3 Bugs can be reported on the help mailing list
4 sbcl-help@lists.sourceforge.net
5 or on the development mailing list
6 sbcl-devel@lists.sourceforge.net
8 Please include enough information in a bug report that someone reading
9 it can reproduce the problem, i.e. don't write
10 Subject: apparent bug in PRINT-OBJECT (or *PRINT-LENGTH*?)
11 PRINT-OBJECT doesn't seem to work with *PRINT-LENGTH*. Is this a bug?
13 Subject: apparent bug in PRINT-OBJECT (or *PRINT-LENGTH*?)
14 In sbcl-1.2.3 running under OpenBSD 4.5 on my Alpha box, when
15 I compile and load the file
16 (DEFSTRUCT (FOO (:PRINT-OBJECT (LAMBDA (X Y)
17 (LET ((*PRINT-LENGTH* 4))
20 then at the command line type
22 the program loops endlessly instead of printing the object.
27 There is also some information on bugs in the manual page and
28 in the TODO file. Eventually more such information may move here.
30 The gaps in the number sequence belong to old bug descriptions which
31 have gone away (typically because they were fixed, but sometimes for
32 other reasons, e.g. because they were moved elsewhere).
35 KNOWN BUGS OF NO SPECIAL CLASS:
38 DEFSTRUCT almost certainly should overwrite the old LAYOUT information
39 instead of just punting when a contradictory structure definition
40 is loaded. As it is, if you redefine DEFSTRUCTs in a way which
41 changes their layout, you probably have to rebuild your entire
42 program, even if you know or guess enough about the internals of
43 SBCL to wager that this (undefined in ANSI) operation would be safe.
46 ANSI specifies that a type mismatch in a structure slot
47 initialization value should not cause a warning.
49 This one might not be fixed for a while because while we're big
50 believers in ANSI compatibility and all, (1) there's no obvious
51 simple way to do it (short of disabling all warnings for type
52 mismatches everywhere), and (2) there's a good portable
53 workaround, and (3) by their own reasoning, it looks as though
54 ANSI may have gotten it wrong. ANSI justifies this specification
56 The restriction against issuing a warning for type mismatches
57 between a slot-initform and the corresponding slot's :TYPE
58 option is necessary because a slot-initform must be specified
59 in order to specify slot options; in some cases, no suitable
61 However, in SBCL (as in CMU CL or, for that matter, any compiler
62 which really understands Common Lisp types) a suitable default
63 does exist, in all cases, because the compiler understands the
64 concept of functions which never return (i.e. has return type NIL).
65 Thus, as a portable workaround, you can use a call to some
66 known-never-to-return function as the default. E.g.
68 (BAR (ERROR "missing :BAR argument")
69 :TYPE SOME-TYPE-TOO-HAIRY-TO-CONSTRUCT-AN-INSTANCE-OF))
71 (DECLAIM (FTYPE (FUNCTION () NIL) MISSING-ARG))
72 (DEFUN REQUIRED-ARG () ; workaround for SBCL non-ANSI slot init typing
73 (ERROR "missing required argument"))
75 (BAR (REQUIRED-ARG) :TYPE TRICKY-TYPE-OF-SOME-SORT)
76 (BLETCH (REQUIRED-ARG) :TYPE TRICKY-TYPE-OF-SOME-SORT)
77 (N-REFS-SO-FAR 0 :TYPE (INTEGER 0)))
78 Such code should compile without complaint and work correctly either
79 on SBCL or on any other completely compliant Common Lisp system.
82 bogus warnings about undefined functions for magic functions like
83 SB!C::%%DEFUN and SB!C::%DEFCONSTANT when cross-compiling files
84 like src/code/float.lisp. Fixing this will probably require
85 straightening out enough bootstrap consistency issues that
86 the cross-compiler can run with *TYPE-SYSTEM-INITIALIZED*.
87 Instead, the cross-compiler runs in a slightly flaky state
88 which is sane enough to compile SBCL itself, but which is
89 also unstable in several ways, including its inability
90 to really grok function declarations.
92 As of sbcl-0.7.5, sbcl's cross-compiler does run with
93 *TYPE-SYSTEM-INITIALIZED*; however, this bug remains.
96 The "compiling top-level form:" output ought to be condensed.
97 Perhaps any number of such consecutive lines ought to turn into a
98 single "compiling top-level forms:" line.
101 The way that the compiler munges types with arguments together
102 with types with no arguments (in e.g. TYPE-EXPAND) leads to
103 weirdness visible to the user:
104 (DEFTYPE FOO () 'FIXNUM)
106 (TYPEP 11 '(FOO)) => T, which seems weird
107 (TYPEP 11 'FIXNUM) => T
108 (TYPEP 11 '(FIXNUM)) signals an error, as it should
109 The situation is complicated by the presence of Common Lisp types
110 like UNSIGNED-BYTE (which can either be used in list form or alone)
111 so I'm not 100% sure that the behavior above is actually illegal.
112 But I'm 90+% sure, and the following related behavior,
114 treating the bare symbol AND as equivalent to '(AND), is specifically
115 forbidden (by the ANSI specification of the AND type).
118 It would be nice if the
120 (during macroexpansion)
121 said what macroexpansion was at fault, e.g.
123 (during macroexpansion of IN-PACKAGE,
124 during macroexpansion of DEFFOO)
127 (SUBTYPEP '(FUNCTION (T BOOLEAN) NIL)
128 '(FUNCTION (FIXNUM FIXNUM) NIL)) => T, T
129 (Also, when this is fixed, we can enable the code in PROCLAIM which
130 checks for incompatible FTYPE redeclarations.)
133 (I *think* this is a bug. It certainly seems like strange behavior. But
134 the ANSI spec is scary, dark, and deep.. -- WHN)
135 (FORMAT NIL "~,1G" 1.4) => "1. "
136 (FORMAT NIL "~3,1G" 1.4) => "1. "
139 from Marco Antoniotti on cmucl-imp mailing list 1 Mar 2000:
141 (setf (find-class 'ccc1) (find-class 'ccc))
142 (defmethod zut ((c ccc1)) 123)
143 In sbcl-0.7.1.13, this gives an error,
144 There is no class named CCC1.
145 DTC's recommended workaround from the mailing list 3 Mar 2000:
146 (setf (pcl::find-class 'ccc1) (pcl::find-class 'ccc))
149 Sometimes (SB-EXT:QUIT) fails with
150 Argh! maximum interrupt nesting depth (4096) exceeded, exiting
151 Process inferior-lisp exited abnormally with code 1
152 I haven't noticed a repeatable case of this yet.
155 The printer doesn't report closures very well. This is true in
159 #<Closure Over Function "DEFUN STRUCTURE-SLOT-ACCESSOR" {134D1A1}>
160 It would be nice to make closures have a settable name slot,
161 and make things like DEFSTRUCT and FLET, which create closures,
162 set helpful values into this slot.
