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 bugs which have been
34 KNOWN BUGS OF NO SPECIAL CLASS:
37 DEFSTRUCT should almost certainly overwrite the old LAYOUT information
38 instead of just punting when a contradictory structure definition
39 is loaded. As it is, if you redefine DEFSTRUCTs in a way which
40 changes their layout, you probably have to rebuild your entire
41 program, even if you know or guess enough about the internals of
42 SBCL to wager that this (undefined in ANSI) operation would be safe.
45 It should cause a STYLE-WARNING, not a full WARNING, when a structure
46 slot default value does not match the declared structure slot type.
47 (The current behavior is consistent with SBCL's behavior elsewhere,
48 and would not be a problem, except that the other behavior is
49 specifically required by the ANSI spec.)
52 It should cause a note, not a WARNING, when the system ignores
53 an FTYPE proclamation for a slot accessor.
56 Error reporting on various stream-requiring operations is not
57 very good when the stream argument has the wrong type, because
58 the operation tries to fall through to Gray stream code, and then
59 dies because it's undefined. E.g.
60 (PRINT-UNREADABLE-OBJECT (*STANDARD-OUTPUT* 1)) ..)
61 gives the error message
62 error in SB-KERNEL::UNDEFINED-SYMBOL-ERROR-HANDLER:
63 The function SB-IMPL::STREAM-WRITE-STRING is undefined.
64 It would be more useful and correct to signal a TYPE-ERROR:
66 (It wouldn't be terribly difficult to write stubs for all the
67 Gray stream functions that the old CMU CL code expects, with
68 each stub just raising the appropriate TYPE-ERROR.)
71 bogus warnings about undefined functions for magic functions like
72 SB!C::%%DEFUN and SB!C::%DEFCONSTANT when cross-compiling files
73 like src/code/float.lisp
76 The "byte compiling top-level form:" output ought to be condensed.
77 Perhaps any number of such consecutive lines ought to turn into a
78 single "byte compiling top-level forms:" line.
81 Compiling a file containing the erroneous program
85 (DEFSTRUCT (BAR (:INCLUDE FOO))
88 gives only the not-very-useful message
90 (during macroexpansion)
91 Condition PROGRAM-ERROR was signalled.
92 (The specific message which says that the problem was duplicate
93 slot names gets lost.)
96 The handling of IGNORE declarations on lambda list arguments of
97 DEFMETHOD is at least weird, and in fact seems broken and useless.
98 I should fix up another layer of binding, declared IGNORABLE, for
99 typed lambda list arguments.
102 The way that the compiler munges types with arguments together
103 with types with no arguments (in e.g. TYPE-EXPAND) leads to
104 weirdness visible to the user:
105 (DEFTYPE FOO () 'FIXNUM)
107 (TYPEP 11 '(FOO)) => T, which seems weird
108 (TYPEP 11 'FIXNUM) => T
109 (TYPEP 11 '(FIXNUM)) signals an error, as it should
110 The situation is complicated by the presence of Common Lisp types
111 like UNSIGNED-BYTE (which can either be used in list form or alone)
112 so I'm not 100% sure that the behavior above is actually illegal.
113 But I'm 90+% sure, and someday perhaps I'll be motivated to look it up..
116 It would be nice if the
118 (during macroexpansion)
119 said what macroexpansion was at fault, e.g.
121 (during macroexpansion of IN-PACKAGE,
122 during macroexpansion of DEFFOO)
125 The type system doesn't understand the KEYWORD type very well:
126 (SUBTYPEP 'KEYWORD 'SYMBOL) => NIL, NIL
127 It might be possible to fix this by changing the definition of
128 KEYWORD to (AND SYMBOL (SATISFIES KEYWORDP)), but the type system
129 would need to be a bit smarter about AND types, too:
130 (SUBTYPEP '(AND SYMBOL KEYWORD) 'SYMBOL) => NIL, NIL
131 (The type system does know something about AND types already,
132 (SUBTYPEP '(AND INTEGER FLOAT) 'NUMBER) => T, T
133 (SUBTYPEP '(AND INTEGER FIXNUM) 'NUMBER) =>T, T
134 so likely this is a small patch.)
137 Floating point infinities are screwed up. [When I was converting CMU CL
138 to SBCL, I was looking for complexity to delete, and I thought it was safe
139 to just delete support for floating point infinities. It wasn't: they're
140 generated by the floating point hardware even when we remove support
141 for them in software. -- WHN] Support for them should be restored.
