[I posted a query
https://lists.gnu.org/archive/html/help-gnu-emacs/2025-03/msg00005.html
on help-gnu-emacs but the discussion didn't proceed, so I'm reposting
here]

Attached is a test case named `debug-font-lock-foo-spec-mode.el' which
sets up a minor-mode foo-spec-mode for *.foospec files which
highlights the string "FOO-SPEC" with font-lock-warning-space.

On invoking the following sequence
```
echo FOO-SPEC > /tmp/foo.foospec
emacs -Q -l debug-font-lock-foo-spec-mode.el /tmp/foo.foospec
```

I expect to open "/tmp/foo.foospec" in a buffer and have it fontified
in red.  However the buffer ends up with font-lock-mode: nil even
though global-font-lock-mode is t (under emacs -Q).

I think this used to work until recently. Is this a possible
regression or is the testcase flawed because of some recent changes?
--Best Regards, Madhu