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 h: (MAKE-CONCATENATED-STREAM (MAKE-STRING-OUTPUT-STREAM))
259 should signal TYPE-ERROR.
260 i: MAKE-TWO-WAY-STREAM doesn't check that its arguments can
261 be used for input and output as needed. It should fail with
262 TYPE-ERROR when handed e.g. the results of
263 MAKE-STRING-INPUT-STREAM or MAKE-STRING-OUTPUT-STREAM in
264 the inappropriate positions, but doesn't.
265 k: READ-BYTE is supposed to signal TYPE-ERROR when its argument is
266 not a binary input stream, but instead cheerfully reads from
267 character streams, e.g. (MAKE-STRING-INPUT-STREAM "abc").
270 DEFCLASS bugs reported by Peter Van Eynde July 25, 2000:
271 d: (DEFGENERIC IF (X)) should signal a PROGRAM-ERROR, but instead
272 causes a COMPILER-ERROR.
275 SYMBOL-MACROLET bugs reported by Peter Van Eynde July 25, 2000:
276 c: SYMBOL-MACROLET should signal PROGRAM-ERROR if something
277 it binds is declared SPECIAL inside.
280 miscellaneous errors reported by Peter Van Eynde July 25, 2000:
282 (DEFGENERIC FOO02 (X))
283 (DEFMETHOD FOO02 ((X NUMBER)) T)
284 (LET ((M (FIND-METHOD (FUNCTION FOO02)
286 (LIST (FIND-CLASS (QUOTE NUMBER))))))
287 (REMOVE-METHOD (FUNCTION FOO02) M)
288 (DEFGENERIC FOO03 (X))
289 (ADD-METHOD (FUNCTION FOO03) M)))
290 should give an error, but SBCL allows it.
293 It has been reported (e.g. by Peter Van Eynde) that there are
294 several metaobject protocol "errors". (In order to fix them, we might
295 need to document exactly what metaobject protocol specification
296 we're following -- the current code is just inherited from PCL.)
299 The implementation of #'+ returns its single argument without
300 type checking, e.g. (+ "illegal") => "illegal".
303 The debugger LIST-LOCATIONS command doesn't work properly.
306 Compiling and loading
307 (DEFUN FAIL (X) (THROW 'FAIL-TAG X))
309 then requesting a BACKTRACE at the debugger prompt gives no information
310 about where in the user program the problem occurred.
313 The compiler is supposed to do type inference well enough that
316 ((SIMPLE-ARRAY SINGLE-FLOAT)
318 (DECLARE (TYPE (SIMPLE-ARRAY SINGLE-FLOAT) X))
321 is redundant. However, as reported by Juan Jose Garcia Ripoll for
322 CMU CL, it sometimes doesn't. Adding declarations is a pretty good
323 workaround for the problem for now, but can't be done by the TYPECASE
324 macros themselves, since it's too hard for the macro to detect
325 assignments to the variable within the clause.
326 Note: The compiler *is* smart enough to do the type inference in
327 many cases. This case, derived from a couple of MACROEXPAND-1
328 calls on Ripoll's original test case,
330 (DECLARE (OPTIMIZE SPEED (SAFETY 0)))
331 (COND ((TYPEP A '(SIMPLE-ARRAY SINGLE-FLOAT)) NIL
332 (LET ((LENGTH (ARRAY-TOTAL-SIZE A)))
333 (LET ((I 0) (G2554 LENGTH))
334 (DECLARE (TYPE REAL G2554) (TYPE REAL I))
337 (WHEN (>= I G2554) (GO SB-LOOP::END-LOOP))
338 (SETF (ROW-MAJOR-AREF A I) (- (ROW-MAJOR-AREF A I)))
339 (GO SB-LOOP::NEXT-LOOP)
340 SB-LOOP::END-LOOP))))))
341 demonstrates the problem; but the problem goes away if the TAGBODY
342 and GO forms are removed (leaving the SETF in ordinary, non-looping
343 code), or if the TAGBODY and GO forms are retained, but the
344 assigned value becomes 0.0 instead of (- (ROW-MAJOR-AREF A I)).
347 Paul Werkowski wrote on cmucl-imp@cons.org 2000-11-15
348 I am looking into this problem that showed up on the cmucl-help
349 list. It seems to me that the "implementation specific environment
350 hacking functions" found in pcl/walker.lisp are completely messed
351 up. The good thing is that they appear to be barely used within
352 PCL and the munged environment object is passed to cmucl only
353 in calls to macroexpand-1, which is probably why this case fails.
354 SBCL uses essentially the same code, so if the environment hacking
355 is screwed up, it affects us too.
358 Using the pretty-printer from the command prompt gives funny
359 results, apparently because the pretty-printer doesn't know
360 about user's command input, including the user's carriage return
361 that the user, and therefore the pretty-printer thinks that
362 the new output block should start indented 2 or more characters
363 rightward of the correct location.
366 As reported by Winton Davies on a CMU CL mailing list 2000-01-10,
367 and reported for SBCL by Martin Atzmueller 2000-10-20: (TRACE GETHASH)
368 crashes SBCL. In general tracing anything which is used in the
369 implementation of TRACE is likely to have the same problem.
