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 please please include enough information in a bug report
9 that someone reading 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.
25 KNOWN PORT-SPECIFIC BUGS
28 The breakpoint-based TRACE facility doesn't work properly
29 in the OpenBSD port of sbcl-0.6.7.
31 KNOWN BUGS RELATED TO THE IR1 INTERPRETER
33 At some point, the pure interpreter (aka the "IR1 interpreter") will
34 probably go away (replaced by constructs like
35 (DEFUN EVAL (X) (FUNCALL (COMPILE NIL (LAMBDA ..)))))
36 and at that time these bugs should go away automatically. Until then,
37 they'll probably remain, since they're not considered urgent.
40 The FUNCTION special operator doesn't check properly whether its
41 argument is a function name. E.g. (FUNCTION (X Y)) returns a value
42 instead of failing with an error. (Later attempting to funcall the
43 value does cause an error.)
46 COMPILED-FUNCTION-P bogusly reports T for interpreted functions:
47 * (DEFUN FOO (X) (- 12 X))
49 * (COMPILED-FUNCTION-P #'FOO)
54 (There is also some information on bugs in the manual page and in the
55 TODO file. Eventually more such information may move here.)
58 Failure in initialization files is not handled gracefully -- it's
59 a throw to TOP-LEVEL-CATCHER, which is not caught until we enter
60 TOPLEVEL-REPL. Code should be added to catch such THROWs even when
61 we're not in TOPLEVEL-REPL and do *something* with them (probably
62 complaining about an error outside TOPLEVEL-REPL, perhaps printing
63 a BACKTRACE, then terminating execution of SBCL).
66 DEFSTRUCT should almost certainly overwrite the old LAYOUT information
67 instead of just punting when a contradictory structure definition
71 It should cause a STYLE-WARNING, not a full WARNING, when a structure
72 slot default value does not match the declared structure slot type.
73 (The current behavior is consistent with SBCL's behavior elsewhere,
74 and would not be a problem, except that the other behavior is
75 specifically required by the ANSI spec.)
78 It should cause a STYLE-WARNING, not a WARNING, when the system ignores
79 an FTYPE proclamation for a slot accessor.
82 Error reporting on various stream-requiring operations is not
83 very good when the stream argument has the wrong type, because
84 the operation tries to fall through to Gray stream code, and then
85 dies because it's undefined. E.g.
86 (PRINT-UNREADABLE-OBJECT (*STANDARD-OUTPUT* 1))
87 gives the error message
88 error in SB-KERNEL::UNDEFINED-SYMBOL-ERROR-HANDLER:
89 The function SB-IMPL::STREAM-WRITE-STRING is undefined.
90 It would be more useful and correct to signal a TYPE-ERROR:
92 (It wouldn't be terribly difficult to write stubs for all the
93 Gray stream functions that the old CMU CL code expects, with
94 each stub just raising the appropriate TYPE-ERROR.)
97 bogus warnings about undefined functions for magic functions like
98 SB!C::%%DEFUN and SB!C::%DEFCONSTANT when cross-compiling files
99 like src/code/float.lisp
102 The "byte compiling top-level form:" output ought to be condensed.
103 Perhaps any number of such consecutive lines ought to turn into a
104 single "byte compiling top-level forms:" line.
107 Compiling a file containing the erroneous program
111 (DEFSTRUCT (BAR (:INCLUDE FOO))
114 gives only the not-very-useful message
116 (during macroexpansion)
117 Condition PROGRAM-ERROR was signalled.
118 (The specific message which says that the problem was duplicate
119 slot names gets lost.)
122 The handling of IGNORE declarations on lambda list arguments of
123 DEFMETHOD is at least weird, and in fact seems broken and useless.
124 I should fix up another layer of binding, declared IGNORABLE, for
125 typed lambda list arguments.
