perf-quick-wins - Branch - byteoak/py4j

perf: decode_bytearray (closes #570) + hot-path smart_decode shortcuts (credit @markjm #575)

#5Merged

Comparing

perf-quick-wins

(

e7b1485

) with

master

(

3c2f164

)

+91%

Improvement: 2

Untouched: 16

Skipped: 7

Benchmarks

25 total

test_macro_scenario[X7-16k]

py4j-python/src/py4j/tests/perf/scenarios/codspeed_macros.py

×3.3

102.9 ms31 ms

test_m5c_encode_float

py4j-python/src/py4j/tests/perf/scenarios/micro.py

+10%

3 µs2.7 µs

test_m7b_unescape

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-2%

21.5 µs22 µs

test_m5a_encode_int

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-2%

2.1 µs2.1 µs

test_m1_static_call_no_args

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-2%

95.2 µs97.4 µs

test_macro_scenario[X6]

py4j-python/src/py4j/tests/perf/scenarios/codspeed_macros.py

+2%

69.2 ms67.8 ms

test_macro_scenario[X2-10k]

py4j-python/src/py4j/tests/perf/scenarios/codspeed_macros.py

-2%

884.1 ms902.8 ms

test_macro_scenario[X1-1]

py4j-python/src/py4j/tests/perf/scenarios/codspeed_macros.py

-2%

946.2 ms965.4 ms

test_m3_jvmview_class_resolution

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-2%

390.1 µs396.5 µs

test_m4_constructor_and_finalize

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-1%

582.7 µs588 µs

test_m6b_decode_string

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-1%

5.2 µs5.2 µs

test_m2b_instance_append_str

py4j-python/src/py4j/tests/perf/scenarios/micro.py

-1%

190.5 µs191.9 µs

test_m6a_decode_int

py4j-python/src/py4j/tests/perf/scenarios/micro.py

1.9 µs1.9 µs

test_macro_scenario[X4]

py4j-python/src/py4j/tests/perf/scenarios/codspeed_macros.py

41.8 ms41.9 ms

test_m7a_escape

py4j-python/src/py4j/tests/perf/scenarios/micro.py

4.5 µs4.5 µs

test_m2a_instance_append_int

py4j-python/src/py4j/tests/perf/scenarios/micro.py

190.4 µs190.6 µs

test_m5b_encode_string

py4j-python/src/py4j/tests/perf/scenarios/micro.py

2.6 µs2.6 µs

test_macro_scenario[X8-16k]

py4j-python/src/py4j/tests/perf/scenarios/codspeed_macros.py

46.2 ms46.1 ms

The benchmarks below were skipped, so their baseline results are used instead. If they were deleted in your codebase, archive them to remove them from the performance reports.
Learn more about archiving benchmarks

test_m5b_encode_string

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

28.2 µs*

test_m5c_encode_float

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

27.8 µs*

test_m7a_escape

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

22.2 µs*

test_m6a_decode_int

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

22.9 µs*

test_m7b_unescape

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

62.4 µs*

test_m6b_decode_string

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

37.8 µs*

test_m5a_encode_int

py4j-python/src/py4j/tests/perf/scenarios/micro.py

Skipped

25.7 µs*

Commits

Click on a commit to change the comparison range

Base

master

3c2f164

+91.12%

perf: drop smart_decode dispatch on hot recv + minor protocol cleanups

e7b1485

4 months ago

by Tagar

Home Terms Privacy Docs