), (STRING 2), (SIMPLE-BIT-VECTOR 3), or whatever.
289 b: CONCATENATE has the same problem of not checking the length
290 of specified output array types. MAKE-SEQUENCE and MAP and
291 MERGE also have the same problem.
292 c: (COERCE 'AND 'FUNCTION) returns something related to
293 (MACRO-FUNCTION 'AND), but ANSI says it should raise an error.
294 f: (FLOAT-RADIX 2/3) should signal an error instead of
296 g: (LOAD "*.lsp") should signal FILE-ERROR.
297 h: (MAKE-CONCATENATED-STREAM (MAKE-STRING-OUTPUT-STREAM))
298 should signal TYPE-ERROR.
299 i: MAKE-TWO-WAY-STREAM doesn't check that its arguments can
300 be used for input and output as needed. It should fail with
301 TYPE-ERROR when handed e.g. the results of
302 MAKE-STRING-INPUT-STREAM or MAKE-STRING-OUTPUT-STREAM in
303 the inappropriate positions, but doesn't.
304 j: (PARSE-NAMESTRING (COERCE (LIST #\f #\o #\o (CODE-CHAR 0) #\4 #\8)
306 should probably signal an error instead of making a pathname with
308 k: READ-BYTE is supposed to signal TYPE-ERROR when its argument is
309 not a binary input stream, but instead cheerfully reads from
310 character streams, e.g. (MAKE-STRING-INPUT-STREAM "abc").
313 DEFCLASS bugs reported by Peter Van Eynde July 25, 2000:
314 a: (DEFCLASS FOO () (A B A)) should signal a PROGRAM-ERROR, and
316 b: (DEFCLASS FOO () (A B A) (:DEFAULT-INITARGS X A X B)) should
317 signal a PROGRAM-ERROR, and doesn't.
318 c: (DEFCLASS FOO07 NIL ((A :ALLOCATION :CLASS :ALLOCATION :CLASS))),
319 and other DEFCLASS forms with duplicate specifications in their
320 slots, should signal a PROGRAM-ERROR, and doesn't.
321 d: (DEFGENERIC IF (X)) should signal a PROGRAM-ERROR, but instead
322 causes a COMPILER-ERROR.
325 SYMBOL-MACROLET bugs reported by Peter Van Eynde July 25, 2000:
326 a: (SYMBOL-MACROLET ((T TRUE)) ..) should probably signal
327 PROGRAM-ERROR, but SBCL accepts it instead.
328 b: SYMBOL-MACROLET should refuse to bind something which is
329 declared as a global variable, signalling PROGRAM-ERROR.
330 c: SYMBOL-MACROLET should signal PROGRAM-ERROR if something
331 it binds is declared SPECIAL inside.
334 type system errors reported by Peter Van Eynde July 25, 2000:
335 c: (SUBTYPEP '(INTEGER (0) (0)) 'NIL) dies with nested errors.
336 d: In general, the system doesn't like '(INTEGER (0) (0)) -- it
337 blows up at the level of SPECIFIER-TYPE with
338 "Lower bound (0) is greater than upper bound (0)." Probably
339 SPECIFIER-TYPE should return NIL instead.
340 e: (TYPEP 0 '(COMPLEX (EQL 0)) fails with
341 "Component type for Complex is not numeric: (EQL 0)."
342 This might be easy to fix; the type system already knows
343 that (SUBTYPEP '(EQL 0) 'NUMBER) is true.
344 g: The type system isn't all that smart about relationships
345 between hairy types, as shown in the type.erg test results,
346 e.g. (SUBTYPEP 'CONS '(NOT ATOM)) => NIL, NIL.
349 miscellaneous errors reported by Peter Van Eynde July 25, 2000:
351 (DEFGENERIC FOO02 (X))
352 (DEFMETHOD FOO02 ((X NUMBER)) T)
353 (LET ((M (FIND-METHOD (FUNCTION FOO02)
355 (LIST (FIND-CLASS (QUOTE NUMBER))))))
356 (REMOVE-METHOD (FUNCTION FOO02) M)
357 (DEFGENERIC FOO03 (X))
358 (ADD-METHOD (FUNCTION FOO03) M)))
359 should give an error, but SBCL allows it.
360 b: READ should probably return READER-ERROR, not the bare
361 arithmetic error, when input a la "1/0" or "1e1000" causes
365 It has been reported (e.g. by Peter Van Eynde) that there are
366 several metaobject protocol "errors". (In order to fix them, we might
367 need to document exactly what metaobject protocol specification
368 we're following -- the current code is just inherited from PCL.)
371 The implementation of #'+ returns its single argument without
372 type checking, e.g. (+ "illegal") => "illegal".
375 Attempting to use COMPILE on something defined by DEFMACRO fails:
376 (DEFMACRO FOO (X) (CONS X X))
378 Error in function C::GET-LAMBDA-TO-COMPILE:
379 #<Closure Over Function "DEFUN (SETF MACRO-FUNCTION)" {480E21B1}> was defined in a non-null environment.
382 (SUBTYPEP '(AND ZILCH INTEGER) 'ZILCH) => NIL, NIL
383 Note: I looked into fixing this in 0.6.11.15, but gave up. The
384 problem seems to be that there are two relevant type methods for
385 the subtypep operation, HAIRY :COMPLEX-SUBTYPEP-ARG2 and
386 INTERSECTION :COMPLEX-SUBTYPEP-ARG1, and only the first is
387 called. This could be fixed, but type dispatch is messy and
388 confusing enough already, I don't want to complicate it further.