144 The ANSI syntax for non-STANDARD method combination types in CLOS is
145 (DEFGENERIC FOO (X) (:METHOD-COMBINATION PROGN))
146 (DEFMETHOD FOO PROGN ((X BAR)) (PRINT 'NUMBER))
147 If you mess this up, omitting the PROGN qualifier in in DEFMETHOD,
148 (DEFGENERIC FOO (X) (:METHOD-COMBINATION PROGN))
149 (DEFMETHOD FOO ((X BAR)) (PRINT 'NUMBER))
150 the error mesage is not easy to understand:
151 INVALID-METHOD-ERROR was called outside the dynamic scope
152 of a method combination function (inside the body of
153 DEFINE-METHOD-COMBINATION or a method on the generic
154 function COMPUTE-EFFECTIVE-METHOD).
155 It would be better if it were more informative, a la
156 The method combination type for this method (STANDARD) does
157 not match the method combination type for the generic function
159 Also, after you make the mistake of omitting the PROGN qualifier
160 on a DEFMETHOD, doing a new DEFMETHOD with the correct qualifier
162 (DEFMETHOD FOO PROGN ((X BAR)) (PRINT 'NUMBER))
164 INVALID-METHOD-ERROR was called outside the dynamic scope
165 of a method combination function (inside the body of
166 DEFINE-METHOD-COMBINATION or a method on the generic
167 function COMPUTE-EFFECTIVE-METHOD).
168 This is not very helpful..
171 (SUBTYPEP '(FUNCTION (T BOOLEAN) NIL)
172 '(FUNCTION (FIXNUM FIXNUM) NIL)) => T, T
173 (Also, when this is fixed, we can enable the code in PROCLAIM which
174 checks for incompatible FTYPE redeclarations.)
177 The ANSI spec says that CONS can be a compound type spec, e.g.
178 (CONS FIXNUM REAL). SBCL doesn't support this.
181 from Paolo Amoroso on the CMU CL mailing list 27 Feb 2000:
182 I use CMU CL 18b under Linux. When COMPILE-FILE is supplied a physical
183 pathname, the type of the corresponding compiled file is X86F:
184 * (compile-file "/home/paolo/lisp/tools/foo")
185 Python version 1.0, VM version Intel x86 on 27 FEB 0 06:00:46 pm.
186 Compiling: /home/paolo/lisp/tools/foo.lisp 27 FEB 0 05:57:42 pm
188 Compiling DEFUN SQUARE:
189 Byte Compiling Top-Level Form:
190 /home/paolo/lisp/tools/foo.x86f written.
191 Compilation finished in 0:00:00.
192 #p"/home/paolo/lisp/tools/foo.x86f"
195 But when the function is called with a logical pathname, the file type
197 * (compile-file "tools:foo")
198 Python version 1.0, VM version Intel x86 on 27 FEB 0 06:01:04 pm.
199 Compiling: /home/paolo/lisp/tools/foo.lisp 27 FEB 0 05:57:42 pm
201 Compiling DEFUN SQUARE:
202 Byte Compiling Top-Level Form:
203 TOOLS:FOO.FASL written.
204 Compilation finished in 0:00:00.
205 #p"/home/paolo/lisp/tools/foo.fasl"
210 from DTC on the CMU CL mailing list 25 Feb 2000:
211 ;;; Compiler fails when this file is compiled.
213 ;;; Problem shows up in delete-block within ir1util.lisp. The assertion
214 ;;; (assert (member (functional-kind lambda) '(:let :mv-let :assignment)))
215 ;;; fails within bind node branch.
217 ;;; Note that if c::*check-consistency* is enabled then an un-reached
218 ;;; entry is also reported.
221 (declare (values nil))
238 (let ((ttt #'(lambda () (go cccc))))
239 (declare (special ttt))
240 (return-from bbbb nil))
243 (return-from bbbb nil))))))
246 (I *think* this is a bug. It certainly seems like strange behavior. But
247 the ANSI spec is scary, dark, and deep..)
248 (FORMAT NIL "~,1G" 1.4) => "1. "
249 (FORMAT NIL "~3,1G" 1.4) => "1. "
252 from Marco Antoniotti on cmucl-imp mailing list 1 Mar 2000:
254 (setf (find-class 'ccc1) (find-class 'ccc))
255 (defmethod zut ((c ccc1)) 123)
256 DTC's recommended workaround from the mailing list 3 Mar 2000:
257 (setf (pcl::find-class 'ccc1) (pcl::find-class 'ccc))
260 The ANSI spec, in section "22.3.5.2 Tilde Less-Than-Sign: Logical Block",
261 says that an error is signalled if ~W, ~_, ~<...~:>, ~I, or ~:T is used
262 inside "~<..~>" (without the colon modifier on the closing syntax).
