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 C++ Python C++ : Python
fib_iter 8 888 0.009
modules 2 179 0.010
fib_recursive 11 893 0.012
loops 4 294 0.013
asdl_generated 11 381 0.029
parse 26 814 0.032
scoped_resource 48 1,035 0.046
tuple_return_value 16 195 0.082
files 8 64 0.119
containers 16 121 0.131
length 37 205 0.181
classes 3 16 0.212
escape 95 432 0.221
cartesian 82 335 0.246
cgi 248 523 0.474
varargs 21 19 1.076
control_flow 210 112 1.880

Max Resident Set Size (MB)

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

example name C++ Python C++ : Python
classes 4.2 10.9 0.39
parse 3.8 7.5 0.51
escape 3.5 6.9 0.51
modules 3.5 6.9 0.51
loops 3.7 7.1 0.52
scoped_resource 3.7 7.1 0.52
cgi 3.7 6.9 0.53
length 3.7 6.9 0.53
cartesian 3.8 7.1 0.54
fib_recursive 3.8 7.1 0.54
asdl_generated 3.7 6.8 0.54
files 3.7 6.8 0.54
control_flow 3.8 6.9 0.55
fib_iter 3.8 6.9 0.55
tuple_return_value 3.8 6.8 0.56
containers 28.7 47.6 0.60
varargs 5.5 6.9 0.79

System Time (milliseconds)

Lower ratios are better.

example name C++ Python C++ : Python
classes 0 16 0.000
containers 0 16 0.000
control_flow 0 4 0.000
fib_iter 0 12 0.000
fib_recursive 0 8 0.000
files 0 12 0.000
loops 0 4 0.000
modules 0 8 0.000
scoped_resource 0 12 0.000
tuple_return_value 4 8 0.504
varargs 46 54 0.846
parse 4 4 0.930
cgi 8 8 1.002
length 8 4 2.054
escape 12 4 2.985
cartesian 8 0 inf
asdl_generated 0 0 NA

raw benchmark files