| Age | Commit message (Expand) | Author |
|---|
| 34 hours | variable.c: avoid out of bound write in `generic_field_set` | Jean Boussier |
| 4 days | Fix missing write barrier on class fields | John Hawthorn |
| 5 days | Add missing write barriers in `rb_imemo_fields_clone`. | Jean Boussier |
| 5 days | Refactor generic fields to use `T_IMEMO/fields` objects. | Jean Boussier |
| 5 days | Refactor `rb_imemo_fields_new` to not assume T_CLASS | Jean Boussier |
| 5 days | Rename `imemo_class_fields` -> `imemo_fields` | Jean Boussier |
| 6 days | Adjust indent [ci] | Nobuyoshi Nakada |
| 6 days | Free after insert in generic_ivar_set_shape_fields | John Hawthorn |
| 9 days | Remove fields_tbl in gen_fields_lookup_ensure_size | John Hawthorn |
| 9 days | Fix generic_ivar_set_shape_field for table rebuild | John Hawthorn |
| 9 days | Get rid of FL_EXIVAR | Jean Boussier |
| 9 days | Use the `shape_id` rather than `FL_EXIVAR` | Jean Boussier |
| 9 days | Enforce consistency between shape_id and FL_EXIVAR | Jean Boussier |
| 10 days | Make setting and accessing class ivars lock-free | Jean Boussier |
| 10 days | Fix class instance variable inside namespaces | Jean Boussier |
| 11 days | Turn `rb_classext_t.fields` into a T_IMEMO/class_fields | Jean Boussier |
| 11 days | Refactor the last references to `rb_shape_t` | Jean Boussier |
| 13 days | Get rid of `gen_fields_tbl.fields_count` | Jean Boussier |
| 2025-06-07 | Get rid of SHAPE_T_OBJECT | Jean Boussier |
| 2025-06-07 | shape.c: ensure heap_index is consistent for complex shapes | Jean Boussier |
| 2025-06-05 | Refactor raw accesses to rb_shape_t.capacity | Jean Boussier |
| 2025-06-05 | Suppress warnings by gcc-13 with `-Og` | Nobuyoshi Nakada |
| 2025-06-05 | Get rid of `rb_shape_t.flags` | Jean Boussier |
| 2025-06-04 | Get rid of TOO_COMPLEX shape type | Jean Boussier |
| 2025-06-04 | Add missing lock in `rb_ivar_defined` | Jean Boussier |
| 2025-06-04 | Get rid of frozen shapes. | Jean Boussier |
| 2025-05-27 | Rename `rb_shape_id_canonical_p` -> `rb_shape_canonical_p` | Jean Boussier |
| 2025-05-27 | Rename `rb_shape_set_shape_id` in `rb_obj_set_shape_id` | Jean Boussier |
| 2025-05-27 | Refactor `rb_shape_too_complex_p` to take a `shape_id_t`. | Jean Boussier |
| 2025-05-27 | Refactor `rb_obj_shape` out. | Jean Boussier |
| 2025-05-27 | Refactor `rb_shape_get_iv_index` to take a `shape_id_t` | Jean Boussier |
| 2025-05-27 | Get rid of `rb_shape_canonical_p` | Jean Boussier |
| 2025-05-27 | Refactor `rb_shape_rebuild_shape` to stop exposing `rb_shape_t` | Jean Boussier |
| 2025-05-26 | Add shape_id to RBasic under 32 bit | John Hawthorn |
| 2025-05-25 | Use RB_VM_LOCKING | Nobuyoshi Nakada |
| 2025-05-23 | Add missing lock for `Module#remove_instance_variable` | Jean Boussier |
| 2025-05-23 | Refactor `rb_shape_transition_remove_ivar` | Jean Boussier |
| 2025-05-23 | lock vm around `rb_free_generic_ivar` | Luke Gruber |
| 2025-05-21 | Use shape_id for determining "too complex" | Aaron Patterson |
| 2025-05-16 | rb_copy_generic_ivar: reset shape_id when no ivar are present | Jean Boussier |
| 2025-05-13 | variable.c: Refactor rb_obj_field_* to take shape_id_t | Jean Boussier |
| 2025-05-11 | Add missing gc_mark in `autoload_const_mark` | Jean Boussier |
| 2025-05-11 | Unexpectedly miss-fixed macro during rebase | Satoshi Tagomori |
| 2025-05-11 | Follow the code style about else | Satoshi Tagomori |
| 2025-05-11 | namespace on read | Satoshi Tagomori |
| 2025-05-10 | Refactor `rb_obj_field_get` to handle complex shapes | Jean Boussier |
| 2025-05-10 | Rename `rb_field_get` -> `rb_obj_field_get` | Jean Boussier |
| 2025-05-09 | Rename `RB_OBJ_SHAPE` -> `rb_obj_shape` | Jean Boussier |
| 2025-05-09 | Get rid of `rb_shape_get_parent`. | Jean Boussier |
| 2025-05-09 | Rename `rb_shape_get_shape_id` -> `RB_OBJ_SHAPE_ID` | Jean Boussier |
