[Gcl-devel] Test failures (complete list)

[Paul F. Dietz]

Here's the complete test failure classification.  I've
removed the make-array tests that now pass.



Bugs causing test failures in gcl (2/10/03):

1. LOAD-TIME-VALUE and SYMBOL-MACROLET should be special operators
  (special-operator-p.1)

2. DECLARE shouldn't be a special operator (special-operator-p.2)

3. The following builtin functions do not exist:

(ARITHMETIC-ERROR-OPERANDS BROADCAST-STREAM-STREAMS
 CELL-ERROR-NAME CONCATENATED-STREAM-STREAMS
 COPY-PPRINT-DISPATCH ECHO-STREAM-INPUT-STREAM
 ECHO-STREAM-OUTPUT-STREAM ENSURE-DIRECTORIES-EXIST
 FUNCTION-LAMBDA-EXPRESSION GET-SETF-EXPANSION
 HASH-TABLE-REHASH-SIZE HASH-TABLE-REHASH-THRESHOLD
 INTERACTIVE-STREAM-P
 LOAD-LOGICAL-PATHNAME-TRANSLATIONS LOGICAL-PATHNAME
 MAKE-LOAD-FORM-SAVING-SLOTS OPEN-STREAM-P
 PATHNAME-MATCH-P PPRINT-DISPATCH PPRINT-FILL
 PPRINT-INDENT PPRINT-LINEAR PPRINT-NEWLINE PPRINT-TAB
 PPRINT-TABULAR PRINT-NOT-READABLE-OBJECT
 READ-SEQUENCE SET-PPRINT-DISPATCH
 SIMPLE-CONDITION-FORMAT-CONTROL
 STREAM-EXTERNAL-FORMAT SYNONYM-STREAM-SYMBOL
 TRANSLATE-LOGICAL-PATHNAME TRANSLATE-PATHNAME
 TWO-WAY-STREAM-INPUT-STREAM
 TWO-WAY-STREAM-OUTPUT-STREAM UNBOUND-SLOT-INSTANCE
 UPGRADED-ARRAY-ELEMENT-TYPE
 UPGRADED-COMPLEX-PART-TYPE WILD-PATHNAME-P
 WRITE-SEQUENCE COMPILER-MACRO-FUNCTION FDEFINITION
 LOGICAL-PATHNAME-TRANSLATIONS READTABLE-CASE)

4. the following macros don't exist:

(DEFINE-COMPILER-MACRO DEFINE-SETF-EXPANDER
 DEFINE-SYMBOL-MACRO FORMATTER IN-PACKAGE
 PPRINT-LOGICAL-BLOCK PRINT-UNREADABLE-OBJECT
 WITH-COMPILATION-UNIT WITH-CONDITION-RESTARTS
 WITH-STANDARD-IO-SYNTAX)

5. The following variables don't exist:

(*COMPILE-FILE-PATHNAME* *COMPILE-FILE-TRUENAME*
 *COMPILE-PRINT* *COMPILE-VERBOSE* *LOAD-PATHNAME*
 *LOAD-PRINT* *LOAD-TRUENAME* *PRINT-LINES*
 *PRINT-MISER-WIDTH* *PRINT-PPRINT-DISPATCH*
 *PRINT-RIGHT-MARGIN* *READ-EVAL*)

6. The following classes don't exist

(BROADCAST-STREAM CONCATENATED-STREAM ECHO-STREAM
 FILE-STREAM FLOATING-POINT-INEXACT
 FLOATING-POINT-INVALID-OPERATION FUNCTION HASH-TABLE
 LOGICAL-PATHNAME PACKAGE PARSE-ERROR PATHNAME
 PRINT-NOT-READABLE RANDOM-STATE READER-ERROR
 READTABLE REAL STREAM STRING-STREAM STYLE-WARNING
 SYNONYM-STREAM TWO-WAY-STREAM UNBOUND-SLOT)

7. COMPILE should return three results, not one.  (COMPILE.2)

8. COMPILE should preserve object identity (EQL-ness)
  in the term being compiled.  (COMPILE.5,6,7,8)

9. CONSTANTP should take an environment (or NIL) as its
  optional second argument.

10, SETF of VALUES is getting the order of evaluation of
 subforms wrong (SETF-VALUES.5)

