The restriction was introduced in commit b563bcb7ff to resolve https://github.com/mruby/mruby/issues/3462.
Subsequently, the `RBreak` object, introduced by mruby 1.3.0, allowed crossing the C boundary.
```ruby
def cross; Class.new { proc { return }.call }; end; cross
# => unexpected return (LocalJumpError) # without this patch
# => nothing raised # with this patch
```
If `Class#allocate` is prohibited, subclasses should also be implicitly prohibited.
```ruby
p Class.new(Struct).allocate.class
# => #<Class:0x82362ac00> by #6122
# => allocator undefined for #<Class:0x000000083a983220> (TypeError) by Ruby 3.2
```
Added `MRB_DEFINE_ALLOCATOR()` to allow subclasses to use `Class#allocate`.
Supplement to #6122.
The method introduced by #5979 causes a fault by swapping classes.
```console
% bin/mruby -e 'Method = Proc; p Object.method(:inspect)'
zsh: segmentation fault (core dumped) bin/mruby -e 'Method = Proc; p Object.method(:inspect)'
```
After applying this patch, a `TypeError` exception will be raised.
```console
% bin/mruby -e 'Method = Proc; p Object.method(:inspect)'
trace (most recent call last):
[1] -e:1
-e:1:in method: allocation failure of Proc (TypeError)
```
However, if the `mrb_vtype` is the same object, the same care must still be taken as before.
```console
% bin/mruby -e 'Method = Binding; p method(:puts).eval("12345")'
trace (most recent call last):
[1] -e:1
-e:1:in eval: wrong argument type nil (expected Proc) (TypeError)
```
Now all functions with `io_buf` takes `mrb_io_buf` as an argument.
Renamed functions (old names):
- io_init_buf (io_buf_init)
- io_fill_buf (io_buf_fill)
- io_fill_buf_comp (io_buf_fill_comp) for UTF-8 encoding
If `outbuf` is `nil` we allocate a buffer string, if `outbuf` is a
string, we resize it to zero length. In the function `io_read`, this
condition is done twice, so we refactor out to `io_reset_outbuf()`.
I hit the following two problems.
- `io.read(0, buf)` always returned a new empty string object.
- `io.read(num, buf)` was appending data to the given `buf`.
This also meant that `buf` was never empty if EOF was reached.
We count `Proc` upper links then check for upper bounds (default:20),
but now we raise an exception as soon as the count exceeds the limit so
that we can avoid traverse unnecessary upper links.
Since Fiber#to_s did not include location information before #6105,
when location information cannot be retrieved for terminated fibers,
instead of meaningless information such as "(unknown):0", we omit
positional information altogether for terminated fibers; ref #6111
When fiber is terminated, the pointer indicated by `f->cxt->cibase->proc` is not protected from GC.
I should have done this patch way as of commit 9af6264d6b (#6105).
Yukihiro "Matz" Matsumoto