It makes no sense to specify both:
- install_dir, which overrides the -Dincludedir= builtin option
- subdir, which suffixes the -Dincludedir= builtin option
We've always silently ignored the subdir in this case, which is really
surprising if someone actually passed it and expected it to do
something. We also confusingly didn't say anything in the documentation
about it.
Document that the options are incompatible, and explicitly check to see
if they are both passed -- if so, raise an error message pointing out
that only install_dir should be used.
Fixes#10046
JNI is a more apt name because it currently only supports the JNI. I
also believe that CMake uses the terminology JNI here as well.
JNI is currently the only way to interact with the JVM through native
code, but there is a project called "Project Panama" which aims to be
another way for native code to interact with the JVM.
GCC with optimization set to 0 does not actually result in no
optimizations, which can be annoying when trying to use a debugger like
gdb, and finding that your variable has been optimized out. We already
do this with clang, so gcc is a bit of an outlier here.
Unfortunately, checking for strings without context is exceedingly prone
to false positives, while missing anything that indirectly opens a file.
Python 3.10 has a feature to warn about this though -- and it uses a
runtime check which runs at the same time that the code fails to open
files in the broken Windows locale. Set this up automatically when
running the testsuite.
Sadly, Python's builtin feature to change the warning level, e.g. by
setting EncodingWarning to error at startup, is utterly broken if you
want to limit it to only certain modules. This is tracked in order to be
more efficiently ignored at https://bugs.python.org/issue34624 and
https://github.com/python/cpython/pull/9358
It is also very trigger happy and passing stuff around via environment
variable either messes with the testsuite, or with thirdparty programs
which are implemented in python *such as lots of gnome*, or perhaps
both.
Instead, add runtime code to meson itself, to add a hidden "feature".
In the application source code, running the 'warnings' module, you can
actually get the expected behavior that $PYTHONWARNINGS doesn't have. So
check for a magic testsuite variable every time meson starts up, and if
it does, then go ahead and initialize a warnings filter that makes
EncodingWarning fatal, but *only* when triggered via Meson and not
arbitrary subprocess scripts.
We didn't consider that it has arguments following it, so the resulting
compiler command line ended up with stuff like:
-L=-rpath-link -L=-L=/path/to/directory -L=more-args
and the directory for rpath-link got eaten up as a regular -L path to
the compiler rather than being passed as -Xlinker to the linker.
Then the -rpath-link would consume the next -Xlinker argument, end up
with the wrong rpath-link (may or may not cause link errors) and then
disappear arguments we need.
As an example failure mode, if the next argument is -soname this treats
the soname text as an input file, which probably does not exist if it
was generated in a subdirectory, and also because it can never be
successfully built in the first place -- though if it did, it would link
to itself which is very wrong.
After implementing a much more extensive Java native module than what
currently exists in the tests, I found shortcomings.
1. You need to be able to pass multiple Java files.
2. Meson needs more information to better track the generated native
headers.
3. Meson wasn't tracking the header files generated from inner classes.
This new function should fix all the issues the old function had with
room to grow should more functionality need to be added. What I
implemented here in this new function is essentially what I have done in
the Heterogeneous-Memory Storage Engine's Java bindings.
It's a dumb check which doesn't know the difference between python
functions and string data. Work around this by changing the message
output to not include an opening parenthesis, and changing other similar
strings the same way for consistency.
All these dependencies are impossible to find on versions of Meson older
than the ones they got custom lookups in, because they don't provide
pkg-config files. So they should not / cannot be reasonably used on
older versions, and it is easy to say "yep, these should emit a
FeatureNew for the versions they got added in".
It would be too difficult and probably a layering violation to give an
interpreter handle to the inner guts of every dependency. What we can do
instead is let every dependency track:
- the Feature checks it can produce,
- the version attribute for when it was implemented
while leaving the interpreter in charge of actually emitting them.
The point of a .use() function is because we don't always have the
information we need to use a feature check, so we allow creating the
feature and then storing it for later use. When implementing location
checks, although it is optional, actually using it violated that design.
Move the location out of the init method for FeatureCheck itself. It
remains compatible with all cases of .single_use(), but fix the rest up.
When the project instals GDB helper scripts, copy them into
meson-private directory with the right tree layout and write a .gdbinit
script to load them automatically.
Do not traceback when trying to update a wrap that isn't a [wrap-file],
just report the problem.
Do not traceback on perfectly valid WrapDB wraps that don't have a
patch_url because they have upstream meson.build, instead try to parse
the version from the source tarball filename.
There's no reason it needs a concrete list, any iterable is fine. This
does mean that we need to explicitly check for `str | bytes` in the
listify pass, as `str` and `bytes` are both iterable.
The default behavior of installing relative to prefix may be unexpected,
and is definitely wrong in many cases.
Give users control in order to specify that yes, they actually want to
install to a venv.
This is particularly useful for projects that use meson as a build
system for a python module, where *all* files shall be installed into
the python site-packages.
For a test to be displayed to stdout without buffering, it has to be
1) in verbose mode 2) not executed in parallel with others
3) not parsed, i.e. not TAP or Rust. Include these three
conditions in a new property of stdout and use it in the
three places where it matters: when printing the initial
and final delimiters of the output, and when deciding the
console mode to use.
Suggested-by: Jussi Pakkanen <jpakkane@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Currently it tries to run "cmd" by default in a MSYS2 bash.
Passing "bash" doesn't work since that defaults to WSL and
one has to pass an absolute path to bash instead, and even then
one misses out on the PS1 override.
In case $SHELL is set and the contained path exists prefer it even if
we are on Windows.
To make the PS1 override work we can't use Unix paths in Python since
we might be on Windows, so move the .bashrc check into the temporary
bash script itself.
This makes "meson devenv" work the same under MSYS2 as on Linux.
If --no-rebuild is used, the test program might not exist, spawning a
FileNotFoundError inside a long traceback rooted in subprocess.Popen
trying to run that test program. Current versions of Meson even say it's
an "unhandled python exception".
But we can do one better and actually tell the user what is wrong, why,
and what to do to fix it. And we can do so before getting waist deep in
partially running tests.
Fixes#10006
Verbose, non-parallel tests generally print their output as they run, rather than
after they finish. This however is not the case if stdout of the test is parsed
as is the case for TAP and Rust tests. In this case, the output during the run
is the list of the subtests, but stderr still has to be printed after the test
finishes.
Remi Thebault
Eli Schwartz
Tristan Partin
Dylan Baker
Xavier Claessens
Christian Wendt
Mark Bolhuis
Andrea Pappacoda
Paolo Bonzini
Christoph Reiter