| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Embedding reduce memory consumption, sacrificing precision. It clips least
significant 2 bits from `mrb_float`, so if you need to keep float precision,
define `MRB_USE_FLOAT_FULL_PRECISION`.
`MRB_WORD_BOXING` and `MRB_INT64`:
`mrb_float` (`double`) is embedded in `mrb_value` clipped last 2 bits.
`MRB_WORD_BOXING` and `MRB_INT64` and `MRB_USE_FLOAT_FULL_PRECISION`:
`mrb_float` is allocated in the heaps wrapped by `struct RFloat`.
`MRB_WORD_BOXING` and `MRB_INT32` and `MRB_USE_FLOAT32`:
`mrb_float` (`float`) is embedded in `mrb_value` clipped last 2 bits.
In addition, to reserve bit space in the `mrb_value`, maximum inline
symbol length become 4 (instead of 5) in the configuration.
`MRB_WORD_BOXING` and `MRB_INT32`:
Assume `MRB_USE_FLOAT_FULL_PRECISION` and allocate Float values in heap.
|
|
In additions:
* use bitmap for flags
* BER integer compression for length
|
|
|
|
|
|
- mrb_packed_int_len()
- mrb_packed_int_encode()
- mrb_packed_int_decode()
|
|
|
|
Previously, the `I` specifier only checked if the object was `MRB_TT_ISTRUCT`.
So it was at risk of getting pointers to different C structs if multiple classes were to use the `MRB_TT_ISTRUCT` instance.
Change this behavior and change the C argument corresponding to the `I` specifier to `(void *, struct RClass)`.
This change is not compatible with the previous mruby.
Please note that if the user uses the previous specifications, `SIGSEGV` may occur or the machine stack may be destroyed.
resolve #5527
|
|
Fix #5528
This reverts commit 59201b59046b9e73c309508350cd3c0fafd20e4d; #5497
|
|
|
|
Recent peephole optimization made `ADDI/SUBI` destinations possibly
local variables.
|
|
|
|
|
|
`-1 * (1<<63)` causes overflow, but `-1<<63` is a valid value.
|
|
|
|
This reverts commit 3738d62a86a54524d5d7738ddbafe4700ca6a889.
The change changed the behavior with floating point numbers.
|
|
|
|
Use `mrb_fixnum_value()` only when you are absolutely sure that the
value is within `Fixnum` range, i.e. 31 bits signed integer at least.
|
|
The `Fixnum` class is no longer provided by `mruby`.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- stdlib.h
- stddef.h
- stdint.h
- stdarg.h
- limits.h
- float.h
|
|
|
|
- OP_GETGV
- OP_SETGV
- OP_GETSV
- OP_SETSV
- OP_GETIV
- OP_SETIV
- OP_GETCV
- OP_SETCV
- OP_GETCONST
- OP_SETCONST
- OP_GETMCNST
- OP_SETMCNST
- OP_GETUPVAR
- OP_SETUPVAR
|
|
So that the function behave consistently with `mrb_load_proc()`.
|
|
|
|
- String#__lines
- Array#__ary_eq
- Array#__ary_cmp
- Hash#__delete
- Kernel#__case_eqq
- Integer#__coerce_step_counter
|
|
|
|
A change in `load.c` is left uncommitted.
|
|
|
|
|
|
|
|
|
|
That means it's a method not to be included in the backtrace, for
example `raise`.
|
|
It uses BER number compression of delta of instruction positions and line
numbers. BER compression is a variable length number representation.
* `mrb_debug_line_ary`: array of line numbers represented in `uint16_t`.
`[lineno, lineno, ...]`
* `mrb_debug_line_flat_map`: array of `mrb_irep_debug_info_line`, which
is `struct {uint32_t pos; uint16_t lineno}`, for each line.
* `mrb_debug_line_packed_map` [new]: sequence of BER compressed 2
numbers, `pos_delta, lineno_delta`. Deltas are differences from
previous values (starting `0`). `line_entry_counts` represents total
length of a packed map string for this type.
|
|
The code generator no longer need to emit `OP_LOADSYM` after `OP_DEF`.
`doc/opcode.md` is also updated.
|
|
|
|
|
|
This reverts commit fd10c7231906ca48cb35892d2a86460004b62249.
I thought it was OK to restrict index value within 1 byte, but in some
cases index value could be 16 bits (2 bytes). I had several ideas to
address the issue, but reverting `fd10c72` is the easiest way. The
biggest reason is `mruby/c` still supports `OP_EXT[123]`, so that they
don't need any additional work.
|
|
It does not need to hold an anonymous proc for constant search.
Also, this change can be expected to cause an anonymous proc to be GC'd.
This is useful for metaprogramming that makes heavy use of the `class`/`module`/`def` syntax in the `class_eval`/`eval` method.
Example:
- code
```ruby
p ObjectSpace.count_objects
String.class_eval do
def a
end
end
p ObjectSpace.count_objects
String.class_eval do
eval <<~CODE
def b
end
CODE
end
p ObjectSpace.count_objects
```
- result of building mruby-head (d63c0df6b) with `build_config/default.rb`
```
{:TOTAL=>1024, :FREE=>262, :T_PROC=>495, :T_ENV=>61, ...}
{:TOTAL=>1024, :FREE=>259, :T_PROC=>497, :T_ENV=>62, ...}
{:TOTAL=>1024, :FREE=>255, :T_PROC=>500, :T_ENV=>63, ...}
```
- result of building mruby with this patch and `build_config/default.rb`
```
{:TOTAL=>1024, :FREE=>264, :T_PROC=>494, :T_ENV=>60, ...}
{:TOTAL=>1024, :FREE=>262, :T_PROC=>495, :T_ENV=>61, ...}
{:TOTAL=>1024, :FREE=>261, :T_PROC=>496, :T_ENV=>61, ...}
```
|
|
Previously the following code did not produce the expected results:
```ruby
bx = binding
block = bx.eval("a = 1; proc { a }")
bx.eval("a = 2")
p block.call # Expect 2 but return 1 due to a bug
```
The previous implementation of `Binding#eval` evaluated the code and then merged the top layer variables.
This patch will parse and expand the variable space before making a call to `eval`.
This means that the call to `Binding#eval` will do the parsing twice.
In addition, the following changes will be made:
- Make `mrb_parser_foreach_top_variable()`, `mrb_binding_extract_proc()` and `mrb_binding_extract_env()` functions private global functions.
- Remove the `posthook` argument from `mrb_exec_irep()`.
The `posthook` argument was introduced to implement the `binding` method.
This patch is unnecessary because it uses a different implementation method.
ref #5362
fixed #5491
|
|
|
