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. ANSI justifies this specification by saying
54 The restriction against issuing a warning for type mismatches
55 between a slot-initform and the corresponding slot's :TYPE
56 option is necessary because a slot-initform must be specified
57 in order to specify slot options; in some cases, no suitable
59 In SBCL, as in CMU CL (or, for that matter, any compiler which
60 really understands Common Lisp types) a suitable default does
61 exist, in all cases, because the compiler understands the concept
62 of functions which never return (i.e. has return type NIL, e.g.
63 ERROR). Thus, as a portable workaround, you can use a call to
64 some known-never-to-return function as the default. E.g.
66 (BAR (ERROR "missing :BAR argument")
67 :TYPE SOME-TYPE-TOO-HAIRY-TO-CONSTRUCT-AN-INSTANCE-OF))
69 (DECLAIM (FTYPE () NIL) MISSING-ARG)
70 (DEFUN REQUIRED-ARG () ; workaround for SBCL non-ANSI slot init typing
71 (ERROR "missing required argument"))
73 (BAR (REQUIRED-ARG) :TYPE TRICKY-TYPE-OF-SOME-SORT)
74 (BLETCH (REQUIRED-ARG) :TYPE TRICKY-TYPE-OF-SOME-SORT)
75 (N-REFS-SO-FAR 0 :TYPE (INTEGER 0)))
76 Such code will compile without complaint and work correctly either
77 on SBCL or on a completely compliant Common Lisp system.
80 bogus warnings about undefined functions for magic functions like
81 SB!C::%%DEFUN and SB!C::%DEFCONSTANT when cross-compiling files
82 like src/code/float.lisp. Fixing this will probably require
83 straightening out enough bootstrap consistency issues that
84 the cross-compiler can run with *TYPE-SYSTEM-INITIALIZED*.
85 Instead, the cross-compiler runs in a slightly flaky state
86 which is sane enough to compile SBCL itself, but which is
87 also unstable in several ways, including its inability
88 to really grok function declarations.
91 The "byte compiling top-level form:" output ought to be condensed.
92 Perhaps any number of such consecutive lines ought to turn into a
93 single "byte compiling top-level forms:" line.
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 from DTC on the CMU CL mailing list 25 Feb 2000:
178 ;;; Compiler fails when this file is compiled.
180 ;;; Problem shows up in delete-block within ir1util.lisp. The assertion
181 ;;; (assert (member (functional-kind lambda) '(:let :mv-let :assignment)))
182 ;;; fails within bind node branch.
184 ;;; Note that if c::*check-consistency* is enabled then an un-reached
185 ;;; entry is also reported.
188 (declare (values nil))
205 (let ((ttt #'(lambda () (go cccc))))
206 (declare (special ttt))
207 (return-from bbbb nil))
210 (return-from bbbb nil))))))
213 (I *think* this is a bug. It certainly seems like strange behavior. But
214 the ANSI spec is scary, dark, and deep..)
215 (FORMAT NIL "~,1G" 1.4) => "1. "
216 (FORMAT NIL "~3,1G" 1.4) => "1. "
219 from Marco Antoniotti on cmucl-imp mailing list 1 Mar 2000:
221 (setf (find-class 'ccc1) (find-class 'ccc))
222 (defmethod zut ((c ccc1)) 123)
223 DTC's recommended workaround from the mailing list 3 Mar 2000:
224 (setf (pcl::find-class 'ccc1) (pcl::find-class 'ccc))
227 The ANSI spec, in section "22.3.5.2 Tilde Less-Than-Sign: Logical Block",
228 says that an error is signalled if ~W, ~_, ~<...~:>, ~I, or ~:T is used
229 inside "~<..~>" (without the colon modifier on the closing syntax).
230 However, SBCL doesn't do this:
231 * (FORMAT T "~<munge~wegnum~>" 12)
236 When too many files are opened, OPEN will fail with an
237 uninformative error message
238 error in function OPEN: error opening #P"/tmp/foo.lisp": NIL
239 instead of saying that too many files are open.
