mycpp Code Generation

Measure the speedup from mycpp, and the resource usage.

Source code: oil/mycpp/examples

User Time (milliseconds)

Lower ratios are better.

example name gen C++ Python C++ : Python
classes gen 0 16 0.000
fib_iter gen 4 891 0.004
modules gen 2 186 0.010
fib_recursive gen 11 886 0.012
loops gen 4 296 0.013
asdl_generated gen 11 379 0.029
parse gen 30 778 0.038
scoped_resource gen 43 1,025 0.042
files gen 7 69 0.107
tuple_return_value gen 20 183 0.108
containers gen 15 113 0.135
gc_stack_roots gen 2 12 0.147
length gen 45 203 0.220
cartesian gen 91 320 0.283
escape gen 106 348 0.305
cgi gen 264 517 0.511
varargs gen 23 20 1.176
control_flow gen 209 107 1.943

Max Resident Set Size (MB)

Lower ratios are better. We use MB (powers of 10), not MiB (powers of 2).

example name gen C++ Python C++ : Python
classes gen 4.5 10.8 0.41
gc_stack_roots gen 3.4 7.1 0.48
cgi gen 3.5 7.1 0.50
loops gen 3.5 7.1 0.50
parse gen 3.8 7.6 0.50
cartesian gen 3.7 6.9 0.53
control_flow gen 3.7 6.9 0.53
escape gen 3.7 6.9 0.53
modules gen 3.7 6.9 0.53
scoped_resource gen 3.8 7.1 0.54
asdl_generated gen 3.8 6.9 0.55
fib_recursive gen 3.8 6.9 0.55
files gen 3.8 6.9 0.55
length gen 3.8 6.9 0.55
tuple_return_value gen 3.8 6.8 0.56
fib_iter gen 3.8 6.7 0.57
containers gen 28.5 47.8 0.60
varargs gen 5.4 7.1 0.76

System Time (milliseconds)

Lower ratios are better.

example name gen C++ Python C++ : Python
asdl_generated gen 0 4 0.000
cartesian gen 0 8 0.000
containers gen 0 20 0.000
control_flow gen 0 8 0.000
escape gen 0 8 0.000
fib_recursive gen 0 4 0.000
files gen 0 8 0.000
length gen 0 12 0.000
loops gen 0 8 0.000
modules gen 0 4 0.000
parse gen 0 4 0.000
tuple_return_value gen 0 8 0.000
classes gen 3 16 0.206
scoped_resource gen 4 12 0.326
cgi gen 4 12 0.336
varargs gen 43 52 0.829
fib_iter gen 4 4 0.943
gc_stack_roots gen 0 0 NA

raw benchmark files