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 19 0.000
modules gen 0 132 0.000
fib_iter gen 7 742 0.009
fib_recursive gen 8 666 0.012
loops gen 3 207 0.013
asdl_generated gen 8 273 0.029
parse gen 21 560 0.037
files gen 3 51 0.052
scoped_resource gen 42 736 0.057
tuple_return_value gen 14 137 0.104
containers gen 9 80 0.106
length gen 28 150 0.186
escape gen 71 246 0.290
gc_stack_roots gen 1 4 0.358
cartesian gen 102 232 0.439
cgi gen 194 377 0.514
varargs gen 12 8 1.464
control_flow gen 151 78 1.938

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
parse gen 3.9 7.7 0.51
gc_stack_roots gen 3.7 7.1 0.52
loops gen 3.8 7.3 0.52
scoped_resource gen 3.8 7.1 0.53
cartesian gen 3.8 7.1 0.53
files gen 3.8 7.2 0.53
asdl_generated gen 3.7 7.1 0.53
fib_recursive gen 3.8 7.1 0.53
length gen 3.8 7.2 0.53
cgi gen 3.8 7.0 0.54
tuple_return_value gen 3.8 7.1 0.54
modules gen 3.9 7.2 0.54
escape gen 3.8 7.1 0.54
fib_iter gen 3.8 7.0 0.54
control_flow gen 3.9 7.1 0.55
containers gen 28.5 47.8 0.60
varargs gen 5.6 7.1 0.79

System Time (milliseconds)

Lower ratios are better.

example name gen C++ Python C++ : Python
cgi gen 0 4 0.000
control_flow gen 0 4 0.000
gc_stack_roots gen 0 4 0.000
loops gen 0 8 0.000
parse gen 0 4 0.000
tuple_return_value gen 0 4 0.000
containers gen 4 20 0.213
files gen 3 4 0.676
varargs gen 35 44 0.798
fib_iter gen 3 4 0.826
classes gen 3 4 0.897
scoped_resource gen 8 8 1.046
cartesian gen 12 8 1.470
escape gen 8 4 1.964
length gen 8 0 inf
modules gen 1 0 inf
asdl_generated gen 0 0 NA
fib_recursive gen 0 0 NA

raw benchmark files