242 Right now, when COMPILE-FILE has a read error, it actually pops
243 you into the debugger before giving up on the file. It should
244 instead handle the error, perhaps issuing (and handling)
245 a secondary error "caught ERROR: unrecoverable error during compilation"
246 and then return with FAILURE-P true,
249 from CMU CL mailing list 01 May 2000
251 I realize I can take care of this by doing (proclaim (ignore pcl::.slots1.))
252 but seeing as .slots0. is not-exported, shouldn't it be ignored within the
256 In: DEFMETHOD FOO-BAR-BAZ (RESOURCE-TYPE)
257 (DEFMETHOD FOO-BAR-BAZ
258 ((SELF RESOURCE-TYPE))
259 (SETF (SLOT-VALUE SELF 'NAME) 3))
260 --> BLOCK MACROLET PCL::FAST-LEXICAL-METHOD-FUNCTIONS
261 --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET
262 --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS LET PCL::BIND-ARGS LET* PCL::PV-BINDING
263 --> PCL::PV-BINDING1 PCL::PV-ENV LET
265 (LET ((PCL::.SLOTS0. #))
270 Warning: Variable PCL::.SLOTS0. defined but never used.
272 Compilation unit finished.
275 #<Standard-Method FOO-BAR-BAZ (RESOURCE-TYPE) {480918FD}>
278 reported by Sam Steingold on the cmucl-imp mailing list 12 May 2000:
280 Also, there is another bug: `array-displacement' should return an array
281 or nil as first value (as per ANSI CL), while CMUCL declares it as
282 returning an array as first value always.
285 Sometimes (SB-EXT:QUIT) fails with
286 Argh! maximum interrupt nesting depth (4096) exceeded, exiting
287 Process inferior-lisp exited abnormally with code 1
288 I haven't noticed a repeatable case of this yet.
291 some sort of bug in inlining and RETURN-FROM in sbcl-0.6.5: Compiling
294 (BLOCK USED-BY-SOME-Y?
297 (UNLESS (REJECTED? Y)
298 (RETURN-FROM USED-BY-SOME-Y? T)))))
299 (DECLARE (INLINE FROB))
304 error in function SB-KERNEL:ASSERT-ERROR:
305 The assertion (EQ (SB-C::CONTINUATION-KIND SB-C::CONT) :BLOCK-START) failed.
306 This is still present in sbcl-0.6.8.
309 In some cases the compiler believes type declarations on array
310 elements without checking them, e.g.
311 (DECLAIM (OPTIMIZE (SAFETY 3) (SPEED 1) (SPACE 1)))
314 (DECLARE (TYPE (SIMPLE-ARRAY CONS 1) X))
315 (WHEN (CONSP (AREF X 0))
317 (BAR (VECTOR (MAKE-FOO :A 11 :B 12)))
320 in SBCL 0.6.5 (and also in CMU CL 18b). This does not happen for
321 all cases, e.g. the type assumption *is* checked if the array
322 elements are declared to be of some structure type instead of CONS.
325 The printer doesn't report closures very well. This is true in
329 #<Closure Over Function "DEFUN STRUCTURE-SLOT-ACCESSOR" {134D1A1}>
330 It would be nice to make closures have a settable name slot,
331 and make things like DEFSTRUCT and FLET, which create closures,
332 set helpful values into this slot.
335 And as long as we're wishing, it would be awfully nice if INSPECT could
336 also report on closures, telling about the values of the bound variables.
339 WHN test case: Compile this file:
340 (eval-when (:compile-toplevel :load-toplevel :execute)
341 (defclass a-class () (a)))
342 (defconstant +a-constant+ (make-instance 'a-class))
343 (defconstant +another-constant+ (vector +a-constant+))
344 as reported by Robert Strandh on the CMU CL mailing list 12 Jun 2000:
346 (defconstant +a-constant+ (make-instance 'a-class))
347 (defconstant +another-constant+ (vector +a-constant+))
349 CMU Common Lisp release x86-linux 2.4.19 8 February 2000 build 456,
352 Send bug reports and questions to your local CMU CL maintainer,
353 or to pvaneynd@debian.org
354 or to cmucl-help@cons.org. (prefered)
355 type (help) for help, (quit) to exit, and (demo) to see the demos
357 Python 1.0, target Intel x86
358 CLOS based on PCL version: September 16 92 PCL (f)
359 * (defclass a-class () ())
360 #<STANDARD-CLASS A-CLASS {48027BD5}>
361 * (compile-file "xx.lisp")
362 Python version 1.0, VM version Intel x86 on 12 JUN 00 08:12:55 am.
364 /home/strandh/Research/Functional/Common-Lisp/CLIM/Development/McCLIM
365 /xx.lisp 12 JUN 00 07:47:14 am
366 Compiling Load Time Value of (PCL::GET-MAKE-INSTANCE-FUNCTION-SYMBOL
368 Byte Compiling Top-Level Form:
369 Error in function C::DUMP-STRUCTURE: Attempt to dump invalid
371 #<A-CLASS {4803A5B5}>
375 The compiler assumes that any time a function of declared FTYPE
376 doesn't signal an error, its arguments were of the declared type.
