1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
|
#!/usr/bin/env python3
"""
eval-accounting-overhead -- evaluate overhead of various accounting methods and energy/power/timestamp integer sizes
Usage:
PYTHONPATH=lib bin/eval-accounting-overhead.py <files ...>
Data Generation:
for accounting in static_state_immediate 'static_state' 'static_statetransition_immediate' 'static_statetransition'; do for intsize in uint16_t uint32_t uint64_t; do PYTHONPATH=~/var/ess/aemr/dfatool/lib ~/var/ess/aemr/dfatool/bin/generate-dfa-benchmark.py --timer-pin=GPIO::p1_0 --sleep=30 --repeat=10 --depth=10 --arch=msp430fr5994lp --app=test_benchmark --trace-filter='setup,setAutoAck,write,getEnergy,$' --timing --dummy= --accounting=${accounting},ts_type=${intsize},power_type=${intsize},energy_type=${intsize} model/driver/nrf24l01.dfa ~/var/projects/multipass/src/app/test_benchmark/main.cc; done; done
Feed the resulting files into this script, output is one line per file
providing overhead per transition and getEnergy overhead
"""
from dfatool.dfatool import AnalyticModel, TimingData, pta_trace_to_aggregate
import json
import sys
for filename in sys.argv[1:]:
with open(filename, "r") as f:
measurement = json.load(f)
raw_data = TimingData([filename])
preprocessed_data = raw_data.get_preprocessed_data()
by_name, parameters, arg_count = pta_trace_to_aggregate(preprocessed_data)
model = AnalyticModel(by_name, parameters, arg_count)
static_model = model.get_static()
if "setup" in model.names:
transition_duration = static_model("setup", "duration")
elif "init" in model.names:
transition_duration = static_model("init", "duration")
get_energy_duration = static_model("getEnergy", "duration")
print(
"{:60s}: {:.0f} / {:.0f} µs".format(
measurement["opt"]["accounting"], transition_duration, get_energy_duration
)
)
|