389 Perhaps someday we can make CLOS cross-compiled (instead of compiled
390 after bootstrapping) so that we don't need to have the type system
391 available before CLOS, and then we can rewrite the type methods to
392 CLOS methods, and then expressing the solutions to stuff like this
393 should become much more straightforward. -- WHN 2001-03-14
396 The debugger LIST-LOCATIONS command doesn't work properly.
399 Compiling and loading
400 (DEFUN FAIL (X) (THROW 'FAIL-TAG X))
402 then requesting a BACKTRACE at the debugger prompt gives no information
403 about where in the user program the problem occurred.
406 The compiler is supposed to do type inference well enough that
409 ((SIMPLE-ARRAY SINGLE-FLOAT)
411 (DECLARE (TYPE (SIMPLE-ARRAY SINGLE-FLOAT) X))
414 is redundant. However, as reported by Juan Jose Garcia Ripoll for
415 CMU CL, it sometimes doesn't. Adding declarations is a pretty good
416 workaround for the problem for now, but can't be done by the TYPECASE
417 macros themselves, since it's too hard for the macro to detect
418 assignments to the variable within the clause.
419 Note: The compiler *is* smart enough to do the type inference in
420 many cases. This case, derived from a couple of MACROEXPAND-1
421 calls on Ripoll's original test case,
423 (DECLARE (OPTIMIZE SPEED (SAFETY 0)))
424 (COND ((TYPEP A '(SIMPLE-ARRAY SINGLE-FLOAT)) NIL
425 (LET ((LENGTH (ARRAY-TOTAL-SIZE A)))
426 (LET ((I 0) (G2554 LENGTH))
427 (DECLARE (TYPE REAL G2554) (TYPE REAL I))
430 (WHEN (>= I G2554) (GO SB-LOOP::END-LOOP))
431 (SETF (ROW-MAJOR-AREF A I) (- (ROW-MAJOR-AREF A I)))
432 (GO SB-LOOP::NEXT-LOOP)
433 SB-LOOP::END-LOOP))))))
434 demonstrates the problem; but the problem goes away if the TAGBODY
435 and GO forms are removed (leaving the SETF in ordinary, non-looping
436 code), or if the TAGBODY and GO forms are retained, but the
437 assigned value becomes 0.0 instead of (- (ROW-MAJOR-AREF A I)).
440 Paul Werkowski wrote on cmucl-imp@cons.org 2000-11-15
441 I am looking into this problem that showed up on the cmucl-help
442 list. It seems to me that the "implementation specific environment
443 hacking functions" found in pcl/walker.lisp are completely messed
444 up. The good thing is that they appear to be barely used within
445 PCL and the munged environment object is passed to cmucl only
446 in calls to macroexpand-1, which is probably why this case fails.
447 SBCL uses essentially the same code, so if the environment hacking
448 is screwed up, it affects us too.
451 Using the pretty-printer from the command prompt gives funny
452 results, apparently because the pretty-printer doesn't know
453 about user's command input, including the user's carriage return
454 that the user, and therefore the pretty-printer thinks that
455 the new output block should start indented 2 or more characters
456 rightward of the correct location.
459 (probably related to bug #70; maybe related to bug #109)
460 As reported by Carl Witty on submit@bugs.debian.org 1999-05-08,
462 (in-package "CL-USER")
463 (defun equal-terms (termx termy)
465 ((alpha-equal-bound-term-lists (listx listy)
466 (or (and (null listx) (null listy))
468 (let ((bindings-x (bindings-of-bound-term (car listx)))
469 (bindings-y (bindings-of-bound-term (car listy))))
470 (if (and (null bindings-x) (null bindings-y))
471 (alpha-equal-terms (term-of-bound-term (car listx))
472 (term-of-bound-term (car listy)))
473 (and (= (length bindings-x) (length bindings-y))
475 (enter-binding-pairs (bindings-of-bound-term (car listx))
476 (bindings-of-bound-term (car listy)))
477 (alpha-equal-terms (term-of-bound-term (car listx))
478 (term-of-bound-term (car listy)))
479 (exit-binding-pairs (bindings-of-bound-term (car listx))
480 (bindings-of-bound-term (car listy)))))))
481 (alpha-equal-bound-term-lists (cdr listx) (cdr listy)))))
483 (alpha-equal-terms (termx termy)
484 (if (and (variable-p termx)
486 (equal-bindings (id-of-variable-term termx)
487 (id-of-variable-term termy))
488 (and (equal-operators-p (operator-of-term termx) (operator-of-term termy))
489 (alpha-equal-bound-term-lists (bound-terms-of-term termx)
490 (bound-terms-of-term termy))))))
494 (with-variable-invocation (alpha-equal-terms termx termy))))))
495 causes an assertion failure
496 The assertion (EQ (C::LAMBDA-TAIL-SET C::CALLER)
497 (C::LAMBDA-TAIL-SET (C::LAMBDA-HOME C::CALLEE))) failed.
499 Bob Rogers reports (1999-07-28 on cmucl-imp@cons.org) a smaller test
500 case with the same problem:
501 (defun parse-fssp-alignment ()
502 ;; Given an FSSP alignment file named by the argument . . .
503 (labels ((get-fssp-char ()
507 ;; Stub body, enough to tickle the bug.