263 However, SBCL doesn't do this:
264 * (FORMAT T "~<munge~wegnum~>" 12)
269 When too many files are opened, OPEN will fail with an
270 uninformative error message
271 error in function OPEN: error opening #P"/tmp/foo.lisp": NIL
272 instead of saying that too many files are open.
275 Right now, when COMPILE-FILE has a read error, it actually pops
276 you into the debugger before giving up on the file. It should
277 instead handle the error, perhaps issuing (and handling)
278 a secondary error "caught ERROR: unrecoverable error during compilation"
279 and then return with FAILURE-P true,
282 from CMU CL mailing list 01 May 2000
284 I realize I can take care of this by doing (proclaim (ignore pcl::.slots1.))
285 but seeing as .slots0. is not-exported, shouldn't it be ignored within the
289 In: DEFMETHOD FOO-BAR-BAZ (RESOURCE-TYPE)
290 (DEFMETHOD FOO-BAR-BAZ
291 ((SELF RESOURCE-TYPE))
292 (SETF (SLOT-VALUE SELF 'NAME) 3))
293 --> BLOCK MACROLET PCL::FAST-LEXICAL-METHOD-FUNCTIONS
294 --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET
295 --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS LET PCL::BIND-ARGS LET* PCL::PV-BINDING
296 --> PCL::PV-BINDING1 PCL::PV-ENV LET
298 (LET ((PCL::.SLOTS0. #))
303 Warning: Variable PCL::.SLOTS0. defined but never used.
305 Compilation unit finished.
308 #<Standard-Method FOO-BAR-BAZ (RESOURCE-TYPE) {480918FD}>
311 reported by Sam Steingold on the cmucl-imp mailing list 12 May 2000:
313 Also, there is another bug: `array-displacement' should return an array
314 or nil as first value (as per ANSI CL), while CMUCL declares it as
315 returning an array as first value always.
318 Sometimes (SB-EXT:QUIT) fails with
319 Argh! maximum interrupt nesting depth (4096) exceeded, exiting
320 Process inferior-lisp exited abnormally with code 1
321 I haven't noticed a repeatable case of this yet.
324 The system accepts DECLAIM in most places where DECLARE would be
325 accepted, without even issuing a warning. ANSI allows this, but since
326 it's fairly easy to mistype DECLAIM instead of DECLARE, and the
327 meaning is rather different, and it's unlikely that the user
328 has a good reason for doing DECLAIM not at top level, it would be
329 good to issue a STYLE-WARNING when this happens. A possible
330 fix would be to issue STYLE-WARNINGs for DECLAIMs not at top level,
331 or perhaps to issue STYLE-WARNINGs for any EVAL-WHEN not at top level.
334 some sort of bug in inlining and RETURN-FROM in sbcl-0.6.5: Compiling
337 (BLOCK USED-BY-SOME-Y?
340 (UNLESS (REJECTED? Y)
341 (RETURN-FROM USED-BY-SOME-Y? T)))))
342 (DECLARE (INLINE FROB))
347 error in function SB-KERNEL:ASSERT-ERROR:
348 The assertion (EQ (SB-C::CONTINUATION-KIND SB-C::CONT) :BLOCK-START) failed.
349 This is still present in sbcl-0.6.8.
352 The CMU CL reader code takes liberties in binding the standard read table
353 when reading the names of characters. Tim Moore posted a patch to the
354 CMU CL mailing list Mon, 22 May 2000 21:30:41 -0700.
357 In some cases the compiler believes type declarations on array
358 elements without checking them, e.g.
359 (DECLAIM (OPTIMIZE (SAFETY 3) (SPEED 1) (SPACE 1)))
362 (DECLARE (TYPE (SIMPLE-ARRAY CONS 1) X))
363 (WHEN (CONSP (AREF X 0))
365 (BAR (VECTOR (MAKE-FOO :A 11 :B 12)))
368 in SBCL 0.6.5 (and also in CMU CL 18b). This does not happen for
369 all cases, e.g. the type assumption *is* checked if the array
370 elements are declared to be of some structure type instead of CONS.
373 The printer doesn't report closures very well. This is true in
377 #<Closure Over Function "DEFUN STRUCTURE-SLOT-ACCESSOR" {134D1A1}>
378 It would be nice to make closures have a settable name slot,
379 and make things like DEFSTRUCT and FLET, which create closures,
380 set helpful values into this slot.