128 The way that the compiler munges types with arguments together
129 with types with no arguments (in e.g. TYPE-EXPAND) leads to
130 weirdness visible to the user:
131 (DEFTYPE FOO () 'FIXNUM)
133 (TYPEP 11 '(FOO)) => T, which seems weird
134 (TYPEP 11 'FIXNUM) => T
135 (TYPEP 11 '(FIXNUM)) signals an error, as it should
136 The situation is complicated by the presence of Common Lisp types
137 like UNSIGNED-BYTE (which can either be used in list form or alone)
138 so I'm not 100% sure that the behavior above is actually illegal.
139 But I'm 90+% sure, and someday perhaps I'll be motivated to look it up..
142 It would be nice if the
144 (during macroexpansion)
145 said what macroexpansion was at fault, e.g.
147 (during macroexpansion of IN-PACKAGE,
148 during macroexpansion of DEFFOO)
151 The type system doesn't understand the KEYWORD type very well:
152 (SUBTYPEP 'KEYWORD 'SYMBOL) => NIL, NIL
153 It might be possible to fix this by changing the definition of
154 KEYWORD to (AND SYMBOL (SATISFIES KEYWORDP)), but the type system
155 would need to be a bit smarter about AND types, too:
156 (SUBTYPEP '(AND SYMBOL KEYWORD) 'SYMBOL) => NIL, NIL
157 (The type system does know something about AND types already,
158 (SUBTYPEP '(AND INTEGER FLOAT) 'NUMBER) => T, T
159 (SUBTYPEP '(AND INTEGER FIXNUM) 'NUMBER) =>T, T
160 so likely this is a small patch.)
163 Floating point infinities are screwed up. [When I was converting CMU CL
164 to SBCL, I was looking for complexity to delete, and I thought it was safe
165 to just delete support for floating point infinities. It wasn't: they're
166 generated by the floating point hardware even when we remove support
167 for them in software. -- WHN] Support for them should be restored.
170 The ANSI syntax for non-STANDARD method combination types in CLOS is
171 (DEFGENERIC FOO (X) (:METHOD-COMBINATION PROGN))
172 (DEFMETHOD FOO PROGN ((X BAR)) (PRINT 'NUMBER))
173 If you mess this up, omitting the PROGN qualifier in in DEFMETHOD,
174 (DEFGENERIC FOO (X) (:METHOD-COMBINATION PROGN))
175 (DEFMETHOD FOO ((X BAR)) (PRINT 'NUMBER))
176 the error mesage is not easy to understand:
177 INVALID-METHOD-ERROR was called outside the dynamic scope
178 of a method combination function (inside the body of
179 DEFINE-METHOD-COMBINATION or a method on the generic
180 function COMPUTE-EFFECTIVE-METHOD).
181 It would be better if it were more informative, a la
182 The method combination type for this method (STANDARD) does
183 not match the method combination type for the generic function
185 Also, after you make the mistake of omitting the PROGN qualifier
186 on a DEFMETHOD, doing a new DEFMETHOD with the correct qualifier
188 (DEFMETHOD FOO PROGN ((X BAR)) (PRINT 'NUMBER))
190 INVALID-METHOD-ERROR was called outside the dynamic scope
191 of a method combination function (inside the body of
192 DEFINE-METHOD-COMBINATION or a method on the generic
193 function COMPUTE-EFFECTIVE-METHOD).
194 This is not very helpful..
197 (SUBTYPEP '(FUNCTION (T BOOLEAN) NIL)
198 '(FUNCTION (FIXNUM FIXNUM) NIL)) => T, T
199 (Also, when this is fixed, we can enable the code in PROCLAIM which
200 checks for incompatible FTYPE redeclarations.)
203 The ANSI spec says that CONS can be a compound type spec, e.g.
204 (CONS FIXNUM REAL). SBCL doesn't support this.
207 from Paolo Amoroso on the CMU CL mailing list 27 Feb 2000:
208 I use CMU CL 18b under Linux. When COMPILE-FILE is supplied a physical
209 pathname, the type of the corresponding compiled file is X86F:
210 * (compile-file "/home/paolo/lisp/tools/foo")
211 Python version 1.0, VM version Intel x86 on 27 FEB 0 06:00:46 pm.
212 Compiling: /home/paolo/lisp/tools/foo.lisp 27 FEB 0 05:57:42 pm
214 Compiling DEFUN SQUARE:
215 Byte Compiling Top-Level Form:
216 /home/paolo/lisp/tools/foo.x86f written.
