JRuby 9000 - Taipei Ruby User's Group

Talk on JRuby 9000 delivered at 5xRuby in Taipei on December 1 2015.

Transcript

1. JRuby 9000 Charles O. Nutter JRuby; Red Hat
2. JRuby is Ruby… (on the JVM...shhhh!)
3. Why the JVM is great for Ruby!
4. Shoulders of Giants JVM J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose Hiro Marcin Nahi Subbu Douglas Christian Dmitry Tom Charlie JRuby
5. All the stuff! JVM J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose J. Rose Garbage Collection Native JIT Profiled Optimizations Native Threading Tooling Cross Platform
6. Can leverage Java Ecosystem 47k libraries in Maven Hadoop EHCache Selenium Sitemesh Lucene Neo4j JMonkeyEngine
7. Polyglot Clojure Scala Groovy Jython Rhino/Nashorn/ DynJS (JavaScript) Micro Focus JVM Visual COBOL Java
8. red/black tree, pure Ruby versus native ruby-2.0.0 + Ruby ruby-2.0.0 + C ext jruby + Ruby Runtime per iteration 0 0.75 1.5 2.25 3 0.29s 0.51s 2.48s
9. GC
10. GC Matters • Applications grow over time • Ruby is very object-heavy • Multiprocess multiplies the problem • You will eventually have issues
11. gc_demo.rb • Heavy GC, mix of old and young • Steadily growing heap use
12. class Simple attr_accessor :next end top = Simple.new puts Benchmark.measure { outer = 10 total = 100000 per = 100 outer.times do total.times do per.times { Simple.new } s = Simple.new top.next = s top = s end end }
13. 0 750 1500 2250 3000 GC count Ruby 2.1.1 JRuby
14. 1 10 100 1000 10000 GC count Ruby 2.1.1 JRuby
15. 0s 0.45s 0.9s 1.35s 1.8s GC time % Ruby 2.2.2 JRuby
16. Threads
17. Real Parallellism • Ruby thread = JVM thread = native thread • One process can use all cores • One server can handle all requests
18. Ruby 2.2 unthreaded Ruby 2.2 threaded JRuby unthreaded JRuby threaded
19. Per-iteration time versus thread count 0.2s 0.35s 0.5s 0.65s 0.8s one thread two threads three threads four threads threaded_reverse
20. Tools
21. Profiling • Java profilers • VisualVM, YourKit, NetBeans, JXInsight • jruby [--profile | --profile.graph] • JVM command-line profilers
22. VisualVM • CPU, memory, thread monitoring • CPU and memory profiling • VisualGC • Heap analysis
26. Scripting Java
27. Purugin • Nearly 100% Ruby wrapper • Thin shim makes Java feel very Ruby-like • It’s Minecraft!
28. Egg Madness
29. Egg Madness class EggMadnessPlugin include Purugin::Plugin description 'EggMadness', 0.1 def on_enable event(:player_egg_throw) do |e| e.hatching = true e.num_hatches = 50 e.hatching_type = :chicken end end end
30. 1.7.23 JRuby Roadmap master master 1.7.3 1.7.4 1.7.5 1.7.6 ... 1.7.7 1.7.22 ... 9.0.4 2.2 1.8, 1.9 jruby-1_7 jruby-1_7 jruby-1_7 master 2.2 1.8, 1.9 End of week Last Friday
31. JRuby 9000 • Ruby 2.2 • New runtime (IR) • Major IO and Encodings overhaul
32. “It’s over 9000!!!!”
33. Now What?
34. PERFORMANCE WORK topic of this talk Now and into not so distant future!
35. Recent Wins • JITable blocks • define_method performance • Reduced-cost transient exceptions
36. Block Jitting • JRuby 1.7 only jitted methods • Not free-standing procs/lambdas • Not define_method blocks • Easier to do now with 9000's IR • Blocks JIT now in 9.0.4.0
37. Jitting is Winning Performance of define_method in loaded file 0k iters/s 750k iters/s 1500k iters/s 2250k iters/s 3000k iters/s MRI JRuby 9.0.1.0 JRuby 9.0.4.0 normal method define_method method ruby -e 'load "bench_define_method.rb"'
38. define_method Convenient for metaprogramming, but blocks have more overhead than methods. define_method(:add) do |a, b|
 a + b
 end names.each do |name|
 define_method(name) { send :"do_#{name}" }
 end
39. :-( 0k iters/s 1000k iters/s 2000k iters/s 3000k iters/s 4000k iters/s MRI JRuby 9.0.1.0 def define_method define_method w/ capture
40. Optimizing define_method • Noncapturing • Treat as method in compiler • Ignore surrounding scope • Capturing (future work) • Lift read-only variables as constant
41. Getting Better! 0k iters/s 1000k iters/s 2000k iters/s 3000k iters/s 4000k iters/s MRI JRuby 9.0.1.0 JRuby 9.0.4.0 def define_method define_method w/ capture
42. Reduced-cost Exceptions • Backtrace cost is VERY high on JVM • Heavily optimized, lots of work to build • Exceptions frequently ignored • ...or used as flow control (shame!) • If ignored, backtrace is not needed!
43. Postfix Antipattern foo rescue nil Exception raised StandardError rescued Exception ignored Result is simple expression, so exception is never visible.
44. csv.rb Converters Converters = { integer: lambda { |f|
 Integer(f.encode(ConverterEncoding)) rescue f
 },
 float: lambda { |f|
 Float(f.encode(ConverterEncoding)) rescue f
 },
 ... All trivial rescues, no traces needed.
45. Simple rescue Improvement 0 150000 300000 450000 600000 Iters/second 524,475 10,700
46. Nearly Two Magnitudes! 1 10 100 1000 10000 100000 1000000 Iters/second 524,475 10,700
47. New Runtime? • AST to semantic representation • Traditional Compiler Design • Wanted Architectural longevity