165 And as long as we're wishing, it would be awfully nice if INSPECT could
166 also report on closures, telling about the values of the bound variables.
169 The compiler assumes that any time a function of declared FTYPE
170 doesn't signal an error, its arguments were of the declared type.
171 E.g. compiling and loading
172 (DECLAIM (OPTIMIZE (SAFETY 3)))
173 (DEFUN FACTORIAL (X) (GAMMA (1+ X)))
175 (DECLAIM (FTYPE (FUNCTION (UNSIGNED-BYTE)) FACTORIAL))
177 (COND ((> (FACTORIAL X) 1.0E6)
178 (FORMAT T "too big~%"))
180 (FORMAT T "exactly ~S~%" (FACTORIAL X)))
182 (FORMAT T "approximately ~S~%" (FACTORIAL X)))))
185 will cause the INTEGERP case to be selected, giving bogus output a la
187 This violates the "declarations are assertions" principle.
188 According to the ANSI spec, in the section "System Class FUNCTION",
189 this is a case of "lying to the compiler", but the lying is done
190 by the code which calls FACTORIAL with non-UNSIGNED-BYTE arguments,
191 not by the unexpectedly general definition of FACTORIAL. In any case,
192 "declarations are assertions" means that lying to the compiler should
193 cause an error to be signalled, and should not cause a bogus
194 result to be returned. Thus, the compiler should not assume
195 that arbitrary functions check their argument types. (It might
196 make sense to add another flag (CHECKED?) to DEFKNOWN to
197 identify functions which *do* check their argument types.)
198 (Also, verify that the compiler handles declared function
199 return types as assertions.)
202 TYPEP of VALUES types is sometimes implemented very inefficiently, e.g. in
203 (DEFTYPE INDEXOID () '(INTEGER 0 1000))
205 (DECLARE (TYPE INDEXOID X))
206 (THE (VALUES INDEXOID)
208 where the implementation of the type check in function FOO
209 includes a full call to %TYPEP. There are also some fundamental problems
210 with the interpretation of VALUES types (inherited from CMU CL, and
211 from the ANSI CL standard) as discussed on the cmucl-imp@cons.org
212 mailing list, e.g. in Robert Maclachlan's post of 21 Jun 2000.
215 The definitions of SIGCONTEXT-FLOAT-REGISTER and
216 %SET-SIGCONTEXT-FLOAT-REGISTER in x86-vm.lisp say they're not
217 supported on FreeBSD because the floating point state is not saved,
218 but at least as of FreeBSD 4.0, the floating point state *is* saved,
219 so they could be supported after all. Very likely
220 SIGCONTEXT-FLOATING-POINT-MODES could now be supported, too.
223 (as discussed by Douglas Crosher on the cmucl-imp mailing list ca.
224 Aug. 10, 2000): CMUCL currently interprets 'member as '(member); same
225 issue with 'union, 'and, 'or etc. So even though according to the
226 ANSI spec, bare 'MEMBER, 'AND, and 'OR are not legal types, CMUCL
227 (and now SBCL) interpret them as legal types.
230 a slew of floating-point-related errors reported by Peter Van Eynde
232 b: SBCL's value for LEAST-POSITIVE-SHORT-FLOAT is bogus, and
233 should probably be 1.4012985e-45. In SBCL,
234 (/ LEAST-POSITIVE-SHORT-FLOAT 2) returns a number smaller
235 than LEAST-POSITIVE-SHORT-FLOAT. Similar problems
236 exist for LEAST-NEGATIVE-SHORT-FLOAT, LEAST-POSITIVE-LONG-FLOAT,
237 and LEAST-NEGATIVE-LONG-FLOAT.
238 c: Many expressions generate floating infinity on x86/Linux:
243 PVE's regression tests want them to raise errors. sbcl-0.7.0.5
244 on x86/Linux generates the infinities instead. That might or
245 might not be conforming behavior, but it's also inconsistent,
246 which is almost certainly wrong. (Inconsistency: (/ 1 0.0)
247 should give the same result as (/ 1.0 0.0), but instead (/ 1 0.0)
248 generates SINGLE-FLOAT-POSITIVE-INFINITY and (/ 1.0 0.0)
250 d: (in section12.erg) various forms a la
251 (FLOAT 1 DOUBLE-FLOAT-EPSILON)
252 don't give the right behavior.
255 type safety errors reported by Peter Van Eynde July 25, 2000:
256 c: (COERCE 'AND 'FUNCTION) returns something related to
257 (MACRO-FUNCTION 'AND), but ANSI says it should raise an error.
258 k: READ-BYTE is supposed to signal TYPE-ERROR when its argument is
259 not a binary input stream, but instead cheerfully reads from
260 character streams, e.g. (MAKE-STRING-INPUT-STREAM "abc").
263 DEFCLASS bugs reported by Peter Van Eynde July 25, 2000:
264 d: (DEFGENERIC IF (X)) should signal a PROGRAM-ERROR, but instead
265 causes a COMPILER-ERROR.
268 SYMBOL-MACROLET bugs reported by Peter Van Eynde July 25, 2000:
269 c: SYMBOL-MACROLET should signal PROGRAM-ERROR if something
270 it binds is declared SPECIAL inside.
273 miscellaneous errors reported by Peter Van Eynde July 25, 2000:
275 (DEFGENERIC FOO02 (X))
276 (DEFMETHOD FOO02 ((X NUMBER)) T)
277 (LET ((M (FIND-METHOD (FUNCTION FOO02)
279 (LIST (FIND-CLASS (QUOTE NUMBER))))))
280 (REMOVE-METHOD (FUNCTION FOO02) M)
281 (DEFGENERIC FOO03 (X))
282 (ADD-METHOD (FUNCTION FOO03) M)))
283 should give an error, but SBCL allows it.
286 It has been reported (e.g. by Peter Van Eynde) that there are
287 several metaobject protocol "errors". (In order to fix them, we might
288 need to document exactly what metaobject protocol specification
289 we're following -- the current code is just inherited from PCL.)