217 Compilation finished in 0:00:00.
218 #p"/home/paolo/lisp/tools/foo.x86f"
221 But when the function is called with a logical pathname, the file type
223 * (compile-file "tools:foo")
224 Python version 1.0, VM version Intel x86 on 27 FEB 0 06:01:04 pm.
225 Compiling: /home/paolo/lisp/tools/foo.lisp 27 FEB 0 05:57:42 pm
227 Compiling DEFUN SQUARE:
228 Byte Compiling Top-Level Form:
229 TOOLS:FOO.FASL written.
230 Compilation finished in 0:00:00.
231 #p"/home/paolo/lisp/tools/foo.fasl"
236 from DTC on the CMU CL mailing list 25 Feb 2000:
237 ;;; Compiler fails when this file is compiled.
239 ;;; Problem shows up in delete-block within ir1util.lisp. The assertion
240 ;;; (assert (member (functional-kind lambda) '(:let :mv-let :assignment)))
241 ;;; fails within bind node branch.
243 ;;; Note that if c::*check-consistency* is enabled then an un-reached
244 ;;; entry is also reported.
247 (declare (values nil))
264 (let ((ttt #'(lambda () (go cccc))))
265 (declare (special ttt))
266 (return-from bbbb nil))
269 (return-from bbbb nil))))))
272 (I *think* this is a bug. It certainly seems like strange behavior. But
273 the ANSI spec is scary, dark, and deep..)
274 (FORMAT NIL "~,1G" 1.4) => "1. "
275 (FORMAT NIL "~3,1G" 1.4) => "1. "
278 from Marco Antoniotti on cmucl-imp mailing list 1 Mar 2000:
280 (setf (find-class 'ccc1) (find-class 'ccc))
281 (defmethod zut ((c ccc1)) 123)
282 DTC's recommended workaround from the mailing list 3 Mar 2000:
283 (setf (pcl::find-class 'ccc1) (pcl::find-class 'ccc))
286 There's probably a bug in the compiler handling of special variables
287 in closures, inherited from the CMU CL code, as reported on the
288 CMU CL mailing list. There's a patch for this on the CMU CL
290 Message-ID: <38C8E188.A1E38B5E@jeack.com.au>
291 Date: Fri, 10 Mar 2000 22:50:32 +1100
292 From: "Douglas T. Crosher" <dtc@jeack.com.au>
295 The ANSI spec, in section "22.3.5.2 Tilde Less-Than-Sign: Logical Block",
296 says that an error is signalled if ~W, ~_, ~<...~:>, ~I, or ~:T is used
297 inside "~<..~>" (without the colon modifier on the closing syntax).
298 However, SBCL doesn't do this:
299 * (FORMAT T "~<munge~wegnum~>" 12)
304 When too many files are opened, OPEN will fail with an
305 uninformative error message
306 error in function OPEN: error opening #P"/tmp/foo.lisp": NIL
307 instead of saying that too many files are open.
310 Right now, when COMPILE-FILE has a read error, it actually pops
311 you into the debugger before giving up on the file. It should
312 instead handle the error, perhaps issuing (and handling)
313 a secondary error "caught ERROR: unrecoverable error during compilation"
