When only a single input file shows up in an arglist, it makes no sense
to inject `-W,--start-group -lone -Wl,--end-group`, since there is
nothing being grouped together. It's just longer command lines for
nothing.
For instance, on Windows, if the terminal is in cp65001, the subprocess
output is not encoded correctly and it results in error when running
unit test.
Python provides some nifty tools for mocking, without relying on
altering running code. We should use these to simplify the actual run
paths and move the complicated logic into tests.
Meson will implicit rpaths when *.so/*.dll/etc. files are injected onto
the link line from pkg-config and (now) cmake dependencies.
Extend the "prebuilt shared" tests to test that these are added.
This replaces all of the Apache blurbs at the start of each file with an
`# SPDX-License-Identifier: Apache-2.0` string. It also fixes existing
uses to be consistent in capitalization, and to be placed above any
copyright notices.
This removes nearly 3000 lines of boilerplate from the project (only
python files), which no developer cares to look at.
SPDX is in common use, particularly in the Linux kernel, and is the
recommended format for Meson's own `project(license: )` field
The unit test infrastructure hardcodes the architecture as x86_64. On
macos, the test_pkgconfig_parse_libs() test creates a few libraries
for testing using clang. This leads to the test failing on arm based
macs, as darwin_get_object_archs() will skip over these libraries,
which then leads to the test failing because of using -l instead of
referencing the libraries by their full path.
I am not at all sure this is the best approach.
I am also somewhat confused why nobody else has encountered this? CI
apparently just runs on x86_64 macs?
CC: @tristan957
The former has rust dependencies, which lead to max capping on Cygwin
since there is no rust compiler there. But it turns out there are other
disadvantages of jsonschema:
- it involves installing 5 wheels, instead of just 1
- it is much slower
To give some perspective to the latter issue, this is what it looks like
when I test with jsonschema:
```
===== 1 passed, 509 deselected in 3.07s =====
Total time: 3.341 seconds
```
And here's what it looks like when I test with fastjsonschema:
```
===== 1 passed, 509 deselected, 1 warning in 0.28s =====
Total time: 0.550 seconds
```
I cannot think of a good reason to use the former. Although in order to
work on old CI images, we'll support it as a fallback mechanism
This was originally added for vala only, with the rationale that vala
generates bad code that has warnings. Unfortunately, the rationale was
fatally flawed. The compiler warns about a number of things, which the
user can control depending on their code (or their code generator's
code), but some of those things are absolutely critical to warn about.
In particular, GCC 14 and clang 17 are updating their defaults to warn
-- and error by default for -- invalid C code that breaks the standard,
but has been silently accepted for over 20 years "because lots of people
do it". The code in question is UB, and compilers will generate faulty
machine code that behaves erroneously and probably has a mass of CVEs
waiting to happen.
Compiler warnings are NOT safe to just... universally turn off. Compiler
warnings could be either:
- coding style lints
- threatening statements that the code is factually and behaviorally wrong
There is no magic bullet to ignore the former while respecting the
latter. And the very last thing we should ever do is pass `-w`, since
that causes ALL warnings to be disabled, even the manually added
`-Werror=XXX`.
If vala generated code creates warnings, then the vala compiler can
decrease the log level by generating better code, or by adding warning
suppression pragmas for *specific* issues, such as unused functions.
We try to backtrack through the filesystem to find the correct directory
to build in, and suggest this as a possible diagnostic. However, our
current heuristic relies on parsing the raw file with string matching to
see if it starts with `project(`, and this may or may not actually work.
Instead, do a bit of recursion and parse each candidate with mparser,
then check if the first node of *that* file is a project() function.
This makes us resilient to a common case: where the root meson.build is
entirely valid, but, the first line is a comment containing e.g. SPDX
license headers and a simple string comparison simply does not cut it.
Fixes the bad error message from #12441, which was supposed to provide
more guidance but did not.
This does require hacking up the test pretty badly, since we need to not
ever pass GCC these invalid values. But it's preferable to writing
another project test I think.
Co-Authored-by: Nirbheek Chauhan <nirbheek@centricular.com>
Which allow passing arguments specifically to the static or shared
libraries.
For design, this is all handled in the interpreter, by the build layer
the arguments are combined into the existing fields. This limits changes
required in the mid and backend layers
Since the previous commit allows for more scenarios with name
collisions, it makes sense to expand the compile command so that it can
also take into account suffixes. i.e. meson compile -C build foo.exe can
now work if the executable has an exe suffix along with being named foo.
When checking target names, meson explictly forbids having multiple
targets with the same name. This is good, but it is strict and it is
impossible to have targets with the same basename and differing suffixes
(e.g. foo and foo.bin) in the same directory. Allow this for executables
by including the suffix (if it exists) in the interal target id. So foo
would be foo@exe and foo.bin would be foo.bin@exe.
The unittest case for `clang-tidy-fix` checks if the whole project is in
git or not, and skips if not.
Fix this by creating a temporary git repo, copy the test files and run
the tests, following how `clang-format` does.
It also reverts some help code introduced in the previous test.
