Merge pull request #749 from pypa/docs-setup

docs: Rework the setup page to include information about local builds before going to CI
This commit is contained in:
Joe Rickerby
2021-09-09 15:41:52 +01:00
committed by GitHub
3 changed files with 230 additions and 101 deletions
+9 -3
View File
@@ -10,6 +10,14 @@ p {
line-height: 26px;
}
pre {
margin: 1em 0;
}
pre:first-child {
margin-top: 0;
}
code {
font-size: 85%;
background-color: rgba(27,31,35,.05);
@@ -22,6 +30,7 @@ code {
code, pre, tt {
font-family: SFMono-Regular, Consolas, Liberation Mono, Menlo, monospace;
line-height: 1.4;
}
pre code, pre code.hljs {
@@ -216,6 +225,3 @@ h1, h2, h3, h4, h5, h6 {
padding-right: 0;
margin-bottom: 0.5em;
}
.tabs.examples pre:first-child {
margin-top: 0;
}
+138 -91
View File
@@ -2,35 +2,23 @@
title: Tips and tricks
---
### Troubleshooting
If your wheel didn't compile, check the list below for some debugging tips.
- A mistake in your config. To quickly test your config without doing a git push and waiting for your code to build on CI, you can test the Linux build in a Docker container. On Mac or Linux, with Docker running, try `cibuildwheel --platform linux`. You'll have to bring your config into the current environment first.
- Missing dependency. You might need to install something on the build machine. You can do this in `.travis.yml`, `appveyor.yml`, or `.circleci/config.yml`, with apt-get, brew or choco. Given how the Linux build works, you'll need to use the [`CIBW_BEFORE_BUILD`](options.md#before-build) option.
- Windows: missing C feature. The Windows C compiler doesn't support C language features invented after 1990, so you'll have to backport your C code to C90. For me, this mostly involved putting my variable declarations at the top of the function like an animal.
- MacOS: calling cibuildwheel from a python3 script and getting a `ModuleNotFoundError`? Due to a (fixed) [bug](https://bugs.python.org/issue22490) in CPython, you'll need to [unset the `__PYVENV_LAUNCHER__` variable](https://github.com/pypa/cibuildwheel/issues/133#issuecomment-478288597) before activating a venv.
### Building Python 2.7 / PyPy2 wheels
See the [cibuildwheel version 1 docs](https://cibuildwheel.readthedocs.io/en/1.x/) for information about building Python 2.7 or PyPy2 wheels. There are lots of tricks and workaround there that are no longer required for Python 3 in cibuildwheel 2.
## Tips
### Linux builds on Docker
Linux wheels are built in the [`manylinux` docker images](https://github.com/pypa/manylinux) to provide binary compatible wheels on Linux, according to [PEP 571](https://www.python.org/dev/peps/pep-0571/). Because of this, when building with `cibuildwheel` on Linux, a few things should be taken into account:
- Programs and libraries are not installed on the Travis CI Ubuntu host, but rather should be installed inside of the Docker image (using `yum` for `manylinux2010` or `manylinux2014`, and `apt-get` for `manylinux_2_24`) or manually. The same goes for environment variables that are potentially needed to customize the wheel building. `cibuildwheel` supports this by providing the `CIBW_ENVIRONMENT` and `CIBW_BEFORE_BUILD` options to setup the build environment inside the running Docker image. See [the options docs](options.md#build-environment) for details on these options.
- Programs and libraries are not installed on the CI runner host, but rather should be installed inside of the Docker image - using `yum` for `manylinux2010` or `manylinux2014`, and `apt-get` for `manylinux_2_24`, or manually. The same goes for environment variables that are potentially needed to customize the wheel building.
- The project directory is mounted in the running Docker instance as `/project`, the output directory for the wheels as `/output`. In general, this is handled transparently by `cibuildwheel`. For a more finegrained level of control however, the root of the host file system is mounted as `/host`, allowing for example to access shared files, caches, etc. on the host file system. Note that this is not available on CircleCI due to their Docker policies.
`cibuildwheel` supports this by providing the [`CIBW_ENVIRONMENT`](options.md#environment) and [`CIBW_BEFORE_ALL`](options.md#before-all) options to setup the build environment inside the running Docker image.
- Alternative dockers images can be specified with the `CIBW_MANYLINUX_X86_64_IMAGE`, `CIBW_MANYLINUX_I686_IMAGE`, and `CIBW_MANYLINUX_PYPY_X86_64_IMAGE` options to allow for a custom, preconfigured build environment for the Linux builds. See [options](options.md#manylinux-image) for more details.