508 (list (read-fssp-char)
512 ANSI specifies that the RESULT-TYPE argument of CONCATENATE must be
513 a subtype of SEQUENCE, but CONCATENATE doesn't check this properly:
514 (CONCATENATE 'SIMPLE-ARRAY #(1 2) '(3)) => #(1 2 3)
515 This also leads to funny behavior when derived type specifiers
516 are used, as originally reported by Milan Zamazal for CMU CL (on the
517 Debian bugs mailing list (?) 2000-02-27), then reported by Martin
518 Atzmueller for SBCL (2000-10-01 on sbcl-devel@lists.sourceforge.net):
519 (DEFTYPE FOO () 'SIMPLE-ARRAY)
520 (CONCATENATE 'FOO #(1 2) '(3))
521 => #<ARRAY-TYPE SIMPLE-ARRAY> is a bad type specifier for
523 The derived type specifier FOO should act the same way as the
524 built-in type SIMPLE-ARRAY here, but it doesn't. That problem
525 doesn't seem to exist for sequence types:
526 (DEFTYPE BAR () 'SIMPLE-VECTOR)
527 (CONCATENATE 'BAR #(1 2) '(3)) => #(1 2 3)
530 As reported by Winton Davies on a CMU CL mailing list 2000-01-10,
531 and reported for SBCL by Martin Atzmueller 2000-10-20: (TRACE GETHASH)
532 crashes SBCL. In general tracing anything which is used in the
533 implementation of TRACE is likely to have the same problem.
536 As reported by Daniel Solaz on cmucl-help@cons.org 2000-11-23,
537 SXHASH returns the same value for all non-STRUCTURE-OBJECT instances,
538 notably including all PCL instances. There's a limit to how much
539 SXHASH can do to return unique values for instances, but at least
540 it should probably look at the class name, the way that it does
541 for STRUCTURE-OBJECTs.
544 As reported by Martin Atzmueller on the sbcl-devel list 2000-11-22,
545 > There remains one issue, that is a bug in SBCL:
546 > According to my interpretation of the spec, the ":" and "@" modifiers
547 > should appear _after_ the comma-seperated arguments.
548 > Well, SBCL (and CMUCL for that matter) accept
549 > (ASSERT (STRING= (FORMAT NIL "~:8D" 1) " 1"))
550 > where the correct way (IMHO) should be
551 > (ASSERT (STRING= (FORMAT NIL "~8:D" 1) " 1"))
552 Probably SBCL should stop accepting the "~:8D"-style format arguments,
553 or at least issue a warning.
556 (probably related to bug #65; maybe related to bug #109)
557 The compiler doesn't like &OPTIONAL arguments in LABELS and FLET
559 (DEFUN FIND-BEFORE (ITEM SEQUENCE &KEY (TEST #'EQL))
560 (LABELS ((FIND-ITEM (OBJ SEQ TEST &OPTIONAL (VAL NIL))
561 (LET ((ITEM (FIRST SEQ)))
564 ((FUNCALL TEST OBJ ITEM)
567 (FIND-ITEM OBJ (REST SEQ) TEST (NCONC VAL `(,ITEM))))))))
568 (FIND-ITEM ITEM SEQUENCE TEST)))
569 from David Young's bug report on cmucl-help@cons.org 30 Nov 2000
570 causes sbcl-0.6.9 to fail with
571 error in function SB-KERNEL:ASSERT-ERROR:
572 The assertion (EQ (SB-C::LAMBDA-TAIL-SET SB-C::CALLER)
573 (SB-C::LAMBDA-TAIL-SET
574 (SB-C::LAMBDA-HOME SB-C::CALLEE))) failed.
577 (DECLAIM (OPTIMIZE ..)) doesn't work. E.g. even after
578 (DECLAIM (OPTIMIZE (SPEED 3))), things are still optimized with
579 the previous SPEED policy. This bug will probably get fixed in
580 0.6.9.x in a general cleanup of optimization policy.
583 (DECLAIM (OPTIMIZE ..)) doesn't work properly inside LOCALLY forms.
586 As reported by Martin Atzmueller on sbcl-devel 26 Dec 2000,
587 ANSI says that WITH-OUTPUT-TO-STRING should have a keyword
588 :ELEMENT-TYPE, but in sbcl-0.6.9 this is not defined for
589 WITH-OUTPUT-TO-STRING.
592 ANSI says in one place that type declarations can be abbreviated even
593 when the type name is not a symbol, e.g.
594 (DECLAIM ((VECTOR T) *FOOVECTOR*))
595 SBCL doesn't support this. But ANSI says in another place that this
596 isn't allowed. So it's not clear this is a bug after all. (See the
597 e-mail on cmucl-help@cons.org on 2001-01-16 and 2001-01-17 from WHN
601 as pointed out by Dan Barlow on sbcl-devel 2000-07-02:
602 The PICK-TEMPORARY-FILE-NAME utility used by LOAD-FOREIGN uses
603 an easily guessable temporary filename in a way which might open
604 applications using LOAD-FOREIGN to hijacking by malicious users
605 on the same machine. Incantations for doing this safely are
606 floating around the net in various "how to write secure programs
607 despite Unix" documents, and it would be good to (1) fix this in
608 LOAD-FOREIGN, and (2) hunt for any other code which uses temporary
609 files and make it share the same new safe logic.
