Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | YJIT: Compile exception handlers (#8171) | Takashi Kokubun | 2023-08-08 | 1 | -2/+2 |
| | | | Co-authored-by: Maxime Chevalier-Boisvert <maximechevalierb@gmail.com> | ||||
* | Remove __bp__ and speed-up bmethod calls (#8060) | Alan Wu | 2023-07-17 | 1 | -1/+0 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove rb_control_frame_t::__bp__ and optimize bmethod calls This commit removes the __bp__ field from rb_control_frame_t. It was introduced to help MJIT, but since MJIT was replaced by RJIT, we can use vm_base_ptr() to compute it from the SP of the previous control frame instead. Removing the field avoids needing to set it up when pushing new frames. Simply removing __bp__ would cause crashes since RJIT and YJIT used a slightly different stack layout for bmethod calls than the interpreter. At the moment of the call, the two layouts looked as follows: ┌────────────┐ ┌────────────┐ │ frame_base │ │ frame_base │ ├────────────┤ ├────────────┤ │ ... │ │ ... │ ├────────────┤ ├────────────┤ │ args │ │ args │ ├────────────┤ └────────────┘<─prev_frame_sp │ receiver │ prev_frame_sp─>└────────────┘ RJIT & YJIT interpreter Essentially, vm_base_ptr() needs to compute the address to frame_base given prev_frame_sp in the diagrams. The presence of the receiver created an off-by-one situation. Make the interpreter use the layout the JITs use for iseq-to-iseq bmethod calls. Doing so removes unnecessary argument shifting and vm_exec_core() re-entry from the interpreter, yielding a speed improvement visible through `benchmark/vm_defined_method.yml`: patched: 7578743.1 i/s master: 4796596.3 i/s - 1.58x slower C-to-iseq bmethod calls now store one more VALUE than before, but that should have negligible impact on overall performance. Note that re-entering vm_exec_core() used to be necessary for firing TracePoint events, but that's no longer the case since 9121e57a5f50bc91bae48b3b91edb283bf96cb6b. Closes ruby/ruby#6428 | ||||
* | Expose rb_hash_resurrect | Aaron Patterson | 2023-06-23 | 1 | -0/+4 |
| | | | | This is for implementing the `duphash` instruction | ||||
* | Unify length field for embedded and heap strings (#7908) | Peter Zhu | 2023-06-06 | 1 | -2/+1 |
| | | | | | | | | * Unify length field for embedded and heap strings The length field is of the same type and position in RString for both embedded and heap allocated strings, so we can unify it. * Remove RSTRING_EMBED_LEN | ||||
* | Update RJIT to support newarray_send | Aaron Patterson | 2023-04-18 | 1 | -0/+8 |
| | | | | This also adds max / hash support | ||||
* | Move `catch_except_p` to `compile_data` | eileencodes | 2023-04-11 | 1 | -5/+4 |
| | | | | | | | | | | | | | | The `catch_except_p` flag is used for communicating between parent and child iseq's that a throw instruction was emitted. So for example if a child iseq has a throw in it and the parent wants to catch the throw, we use this flag to communicate to the parent iseq that a throw instruction was emitted. This flag is only useful at compile time, it only impacts the compilation process so it seems to be fine to move it from the iseq body to the compile_data struct. Co-authored-by: Aaron Patterson <tenderlove@ruby-lang.org> | ||||
* | Expose rb_sym_to_proc via RJIT | Aaron Patterson | 2023-04-07 | 1 | -0/+4 |
| | | | | This is needed for getblockparamproxy | ||||
* | [Feature #19579] Remove !USE_RVARGC code (#7655) | Peter Zhu | 2023-04-04 | 1 | -0/+5 |
| | | | | | | | | | | | Remove !USE_RVARGC code [Feature #19579] The Variable Width Allocation feature was turned on by default in Ruby 3.2. Since then, we haven't received bug reports or backports to the non-Variable Width Allocation code paths, so we assume that nobody is using it. We also don't plan on maintaining the non-Variable Width Allocation code, so we are going to remove it. | ||||
* | RJIT: Add --rjit-verify-ctx option | Takashi Kokubun | 2023-04-04 | 1 | -0/+1 |
| | |||||
