Running benchmarks of context-threaded Ruby vs. normal Ruby. ctruby build with regular ./configure and no special flags. Benchmarks performed on a MacBook Air mid-2011.

gistfile1.txt

$ ruby -v
ruby 2.0.0p247 (2013-06-27 revision 41674) [x86_64-darwin12.4.0]

$ ./ruby -v
ruby 2.1.0dev (2013-10-23 trunk 43398) [x86_64-darwin13.0.0]

$ ruby benchmark/driver.rb --executables="normal_ruby::/opt/boxen/rbenv/shims/ruby; ctruby::/Users/edward/Code/C/ruby/ruby -Ilib:.:.ext/common:.ext/x86_64-darwin13.0.0/" --pattern='bm_' --directory=/Users/edward/Code/C/ruby/benchmark -r 5
preparing benchmark/fasta.output.100000
preparing benchmark/fasta.output.2500000

-----------------------------------------------------------
benchmark results:
Execution time (sec)
name	normal_ruby	ctruby
app_answer	0.206	0.227
app_aobench	112.314	220.067
app_erb	2.274	4.616
app_factorial	2.303	5.646
app_fib	1.216	2.078
app_mandelbrot	2.316	6.093
app_pentomino	32.636	70.389
app_raise	0.775	1.523
app_strconcat	1.953	5.325
app_tak	1.608	2.866
app_tarai	1.290	2.340
app_uri	1.836	3.385
hash_shift	6.037	0.156
io_file_create	5.496	6.338
io_file_read	4.336	5.473
io_file_write	1.569	2.047
io_select	4.233	5.691
io_select2	9.145	10.620
io_select3	0.234	0.225
loop_for	2.598	5.693
loop_generator	0.945	1.768
loop_times	2.320	5.116
loop_whileloop	1.017	1.884
loop_whileloop2	0.296	0.439
so_ackermann	1.375	2.380
so_array	2.143	4.851
so_binary_trees	13.412	27.378
so_concatenate	7.534	16.212
so_count_words	0.508	0.834
so_exception	0.752	1.494
so_fannkuch	2.933	5.210
so_fasta	3.814	8.254
so_k_nucleotide	2.327	5.388
so_lists	1.293	2.804
so_mandelbrot	4.692	9.185
so_matrix	1.392	3.122
so_meteor_contest	5.561	13.407
so_nbody	3.054	5.692
so_nested_loop	2.040	4.750
so_nsieve	3.912	9.161
so_nsieve_bits	4.138	10.488
so_object	1.630	3.596
so_partial_sums	3.718	6.599
so_pidigits	1.581	3.563
so_random	0.772	1.528
so_reverse_complement	2.536	4.988
so_sieve	1.349	2.884
so_spectralnorm	3.670	8.030
vm1_attr_ivar*	2.046	3.699
vm1_attr_ivar_set*	2.561	4.744
vm1_block*	3.401	7.612
vm1_const*	0.528	1.262
vm1_ensure*	0.302	0.537
vm1_float_simple*	8.102	17.805
vm1_ivar*	1.608	2.339
vm1_ivar_set*	1.950	2.760
vm1_length*	1.166	2.648
vm1_lvar_init*	2.313	12.499
vm1_lvar_set*	3.510	7.086
vm1_neq*	0.844	2.311
vm1_not*	0.415	1.280
vm1_rescue*	0.199	0.530
vm1_simplereturn*	1.647	3.515
vm1_swap*	0.399	0.997
vm1_yield*	1.675	4.162
vm2_array*	1.692	2.751
vm2_bigarray*	28.032	14.551
vm2_bighash*	9.813	14.773
vm2_case*	0.221	0.505
vm2_defined_method*	5.216	9.901
vm2_dstr*	2.265	6.053
vm2_eval*	43.876	251.476
vm2_method*	2.553	5.170
vm2_method_missing*	3.871	9.723
vm2_method_with_block*	2.943	5.717
vm2_mutex*	1.797	3.376
vm2_poly_method*	4.213	10.465
vm2_poly_method_ov*	0.324	0.554
vm2_proc*	0.841	1.764
vm2_raise1*	13.720	32.739
vm2_raise2*	19.382	44.209
vm2_regexp*	1.829	4.610
vm2_send*	0.574	1.262
vm2_super*	0.863	2.004
vm2_unif1*	0.322	0.731
vm2_zsuper*	0.912	2.049
vm3_backtrace	0.473	0.656
vm3_clearmethodcache	1.084	1.647
vm3_gc	16.075	10.622
vm_thread_alive_check1	0.406	0.455
vm_thread_create_join	4.611	5.386
vm_thread_mutex1	1.536	2.835
vm_thread_mutex2	2.712	4.948
vm_thread_mutex3	171.055	207.535
vm_thread_pass	1.281	1.677
vm_thread_pass_flood	0.674	0.297
vm_thread_pipe	0.798	1.401

