mirror of
https://github.com/ruby/ruby.git
synced 2022-11-09 12:17:21 -05:00
ed935aa5be
which has been developed by Takashi Kokubun <takashikkbn@gmail> as YARV-MJIT. Many of its bugs are fixed by wanabe <s.wanabe@gmail.com>. This JIT compiler is designed to be a safe migration path to introduce JIT compiler to MRI. So this commit does not include any bytecode changes or dynamic instruction modifications, which are done in original MJIT. This commit even strips off some aggressive optimizations from YARV-MJIT, and thus it's slower than YARV-MJIT too. But it's still fairly faster than Ruby 2.5 in some benchmarks (attached below). Note that this JIT compiler passes `make test`, `make test-all`, `make test-spec` without JIT, and even with JIT. Not only it's perfectly safe with JIT disabled because it does not replace VM instructions unlike MJIT, but also with JIT enabled it stably runs Ruby applications including Rails applications. I'm expecting this version as just "initial" JIT compiler. I have many optimization ideas which are skipped for initial merging, and you may easily replace this JIT compiler with a faster one by just replacing mjit_compile.c. `mjit_compile` interface is designed for the purpose. common.mk: update dependencies for mjit_compile.c. internal.h: declare `rb_vm_insn_addr2insn` for MJIT. vm.c: exclude some definitions if `-DMJIT_HEADER` is provided to compiler. This avoids to include some functions which take a long time to compile, e.g. vm_exec_core. Some of the purpose is achieved in transform_mjit_header.rb (see `IGNORED_FUNCTIONS`) but others are manually resolved for now. Load mjit_helper.h for MJIT header. mjit_helper.h: New. This is a file used only by JIT-ed code. I'll refactor `mjit_call_cfunc` later. vm_eval.c: add some #ifdef switches to skip compiling some functions like Init_vm_eval. win32/mkexports.rb: export thread/ec functions, which are used by MJIT. include/ruby/defines.h: add MJIT_FUNC_EXPORTED macro alis to clarify that a function is exported only for MJIT. array.c: export a function used by MJIT. bignum.c: ditto. class.c: ditto. compile.c: ditto. error.c: ditto. gc.c: ditto. hash.c: ditto. iseq.c: ditto. numeric.c: ditto. object.c: ditto. proc.c: ditto. re.c: ditto. st.c: ditto. string.c: ditto. thread.c: ditto. variable.c: ditto. vm_backtrace.c: ditto. vm_insnhelper.c: ditto. vm_method.c: ditto. I would like to improve maintainability of function exports, but I believe this way is acceptable as initial merging if we clarify the new exports are for MJIT (so that we can use them as TODO list to fix) and add unit tests to detect unresolved symbols. I'll add unit tests of JIT compilations in succeeding commits. Author: Takashi Kokubun <takashikkbn@gmail.com> Contributor: wanabe <s.wanabe@gmail.com> Part of [Feature #14235] --- * Known issues * Code generated by gcc is faster than clang. The benchmark may be worse in macOS. Following benchmark result is provided by gcc w/ Linux. * Performance is decreased when Google Chrome is running * JIT can work on MinGW, but it doesn't improve performance at least in short running benchmark. * Currently it doesn't perform well with Rails. We'll try to fix this before release. --- * Benchmark reslts Benchmarked with: Intel 4.0GHz i7-4790K with 16GB memory under x86-64 Ubuntu 8 Cores - 2.0.0-p0: Ruby 2.0.0-p0 - r62186: Ruby trunk (early 2.6.0), before MJIT changes - JIT off: On this commit, but without `--jit` option - JIT on: On this commit, and with `--jit` option ** Optcarrot fps Benchmark: https://github.com/mame/optcarrot | |2.0.0-p0 |r62186 |JIT off |JIT on | |:--------|:--------|:--------|:--------|:--------| |fps |37.32 |51.46 |51.31 |58.88 | |vs 2.0.0 |1.00x |1.38x |1.37x |1.58x | ** MJIT benchmarks Benchmark: https://github.com/benchmark-driver/mjit-benchmarks (Original: https://github.com/vnmakarov/ruby/tree/rtl_mjit_branch/MJIT-benchmarks) | |2.0.0-p0 |r62186 |JIT off |JIT on | |:----------|:--------|:--------|:--------|:--------| |aread |1.00 |1.09 |1.07 |2.19 | |aref |1.00 |1.13 |1.11 |2.22 | |aset |1.00 |1.50 |1.45 |2.64 | |awrite |1.00 |1.17 |1.13 |2.20 | |call |1.00 |1.29 |1.26 |2.02 | |const2 |1.00 |1.10 |1.10 |2.19 | |const |1.00 |1.11 |1.10 |2.19 | |fannk |1.00 |1.04 |1.02 |1.00 | |fib |1.00 |1.32 |1.31 |1.84 | |ivread |1.00 |1.13 |1.12 |2.43 | |ivwrite |1.00 |1.23 |1.21 |2.40 | |mandelbrot |1.00 |1.13 |1.16 |1.28 | |meteor |1.00 |2.97 |2.92 |3.17 | |nbody |1.00 |1.17 |1.15 |1.49 | |nest-ntimes|1.00 |1.22 |1.20 |1.39 | |nest-while |1.00 |1.10 |1.10 |1.37 | |norm |1.00 |1.18 |1.16 |1.24 | |nsvb |1.00 |1.16 |1.16 |1.17 | |red-black |1.00 |1.02 |0.99 |1.12 | |sieve |1.00 |1.30 |1.28 |1.62 | |trees |1.00 |1.14 |1.13 |1.19 | |while |1.00 |1.12 |1.11 |2.41 | ** Discourse's script/bench.rb Benchmark: https://github.com/discourse/discourse/blob/v1.8.7/script/bench.rb NOTE: Rails performance was somehow a little degraded with JIT for now. We should fix this. (At least I know opt_aref is performing badly in JIT and I have an idea to fix it. Please wait for the fix.) *** JIT off Your Results: (note for timings- percentile is first, duration is second in millisecs) categories_admin: 50: 17 75: 18 90: 22 99: 29 home_admin: 50: 21 75: 21 90: 27 99: 40 topic_admin: 50: 17 75: 18 90: 22 99: 32 categories: 50: 35 75: 41 90: 43 99: 77 home: 50: 39 75: 46 90: 49 99: 95 topic: 50: 46 75: 52 90: 56 99: 101 *** JIT on Your Results: (note for timings- percentile is first, duration is second in millisecs) categories_admin: 50: 19 75: 21 90: 25 99: 33 home_admin: 50: 24 75: 26 90: 30 99: 35 topic_admin: 50: 19 75: 20 90: 25 99: 30 categories: 50: 40 75: 44 90: 48 99: 76 home: 50: 42 75: 48 90: 51 99: 89 topic: 50: 49 75: 55 90: 58 99: 99 git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62197 b2dd03c8-39d4-4d8f-98ff-823fe69b080e |
||
---|---|---|
.. | ||
autorun.rb | ||
benchmark.rb | ||
mock.rb | ||
README.txt | ||
unit.rb |
= minitest/{unit,spec,mock,benchmark} home :: https://github.com/seattlerb/minitest rdoc :: http://docs.seattlerb.org/minitest vim :: https://github.com/sunaku/vim-ruby-minitest == DESCRIPTION: minitest provides a complete suite of testing facilities supporting TDD, BDD, mocking, and benchmarking. "I had a class with Jim Weirich on testing last week and we were allowed to choose our testing frameworks. Kirk Haines and I were paired up and we cracked open the code for a few test frameworks... I MUST say that minitest is *very* readable / understandable compared to the 'other two' options we looked at. Nicely done and thank you for helping us keep our mental sanity." -- Wayne E. Seguin minitest/unit is a small and incredibly fast unit testing framework. It provides a rich set of assertions to make your tests clean and readable. minitest/spec is a functionally complete spec engine. It hooks onto minitest/unit and seamlessly bridges test assertions over to spec expectations. minitest/benchmark is an awesome way to assert the performance of your algorithms in a repeatable manner. Now you can assert that your newb co-worker doesn't replace your linear algorithm with an exponential one! minitest/mock by Steven Baker, is a beautifully tiny mock (and stub) object framework. minitest/pride shows pride in testing and adds coloring to your test output. I guess it is an example of how to write IO pipes too. :P minitest/unit is meant to have a clean implementation for language implementors that need a minimal set of methods to bootstrap a working test suite. For example, there is no magic involved for test-case discovery. "Again, I can't praise enough the idea of a testing/specing framework that I can actually read in full in one sitting!" -- Piotr Szotkowski Comparing to rspec: rspec is a testing DSL. minitest is ruby. -- Adam Hawkins, "Bow Before MiniTest" minitest doesn't reinvent anything that ruby already provides, like: classes, modules, inheritance, methods. This means you only have to learn ruby to use minitest and all of your regular OO practices like extract-method refactorings still apply. == FEATURES/PROBLEMS: * minitest/autorun - the easy and explicit way to run all your tests. * minitest/unit - a very fast, simple, and clean test system. * minitest/spec - a very fast, simple, and clean spec system. * minitest/mock - a simple and clean mock/stub system. * minitest/benchmark - an awesome way to assert your algorithm's performance. * minitest/pride - show your pride in testing! * Incredibly small and fast runner, but no bells and whistles. == RATIONALE: See design_rationale.rb to see how specs and tests work in minitest. == SYNOPSIS: Given that you'd like to test the following class: class Meme def i_can_has_cheezburger? "OHAI!" end def will_it_blend? "YES!" end end === Unit tests require 'minitest/autorun' class TestMeme < MiniTest::Unit::TestCase def setup @meme = Meme.new end def test_that_kitty_can_eat assert_equal "OHAI!", @meme.i_can_has_cheezburger? end def test_that_it_will_not_blend refute_match /^no/i, @meme.will_it_blend? end def