bug#77924: 31.0.50; [Feature branch] Change marker implementation

[Stefan Monnier]

>>> So the benchmark tests scroll time, including jit/font-lock time.
>>> It uses its own copy of a major mode, so that you can compare "scroll +
>>> font-lock" performance between different Emacs releases without being
>>> affected by improvements/regressions in CC-mode itself.
>>
>> So maybe it would make sense to run both that benchmark as-is plus one
>> without smie but with tamil.txt, I guess. Or in other words, now only
>> the as-is benchmark, because the tamil.txt results I've already posted.
>
> Before including tamil.txt, I would like to make clear that the file
> contents is volume 1 of a (popular) novel which was released in the 50s.
> I got the text itself from Project Madurai which states that [1]:

It'd be good to have a benchmark for multibyte "human text", indeed.
I'll let others figure out the copyright issues (hopefully none) and
then write the corresponding benchmark.

In the mean time I took that `xmenu.c` code used for the scroll and SMIE
benchmarks and made it multibyte by replacing all the ASCII letter with
arbitrary non-ASCII characters.
I tried to do it in such a way that the keywords are preserved so it's
still a "valid" C file that is still highlighted and indented in the
same way.

The resulting `scroll-nonascii` and `smie-nonascii` benchmarks are now
in `elisp-benchmark` in `elpa.git`.


        Stefan