383 And as long as we're wishing, it would be awfully nice if INSPECT could
384 also report on closures, telling about the values of the bound variables.
387 as reported by Robert Strandh on the CMU CL mailing list 12 Jun 2000:
389 (defconstant +a-constant+ (make-instance 'a-class))
390 (defconstant +another-constant+ (vector +a-constant+))
392 CMU Common Lisp release x86-linux 2.4.19 8 February 2000 build 456,
395 Send bug reports and questions to your local CMU CL maintainer,
396 or to pvaneynd@debian.org
397 or to cmucl-help@cons.org. (prefered)
398 type (help) for help, (quit) to exit, and (demo) to see the demos
400 Python 1.0, target Intel x86
401 CLOS based on PCL version: September 16 92 PCL (f)
402 * (defclass a-class () ())
403 #<STANDARD-CLASS A-CLASS {48027BD5}>
404 * (compile-file "xx.lisp")
405 Python version 1.0, VM version Intel x86 on 12 JUN 00 08:12:55 am.
407 /home/strandh/Research/Functional/Common-Lisp/CLIM/Development/McCLIM
408 /xx.lisp 12 JUN 00 07:47:14 am
409 Compiling Load Time Value of (PCL::GET-MAKE-INSTANCE-FUNCTION-SYMBOL
411 Byte Compiling Top-Level Form:
412 Error in function C::DUMP-STRUCTURE: Attempt to dump invalid
414 #<A-CLASS {4803A5B5}>
418 The compiler assumes that any time a function of declared FTYPE
419 doesn't signal an error, its arguments were of the declared type.
420 E.g. compiling and loading
421 (DECLAIM (OPTIMIZE (SAFETY 3)))
422 (DEFUN FACTORIAL (X) (GAMMA (1+ X)))
423 (DECLAIM (FTYPE (FUNCTION (UNSIGNED-BYTE) FACTORIAL)))
425 (COND ((> (FACTORIAL X) 1.0E6)
426 (FORMAT T "too big~%"))
428 (FORMAT T "exactly ~S~%" (FACTORIAL X)))
430 (FORMAT T "approximately ~S~%" (FACTORIAL X)))))
433 will cause the INTEGERP case to be selected, giving bogus output a la
435 This violates the "declarations are assertions" principle.
436 According to the ANSI spec, in the section "System Class FUNCTION",
437 this is a case of "lying to the compiler", but the lying is done
438 by the code which calls FACTORIAL with non-UNSIGNED-BYTE arguments,
439 not by the unexpectedly general definition of FACTORIAL. In any case,
440 "declarations are assertions" means that lying to the compiler should
441 cause an error to be signalled, and should not cause a bogus
442 result to be returned. Thus, the compiler should not assume
443 that arbitrary functions check their argument types. (It might
444 make sense to add another flag (CHECKED?) to DEFKNOWN to
445 identify functions which *do* check their argument types.)
448 As pointed out by Martin Cracauer on the CMU CL mailing list
449 13 Jun 2000, the :FILE-LENGTH operation for
450 FD-STREAM-MISC-ROUTINE is broken for large files: it says
451 (THE INDEX SIZE) even though SIZE can be larger than INDEX.
454 In SBCL 0.6.5 (and CMU CL 18b) compiling and loading
455 (in-package :cl-user)
456 (declaim (optimize (safety 3)
458 (compilation-speed 2)
461 #+nil (sb-ext:inhibit-warnings 2)))
462 (declaim (ftype (function * (values)) emptyvalues))
463 (defun emptyvalues (&rest rest) (declare (ignore rest)) (values))
465 (defgeneric assertoid ((x t)))
466 (defmethod assertoid ((x t)) "just a placeholder")
468 (declare (type hash-table ht))
474 (assertoid (hash-table-count ht)))))))
475 (unless (typep res 'foo)
477 (common-lisp-user::bad-result-from-assertive-typed-fun
481 (bar (make-hash-table))
483 Error in KERNEL::UNDEFINED-SYMBOL-ERROR-HANDLER:
484 the function C::%INSTANCE-TYPEP is undefined.
485 %INSTANCE-TYPEP is always supposed to be IR1-transformed away, but for
486 some reason -- the (VALUES) return value declaration? -- the optimizer is
487 confused and compiles a full call to %INSTANCE-TYPEP (which doesn't exist
488 as a function) instead.