292 The implementation of #'+ returns its single argument without
293 type checking, e.g. (+ "illegal") => "illegal".
296 The debugger LIST-LOCATIONS command doesn't work properly.
299 Compiling and loading
300 (DEFUN FAIL (X) (THROW 'FAIL-TAG X))
302 then requesting a BACKTRACE at the debugger prompt gives no information
303 about where in the user program the problem occurred.
306 The compiler is supposed to do type inference well enough that
309 ((SIMPLE-ARRAY SINGLE-FLOAT)
311 (DECLARE (TYPE (SIMPLE-ARRAY SINGLE-FLOAT) X))
314 is redundant. However, as reported by Juan Jose Garcia Ripoll for
315 CMU CL, it sometimes doesn't. Adding declarations is a pretty good
316 workaround for the problem for now, but can't be done by the TYPECASE
317 macros themselves, since it's too hard for the macro to detect
318 assignments to the variable within the clause.
319 Note: The compiler *is* smart enough to do the type inference in
320 many cases. This case, derived from a couple of MACROEXPAND-1
321 calls on Ripoll's original test case,
323 (DECLARE (OPTIMIZE SPEED (SAFETY 0)))
324 (COND ((TYPEP A '(SIMPLE-ARRAY SINGLE-FLOAT)) NIL
325 (LET ((LENGTH (ARRAY-TOTAL-SIZE A)))
326 (LET ((I 0) (G2554 LENGTH))
327 (DECLARE (TYPE REAL G2554) (TYPE REAL I))
330 (WHEN (>= I G2554) (GO SB-LOOP::END-LOOP))
331 (SETF (ROW-MAJOR-AREF A I) (- (ROW-MAJOR-AREF A I)))
332 (GO SB-LOOP::NEXT-LOOP)
333 SB-LOOP::END-LOOP))))))
334 demonstrates the problem; but the problem goes away if the TAGBODY
335 and GO forms are removed (leaving the SETF in ordinary, non-looping
336 code), or if the TAGBODY and GO forms are retained, but the
337 assigned value becomes 0.0 instead of (- (ROW-MAJOR-AREF A I)).
340 Paul Werkowski wrote on cmucl-imp@cons.org 2000-11-15
341 I am looking into this problem that showed up on the cmucl-help
342 list. It seems to me that the "implementation specific environment
343 hacking functions" found in pcl/walker.lisp are completely messed
344 up. The good thing is that they appear to be barely used within
345 PCL and the munged environment object is passed to cmucl only
346 in calls to macroexpand-1, which is probably why this case fails.
347 SBCL uses essentially the same code, so if the environment hacking
348 is screwed up, it affects us too.
351 Using the pretty-printer from the command prompt gives funny
352 results, apparently because the pretty-printer doesn't know
353 about user's command input, including the user's carriage return
354 that the user, and therefore the pretty-printer thinks that
355 the new output block should start indented 2 or more characters
356 rightward of the correct location.
359 As reported by Winton Davies on a CMU CL mailing list 2000-01-10,
360 and reported for SBCL by Martin Atzmueller 2000-10-20: (TRACE GETHASH)
361 crashes SBCL. In general tracing anything which is used in the
362 implementation of TRACE is likely to have the same problem.
365 (DECLAIM (OPTIMIZE ..)) doesn't work properly inside LOCALLY forms.
368 As reported by Martin Atzmueller on sbcl-devel 26 Dec 2000,
369 ANSI says that WITH-OUTPUT-TO-STRING should have a keyword
370 :ELEMENT-TYPE, but in sbcl-0.6.9 this is not defined for
371 WITH-OUTPUT-TO-STRING.
374 ANSI says in one place that type declarations can be abbreviated even
375 when the type name is not a symbol, e.g.
376 (DECLAIM ((VECTOR T) *FOOVECTOR*))
377 SBCL doesn't support this. But ANSI says in another place that this
378 isn't allowed. So it's not clear this is a bug after all. (See the
379 e-mail on cmucl-help@cons.org on 2001-01-16 and 2001-01-17 from WHN
383 as pointed out by Dan Barlow on sbcl-devel 2000-07-02:
384 The PICK-TEMPORARY-FILE-NAME utility used by LOAD-FOREIGN uses
385 an easily guessable temporary filename in a way which might open
386 applications using LOAD-FOREIGN to hijacking by malicious users
387 on the same machine. Incantations for doing this safely are
388 floating around the net in various "how to write secure programs
389 despite Unix" documents, and it would be good to (1) fix this in
390 LOAD-FOREIGN, and (2) hunt for any other code which uses temporary
391 files and make it share the same new safe logic.
394 Functions are assigned names based on the context in which they're
395 defined. This is less than ideal for the functions which are
396 used to implement CLOS methods. E.g. the output of
397 (DESCRIBE 'PRINT-OBJECT) lists functions like
398 #<FUNCTION "DEF!STRUCT (TRACE-INFO (:MAKE-LOAD-FORM-FUN SB-KERNEL:JUST-DUMP-IT-NORMALLY) (:PRINT-OBJECT #))" {1020E49}>
400 #<FUNCTION "MACROLET ((FORCE-DELAYED-DEF!METHODS NIL #))" {1242871}>
401 It would be better if these functions' names always identified
402 them as methods, and identified their generic functions and
406 RANDOM-INTEGER-EXTRA-BITS=10 may not be large enough for the RANDOM
407 RNG to be high quality near RANDOM-FIXNUM-MAX; it looks as though
408 the mean of the distribution can be systematically O(0.1%) wrong.
409 Just increasing R-I-E-B is probably not a good solution, since
410 it would decrease efficiency more than is probably necessary. Perhaps
411 using some sort of accept/reject method would be better.
414 Internally the compiler sometimes evaluates
415 (sb-kernel:type/= (specifier-type '*) (specifier-type t))
416 (I stumbled across this when I added an
417 (assert (not (eq type1 *wild-type*)))
418 in the NAMED :SIMPLE-= type method.) '* isn't really a type, and
419 in a type context should probably be translated to T, and so it's
420 probably wrong to ask whether it's equal to the T type and then (using
421 the EQ type comparison in the NAMED :SIMPLE-= type method) return NIL.
422 (I haven't tried to investigate this bug enough to guess whether
423 there might be any user-level symptoms.)
425 In fact, the type system is likely to depend on this inequality not
426 holding... * is not equivalent to T in many cases, such as
427 (VECTOR *) /= (VECTOR T).