48. Lexical Analysis Parsing Semantic Analysis Optimization Bytecode Generation Interpret AST IR Instructions CFG DFG ... JRuby 1.7.x 9000+ Dalvik Generation ...
49. def foo(a, b) c = 1 d = a + c end 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 2 b = recv_pre_reqd_arg(1) 3 %block = recv_closure 4 thread_poll 5 line_num(1) 6 c = 1 7 line_num(2) 8 %v_0 = call(:+, a, [c]) 9 d = copy(%v_0) 10 return(%v_0) Register-based 3 address format IR Instructions Semantic Analysis
50. -Xir.passes=LocalOptimizationPass, DeadCodeElimination def foo(a, b) c = 1 d = a + c end 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 2 b = recv_pre_reqd_arg(1) 3 %block = recv_closure 4 thread_poll 5 line_num(1) 6 c = 1 7 line_num(2) 8 %v_0 = call(:+, a, [c]) 9 d = copy(%v_0) 10 return(%v_0) Optimization
51. def foo(a, b) c = 1 d = a + c end 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 4 thread_poll 5 line_num(1) 6 c = 1 7 line_num(2) 8 %v_0 = call(:+, a, [c]) 9 d = copy(%v_0) 10 return(%v_0) 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 4 thread_poll 5 line_num(1) 6 c = 7 line_num(2) 8 %v_0 = call(:+, a, [ ]) 9 d = copy(%v_0) 10 return(%v_0) 1 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 4 thread_poll 5 line_num(1) 7 line_num(2) 8 %v_0 = call(:+, a, [1]) 9 d = copy(%v_0) 10 return(%v_0) Optimization -Xir.passes=LocalOptimizationPass, DeadCodeElimination
52. 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 4 thread_poll 5 line_num(1) 7 line_num(2) 8 %v_0 = call(:+, a, [1]) 9 d = copy(%v_0) 10 return(%v_0) 0 check_arity(2, 0, -1) 1 a = recv_pre_reqd_arg(0) 4 thread_poll 7 line_num(2) 8 %v_0 = call(:+, a, [1]) 9 d = copy(%v_0) 10 return(%v_0) Optimization -Xir.passes=LocalOptimizationPass, DeadCodeElimination
53. Inlining • 500 pound gorilla of optimizations • shove method/closure back to callsite • eliminate stack frame • eliminate parameter passing/return • eliminate additional allocation Optimization
54. Today’s Inliner def decrement_one(i) i - 1 end i = 1_000_000 while i > 0 i = decrement_one(i) end def decrement_one(i) i - 1 end i = 1_000_000 while i < 0 if guard_same? self i = i - 1 else i = decrement_one(i) end end
55. Numeric Specialization • Everything's an object • JVM has only references and primitives • Not compatible in bytecode • Need to optimize numerics as primitive
56. def looper(n)
 i = 0
 while i < n
 do_something(i)
 i += 1
 end
 end Cached object Call with i New Fixnum i + 1 Probably a Fixnum?
57. def looper(n)
 i = 0
 while i < n
 do_something(i)
 i += 1
 end
 end def looper(long n)
 long i = 0
 while i < n
 do_something(i)
 i += 1
 end
 end Specialize n, i to long def looper(n)
 i = 0
 while i < n
 do_something(i)
 i += 1
 end
 end Deopt to object version if n or i + 1 is not Fixnum
58. JVM Futures • We're good friends with OpenJDK folks • Working to improve JVM as well • FFI being added at JVM level • AOT compilation for startup perf
59. FFI in JVM • Project Panama (JEP-191) • Native support for FFI • Code generators for binding • JIT support for calling • API support for userland
60. User Code JNI call JNI impl Target Library Java C/native
61. User Code JNR stub JNI call JNI impl libffi Target Library Java C/native
62. User Code Panama Target Library Java C/native JIT knows about both sides
63. JIT Magic callq <getpid address> ; - libSystem.B.dylib ;*invokeinterface getpid ; - GetPidJNRExample::benchGetPid@13 (line 26) ; {optimized virtual_call} Direct call from JITed Ruby code
64. Startup Time • By far our greatest challenge • Everything starts cold: parser, interpreter, compiler, core classes, boot logic • Increasing amount of Ruby in JRuby • Aggravates the problem
65. JRuby Startup C Ruby JRuby Time in seconds (lower is better) 0s 3.5s 7s 10.5s 14s -e 1 gem --list rake -T in Rals app
66. --dev • Disables JRuby JIT • Sets JVM to reduced optimization mode • 50% reduction in startup time • Much lower peak perf
67. JRuby --dev C Ruby JRuby JRuby --dev Time in seconds (lower is better) 0s 3.5s 7s 10.5s 14s -e 1 gem --list rake -T in Rals app
68. AOT • Precompile JVM bytecode to native • Focus on hot code • Save original structure for optimization • Get JRuby running native right away • AOT compile Ruby to native in future
69. Getting There C Ruby JRuby JRuby --dev Non-opto AOT Opto AOT Time in seconds (lower is better) 0s 3.5s 7s 10.5s 14s rake -T in Rails app
70. AOT Future • AOT might be available in Java 9 • Many tweaks we can make to help it • Ideal: all code run at boot runs native • Should get closer to MRI
73. Thank You • @headius • @tom_enebo • http://jruby.org