bug#18190: 24.4.50; doc string of `auth-source-backend-child-p'

[Drew Adams <drew.adams@oracle.com>]

> > Incomprehensible (gibberish).  Probably a typo or two in that sentence
> > somewhere.
> 
> Looks fine now, at least:
> 
> auth-source-backend-child-p is an alias for
> 'auth-source-backend--eieio-childp'.
> 
> (auth-source-backend-child-p OBJ)
> 
> This function is obsolete since 25.1;
> use (cl-typep ... \='auth-source-backend) instead
> 
> Return non-nil if OBJ is an object of type 'auth-source-backend' or a
> subclass.

That's certainly a lot better than what it was when I reported
the bug:

 Test OBJ to see if it an object is a child of type auth-source-backend

But if you try to understand what it says now, `C-h f cl-typep'
tells you only:

  This function has a compiler macro 'cl-typep--inliner'.

  Not documented.

One problem at a time, I guess...

Is there perhaps a typo in this part?

  (cl-typep ... \='auth-source-backend)

What does the ... correspond to?  And what's with \='?