430 Inconsistencies between derived and declared VALUES return types for
431 DEFUN aren't checked very well. E.g. the logic which successfully
432 catches problems like
433 (declaim (ftype (function (fixnum) float) foo))
435 (declare (type integer x))
436 (values x)) ; wrong return type, detected, gives warning, good!
438 (declaim (ftype (function (t) (values t t)) bar))
440 (values x)) ; wrong number of return values, no warning, bad!
441 The cause of this is seems to be that (1) the internal function
442 VALUES-TYPES-EQUAL-OR-INTERSECT used to make the check handles its
443 arguments symmetrically, and (2) when the type checking code was
444 written back when when SBCL's code was still CMU CL, the intent
446 (declaim (ftype (function (t) t) bar))
448 (values x x)) ; wrong number of return values; should give warning?
449 not be warned for, because a two-valued return value is considered
450 to be compatible with callers who expects a single value to be
451 returned. That intent is probably not appropriate for modern ANSI
452 Common Lisp, but fixing this might be complicated because of other
453 divergences between auld-style and new-style handling of
454 multiple-VALUES types. (Some issues related to this were discussed
455 on cmucl-imp at some length sometime in 2000.)
458 The facility for dumping a running Lisp image to disk gets confused
459 when run without the PURIFY option, and creates an unnecessarily large
460 core file (apparently representing memory usage up to the previous
461 high-water mark). Moreover, when the file is loaded, it confuses the
462 GC, so that thereafter memory usage can never be reduced below that
466 In sbcl-0.6.11.41 (and in all earlier SBCL, and in CMU
467 CL), out-of-line structure slot setters are horribly inefficient
468 whenever the type of the slot is declared, because out-of-line
469 structure slot setters are implemented as closures to save space,
470 so the compiler doesn't compile the type test into code, but
471 instead just saves the type in a lexical closure and interprets it
473 A proper solution involves deciding whether it's really worth
474 saving space by implementing structure slot accessors as closures.
475 (If it's not worth it, the problem vanishes automatically. If it
476 is worth it, there are hacks we could use to force type tests to
477 be compiled anyway, and even shared. E.g. we could implement
478 an EQUAL hash table mapping from types to compiled type tests,
479 and save the appropriate compiled type test as part of each lexical
480 closure; or we could make the lexical closures be placeholders
481 which overwrite their old definition as a lexical closure with
482 a new compiled definition the first time that they're called.)
483 As a workaround for the problem, #'(SETF FOO) expressions can
484 be replaced with (EFFICIENT-SETF-FUNCTION FOO), where
485 (defmacro efficient-setf-function (place-function-name)
486 (or #+sbcl (and (sb-impl::info :function :accessor-for place-function-name)
487 ;; a workaround for the problem, encouraging the
488 ;; inline expansion of the structure accessor, so
489 ;; that the compiler can optimize its type test
490 (let ((new-value (gensym "NEW-VALUE-"))
491 (structure-value (gensym "STRUCTURE-VALUE-")))
492 `(lambda (,new-value ,structure-value)
493 (setf (,place-function-name ,structure-value)
495 ;; no problem, can just use the ordinary expansion
496 `(function (setf ,place-function-name))))
499 There's apparently a bug in CEILING optimization which caused
500 Douglas Crosher to patch the CMU CL version. Martin Atzmueller
501 applied the patches to SBCL and they didn't seem to cause problems
502 (as reported sbcl-devel 2001-05-04). However, since the patches
503 modify nontrivial code which was apparently written incorrectly
504 the first time around, until regression tests are written I'm not
505 comfortable merging the patches in the CVS version of SBCL.
508 (DESCRIBE 'SB-ALIEN:DEF-ALIEN-TYPE) reports the macro argument list
512 in #<PACKAGE "SB-ALIEN">.
513 Macro-function: #<FUNCTION "DEF!MACRO DEF-ALIEN-TYPE" {19F4A39}>
514 Macro arguments: (#:whole-470 #:environment-471)
515 On Sat, May 26, 2001 09:45:57 AM CDT it was compiled from:
516 /usr/stuff/sbcl/src/code/host-alieneval.lisp
517 Created: Monday, March 12, 2001 07:47:43 AM CST
520 (TIME (ROOM T)) reports more than 200 Mbytes consed even for
521 a clean, just-started SBCL system. And it seems to be right:
522 (ROOM T) can bring a small computer to its knees for a *long*
523 time trying to GC afterwards. Surely there's some more economical
524 way to implement (ROOM T).
527 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
529 (in-package :cl-user)
530 ;;; The following invokes a compiler error.
531 (declaim (optimize (speed 2) (debug 3)))
534 (unwind-protect nil)))
538 The error message in sbcl-0.6.12.42 is
539 internal error, failed AVER:
540 "(COMMON-LISP:EQ (SB!C::TN-ENVIRONMENT SB!C:TN) SB!C::TN-ENV)"
543 When the compiler inline expands functions, it may be that different
544 kinds of return values are generated from different code branches.
545 E.g. an inline expansion of POSITION generates integer results
546 from one branch, and NIL results from another. When that inline
547 expansion is used in a context where only one of those results
550 (aref *a1* (position x *a2*)))
551 and the compiler can't prove that the unacceptable branch is
552 never taken, then bogus type mismatch warnings can be generated.
553 If you need to suppress the type mismatch warnings, you can
554 suppress the inline expansion,
556 #+sbcl (declare (notinline position)) ; to suppress bug 117 bogowarnings
557 (aref *a1* (position x *a2*)))
558 or, sometimes, suppress them by declaring the result to be of an
561 (aref *a1* (the integer (position x *a2*))))
563 This is not a new compiler problem in 0.7.0, but the new compiler
564 transforms for FIND, POSITION, FIND-IF, and POSITION-IF make it
565 more conspicuous. If you don't need performance from these functions,
566 and the bogus warnings are a nuisance for you, you can return to
567 your pre-0.7.0 state of grace with
568 #+sbcl (declaim (notinline find position find-if position-if)) ; bug 117..
571 as reported by Eric Marsden on cmucl-imp@cons.org 2001-08-14:
572 (= (FLOAT 1 DOUBLE-FLOAT-EPSILON)
573 (+ (FLOAT 1 DOUBLE-FLOAT-EPSILON) DOUBLE-FLOAT-EPSILON)) => T
574 when of course it should be NIL. (He says it only fails for X86,
575 not SPARC; dunno about Alpha.)