314 and then return with FAILURE-P true,
317 from CMU CL mailing list 01 May 2000
319 I realize I can take care of this by doing (proclaim (ignore pcl::.slots1.))
320 but seeing as .slots0. is not-exported, shouldn't it be ignored within the
324 In: DEFMETHOD FOO-BAR-BAZ (RESOURCE-TYPE)
325 (DEFMETHOD FOO-BAR-BAZ
326 ((SELF RESOURCE-TYPE))
327 (SETF (SLOT-VALUE SELF 'NAME) 3))
328 --> BLOCK MACROLET PCL::FAST-LEXICAL-METHOD-FUNCTIONS
329 --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET
330 --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS LET PCL::BIND-ARGS LET* PCL::PV-BINDING
331 --> PCL::PV-BINDING1 PCL::PV-ENV LET
333 (LET ((PCL::.SLOTS0. #))
338 Warning: Variable PCL::.SLOTS0. defined but never used.
340 Compilation unit finished.
343 #<Standard-Method FOO-BAR-BAZ (RESOURCE-TYPE) {480918FD}>
346 reported by Sam Steingold on the cmucl-imp mailing list 12 May 2000:
348 Also, there is another bug: `array-displacement' should return an array
349 or nil as first value (as per ANSI CL), while CMUCL declares it as
350 returning an array as first value always.
353 Sometimes (SB-EXT:QUIT) fails with
354 Argh! maximum interrupt nesting depth (4096) exceeded, exiting
355 Process inferior-lisp exited abnormally with code 1
356 I haven't noticed a repeatable case of this yet.
359 The system accepts DECLAIM in most places where DECLARE would be
360 accepted, without even issuing a warning. ANSI allows this, but since
361 it's fairly easy to mistype DECLAIM instead of DECLARE, and the
362 meaning is rather different, and it's unlikely that the user
363 has a good reason for doing DECLAIM not at top level, it would be
364 good to issue a STYLE-WARNING when this happens. A possible
365 fix would be to issue STYLE-WARNINGs for DECLAIMs not at top level,
366 or perhaps to issue STYLE-WARNINGs for any EVAL-WHEN not at top level.
369 some sort of bug in inlining and RETURN-FROM in sbcl-0.6.5: Compiling
372 (BLOCK USED-BY-SOME-Y?
375 (UNLESS (REJECTED? Y)
376 (RETURN-FROM USED-BY-SOME-Y? T)))))
377 (DECLARE (INLINE FROB))
382 error in function SB-KERNEL:ASSERT-ERROR:
383 The assertion (EQ (SB-C::CONTINUATION-KIND SB-C::CONT) :BLOCK-START) failed.
384 This is still present in sbcl-0.6.8.
387 The CMU CL reader code takes liberties in binding the standard read table
388 when reading the names of characters. Tim Moore posted a patch to the
389 CMU CL mailing list Mon, 22 May 2000 21:30:41 -0700.
392 In some cases the compiler believes type declarations on array
393 elements without checking them, e.g.
394 (DECLAIM (OPTIMIZE (SAFETY 3) (SPEED 1) (SPACE 1)))
397 (DECLARE (TYPE (SIMPLE-ARRAY CONS 1) X))
398 (WHEN (CONSP (AREF X 0))
400 (BAR (VECTOR (MAKE-FOO :A 11 :B 12)))
403 in SBCL 0.6.5 (and also in CMU CL 18b). This does not happen for
404 all cases, e.g. the type assumption *is* checked if the array
405 elements are declared to be of some structure type instead of CONS.
408 The printer doesn't report closures very well. This is true in
412 #<Closure Over Function "DEFUN STRUCTURE-SLOT-ACCESSOR" {134D1A1}>
413 It would be nice to make closures have a settable name slot,
414 and make things like DEFSTRUCT and FLET, which create closures,
415 set helpful values into this slot.
418 And as long as we're wishing, it would be awfully nice if INSPECT could
419 also report on closures, telling about the values of the bound variables.