612 Functions are assigned names based on the context in which they're
613 defined. This is less than ideal for the functions which are
614 used to implement CLOS methods. E.g. the output of
615 (DESCRIBE 'PRINT-OBJECT) lists functions like
616 #<FUNCTION "DEF!STRUCT (TRACE-INFO (:MAKE-LOAD-FORM-FUN SB-KERNEL:JUST-DUMP-IT-NORMALLY) (:PRINT-OBJECT #))" {1020E49}>
618 #<FUNCTION "MACROLET ((FORCE-DELAYED-DEF!METHODS NIL #))" {1242871}>
619 It would be better if these functions' names always identified
620 them as methods, and identified their generic functions and
624 RANDOM-INTEGER-EXTRA-BITS=10 may not be large enough for the RANDOM
625 RNG to be high quality near RANDOM-FIXNUM-MAX; it looks as though
626 the mean of the distribution can be systematically O(0.1%) wrong.
627 Just increasing R-I-E-B is probably not a good solution, since
628 it would decrease efficiency more than is probably necessary. Perhaps
629 using some sort of accept/reject method would be better.
632 (SUBTYPEP '(SATISFIES SOME-UNDEFINED-FUN) NIL)=>NIL,T (should be NIL,NIL)
635 Internally the compiler sometimes evaluates
636 (sb-kernel:type/= (specifier-type '*) (specifier-type t))
637 (I stumbled across this when I added an
638 (assert (not (eq type1 *wild-type*)))
639 in the NAMED :SIMPLE-= type method.) '* isn't really a type, and
640 in a type context should probably be translated to T, and so it's
641 probably to ask whether it's equal to the T type and then (using the
642 EQ type comparison in the NAMED :SIMPLE-= type method) return NIL.
643 (I haven't tried to investigate this bug enough to guess whether
644 there might be any user-level symptoms.)
647 a latent cross-compilation/bootstrapping bug: The cross-compilation
648 host's CL:CHAR-CODE-LIMIT is used in target code in readtable.lisp
649 and possibly elsewhere. Instead, we should use the target system's
650 CHAR-CODE-LIMIT. This will probably cause problems if we try to
651 bootstrap on a system which uses a different value of CHAR-CODE-LIMIT
655 (subtypep '(or (integer -1 1)
659 (integer -1 1))) => NIL,T
660 An analogous problem with SINGLE-FLOAT and REAL types was fixed in
661 sbcl-0.6.11.22, but some peculiarites of the RATIO type make it
662 awkward to generalize the fix to INTEGER and RATIONAL. It's not
663 clear what's the best fix. (See the "bug in type handling" discussion
664 on cmucl-imp ca. 2001-03-22 and ca. 2001-02-12.)
667 In sbcl-0.6.11.26, (COMPILE 'IN-HOST-COMPILATION-MODE) in
668 src/cold/shared.lisp doesn't correctly translate the
670 (defun in-host-compilation-mode (fn)
671 (let ((*features* (cons :sb-xc-host *features*))
672 ;; the CROSS-FLOAT-INFINITY-KLUDGE, as documented in
673 ;; base-target-features.lisp-expr:
674 (*shebang-features* (set-difference *shebang-features*
675 '(:sb-propagate-float-type
676 :sb-propagate-fun-type))))
677 (with-additional-nickname ("SB-XC" "SB!XC")
679 No error is reported by the compiler, but when the function is executed,
681 TYPE-ERROR in SB-KERNEL::OBJECT-NOT-TYPE-ERROR-HANDLER:
682 (:LINUX :X86 :IEEE-FLOATING-POINT :SB-CONSTRAIN-FLOAT-TYPE :SB-TEST
683 :SB-INTERPRETER :SB-DOC :UNIX ...) is not of type SYMBOL.
686 Inconsistencies between derived and declared VALUES return types for
687 DEFUN aren't checked very well. E.g. the logic which successfully
688 catches problems like
689 (declaim (ftype (function (fixnum) float) foo))
691 (declare (type integer x))
692 (values x)) ; wrong return type, detected, gives warning, good!
694 (declaim (ftype (function (t) (values t t)) bar))
696 (values x)) ; wrong number of return values, no warning, bad!
697 The cause of this is seems to be that (1) the internal function
698 VALUES-TYPES-EQUAL-OR-INTERSECT used to make the check handles its
699 arguments symmetrically, and (2) when the type checking code was
700 written back when when SBCL's code was still CMU CL, the intent
702 (declaim (ftype (function (t) t) bar))
704 (values x x)) ; wrong number of return values; should give warning?
705 not be warned for, because a two-valued return value is considered
706 to be compatible with callers who expects a single value to be
707 returned. That intent is probably not appropriate for modern ANSI
708 Common Lisp, but fixing this might be complicated because of other
709 divergences between auld-style and new-style handling of
710 multiple-VALUES types. (Some issues related to this were discussed
711 on cmucl-imp at some length sometime in 2000.)
714 The facility for dumping a running Lisp image to disk gets confused
715 when run without the PURIFY option, and creates an unnecessarily large
716 core file (apparently representing memory usage up to the previous
717 high-water mark). Moreover, when the file is loaded, it confuses the
718 GC, so that thereafter memory usage can never be reduced below that
722 The TRACE facility can't be used on some kinds of functions.
723 (Basically, the breakpoint facility was incompletely implemented
724 in the X86 port of CMU CL, and hasn't been fixed in SBCL.)
727 In sbcl-0.6.11.41 (and in all earlier SBCL, and in CMU
728 CL), out-of-line structure slot setters are horribly inefficient
729 whenever the type of the slot is declared, because out-of-line
730 structure slot setters are implemented as closures to save space,
731 so the compiler doesn't compile the type test into code, but
732 instead just saves the type in a lexical closure and interprets it
734 A proper solution involves deciding whether it's really worth
735 saving space by implementing structure slot accessors as closures.