577 Also, "the same problem exists for LONG-FLOAT-EPSILON,
578 DOUBLE-FLOAT-NEGATIVE-EPSILON, LONG-FLOAT-NEGATIVE-EPSILON (though
579 for the -negative- the + is replaced by a - in the test)."
581 Raymond Toy comments that this is tricky on the X86 since its FPU
582 uses 80-bit precision internally.
585 Even in sbcl-0.pre7.x, which is supposed to be free of the old
586 non-ANSI behavior of treating the function return type inferred
587 from the current function definition as a declaration of the
588 return type from any function of that name, the return type of NIL
589 is attached to FOO in 120a above, and used to optimize code which
593 There was some sort of screwup in handling of
594 (IF (NOT (IGNORE-ERRORS ..))). E.g.
596 (if (not (ignore-errors
597 (make-pathname :host "foo"
601 (error "notunlessnot")))
602 The (NOT (IGNORE-ERRORS ..)) form evaluates to T, so this should be
603 printing "ok", but instead it's going to the ERROR. This problem
604 seems to've been introduced by MNA's HANDLER-CASE patch (sbcl-devel
605 2001-07-17) and as a workaround (put in sbcl-0.pre7.14.flaky4.12)
606 I reverted back to the old weird HANDLER-CASE code. However, I
607 think the problem looks like a compiler bug in handling RETURN-FROM,
608 so I left the MNA-patched code in HANDLER-CASE (suppressed with
609 #+NIL) and I'd like to go back to see whether this really is
610 a compiler bug before I delete this BUGS entry.
613 As of version 0.pre7.14, SBCL's implementation of MACROLET makes
614 the entire lexical environment at the point of MACROLET available
615 in the bodies of the macroexpander functions. In particular, it
616 allows the function bodies (which run at compile time) to try to
617 access lexical variables (which are only defined at runtime).
618 It doesn't even issue a warning, which is bad.
620 The SBCL behavior arguably conforms to the ANSI spec (since the
621 spec says that the behavior is undefined, ergo anything conforms).
622 However, it would be better to issue a compile-time error.
623 Unfortunately I (WHN) don't see any simple way to detect this
624 condition in order to issue such an error, so for the meantime
625 SBCL just does this weird broken "conforming" thing.
627 The ANSI standard says, in the definition of the special operator
629 The macro-expansion functions defined by MACROLET are defined
630 in the lexical environment in which the MACROLET form appears.
631 Declarations and MACROLET and SYMBOL-MACROLET definitions affect
632 the local macro definitions in a MACROLET, but the consequences
633 are undefined if the local macro definitions reference any
634 local variable or function bindings that are visible in that
636 Then it seems to contradict itself by giving the example
638 (macrolet ((fudge (z)
639 ;The parameters x and flag are not accessible
640 ; at this point; a reference to flag would be to
641 ; the global variable of that name.
642 ` (if flag (* ,z ,z) ,z)))
643 ;The parameters x and flag are accessible here.
647 The comment "a reference to flag would be to the global variable
648 of the same name" sounds like good behavior for the system to have.
649 but actual specification quoted above says that the actual behavior
653 (as reported by Gabe Garza on cmucl-help 2001-09-21)
655 (defun test-pred (x y)
659 (func (lambda () x)))
660 (print (eq func func))
661 (print (test-pred func func))
662 (delete func (list func))))
663 Now calling (TEST-CASE) gives output
666 (#<FUNCTION {500A9EF9}>)
667 Evidently Python thinks of the lambda as a code transformation so
668 much that it forgets that it's also an object.
671 The DEFSTRUCT section of the ANSI spec, in the :CONC-NAME section,
672 specifies a precedence rule for name collisions between slot accessors of
673 structure classes related by inheritance. As of 0.7.0, SBCL still
677 Ideally, uninterning a symbol would allow it, and its associated
678 FDEFINITION and PROCLAIM data, to be reclaimed by the GC. However,
679 at least as of sbcl-0.7.0, this isn't the case. Information about
680 FDEFINITIONs and PROCLAIMed properties is stored in globaldb.lisp
681 essentially in ordinary (non-weak) hash tables keyed by symbols.
682 Thus, once a system has an entry in this system, it tends to live
683 forever, even when it is uninterned and all other references to it
687 (reported by Arnaud Rouanet on cmucl-imp 2001-12-18)
688 (defmethod foo ((x integer))
690 (defmethod foo :around ((x integer))
693 Now (FOO 3) should return 3, but instead it returns 4.
696 (reported by Alexey Dejneka sbcl-devel 2002-01-03)
698 SUBTYPEP does not work well with redefined classes:
707 * (defclass b (a) ())
720 This bug was fixed in sbcl-0.7.4.1 by invalidating the PCL wrapper
721 class upon redefinition. Unfortunately, doing so causes bug #176 to
722 appear. Pending further investigation, one or other of these bugs
723 might be present at any given time.
726 Pretty-printing nested backquotes doesn't work right, as
727 reported by Alexey Dejneka sbcl-devel 2002-01-13:
729 ``(FOO SB-IMPL::BACKQ-COMMA-AT S)
730 * (lisp-implementation-version)
734 (reported by Jesse Bouwman 2001-10-24 through the unfortunately
735 prominent SourceForge web/db bug tracking system, which is
736 unfortunately not a reliable way to get a timely response from
737 the SBCL maintainers)
738 In the course of trying to build a test case for an
739 application error, I encountered this behavior:
740 If you start up sbcl, and then lay on CTRL-C for a
741 minute or two, the lisp process will eventually say:
742 %PRIMITIVE HALT called; the party is over.
743 and throw you into the monitor. If I start up lisp,
744 attach to the process with strace, and then do the same
745 (abusive) thing, I get instead:
746 access failure in heap page not marked as write-protected
747 and the monitor again. I don't know enough to have the
748 faintest idea of what is going on here.
749 This is with sbcl 6.12, uname -a reports:
750 Linux prep 2.2.19 #4 SMP Tue Apr 24 13:59:52 CDT 2001 i686 unknown
751 I (WHN) have verified that the same thing occurs on sbcl-0.pre7.141
752 under OpenBSD 2.9 on my X86 laptop. Do be patient when you try it:
753 it took more than two minutes (but less than five) for me.
756 (This was once known as IR1-4, but it lived on even after the
757 IR1 interpreter went to the big bit bucket in the sky.)
758 The system accepts DECLAIM in most places where DECLARE would be
759 accepted, without even issuing a warning. ANSI allows this, but since
760 it's fairly easy to mistype DECLAIM instead of DECLARE, and the
761 meaning is rather different, and it's unlikely that the user
762 has a good reason for doing DECLAIM not at top level, it would be
763 good to issue a STYLE-WARNING when this happens. A possible
764 fix would be to issue STYLE-WARNINGs for DECLAIMs not at top level,
765 or perhaps to issue STYLE-WARNINGs for any EVAL-WHEN not at top level.