422 as reported by Robert Strandh on the CMU CL mailing list 12 Jun 2000:
424 (defconstant +a-constant+ (make-instance 'a-class))
425 (defconstant +another-constant+ (vector +a-constant+))
427 CMU Common Lisp release x86-linux 2.4.19 8 February 2000 build 456,
430 Send bug reports and questions to your local CMU CL maintainer,
431 or to pvaneynd@debian.org
432 or to cmucl-help@cons.org. (prefered)
433 type (help) for help, (quit) to exit, and (demo) to see the demos
435 Python 1.0, target Intel x86
436 CLOS based on PCL version: September 16 92 PCL (f)
437 * (defclass a-class () ())
438 #<STANDARD-CLASS A-CLASS {48027BD5}>
439 * (compile-file "xx.lisp")
440 Python version 1.0, VM version Intel x86 on 12 JUN 00 08:12:55 am.
442 /home/strandh/Research/Functional/Common-Lisp/CLIM/Development/McCLIM
443 /xx.lisp 12 JUN 00 07:47:14 am
444 Compiling Load Time Value of (PCL::GET-MAKE-INSTANCE-FUNCTION-SYMBOL
446 Byte Compiling Top-Level Form:
447 Error in function C::DUMP-STRUCTURE: Attempt to dump invalid
449 #<A-CLASS {4803A5B5}>
453 The compiler assumes that any time a function of declared FTYPE
454 doesn't signal an error, its arguments were of the declared type.
455 E.g. compiling and loading
456 (DECLAIM (OPTIMIZE (SAFETY 3)))
457 (DEFUN FACTORIAL (X) (GAMMA (1+ X)))
458 (DECLAIM (FTYPE (FUNCTION (UNSIGNED-BYTE) FACTORIAL)))
460 (COND ((> (FACTORIAL X) 1.0E6)
461 (FORMAT T "too big~%"))
463 (FORMAT T "exactly ~S~%" (FACTORIAL X)))
465 (FORMAT T "approximately ~S~%" (FACTORIAL X)))))
468 will cause the INTEGERP case to be selected, giving bogus output a la
470 This violates the "declarations are assertions" principle.
471 According to the ANSI spec, in the section "System Class FUNCTION",
472 this is a case of "lying to the compiler", but the lying is done
473 by the code which calls FACTORIAL with non-UNSIGNED-BYTE arguments,
474 not by the unexpectedly general definition of FACTORIAL. In any case,
475 "declarations are assertions" means that lying to the compiler should
476 cause an error to be signalled, and should not cause a bogus
477 result to be returned. Thus, the compiler should not assume
478 that arbitrary functions check their argument types. (It might
479 make sense to add another flag (CHECKED?) to DEFKNOWN to
480 identify functions which *do* check their argument types.)
483 As pointed out by Martin Cracauer on the CMU CL mailing list
484 13 Jun 2000, the :FILE-LENGTH operation for
485 FD-STREAM-MISC-ROUTINE is broken for large files: it says
486 (THE INDEX SIZE) even though SIZE can be larger than INDEX.
489 In SBCL 0.6.5 (and CMU CL 18b) compiling and loading
490 (in-package :cl-user)
491 (declaim (optimize (safety 3)
493 (compilation-speed 2)
496 #+nil (sb-ext:inhibit-warnings 2)))
497 (declaim (ftype (function * (values)) emptyvalues))
498 (defun emptyvalues (&rest rest) (declare (ignore rest)) (values))
500 (defgeneric assertoid ((x t)))
501 (defmethod assertoid ((x t)) "just a placeholder")
503 (declare (type hash-table ht))
509 (assertoid (hash-table-count ht)))))))
510 (unless (typep res 'foo)
512 (common-lisp-user::bad-result-from-assertive-typed-fun
516 (bar (make-hash-table))
518 Error in KERNEL::UNDEFINED-SYMBOL-ERROR-HANDLER:
519 the function C::%INSTANCE-TYPEP is undefined.
520 %INSTANCE-TYPEP is always supposed to be IR1-transformed away, but for
521 some reason -- the (VALUES) return value declaration? -- the optimizer is
522 confused and compiles a full call to %INSTANCE-TYPEP (which doesn't exist
523 as a function) instead.