736 (If it's not worth it, the problem vanishes automatically. If it
737 is worth it, there are hacks we could use to force type tests to
738 be compiled anyway, and even shared. E.g. we could implement
739 an EQUAL hash table mapping from types to compiled type tests,
740 and save the appropriate compiled type test as part of each lexical
741 closure; or we could make the lexical closures be placeholders
742 which overwrite their old definition as a lexical closure with
743 a new compiled definition the first time that they're called.)
744 As a workaround for the problem, #'(SETF FOO) expressions can
745 be replaced with (EFFICIENT-SETF-FUNCTION FOO), where
746 (defmacro efficient-setf-function (place-function-name)
747 (or #+sbcl (and (sb-impl::info :function :accessor-for place-function-name)
748 ;; a workaround for the problem, encouraging the
749 ;; inline expansion of the structure accessor, so
750 ;; that the compiler can optimize its type test
751 (let ((new-value (gensym "NEW-VALUE-"))
752 (structure-value (gensym "STRUCTURE-VALUE-")))
753 `(lambda (,new-value ,structure-value)
754 (setf (,place-function-name ,structure-value)
756 ;; no problem, can just use the ordinary expansion
757 `(function (setf ,place-function-name))))
760 DESCRIBE interacts poorly with *PRINT-CIRCLE*, e.g. the output from
761 (let ((*print-circle* t)) (describe (make-hash-table)))
763 #<HASH-TABLE :TEST EQL :COUNT 0 {90BBFC5}> is an . (EQL)
765 Its REHASH-SIZE is 1.5. Its REHASH-THRESHOLD is . (1.0)
766 It holds 0 key/value pairs.
767 where the ". (EQL)" and ". (1.0)" substrings are screwups.
768 (This is likely a pretty-printer problem which happens to
769 be exercised by DESCRIBE, not actually a DESCRIBE problem.)
772 There's apparently a bug in CEILING optimization which caused
773 Douglas Crosher to patch the CMU CL version. Martin Atzmueller
774 applied the patches to SBCL and they didn't seem to cause problems
775 (as reported sbcl-devel 2001-05-04). However, since the patches
776 modify nontrivial code which was apparently written incorrectly
777 the first time around, until regression tests are written I'm not
778 comfortable merging the patches in the CVS version of SBCL.
781 As reported by Arthur Lemmens sbcl-devel 2001-05-05, ANSI
782 requires that SYMBOL-MACROLET refuse to rebind special variables,
783 but SBCL doesn't do this. (Also as reported by AL in the same
784 message, SBCL depended on this nonconforming behavior to build
785 itself, because of the way that **CURRENT-SEGMENT** was implemented.
786 As of sbcl-0.6.12.x, this dependence on the nonconforming behavior
787 has been fixed, but the nonconforming behavior remains.)
790 (DESCRIBE 'SB-ALIEN:DEF-ALIEN-TYPE) reports the macro argument list
794 in #<PACKAGE "SB-ALIEN">.
795 Macro-function: #<FUNCTION "DEF!MACRO DEF-ALIEN-TYPE" {19F4A39}>
796 Macro arguments: (#:whole-470 #:environment-471)
797 On Sat, May 26, 2001 09:45:57 AM CDT it was compiled from:
798 /usr/stuff/sbcl/src/code/host-alieneval.lisp
799 Created: Monday, March 12, 2001 07:47:43 AM CST
802 (DESCRIBE 'STREAM-READ-BYTE)
805 (reported by Eric Marsden on cmucl-imp 2001-06-15)
806 (and APD pointed out on sbcl-devel 2001-12-29 that it's the same
810 (TIME (ROOM T)) reports more than 200 Mbytes consed even for
811 a clean, just-started SBCL system. And it seems to be right:
812 (ROOM T) can bring a small computer to its knees for a *long*
813 time trying to GC afterwards. Surely there's some more economical
814 way to implement (ROOM T).
817 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
819 ;;; This file fails to compile.
820 ;;; Maybe this bug is related to bugs #65, #70 in the BUGS file.
821 (in-package :cl-user)
827 ;; Uncomment and it works
830 In SBCL 0.6.12.42, the problem is
831 internal error, failed AVER:
832 "(COMMON-LISP:EQ (SB!C::LAMBDA-TAIL-SET SB!C::CALLER)
833 (SB!C::LAMBDA-TAIL-SET (SB!C::LAMBDA-HOME SB!C::CALLEE)))"
836 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
838 ;;; The compiler is flushing the argument type test, and the default
839 ;;; case in the cond, so that calling with say a fixnum 0 causes a
841 (declaim (optimize (safety 2) (speed 3)))
843 (declare (type (or string stream) x))
844 (cond ((typep x 'string) 'string)
845 ((typep x 'stream) 'stream)
848 The symptom in sbcl-0.6.12.42 on OpenBSD is actually (TST 0)=>STREAM
849 (not the SIGBUS reported in the comment) but that's broken too;
850 type declarations are supposed to be treated as assertions unless
851 SAFETY 0, so we should be getting a TYPE-ERROR.
854 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
856 (in-package :cl-user)
857 ;;; Produces an assertion failures when compiled.