377 E.g. compiling and loading
378 (DECLAIM (OPTIMIZE (SAFETY 3)))
379 (DEFUN FACTORIAL (X) (GAMMA (1+ X)))
380 (DECLAIM (FTYPE (FUNCTION (UNSIGNED-BYTE) FACTORIAL)))
382 (COND ((> (FACTORIAL X) 1.0E6)
383 (FORMAT T "too big~%"))
385 (FORMAT T "exactly ~S~%" (FACTORIAL X)))
387 (FORMAT T "approximately ~S~%" (FACTORIAL X)))))
390 will cause the INTEGERP case to be selected, giving bogus output a la
392 This violates the "declarations are assertions" principle.
393 According to the ANSI spec, in the section "System Class FUNCTION",
394 this is a case of "lying to the compiler", but the lying is done
395 by the code which calls FACTORIAL with non-UNSIGNED-BYTE arguments,
396 not by the unexpectedly general definition of FACTORIAL. In any case,
397 "declarations are assertions" means that lying to the compiler should
398 cause an error to be signalled, and should not cause a bogus
399 result to be returned. Thus, the compiler should not assume
400 that arbitrary functions check their argument types. (It might
401 make sense to add another flag (CHECKED?) to DEFKNOWN to
402 identify functions which *do* check their argument types.)
405 DEFMETHOD doesn't check the syntax of &REST argument lists properly,
406 accepting &REST even when it's not followed by an argument name:
407 (DEFMETHOD FOO ((X T) &REST) NIL)
410 TYPEP of VALUES types is sometimes implemented very inefficiently, e.g. in
411 (DEFTYPE INDEXOID () '(INTEGER 0 1000))
413 (DECLARE (TYPE INDEXOID X))
414 (THE (VALUES INDEXOID)
416 where the implementation of the type check in function FOO
417 includes a full call to %TYPEP. There are also some fundamental problems
418 with the interpretation of VALUES types (inherited from CMU CL, and
419 from the ANSI CL standard) as discussed on the cmucl-imp@cons.org
420 mailing list, e.g. in Robert Maclachlan's post of 21 Jun 2000.
423 The definitions of SIGCONTEXT-FLOAT-REGISTER and
424 %SET-SIGCONTEXT-FLOAT-REGISTER in x86-vm.lisp say they're not
425 supported on FreeBSD because the floating point state is not saved,
426 but at least as of FreeBSD 4.0, the floating point state *is* saved,
427 so they could be supported after all. Very likely
428 SIGCONTEXT-FLOATING-POINT-MODES could now be supported, too.
431 (as discussed by Douglas Crosher on the cmucl-imp mailing list ca.
432 Aug. 10, 2000): CMUCL currently interprets 'member as '(member); same
433 issue with 'union, 'and, 'or etc. So even though according to the
434 ANSI spec, bare 'MEMBER, 'AND, and 'OR are not legal types, CMUCL
435 (and now SBCL) interpret them as legal types.
438 ANSI specifies DEFINE-SYMBOL-MACRO, but it's not defined in SBCL.
439 CMU CL added it ca. Aug 13, 2000, after some discussion on the mailing
440 list, and it is probably possible to use substantially the same
441 patches to add it to SBCL.
444 a slew of floating-point-related errors reported by Peter Van Eynde
446 a: (SQRT -9.0) fails, because SB-KERNEL::COMPLEX-SQRT is undefined.
447 Similarly, COMPLEX-ASIN, COMPLEX-ACOS, COMPLEX-ACOSH, and others
449 b: SBCL's value for LEAST-POSITIVE-SHORT-FLOAT is bogus, and
450 should probably be 1.4012985e-45. In SBCL,
451 (/ LEAST-POSITIVE-SHORT-FLOAT 2) returns a number smaller
452 than LEAST-POSITIVE-SHORT-FLOAT. Similar problems
453 exist for LEAST-NEGATIVE-SHORT-FLOAT, LEAST-POSITIVE-LONG-FLOAT,
454 and LEAST-NEGATIVE-LONG-FLOAT.
