We have added internal convenience method `__to_str` which
does string type check.
The issue #3854 was fixed but fundamental flaw of lack of stack
depth check along with fibers still remains. Use `MRB_GC_FIXED_ARENA`
for workaround.
We assume meta-programming is less used in embedded environments.
We have moved following methods:
* Kernel module
global_variables, local_variables, singleton_class,
instance_variables, instance_variables_defined?, instance_variable_get,
instance_variable_set, methods, private_methods, public_methods,
protected_methods, singleton_methods, define_singleton_methods
* Module class
class_variables, class_variables_defined?, class_variable_get,
class_variable_set, remove_class_variable, included_modules,
instance_methods, remove_method, method_removed, constants
* Module class methods
constants, nesting
Note:
Following meta-programming methods are kept in the core:
* Module class
alias_method, undef_method, ancestors, const_defined?, const_get,
const_set, remove_const, method_defined?, define_method
* Toplevel object
define_method
`mruby-metaprog` gem is linked by default (specified in default.gembox).
When it is removed, it will save 40KB (stripped:8KB) on x86-64
environment last time I measured.
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.
Make `block_given?` to search for the top of the scope first.
The top of the scope means either:
* the top method body
* the enclosing class body
* the top-level
The special case is the method defined by `define_method` with a
block as in #3841. In cases like this, the method body (given by
a block) is not considered as the top of the scope. You need to use
`&block` in the block parameter if you want to know if a block is
given to the method.
This commit also changes the behavior of `MRB_PROC_SCOPE` flag.
Now it is only set if the `proc` is either a class body or a method
body defined in Ruby. It is no longer set for a block that given to
`define_method`.
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`.
To enable method cache, define `MRB_METHOD_CACHE` or
`MRB_METHOD_CACHE_SIZE`. The cache size must be power of 2.
The default cache size is 128.
The measurement:
I measured simple benchmarks found in benchmark/ directory. With
method cache enabled, we gained 6-8% performance improvement, with
97-99% cache hit rate.
The change removes several internal instance variables used by
class name management. The variables `__classid__` and `__classpath__`
are no longer available. `__outer__` is used only for unnamed outer
classes/modules (and will be removed after they are named).
[Important note]
Along with this change we removed several public functions.
- mrb_class_outer_module()
- mrb_class_sym()
We believe no one have used those functions, but if you do, please
ask us for the workaround.
It reduces the memory consumption and sometimes improve the
performance as well. For example, the consumed memory size
of `bench/bm_ao_render.rb` is reduced from 1.2GB to 1GB, and
its total execution time become 18.795 sec from 22.229 sec.