859 (declare (type (or (function (t) t) null) z))
860 (let ((z (or z #'identity)))
861 (declare (type (function (t) t) z))
863 The error in sbcl-0.6.12.42 is
864 internal error, failed AVER:
865 "(COMMON-LISP:NOT (COMMON-LISP:EQ SB!C::CHECK COMMON-LISP:T))"
868 reported by Martin Atzmueller 2001-06-25; taken from CMU CL bugs
869 collection; apparently originally reported by Bruno Haible
870 (in-package :cl-user)
871 ;;; From: Bruno Haible
872 ;;; Subject: scope of SPECIAL declarations
873 ;;; It seems CMUCL has a bug relating to the scope of SPECIAL
874 ;;; declarations. I observe this with "CMU Common Lisp 18a x86-linux
877 (declare (special x))
880 (declare (special x)) y)))
881 ;;; Gives: 0 (this should return 1 according to CLHS)
883 (declare (special x))
886 (declare (special x)) y)))
887 ;;; Gives: 1 (correct).
888 The reported results match what we get from the interpreter
892 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
894 (in-package :cl-user)
895 ;;; From: David Gadbois <gadbois@cyc.com>
897 ;;; Logical pathnames aren't externalizable.
899 (let ((tempfile "/tmp/test.lisp"))
900 (setf (logical-pathname-translations "XXX")
901 '(("XXX:**;*.*" "/tmp/**/*.*")))
902 (with-open-file (out tempfile :direction :output)
903 (write-string "(defvar *path* #P\"XXX:XXX;FOO.LISP\")" out))
904 (compile-file tempfile))
905 The error message in sbcl-0.6.12.42 is
907 ; (while making load form for #<SB-IMPL::LOGICAL-HOST "XXX">)
908 ; A logical host can't be dumped as a constant: #<SB-IMPL::LOGICAL-HOST "XXX">
911 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
913 (in-package :cl-user)
914 ;;; The following invokes a compiler error.
915 (declaim (optimize (speed 2) (debug 3)))
918 (unwind-protect nil)))
922 The error message in sbcl-0.6.12.42 is
923 internal error, failed AVER:
924 "(COMMON-LISP:EQ (SB!C::TN-ENVIRONMENT SB!C:TN) SB!C::TN-ENV)"
927 When the compiler inline expands functions, it may be that different
928 kinds of return values are generated from different code branches.
929 E.g. an inline expansion of POSITION generates integer results
930 from one branch, and NIL results from another. When that inline
931 expansion is used in a context where only one of those results
934 (aref *a1* (position x *a2*)))
935 and the compiler can't prove that the unacceptable branch is
936 never taken, then bogus type mismatch warnings can be generated.
937 If you need to suppress the type mismatch warnings, you can
938 suppress the inline expansion,
940 #+sbcl (declare (notinline position)) ; to suppress bug 117 bogowarnings
941 (aref *a1* (position x *a2*)))
942 or, sometimes, suppress them by declaring the result to be of an
945 (aref *a1* (the integer (position x *a2*))))
947 This is not a new compiler problem in 0.7.0, but the new compiler
948 transforms for FIND, POSITION, FIND-IF, and POSITION-IF make it
949 more conspicuous. If you don't need performance from these functions,
950 and the bogus warnings are a nuisance for you, you can return to
951 your pre-0.7.0 state of grace with
952 #+sbcl (declaim (notinline find position find-if position-if)) ; bug 117..
955 as reported by Eric Marsden on cmucl-imp@cons.org 2001-08-14:
956 (= (FLOAT 1 DOUBLE-FLOAT-EPSILON)
957 (+ (FLOAT 1 DOUBLE-FLOAT-EPSILON) DOUBLE-FLOAT-EPSILON)) => T
958 when of course it should be NIL. (He says it only fails for X86,
959 not SPARC; dunno about Alpha.)
961 Also, "the same problem exists for LONG-FLOAT-EPSILON,
962 DOUBLE-FLOAT-NEGATIVE-EPSILON, LONG-FLOAT-NEGATIVE-EPSILON (though
963 for the -negative- the + is replaced by a - in the test)."
965 Raymond Toy comments that this is tricky on the X86 since its FPU
966 uses 80-bit precision internally.
969 The compiler incorrectly figures the return type of
970 (DEFUN FOO (FRAME UP-FRAME)
977 This problem exists in CMU CL 18c too. When I reported it on
978 cmucl-imp@cons.org, Raymond Toy replied 23 Aug 2001 with
979 a partial explanation, but no fix has been found yet.
982 Even in sbcl-0.pre7.x, which is supposed to be free of the old
983 non-ANSI behavior of treating the function return type inferred
984 from the current function definition as a declaration of the
985 return type from any function of that name, the return type of NIL
986 is attached to FOO in 120a above, and used to optimize code which
990 There was some sort of screwup in handling of
991 (IF (NOT (IGNORE-ERRORS ..))). E.g.
993 (if (not (ignore-errors
994 (make-pathname :host "foo" :directory "!bla" :name "bar")))
996 (error "notunlessnot")))
997 The (NOT (IGNORE-ERRORS ..)) form evaluates to T, so this should be
998 printing "ok", but instead it's going to the ERROR. This problem
999 seems to've been introduced by MNA's HANDLER-CASE patch (sbcl-devel
1000 2001-07-17) and as a workaround (put in sbcl-0.pre7.14.flaky4.12)
1001 I reverted back to the old weird HANDLER-CASE code. However, I
1002 think the problem looks like a compiler bug in handling RETURN-FROM,
1003 so I left the MNA-patched code in HANDLER-CASE (suppressed with
1004 #+NIL) and I'd like to go back to see whether this really is
1005 a compiler bug before I delete this BUGS entry.