455 c: Many expressions generate floating infinity:
460 PVE's regression tests want them to raise errors. SBCL
461 generates the infinities instead, which may or may not be
462 conforming behavior, but then blow it by being unable to
463 output the infinities, since support for infinities is generally
464 broken, and in particular SB-IMPL::OUTPUT-FLOAT-INFINITY is
466 d: (in section12.erg) various forms a la
467 (FLOAT 1 DOUBLE-FLOAT-EPSILON)
468 don't give the right behavior.
471 type safety errors reported by Peter Van Eynde July 25, 2000:
472 a: (COERCE (QUOTE (A B C)) (QUOTE (VECTOR * 4)))
474 In general lengths of array type specifications aren't
475 checked by COERCE, so it fails when the spec is
476 (VECTOR 4), (STRING 2), (SIMPLE-BIT-VECTOR 3), or whatever.
477 b: CONCATENATE has the same problem of not checking the length
478 of specified output array types. MAKE-SEQUENCE and MAP and
479 MERGE also have the same problem.
480 c: (COERCE 'AND 'FUNCTION) returns something related to
481 (MACRO-FUNCTION 'AND), but ANSI says it should raise an error.
482 d: ELT signals SIMPLE-ERROR if its index argument
483 isn't a valid index for its sequence argument, but should
484 signal TYPE-ERROR instead.
485 e: FILE-LENGTH is supposed to signal a type error when its
486 argument is not a stream associated with a file, but doesn't.
487 f: (FLOAT-RADIX 2/3) should signal an error instead of
489 g: (LOAD "*.lsp") should signal FILE-ERROR.
490 h: (MAKE-CONCATENATED-STREAM (MAKE-STRING-OUTPUT-STREAM))
491 should signal TYPE-ERROR.
492 i: MAKE-TWO-WAY-STREAM doesn't check that its arguments can
493 be used for input and output as needed. It should fail with
494 TYPE-ERROR when handed e.g. the results of
495 MAKE-STRING-INPUT-STREAM or MAKE-STRING-OUTPUT-STREAM in
496 the inappropriate positions, but doesn't.
497 j: (PARSE-NAMESTRING (COERCE (LIST #\f #\o #\o (CODE-CHAR 0) #\4 #\8)
499 should probably signal an error instead of making a pathname with
501 k: READ-BYTE is supposed to signal TYPE-ERROR when its argument is
502 not a binary input stream, but instead cheerfully reads from
503 character streams, e.g. (MAKE-STRING-INPUT-STREAM "abc").
506 DEFCLASS bugs reported by Peter Van Eynde July 25, 2000:
507 a: (DEFCLASS FOO () (A B A)) should signal a PROGRAM-ERROR, and
509 b: (DEFCLASS FOO () (A B A) (:DEFAULT-INITARGS X A X B)) should
510 signal a PROGRAM-ERROR, and doesn't.
511 c: (DEFCLASS FOO07 NIL ((A :ALLOCATION :CLASS :ALLOCATION :CLASS))),
512 and other DEFCLASS forms with duplicate specifications in their
513 slots, should signal a PROGRAM-ERROR, and doesn't.
514 d: (DEFGENERIC IF (X)) should signal a PROGRAM-ERROR, but instead
515 causes a COMPILER-ERROR.
518 SYMBOL-MACROLET bugs reported by Peter Van Eynde July 25, 2000:
519 a: (SYMBOL-MACROLET ((T TRUE)) ..) should probably signal
520 PROGRAM-ERROR, but SBCL accepts it instead.
521 b: SYMBOL-MACROLET should refuse to bind something which is
522 declared as a global variable, signalling PROGRAM-ERROR.
523 c: SYMBOL-MACROLET should signal PROGRAM-ERROR if something
524 it binds is declared SPECIAL inside.
527 LOOP bugs reported by Peter Van Eynde July 25, 2000:
528 a: (LOOP WITH (A B) DO (PRINT 1)) is a syntax error according to
529 the definition of WITH clauses given in the ANSI spec, but
530 compiles and runs happily in SBCL.
531 b: a messy one involving package iteration:
532 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<))))
533 Should be: (("blah" "blah2") ("blah2"))
534 SBCL: (("blah") ("blah2"))
535 * (LET ((X 1)) (LOOP FOR I BY (INCF X) FROM X TO 10 COLLECT I))
536 doesn't work -- SBCL's LOOP says BY isn't allowed in a FOR clause.
539 type system errors reported by Peter Van Eynde July 25, 2000:
540 a: (SUBTYPEP 'BIGNUM 'INTEGER) => NIL, NIL
541 but should be (VALUES T T) instead.