372 (DECLAIM (OPTIMIZE ..)) doesn't work properly inside LOCALLY forms.
375 As reported by Martin Atzmueller on sbcl-devel 26 Dec 2000,
376 ANSI says that WITH-OUTPUT-TO-STRING should have a keyword
377 :ELEMENT-TYPE, but in sbcl-0.6.9 this is not defined for
378 WITH-OUTPUT-TO-STRING.
381 ANSI says in one place that type declarations can be abbreviated even
382 when the type name is not a symbol, e.g.
383 (DECLAIM ((VECTOR T) *FOOVECTOR*))
384 SBCL doesn't support this. But ANSI says in another place that this
385 isn't allowed. So it's not clear this is a bug after all. (See the
386 e-mail on cmucl-help@cons.org on 2001-01-16 and 2001-01-17 from WHN
390 as pointed out by Dan Barlow on sbcl-devel 2000-07-02:
391 The PICK-TEMPORARY-FILE-NAME utility used by LOAD-FOREIGN uses
392 an easily guessable temporary filename in a way which might open
393 applications using LOAD-FOREIGN to hijacking by malicious users
394 on the same machine. Incantations for doing this safely are
395 floating around the net in various "how to write secure programs
396 despite Unix" documents, and it would be good to (1) fix this in
397 LOAD-FOREIGN, and (2) hunt for any other code which uses temporary
398 files and make it share the same new safe logic.
401 Functions are assigned names based on the context in which they're
402 defined. This is less than ideal for the functions which are
403 used to implement CLOS methods. E.g. the output of
404 (DESCRIBE 'PRINT-OBJECT) lists functions like
405 #<FUNCTION "DEF!STRUCT (TRACE-INFO (:MAKE-LOAD-FORM-FUN SB-KERNEL:JUST-DUMP-IT-NORMALLY) (:PRINT-OBJECT #))" {1020E49}>
407 #<FUNCTION "MACROLET ((FORCE-DELAYED-DEF!METHODS NIL #))" {1242871}>
408 It would be better if these functions' names always identified
409 them as methods, and identified their generic functions and
413 RANDOM-INTEGER-EXTRA-BITS=10 may not be large enough for the RANDOM
414 RNG to be high quality near RANDOM-FIXNUM-MAX; it looks as though
415 the mean of the distribution can be systematically O(0.1%) wrong.
416 Just increasing R-I-E-B is probably not a good solution, since
417 it would decrease efficiency more than is probably necessary. Perhaps
418 using some sort of accept/reject method would be better.
421 Internally the compiler sometimes evaluates
422 (sb-kernel:type/= (specifier-type '*) (specifier-type t))
423 (I stumbled across this when I added an
424 (assert (not (eq type1 *wild-type*)))
425 in the NAMED :SIMPLE-= type method.) '* isn't really a type, and
426 in a type context should probably be translated to T, and so it's
427 probably wrong to ask whether it's equal to the T type and then (using
428 the EQ type comparison in the NAMED :SIMPLE-= type method) return NIL.
429 (I haven't tried to investigate this bug enough to guess whether
430 there might be any user-level symptoms.)
433 Inconsistencies between derived and declared VALUES return types for
434 DEFUN aren't checked very well. E.g. the logic which successfully
435 catches problems like
436 (declaim (ftype (function (fixnum) float) foo))
438 (declare (type integer x))
439 (values x)) ; wrong return type, detected, gives warning, good!
441 (declaim (ftype (function (t) (values t t)) bar))
443 (values x)) ; wrong number of return values, no warning, bad!
444 The cause of this is seems to be that (1) the internal function
445 VALUES-TYPES-EQUAL-OR-INTERSECT used to make the check handles its
446 arguments symmetrically, and (2) when the type checking code was
447 written back when when SBCL's code was still CMU CL, the intent
449 (declaim (ftype (function (t) t) bar))
451 (values x x)) ; wrong number of return values; should give warning?
452 not be warned for, because a two-valued return value is considered
453 to be compatible with callers who expects a single value to be
454 returned. That intent is probably not appropriate for modern ANSI
455 Common Lisp, but fixing this might be complicated because of other
456 divergences between auld-style and new-style handling of
457 multiple-VALUES types. (Some issues related to this were discussed
458 on cmucl-imp at some length sometime in 2000.)
461 The facility for dumping a running Lisp image to disk gets confused
462 when run without the PURIFY option, and creates an unnecessarily large
463 core file (apparently representing memory usage up to the previous
464 high-water mark). Moreover, when the file is loaded, it confuses the
465 GC, so that thereafter memory usage can never be reduced below that
469 In sbcl-0.6.11.41 (and in all earlier SBCL, and in CMU
470 CL), out-of-line structure slot setters are horribly inefficient
471 whenever the type of the slot is declared, because out-of-line
472 structure slot setters are implemented as closures to save space,
473 so the compiler doesn't compile the type test into code, but
474 instead just saves the type in a lexical closure and interprets it
476 A proper solution involves deciding whether it's really worth
477 saving space by implementing structure slot accessors as closures.