1008 The *USE-IMPLEMENTATION-TYPES* hack causes bugs, particularly
1009 (IN-PACKAGE :SB-KERNEL)
1010 (TYPE= (SPECIFIER-TYPE '(VECTOR T))
1011 (SPECIFIER-TYPE '(VECTOR UNDEFTYPE)))
1012 Then because of this, the compiler bogusly optimizes
1013 (TYPEP #(11) '(SIMPLE-ARRAY UNDEF-TYPE 1))
1014 to T. Unfortunately, just setting *USE-IMPLEMENTATION-TYPES* to
1015 NIL around sbcl-0.pre7.14.flaky4.12 didn't work: the compiler complained
1016 about type mismatches (probably harmlessly, another instance of bug 117);
1017 and then cold init died with a segmentation fault.
1020 As of version 0.pre7.14, SBCL's implementation of MACROLET makes
1021 the entire lexical environment at the point of MACROLET available
1022 in the bodies of the macroexpander functions. In particular, it
1023 allows the function bodies (which run at compile time) to try to
1024 access lexical variables (which are only defined at runtime).
1025 It doesn't even issue a warning, which is bad.
1027 The SBCL behavior arguably conforms to the ANSI spec (since the
1028 spec says that the behavior is undefined, ergo anything conforms).
1029 However, it would be better to issue a compile-time error.
1030 Unfortunately I (WHN) don't see any simple way to detect this
1031 condition in order to issue such an error, so for the meantime
1032 SBCL just does this weird broken "conforming" thing.
1034 The ANSI standard says, in the definition of the special operator
1036 The macro-expansion functions defined by MACROLET are defined
1037 in the lexical environment in which the MACROLET form appears.
1038 Declarations and MACROLET and SYMBOL-MACROLET definitions affect
1039 the local macro definitions in a MACROLET, but the consequences
1040 are undefined if the local macro definitions reference any
1041 local variable or function bindings that are visible in that
1042 lexical environment.
1043 Then it seems to contradict itself by giving the example
1045 (macrolet ((fudge (z)
1046 ;The parameters x and flag are not accessible
1047 ; at this point; a reference to flag would be to
1048 ; the global variable of that name.
1049 ` (if flag (* ,z ,z) ,z)))
1050 ;The parameters x and flag are accessible here.
1054 The comment "a reference to flag would be to the global variable
1055 of the same name" sounds like good behavior for the system to have.
1056 but actual specification quoted above says that the actual behavior
1060 (as reported by Gabe Garza on cmucl-help 2001-09-21)
1062 (defun test-pred (x y)
1066 (func (lambda () x)))
1067 (print (eq func func))
1068 (print (test-pred func func))
1069 (delete func (list func))))
1070 Now calling (TEST-CASE) gives output
1073 (#<FUNCTION {500A9EF9}>)
1074 Evidently Python thinks of the lambda as a code transformation so
1075 much that it forgets that it's also an object.
1078 (fixed in 0.pre7.41)
1081 The DEFSTRUCT section of the ANSI spec, in the :CONC-NAME section,
1082 specifies a precedence rule for name collisions between slot accessors of
1083 structure classes related by inheritance. As of 0.7.0, SBCL still
1087 insufficient syntax checking in MACROLET:
1089 (macrolet ((defmacro bar (z) `(+ z z)))
1091 shouldn't compile without error (because of the extra DEFMACRO symbol).
1094 reported by Alexey Dejneka on sbcl-devel 2001-11-03
1096 "Return X if X is a non-negative integer."
1097 (let ((step (lambda (%funcall)
1100 (t (1+ (funcall %funcall (1- n)))))))))
1103 (funcall step (lambda (n)
1104 (funcall (funcall a a) n))))
1106 (funcall step (lambda (n)
1107 (funcall (funcall a a) n)))))
1109 This function returns its argument. But after removing percents it
1110 does not work: "Result of (1- n) is not a function".
1113 As of sbcl-0.pre7.86.flaky7.3, the cross-compiler, and probably
1114 the CL:COMPILE function (which is based on the same %COMPILE
1115 mechanism) get confused by
1117 (labels ((sxhash-number (x)
1119 (fixnum (sxhash x)) ; through DEFTRANSFORM
1120 (integer (sb!bignum:sxhash-bignum x))
1121 (single-float (sxhash x)) ; through DEFTRANSFORM
1122 (double-float (sxhash x)) ; through DEFTRANSFORM
1123 #!+long-float (long-float (error "stub: no LONG-FLOAT"))
1124 (ratio (let ((result 127810327))
1125 (declare (type fixnum result))
1126 (mixf result (sxhash-number (numerator x)))
1127 (mixf result (sxhash-number (denominator x)))
1129 (complex (let ((result 535698211))
1130 (declare (type fixnum result))
1131 (mixf result (sxhash-number (realpart x)))
1132 (mixf result (sxhash-number (imagpart x)))
1134 (sxhash-recurse (x &optional (depthoid +max-hash-depthoid+))
1135 (declare (type index depthoid))
1138 (if (plusp depthoid)
1139 (mix (sxhash-recurse (car x) (1- depthoid))
1140 (sxhash-recurse (cdr x) (1- depthoid)))
1143 (if (typep x 'structure-object)
1145 (sxhash ; through DEFTRANSFORM
1146 (class-name (layout-class (%instance-layout x)))))
1148 (symbol (sxhash x)) ; through DEFTRANSFORM
1149 (number (sxhash-number x))
1152 (simple-string (sxhash x)) ; through DEFTRANSFORM
1153 (string (%sxhash-substring x))
1154 (bit-vector (let ((result 410823708))
1155 (declare (type fixnum result))
1156 (dotimes (i (min depthoid (length x)))
1157 (mixf result (aref x i)))
1159 (t (logxor 191020317 (sxhash (array-rank x))))))
1162 (sxhash (char-code x)))) ; through DEFTRANSFORM
1164 (sxhash-recurse x)))
1165 complaining "function called with two arguments, but wants exactly
1166 one" about SXHASH-RECURSE. (This might not be strictly a new bug,
1167 since IIRC post-fork CMU CL has also had problems with &OPTIONAL
1168 arguments in FLET/LABELS: it might be an old Python bug which is
1169 only exercised by the new arrangement of the SBCL compiler.)