766 [This is considered an IR1-interpreter-related bug because until
767 EVAL-WHEN is rewritten, which won't happen until after the IR1
768 interpreter is gone, the system's notion of what's a top-level form
769 and what's not will remain too confused to fix this problem.]
772 ANSI allows types `(COMPLEX ,FOO) to use very hairy values for
773 FOO, e.g. (COMPLEX (AND REAL (SATISFIES ODDP))). The old CMU CL
774 COMPLEX implementation didn't deal with this, and hasn't been
775 upgraded to do so. (This doesn't seem to be a high priority
776 conformance problem, since seems hard to construct useful code
780 Floating point errors are reported poorly. E.g. on x86 OpenBSD
783 debugger invoked on condition of type SB-KERNEL:FLOATING-POINT-EXCEPTION:
784 An arithmetic error SB-KERNEL:FLOATING-POINT-EXCEPTION was signalled.
785 No traps are enabled? How can this be?
786 It should be possible to be much more specific (overflow, division
787 by zero, etc.) and of course the "How can this be?" should be fixable.
789 See also bugs #45.c and #183
792 In sbcl-0.7.1.3 on x86, COMPILE-FILE on the file
793 (in-package :cl-user)
796 (defstruct foo bar bletch)
798 (labels ((kidify1 (kid)
806 (declare (inline kid-frob))
809 (the simple-vector (foo-bar perd)))))
811 debugger invoked on condition of type TYPE-ERROR:
812 The value NIL is not of type SB-C::NODE.
813 The location of this failure has moved around as various related
814 issues were cleaned up. As of sbcl-0.7.1.9, it occurs in
815 NODE-BLOCK called by LAMBDA-COMPONENT called by IR2-CONVERT-CLOSURE.
817 (Python LET-converts KIDIFY1 into KID-FROB, then tries to inline
818 expand KID-FROB into %ZEEP. Having partially done it, it sees a call
819 of KIDIFY1, which already does not exist. So it gives up on
820 expansion, leaving garbage consisting of infinished blocks of the
821 partially converted function.)
824 Functions SUBTYPEP, TYPEP, UPGRADED-ARRAY-ELEMENT-TYPE, and
825 UPGRADED-COMPLEX-PART-TYPE should have an optional environment argument.
826 (reported by Alexey Dejneka sbcl-devel 2002-04-12)
829 (reported by Robert E. Brown 2002-04-16)
830 When a function is called with too few arguments, causing the
831 debugger to be entered, the uninitialized slots in the bad call frame
832 seem to cause GCish problems, being interpreted as tagged data even
833 though they're not. In particular, executing ROOM in the
834 debugger at that point causes AVER failures:
837 * (lisp-implementation-version)
843 failed AVER: "(SAP= CURRENT END)"
844 (Christophe Rhodes reports that this doesn't occur on the SPARC, which
845 isn't too surprising since there are many differences in stack
846 implementation and GC conservatism between the X86 and other ports.)
850 (in-package :cl-user)
852 (defmethod permanentize ((uustk uustk))
853 (flet ((frob (hash-table test-for-deletion)
855 (obj-entry.stale? (oe)
856 (destructuring-bind (key . datum) oe
857 (declare (type simple-vector key))
858 (deny0 (void? datum))
859 (some #'stale? key))))
860 (declare (inline frob obj-entry.stale?))
861 (frob (uustk.args-hash->obj-alist uustk)
863 (frob (uustk.hash->memoized-objs-list uustk)
866 in sbcl-0.7.3.11 causes an assertion failure,
869 (AND (NULL (BLOCK-SUCC B))
870 (NOT (BLOCK-DELETE-P B))
871 (NOT (EQ B (COMPONENT-HEAD #)))))"
874 In sbcl-0.7.3.11, compiling the (illegal) code
875 (in-package :cl-user)
876 (defmethod prove ((uustk uustk))
879 gives the (not terribly clear) error message
881 ; (during macroexpansion of (DEFMETHOD PROVE ...))
882 ; can't get template for (FROB NIL NIL)
883 The problem seems to be that the code walker used by the DEFMETHOD
884 macro is unhappy with the illegal syntax in the method body, and
885 is giving an unclear error message.
888 sbcl's treatment of at least macro lambda lists is too permissive;
889 e.g., in sbcl-0.7.3.7:
890 (defmacro foo (&rest rest bar) `(,bar ,rest))
891 (macroexpand '(foo quux zot)) -> (QUUX (QUUX ZOT))
892 whereas section 3.4.4 of the CLHS doesn't allow required parameters
893 to come after the rest argument.
896 The compiler sometimes tries to constant-fold expressions before
897 it checks to see whether they can be reached. This can lead to
898 bogus warnings about errors in the constant folding, e.g. in code
901 (WRITE-STRING (> X 0) "+" "0"))
902 compiled in a context where the compiler can prove that X is NIL,
903 and the compiler complains that (> X 0) causes a type error because
904 NIL isn't a valid argument to #'>. Until sbcl-0.7.4.10 or so this
905 caused a full WARNING, which made the bug really annoying because then
906 COMPILE and COMPILE-FILE returned FAILURE-P=T for perfectly legal
907 code. Since then the warning has been downgraded to STYLE-WARNING,
908 so it's still a bug but at least it's a little less annoying.
911 reported by Alexey Dejneka 08 Jun 2002 in sbcl-devel:
912 Playing with McCLIM, I've received an error "Unbound variable WRAPPER
913 in SB-PCL::CHECK-WRAPPER-VALIDITY".
914 (defun check-wrapper-validity (instance)
915 (let* ((owrapper (wrapper-of instance)))
916 (if (not (invalid-wrapper-p owrapper))
918 (let* ((state (wrapper-state wrapper)) ; !!!
920 I've tried to replace it with OWRAPPER, but now OBSOLETE-INSTANCE-TRAP
921 breaks with "NIL is not of type SB-KERNEL:LAYOUT".
923 partial fix: The undefined variable WRAPPER resulted from an error
924 in recent refactoring, as can be seen by comparing to the code in e.g.