- The project directory is mounted in the running Docker instance as `/project`, the output directory for the wheels as `/output`. In general, this is handled transparently by `cibuildwheel`. For a more finegrained level of control however, the root of the host file system is mounted as `/host`, allowing for example to access shared files, caches, etc. on the host file system. Note that `/host` is not available on CircleCI due to their Docker policies.
- Alternative Docker images can be specified with the `CIBW_MANYLINUX_*_IMAGE` options to allow for a custom, preconfigured build environment for the Linux builds. See [options](options.md#manylinux-image) for more details.
### Building macOS wheels for Apple Silicon {: #apple-silicon}
`cibuildwheel` supports cross-compiling `universal2` and `arm64` wheels on `x86_64` runners. With the introduction of Apple Silicon, you now have several choices for wheels for Python 3.9+:
`cibuildwheel` supports cross-compiling `universal2` and `arm64` wheels on `x86_64` runners. With the introduction of Apple Silicon, you now have several choices for wheels for Python 3.8+:
#### `x86_64`
@@ -89,6 +77,7 @@ by default, and require opt-in by setting `CIBW_ARCHS_MACOS`.
Here's an example GitHub Actions workflow with a job that builds for Apple Silicon:
> .github/workflows/build_macos.yml
```yml
{% include "../examples/github-apple-silicon.yml" %}
```
@@ -114,33 +103,6 @@ Linux), and the other architectures are emulated automatically.
{% include "../examples/github-with-qemu.yml" %}
```
### Building Apple Silicon wheels on Intel {: #apple-silicon}
`cibuildwheel` supports cross-compiling `universal2` and `arm64` wheels on `x86_64` runners.
These wheels are not built by default, but can be enabled by setting the [`CIBW_ARCHS_MACOS` option](options.md#archs) to `x86_64 arm64 universal2`. Cross-compilation is provided by the Xcode toolchain.
!!! important
When cross-compiling on Intel, it is not possible to test `arm64` and the `arm64` part of a `universal2` wheel.
`cibuildwheel` will raise a warning to notify you of this - these warnings be be silenced by skipping testing on these platforms: `CIBW_TEST_SKIP: *_arm64 *_universal2:arm64`.
Hopefully, this is a temporary situation. Once we have widely available Apple Silicon CI runners, we can build and test `arm64` and `universal2` wheels more natively. That's why `universal2` wheels are not yet built by default, and require opt-in by setting `CIBW_ARCHS_MACOS`.
!!! note
Your runner image needs Xcode Command Line Tools 12.2 or later to build `universal2` and `arm64`.
So far, only CPython 3.9 supports `universal2` and `arm64` wheels.
Here's an example GitHub Actions workflow with a job that builds for Apple Silicon:
> .github/workflows/build_macos.yml
```yml
{% include "../examples/github-apple-silicon.yml" %}
```
### Building packages with optional C extensions
`cibuildwheel` defines the environment variable `CIBUILDWHEEL` to the value `1` allowing projects for which the C extension is optional to make it mandatory when building wheels.
@@ -155,51 +117,6 @@ myextension = Extension(
)
```
### 'No module named XYZ' errors after running cibuildwheel on macOS
`cibuildwheel` on Mac installs the distributions from Python.org system-wide during its operation. This is necessary, but it can cause some confusing errors after cibuildwheel has finished.
Consider the build script:
```bash
python3 -m pip install twine cibuildwheel
python3 -m cibuildwheel --output-dir wheelhouse
python3 -m twine upload wheelhouse/*.whl
# error: no module named 'twine'
```
This doesn't work because while `cibuildwheel` was running, it installed a few new versions of 'python3', so the `python3` run on line 3 isn't the same as the `python3` that ran on line 1.
Solutions to this vary, but the simplest is to install tools immediately before they're used:
```bash
python3 -m pip install cibuildwheel
python3 -m cibuildwheel --output-dir wheelhouse
python3 -m pip install twine
python3 -m twine upload wheelhouse/*.whl
```
### 'ImportError: DLL load failed: The specific module could not be found' error on Windows
Visual Studio and MSVC link the compiled binary wheels to the Microsoft Visual C++ Runtime. Normally, these are included with Python, but when compiling with a newer version of Visual Studio, it is possible users will run into problems on systems that do not have these runtime libraries installed. The solution is to ask users to download the corresponding Visual C++ Redistributable from the [Microsoft website](https://support.microsoft.com/en-us/help/2977003/the-latest-supported-visual-c-downloads) and install it. Since a Python installation normally includes these VC++ Redistributable files for [the version of the MSVC compiler used to compile Python](https://wiki.python.org/moin/WindowsCompilers), this is typically only a problem when compiling a Python C extension with a newer compiler.