491 The %INSTANCE-TYPEP problem in bug 37 comes up also when compiling
493 (IN-PACKAGE :CL-USER)
495 (DECLARE (OPTIMIZE (SAFETY 3) (SPEED 2) (SPACE 2)))
496 (DECLAIM (FTYPE (FUNCTION (&REST T) (VALUES)) EMPTYVALUES))
497 (DEFUN EMPTYVALUES (&REST REST)
498 (DECLARE (IGNORE REST))
500 (DEFSTRUCT DUMMYSTRUCT X Y)
501 (DEFUN FROB-EMPTYVALUES (X)
502 (LET ((RES (EMPTYVALUES X X X)))
503 (UNLESS (TYPEP RES 'DUMMYSTRUCT)
504 'EXPECTED-RETURN-VALUE))))
505 (ASSERT (EQ (FROB-EMPTYVALUES 11) 'EXPECTED-RETURN-VALUE))
509 DEFMETHOD doesn't check the syntax of &REST argument lists properly,
510 accepting &REST even when it's not followed by an argument name:
511 (DEFMETHOD FOO ((X T) &REST) NIL)
514 On the CMU CL mailing list 26 June 2000, Douglas Crosher wrote
516 Hannu Rummukainen wrote:
518 > There's something weird going on with the compilation of the attached
519 > code. Compiling and loading the file in a fresh lisp, then invoking
521 Thanks for the bug report, nice to have this one fixed. It was a bug
522 in the x86 backend, the < VOP. A fix has been committed to the main
523 source, see the file compiler/x86/float.lisp.
525 Probably the same bug exists in SBCL.
528 TYPEP treats the result of UPGRADED-ARRAY-ELEMENT-TYPE as gospel,
529 so that (TYPEP (MAKE-ARRAY 3) '(VECTOR SOMETHING-NOT-DEFINED-YET))
530 returns (VALUES T T). Probably it should be an error instead,
531 complaining that the type SOMETHING-NOT-DEFINED-YET is not defined.
534 TYPEP of VALUES types is sometimes implemented very inefficiently, e.g. in
535 (DEFTYPE INDEXOID () '(INTEGER 0 1000))
537 (DECLARE (TYPE INDEXOID X))
538 (THE (VALUES INDEXOID)
540 where the implementation of the type check in function FOO
541 includes a full call to %TYPEP. There are also some fundamental problems
542 with the interpretation of VALUES types (inherited from CMU CL, and
543 from the ANSI CL standard) as discussed on the cmucl-imp@cons.org
544 mailing list, e.g. in Robert Maclachlan's post of 21 Jun 2000.
547 The definitions of SIGCONTEXT-FLOAT-REGISTER and
548 %SET-SIGCONTEXT-FLOAT-REGISTER in x86-vm.lisp say they're not
549 supported on FreeBSD because the floating point state is not saved,
550 but at least as of FreeBSD 4.0, the floating point state *is* saved,
551 so they could be supported after all. Very likely
552 SIGCONTEXT-FLOATING-POINT-MODES could now be supported, too.
555 (as discussed by Douglas Crosher on the cmucl-imp mailing list ca.
556 Aug. 10, 2000): CMUCL currently interprets 'member as '(member); same
557 issue with 'union, 'and, 'or etc. So even though according to the
558 ANSI spec, bare 'MEMBER, 'AND, and 'OR are not legal types, CMUCL
559 (and now SBCL) interpret them as legal types.
562 ANSI specifies DEFINE-SYMBOL-MACRO, but it's not defined in SBCL.
563 CMU CL added it ca. Aug 13, 2000, after some discussion on the mailing
564 list, and it is probably possible to use substantially the same
565 patches to add it to SBCL.
568 a slew of floating-point-related errors reported by Peter Van Eynde
570 a: (SQRT -9.0) fails, because SB-KERNEL::COMPLEX-SQRT is undefined.
571 Similarly, COMPLEX-ASIN, COMPLEX-ACOS, COMPLEX-ACOSH, and others
573 b: SBCL's value for LEAST-POSITIVE-SHORT-FLOAT is bogus, and
574 should probably be 1.4012985e-45. In SBCL,
575 (/ LEAST-POSITIVE-SHORT-FLOAT 2) returns a number smaller
576 than LEAST-POSITIVE-SHORT-FLOAT. Similar problems
577 exist for LEAST-NEGATIVE-SHORT-FLOAT, LEAST-POSITIVE-LONG-FLOAT,
578 and LEAST-NEGATIVE-LONG-FLOAT.