Speedup ratio: compare with the result of `normal_ruby' (greater is better)
name	ctruby
app_answer	0.908
app_aobench	0.510
app_erb	0.493
app_factorial	0.408
app_fib	0.585
app_mandelbrot	0.380
app_pentomino	0.464
app_raise	0.509
app_strconcat	0.367
app_tak	0.561
app_tarai	0.551
app_uri	0.542
hash_shift	38.782
io_file_create	0.867
io_file_read	0.792
io_file_write	0.766
io_select	0.744
io_select2	0.861
io_select3	1.042
loop_for	0.456
loop_generator	0.534
loop_times	0.453
loop_whileloop	0.540
loop_whileloop2	0.674
so_ackermann	0.578
so_array	0.442
so_binary_trees	0.490
so_concatenate	0.465
so_count_words	0.610
so_exception	0.503
so_fannkuch	0.563
so_fasta	0.462
so_k_nucleotide	0.432
so_lists	0.461
so_mandelbrot	0.511
so_matrix	0.446
so_meteor_contest	0.415
so_nbody	0.536
so_nested_loop	0.430
so_nsieve	0.427
so_nsieve_bits	0.395
so_object	0.453
so_partial_sums	0.563
so_pidigits	0.444
so_random	0.506
so_reverse_complement	0.508
so_sieve	0.468
so_spectralnorm	0.457
vm1_attr_ivar*	0.553
vm1_attr_ivar_set*	0.540
vm1_block*	0.447
vm1_const*	0.418
vm1_ensure*	0.563
vm1_float_simple*	0.455
vm1_ivar*	0.687
vm1_ivar_set*	0.706
vm1_length*	0.440
vm1_lvar_init*	0.185
vm1_lvar_set*	0.495
vm1_neq*	0.365
vm1_not*	0.324
vm1_rescue*	0.374
vm1_simplereturn*	0.468
vm1_swap*	0.400
vm1_yield*	0.403
vm2_array*	0.615
vm2_bigarray*	1.926
vm2_bighash*	0.664
vm2_case*	0.437
vm2_defined_method*	0.527
vm2_dstr*	0.374
vm2_eval*	0.174
vm2_method*	0.494
vm2_method_missing*	0.398
vm2_method_with_block*	0.515
vm2_mutex*	0.532
vm2_poly_method*	0.403
vm2_poly_method_ov*	0.584
vm2_proc*	0.477
vm2_raise1*	0.419
vm2_raise2*	0.438
vm2_regexp*	0.397
vm2_send*	0.455
vm2_super*	0.430
vm2_unif1*	0.441
vm2_zsuper*	0.445
vm3_backtrace	0.720
vm3_clearmethodcache	0.658
vm3_gc	1.513
vm_thread_alive_check1	0.894
vm_thread_create_join	0.856
vm_thread_mutex1	0.542
vm_thread_mutex2	0.548
vm_thread_mutex3	0.824
vm_thread_pass	0.764
vm_thread_pass_flood	2.267
vm_thread_pipe	0.569

Log file: bmlog-20131029-020754.98319

I’m going to try running with CFLAGS=-g3 ./configure --enable-debug-env just to see.