1173 (DEFUN FOO () (CATCH 0 (PRINT 1331)))
1175 #<SB-C:TN '0!1> is not valid as the second argument to VOP:
1176 SB-C:MAKE-CATCH-BLOCK,
1177 since the TN's primitive type SB-VM::POSITIVE-FIXNUM doesn't allow
1178 any of the SCs allowed by the operand restriction:
1179 (SB-VM::DESCRIPTOR-REG)
1180 The (CATCH 0 ...) construct is bad style (because of unportability
1181 of EQ testing of numbers) but it is legal, and shouldn't cause an
1182 internal compiler error. (This error occurs in sbcl-0.6.13 and in
1183 0.pre7.86.flaky7.14.)
1186 Trying to compile something like
1187 (sb!alien:def-alien-routine "breakpoint_remove" sb!c-call:void
1188 (code-obj sb!c-call:unsigned-long)
1189 (pc-offset sb!c-call:int)
1190 (old-inst sb!c-call:unsigned-long))
1191 in SBCL-0.pre7.86.flaky7.22 after warm init fails with an error
1192 cannot use values types here
1193 probably because the SB-C-CALL:VOID type gets translated to (VALUES).
1194 It should be valid to use VOID for a function return type, so perhaps
1195 instead of calling SPECIFIER-TYPE (which excludes all VALUES types
1196 automatically) we should call VALUES-SPECIFIER-TYPE and handle VALUES
1197 types manually, allowing the special case (VALUES) but still excluding
1198 all more-complex VALUES types.
1201 Ideally, uninterning a symbol would allow it, and its associated
1202 FDEFINITION and PROCLAIM data, to be reclaimed by the GC. However,
1203 at least as of sbcl-0.7.0, this isn't the case. Information about
1204 FDEFINITIONs and PROCLAIMed properties is stored in globaldb.lisp
1205 essentially in ordinary (non-weak) hash tables keyed by symbols.
1206 Thus, once a system has an entry in this system, it tends to live
1207 forever, even when it is uninterned and all other references to it
1211 (reported by Arnaud Rouanet on cmucl-imp 2001-12-18)
1212 (defmethod foo ((x integer))
1214 (defmethod foo :around ((x integer))
1216 (call-next-method)))
1217 Now (FOO 3) should return 3, but instead it returns 4.
1220 (SB-DEBUG:BACKTRACE) output should start with something
1221 including the name BACKTRACE, not (as in 0.pre7.88)
1222 just "0: (\"hairy arg processor\" ...)". Until about
1223 sbcl-0.pre7.109, the names in BACKTRACE were all screwed
1224 up compared to the nice useful names in sbcl-0.6.13.
1225 Around sbcl-0.pre7.109, they were mostly fixed by using
1226 NAMED-LAMBDA to implement DEFUN. However, there are still
1227 some screwups left, e.g. as of sbcl-0.pre7.109, there are
1228 still some functions named "hairy arg processor" and
1229 "SB-INT:&MORE processor".
1232 (reported by Alexey Dejneka sbcl-devel 2002-01-03)
1234 SUBTYPEP does not work well with redefined classes:
1236 * (defclass a () ())
1238 * (defclass b () ())
1243 * (defclass b (a) ())
1248 * (defclass b () ())
1257 KNOWN BUGS RELATED TO THE IR1 INTERPRETER
1259 (Now that the IR1 interpreter has gone away, these should be
1260 relatively straightforward to fix.)
1263 The system accepts DECLAIM in most places where DECLARE would be
1264 accepted, without even issuing a warning. ANSI allows this, but since
1265 it's fairly easy to mistype DECLAIM instead of DECLARE, and the
1266 meaning is rather different, and it's unlikely that the user
1267 has a good reason for doing DECLAIM not at top level, it would be
1268 good to issue a STYLE-WARNING when this happens. A possible
1269 fix would be to issue STYLE-WARNINGs for DECLAIMs not at top level,
1270 or perhaps to issue STYLE-WARNINGs for any EVAL-WHEN not at top level.
1271 [This is considered an IR1-interpreter-related bug because until
1272 EVAL-WHEN is rewritten, which won't happen until after the IR1
1273 interpreter is gone, the system's notion of what's a top-level form
1274 and what's not will remain too confused to fix this problem.]
1277 (another wishlist thing..) Reimplement DEFMACRO to be basically
1278 like DEFMACRO-MUNDANELY, just using EVAL-WHEN.