Additionally, Visual Studio 2019 started linking to an even newer DLL, `VCRUNTIME140_1.dll`, besides the `VCRUNTIME140.dll` that is included with recent Python versions (starting from Python 3.5; see [here](https://wiki.python.org/moin/WindowsCompilers) for more details on the corresponding Visual Studio & MSVC versions used to compile the different Python versions). To avoid this extra dependency on `VCRUNTIME140_1.dll`, the [`/d2FH4-` flag](https://devblogs.microsoft.com/cppblog/making-cpp-exception-handling-smaller-x64/) can be added to the MSVC invocations (check out [this issue](https://github.com/pypa/cibuildwheel/issues/423) for details and references).
To add the `/d2FH4-` flag to a standard `setup.py` using `setuptools`, the `extra_compile_args` option can be used:
```python
ext_modules=[
Extension(
'c_module',
sources=['extension.c'],
extra_compile_args=['/d2FH4-'] if sys.platform == 'win32' else []
)
],
```
To investigate the dependencies of a C extension (i.e., the `.pyd` file, a DLL in disguise) on Windows, [Dependency Walker](http://www.dependencywalker.com/) is a great tool.
### Automatic updates {: #automatic-updates}
Selecting a moving target (like the latest release) is generally a bad idea in CI. If something breaks, you can't tell whether it was your code or an upstream update that caused the breakage, and in a worse-case scenario, it could occur during a release.
@@ -251,3 +168,133 @@ updates:
```
This will also try to update other pins in all requirement files, so be sure you want to do that. The only control you have over the files used is via the directory option.
### Alternatives to cibuildwheel options {: #cibw-options-alternatives}
cibuildwheel provides lots of opportunities to configure the build
environment. However, you might consider adding this build configuration into
the package itself - in general, this is preferred, because users of your
package 'sdist' will also benefit.
#### Missing build dependencies {: #cibw-options-alternatives-deps}
If your build needs Python dependencies, rather than using CIBW_BEFORE_BUILD, it's best to add these to the
[`build-system.requires`](https://www.python.org/dev/peps/pep-0518/#build-system-table)
section of your pyproject.toml. For example, if your project requires Cython
to build, your pyproject.toml might include a section like this:
```toml
[build-system]
requires = [
"setuptools>=42",
"wheel",
"Cython",
]
build-backend = "setuptools.build_meta"
```
#### Actions you need to perform before building
You might need to run some other commands before building, like running a
script that performs codegen or downloading some data that's not stored in
your source tree.
Rather than using CIBW_BEFORE_ALL or CIBW_BEFORE_BUILD, you could incorporate
these steps into your package's build process. For example, if you're using
setuptools, you can add steps to your package's `setup.py` using a structure
like this:
```python
import subprocess
import setuptools
import setuptools.command.build_py
class BuildPyCommand(setuptools.command.build_py.build_py):
"""Custom build command."""
def run(self):
# your custom build steps here
# e.g.
# subprocess.run(['python', 'scripts/my_custom_script.py'], check=True)
setuptools.command.build_py.build_py.run(self)
setuptools.setup(
cmdclass={
'build_py': BuildPyCommand,
},
# Usual setup() args.
# ...
)
```
#### Compiler flags
Your build might need some compiler flags to be set through environment variables.
Consider incorporating these into your package, for example, in `setup.py` using [`extra_compile_args` or
`extra_link_args`](https://docs.python.org/3/distutils/setupscript.html#other-options).
### Python 2.7 / PyPy2 wheels
See the [cibuildwheel version 1 docs](https://cibuildwheel.readthedocs.io/en/1.x/) for information about building Python 2.7 or PyPy2 wheels. There are lots of tricks and workaround there that are no longer required for Python 3 in cibuildwheel 2.
## Troubleshooting
If your wheel didn't compile, you might have a mistake in your config.
To quickly test your config without doing a git push and waiting for your code to build on CI, you can [test the Linux build in a local Docker container](setup.md#local).
