Usage¶
Installation¶
Command to install pyperformance:
python3 -m pip install pyperformance
The command installs a new pyperformance
program.
If needed, pyperf
and six
dependencies are installed automatically.
pyperformance works on Python 3.6 and newer, but it may work on Python 3.4 and 3.5.
At runtime, Python development files (header files) may be needed to install
some dependencies like dulwich_log
or psutil
, to build their C
extension. Commands on Fedora to install dependencies:
- Python 3:
sudo dnf install python3-devel
- PyPy:
sudo dnf install pypy-devel
Windows notes¶
On Windows, to allow pyperformance to build dependencies from source
like greenlet
, dulwich
or psutil
, if you want to use a
python.exe
built from source, you should not use the python.exe
directly. Instead, you must run the little-known command PC\layout
to create a filesystem layout that resembles an installed Python:
.\python.bat -m PC.layout --preset-default --copy installed -v
(Use the --help
flag for more info about PC\layout
.)
Now you can use the “installed” Python executable:
installed\python.exe -m pip install pyperformance
installed\python.exe -m pyperformance run ...
Using an actually installed Python executable (e.g. via py
)
works fine too.
Run benchmarks¶
Commands to compare Python 3.6 and Python 3.7 performance:
pyperformance run --python=python3.6 -o py36.json
pyperformance run --python=python3.7 -o py38.json
pyperformance compare py36.json py38.json
Note: python3 -m pyperformance ...
syntax works as well (ex: python3 -m
pyperformance run -o py38.json
), but requires to install pyperformance on each
tested Python version.
JSON files are produced by the pyperf module and so can be analyzed using pyperf commands:
python3 -m pyperf show py36.json
python3 -m pyperf check py36.json
python3 -m pyperf metadata py36.json
python3 -m pyperf stats py36.json
python3 -m pyperf hist py36.json
python3 -m pyperf dump py36.json
(...)
It’s also possible to use pyperf to compare results of two JSON files:
python3 -m pyperf compare_to py36.json py38.json --table
Basic commands¶
pyperformance actions:
run Run benchmarks on the running python
show Display a benchmark file
compare Compare two benchmark files
list List benchmarks of the running Python
list_groups List benchmark groups of the running Python
venv Actions on the virtual environment
run¶
Run benchmarks on the running python.
Usage:
pyperformance run [-h] [-r] [-f] [--debug-single-value] [-v] [-m]
[--affinity CPU_LIST] [-o FILENAME]
[--append FILENAME] [--manifest MANIFEST]
[-b BM_LIST] [--inherit-environ VAR_LIST]
[-p PYTHON]
options:
-h, --help show this help message and exit
-r, --rigorous Spend longer running tests to get more
accurate results
-f, --fast Get rough answers quickly
--debug-single-value Debug: fastest mode, only compute a single
value
-v, --verbose Print more output
-m, --track-memory Track memory usage. This only works on Linux.
--affinity CPU_LIST Specify CPU affinity for benchmark runs. This
way, benchmarks can be forced to run on a
given CPU to minimize run to run variation.
-o FILENAME, --output FILENAME
Run the benchmarks on only one interpreter and
write benchmark into FILENAME. Provide only
baseline_python, not changed_python.
--append FILENAME Add runs to an existing file, or create it if
it doesn't exist
--manifest MANIFEST benchmark manifest file to use
-b BM_LIST, --benchmarks BM_LIST
Comma-separated list of benchmarks to run. Can
contain both positive and negative arguments:
--benchmarks=run_this,also_this,-not_this. If
there are no positive arguments, we'll run all
benchmarks except the negative arguments.
Otherwise we run only the positive arguments.
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
compare¶
Compare two benchmark files.
Usage:
pyperformance compare [-h] [-v] [-O STYLE] [--csv CSV_FILE]
[--inherit-environ VAR_LIST] [-p PYTHON]
baseline_file.json changed_file.json
positional arguments:
baseline_file.json
changed_file.json
options:
-v, --verbose Print more output
-O STYLE, --output_style STYLE
What style the benchmark output should take.
