bug#65092: 30.0.50; C-h f shows malformed description for vc-mode

[Eli Zaretskii <eliz@gnu.org>]

> Date: Sat, 05 Aug 2023 14:25:11 +0300
> From:  Eshel Yaron via "Bug reports for GNU Emacs,
>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
> 
> 
> I seems to have stumbled upon an issue with type inference on the master
> branch that causes `describe-function` to display the following
> malformed description for the function `vc-mode`:
> 
>     vc-mode is a native-compiled Lisp function in ‘vc-hooks.el’.
> 
>     (vc-mode &optional ARG)
> 
>     Inferred type: (function (&optional t) (member Version Control minor
>     mode.  This minor mode is automatically activated whenever you visit a
>     file under control of one of the revision control systems in
>     `vc-handled-backends'.  VC commands are globally reachable under the
>     prefix `\[vc-prefix-map]': \{vc-prefix-map}))
> 
>     Version Control minor mode.
>     This minor mode is automatically activated whenever you...
> 
> To reproduce, do `C-h f vc-mode RET` in `emacs -Q`.
> 
> Basically, it looks like the inferred type somehow includes the
> function's docstring verbatim:
> 
> --8<---------------cut here---------------start------------->8---
> (comp-function-type-spec 'vc-mode)
> => ((function (&optional t) (member "Version Control minor mode.
> This minor mode is automatically activated whenever you visit a file under
> control of one of the revision control systems in `vc-handled-backends'.
> VC commands are globally reachable under the prefix `\\[vc-prefix-map]':
> \\{vc-prefix-map}")) . inferred)
> --8<---------------cut here---------------end--------------->8---

Probably because vc-mode is a "dummy function", whatever that means.
IOW, we are shooting ourselves in the foot, it seems like.

Adding Andrea.