478 (If it's not worth it, the problem vanishes automatically. If it
479 is worth it, there are hacks we could use to force type tests to
480 be compiled anyway, and even shared. E.g. we could implement
481 an EQUAL hash table mapping from types to compiled type tests,
482 and save the appropriate compiled type test as part of each lexical
483 closure; or we could make the lexical closures be placeholders
484 which overwrite their old definition as a lexical closure with
485 a new compiled definition the first time that they're called.)
486 As a workaround for the problem, #'(SETF FOO) expressions can
487 be replaced with (EFFICIENT-SETF-FUNCTION FOO), where
488 (defmacro efficient-setf-function (place-function-name)
489 (or #+sbcl (and (sb-impl::info :function :accessor-for place-function-name)
490 ;; a workaround for the problem, encouraging the
491 ;; inline expansion of the structure accessor, so
492 ;; that the compiler can optimize its type test
493 (let ((new-value (gensym "NEW-VALUE-"))
494 (structure-value (gensym "STRUCTURE-VALUE-")))
495 `(lambda (,new-value ,structure-value)
496 (setf (,place-function-name ,structure-value)
498 ;; no problem, can just use the ordinary expansion
499 `(function (setf ,place-function-name))))
502 There's apparently a bug in CEILING optimization which caused
503 Douglas Crosher to patch the CMU CL version. Martin Atzmueller
504 applied the patches to SBCL and they didn't seem to cause problems
505 (as reported sbcl-devel 2001-05-04). However, since the patches
506 modify nontrivial code which was apparently written incorrectly
507 the first time around, until regression tests are written I'm not
508 comfortable merging the patches in the CVS version of SBCL.
511 (DESCRIBE 'SB-ALIEN:DEF-ALIEN-TYPE) reports the macro argument list
515 in #<PACKAGE "SB-ALIEN">.
516 Macro-function: #<FUNCTION "DEF!MACRO DEF-ALIEN-TYPE" {19F4A39}>
517 Macro arguments: (#:whole-470 #:environment-471)
518 On Sat, May 26, 2001 09:45:57 AM CDT it was compiled from:
519 /usr/stuff/sbcl/src/code/host-alieneval.lisp
520 Created: Monday, March 12, 2001 07:47:43 AM CST
523 (TIME (ROOM T)) reports more than 200 Mbytes consed even for
524 a clean, just-started SBCL system. And it seems to be right:
525 (ROOM T) can bring a small computer to its knees for a *long*
526 time trying to GC afterwards. Surely there's some more economical
527 way to implement (ROOM T).
530 reported by Martin Atzmueller 2001-06-25; originally from CMU CL bugs
532 (in-package :cl-user)
533 ;;; The following invokes a compiler error.
534 (declaim (optimize (speed 2) (debug 3)))
537 (unwind-protect nil)))
541 The error message in sbcl-0.6.12.42 is
542 internal error, failed AVER:
543 "(COMMON-LISP:EQ (SB!C::TN-ENVIRONMENT SB!C:TN) SB!C::TN-ENV)"
546 When the compiler inline expands functions, it may be that different
547 kinds of return values are generated from different code branches.
548 E.g. an inline expansion of POSITION generates integer results
549 from one branch, and NIL results from another. When that inline
550 expansion is used in a context where only one of those results
553 (aref *a1* (position x *a2*)))
554 and the compiler can't prove that the unacceptable branch is
555 never taken, then bogus type mismatch warnings can be generated.
556 If you need to suppress the type mismatch warnings, you can
557 suppress the inline expansion,
559 #+sbcl (declare (notinline position)) ; to suppress bug 117 bogowarnings
560 (aref *a1* (position x *a2*)))
561 or, sometimes, suppress them by declaring the result to be of an
564 (aref *a1* (the integer (position x *a2*))))
566 This is not a new compiler problem in 0.7.0, but the new compiler
567 transforms for FIND, POSITION, FIND-IF, and POSITION-IF make it
568 more conspicuous. If you don't need performance from these functions,
569 and the bogus warnings are a nuisance for you, you can return to
570 your pre-0.7.0 state of grace with
571 #+sbcl (declaim (notinline find position find-if position-if)) ; bug 117..
574 as reported by Eric Marsden on cmucl-imp@cons.org 2001-08-14:
575 (= (FLOAT 1 DOUBLE-FLOAT-EPSILON)
576 (+ (FLOAT 1 DOUBLE-FLOAT-EPSILON) DOUBLE-FLOAT-EPSILON)) => T
577 when of course it should be NIL. (He says it only fails for X86,
578 not SPARC; dunno about Alpha.)
580 Also, "the same problem exists for LONG-FLOAT-EPSILON,
581 DOUBLE-FLOAT-NEGATIVE-EPSILON, LONG-FLOAT-NEGATIVE-EPSILON (though
582 for the -negative- the + is replaced by a - in the test)."
584 Raymond Toy comments that this is tricky on the X86 since its FPU
585 uses 80-bit precision internally.