925 sbcl-0.7.2. Replacing WRAPPER with OWRAPPER (done by WHN in sbcl-0.7.4.22)
926 should bring the code back to its behavior as of sbcl-0.7.2, but
927 that still leaves the OBSOLETE-INSTANCE-TRAP bug. An example of
928 input which triggers that bug is
930 (let ((lastname (intern (format nil "C~D" (1- i))))
931 (name (intern (format nil "C~D" i))))
932 (eval `(defclass ,name
933 (,@(if (= i 0) nil (list lastname)))
935 (eval `(defmethod initialize-instance :after ((x ,name) &rest any)
936 (declare (ignore any))))))
943 178: "AVER failure compiling confused THEs in FUNCALL"
944 In sbcl-0.7.4.24, compiling
946 (funcall (the function (the standard-object x))))
949 "(AND (EQ (IR2-CONTINUATION-PRIMITIVE-TYPE 2CONT) FUNCTION-PTYPE) (EQ CHECK T))"
950 This variant compiles OK, though:
951 (defun bug178alternative (x)
952 (funcall (the nil x)))
954 (since 0.7.8.9 it does not signal an error; see also bug 199)
956 183: "IEEE floating point issues"
957 Even where floating point handling is being dealt with relatively
958 well (as of sbcl-0.7.5, on sparc/sunos and alpha; see bug #146), the
959 accrued-exceptions and current-exceptions part of the fp control
960 word don't seem to bear much relation to reality. E.g. on
964 debugger invoked on condition of type DIVISION-BY-ZERO:
965 arithmetic error DIVISION-BY-ZERO signalled
966 0] (sb-vm::get-floating-point-modes)
968 (:TRAPS (:OVERFLOW :INVALID :DIVIDE-BY-ZERO)
969 :ROUNDING-MODE :NEAREST
970 :CURRENT-EXCEPTIONS NIL
971 :ACCRUED-EXCEPTIONS (:INEXACT)
974 * (sb-vm::get-floating-point-modes)
975 (:TRAPS (:OVERFLOW :INVALID :DIVIDE-BY-ZERO)
976 :ROUNDING-MODE :NEAREST
977 :CURRENT-EXCEPTIONS (:INEXACT)
978 :ACCRUED-EXCEPTIONS (:INEXACT)
981 185: "top-level forms at the REPL"
982 * (locally (defstruct foo (a 0 :type fixnum)))
985 ; (in macroexpansion of (SB-KERNEL::%DELAYED-GET-COMPILER-LAYOUT BAR))
986 however, compiling and loading the same expression in a file works
989 187: "type inference confusion around DEFTRANSFORM time"
990 (reported even more verbosely on sbcl-devel 2002-06-28 as "strange
991 bug in DEFTRANSFORM")
992 After the file below is compiled and loaded in sbcl-0.7.5, executing
993 (TCX (MAKE-ARRAY 4 :FILL-POINTER 2) 0)
994 at the REPL returns an adjustable vector, which is wrong. Presumably
995 somehow the DERIVE-TYPE information for the output values of %WAD is
996 being mispropagated as a type constraint on the input values of %WAD,
997 and so causing the type test to be optimized away. It's unclear how
998 hand-expanding the DEFTRANSFORM would change this, but it suggests
999 the DEFTRANSFORM machinery (or at least the way DEFTRANSFORMs are
1000 invoked at a particular phase) is involved.
1001 (cl:in-package :sb-c)
1002 (eval-when (:compile-toplevel)
1003 ;;; standin for %DATA-VECTOR-AND-INDEX
1004 (defknown %dvai (array index)
1006 (foldable flushable))
1007 (deftransform %dvai ((array index)
1011 (let* ((atype (continuation-type array))
1012 (eltype (array-type-specialized-element-type atype)))
1013 (when (eq eltype *wild-type*)
1014 (give-up-ir1-transform
1015 "specialized array element type not known at compile-time"))
1016 (when (not (array-type-complexp atype))
1017 (give-up-ir1-transform "SIMPLE array!"))
1018 `(if (array-header-p array)
1019 (%wad array index nil)
1020 (values array index))))
1021 ;;; standin for %WITH-ARRAY-DATA
1022 (defknown %wad (array index (or index null))
1023 (values (simple-array * (*)) index index index)
1024 (foldable flushable))
1025 ;;; (Commenting out this optimizer causes the bug to go away.)
1026 (defoptimizer (%wad derive-type) ((array start end))
1027 (let ((atype (continuation-type array)))
1028 (when (array-type-p atype)
1029 (values-specifier-type
1030 `(values (simple-array ,(type-specifier
1031 (array-type-specialized-element-type atype))
1033 index index index)))))
1035 (defun %wad (array start end)
1036 (format t "~&in %WAD~%")
1037 (%with-array-data array start end))
1038 (cl:in-package :cl-user)
1040 (declare (type (vector t) v))
1041 (declare (notinline sb-kernel::%with-array-data))
1042 ;; (Hand-expending DEFTRANSFORM %DVAI here also causes the bug to
1046 188: "compiler performance fiasco involving type inference and UNION-TYPE"
1047 (In sbcl-0.7.6.10, DEFTRANSFORM CONCATENATE was commented out until this
1048 bug could be fixed properly, so you won't see the bug unless you restore
1049 the DEFTRANSFORM by hand.) In sbcl-0.7.5.11 on a 700 MHz Pentium III,
1053 (declare (optimize (safety 3)))
1054 (declare (optimize (compilation-speed 2)))
1055 (declare (optimize (speed 1) (debug 1) (space 1)))
1056 (let ((fn "if-this-file-exists-the-universe-is-strange"))
1057 (load fn :if-does-not-exist nil)
1058 (load (concatenate 'string fn ".lisp") :if-does-not-exist nil)
1059 (load (concatenate 'string fn ".fasl") :if-does-not-exist nil)
1060 (load (concatenate 'string fn ".misc-garbage")
1061 :if-does-not-exist nil)))))
1063 134.552 seconds of real time
1064 133.35156 seconds of user run time
1065 0.03125 seconds of system run time
1066 [Run times include 2.787 seconds GC run time.]
1068 246883368 bytes consed.
1069 BACKTRACE from Ctrl-C in the compilation shows that the compiler is
1070 thinking about type relationships involving types like
1072 (OR (INTEGER 576 576)
1083 190: "PPC/Linux pipe? buffer? bug"
1084 In sbcl-0.7.6, the run-program.test.sh test script sometimes hangs
1085 on the PPC/Linux platform, waiting for a zombie env process. This
1086 is a classic symptom of buffer filling and deadlock, but it seems
1087 only sporadically reproducible.