542 b: (SUBTYPEP 'EXTENDED-CHAR 'CHARACTER) => NIL, NIL
543 but should be (VALUES T T) instead.
544 c: (SUBTYPEP '(INTEGER (0) (0)) 'NIL) dies with nested errors.
545 d: In general, the system doesn't like '(INTEGER (0) (0)) -- it
546 blows up at the level of SPECIFIER-TYPE with
547 "Lower bound (0) is greater than upper bound (0)." Probably
548 SPECIFIER-TYPE should return NIL instead.
549 e: (TYPEP 0 '(COMPLEX (EQL 0)) fails with
550 "Component type for Complex is not numeric: (EQL 0)."
551 This might be easy to fix; the type system already knows
552 that (SUBTYPEP '(EQL 0) 'NUMBER) is true.
553 f: The type system doesn't know about the condition system,
554 so that e.g. (TYPEP 'SIMPLE-ERROR 'ERROR)=>NIL.
555 g: The type system isn't all that smart about relationships
556 between hairy types, as shown in the type.erg test results,
557 e.g. (SUBTYPEP 'CONS '(NOT ATOM)) => NIL, NIL.
560 miscellaneous errors reported by Peter Van Eynde July 25, 2000:
562 (DEFGENERIC FOO02 (X))
563 (DEFMETHOD FOO02 ((X NUMBER)) T)
564 (LET ((M (FIND-METHOD (FUNCTION FOO02)
566 (LIST (FIND-CLASS (QUOTE NUMBER))))))
567 (REMOVE-METHOD (FUNCTION FOO02) M)
568 (DEFGENERIC FOO03 (X))
569 (ADD-METHOD (FUNCTION FOO03) M)))
570 should give an error, but SBCL allows it.
571 b: READ should probably return READER-ERROR, not the bare
572 arithmetic error, when input a la "1/0" or "1e1000" causes
576 It has been reported (e.g. by Peter Van Eynde) that there are
577 several metaobject protocol "errors". (In order to fix them, we might
578 need to document exactly what metaobject protocol specification
579 we're following -- the current code is just inherited from PCL.)
582 another error from Peter Van Eynde 5 September 2000:
583 (FORMAT NIL "~F" "FOO") should work, but instead reports an error.
584 PVE submitted a patch to deal with this bug, but it exposes other
585 comparably serious bugs, so I didn't apply it. It looks as though
586 the FORMAT code needs a fair amount of rewriting in order to comply
587 with the various details of the ANSI spec.
590 The implementation of #'+ returns its single argument without
591 type checking, e.g. (+ "illegal") => "illegal".
594 In sbcl-0.6.7, there is no doc string for CL:PUSH, probably
595 because it's defined with the DEFMACRO-MUNDANELY macro and something
596 is wrong with doc string setting in that macro.
599 Attempting to use COMPILE on something defined by DEFMACRO fails:
600 (DEFMACRO FOO (X) (CONS X X))
602 Error in function C::GET-LAMBDA-TO-COMPILE:
603 #<Closure Over Function "DEFUN (SETF MACRO-FUNCTION)" {480E21B1}> was defined in a non-null environment.
606 (SUBTYPEP '(AND ZILCH INTEGER) 'ZILCH)
610 CL:*DEFAULT-PATHNAME-DEFAULTS* doesn't behave as ANSI suggests (reflecting
611 current working directory). And there's no supported way to update
612 or query the current working directory (a la Unix "chdir" and "pwd"),
613 which is functionality that ILISP needs (and currently gets with low-level
617 The debugger LIST-LOCATIONS command doesn't work properly.
620 Compiling and loading
621 (DEFUN FAIL (X) (THROW 'FAIL-TAG X))
623 then requesting a BACKTRACE at the debugger prompt gives no information
624 about where in the user program the problem occurred.
627 The compiler is supposed to do type inference well enough that
630 ((SIMPLE-ARRAY SINGLE-FLOAT)
632 (DECLARE (TYPE (SIMPLE-ARRAY SINGLE-FLOAT) X))
635 is redundant. However, as reported by Juan Jose Garcia Ripoll for
636 CMU CL, it sometimes doesn't. Adding declarations is a pretty good
637 workaround for the problem for now, but can't be done by the TYPECASE
638 macros themselves, since it's too hard for the macro to detect
639 assignments to the variable within the clause.