588 Even in sbcl-0.pre7.x, which is supposed to be free of the old
589 non-ANSI behavior of treating the function return type inferred
590 from the current function definition as a declaration of the
591 return type from any function of that name, the return type of NIL
592 is attached to FOO in 120a above, and used to optimize code which
596 There was some sort of screwup in handling of
597 (IF (NOT (IGNORE-ERRORS ..))). E.g.
599 (if (not (ignore-errors
600 (make-pathname :host "foo"
604 (error "notunlessnot")))
605 The (NOT (IGNORE-ERRORS ..)) form evaluates to T, so this should be
606 printing "ok", but instead it's going to the ERROR. This problem
607 seems to've been introduced by MNA's HANDLER-CASE patch (sbcl-devel
608 2001-07-17) and as a workaround (put in sbcl-0.pre7.14.flaky4.12)
609 I reverted back to the old weird HANDLER-CASE code. However, I
610 think the problem looks like a compiler bug in handling RETURN-FROM,
611 so I left the MNA-patched code in HANDLER-CASE (suppressed with
612 #+NIL) and I'd like to go back to see whether this really is
613 a compiler bug before I delete this BUGS entry.
616 As of version 0.pre7.14, SBCL's implementation of MACROLET makes
617 the entire lexical environment at the point of MACROLET available
618 in the bodies of the macroexpander functions. In particular, it
619 allows the function bodies (which run at compile time) to try to
620 access lexical variables (which are only defined at runtime).
621 It doesn't even issue a warning, which is bad.
623 The SBCL behavior arguably conforms to the ANSI spec (since the
624 spec says that the behavior is undefined, ergo anything conforms).
625 However, it would be better to issue a compile-time error.
626 Unfortunately I (WHN) don't see any simple way to detect this
627 condition in order to issue such an error, so for the meantime
628 SBCL just does this weird broken "conforming" thing.
630 The ANSI standard says, in the definition of the special operator
632 The macro-expansion functions defined by MACROLET are defined
633 in the lexical environment in which the MACROLET form appears.
634 Declarations and MACROLET and SYMBOL-MACROLET definitions affect
635 the local macro definitions in a MACROLET, but the consequences
636 are undefined if the local macro definitions reference any
637 local variable or function bindings that are visible in that
639 Then it seems to contradict itself by giving the example
641 (macrolet ((fudge (z)
642 ;The parameters x and flag are not accessible
643 ; at this point; a reference to flag would be to
644 ; the global variable of that name.
645 ` (if flag (* ,z ,z) ,z)))
646 ;The parameters x and flag are accessible here.
650 The comment "a reference to flag would be to the global variable
651 of the same name" sounds like good behavior for the system to have.
652 but actual specification quoted above says that the actual behavior
656 (as reported by Gabe Garza on cmucl-help 2001-09-21)
658 (defun test-pred (x y)
662 (func (lambda () x)))
663 (print (eq func func))
664 (print (test-pred func func))
665 (delete func (list func))))
666 Now calling (TEST-CASE) gives output
669 (#<FUNCTION {500A9EF9}>)
670 Evidently Python thinks of the lambda as a code transformation so
671 much that it forgets that it's also an object.
674 The DEFSTRUCT section of the ANSI spec, in the :CONC-NAME section,
675 specifies a precedence rule for name collisions between slot accessors of
676 structure classes related by inheritance. As of 0.7.0, SBCL still
680 insufficient syntax checking in MACROLET:
682 (macrolet ((defmacro bar (z) `(+ z z)))
684 shouldn't compile without error (because of the extra DEFMACRO symbol).
687 Ideally, uninterning a symbol would allow it, and its associated
688 FDEFINITION and PROCLAIM data, to be reclaimed by the GC. However,
689 at least as of sbcl-0.7.0, this isn't the case. Information about
690 FDEFINITIONs and PROCLAIMed properties is stored in globaldb.lisp
691 essentially in ordinary (non-weak) hash tables keyed by symbols.
692 Thus, once a system has an entry in this system, it tends to live
693 forever, even when it is uninterned and all other references to it
697 (reported by Arnaud Rouanet on cmucl-imp 2001-12-18)
698 (defmethod foo ((x integer))
700 (defmethod foo :around ((x integer))
703 Now (FOO 3) should return 3, but instead it returns 4.
706 (reported by Alexey Dejneka sbcl-devel 2002-01-03)
708 SUBTYPEP does not work well with redefined classes:
717 * (defclass b (a) ())
730 This bug was fixed in sbcl-0.7.4.1 by invalidating the PCL wrapper
731 class upon redefinition. Unfortunately, doing so causes bug #176 to
732 appear. Pending further investigation, one or other of these bugs
733 might be present at any given time.