* | RJIT: Store type information in Context | Takashi Kokubun | 2023-04-02 | 1 | -0/+8 |
| | |||||
* | RJIT: Support entry with different PCs | Takashi Kokubun | 2023-04-02 | 1 | -8/+8 |
| | |||||
* | RJIT: Support has_opt ISEQs | Takashi Kokubun | 2023-04-02 | 1 | -0/+2 |
| | |||||
* | RJIT: Simplify cfunc implementation | Takashi Kokubun | 2023-04-02 | 1 | -0/+14 |
| | |||||
* | RJIT: Simplify invokesuper implementation | Takashi Kokubun | 2023-04-02 | 1 | -0/+2 |
| | |||||
* | RJIT: Group blockarg exit reasons | Takashi Kokubun | 2023-04-02 | 1 | -4/+1 |
| | |||||
* | RJIT: Support splat args | Takashi Kokubun | 2023-04-02 | 1 | -1/+2 |
| | |||||
* | RJIT: Update exit reasons | Takashi Kokubun | 2023-04-02 | 1 | -0/+4 |
| | |||||
* | Remove an unneeded function copy | Takashi Kokubun | 2023-04-01 | 1 | -4/+4 |
| | |||||
* | RJIT: Support rest args | Takashi Kokubun | 2023-04-01 | 1 | -0/+12 |
| | |||||
* | RJIT: Fix has_rest exit conditions | Takashi Kokubun | 2023-04-01 | 1 | -1/+1 |
| | |||||
* | RJIT: Remove unused counters | Takashi Kokubun | 2023-04-01 | 1 | -12/+3 |
| | |||||
* | RJIT: Start moving away from VM-like ISEQ handling | Takashi Kokubun | 2023-04-01 | 1 | -4/+29 |
| | |||||
* | RJIT: Implement leaf builtin call | Takashi Kokubun | 2023-03-26 | 1 | -0/+2 |
| | |||||
* | RJIT: Implement attr_writer | Takashi Kokubun | 2023-03-26 | 1 | -3/+9 |
| | |||||
* | RJIT: Put a guard for splat w/ var-arg cfunc | Takashi Kokubun | 2023-03-25 | 1 | -0/+1 |
| | |||||
* | RJIT: Support optional params on splat | Takashi Kokubun | 2023-03-25 | 1 | -1/+0 |
| | |||||
* | RJIT: Remove send_iseq_complex_splat exit | Takashi Kokubun | 2023-03-25 | 1 | -1/+0 |
| | |||||
* | RJIT: Initial support of splat | Takashi Kokubun | 2023-03-25 | 1 | -0/+7 |
| | |||||
* | `vm_call_single_noarg_inline_builtin` | Koichi Sasada | 2023-03-23 | 1 | -0/+1 |
| | | | | | | | | If the iseq only contains `opt_invokebuiltin_delegate_leave` insn and the builtin-function (bf) is inline-able, the caller doesn't need to build a method frame. `vm_call_single_noarg_inline_builtin` is fast path for such cases. | ||||
* | RJIT: Update binding | Takashi Kokubun | 2023-03-21 | 1 | -1/+2 |
| | |||||
* | RJIT: Split has_rest_or_post exit reasons | Takashi Kokubun | 2023-03-21 | 1 | -2/+2 |
| | |||||
* | RJIT: Fix invokesuper | Takashi Kokubun | 2023-03-21 | 1 | -0/+4 |
| | |||||
* | RJIT: Break up RJIT send_iseq_complex exit reasons | Takashi Kokubun | 2023-03-19 | 1 | -1/+14 |
| | |||||
* | RJIT: Implement ifunc invokeblock | Takashi Kokubun | 2023-03-19 | 1 | -1/+6 |
| | |||||
* | RJIT: Fix ISeq invokeblock | Takashi Kokubun | 2023-03-19 | 1 | -0/+2 |
| | |||||
* | RJIT: Implement invokeblock with ISeq | Takashi Kokubun | 2023-03-19 | 1 | -1/+2 |
| | |||||
* | RJIT: Optimize Kernel#respond_to? | Takashi Kokubun | 2023-03-19 | 1 | -0/+24 |
| | |||||
* | RJIT: Optimize String#+@ | Takashi Kokubun | 2023-03-19 | 1 | -0/+5 |
| | |||||
* | RJIT: Optimize String#<< | Takashi Kokubun | 2023-03-19 | 1 | -0/+9 |
| | |||||
* | RJIT: Drop duplicated obj_is_kind_of | Takashi Kokubun | 2023-03-19 | 1 | -4/+0 |
| | |||||
* | RJIT: Workaround USE_RVARGC=0 CI | Takashi Kokubun | 2023-03-18 | 1 | -5/+0 |
| | |||||
* | RJIT: Optimize String#bytesize | Takashi Kokubun | 2023-03-18 | 1 | -0/+4 |
| | |||||
* | RJIT: Optimize String#empty? | Takashi Kokubun | 2023-03-18 | 1 | -0/+25 |
| | |||||
* | RJIT: Optimize Kernel#instance_of? | Takashi Kokubun | 2023-03-18 | 1 | -0/+5 |
| | |||||
* | RJIT: Optimize Kernel#is_a? | Takashi Kokubun | 2023-03-18 | 1 | -0/+6 |
| | |||||
* | RJIT: Reorder opt_case_dispatch branches | Takashi Kokubun | 2023-03-18 | 1 | -3/+20 |
| | |||||
* | RJIT: Implement setclassvariable | Takashi Kokubun | 2023-03-18 | 1 | -0/+4 |
| | |||||
* | RJIT: Implement intern | Takashi Kokubun | 2023-03-18 | 1 | -0/+4 |
| | |||||
* | RJIT: Implement toregexp | Takashi Kokubun | 2023-03-18 | 1 | -0/+12 |
| | |||||
* | RJIT: Prefix rjit_options with rb_ | Takashi Kokubun | 2023-03-18 | 1 | -14/+14 |
| |