640 Note: The compiler *is* smart enough to do the type inference in
641 many cases. This case, derived from a couple of MACROEXPAND-1
642 calls on Ripoll's original test case,
644 (DECLARE (OPTIMIZE SPEED (SAFETY 0)))
645 (COND ((TYPEP A '(SIMPLE-ARRAY SINGLE-FLOAT)) NIL
646 (LET ((LENGTH (ARRAY-TOTAL-SIZE A)))
647 (LET ((I 0) (G2554 LENGTH))
648 (DECLARE (TYPE REAL G2554) (TYPE REAL I))
651 (WHEN (>= I G2554) (GO SB-LOOP::END-LOOP))
652 (SETF (ROW-MAJOR-AREF A I) (- (ROW-MAJOR-AREF A I)))
653 (GO SB-LOOP::NEXT-LOOP)
654 SB-LOOP::END-LOOP))))))
655 demonstrates the problem; but the problem goes away if the TAGBODY
656 and GO forms are removed (leaving the SETF in ordinary, non-looping
657 code), or if the TAGBODY and GO forms are retained, but the
658 assigned value becomes 0.0 instead of (- (ROW-MAJOR-AREF A I)).
661 Paul Werkowski wrote on cmucl-imp@cons.org 2000-11-15
662 I am looking into this problem that showed up on the cmucl-help
663 list. It seems to me that the "implementation specific environment
664 hacking functions" found in pcl/walker.lisp are completely messed
665 up. The good thing is that they appear to be barely used within
666 PCL and the munged environment object is passed to cmucl only
667 in calls to macroexpand-1, which is probably why this case fails.
668 SBCL uses essentially the same code, so if the environment hacking
669 is screwed up, it affects us too.
672 Using the pretty-printer from the command prompt gives funny
673 results, apparently because the pretty-printer doesn't know
674 about user's command input, including the user's carriage return
675 that the user, and therefore the pretty-printer thinks that
676 the new output block should start indented 2 or more characters
677 rightward of the correct location.
680 (probably related to bug #70)
681 As reported by Carl Witty on submit@bugs.debian.org 1999-05-08,
683 (in-package "CL-USER")
684 (defun equal-terms (termx termy)
686 ((alpha-equal-bound-term-lists (listx listy)
687 (or (and (null listx) (null listy))
689 (let ((bindings-x (bindings-of-bound-term (car listx)))
690 (bindings-y (bindings-of-bound-term (car listy))))
691 (if (and (null bindings-x) (null bindings-y))
692 (alpha-equal-terms (term-of-bound-term (car listx))
693 (term-of-bound-term (car listy)))
694 (and (= (length bindings-x) (length bindings-y))
696 (enter-binding-pairs (bindings-of-bound-term (car listx))
697 (bindings-of-bound-term (car listy)))
698 (alpha-equal-terms (term-of-bound-term (car listx))
699 (term-of-bound-term (car listy)))
700 (exit-binding-pairs (bindings-of-bound-term (car listx))
701 (bindings-of-bound-term (car listy)))))))
702 (alpha-equal-bound-term-lists (cdr listx) (cdr listy)))))
704 (alpha-equal-terms (termx termy)
705 (if (and (variable-p termx)
707 (equal-bindings (id-of-variable-term termx)
708 (id-of-variable-term termy))
709 (and (equal-operators-p (operator-of-term termx) (operator-of-term termy))
710 (alpha-equal-bound-term-lists (bound-terms-of-term termx)
711 (bound-terms-of-term termy))))))
715 (with-variable-invocation (alpha-equal-terms termx termy))))))
716 causes an assertion failure
717 The assertion (EQ (C::LAMBDA-TAIL-SET C::CALLER)
718 (C::LAMBDA-TAIL-SET (C::LAMBDA-HOME C::CALLEE))) failed.
720 Bob Rogers reports (1999-07-28 on cmucl-imp@cons.org) a smaller test
721 case with the same problem:
722 (defun parse-fssp-alignment ()
723 ;; Given an FSSP alignment file named by the argument . . .
724 (labels ((get-fssp-char ()
728 ;; Stub body, enough to tickle the bug.
