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

Where fixnum overflow can happen.

Since it's not supported on VC without `/std:c++latest`. That means it
doesn't work for `cxx_api` build on Windows VC.

- `pool`
- `syms`
- `reps`

`mrb_get_arg1()` raises `ArgumentError` if the method does not receive one
argument.

And replaces all `mrb_get_args(mrb, "o", &arg)` by the new function.

commit 2d7d545c4c4bfce7fdcbcbe9baaeb437915742f0
Merge: 625a1249 b178914b
Author: Yukihiro "Matz" Matsumoto <matz@ruby.or.jp>
Date:   Fri Jun 5 14:35:13 2020 +0900

    Merge branch 'fix-mrb_open-with-nomem' of https://github.com/dearblue/mruby into dearblue-fix-mrb_open-with-nomem

commit b178914b111dda79a8f36ec4eb3e9d37b76f982e
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sat Jan 19 22:22:44 2019 +0900

    Fix invalid pointer free inside other heap's block

     1. `e = mrb_obj_alloc(...)`
     2. `e->stack = mrb->c->stack` (`mrb->c->stack` is anywhere in the range `stbase...stend`)
     3. And raised exception by `mrb_malloc()`!
     4. `mrb_free(e->stack)` by GC part (wrong free)

commit 52e3d5d8585daf86af3ed12db5ab0efefbc9b956
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sat Jan 19 21:55:36 2019 +0900

    Fix memory leak for temporary symbols when out of memory

commit 4c5499b88e47cc6012ad7d7379cb6bc74c6a0b60
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 11:42:07 2019 +0900

    Fix uninitialized pointer dereference for debug section

commit 8e993167dec62a9709d6faacd517729ddcedf4f9
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 11:41:09 2019 +0900

    Fix memory leak for temporary filenames when out of memory

commit 8b422577e6eae68a28121b88421d937e8707b487
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 10:57:51 2019 +0900

    Fix memory leak for irep when out of memory

commit 6b35ebf49a0aa3edb6bbda770ed58681e9c2e6af
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 10:55:50 2019 +0900

    Fix uninitialized pointer dereference when do not finished initializing irep

commit 2531f2631e67e0462749618e2344c733a29238f0
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 10:48:15 2019 +0900

    Fix NULL pointer dereference when do not finished initializing irep

commit e2d6896ebad13694800af49c2625e106b8440ddf
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sat Jan 19 12:54:19 2019 +0900

    Fix memory leak for irep when out of memory by `mrb_proc_new()`

commit b6214ff8a0a1c73bc9554e39053878ac50bb683f
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sat Jan 19 12:53:07 2019 +0900

    Fix memory leak for `khash_t` in `kh_init_size()` when out of memory by `kh_alloc()`

commit 19162dd6c11f0093d0011e7cab83b8f9e84c2c07
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 02:15:07 2019 +0900

    Fix memory leak for symbol string when out of memory in `kh_put()`

commit 15e67297ff54bc14ef359d6d1e745d760a4a255a
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 02:12:24 2019 +0900

    Fix keep wrong symbol index when out of memory

commit 3f8e2b375244f5441e8d62efa13c6e6a9afecb14
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 02:08:13 2019 +0900

    Fix keep wrong symbol capacity when out of memory

commit a3cfe755ab3e758046c3f4e30938ac8d567ed046
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sat Jan 19 10:11:37 2019 +0900

    Fix NULL pointer dereference `mrb->c` by `mark_context()`

commit d9c7b6be6eb54630b64eea5c35be241e551676e5
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 15:25:09 2019 +0900

    Fix protect exception for print error message

commit 100642750e4d549f2e8050f8d6cabdf8825d4495
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 11:59:02 2019 +0900

    Protect exception for mruby core initialization

commit 7a0418304ec70764fa215bef3599f5f735222075
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Fri Jan 18 20:38:27 2019 +0900

    Fix memory leak for string object when out of memory

    The `mrb_str_pool()` function has a path to call `malloc()` twice.

    If occurs `NoMemoryError` exception in second `malloc()`,
    first `malloc()` pointer is not freed.

commit fef1c152ce4e52b9e4a34dc23aca5b02907ac639
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sat Jan 19 13:05:09 2019 +0900

    Fix stack overflow when out of memory

    As a result of this change, no backtrace information is set
    for NoMemoryError (`mrb->nomem_err`).

    Detailes:

    When generating a backtrace, called `mrb_intern_lit()`,
    `mrb_str_new_cstr()` and `mrb_obj_iv_set()` function with
    `exc_debug_info()` function in `src/error.c`.