Valid options are 'normal' and 'table'.
Default is normal.
--csv CSV_FILE Name of a file the results will be written to,
as a three-column CSV file containing minimum
runtimes for each benchmark.
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
list¶
List benchmarks of the running Python.
Usage:
pyperformance list [-h] [--manifest MANIFEST] [-b BM_LIST]
[--inherit-environ VAR_LIST] [-p PYTHON]
options:
--manifest MANIFEST benchmark manifest file to use
-b BM_LIST, --benchmarks BM_LIST
Comma-separated list of benchmarks to run. Can
contain both positive and negative arguments:
--benchmarks=run_this,also_this,-not_this. If
there are no positive arguments, we'll run all
benchmarks except the negative arguments.
Otherwise we run only the positive arguments.
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
Use python3 -m pyperformance list -b all
to list all benchmarks.
list_groups¶
List benchmark groups of the running Python.
Usage:
pyperformance list_groups [-h] [--manifest MANIFEST]
[--inherit-environ VAR_LIST]
[-p PYTHON]
options:
--manifest MANIFEST benchmark manifest file to use
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
venv¶
Actions on the virtual environment.
Actions:
show Display the path to the virtual environment and its status
(created or not)
create Create the virtual environment
recreate Force the recreation of the the virtual environment
remove Remove the virtual environment
Common options:
--venv VENV Path to the virtual environment
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
venv show¶
Display the path to the virtual environment and its status (created or not).
Usage:
pyperformance venv show [-h] [--venv VENV]
[--inherit-environ VAR_LIST] [-p PYTHON]
venv create¶
Create the virtual environment.
Usage:
pyperformance venv create [-h] [--venv VENV]
[--manifest MANIFEST] [-b BM_LIST]
[--inherit-environ VAR_LIST]
[-p PYTHON]
options:
--manifest MANIFEST benchmark manifest file to use
-b BM_LIST, --benchmarks BM_LIST
Comma-separated list of benchmarks to run. Can
contain both positive and negative arguments:
--benchmarks=run_this,also_this,-not_this. If
there are no positive arguments, we'll run all
benchmarks except the negative arguments.
Otherwise we run only the positive arguments.
venv recreate¶
Force the recreation of the the virtual environment.
Usage:
pyperformance venv recreate [-h] [--venv VENV]
[--manifest MANIFEST] [-b BM_LIST]
[--inherit-environ VAR_LIST]
[-p PYTHON]
options:
--manifest MANIFEST benchmark manifest file to use
-b BM_LIST, --benchmarks BM_LIST
Comma-separated list of benchmarks to run. Can
contain both positive and negative arguments:
--benchmarks=run_this,also_this,-not_this. If
there are no positive arguments, we'll run all
benchmarks except the negative arguments.
Otherwise we run only the positive arguments.
venv remove¶
Remove the virtual environment.
Usage:
pyperformance venv remove [-h] [--venv VENV]
[--inherit-environ VAR_LIST]
[-p PYTHON]
Compile Python to run benchmarks¶
pyperformance actions:
compile Compile and install CPython and run benchmarks on
installed Python
compile_all Compile and install CPython and run benchmarks on
installed Python on all branches and revisions of
CONFIG_FILE
upload Upload JSON results to a Codespeed website
All these commands require a configuration file.