11. SYMBOL-MACROLET interaction with PSETQ/PSETF/MULTIPLE-VALUE-SETQ
  (PSETQ.7, PSETF.7, MULTIPLE-VALUE-SETQ.5,8)

12. THROW to a nonexistant catch tag should cause a CONTROL-ERROR,
  not a STREAM-ERROR (test THROW-ERROR).

13. FBOUDNP isn't working on names of the form (SETF ...).  (FBOUNDP.6,7,
   FMAKUNBOUND.4, and three STRUCT tests) FLET doesn't work on names
   of that form.  (FLET 17)  #' doesn't either (FUNCTION.7).  FUNCTIONP
   does not (FUNCTIONP.7)  LABELS does not (LABELS.17)

14. FLET isn't handling :ALLOW-OTHER-KEYS arguments properly.
  (FLET.27)

15. FLET isn't handling the (keyword variable) syntax of ordinary
  lambda lists (FLET.30)

16. FUNCTION-LAMBDA-EXPRESSION and FDEFINITION are undefined

17. symbols and (lambda ...) forms are being treated as elements
  of type FUNCTION (this is an incompatible CLTL1/ANSI CL change).
  (FUNCTION.2,6; FUNCTIONP.2,6

18. The types FUNCTION and CONS should be disjoint, but they aren't.
  (FUNCTION.10, FUNCTIONP.10)

19, GET-SETF-EXPANSION is undefined.

20. MACROLET doesn't understand &WHOLE arguments.  (MACROLET.5)

21. SYMBOL-MACROLET and MACROLET expansion interact incorrectly.
   (MACROLET.13,14)

22. LOOP ACROSS on displaced bit vectors is ignoring the displacement
   (LOOP.5.38)

23. (LOOP FOR NIL BEING THE HASH-KEYS OF ...) is generating some
   hard to interpret error (LOOP.6.19,20)

24. (LOOP ... USING (HASH-KEY (k1 . k2)) ...) isn't working.
   (LOOP.6.37); similarly for (HASH-VALUE (v1 . v2)) (LOOP.6.38)

25. Various errors for LOOP on the symbols of packages:  the iteration
   control forms do not allow *PACKAGE* to be used as the default package.
   The forms should throw a PACKAGE-ERROR when the package specifier
   does not specify an existing package, not a TYPE-ERROR.  NIL is not
   being accepted as a destructing bind list in these forms (meaning:
   don't bind the symbol to anything.)  (various LOOP.7.* failures)

26. Destructuring problems in LOOP WITH clauses (LOOP.8.19,20,21,22)

27. Failure to detect conflicting bindings of the same name in LOOP
   forms (LOOP.8.ERROR.2, LOOP.9.10,11,27,28,37,38, LOOP.10.9,10,37,38,
   57,58,85,86

28. REPEAT clauses are not being allowed where they should be.
   LOOP.11.9

29. 'IT' in loop clauses applies only to the first subclause in
   a conditional, not to the one after the 'AND'.  (LOOP.14.29)

30. SUBTYPEP failures on conditions.  The following pairs should
   have the subtype relationship, but don't (CONDITION.2):

   (SIMPLE-ERROR SIMPLE-CONDITION)
   (SIMPLE-ERROR ERROR)
   (SIMPLE-ERROR SERIOUS-CONDITION)
   (SIMPLE-ERROR CONDITION)

31. CELL-ERROR-NAME is undefined.

32. SIMPLE-CONDITION-FORMAT-CONTROL is undefined.

33. Multiple :ALLOW-OTHER-KEYS arguments broken in MAKE-LIST, ADJOIN, MEMBER,


34. FIRST should throw a program error at a safe call with no args,
   or with too many args.

35. LAST should throw a program error at a safe call with no args,
   or with too many args.

36. (SETF (GETF ...) ...) is evaluating arguments in the wrong order
   when the ignored third argument to GETF is present. (SETF-GETF.7)

37. (ARRAY * 1) is not being treated by TYPEP as a valid type specifier
   (ARRAY.4.1,3,4,5,6)  Similar problems in ARRAY.5.1,2,3,4,5,6, and
   ARRAY.7.1-6 and ARRAY-T.[456].*

38. (fixed)

39. 'ARRAY should be equivalent to the class ARRAY in SUBTYPEP.
    (ARRAY-AS-CLASS.7,8)

40. Various failures of things to be simple-arrays (simple-array.[457].*;
    simple-array-t.[457].*)

41. Large numbers of MAKE-ARRAY.* tests due to the :FAIL-NOT-ARRAY-CLASS
    subtest.  This is a test that the array is in (FIND-CLASS 'ARRAY).