729 (list (read-fssp-char)
733 ANSI specifies that the RESULT-TYPE argument of CONCATENATE must be
734 a subtype of SEQUENCE, but CONCATENATE doesn't check this properly:
735 (CONCATENATE 'SIMPLE-ARRAY #(1 2) '(3)) => #(1 2 3)
736 This also leads to funny behavior when derived type specifiers
737 are used, as originally reported by Milan Zamazal for CMU CL (on the
738 Debian bugs mailing list (?) 2000-02-27), then reported by Martin
739 Atzmueller for SBCL (2000-10-01 on sbcl-devel@lists.sourceforge.net):
740 (DEFTYPE FOO () 'SIMPLE-ARRAY)
741 (CONCATENATE 'FOO #(1 2) '(3))
742 => #<ARRAY-TYPE SIMPLE-ARRAY> is a bad type specifier for
744 The derived type specifier FOO should act the same way as the
745 built-in type SIMPLE-ARRAY here, but it doesn't. That problem
746 doesn't seem to exist for sequence types:
747 (DEFTYPE BAR () 'SIMPLE-VECTOR)
748 (CONCATENATE 'BAR #(1 2) '(3)) => #(1 2 3)
751 As reported by Winton Davies on a CMU CL mailing list 2000-01-10,
752 and reported for SBCL by Martin Atzmueller 2000-10-20: (TRACE GETHASH)
753 crashes SBCL. In general tracing anything which is used in the
754 implementation of TRACE is likely to have the same problem.
757 As reported by Daniel Solaz on cmucl-help@cons.org 2000-11-23,
758 SXHASH returns the same value for all non-STRUCTURE-OBJECT instances,
759 notably including all PCL instances. There's a limit to how much
760 SXHASH can do to return unique values for instances, but at least
761 it should probably look at the class name, the way that it does
762 for STRUCTURE-OBJECTs.
765 As reported by Martin Atzmueller on the sbcl-devel list 2000-11-22,
766 > There remains one issue, that is a bug in SBCL:
767 > According to my interpretation of the spec, the ":" and "@" modifiers
768 > should appear _after_ the comma-seperated arguments.
769 > Well, SBCL (and CMUCL for that matter) accept
770 > (ASSERT (STRING= (FORMAT NIL "~:8D" 1) " 1"))
771 > where the correct way (IMHO) should be
772 > (ASSERT (STRING= (FORMAT NIL "~8:D" 1) " 1"))
773 Probably SBCL should stop accepting the "~:8D"-style format arguments,
774 or at least issue a warning.
777 (probably related to bug #65)
778 The compiler doesn't like &OPTIONAL arguments in LABELS and FLET
780 (DEFUN FIND-BEFORE (ITEM SEQUENCE &KEY (TEST #'EQL))
781 (LABELS ((FIND-ITEM (OBJ SEQ TEST &OPTIONAL (VAL NIL))
782 (LET ((ITEM (FIRST SEQ)))
785 ((FUNCALL TEST OBJ ITEM)
788 (FIND-ITEM OBJ (REST SEQ) TEST (NCONC VAL `(,ITEM))))))))
789 (FIND-ITEM ITEM SEQUENCE TEST)))
790 from David Young's bug report on cmucl-help@cons.org 30 Nov 2000
791 causes sbcl-0.6.9 to fail with
792 error in function SB-KERNEL:ASSERT-ERROR:
793 The assertion (EQ (SB-C::LAMBDA-TAIL-SET SB-C::CALLER)
794 (SB-C::LAMBDA-TAIL-SET
795 (SB-C::LAMBDA-HOME SB-C::CALLEE))) failed.
798 (DECLAIM (OPTIMIZE ..)) doesn't work. E.g. even after
799 (DECLAIM (OPTIMIZE (SPEED 3))), things are still optimized with
800 the previous SPEED policy. This bug will probably get fixed in
801 0.6.9.x in a general cleanup of optimization policy.
804 (DECLAIM (OPTIMIZE ..)) doesn't work properly inside LOCALLY forms.
807 As noted in the ANSI specification for COERCE, (COERCE 3 'COMPLEX)
808 gives a result which isn't COMPLEX. The result type optimizer
809 for COERCE doesn't know this, perhaps because it was written before
810 ANSI threw this curveball: the optimizer thinks that COERCE always
811 returns a result of the specified type. Thus while the interpreted
813 (DEFUN TRICKY (X) (TYPEP (COERCE X 'COMPLEX) 'COMPLEX))
814 returns the correct result,
816 the compiled function
822 As reported by Martin Atzmueller on sbcl-devel 26 Dec 2000,
823 ANSI says that WITH-OUTPUT-TO-STRING should have a keyword
824 :ELEMENT-TYPE, but in sbcl-0.6.9 this is not defined for
825 WITH-OUTPUT-TO-STRING.
828 ANSI says in one place that type declarations can be abbreviated even
829 when the type name is not a symbol, e.g.