### Missing dependencies
You might need to install something on the build machine. You can do this with apt/yum, brew or choco, using the [`CIBW_BEFORE_ALL`](options.md#before-all) option. Or, for a Python dependency, consider [adding it to pyproject.toml](#cibw-options-alternatives-deps).
### ModuleNotFoundError on macOS
Calling cibuildwheel from a python3 script and getting a `ModuleNotFoundError`? Due to a (fixed) [bug](https://bugs.python.org/issue22490) in CPython, you'll need to [unset the `__PYVENV_LAUNCHER__` variable](https://github.com/pypa/cibuildwheel/issues/133#issuecomment-478288597) before activating a venv.
### 'No module named XYZ' errors after running cibuildwheel on macOS
`cibuildwheel` on Mac installs the distributions from Python.org system-wide during its operation. This is necessary, but it can cause some confusing errors after cibuildwheel has finished.
Consider the build script:
```bash
python3 -m pip install twine cibuildwheel
python3 -m cibuildwheel --output-dir wheelhouse
python3 -m twine upload wheelhouse/*.whl
# error: no module named 'twine'
```
This doesn't work because while `cibuildwheel` was running, it installed a few new versions of 'python3', so the `python3` run on line 3 isn't the same as the `python3` that ran on line 1.
Solutions to this vary, but the simplest is to install tools immediately before they're used:
```bash
python3 -m pip install cibuildwheel
python3 -m cibuildwheel --output-dir wheelhouse
python3 -m pip install twine
python3 -m twine upload wheelhouse/*.whl
```
### 'ImportError: DLL load failed: The specific module could not be found' error on Windows
Visual Studio and MSVC link the compiled binary wheels to the Microsoft Visual C++ Runtime. Normally, these are included with Python, but when compiling with a newer version of Visual Studio, it is possible users will run into problems on systems that do not have these runtime libraries installed. The solution is to ask users to download the corresponding Visual C++ Redistributable from the [Microsoft website](https://support.microsoft.com/en-us/help/2977003/the-latest-supported-visual-c-downloads) and install it. Since a Python installation normally includes these VC++ Redistributable files for [the version of the MSVC compiler used to compile Python](https://wiki.python.org/moin/WindowsCompilers), this is typically only a problem when compiling a Python C extension with a newer compiler.
Additionally, Visual Studio 2019 started linking to an even newer DLL, `VCRUNTIME140_1.dll`, besides the `VCRUNTIME140.dll` that is included with recent Python versions (starting from Python 3.5; see [here](https://wiki.python.org/moin/WindowsCompilers) for more details on the corresponding Visual Studio & MSVC versions used to compile the different Python versions). To avoid this extra dependency on `VCRUNTIME140_1.dll`, the [`/d2FH4-` flag](https://devblogs.microsoft.com/cppblog/making-cpp-exception-handling-smaller-x64/) can be added to the MSVC invocations (check out [this issue](https://github.com/pypa/cibuildwheel/issues/423) for details and references).
To add the `/d2FH4-` flag to a standard `setup.py` using `setuptools`, the `extra_compile_args` option can be used:
```python
ext_modules=[
Extension(
'c_module',
sources=['extension.c'],
extra_compile_args=['/d2FH4-'] if sys.platform == 'win32' else []
)
],
```
To investigate the dependencies of a C extension (i.e., the `.pyd` file, a DLL in disguise) on Windows, [Dependency Walker](http://www.dependencywalker.com/) is a great tool.
+83 -7
View File
@@ -2,7 +2,79 @@
title: 'Setup'
---
# GitHub Actions [linux/mac/windows] {: #github-actions}
# Run cibuildwheel locally (optional) {: #local}
Before getting to CI setup, it can be convenient to test cibuildwheel
locally to quickly iterate and track down issues. If you've got
[Docker](https://www.docker.com/products/docker-desktop) installed on
your development machine, you can run a Linux build without even
touching CI.
!!! tip
You can run the Linux build on any platform. Even Windows can run
Linux containers these days, but there are a few hoops to jump
through. Check [this document](https://docs.microsoft.com/en-us/virtualization/windowscontainers/quick-start/quick-start-windows-10-linux)
for more info.
This is convenient as it's quicker to iterate, and because the builds are
happening in manylinux Docker containers, they're perfectly reproducible.