579 c: Many expressions generate floating infinity:
584 PVE's regression tests want them to raise errors. SBCL
585 generates the infinities instead, which may or may not be
586 conforming behavior, but then blow it by being unable to
587 output the infinities, since support for infinities is generally
588 broken, and in particular SB-IMPL::OUTPUT-FLOAT-INFINITY is
590 d: (in section12.erg) various forms a la
591 (FLOAT 1 DOUBLE-FLOAT-EPSILON)
592 don't give the right behavior.
595 type safety errors reported by Peter Van Eynde July 25, 2000:
596 a: (COERCE (QUOTE (A B C)) (QUOTE (VECTOR * 4)))
598 In general lengths of array type specifications aren't
599 checked by COERCE, so it fails when the spec is
600 (VECTOR 4), (STRING 2), (SIMPLE-BIT-VECTOR 3), or whatever.
601 b: CONCATENATE has the same problem of not checking the length
602 of specified output array types. MAKE-SEQUENCE and MAP and
603 MERGE also have the same problem.
604 c: (COERCE 'AND 'FUNCTION) returns something related to
605 (MACRO-FUNCTION 'AND), but ANSI says it should raise an error.
606 d: ELT signals SIMPLE-ERROR if its index argument
607 isn't a valid index for its sequence argument, but should
608 signal TYPE-ERROR instead.
609 e: FILE-LENGTH is supposed to signal a type error when its
610 argument is not a stream associated with a file, but doesn't.
611 f: (FLOAT-RADIX 2/3) should signal an error instead of
613 g: (LOAD "*.lsp") should signal FILE-ERROR.
614 h: (MAKE-CONCATENATED-STREAM (MAKE-STRING-OUTPUT-STREAM))
615 should signal TYPE-ERROR.
616 i: MAKE-TWO-WAY-STREAM doesn't check that its arguments can
617 be used for input and output as needed. It should fail with
618 TYPE-ERROR when handed e.g. the results of
619 MAKE-STRING-INPUT-STREAM or MAKE-STRING-OUTPUT-STREAM in
620 the inappropriate positions, but doesn't.
621 j: (PARSE-NAMESTRING (COERCE (LIST #\f #\o #\o (CODE-CHAR 0) #\4 #\8)
623 should probably signal an error instead of making a pathname with
625 k: READ-BYTE is supposed to signal TYPE-ERROR when its argument is
626 not a binary input stream, but instead cheerfully reads from
627 character streams, e.g. (MAKE-STRING-INPUT-STREAM "abc").
630 DEFCLASS bugs reported by Peter Van Eynde July 25, 2000:
631 a: (DEFCLASS FOO () (A B A)) should signal a PROGRAM-ERROR, and
633 b: (DEFCLASS FOO () (A B A) (:DEFAULT-INITARGS X A X B)) should
634 signal a PROGRAM-ERROR, and doesn't.
635 c: (DEFCLASS FOO07 NIL ((A :ALLOCATION :CLASS :ALLOCATION :CLASS))),
636 and other DEFCLASS forms with duplicate specifications in their
637 slots, should signal a PROGRAM-ERROR, and doesn't.
638 d: (DEFGENERIC IF (X)) should signal a PROGRAM-ERROR, but instead
639 causes a COMPILER-ERROR.
642 SYMBOL-MACROLET bugs reported by Peter Van Eynde July 25, 2000:
643 a: (SYMBOL-MACROLET ((T TRUE)) ..) should probably signal
644 PROGRAM-ERROR, but SBCL accepts it instead.
645 b: SYMBOL-MACROLET should refuse to bind something which is
646 declared as a global variable, signalling PROGRAM-ERROR.
647 c: SYMBOL-MACROLET should signal PROGRAM-ERROR if something
648 it binds is declared SPECIAL inside.
651 LOOP bugs reported by Peter Van Eynde July 25, 2000:
652 a: (LOOP WITH (A B) DO (PRINT 1)) is a syntax error according to
653 the definition of WITH clauses given in the ANSI spec, but
654 compiles and runs happily in SBCL.