830 (DECLAIM ((VECTOR T) *FOOVECTOR*))
831 SBCL doesn't support this. But ANSI says in another place that this
832 isn't allowed. So it's not clear this is a bug after all. (See the
833 e-mail on cmucl-help@cons.org on 2001-01-16 and 2001-01-17 from WHN
837 as pointed out by Dan Barlow on sbcl-devel 2000-07-02:
838 The PICK-TEMPORARY-FILE-NAME utility used by LOAD-FOREIGN uses
839 an easily guessable temporary filename in a way which might open
840 applications using LOAD-FOREIGN to hijacking by malicious users
841 on the same machine. Incantations for doing this safely are
842 floating around the net in various "how to write secure programs
843 despite Unix" documents, and it would be good to (1) fix this in
844 LOAD-FOREIGN, and (2) hunt for any other code which uses temporary
845 files and make it share the same new safe logic.
848 The subtle CMU CL bug discussed by Douglas Thomas Crosher on
849 cmucl-imp@cons.org 29 Jan 2001 sounds like something that probably
850 still exists in the corresponding SBCL code.
853 KNOWN BUGS RELATED TO THE IR1 INTERPRETER
855 (Note: At some point, the pure interpreter (actually a semi-pure
856 interpreter aka "the IR1 interpreter") will probably go away, replaced
858 (DEFUN EVAL (X) (FUNCALL (COMPILE NIL (LAMBDA ..)))))
859 and at that time these bugs should either go away automatically or
860 become more tractable to fix. Until then, they'll probably remain,
861 since some of them aren't considered urgent, and the rest are too hard
862 to fix as long as so many special cases remain. After the IR1
863 interpreter goes away is also the preferred time to start
864 systematically exterminating cases where debugging functionality
865 (backtrace, breakpoint, etc.) breaks down, since getting rid of the
866 IR1 interpreter will reduce the number of special cases we need to
870 The FUNCTION special operator doesn't check properly whether its
871 argument is a function name. E.g. (FUNCTION (X Y)) returns a value
872 instead of failing with an error. (Later attempting to funcall the
873 value does cause an error.)
876 COMPILED-FUNCTION-P bogusly reports T for interpreted functions:
877 * (DEFUN FOO (X) (- 12 X))
879 * (COMPILED-FUNCTION-P #'FOO)
884 (DEFVAR *SUPPRESS-P* T)
885 (EVAL '(UNLESS *SUPPRESS-P*
886 (EVAL-WHEN (:COMPILE-TOPLEVEL :LOAD-TOPLEVEL :EXECUTE)
887 (FORMAT T "surprise!"))))
888 prints "surprise!". Probably the entire EVAL-WHEN mechanism ought to be
889 rewritten from scratch to conform to the ANSI definition, abandoning
890 the *ALREADY-EVALED-THIS* hack which is used in sbcl-0.6.8.9 (and
891 in the original CMU CL source, too). This should be easier to do --
892 though still nontrivial -- once the various IR1 interpreter special
896 EVAL-WHEN's idea of what's a toplevel form is even more screwed up
897 than the example in IR1-3 would suggest, since COMPILE-FILE and
898 COMPILE both print both "right now!" messages when compiling the
902 (EVAL-WHEN (:COMPILE-TOPLEVEL :LOAD-TOPLEVEL :EXECUTE)
903 (PRINT "yes! right now!"))
906 (EVAL-WHEN (:COMPILE-TOPLEVEL :LOAD-TOPLEVEL :EXECUTE)
907 (PRINT "no! right now!"))
909 and while EVAL doesn't print the "right now!" messages, the first
910 FUNCALL on the value returned by EVAL causes both of them to be printed.
913 The system accepts DECLAIM in most places where DECLARE would be
914 accepted, without even issuing a warning. ANSI allows this, but since
915 it's fairly easy to mistype DECLAIM instead of DECLARE, and the
916 meaning is rather different, and it's unlikely that the user
917 has a good reason for doing DECLAIM not at top level, it would be
918 good to issue a STYLE-WARNING when this happens. A possible
919 fix would be to issue STYLE-WARNINGs for DECLAIMs not at top level,
920 or perhaps to issue STYLE-WARNINGs for any EVAL-WHEN not at top level.
921 [This is considered an IR1-interpreter-related bug because until
922 EVAL-WHEN is rewritten, which won't happen until after the IR1
923 interpreter is gone, the system's notion of what's a top-level form
924 and what's not will remain too confused to fix this problem.]