New people that want to use Meson for building Python extensions most
probably will read Python module docs first. Direct them to meson-python
and suggest to set `python.install_env=auto`.
Instead of invoking javac for every .java file, pass all of the sources
for a jar target to a single javac invocation. This massively improves
first compilation time and doesn't meaningfully affect incremental builds
(it can even be faster in some cases).
The old approach also had issues where files would not always get recompiled
even though they should, necessitating a clean rebuild in order to see changes
reflected in the build output.
Multiple invocations seem to only make sense if:
- issues with files not getting flagged for rebuild are investigated and fixed
- something like the javaserver buildtool from openjdk sources is used
instead of directly spawning javac processes
- the amount of java files per jar is so large that it is faster to compile
several files one by one than to compile all the files at once (batching may
still make sense to get a reasonable balance)
From python 3.11 [1]:
> The entry in the netrc file no longer needs to contain all tokens. The missing
> tokens' value default to an empty string. All the tokens and their values now
> can contain arbitrary characters, like whitespace and non-ASCII characters.
> If the login name is anonymous, it won't trigger the security check.
[1] 15409c720b
It started failing CI as soon as the default shifted to 1.78. Something
is broken and it prevents running stable CI. Tracking issue opened.
We pin the version because that is the same way we handle CI for linux
-- with the exception that Linux CI can upgrade itself as soon as we
fix issues causing the CI Image Builder to jam itself, whereas
unfortunately Windows will need to be manually unpinned, but such is
life as a Windows supporter.
Bug: #13236
Some settings require "objectVersion" to be set to a certain value or
Xcode will not use it. To fix this, we set it to the highest possible
value, determined by the detected version of Xcode. We also set
"compatibilityVersion", but mainly so it lines up with "objectVersion".
At the same time, we should not be generating Xcode 3.2-compatible
projects by default anyway.
Arch profile.d scripts were converted to use an appending function that
disappears when /etc/profile exits, and overall are simply not suitable
-- any more -- for sourcing individually.
(I will freely admit I'm not really sure what the overall goal of
refraining from sourcing /etc/profile itself is. Arguably it's kind of
misuse of the profile...)
This silently broke the cuda tests, which never ran because the cuda
compiler was not detected as available.
While we are at it, I guess we can convert gentoo to use the same trick
of appending it in install.sh
`configure_file` kwarg `copy` runs at configure time, whereas
`fs.copyfile` runs at build time. Both have use cases, so this
undeprecates the `configure_file` version.
Fixes: #12792
Clang is such a great compiler! Not.
Compilers have enhanced diagnostics for some kinds of "well known"
undeclared identifiers, telling you exactly which header you might have
forgotten to include. The reason why clang needs an option GCC doesn't
need is because clang's fixit suggestions, unlike GCC's actually
*changes the type of the error*, as a result of a fixit of all things.
After the fixit suggestion grants this error the right to be ignored,
we start having to add clang-specific options.
Follow-up to https://github.com/mesonbuild/meson/issues/9140
Upstream clang bug, which appears to be going nowhere:
https://github.com/llvm/llvm-project/issues/33905
This reverts commit aee941559c.
The commit being reverted breaks compilation of a major Meson consumer
(Mesa). As a result, various distros are either pinning to <1.4.0 (before
the commit) or performing this same revert downstream.
Fixing a regression takes priority, so let's revert.
Fixes: https://github.com/mesonbuild/meson/issues/12973
* Previously, cuda would just plainly prepend `-l` to the libname.
* By relying on the host compiler to find libraries, we now get
more subtle failures, such as CUDA modules not being found
anymore.
* We need to simplify these CUDA modules when nvcc is used for
linking, since this may have side-effects from the cuda toolchain.
Closes: #13240
We do:
```
if is_thing:
assignme = value
if_is_thing:
...
else:
assignme = []
```
It is always defined on both sides, but there was no particular reason
we had to assign it in the later copy. pylint reported it as a false
positive, and it may prove confusing in general, and it's harmless to
move, so do so.
It does no control flow analysis, and upgrading to pylint 3.2.0 produced
many incorrect warnings.
Also ignore contextmanager-generator-missing-cleanup for now. It has FP
when there is no code after a yield (and thus no cleanup needs to be
handled), which is what we do. Currently under discussion upstream:
https://github.com/pylint-dev/pylint/issues/9625
We cover every case as if/elif/elif. mypy can handle this fine, but
pylint doesn't do control flow or type checking and thinks in the
missing else case, the variable might not be defined.
For mypy as well, doing this instance check is unnecessary as it can be
inferred. So just micro-optimize the check and allow pylint to safely
analyze the logic.
Emscripten version numbers are unrelated to Clang version numbers,
so it is necessary to change the version checks for `c_std=c17` & co.
Without that, no project that defaults to C17 or newer will build with
Emscripten.
See https://github.com/pyodide/pyodide/discussions/4762 for more
context. Also note that this bug caused defaulting to C17 in
scikit-learn to be reverted (scikit-learn#29015), and it may be a
problem for SciPy 1.14.0 too since that release will upgrade from C99
to C17.
Co-authored-by: Loic Esteve <loic.esteve@ymail.com>
When running our integration tests in systemd we depend on each test
having a unique name. This is always the case unless --repeat is used,
in which case multiple tests with the same name run concurrently which
causes issues when allocating resources that use the test name as the
identifier.
Let's set MESON_TEST_ITERATION to the current iteration of the test so
we can use $TEST_NAME-$TEST_ITERATION as our test identifiers which will
avoid these issues.
Emil Melnikov
Mis012
Jussi Pakkanen
Ruben Gonzalez
Wolfgang Walther
Eli Schwartz
Albert Tang
Filipe Laíns
Emil Melnikov
David Seifert
Bruno
Marc Durdin
Sam James
Michael Gene Brockus
Axel Ricard
Ralf Gommers
Stas Sergeev
sammyj85
Daan De Meyer