526 DEFMETHOD doesn't check the syntax of &REST argument lists properly,
527 accepting &REST even when it's not followed by an argument name:
528 (DEFMETHOD FOO ((X T) &REST) NIL)
531 On the CMU CL mailing list 26 June 2000, Douglas Crosher wrote
533 Hannu Rummukainen wrote:
535 > There's something weird going on with the compilation of the attached
536 > code. Compiling and loading the file in a fresh lisp, then invoking
538 Thanks for the bug report, nice to have this one fixed. It was a bug
539 in the x86 backend, the < VOP. A fix has been committed to the main
540 source, see the file compiler/x86/float.lisp.
542 Probably the same bug exists in SBCL.
545 TYPEP treats the result of UPGRADED-ARRAY-ELEMENT-TYPE as gospel,
546 so that (TYPEP (MAKE-ARRAY 3) '(VECTOR SOMETHING-NOT-DEFINED-YET))
547 returns (VALUES T T). Probably it should be an error instead,
548 complaining that the type SOMETHING-NOT-DEFINED-YET is not defined.
551 TYPEP of VALUES types is sometimes implemented very inefficiently, e.g. in
552 (DEFTYPE INDEXOID () '(INTEGER 0 1000))
554 (DECLARE (TYPE INDEXOID X))
555 (THE (VALUES INDEXOID)
557 where the implementation of the type check in function FOO
558 includes a full call to %TYPEP. There are also some fundamental problems
559 with the interpretation of VALUES types (inherited from CMU CL, and
560 from the ANSI CL standard) as discussed on the cmucl-imp@cons.org
561 mailing list, e.g. in Robert Maclachlan's post of 21 Jun 2000.
564 The definitions of SIGCONTEXT-FLOAT-REGISTER and
565 %SET-SIGCONTEXT-FLOAT-REGISTER in x86-vm.lisp say they're not
566 supported on FreeBSD because the floating point state is not saved,
567 but at least as of FreeBSD 4.0, the floating point state *is* saved,
568 so they could be supported after all. Very likely
569 SIGCONTEXT-FLOATING-POINT-MODES could now be supported, too.
572 (as discussed by Douglas Crosher on the cmucl-imp mailing list ca.
573 Aug. 10, 2000): CMUCL currently interprets 'member as '(member); same
574 issue with 'union, 'and, 'or etc. So even though according to the
575 ANSI spec, bare 'MEMBER, 'AND, and 'OR are not legal types, CMUCL
576 (and now SBCL) interpret them as legal types.
579 ANSI specifies DEFINE-SYMBOL-MACRO, but it's not defined in SBCL.
580 CMU CL added it ca. Aug 13, 2000, after some discussion on the mailing
581 list, and it is probably possible to use substantially the same
582 patches to add it to SBCL.
585 a slew of floating-point-related errors reported by Peter Van Eynde
587 a: (SQRT -9.0) fails, because SB-KERNEL::COMPLEX-SQRT is undefined.
588 Similarly, COMPLEX-ASIN, COMPLEX-ACOS, COMPLEX-ACOSH, and others
590 b: SBCL's value for LEAST-POSITIVE-SHORT-FLOAT is bogus, and
591 should probably be 1.4012985e-45. In SBCL,
592 (/ LEAST-POSITIVE-SHORT-FLOAT 2) returns a number smaller
593 than LEAST-POSITIVE-SHORT-FLOAT. Similar problems
594 exist for LEAST-NEGATIVE-SHORT-FLOAT, LEAST-POSITIVE-LONG-FLOAT,
595 and LEAST-NEGATIVE-LONG-FLOAT.
596 c: Many expressions generate floating infinity:
601 PVE's regression tests want them to raise errors. SBCL
602 generates the infinities instead, which may or may not be
603 conforming behavior, but then blow it by being unable to
604 output the infinities, since support for infinities is generally
605 broken, and in particular SB-IMPL::OUTPUT-FLOAT-INFINITY is
607 d: (in section12.erg) various forms a la
608 (FLOAT 1 DOUBLE-FLOAT-EPSILON)
609 don't give the right behavior.
612 type safety errors reported by Peter Van Eynde July 25, 2000:
613 a: (COERCE (QUOTE (A B C)) (QUOTE (VECTOR * 4)))
615 In general lengths of array type specifications aren't
616 checked by COERCE, so it fails when the spec is
617 (VECTOR 4), (STRING 2), (SIMPLE-BIT-VECTOR 3), or whatever.