test_that_will_be_skipped skip "test this later" end end === Specs require 'minitest/autorun' describe Meme do before do @meme = Meme.new end describe "when asked about cheeseburgers" do it "must respond positively" do @meme.i_can_has_cheezburger?.must_equal "OHAI!" end end describe "when asked about blending possibilities" do it "won't say no" do @meme.will_it_blend?.wont_match /^no/i end end end For matchers support check out: https://github.com/zenspider/minitest-matchers === Benchmarks Add benchmarks to your regular unit tests. If the unit tests fail, the benchmarks won't run. # optionally run benchmarks, good for CI-only work! require 'minitest/benchmark' if ENV["BENCH"] class TestMeme < MiniTest::Unit::TestCase # Override self.bench_range or default range is [1, 10, 100, 1_000, 10_000] def bench_my_algorithm assert_performance_linear 0.9999 do |n| # n is a range value @obj.my_algorithm(n) end end end Or add them to your specs. If you make benchmarks optional, you'll need to wrap your benchmarks in a conditional since the methods won't be defined. describe Meme do if ENV["BENCH"] then bench_performance_linear "my_algorithm", 0.9999 do |n| 100.times do @obj.my_algorithm(n) end end end end outputs something like: # Running benchmarks: TestBlah 100 1000 10000 bench_my_algorithm 0.006167 0.079279 0.786993 bench_other_algorithm 0.061679 0.792797 7.869932 Output is tab-delimited to make it easy to paste into a spreadsheet. === Mocks class MemeAsker def initialize(meme) @meme = meme end def ask(question) method = question.tr(" ","_") + "?" @meme.__send__(method) end end require 'minitest/autorun' describe MemeAsker do before do @meme = MiniTest::Mock.new @meme_asker = MemeAsker.new @meme end describe "#ask" do describe "when passed an unpunctuated question" do it "should invoke the appropriate predicate method on the meme" do @meme.expect :will_it_blend?, :return_value @meme_asker.ask "will it blend" @meme.verify end end end end === Stubs def test_stale_eh obj_under_test = Something.new refute obj_under_test.stale? Time.stub :now, Time.at(0) do # stub goes away once the block is done assert obj_under_test.stale? end end A note on stubbing: In order to stub a method, the method must actually exist prior to stubbing. Use a singleton method to create a new non-existing method: def obj_under_test.fake_method ... end === Customizable Test Runner Types: MiniTest::Unit.runner=(runner) provides an easy way of creating custom test runners for specialized needs. Justin Weiss provides the following real-world example to create an alternative to regular fixture loading: class MiniTestWithHooks::Unit < MiniTest::Unit def before_suites end def after_suites end def _run_suites(suites, type) begin before_suites super(suites, type) ensure after_suites end end def _run_suite(suite, type) begin suite.before_suite super(suite, type) ensure suite.after_suite end end end module MiniTestWithTransactions class Unit < MiniTestWithHooks::Unit include TestSetupHelper def before_suites super setup_nested_transactions # load any data we want available for all tests end def after_suites teardown_nested_transactions super end end end MiniTest::Unit.runner = MiniTestWithTransactions::Unit.new == FAQ === How to test SimpleDelegates? The following implementation and test: class Worker < SimpleDelegator def work end end describe Worker do before do @worker = Worker.new(Object.new) end it "must respond to work" do @worker.must_respond_to :work end end outputs a failure: 1) Failure: Worker#test_0001_must respond to work [bug11.rb:16]: Expected #<Object:0x007f9e7184f0a0> (Object) to respond to #work. Worker is a SimpleDelegate which in 1.9+ is a subclass of BasicObject. Expectations are put on Object (one level down) so the Worker (SimpleDelegate) hits `method_missing` and delegates down to the `Object.new` instance. That object doesn't respond to work so the test fails. You can bypass `SimpleDelegate#method_missing` by extending the worker with `MiniTest::Expectations`. You can either do that in your setup at the instance level, like: before do @worker = Worker.new(Object.new) @worker.extend MiniTest::Expectations end or you can extend the Worker class (within the test file!), like: class Worker include ::MiniTest::Expectations end == Known Extensions: capybara_minitest_spec :: Bridge between Capybara RSpec matchers and MiniTest::Spec expectations (e.g. page.must_have_content('Title')). minispec-metadata :: Metadata for describe/it blocks (e.g. `it 'requires JS