summaryrefslogtreecommitdiffhomepage
path: root/src/vm.c
AgeCommit message (Collapse)Author
2020-10-12Use `mrb_int_value()` instead of `mrb_fixnum_value()`.Yukihiro "Matz" Matsumoto
Where fixnum overflow can happen.
2020-10-12Reorganize `Integer` system.Yukihiro "Matz" Matsumoto
- Integrate `Fixnum` and `Integer` - Remove `Integral` - `int / int -> int` - Replace `mrb_fixnum()` to `mrb_int()` - Replace `mrb_fixnum_value()` to `mrb_int_value()`. - Use `mrb_integer_p()` instead of `mrb_fixnum_p()`
2020-10-12Make division by zero cause `ZeroDivisionError`.Yukihiro "Matz" Matsumoto
As described in ISO 15.2.30.
2020-10-12Rename `MRB_TT_FIXNUM` to `MRB_TT_INTEGER`.Yukihiro "Matz" Matsumoto
We still have `#define MRB_TT_FIXNUM MRB_TT_INTEGER` for compatibility.
2020-10-12Update the PR #4922 according to #3123.Yukihiro "Matz" Matsumoto
close #3123
2020-10-12Update the PR #4992 to enable integer division always.Yukihiro "Matz" Matsumoto
2020-10-12Update vm.cLanza
2020-10-12Rename float configuration option names.Yukihiro "Matz" Matsumoto
- `MRB_WITHOUT_FLOAT` => `MRB_NO_FLOAT` - `MRB_USE_FLOAT` => `MRB_USE_FLOAT32` The former is to use `USE_XXX` naming convention. The latter is to make sure `float` is 32bit float and not floating point number in general.
2020-10-12Skip array embedding if `MRB_NO_BOXING` and `MRB_32BIT`; fix #4382Yukihiro "Matz" Matsumoto
On some platforms, `sizeof(mrb_value) > sizeof(void*)*3`, which makes `MRB_ARY_EMBED_LEN_MAX` zero. And zero sized array cause compile errors.
2020-10-12Rename `OP_JUW` instruction to `OP_JMPUW`.Yukihiro "Matz" Matsumoto
2020-10-12Fix typo `_hander` -> `_handler`.Yukihiro "Matz" Matsumoto
2020-10-12Adjust PR #5060 to the latest `mruby3` branch.Yukihiro "Matz" Matsumoto
2020-10-12Suppress warnings for C++dearblue
Variables in jump destination block separate declaration and assignment.
2020-10-12Restore the variable `pc` after `longjmp()`dearblue
Changes made after `setjmp()` are destroyed and need reassignment. This problem is now caused by the addition of the `OP_JUW` instruction. When actually building on FreeBSD 12.1 with `clang10 -fsanitize=address`, mrbtest "NameError#name [15.2.31.2.1]" is failed. However, qualifying `pc` with `volatile` slows down significantly and increases the object code. Suppress them by qualifying only the variables that restore `pc`.
2020-10-12Replace global jump with catch handler implementationdearblue
When a global jump occurs, look at the catch handler table to determine where to jump. In that case, `pc` already shows the following instruction, but since the table shows `begin_offset ... end_offset`, the comparison is done with `begin_offset < pc && pc <= end_offset`. If there is a corresponding handler, move `pc` to `handler.target_offset` and continue running the VM. When a global jump across `ensure` is made by `return`, `break`, `next`, `redo` and `retry`, the extended `RBreak` object saves and restores the C-level execution position. This extended `RBreak` can have tag information, which makes it a pseudo coroutine (the "tag" mimics CRuby). The implementation of pseudo coroutines by `RBreak` is summarized by `CHECKPOINT_RESTORE ... CHECKPOINT_MAIN ... CHECKPOINT_END` and `throw_tagged_break` / `unwind_ensure` macros. The restart of processing is branched by `RBREAK_TAG_FOREACH(DISPATCH_CHECKPOINTS)`. - Not only `rescue` blocks but also `ensure` blocks are now sandwiched between `OP_EXCEPT` and `OP_RAISEIF`. - Remove the function `ecall()`. It is no longer necessary to re-enter the VM to perform an "ensure block". This will resolves #1888. - Added instruction `OP_JUW` (Jump while UnWind). It jumps unconditionally like `OP_JMP`, but searches the catch handler table and executes the ensure block. Since it searches the catch handler table, it is much heavier than `OP_JMP`.
2020-10-12Removed push/pop instructions for rescue/ensuredearblue
`OP_PUSHERR`, `OP_POPERR`, `OP_EPUSH` and `OP_EPOP` are removed.
2020-10-12Extended `OP_EXCEPT` and `OP_RAISE` (`OP_RAISEIF`) instructionsdearblue
- `OP_EXCEPT` checks if `mrb->exc` is `NULL`, `MRB_TT_EXCEPTION` or `MRB_TT_BREAK`. If `mrb->exc` is `NULL`, it will be replaced with `nil`. - If `OP_RAISE` is `nil`, it does nothing and the immediately following instruction is executed (like `OP_NOP`). Also, in case of `RBreak` object, it moves to the processing for `break`. With this change, the instruction name is changed from `OP_RAISE` to `OP_RAISEIF`.
2020-10-12Initialize `kdict` in `OP_ENTER` in `vm.c`.Yukihiro "Matz" Matsumoto
2020-10-12Add `const` specifier to `cipush()`; ref #5052Yukihiro "Matz" Matsumoto
2020-10-12Fix the bug by the combination with `MRB_64BIT` and `MRB_INT32`.Yukihiro "Matz" Matsumoto
Which is caused by `MRB_NAN_BOXING`.