1089 191: "Miscellaneous PCL deficiencies"
1090 (reported by Alexey Dejneka sbcl-devel 2002-08-04)
1091 a. DEFCLASS does not inform the compiler about generated
1092 functions. Compiling a file with
1093 (DEFCLASS A-CLASS ()
1095 (DEFUN A-CLASS-X (A)
1096 (WITH-SLOTS (A-CLASS-X) A
1098 results in a STYLE-WARNING:
1100 SB-SLOT-ACCESSOR-NAME::|COMMON-LISP-USER A-CLASS-X slot READER|
1102 APD's fix for this was checked in to sbcl-0.7.6.20, but Pierre
1103 Mai points out that the declamation of functions is in fact
1104 incorrect in some cases (most notably for structure
1105 classes). This means that at present erroneous attempts to use
1106 WITH-SLOTS and the like on classes with metaclass STRUCTURE-CLASS
1107 won't get the corresponding STYLE-WARNING.
1108 c. the examples in CLHS 7.6.5.1 (regarding generic function lambda
1109 lists and &KEY arguments) do not signal errors when they should.
1111 192: "Python treats free type declarations as promises."
1112 b. What seemed like the same fundamental problem as bug 192a, but
1113 was not fixed by the same (APD "more strict type checking
1114 sbcl-devel 2002-08-97) patch:
1115 (DOTIMES (I ...) (DOTIMES (J ...) (DECLARE ...) ...)):
1116 (declaim (optimize (speed 1) (safety 3)))
1117 (defun trust-assertion (i)
1119 (declare (type (mod 4) i)) ; when commented out, behavior changes!
1122 (trust-assertion 6) ; prints nothing unless DECLARE is commented out
1126 193: "unhelpful CLOS error reporting when the primary method is missing"
1128 (defmethod foo :before ((x t)) (print x))
1129 is the only method defined on FOO, the error reporting when e.g.
1131 is relatively unhelpful:
1132 There is no primary method for the generic function
1133 #<STANDARD-GENERIC-FUNCTION FOO (1)>.
1134 with the offending argument nowhere visible in the backtrace. This
1135 continues even if there *are* primary methods, just not for the
1136 specified arg type, e.g.
1137 (defmethod foo ((x character)) (print x))
1138 (defmethod foo ((x string)) (print x))
1139 (defmethod foo ((x pathname)) ...)
1140 In that case it could be very helpful to know what argument value is
1141 falling through the cracks of the defined primary methods, but the
1142 error message stays the same (even BACKTRACE doesn't tell you what the
1143 bad argument value is).
1145 194: "no error from (THE REAL '(1 2 3)) in some cases"
1148 (multiple-value-prog1 (progn (the real '(1 2 3))))
1149 returns (1 2 3) instead of signalling an error. This was fixed by
1150 APD's "more strict type checking patch", but although the fixed
1151 code (in sbcl-0.7.7.19) works (signals TYPE-ERROR) interactively,
1152 it's difficult to write a regression test for it, because
1153 (IGNORE-ERRORS (MULTIPLE-VALUE-PROG1 (PROGN (THE REAL '(1 2 3)))))
1154 still returns (1 2 3).
1156 b. (IGNORE-ERRORS (MULTIPLE-VALUE-PROG1 (PROGN (THE REAL '(1 2 3)))))
1157 returns (1 2 3). (As above, this shows up when writing regression
1158 tests for fixed-ness of part a.)
1159 c. Also in sbcl-0.7.7.9, (IGNORE-ERRORS (THE REAL '(1 2 3))) => (1 2 3).
1161 (null (ignore-errors
1163 (arg2 (identity (the real #(1 2 3)))))
1164 (if (< arg1 arg2) arg1 arg2))))
1166 but putting the same expression inside (DEFUN FOO () ...),
1169 * Actually this entry is probably multiple bugs, as
1170 Alexey Dejneka commented on sbcl-devel 2002-09-03:)
1171 I don't think that placing these two bugs in one entry is
1172 a good idea: they have different explanations. The second
1173 (min 1 nil) is caused by flushing of unused code--IDENTITY
1174 can do nothing with it. So it is really bug 122. The first
1175 (min nil) is due to M-V-PROG1: substituting a continuation
1176 for the result, it forgets about type assertion. The purpose
1177 of IDENTITY is to save the restricted continuation from
1178 inaccurate transformations.
1179 * Alexey Dejneka pointed out that
1180 (IGNORE-ERRORS (IDENTITY (THE REAL '(1 2 3))))
1181 works as it should. Also
1182 (IGNORE-ERRORS (VALUES (THE REAL '(1 2 3))))
1183 works as it should. Perhaps this is another case of VALUES type
1184 intersections behaving in non-useful ways?
1186 199: "hairy FUNCTION types confuse the compiler"
1187 (reported by APD sbcl-devel 2002-09-15)
1189 (EQ NIL (FUNCALL F)))
1192 (DECLARE (TYPE (AND FUNCTION (SATISFIES MUR)) F))
1195 fails to compile, printing
1197 "(AND (EQ (IR2-CONTINUATION-PRIMITIVE-TYPE 2CONT) FUNCTION-PTYPE) (EQ CHECK T))"
1199 APD further reports that this bug is not present in CMUCL.
1201 (this case was fixed in 0.7.8.9; see also bug 178)
1203 201: "Incautious type inference from compound CONS types"
1204 (reported by APD sbcl-devel 2002-09-17)
1206 (LET ((Y (CAR (THE (CONS INTEGER *) X))))
1208 (FORMAT NIL "~S IS ~S, Y = ~S"
1215 (FOO ' (1 . 2)) => "NIL IS INTEGER, Y = 1"
1218 Compiler does not check THEs on unused values, e.g. in
1220 (progn (the real (list 1)) t)
1222 This situation may appear during optimizing away degenerate cases of
1223 certain functions: see bugs 54, 192b.
1226 (EVAL-WHEN (:COMPILE-TOPLEVEL) ...) inside MACROLET evaluates its
1227 argument in the null lexical environment. E.g. compiling file with
1229 (macrolet ((def (x) `(print ,x)))
1230 (eval-when (:compile-toplevel)
1235 debugger invoked on condition of type UNDEFINED-FUNCTION:
1236 The function DEF is undefined.
1239 DEFUNCT CATEGORIES OF BUGS
1241 These labels were used for bugs related to the old IR1 interpreter.
1242 The # values reached 6 before the category was closed down.