42. ADJUSTABLE-ARRAY-P should signal a type error if its argument
    is not an array. (ADJUSTABLE-ARRAY-P.ERROR.4,5,6)

43. ARRAY-DISPLACEMENT is not working on displaced strings or bit strings.
    (MAKE-ARRAY.DISPLACED.9,10,11,12,13,14,15,16,
     ARRAY-DISPLACEMENT.9,10,11,12)

44. ARRAY-DISPLACEMENT should throw a program error in safe calls
   on incorrect numbers of arguments. and a type error when the argument
   is not an array. (ARRAY-DISPLACEMENT.ERROR.1,2,3,4,5)

45. ARRAY-DIMENSIONS should throw a type error when the argument
   is not an array.  (ARRAY-DIMENSIONS.3,4,5)

46 ARRAY-RANK should throw a type error when the argument is not an
   array. (ARRAY-RANK.ERROR.3,4,5)

47. ARRAY-TOTAL-SIZE should throw a type error when the argument is not
   an array. (ARRAY-TOTAL-SIZE.ERROR.3,4,5)

48. FILL-POINTER should throw a type error when the argument is not
   a vector with a fill pointer. (FILL-POINTER.ERROR.3,4,5,6,7)

49. UPGRADED-ARRAY-ELEMENT-TYPE isn't defined.
   (UPGRADED-ARRAY-ELEMENT-TYPE.*)

50. TYPEP is messing up on a number of different vector types:
  -- Strings aren't of type (VECTOR T) (neither are bit strings).
  -- The length in vector type specifiers is being ignored.
  -- It isn't handling the class VECTOR properly.
    (VECTOR.TYPE.13,15,17,18,20,21,24,25,27,28,29,30,33,34,39,40)

51. VECTOR-POP should signal a type error on vectors without fill
  pointers. (VECTOR-POP.ERROR.1,5)

52. Bit logical operations are not working on bit arrays of dimension
  2.(BIT-AND.12, BIT-AND.14, BIT-AND.16, BIT-AND.17, BIT-AND.18,
  BIT-AND.19, BIT-ANDC1.12, BIT-ANDC1.14, BIT-ANDC1.16, BIT-ANDC1.17,
  BIT-ANDC1.18, BIT-ANDC1.19, BIT-ANDC2.12, BIT-ANDC2.14, BIT-ANDC2.16,
  BIT-ANDC2.17, BIT-ANDC2.18, BIT-ANDC2.19, BIT-EQV.12, BIT-EQV.14,
  BIT-EQV.16, BIT-EQV.17, BIT-EQV.18, BIT-EQV.19, BIT-IOR.12,
  BIT-IOR.14, BIT-IOR.16, BIT-IOR.17, BIT-IOR.18, BIT-IOR.19,
  BIT-NAND.12, BIT-NAND.14, BIT-NAND.16, BIT-NAND.17, BIT-NAND.18,
  BIT-NAND.19, BIT-NOR.12, BIT-NOR.14, BIT-NOR.16, BIT-NOR.17,
  BIT-NOR.18, BIT-NOR.19, BIT-ORC1.12, BIT-ORC1.14, BIT-ORC1.16,
  BIT-ORC1.17, BIT-ORC1.18, BIT-ORC1.19, BIT-ORC2.12, BIT-ORC2.14,
  BIT-ORC2.16, BIT-ORC2.17, BIT-ORC2.18, BIT-ORC2.19, BIT-XOR.12,
  BIT-XOR.14, BIT-XOR.16, BIT-XOR.17, BIT-XOR.18, BIT-XOR.19,
  BIT-NOT.10, BIT-NOT.11, BIT-NOT.14, BIT-NOT.15)

53. IN-PACKAGE should be a macro, not a function. (IN-PACKAGE-2)

54. The STRUCT-TEST-*/15A tests fail because the classes of
   the structure types are not subclasses of STRUCTURE-OBJECT
   (as reported by SUBTYPEP).