2020-10-12Remove `OP_EXT[123]` from operands.Yukihiro "Matz" Matsumoto
2020-10-12Clarify the use of `MRB_64BIT` and `MRB_INT64` in `dump.c` and `load.c`.Yukihiro "Matz" Matsumoto
- `MRB_64BIT`: the size of a pointer is 64 bits - `MRB_INT64`: the size of `mrb_int` is 64 bits
2020-10-12Generate C struct from `irep` instead of binary dump.Yukihiro "Matz" Matsumoto
2020-10-12Replace entire `irep->pool`.Yukihiro "Matz" Matsumoto
Changes: - `pool format is completely replaced - supported types: `STR`, `INT32`, `INT64`, `FLOAT` - `FLOAT` may be replaced by binary representation in the future - insert `NUL` after string literals in `mrb` files - `irep->pool` no longer store values in `mrb_value` - instead it stores in `mrb_pool_value` - less allocation - `mrb_irep` can be stored in ROM
2020-10-12Constify `irep` members.Yukihiro "Matz" Matsumoto
- `pool` - `syms` - `reps`
2020-10-12Change the arguments of following implicit conversion functions:Yukihiro "Matz" Matsumoto
- `mrb_convert_type` - `mrb_check_convert_type` Those function no longer take `tname` string representation of desired type, and take method symbols instead of `const char*` names. This is incompatible change. I hope no third-party gems use those functions.
2020-10-12Define a new function `mrb_funcall_id()`.Yukihiro "Matz" Matsumoto
`mrb_funcall_id()` takes `mrb_sym` instead of `char*` for a method name. You can use `MRB_SYM()`/`MRB_QSYM()` to specify the method to call.
2020-10-12Use `MRB_QSYM()` instead of `MRB_OPSYM()`.Yukihiro "Matz" Matsumoto
2020-10-12Use `MRB_OPSYM()` instead of `mrb_intern_lit()`.Yukihiro "Matz" Matsumoto
2020-10-12Add `MRB_SYM()` for inline symbols.Yukihiro "Matz" Matsumoto
2020-08-06Use `memcpy` to copy stack.Yukihiro "Matz" Matsumoto
2020-08-06Use `memset()` to clear stack unless `MRB_NAN_BOXING`.Yukihiro "Matz" Matsumoto
2020-08-06Remove block clear of `callinfo`.Yukihiro "Matz" Matsumoto
2020-07-26Extend the `cipush()` and `cipop()` functionsdearblue
- Returns the updated call info. - Unify the processing around `cipush()`. - `cipop()` restores the stack.
2020-07-21Fix the VM stack handling bug in 'mrb_yield_with_class()`; fix #5042Yukihiro "Matz" Matsumoto
2020-06-25Remove unnecessary stack adjustment in `OP_CALL`.Yukihiro "Matz" Matsumoto
2020-06-25Change flag names in preparation of `REnv` refactoring.Yukihiro "Matz" Matsumoto
2020-05-28Check `c->eidx` before decrement in `ecall()`; close #4977Yukihiro "Matz" Matsumoto
2020-05-24Do not destruct rest arguments for __send__Yusuke Endoh
Formerly, `__send__(*args)` modified `args` with `Array#shift`. This bug affects optcarrot. This changeset avoids the array destruction by using `args = args[1, len-1]`.
2020-05-10Get an irep symbol if it's `OP_GETSV` or `OP_SETSV`dearblue
2020-05-07Add a new instruction `OP_LOADI16`.Yukihiro "Matz" Matsumoto
Which loads 16bit integer to the register. The instruction number should be reorder on massive instruction refactoring. The instruction is added for `mruby/c` which had performance issue with `OP_EXT`. With this instruction, `mruby/c` VM can just raise errors on `OP_EXT` extension instructions.
2020-05-07Remove `mrb_run` from `MRB_API`; #4488Yukihiro "Matz" Matsumoto
`mrb_run` requires to push callinfo stack before calling, which is very hard from outside of `vm.c`. So there should be virtually no correct usage of the function, hence the cause of #4488. We removed it. You can use `mrb_top_run(mrb, proc, self, 0)` instead of `mrb_run(mrb, proc self)`.
2020-04-22Make `mrb_obj_instance_eval` to use `mrb_singleton_class_ptr`; #4973Yukihiro "Matz" Matsumoto
2020-04-21Fix the bug by `#instance_eval` called via a method object; fix #4973Yukihiro "Matz" Matsumoto
The tranpoline code in 6a0b68f8b was wrong; reverted.
2020-04-21Simplified `mrb_obj_instance_eval`; ref #4973Yukihiro "Matz" Matsumoto
2020-03-31Fix `mrb_funcall_with_block()` uses more GC arenadearblue
If `mrb->jmp` is `NULL` and the function `mrb_funcall_with_block()` is called, GC Arena is returned from the function with over-used. - A normal (no global exodus) return will consume two GC Arena's. - In the event of an exception, five GC Arena are consumed. This patch reduces consumption in both cases to one.
2020-01-29Removed junk.Yukihiro "Matz" Matsumoto
My cat stepped on the keyboard at the last moment before the commit.
2020-01-29Remove unused local variable `mid`; ref #4936Yukihiro "Matz" Matsumoto
2020-01-27Fixed backtrace message for top-level blocks; fix #4936Yukihiro "Matz" Matsumoto
In top-level, `mid` is `NULL`. We used to ignore 'mid` update for `NULL`.
2020-01-08Add `MRB_WITHOUT_FLOAT` guard to `<math.h>`.Yukihiro "Matz" Matsumoto