618 b: CONCATENATE has the same problem of not checking the length
619 of specified output array types. MAKE-SEQUENCE and MAP and
620 MERGE also have the same problem.
621 c: (COERCE 'AND 'FUNCTION) returns something related to
622 (MACRO-FUNCTION 'AND), but ANSI says it should raise an error.
623 d: ELT signals SIMPLE-ERROR if its index argument
624 isn't a valid index for its sequence argument, but should
625 signal TYPE-ERROR instead.
626 e: FILE-LENGTH is supposed to signal a type error when its
627 argument is not a stream associated with a file, but doesn't.
628 f: (FLOAT-RADIX 2/3) should signal an error instead of
630 g: (LOAD "*.lsp") should signal FILE-ERROR.
631 h: (MAKE-CONCATENATED-STREAM (MAKE-STRING-OUTPUT-STREAM))
632 should signal TYPE-ERROR.
633 i: MAKE-TWO-WAY-STREAM doesn't check that its arguments can
634 be used for input and output as needed. It should fail with
635 TYPE-ERROR when handed e.g. the results of
636 MAKE-STRING-INPUT-STREAM or MAKE-STRING-OUTPUT-STREAM in
637 the inappropriate positions, but doesn't.
638 j: (PARSE-NAMESTRING (COERCE (LIST #\f #\o #\o (CODE-CHAR 0) #\4 #\8)
640 should probably signal an error instead of making a pathname with
642 k: READ-BYTE is supposed to signal TYPE-ERROR when its argument is
643 not a binary input stream, but instead cheerfully reads from
644 character streams, e.g. (MAKE-STRING-INPUT-STREAM "abc").
647 DEFCLASS bugs reported by Peter Van Eynde July 25, 2000:
648 a: (DEFCLASS FOO () (A B A)) should signal a PROGRAM-ERROR, and
650 b: (DEFCLASS FOO () (A B A) (:DEFAULT-INITARGS X A X B)) should
651 signal a PROGRAM-ERROR, and doesn't.
652 c: (DEFCLASS FOO07 NIL ((A :ALLOCATION :CLASS :ALLOCATION :CLASS))),
653 and other DEFCLASS forms with duplicate specifications in their
654 slots, should signal a PROGRAM-ERROR, and doesn't.
655 d: (DEFGENERIC IF (X)) should signal a PROGRAM-ERROR, but instead
656 causes a COMPILER-ERROR.
659 SYMBOL-MACROLET bugs reported by Peter Van Eynde July 25, 2000:
660 a: (SYMBOL-MACROLET ((T TRUE)) ..) should probably signal
661 PROGRAM-ERROR, but SBCL accepts it instead.
662 b: SYMBOL-MACROLET should refuse to bind something which is
663 declared as a global variable, signalling PROGRAM-ERROR.
664 c: SYMBOL-MACROLET should signal PROGRAM-ERROR if something
665 it binds is declared SPECIAL inside.
668 LOOP bugs reported by Peter Van Eynde July 25, 2000:
669 a: (LOOP WITH (A B) DO (PRINT 1)) is a syntax error according to
670 the definition of WITH clauses given in the ANSI spec, but
671 compiles and runs happily in SBCL.
672 b: a messy one involving package iteration:
673 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<))))
674 Should be: (("blah" "blah2") ("blah2"))
675 SBCL: (("blah") ("blah2"))
676 * (LET ((X 1)) (LOOP FOR I BY (INCF X) FROM X TO 10 COLLECT I))
677 doesn't work -- SBCL's LOOP says BY isn't allowed in a FOR clause.
680 type system errors reported by Peter Van Eynde July 25, 2000:
681 a: (SUBTYPEP 'BIGNUM 'INTEGER) => NIL, NIL
682 but should be (VALUES T T) instead.