736 Pretty-printing nested backquotes doesn't work right, as
737 reported by Alexey Dejneka sbcl-devel 2002-01-13:
739 ``(FOO SB-IMPL::BACKQ-COMMA-AT S)
740 * (lisp-implementation-version)
744 (reported by Jesse Bouwman 2001-10-24 through the unfortunately
745 prominent SourceForge web/db bug tracking system, which is
746 unfortunately not a reliable way to get a timely response from
747 the SBCL maintainers)
748 In the course of trying to build a test case for an
749 application error, I encountered this behavior:
750 If you start up sbcl, and then lay on CTRL-C for a
751 minute or two, the lisp process will eventually say:
752 %PRIMITIVE HALT called; the party is over.
753 and throw you into the monitor. If I start up lisp,
754 attach to the process with strace, and then do the same
755 (abusive) thing, I get instead:
756 access failure in heap page not marked as write-protected
757 and the monitor again. I don't know enough to have the
758 faintest idea of what is going on here.
759 This is with sbcl 6.12, uname -a reports:
760 Linux prep 2.2.19 #4 SMP Tue Apr 24 13:59:52 CDT 2001 i686 unknown
761 I (WHN) have verified that the same thing occurs on sbcl-0.pre7.141
762 under OpenBSD 2.9 on my X86 laptop. Do be patient when you try it:
763 it took more than two minutes (but less than five) for me.
766 (This was once known as IR1-4, but it lived on even after the
767 IR1 interpreter went to the big bit bucket in the sky.)
768 The system accepts DECLAIM in most places where DECLARE would be
769 accepted, without even issuing a warning. ANSI allows this, but since
770 it's fairly easy to mistype DECLAIM instead of DECLARE, and the
771 meaning is rather different, and it's unlikely that the user
772 has a good reason for doing DECLAIM not at top level, it would be
773 good to issue a STYLE-WARNING when this happens. A possible
774 fix would be to issue STYLE-WARNINGs for DECLAIMs not at top level,
775 or perhaps to issue STYLE-WARNINGs for any EVAL-WHEN not at top level.
776 [This is considered an IR1-interpreter-related bug because until
777 EVAL-WHEN is rewritten, which won't happen until after the IR1
778 interpreter is gone, the system's notion of what's a top-level form
779 and what's not will remain too confused to fix this problem.]
782 ANSI allows types `(COMPLEX ,FOO) to use very hairy values for
783 FOO, e.g. (COMPLEX (AND REAL (SATISFIES ODDP))). The old CMU CL
784 COMPLEX implementation didn't deal with this, and hasn't been
785 upgraded to do so. (This doesn't seem to be a high priority
786 conformance problem, since seems hard to construct useful code
790 Floating point errors are reported poorly. E.g. on x86 OpenBSD
793 debugger invoked on condition of type SB-KERNEL:FLOATING-POINT-EXCEPTION:
794 An arithmetic error SB-KERNEL:FLOATING-POINT-EXCEPTION was signalled.
795 No traps are enabled? How can this be?
796 It should be possible to be much more specific (overflow, division
797 by zero, etc.) and of course the "How can this be?" should be fixable.
799 See also bugs #45.c and #183
802 In sbcl-0.7.1.3 on x86, COMPILE-FILE on the file
803 (in-package :cl-user)
806 (defstruct foo bar bletch)
808 (labels ((kidify1 (kid)
816 (declare (inline kid-frob))
819 (the simple-vector (foo-bar perd)))))
821 debugger invoked on condition of type TYPE-ERROR:
822 The value NIL is not of type SB-C::NODE.
823 The location of this failure has moved around as various related
824 issues were cleaned up. As of sbcl-0.7.1.9, it occurs in
825 NODE-BLOCK called by LAMBDA-COMPONENT called by IR2-CONVERT-CLOSURE.
828 Functions SUBTYPEP, TYPEP, UPGRADED-ARRAY-ELEMENT-TYPE, and
829 UPGRADED-COMPLEX-PART-TYPE should have an optional environment argument.
830 (reported by Alexey Dejneka sbcl-devel 2002-04-12)
833 (reported by Robert E. Brown 2002-04-16)
834 When a function is called with too few arguments, causing the
835 debugger to be entered, the uninitialized slots in the bad call frame
836 seem to cause GCish problems, being interpreted as tagged data even
837 though they're not. In particular, executing ROOM in the
838 debugger at that point causes AVER failures:
841 * (lisp-implementation-version)
847 failed AVER: "(SAP= CURRENT END)"
848 (Christophe Rhodes reports that this doesn't occur on the SPARC, which
849 isn't too surprising since there are many differences in stack
850 implementation and GC conservatism between the X86 and other ports.)
854 (in-package :cl-user)
856 (defmethod permanentize ((uustk uustk))
857 (flet ((frob (hash-table test-for-deletion)
859 (obj-entry.stale? (oe)
860 (destructuring-bind (key . datum) oe
861 (declare (type simple-vector key))
862 (deny0 (void? datum))
863 (some #'stale? key))))
864 (declare (inline frob obj-entry.stale?))