Simple configuration usable for compile
(but not for compile_all
nor
upload
), doc/benchmark.conf
:
[config]
json_dir = ~/prog/python/bench_json
[scm]
repo_dir = ~/prog/python/master
update = True
[compile]
bench_dir = ~/prog/python/bench_tmpdir
[run_benchmark]
system_tune = True
affinity = 2,3
Configuration file sample with comments, doc/benchmark.conf.sample
:
[config]
# Directory where JSON files are written.
# - uploaded files are moved to json_dir/uploaded/
# - results of patched Python are written into json_dir/patch/
json_dir = ~/json
# If True, compile CPython is debug mode (LTO and PGO disabled),
# run benchmarks with --debug-single-sample, and disable upload.
#
# Use this option used to quickly test a configuration.
debug = False
[scm]
# Directory of CPython source code (Git repository)
repo_dir = ~/cpython
# Update the Git repository (git fetch)?
update = True
# Name of the Git remote, used to create revision of
# the Git branch. For example, use revision 'remotes/origin/3.6'
# for the branch '3.6'.
git_remote = remotes/origin
[compile]
# Create files into bench_dir:
# - bench_dir/bench-xxx.log
# - bench_dir/prefix/: where Python is installed
# - bench_dir/venv/: Virtual environment used by pyperformance
bench_dir = ~/bench_tmpdir
# Link Time Optimization (LTO)?
lto = True
# Profiled Guided Optimization (PGO)?
pgo = True
# The space-separated list of libraries that are package-only,
# i.e., locally installed but not on header and library paths.
# For each such library, determine the install path and add an
# appropriate subpath to CFLAGS and LDFLAGS declarations passed
# to configure. As an exception, the prefix for openssl, if that
# library is present here, is passed via the --with-openssl
# option. Currently, this only works with Homebrew on macOS.
# If running on macOS with Homebrew, you probably want to use:
# pkg_only = openssl readline sqlite3 xz zlib
# The version of zlib shipping with macOS probably works as well,
# as long as Apple's SDK headers are installed.
pkg_only =
# Install Python? If false, run Python from the build directory
#
# WARNING: Running Python from the build directory introduces subtle changes
# compared to running an installed Python. Moreover, creating a virtual
# environment using a Python run from the build directory fails in many cases,
# especially on Python older than 3.4. Only disable installation if you
# really understand what you are doing!
install = True
[run_benchmark]
# Run "sudo python3 -m pyperf system tune" before running benchmarks?
system_tune = True
# --manifest option for 'pyperformance run'
manifest =
# --benchmarks option for 'pyperformance run'
benchmarks =
# --affinity option for 'pyperf system tune' and 'pyperformance run'
affinity =
# Upload generated JSON file?
#
# Upload is disabled on patched Python, in debug mode or if install is
# disabled.
upload = False
# Specify '-j' parameter in 'make' command
jobs = 8
# Configuration to upload results to a Codespeed website
[upload]
url =
environment =
executable =
project =
[compile_all]
# List of CPython Git branches
branches = default 3.6 3.5 2.7
# List of revisions to benchmark by compile_all
[compile_all_revisions]
# list of 'sha1=' (default branch: 'master') or 'sha1=branch'
# used by the "pyperformance compile_all" command
# e.g.:
11159d2c9d6616497ef4cc62953a5c3cc8454afb =
compile¶
Compile Python, install Python and run benchmarks on the installed Python.
Usage:
pyperformance compile [-h] [--patch PATCH] [-U] [-T]
[--inherit-environ VAR_LIST] [-p PYTHON]
config_file revision [branch]
positional arguments:
config_file Configuration filename
revision Python benchmarked revision
branch Git branch
options:
--patch PATCH Patch file
-U, --no-update Don't update the Git repository
-T, --no-tune Don't run 'pyperf system tune' to tune the
system for benchmarks
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
Notes:
- PGO is broken on Ubuntu 14.04 LTS with GCC 4.8.4-2ubuntu1~14.04:
Modules/socketmodule.c:7743:1: internal compiler error: in edge_badness, at ipa-inline.c:895
compile_all¶
Compile all branches and revisions of CONFIG_FILE.
Usage:
pyperformance compile_all [-h] [--inherit-environ VAR_LIST] [-p PYTHON]
config_file
positional arguments:
config_file Configuration filename
options:
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
upload¶
Upload results from a JSON file to a Codespeed website.
