bug#50968: 28.0.60; curly quote in lisp form in docstrings

[David Fussner via "Bug reports for GNU Emacs, the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>]

All looks well here after a bootstrap -- thank you very much for the
quick fix, Lars.

David.

On Sun, 3 Oct 2021 at 12:01, Eli Zaretskii <eliz@gnu.org> wrote:
>
> > From: Lars Ingebrigtsen <larsi@gnus.org>
> > Date: Sun, 03 Oct 2021 11:36:13 +0200
> > Cc: 50968@debbugs.gnu.org
> >
> > David Fussner <dfussner@googlemail.com> writes:
> >
> > > To check whether the minor mode is enabled in the current buffer,
> > > evaluate ‘(default-value ’show-paren-mode)’.
> > >
> > > -- end quotation --
> > >
> > > The straight quote inside the form has become curly (unless gmail has
> > > mangled it), preventing evaluation. This part of the docstring comes
> > > from a template in
> > > easy-mmode.el, so versions of it appear everywhere.
> >
> > Yup, this should now be fixed in Emacs 28 (but requires recompiling all
> > files that use easy-mmode).
>
> I 'touch'ed all the *.el files that use define-minor-mode, but then
> rebuilding Emacs with native-compilation fails with error messages
> that complain about frame-XXX.eln:
>
>     ELC+ELN  ../lisp/international/mule-cmds.elc
>   emacs: can't find function "F616e6f6e796d6f75732d6c616d626461_anonymous_lambda_41" in compilation unit d:\gnu\git\emacs\native-comp\src\../native-lisp/28.0.60-bc5f6bd2/preloaded/frame-b40fc590-734bdcac.eln
>   Makefile:286: recipe for target `../lisp/international/mule-cmds.elc' failed
>
> I saw others report similar problems recently.
>
> I worked around this by removing emacs and bootstrap-emacs executables
> (and the corresponding pdmp files), which causes a re-dump.
>
> There were no such problems in the build without native-compilation.
>
> frame.el was one of the 'touch'ed files, so it was recompiled, and it
> sounds like the problem is with some discrepancy between the old
> frame-XXX.eln file with which emacs was dumped, and the new one?  And
> it's somehow triggered by refreshing the preloaded *.eln files that
> were meanwhile recompiled?
>
> Andrea, what do these error messages mean, and how to avoid them in
> the future?  Do we lack some cleanup or some prerequisites in our
> Makefile's?