driver', js: true do`) minitest-ansi :: Colorize minitest output with ANSI colors. minitest-around :: Around block for minitest. An alternative to setup/teardown dance. minitest-capistrano :: Assertions and expectations for testing Capistrano recipes minitest-capybara :: Capybara matchers support for minitest unit and spec minitest-chef-handler :: Run Minitest suites as Chef report handlers minitest-ci :: CI reporter plugin for MiniTest. minitest-colorize :: Colorize MiniTest output and show failing tests instantly. minitest-context :: Defines contexts for code reuse in MiniTest specs that share common expectations. minitest-debugger :: Wraps assert so failed assertions drop into the ruby debugger. minitest-display :: Patches MiniTest to allow for an easily configurable output. minitest-emoji :: Print out emoji for your test passes, fails, and skips. minitest-english :: Semantically symmetric aliases for assertions and expectations. minitest-excludes :: Clean API for excluding certain tests you don't want to run under certain conditions. minitest-firemock :: Makes your MiniTest mocks more resilient. minitest-great_expectations :: Generally useful additions to minitest's assertions and expectations minitest-growl :: Test notifier for minitest via growl. minitest-implicit-subject :: Implicit declaration of the test subject. minitest-instrument :: Instrument ActiveSupport::Notifications when test method is executed minitest-instrument-db :: Store information about speed of test execution provided by minitest-instrument in database minitest-libnotify :: Test notifier for minitest via libnotify. minitest-macruby :: Provides extensions to minitest for macruby UI testing. minitest-matchers :: Adds support for RSpec-style matchers to minitest. minitest-metadata :: Annotate tests with metadata (key-value). minitest-mongoid :: Mongoid assertion matchers for MiniTest minitest-must_not :: Provides must_not as an alias for wont in MiniTest minitest-nc :: Test notifier for minitest via Mountain Lion's Notification Center minitest-predicates :: Adds support for .predicate? methods minitest-rails :: MiniTest integration for Rails 3.x minitest-rails-capybara :: Capybara integration for MiniTest::Rails minitest-reporters :: Create customizable MiniTest output formats minitest-should_syntax :: RSpec-style +x.should == y+ assertions for MiniTest minitest-shouldify :: Adding all manner of shoulds to MiniTest (bad idea) minitest-spec-context :: Provides rspec-ish context method to MiniTest::Spec minitest-spec-magic :: Minitest::Spec extensions for Rails and beyond minitest-spec-rails :: Drop in MiniTest::Spec superclass for ActiveSupport::TestCase. minitest-stub-const :: Stub constants for the duration of a block minitest-tags :: add tags for minitest minitest-wscolor :: Yet another test colorizer. minitest_owrapper :: Get tests results as a TestResult object. minitest_should :: Shoulda style syntax for minitest test::unit. minitest_tu_shim :: minitest_tu_shim bridges between test/unit and minitest. mongoid-minitest :: MiniTest matchers for Mongoid. pry-rescue :: A pry plugin w/ minitest support. See pry-rescue/minitest.rb. == Unknown Extensions: Authors... Please send me a pull request with a description of your minitest extension. * assay-minitest * detroit-minitest * em-minitest-spec * flexmock-minitest * guard-minitest * guard-minitest-decisiv * minitest-activemodel * minitest-ar-assertions * minitest-capybara-unit * minitest-colorer * minitest-deluxe * minitest-extra-assertions * minitest-rails-shoulda * minitest-spec * minitest-spec-should * minitest-sugar * minitest_should * mongoid-minitest * spork-minitest == REQUIREMENTS: * Ruby 1.8, maybe even 1.6 or lower. No magic is involved. == INSTALL: sudo gem install minitest On 1.9, you already have it. To get newer candy you can still install the gem, but you'll need to activate the gem explicitly to use it: require 'rubygems' gem 'minitest' # ensures you're using the gem, and not the built in MT require 'minitest/autorun' # ... usual testing stuffs ... DO NOTE: There is a serious problem with the way that ruby 1.9/2.0 packages their own gems. They install a gem specification file, but don't install the gem contents in the gem path. This messes up Gem.find_files and many other things (gem which, gem contents, etc). Just install minitest as a gem for real and you'll be happier. == LICENSE: (The MIT License) Copyright (c) Ryan Davis, seattle.rb Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.