655 b: a messy one involving package iteration:
656 interpreted Form: (LET ((PACKAGE (MAKE-PACKAGE "LOOP-TEST"))) (INTERN "blah" PACKAGE) (LET ((BLAH2 (INTERN "blah2" PACKAGE))) (EXPORT BLAH2 PACKAGE)) (LIST (SORT (LOOP FOR SYM BEING EACH PRESENT-SYMBOL OF PACKAGE FOR SYM-NAME = (SYMBOL-NAME SYM) COLLECT SYM-NAME) (FUNCTION STRING<)) (SORT (LOOP FOR SYM BEING EACH EXTERNAL-SYMBOL OF PACKAGE FOR SYM-NAME = (SYMBOL-NAME SYM) COLLECT SYM-NAME) (FUNCTION STRING<))))
657 Should be: (("blah" "blah2") ("blah2"))
658 SBCL: (("blah") ("blah2"))
659 * (LET ((X 1)) (LOOP FOR I BY (INCF X) FROM X TO 10 COLLECT I))
660 doesn't work -- SBCL's LOOP says BY isn't allowed in a FOR clause.
663 type system errors reported by Peter Van Eynde July 25, 2000:
664 a: (SUBTYPEP 'BIGNUM 'INTEGER) => NIL, NIL
665 but should be (VALUES T T) instead.
666 b: (SUBTYPEP 'EXTENDED-CHAR 'CHARACTER) => NIL, NIL
667 but should be (VALUES T T) instead.
668 c: (SUBTYPEP '(INTEGER (0) (0)) 'NIL) dies with nested errors.
669 d: In general, the system doesn't like '(INTEGER (0) (0)) -- it
670 blows up at the level of SPECIFIER-TYPE with
671 "Lower bound (0) is greater than upper bound (0)." Probably
672 SPECIFIER-TYPE should return NIL instead.
673 e: (TYPEP 0 '(COMPLEX (EQL 0)) fails with
674 "Component type for Complex is not numeric: (EQL 0)."
675 This might be easy to fix; the type system already knows
676 that (SUBTYPEP '(EQL 0) 'NUMBER) is true.
677 f: The type system doesn't know about the condition system,
678 so that e.g. (TYPEP 'SIMPLE-ERROR 'ERROR)=>NIL.
679 g: The type system isn't all that smart about relationships
680 between hairy types, as shown in the type.erg test results,
681 e.g. (SUBTYPEP 'CONS '(NOT ATOM)) => NIL, NIL.
684 miscellaneous errors reported by Peter Van Eynde July 25, 2000:
686 (DEFGENERIC FOO02 (X))
687 (DEFMETHOD FOO02 ((X NUMBER)) T)
688 (LET ((M (FIND-METHOD (FUNCTION FOO02)
690 (LIST (FIND-CLASS (QUOTE NUMBER))))))
691 (REMOVE-METHOD (FUNCTION FOO02) M)
692 (DEFGENERIC FOO03 (X))
693 (ADD-METHOD (FUNCTION FOO03) M)))
694 should give an error, but SBCL allows it.
695 b: READ should probably return READER-ERROR, not the bare
696 arithmetic error, when input a la "1/0" or "1e1000" causes
700 It has been reported (e.g. by Peter Van Eynde) that there are
701 several metaobject protocol "errors". (In order to fix them, we might
702 need to document exactly what metaobject protocol specification
703 we're following -- the current code is just inherited from PCL.)
706 another error from Peter Van Eynde 5 September 2000:
707 (FORMAT NIL "~F" "FOO") should work, but instead reports an error.
708 PVE submitted a patch to deal with this bug, but it exposes other
709 comparably serious bugs, so I didn't apply it. It looks as though
710 the FORMAT code needs a fair amount of rewriting in order to comply
711 with the various details of the ANSI spec.
714 The implementation of #'+ returns its single argument without
715 type checking, e.g. (+ "illegal") => "illegal".
718 In sbcl-0.6.7, there is no doc string for CL:PUSH, probably
719 because it's defined with the DEFMACRO-MUNDANELY macro and something
720 is wrong with doc string setting in that macro.
723 Attempting to use COMPILE on something defined by DEFMACRO fails:
724 (DEFMACRO FOO (X) (CONS X X))
726 Error in function C::GET-LAMBDA-TO-COMPILE:
727 #<Closure Over Function "DEFUN (SETF MACRO-FUNCTION)" {480E21B1}> was defined in a non-null environment.
730 (SUBTYPEP '(AND ZILCH INTEGER) 'ZILCH)
734 CL:*DEFAULT-PATHNAME-DEFAULTS* doesn't behave as ANSI suggests (reflecting
735 current working directory). And there's no supported way to update
736 or query the current working directory (a la Unix "chdir" and "pwd"),
737 which is functionality that ILISP needs (and currently gets with low-level
741 The debugger LIST-LOCATIONS command doesn't work properly.
744 Compiling and loading
745 (DEFUN FAIL (X) (THROW 'FAIL-TAG X))
747 then requesting a BACKTRACE at the debugger prompt gives no information
748 about where in the user program the problem occurred.