    If a `NoMemoryError` exception occurs at this time,
    the `exc_debug_info()` function will be called again,
    and in the same way `NoMemoryError` exception raised will result
    in an infinite loop to occurs stack overflow (and SIGSEGV).

commit da7d7f881bbbad9988a3a2b7bad8f2b72ff06bc6
Author: dearblue <dearblue@users.noreply.github.com>
Date:   Sun Jan 20 12:00:38 2019 +0900

    Fix NULL pointer dereference `mrb->nomem_err` when not initialized

    Add internal functions (not `static`):

      * `mrb_raise_nomemory()`
      * `mrb_core_init_abort()`

For efficiency with `MRB_WORD_BOXING` (implement type predicate macros for
all `enum mrb_vtype`).

We needed to preserve the original method name somewhere. We kept it in
the `env` structure pointed from aliased methods. #1457 and #1531 tried
to address this issue. But this patch is more memory efficient.

Limitation: this fix does not support `super` from methods defined by
`define_method`. This limitation may be addressed in the future, but
it's low priority.

Some error messages will be changed.

The binary sizes (gems are only `mruby-bin-mruby`) are reduced slightly in
my environment than before the introduction of new specifiers/modifiers
(5116789a) with this change.

  ------------+-------------------+-------------------+--------
   BINARY     | BEFORE (5116789a) |   AFTER (This PR) |  RATIO
  ------------+-------------------+-------------------+--------
   mruby      |      593416 bytes |      593208 bytes | -0.04%
   libmruby.a |      769048 bytes |      767264 bytes | -0.23%
  ------------+-------------------+-------------------+--------

BTW, I accidentally changed `tasks/toolchains/visualcpp.rake` at #4613,
so I put it back.

Since it is already set in `mrb_proc_new()`.

Now the method tables (in classes/modules and caches) keeps C function
pointers without wrapping in `struct RProc` objects. For the sake of
portability, `mrb_method_t` is represented by the struct and union, but
if the most significant bit of the pointer is not used by the platform,
`mrb_method_t` should be packed in `uintptr_t` to reduce memory usage.

`MRB_METHOD_TABLE_INLINE` is turned on by default for linux.

Instead of `irep` links, we added a `upper` link to `struct RProc`.
To make a space for the `upper` link, we moved `target_class` reference.
If a `Proc` does not have `env`, `target_class` is saved in an `union`
shared with `env` (if a `Proc` has env, you can tell it by `MRB_PROC_ENV_P()).
Otherwise `target_class` is referenced from `env->c`. We removed links
in `env` as well.

This change removes 2 members from `mrb_irep` struct, thus saving 2
words per method/proc/block. This also fixes potential memory leaks
due to the circular references caused by a link from `mrb_irep`.

fix: src\proc.c(96): warning C4244: 'function': conversion from 'mrb_int' to 'int', possible loss of data

It may be better to raise exceptions, but CRuby doesn't.

One for the receiver, the other for the block.

Fix #3356

Reported by @bouk

ref #3215

If a function (such as mrb_read_irep_file()) is declared without MRB_API
in header file (such as include/mruby/dump.h), implementation of the
function in source file (such as src/load.c) should also defined without
MRB_API.

If MRB_API is mismatch, Visual C++ reports link error with C2375 error
code: https://msdn.microsoft.com/en-us/library/5k6kw95a.aspx

mrb_proc_new_cfunc_with_env() allocates RProc with RProc::env as NULL
then allocates REnv and sets it to RProc::env of the allocated RProc. If
incremental GC is ran before "allocates REnv and sets it to RProc::env
of the allocated RProc", the allocated RProc's GC status is
"marked" (Black) and the allocated REnv's GC status is
"unmarked" (White). The next incremental GC sweeps the allocated REnv
without re-marking the allocated RProc. Because the RProc is Black and
the REnv is White.

We need to implement write barrier for the case.

We can force to cause the above situation by the following patch:

    diff --git a/src/proc.c b/src/proc.c
    index f98998f..4f4e25c 100644
    --- a/src/proc.c
    +++ b/src/proc.c
    @@ -92,6 +92,7 @@ mrb_proc_new_cfunc_with_env(mrb_state *mrb, mrb_func_t func, mrb_int argc, const
       struct REnv *e;
       int i;

    +  mrb_incremental_gc(mrb);
       p->env = e = env_new(mrb, argc);
       MRB_ENV_UNSHARE_STACK(e);
       e->stack = (mrb_value*)mrb_malloc(mrb, sizeof(mrb_value) * argc);

With this patch, "rake test" causes segmentation fault.