865 (frob (uustk.args-hash->obj-alist uustk)
867 (frob (uustk.hash->memoized-objs-list uustk)
870 in sbcl-0.7.3.11 causes an assertion failure,
873 (AND (NULL (BLOCK-SUCC B))
874 (NOT (BLOCK-DELETE-P B))
875 (NOT (EQ B (COMPONENT-HEAD #)))))"
878 In sbcl-0.7.3.11, compiling the (illegal) code
879 (in-package :cl-user)
880 (defmethod prove ((uustk uustk))
883 gives the (not terribly clear) error message
885 ; (during macroexpansion of (DEFMETHOD PROVE ...))
886 ; can't get template for (FROB NIL NIL)
887 The problem seems to be that the code walker used by the DEFMETHOD
888 macro is unhappy with the illegal syntax in the method body, and
889 is giving an unclear error message.
892 sbcl's treatment of at least macro lambda lists is too permissive;
893 e.g., in sbcl-0.7.3.7:
894 (defmacro foo (&rest rest bar) `(,bar ,rest))
895 (macroexpand '(foo quux zot)) -> (QUUX (QUUX ZOT))
896 whereas section 3.4.4 of the CLHS doesn't allow required parameters
897 to come after the rest argument.
900 The compiler sometimes tries to constant-fold expressions before
901 it checks to see whether they can be reached. This can lead to
902 bogus warnings about errors in the constant folding, e.g. in code
905 (WRITE-STRING (> X 0) "+" "0"))
906 compiled in a context where the compiler can prove that X is NIL,
907 and the compiler complains that (> X 0) causes a type error because
908 NIL isn't a valid argument to #'>. Until sbcl-0.7.4.10 or so this
909 caused a full WARNING, which made the bug really annoying because then
910 COMPILE and COMPILE-FILE returned FAILURE-P=T for perfectly legal
911 code. Since then the warning has been downgraded to STYLE-WARNING,
912 so it's still a bug but at least it's a little less annoying.
915 reported by Alexey Dejneka 08 Jun 2002 in sbcl-devel:
916 Playing with McCLIM, I've received an error "Unbound variable WRAPPER
917 in SB-PCL::CHECK-WRAPPER-VALIDITY".
918 (defun check-wrapper-validity (instance)
919 (let* ((owrapper (wrapper-of instance)))
920 (if (not (invalid-wrapper-p owrapper))
922 (let* ((state (wrapper-state wrapper)) ; !!!
924 I've tried to replace it with OWRAPPER, but now OBSOLETE-INSTANCE-TRAP
925 breaks with "NIL is not of type SB-KERNEL:LAYOUT".
927 partial fix: The undefined variable WRAPPER resulted from an error
928 in recent refactoring, as can be seen by comparing to the code in e.g.
929 sbcl-0.7.2. Replacing WRAPPER with OWRAPPER (done by WHN in sbcl-0.7.4.22)
930 should bring the code back to its behavior as of sbcl-0.7.2, but
931 that still leaves the OBSOLETE-INSTANCE-TRAP bug. An example of
932 input which triggers that bug is
934 (let ((lastname (intern (format nil "C~D" (1- i))))
935 (name (intern (format nil "C~D" i))))
936 (eval `(defclass ,name
937 (,@(if (= i 0) nil (list lastname)))
939 (eval `(defmethod initialize-instance :after ((x ,name) &rest any)
940 (declare (ignore any))))))
947 178: "AVER failure compiling confused THEs in FUNCALL"
948 In sbcl-0.7.4.24, compiling
950 (funcall (the function (the standard-object x))))
953 "(AND (EQ (IR2-CONTINUATION-PRIMITIVE-TYPE 2CONT) FUNCTION-PTYPE) (EQ CHECK T))"
954 This variant compiles OK, though:
955 (defun bug178alternative (x)
956 (funcall (the nil x)))
958 (since 0.7.8.9 it does not signal an error; see also bug 199)
960 183: "IEEE floating point issues"
961 Even where floating point handling is being dealt with relatively
962 well (as of sbcl-0.7.5, on sparc/sunos and alpha; see bug #146), the
963 accrued-exceptions and current-exceptions part of the fp control
964 word don't seem to bear much relation to reality. E.g. on
968 debugger invoked on condition of type DIVISION-BY-ZERO:
969 arithmetic error DIVISION-BY-ZERO signalled
970 0] (sb-vm::get-floating-point-modes)
972 (:TRAPS (:OVERFLOW :INVALID :DIVIDE-BY-ZERO)
973 :ROUNDING-MODE :NEAREST
974 :CURRENT-EXCEPTIONS NIL
975 :ACCRUED-EXCEPTIONS (:INEXACT)
978 * (sb-vm::get-floating-point-modes)
979 (:TRAPS (:OVERFLOW :INVALID :DIVIDE-BY-ZERO)
980 :ROUNDING-MODE :NEAREST
981 :CURRENT-EXCEPTIONS (:INEXACT)
982 :ACCRUED-EXCEPTIONS (:INEXACT)
985 185: "top-level forms at the REPL"
986 * (locally (defstruct foo (a 0 :type fixnum)))
989 ; (in macroexpansion of (SB-KERNEL::%DELAYED-GET-COMPILER-LAYOUT BAR))
990 however, compiling and loading the same expression in a file works
993 187: "type inference confusion around DEFTRANSFORM time"
994 (reported even more verbosely on sbcl-devel 2002-06-28 as "strange
995 bug in DEFTRANSFORM")
996 After the file below is compiled and loaded in sbcl-0.7.5, executing
997 (TCX (MAKE-ARRAY 4 :FILL-POINTER 2) 0)
998 at the REPL returns an adjustable vector, which is wrong. Presumably
999 somehow the DERIVE-TYPE information for the output values of %WAD is
1000 being mispropagated as a type constraint on the input values of %WAD,
1001 and so causing the type test to be optimized away. It's unclear how
1002 hand-expanding the DEFTRANSFORM would change this, but it suggests
1003 the DEFTRANSFORM machinery (or at least the way DEFTRANSFORMs are
1004 invoked at a particular phase) is involved.
