cl:describe doesn't show documentation for nameless functions (and closures)

[Mikko Markus Torni]

(let ((fun (lambda () "%DOCSTRING%" NIL)))
  (setf (documentation fun 'function) "%SETF-DOCUMENTATION%")
  (describe fun))

This prints:

  #<FUNCTION (lambda ())>
  [...]
  Documentation:
    T
  [...]

I expect to see:

  [...]
  Documentation:
    %SETF-DOCUMENTATION%
  [...]

or even:

  [...]
  Documentation:
    %DOCSTRING%
  [...]

A clue to what's happening is:

(let ((fun (lambda () "%DOCSTRING%" nil)))
  (setf (sb-impl::%fun-name fun) 'CAR)
  (describe fun))

  [...]
  Documentation:
    Return the 1st object in a list.
  [...]

The culprit seems to in DESCRIBE-FUNCTION in src/code/describe.lisp:

(defun describe-function (name function stream)
  (let ((name (if function (%fun-name function) name)))
    ;; CUT ;;
    (describe-documentation name 'function stream)
    ;; ^^^^\ Should perhaps be FUNCTION instead of NAME?
    ;; CUT ;;

I have included a patch making the above change.

$ sbcl --version
SBCL 1.4.16.debian
$ uname -a
Linux premium 4.9.0-2-amd64 #1 SMP Debian 4.9.18-1 (2017-03-30) x86_64 GNU/Linux

CL-USER> *features*
(:CLOSER-MOP :SPLIT-SEQUENCE :NAMED-READTABLES :CL-PPCRE
 ALEXANDRIA.0.DEV::SEQUENCE-EMPTYP :SWANK :QUICKLISP :SB-BSD-SOCKETS-ADDRINFO
 :ASDF3.3 :ASDF3.2 :ASDF3.1 :ASDF3 :ASDF2 :ASDF :OS-UNIX
 :NON-BASE-CHARS-EXIST-P :ASDF-UNICODE :X86-64 :64-BIT :64-BIT-REGISTERS
 :ALIEN-CALLBACKS :ANSI-CL :AVX2 :C-STACK-IS-CONTROL-STACK :CALL-SYMBOL
 :COMMON-LISP :COMPACT-INSTANCE-HEADER :COMPARE-AND-SWAP-VOPS :CYCLE-COUNTER
 :ELF :FP-AND-PC-STANDARD-SAVE :GENCGC :IEEE-FLOATING-POINT :IMMOBILE-CODE
 :IMMOBILE-SPACE :INTEGER-EQL-VOP :LARGEFILE :LINKAGE-TABLE :LINUX
 :LITTLE-ENDIAN :OS-PROVIDES-BLKSIZE-T :OS-PROVIDES-DLADDR :OS-PROVIDES-DLOPEN
 :OS-PROVIDES-GETPROTOBY-R :OS-PROVIDES-POLL :OS-PROVIDES-PUTWC
 :OS-PROVIDES-SUSECONDS-T :PACKAGE-LOCAL-NICKNAMES :RELOCATABLE-HEAP
 :SB-CORE-COMPRESSION :SB-DOC :SB-EVAL :SB-FUTEX :SB-LDB :SB-PACKAGE-LOCKS
 :SB-SIMD-PACK :SB-SIMD-PACK-256 :SB-SOURCE-LOCATIONS :SB-THREAD :SB-UNICODE
 :SB-XREF-FOR-INTERNALS :SBCL :STACK-ALLOCATABLE-CLOSURES
 :STACK-ALLOCATABLE-FIXED-OBJECTS :STACK-ALLOCATABLE-LISTS
 :STACK-ALLOCATABLE-VECTORS :STACK-GROWS-DOWNWARD-NOT-UPWARD
 :UNDEFINED-FUN-RESTARTS :UNIX :UNWIND-TO-FRAME-AND-CALL-VOP)

[Douglas Katzman (dougk)]

I wonder if you wouldn't mind seeing how this affects some strange cases as in the following example? I don't have a good idea of what things "should" do, but it might be nice to put in comments about expected behaviors, rather than just having it be quasi-random/inexplicable and then changing it be a different behavior that is equally so.
(I'm certainly not suggesting that the patch breaks anything - it looks good to me).

* (defun foo () "i am foo" 1)
FOO
* (defparameter thing #'foo)
THING
* (defun foo () "no i am foo" 2)
WARNING: redefining COMMON-LISP-USER::FOO in DEFUN
FOO
* (setf (symbol-function 'bar) #'foo)
#<FUNCTION FOO>
* (setf (documentation #'bar 'function) "here is a string for bar")
"here is a string for bar"

* (describe 'foo)
COMMON-LISP-USER::FOO
  [symbol]
FOO names a compiled function:
  Lambda-list: ()
  Derived type: (FUNCTION NIL (VALUES (INTEGER 2 2) &OPTIONAL))
  Source form:
    (LAMBDA () "no i am foo" (BLOCK FOO 2))
* (documentation #'foo 'function)
"here is a string for bar"

* (describe 'bar)
COMMON-LISP-USER::BAR
  [symbol]
BAR names a compiled function:
  Lambda-list: ()
  Derived type: (FUNCTION NIL (VALUES (INTEGER 2 2) &OPTIONAL))
  Source form:
    (LAMBDA () "no i am foo" (BLOCK FOO 2))
* (describe #'bar)
#<FUNCTION FOO>
  [compiled function]
Lambda-list: ()
Derived type: (FUNCTION NIL (VALUES (INTEGER 2 2) &OPTIONAL))
Documentation:
  here is a string for bar
Source form:
  (LAMBDA () "no i am foo" (BLOCK FOO 2))