683 b: (SUBTYPEP 'EXTENDED-CHAR 'CHARACTER) => NIL, NIL
684 but should be (VALUES T T) instead.
685 c: (SUBTYPEP '(INTEGER (0) (0)) 'NIL) dies with nested errors.
686 d: In general, the system doesn't like '(INTEGER (0) (0)) -- it
687 blows up at the level of SPECIFIER-TYPE with
688 "Lower bound (0) is greater than upper bound (0)." Probably
689 SPECIFIER-TYPE should return NIL instead.
690 e: (TYPEP 0 '(COMPLEX (EQL 0)) fails with
691 "Component type for Complex is not numeric: (EQL 0)."
692 This might be easy to fix; the type system already knows
693 that (SUBTYPEP '(EQL 0) 'NUMBER) is true.
694 f: The type system doesn't know about the condition system,
695 so that e.g. (TYPEP 'SIMPLE-ERROR 'ERROR)=>NIL.
696 g: The type system isn't all that smart about relationships
697 between hairy types, as shown in the type.erg test results,
698 e.g. (SUBTYPEP 'CONS '(NOT ATOM)) => NIL, NIL.
701 miscellaneous errors reported by Peter Van Eynde July 25, 2000:
703 (DEFGENERIC FOO02 (X))
704 (DEFMETHOD FOO02 ((X NUMBER)) T)
705 (LET ((M (FIND-METHOD (FUNCTION FOO02)
707 (LIST (FIND-CLASS (QUOTE NUMBER))))))
708 (REMOVE-METHOD (FUNCTION FOO02) M)
709 (DEFGENERIC FOO03 (X))
710 (ADD-METHOD (FUNCTION FOO03) M)))
711 should give an error, but SBCL allows it.
712 b: READ should probably return READER-ERROR, not the bare
713 arithmetic error, when input a la "1/0" or "1e1000" causes
715 c: (BUTLAST NIL) should return NIL. (This appears to be a compiler
716 bug, since the definition of BUTLAST, when interpreted, does
717 give (BUTLAST NIL)=>NIL.)
720 It has been reported (e.g. by Peter Van Eynde) that there are
721 several metaobject protocol "errors". (In order to fix them, we might
722 need to document exactly what metaobject protocol specification
723 we're following -- the current code is just inherited from PCL.)
726 another error from Peter Van Eynde 5 September 2000:
727 (FORMAT NIL "~F" "FOO") should work, but instead reports an error.
728 PVE submitted a patch to deal with this bug, but it exposes other
729 comparably serious bugs, so I didn't apply it. It looks as though
730 the FORMAT code needs a fair amount of rewriting in order to comply
731 with the various details of the ANSI spec.
734 The implementation of #'+ returns its single argument without
735 type checking, e.g. (+ "illegal") => "illegal".
738 In sbcl-0.6.7, there is no doc string for CL:PUSH, probably
739 because it's defined with the DEFMACRO-MUNDANELY macro and something
740 is wrong with doc string setting in that macro.
743 Attempting to use COMPILE on something defined by DEFMACRO fails:
744 (DEFMACRO FOO (X) (CONS X X))
746 Error in function C::GET-LAMBDA-TO-COMPILE:
747 #<Closure Over Function "DEFUN (SETF MACRO-FUNCTION)" {480E21B1}> was defined in a non-null environment.
750 In sbcl-0.6.7, the compiler accepted a bogus declaration
751 (TYPE INDEX LENGTH) in the definition of BUTLAST, and then died
752 with infinite regress of errors when the BUTLAST function was
753 executed with a LIST=NIL which would cause LENGTH to be -1.
754 I fixed the bogus declaration, but I should come back and see
755 whether the system's inability to recover from the bogus declaration
756 (by signalling a TYPE-ERROR and dropping into the debugger) was
757 a compiler problem which remains to be fixed, or one of the
758 unrelated infinite-regress-errors problems, many related to
759 revised signal handling, which were fixed around the same time.
762 (SUBTYPEP '(AND ZILCH INTEGER) 'ZILCH)