1005 (cl:in-package :sb-c)
1006 (eval-when (:compile-toplevel)
1007 ;;; standin for %DATA-VECTOR-AND-INDEX
1008 (defknown %dvai (array index)
1010 (foldable flushable))
1011 (deftransform %dvai ((array index)
1015 (let* ((atype (continuation-type array))
1016 (eltype (array-type-specialized-element-type atype)))
1017 (when (eq eltype *wild-type*)
1018 (give-up-ir1-transform
1019 "specialized array element type not known at compile-time"))
1020 (when (not (array-type-complexp atype))
1021 (give-up-ir1-transform "SIMPLE array!"))
1022 `(if (array-header-p array)
1023 (%wad array index nil)
1024 (values array index))))
1025 ;;; standin for %WITH-ARRAY-DATA
1026 (defknown %wad (array index (or index null))
1027 (values (simple-array * (*)) index index index)
1028 (foldable flushable))
1029 ;;; (Commenting out this optimizer causes the bug to go away.)
1030 (defoptimizer (%wad derive-type) ((array start end))
1031 (let ((atype (continuation-type array)))
1032 (when (array-type-p atype)
1033 (values-specifier-type
1034 `(values (simple-array ,(type-specifier
1035 (array-type-specialized-element-type atype))
1037 index index index)))))
1039 (defun %wad (array start end)
1040 (format t "~&in %WAD~%")
1041 (%with-array-data array start end))
1042 (cl:in-package :cl-user)
1044 (declare (type (vector t) v))
1045 (declare (notinline sb-kernel::%with-array-data))
1046 ;; (Hand-expending DEFTRANSFORM %DVAI here also causes the bug to
1050 188: "compiler performance fiasco involving type inference and UNION-TYPE"
1051 (In sbcl-0.7.6.10, DEFTRANSFORM CONCATENATE was commented out until this
1052 bug could be fixed properly, so you won't see the bug unless you restore
1053 the DEFTRANSFORM by hand.) In sbcl-0.7.5.11 on a 700 MHz Pentium III,
1057 (declare (optimize (safety 3)))
1058 (declare (optimize (compilation-speed 2)))
1059 (declare (optimize (speed 1) (debug 1) (space 1)))
1060 (let ((fn "if-this-file-exists-the-universe-is-strange"))
1061 (load fn :if-does-not-exist nil)
1062 (load (concatenate 'string fn ".lisp") :if-does-not-exist nil)
1063 (load (concatenate 'string fn ".fasl") :if-does-not-exist nil)
1064 (load (concatenate 'string fn ".misc-garbage")
1065 :if-does-not-exist nil)))))
1067 134.552 seconds of real time
1068 133.35156 seconds of user run time
1069 0.03125 seconds of system run time
1070 [Run times include 2.787 seconds GC run time.]
1072 246883368 bytes consed.
1073 BACKTRACE from Ctrl-C in the compilation shows that the compiler is
1074 thinking about type relationships involving types like
1076 (OR (INTEGER 576 576)
1087 190: "PPC/Linux pipe? buffer? bug"
1088 In sbcl-0.7.6, the run-program.test.sh test script sometimes hangs
1089 on the PPC/Linux platform, waiting for a zombie env process. This
1090 is a classic symptom of buffer filling and deadlock, but it seems
1091 only sporadically reproducible.
1093 191: "Miscellaneous PCL deficiencies"
1094 (reported by Alexey Dejneka sbcl-devel 2002-08-04)
1095 a. DEFCLASS does not inform the compiler about generated
1096 functions. Compiling a file with
1097 (DEFCLASS A-CLASS ()
1099 (DEFUN A-CLASS-X (A)
1100 (WITH-SLOTS (A-CLASS-X) A
1102 results in a STYLE-WARNING:
1104 SB-SLOT-ACCESSOR-NAME::|COMMON-LISP-USER A-CLASS-X slot READER|
1106 APD's fix for this was checked in to sbcl-0.7.6.20, but Pierre
1107 Mai points out that the declamation of functions is in fact
1108 incorrect in some cases (most notably for structure
1109 classes). This means that at present erroneous attempts to use
1110 WITH-SLOTS and the like on classes with metaclass STRUCTURE-CLASS
1111 won't get the corresponding STYLE-WARNING.
