Work on https://github.com/neovim/neovim/issues/459

"cfuncs" was only ever used to wrap luarefs. As vim8script is
finished and will not be developed further, support for "cfuncs"
for other usecases are not planned. This abstraction was immediately
broken anyway in order to get luarefs out of userfuncs again.

Even if a new kind of userfunc needs to be invented in the future,
likely just extending the FC_... flag union directy, instead of
invoking unnecessary heap object and c function pointer indirection,
will be a more straightforward design pattern.

Now nvim_parse_cmd and nvim_create_user_command use a "tab" value which
is the same as the number passed before :tab modifier instead of the
number plus 1, and "tab" value is -1 if :tab modifier is not used.

Co-authored-by: Famiu Haque <famiuhaque@protonmail.com>

`!did_throw` doesn't exactly imply `!current_exception`, as `did_throw = false`
is sometimes used to defer exception handling for later (without forgetting the
exception). E.g: uncaught exception handling in `do_cmdline()` may be deferred
to a different call (e.g: when `try_level > 0`).

In #7881, `current_exception = NULL` in `do_cmdline()` is used as an analogue of
`did_throw = false`, but also causes the pending exception to be lost, which
also leaks as `discard_exception()` wasn't used.

It may be possible to fix this by saving/restoring `current_exception`, but
handling all of `did_throw`'s edge cases seems messier. Maybe not worth
diverging over.

This fix also uncovers a `man_spec.lua` bug on Windows: exceptions are thrown
due to Windows missing `man`, but they're lost; skip these tests if `man` isn't
executable.

Work on https://github.com/neovim/neovim/issues/459

perf(api): allow to use an arena for return values

This is both simpler in client code and more effective (always reuse
block hottest in cache)