`MRB_TRY()` does not work when compiled by C compiler with `cxx_exception`,
due to the mixture of `setjmp()` used by `mirb.c` and `throw` used by
the core.  The original intension of e2e6554b is to protect code from
signal interruption, but the signal interruption is not well-defined in
mruby anyway.

Also fixed the size calculation of `irep` dump, that could cause memory
corruption.

To allow C++ compilation. Fix suggested by @dearblue.

Redirect `mrb_str_to_str` to `mrb_obj_as_string` via C macro.
Inspired by #5082

Those types are `uint16_t` in definition. Also we no longer need padding
for `iseq`.

The fix was proposed by @dearblue

The fix was proposed by @dearblue

The fix was proposed by @dearblue

Need to support `int / int -> int` update.

Target `host-debug` to use `host` internal target with debugging
configuration.

The PR was from @cubicdaiya.

    e.g. symbols like "foo[]=" make invalid C codes

The original PR was skipping Ruby comments as well, but caused some
issues in test suites.

- `mrb_check_intern()` to return `mrb_value`
- `mrb_intern_check()` to return `mrb_sym` [NEW]

Other new functions:

- `mrb_intern_check_cstr()`
- `mrb_intern_check_str()`

Rename new functions:

- `mrb_convert_type(mrb,val,type,tname,method)`
   => `mrb_type_convert(mrb,val,type,tname,method)`
- `mrb_check_convert_type(mrb,val,type,tname,method)`
   => `mrb_type_convert_check(mrb,val,type,tname,method)`

Old names are defined by macros (support `tname` drop and
`char*` => `mrb_sym` conversion).

On platforms where `sizeof(long)` is 4, casting `(long)` can lose data
or sign information.

- Implement `Rational()` in `C`.
- Use `float` to `rational` conversion function taken from:
  https://rosettacode.org/wiki/Convert_decimal_number_to_rational#C

The mixture causes warnings on 64 bit Windows (VC).

You can now use `NEXT` within `switch` statement like 7c087eb.

On non-`gcc` compatible environment, `NEXT` is translated to `break`.

Linux `atan2(3)` man page says:

```
If  y is positive infinity (negative infinity) and x is positive infinity,
+pi/4 (-pi/4) is re‐ turned.
```

But on Microsoft VC/MinGW, `atan2()` returns `NaN` if either of
arguments is infinite. So we skip those tests on the platforms.