751 The compiler is supposed to do type inference well enough that
754 ((SIMPLE-ARRAY SINGLE-FLOAT)
756 (DECLARE (TYPE (SIMPLE-ARRAY SINGLE-FLOAT) X))
759 is redundant. However, as reported by Juan Jose Garcia Ripoll for
760 CMU CL, it sometimes doesn't. Adding declarations is a pretty good
761 workaround for the problem for now, but can't be done by the TYPECASE
762 macros themselves, since it's too hard for the macro to detect
763 assignments to the variable within the clause.
764 Note: The compiler *is* smart enough to do the type inference in
765 many cases. This case, derived from a couple of MACROEXPAND-1
766 calls on Ripoll's original test case,
768 (DECLARE (OPTIMIZE SPEED (SAFETY 0)))
769 (COND ((TYPEP A '(SIMPLE-ARRAY SINGLE-FLOAT)) NIL
770 (LET ((LENGTH (ARRAY-TOTAL-SIZE A)))
771 (LET ((I 0) (G2554 LENGTH))
772 (DECLARE (TYPE REAL G2554) (TYPE REAL I))
775 (WHEN (>= I G2554) (GO SB-LOOP::END-LOOP))
776 (SETF (ROW-MAJOR-AREF A I) (- (ROW-MAJOR-AREF A I)))
777 (GO SB-LOOP::NEXT-LOOP)
778 SB-LOOP::END-LOOP))))))
779 demonstrates the problem; but the problem goes away if the TAGBODY
780 and GO forms are removed (leaving the SETF in ordinary, non-looping
781 code), or if the TAGBODY and GO forms are retained, but the
782 assigned value becomes 0.0 instead of (- (ROW-MAJOR-AREF A I)).
785 KNOWN BUGS RELATED TO THE IR1 INTERPRETER
787 (Note: At some point, the pure interpreter (actually a semi-pure
788 interpreter aka "the IR1 interpreter") will probably go away, replaced
790 (DEFUN EVAL (X) (FUNCALL (COMPILE NIL (LAMBDA ..)))))
791 and at that time these bugs should either go away automatically or
792 become more tractable to fix. Until then, they'll probably remain,
793 since some of them aren't considered urgent, and the rest are too hard
794 to fix as long as so many special cases remain. After the IR1
795 interpreter goes away is also the preferred time to start
796 systematically exterminating cases where debugging functionality
797 (backtrace, breakpoint, etc.) breaks down, since getting rid of the
798 IR1 interpreter will reduce the number of special cases we need to
802 The FUNCTION special operator doesn't check properly whether its
803 argument is a function name. E.g. (FUNCTION (X Y)) returns a value
804 instead of failing with an error. (Later attempting to funcall the
805 value does cause an error.)
808 COMPILED-FUNCTION-P bogusly reports T for interpreted functions:
809 * (DEFUN FOO (X) (- 12 X))
811 * (COMPILED-FUNCTION-P #'FOO)
816 (DEFVAR *SUPPRESS-P* T)
817 (EVAL '(UNLESS *SUPPRESS-P*
818 (EVAL-WHEN (:COMPILE-TOPLEVEL :LOAD-TOPLEVEL :EXECUTE)
819 (FORMAT T "surprise!"))))
820 prints "surprise!". Probably the entire EVAL-WHEN mechanism ought to be
821 rewritten from scratch to conform to the ANSI definition, abandoning
822 the *ALREADY-EVALED-THIS* hack which is used in sbcl-0.6.8.9 (and
823 in the original CMU CL source, too). This should be easier to do --
824 though still nontrivial -- once the various IR1 interpreter special
828 EVAL-WHEN's idea of what's a toplevel form is even more screwed up
829 than the example in IR1-3 would suggest, since COMPILE-FILE and
830 COMPILE both print both "right now!" messages when compiling the
834 (EVAL-WHEN (:COMPILE-TOPLEVEL :LOAD-TOPLEVEL :EXECUTE)
835 (PRINT "yes! right now!"))
838 (EVAL-WHEN (:COMPILE-TOPLEVEL :LOAD-TOPLEVEL :EXECUTE)
839 (PRINT "no! right now!"))
841 and while EVAL doesn't print the "right now!" messages, the first
842 FUNCALL on the value returned by EVAL causes both of them to be printed.