55. STRUCT-TEST-07* tests fail because :CONC-NAME is not accepted
   as an option to defstruct.

56. STRUCT-TEST-07* tests fail because (:CONC-NAME) is not accepted
   as an option to defstruct.

57. Some structure tests are failing because 'non-symbol and
   non-interned symbol slot name accessorsare not yet implemented'
   (STRUCT-TEST-30*)

58. (FIND-CLASS '#:STRUCT-TEST-31) is failing for some reason.
   (CMUCL had problems with these due to the uninterned symbol).

59. STRUCT-TEST-33,34,35 tests: the accessor function may be in the
   wrong package.

60. Improper collision of slot names with special variables
   (the constructor built by DEFSTRUCT should use different parameter
    names to avoid collision with special variables)
   (STRUCTURE-60-1)

61. BOA constructors fail for some reason: SBT-08, SBT-09, SBT-13,
   SBT-14, SBT-15

62. TYPES-3 is failing because of this oddness:

   CL-TEST>(subtypep '(and standard-object t) 'standard-object)

   Error in SUBTYPEP [or a callee]: (STANDARD-OBJECT) is not of type STRING.

63. TYPES-7B fails for a similar reason:

   CL-TEST>(subtypep 'condition '(not cons))

   Error in SUBTYPEP [or a callee]: (CONDITION) is not of type STRING.

64. TYPES-7C fails likewise:

   CL-TEST>(subtypep 'generic-function '(not cons))

   Error in SUBTYPEP [or a callee]: (GENERIC-FUNCTION) is not of type STRING.

  (Other similar failures: USER-CLASS-DISJOINTNESS, USER-STRUCT-DISJOINTNESS,
   TAC-3.16)


65. TYPES-9A fails because lambda closures aren't atoms.

66. Many standard classes are not subtypes of STANDARD-OBJECT

67. Two structure classes are not subtypes of STRUCTURE-OBJECT

68. The following external symbols of the COMMON-LISP package should
  either not be external or should not name classes:

  (STANDARD-SLOT-DEFINITION STACK-OVERFLOW SLOT-DEFINITION
   SPECIALIZER STORAGE-EXHAUSTED
   EFFECTIVE-SLOT-DEFINITION STANDARD-WRITER-METHOD
   STANDARD-READER-METHOD EQL-SPECIALIZER
   DIRECT-SLOT-DEFINITION FORWARD-REFERENCED-CLASS
   STANDARD-EFFECTIVE-SLOT-DEFINITION
   STANDARD-ACCESSOR-METHOD
   STANDARD-DIRECT-SLOT-DEFINITION
   FUNCALLABLE-STANDARD-CLASS)

69. There are five standard generic functions that are either
   not generic functions or are ordinary non-generic functions.

70. TYPE-ERROR-DATUM should throw a program error on an incorrect
   number of arguments

71. TYPEP should take an optional third argument (NIL or an environment)

72. SUBTYPEP does not recognize subclass relationships on user-defined
   classes.  (TAC-3.1,2,3,4,5,6,7,8,9,10,11,12,13,14,15)

73. Should be able to coerce '(1 0) to a vector (COERCE.11)

74. Coercing a symbol to type 'FUNCTION should give a function object,
    not the symbol (CLTL1/ANSI incompatibility, COERCE.21)

75. Coercions that cannot occur should be TYPE-ERRORs, not STREAM-ERRORs
    (COERCE.ERROR.4,9)

76. SUBTYPEP should take an optional third argument (NIL or environment)
    (SUBTYPE.ENV.1,2)

77. SUBTYPEP on integer types has problems with some integer subranges
    that have exclusive endpoints (SUBTYPE.BIGNUM.INTEGER,
    SUBTYPEP.INTEGER.23, SUBTYPEP.INTEGER.24)

78. SUBTYPEP fails on some CONS type specifiers (SUBTYPEP.CONS.2)

79. DIGIT-CHAR should signal a program-error on three arguments,
   not check the third argument for its type. (DIGIT-CHAR.ERROR.2)

80. CODE-CHAR should throw a program-error on excessive arguments
   (CODE-CHAR.ERROR.2)