Usage:
pyperformance upload [-h] [--inherit-environ VAR_LIST] [-p PYTHON]
config_file json_file
positional arguments:
config_file Configuration filename
json_file JSON filename
options:
--inherit-environ VAR_LIST
Comma-separated list of environment variable
names that are inherited from the parent
environment when running benchmarking
subprocesses.
-p PYTHON, --python PYTHON
Python executable (default: use running
Python)
How to get stable benchmarks¶
- Run
python3 -m pyperf system tune
command - Compile Python using LTO (Link Time Optimization) and PGO (profile guided optimizations): use the pyperformance compile command with uses LTO and PGO by default
- See advices of the pyperf documentation: How to get reproductible benchmark results.
pyperformance virtual environment¶
To run benchmarks, pyperformance first creates a virtual environment. It installs
requirements with fixed versions to get a reproductible environment. The system
Python has unknown module installed with unknown versions, and can have
.pth
files run at Python startup which can modify Python behaviour or at
least slow down Python startup.
What is the goal of pyperformance¶
A benchmark is always written for a specific purpose. Depending how the benchmark is written and how the benchmark is run, the result can be different and so have a different meaning.
The pyperformance benchmark suite has multiple goals:
- Help to detect performance regression in a Python implementation
- Validate that an optimization change makes Python faster and don’t performance regressions, or only minor regressions
- Compare two implementations of Python, for example CPython and PyPy
- Showcase of Python performance which ideally would be representative of performances of applications running on production
Don’t disable GC nor ASLR¶
The pyperf module and pyperformance benchmarks are designed to produce
reproductible results, but not at the price of running benchmarks in a special
mode which would not be used to run applications in production. For these
reasons, the Python garbage collector, Python randomized hash function and
system ASLR (Address Space Layout Randomization) are not disabled.
Benchmarks don’t call gc.collect()
neither since CPython implements it with
stop-the-world
and so applications don’t call it to not kill performances.
Include outliers and spikes¶
Moreover, while the pyperf documentation explains how to reduce the random noise of the system and other applications, some benchmarks use the system and so can get different timing depending on the system workload, depending on I/O performances, etc. Outliers and temporary spikes in results are not automatically removed: values are summarized by computing the average (arithmetic mean) and standard deviation which “contains” these spikes, instead of using median and the median absolute deviation for example which to ignore outliers. It is deliberate choice since applications running in production are impacted by such temporary slowdown caused by various things like a garbage collection or a JIT compilation.
Warmups and steady state¶
A borderline issue are the benchmarks “warmups”. The first values of each worker process are always slower: 10% slower in the best case, it can be 1000% slower or more on PyPy. Right now (2017-04-14), pyperformance ignore first values considered as warmup until a benchmark reachs its “steady state”. The “steady state” can include temporary spikes every 5 values (ex: caused by the garbage collector), and it can still imply further JIT compiler optimizations but with a “low” impact on the average pyperformance.
To be clear “warmup” and “steady state” are a work-in-progress and a very complex topic, especially on PyPy and its JIT compiler.
Notes¶
Tool for comparing the performance of two Python implementations.
pyperformance will run Student’s two-tailed T test on the benchmark results at the 95% confidence level to indicate whether the observed difference is statistically significant.
Omitting the -b
option will result in the default group of benchmarks being
run Omitting -b
is the same as specifying -b default.
To run every benchmark pyperformance knows about, use -b all
. To see a full
list of all available benchmarks, use –help.
Negative benchmarks specifications are also supported: -b -2to3 will run every benchmark in the default group except for 2to3 (this is the same as -b default,-2to3). -b all,-django will run all benchmarks except the Django templates benchmark. Negative groups (e.g., -b -default) are not supported. Positive benchmarks are parsed before the negative benchmarks are subtracted.
If --track_memory
is passed, pyperformance will continuously sample the
benchmark’s memory usage. This currently only works on Linux 2.6.16 and higher
or Windows with PyWin32. Because --track_memory
introduces performance
jitter while collecting memory measurements, only memory usage is reported in
the final report.