1112 c. the examples in CLHS 7.6.5.1 (regarding generic function lambda
1113 lists and &KEY arguments) do not signal errors when they should.
1115 192: "Python treats free type declarations as promises."
1116 b. What seemed like the same fundamental problem as bug 192a, but
1117 was not fixed by the same (APD "more strict type checking
1118 sbcl-devel 2002-08-97) patch:
1119 (DOTIMES (I ...) (DOTIMES (J ...) (DECLARE ...) ...)):
1120 (declaim (optimize (speed 1) (safety 3)))
1121 (defun trust-assertion (i)
1123 (declare (type (mod 4) i)) ; when commented out, behavior changes!
1126 (trust-assertion 6) ; prints nothing unless DECLARE is commented out
1128 193: "unhelpful CLOS error reporting when the primary method is missing"
1130 (defmethod foo :before ((x t)) (print x))
1131 is the only method defined on FOO, the error reporting when e.g.
1133 is relatively unhelpful:
1134 There is no primary method for the generic function
1135 #<STANDARD-GENERIC-FUNCTION FOO (1)>.
1136 with the offending argument nowhere visible in the backtrace. This
1137 continues even if there *are* primary methods, just not for the
1138 specified arg type, e.g.
1139 (defmethod foo ((x character)) (print x))
1140 (defmethod foo ((x string)) (print x))
1141 (defmethod foo ((x pathname)) ...)
1142 In that case it could be very helpful to know what argument value is
1143 falling through the cracks of the defined primary methods, but the
1144 error message stays the same (even BACKTRACE doesn't tell you what the
1145 bad argument value is).
1147 194: "no error from (THE REAL '(1 2 3)) in some cases"
1150 (multiple-value-prog1 (progn (the real '(1 2 3))))
1151 returns (1 2 3) instead of signalling an error. This was fixed by
1152 APD's "more strict type checking patch", but although the fixed
1153 code (in sbcl-0.7.7.19) works (signals TYPE-ERROR) interactively,
1154 it's difficult to write a regression test for it, because
1155 (IGNORE-ERRORS (MULTIPLE-VALUE-PROG1 (PROGN (THE REAL '(1 2 3)))))
1156 still returns (1 2 3).
1158 b. (IGNORE-ERRORS (MULTIPLE-VALUE-PROG1 (PROGN (THE REAL '(1 2 3)))))
1159 returns (1 2 3). (As above, this shows up when writing regression
1160 tests for fixed-ness of part a.)
1161 c. Also in sbcl-0.7.7.9, (IGNORE-ERRORS (THE REAL '(1 2 3))) => (1 2 3).
1163 (null (ignore-errors
1165 (arg2 (identity (the real #(1 2 3)))))
1166 (if (< arg1 arg2) arg1 arg2))))
1168 but putting the same expression inside (DEFUN FOO () ...),
1171 * Actually this entry is probably multiple bugs, as
1172 Alexey Dejneka commented on sbcl-devel 2002-09-03:)
1173 I don't think that placing these two bugs in one entry is
1174 a good idea: they have different explanations. The second
1175 (min 1 nil) is caused by flushing of unused code--IDENTITY
1176 can do nothing with it. So it is really bug 122. The first
1177 (min nil) is due to M-V-PROG1: substituting a continuation
1178 for the result, it forgets about type assertion. The purpose
1179 of IDENTITY is to save the restricted continuation from
1180 inaccurate transformations.
1181 * Alexey Dejneka pointed out that
1182 (IGNORE-ERRORS (IDENTITY (THE REAL '(1 2 3))))
1183 works as it should. Also
1184 (IGNORE-ERRORS (VALUES (THE REAL '(1 2 3))))
1185 works as it should. Perhaps this is another case of VALUES type
1186 intersections behaving in non-useful ways?
1188 199: "hairy FUNCTION types confuse the compiler"
1189 (reported by APD sbcl-devel 2002-09-15)
1191 (EQ NIL (FUNCALL F)))
1194 (DECLARE (TYPE (AND FUNCTION (SATISFIES MUR)) F))
1197 fails to compile, printing
1199 "(AND (EQ (IR2-CONTINUATION-PRIMITIVE-TYPE 2CONT) FUNCTION-PTYPE) (EQ CHECK T))"
1201 APD further reports that this bug is not present in CMUCL.
1203 (this case was fixed in 0.7.8.9; see also bug 178)
1205 201: "Incautious type inference from compound CONS types"
1206 (reported by APD sbcl-devel 2002-09-17)
1208 (LET ((Y (CAR (THE (CONS INTEGER *) X))))
1210 (FORMAT NIL "~S IS ~S, Y = ~S"
1217 (FOO ' (1 . 2)) => "NIL IS INTEGER, Y = 1"
1220 Compiler does not check THEs on unused values, e.g. in
1222 (progn (the real (list 1)) t)
1224 This situation may appear during optimizing away degenerate cases of
1225 certain functions: see bugs 54, 192b.
1227 DEFUNCT CATEGORIES OF BUGS
1229 These labels were used for bugs related to the old IR1 interpreter.
1230 The # values reached 6 before the category was closed down.