Install cibuildwheel and run a build like this:
```sh
pip install cibuildwheel
cibuildwheel --platform linux
```
Or, using [pipx](https://github.com/pypa/pipx):
```sh
pipx run cibuildwheel --platform linux
```
You should see the builds taking place. You can experiment with options using environment variables or pyproject.toml.
!!! tab "Environment variables"
cibuildwheel will read config from the environment. Syntax varies, depending on your shell:
> POSIX shell (Linux/macOS)
```sh
# run a command to set up the build system
export CIBW_BEFORE_ALL='apt install libpng-dev'
cibuildwheel --platform linux
```
> CMD (Windows)
```bat
set CIBW_BEFORE_ALL='apt install libpng-dev'
cibuildwheel --platform linux
```
!!! tab "pyproject.toml"
If you write your options into [`pyproject.toml`](options.md#configuration-file), you can work on your options locally, and they'll be automatically picked up when running in CI.
> pyproject.toml
```
[tool.cibuildwheel]
before-all = "apt install libpng-dev"
```
Then invoke cibuildwheel, like:
```console
cibuildwheel --platform linux
```
# Configure a CI service
## GitHub Actions [linux/mac/windows] {: #github-actions}
To build Linux, Mac, and Windows wheels using GitHub Actions, create a `.github/workflows/build_wheels.yml` file in your repo.
@@ -99,7 +171,7 @@ For more info on this file, check out the [docs](https://help.github.com/en/acti
[`examples/github-deploy.yml`](https://github.com/pypa/cibuildwheel/blob/main/examples/github-deploy.yml) extends this minimal example with a demonstration of how to automatically upload the built wheels to PyPI.
# Azure Pipelines [linux/mac/windows] {: #azure-pipelines}
## Azure Pipelines [linux/mac/windows] {: #azure-pipelines}
To build Linux, Mac, and Windows wheels on Azure Pipelines, create a `azure-pipelines.yml` file in your repo.
@@ -113,7 +185,7 @@ Commit this file, enable building of your repo on Azure Pipelines, and push.
Wheels will be stored for you and available through the Pipelines interface. For more info on this file, check out the [docs](https://docs.microsoft.com/en-us/azure/devops/pipelines/yaml-schema).
# Travis CI [linux/windows] {: #travis-ci}
## Travis CI [linux/windows] {: #travis-ci}
To build Linux and Windows wheels on Travis CI, create a `.travis.yml` file in your repo.
@@ -129,7 +201,7 @@ Then setup a deployment method by following the [Travis CI deployment docs](http
[`examples/travis-ci-deploy.yml`](https://github.com/pypa/cibuildwheel/blob/main/examples/travis-ci-deploy.yml) extends this minimal example with a demonstration of how to automatically upload the built wheels to PyPI.
# AppVeyor [linux/mac/windows] {: #appveyor}
## AppVeyor [linux/mac/windows] {: #appveyor}
To build Linux, Mac, and Windows wheels on AppVeyor, create an `appveyor.yml` file in your repo.
@@ -145,7 +217,7 @@ AppVeyor will store the built wheels for you - you can access them from the proj
For more info on this config file, check out the [docs](https://www.appveyor.com/docs/).
# CircleCI [linux/mac] {: #circleci}
## CircleCI [linux/mac] {: #circleci}
To build Linux and Mac wheels on CircleCI, create a `.circleci/config.yml` file in your repo,
@@ -162,7 +234,7 @@ Commit this file, enable building of your repo on CircleCI, and push.
CircleCI will store the built wheels for you - you can access them from the project console. Check out the CircleCI [docs](https://circleci.com/docs/2.0/configuration-reference/#section=configuration) for more info on this config file.
# Gitlab CI [linux] {: #gitlab-ci}
## Gitlab CI [linux] {: #gitlab-ci}
To build Linux wheels on Gitlab CI, create a `.gitlab-ci.yml` file in your repo,
@@ -178,9 +250,13 @@ Gitlab will store the built wheels for you - you can access them from the Pipeli
> ⚠️ Got an error? Check the [FAQ](faq.md).
# Next steps
Once you've got the wheel building successfully, you might want to set up [testing](options.md#test-command) or [automatic releases to PyPI](deliver-to-pypi.md#automatic-method).
<script>
document.addEventListener('DOMContentLoaded', function() {
$('.toctree-l2 a, .rst-content h1').each(function(i, el) {
$('a.toctree-l3, .rst-content h2').each(function(i, el) {
var text = $(el).text()
var match = text.match(/(.*) \[([a-z/]+)\]/);