Tested: Verify the test case passes.
Signed-off-by: Lei YU <yulei.sh@bytedance.com>
It should *be* a version comparison. We are guaranteed to get a
two-element version number, which also parses as a float but a float
doesn't correctly handle version sorting when the second component
differs in number of digits.
The standard way to handle this is by comparing tuples such that each
component is an integer. Do so here.
Fixes#12195
Co-authored-by: George Koehler <xkernigh@netscape.net>
(for unittests)
Add the `clang-tidy-fix` target to apply clang-tidy fixes to the source
code.
This is done by calling `run-clang-tidy` with `-fix` argument.
Add a test case to run `clang-tidy-fix` and verify the file is changed.
Signed-off-by: Lei YU <yulei.sh@bytedance.com>
While at it, make more methods private by storing the version found on
the instance. That avoids having to call check_pkgconfig() as static
method from unittests.
This also makes it more consistent with get_pkgconfig_variable() which
always return empty value instead of failing when the variable does not
exist. Linking that to self.required makes no sense and was never
documented any way.
Allow packagecache to contain already extracted directory to match what
some distro does with Cargo source packages in /usr/share/cargo/registry.
Note that there is no need to lock the cache directory because we
download into a temporary name and atomically rename afterward. It means
we could be downloading the same file twice, but at least integrity is
guaranteed.
Fixes: #12211
`meson setup -Dfoo=bar builddir` command was returning success ignoring
new option values.
This now also update options. It is useful because it means
`meson setup -Dfoo=bar builddir && ninja -C builddir` works regardless
whether builddir already exists or not, and when done in a script,
changing options in the script will automatically trigger a reconfigure
if needed. This was already possible by always passing --reconfigure
argument, but that triggers a reconfigure even when options did not
change.
- allow defines with leading whitespace
- always do replacement for cmakedefine
- output boolean value for cmakedefine01
- correct unittests for cmakedefine
- add cmakedefine specific unittests
When an installed static library A links to an internal static library B
built using a custom_target(), raise an error instead of a warning. This
is because to be usable, A needs to contain B which would require to
extract the archive to get its objects files.
This used to work, but was printing a warning and was installing a
broken static library, because we used to overlink in many cases, and
that got fixed in Meson 1.2.0. It now fails at link time with symbols
from the custom target not being defined. It's better to turn the
warning into a hard error at configure time.
While at it, noticed this situation can happen for any internal custom
or rust target we link to, recursively.
get_internal_static_libraries_recurse() could be called on CustomTarget
objects which do not implement it, and even if we did not call that
method, it would still fail when trying to call extract_all_objects() on
it.
Fixes: #12006
Projects that prefer GNU C but can fallback to ISO C can now set for
example `default_options: 'c_std=gnu11,c11'` and it will use gnu11 when
available, fallback to c11 otherwise. It is an error only if none of the
values are supported by the current compiler.
This allows to deprecate gnuXX values from MSVC compiler, that means
that `default_options: 'c_std=gnu11'` will now print warning with MSVC
but still fallback to 'c11' value. No warning is printed if at least one
of the values is valid, i.e. `default_options: 'c_std=gnu11,c11'`.
In the future that deprecation warning will become an hard error because
`c_std=gnu11` should mean GNU is required, for projects that cannot be
built with MSVC for example.
In some cases the desired result can be different if there are no
compilers at all. The expectations here are based on there being at
least one compiler, so reflect that by providing one; a later test
enhancement can cover the case where there are no compilers provided.
As a result of the mock any_compiler_has_define(), all that matters
will be the distinction between an empty or non-empty dict: the compiler
object itself is unused.
Signed-off-by: Simon McVittie <smcv@debian.org>
We support this in a machine file:
```
[binaries]
pkgconfig = 'pkg-config'
pkg-config = 'pkg-config'
```
and you can use either one, because internally we look up both. If you
only set *one* of them, this plays awkwardly with setting $PKG_CONFIG,
since we don't know which one you set in the machine file and the
*other* one will be initialized from the environment instead.
In commit 22df45a319 we changed program
lookup of config-tool style dependencies to use the regular tool names
and only fall back on the strange internal names. This affected the
pkg-config class too.
The result is that instead of preferring `pkgconfig =` followed by
$PKG_CONFIG followed by `pkg-config =`, we inverted the lookup order.
This is a good idea anyway, because now it behaves consistently with
`find_program('pkg-config')`.
Unfortunately, we documented the wrong name in a bunch of places, and
also used the wrong name in various testsuite bits, which meant that if
you set $PKG_CONFIG and then ran the testsuite, it would fail.
Correct these references, because they are buggy.
One test case expected to find_program() a native copy for convenience
of testing against natively installed glib. Force it to resolve a native
copy.
Charles Brunet
Jussi Pakkanen
Eli Schwartz
Dylan Baker
Jonathon Anderson
Alyssa Ross
Andres Freund
Xavier Claessens
Benjamin Redelings
Dudemanguy
Tristan Partin
Lei YU
Christoph Reiter
Benoit Pierre
Eli Schwartz
Jan200101
Simon McVittie