When projects do not specify a minimum meson version, we used to avoid
giving them the benefit of the Feature checks framework. Instead:
- warn for features that were added after the most recent semver bump,
since they aren't portable to the range of versions people might use
these days
- warn for features that were deprecated before the upcoming semver
bump, i.e. all deprecated features, since they aren't portable to
upcoming semver-compatible versions people might be imminently upgrading
to
Previously, setting `MESON_TESTTHREADS` to a number lower than 1
resulted in unexpected behavior. This commit introduces test for
negative value (with fallback to 1), and fallback to core count in case
it is set to 0.
It improves experience in job-matrix type of CI workflows, where some
jobs within the matrix require single job execution, whereas others can
default to taking core count as the job count.
Signed-off-by: Marek Pikuła <m.pikula@partner.samsung.com>
Although it's not especially common, there are certainly cases where it's
useful to pass the path to an external program to a test program.
Fixes: https://github.com/mesonbuild/meson/issues/3552
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We've documented these lists as being `List[Path]`, but then we have the
potential to insert a None into them via the `rel_path()` function,
which can return `None` in some cases. Currently we fix some (but not
all) of these later, but we should actually remove them from the list
before we assign, so that it's actually a `List[Path]` at all times.
While we're here I've simplified the logic a bit.
Closes: #13551
It was possible (with some frequency) for the clang-tidy/format target
to continue starting new subprocesses after a CTRL-C, because we were
not canceling the already queued tasks and waiting for all of them.
This makes a best-effort attempt to cancel all further subprocesses. It
is not 100%, because there is a race that is hard to avoid (without
major restructuring, at least): new subprocesses may be started after
KeyboardInterrupt (or any other exception) is raised but before we get
to the cancellation.
When the race happens, the calling ninja may exit before Meson exits,
causing some output (from clang-tidy/format and Meson's traceback) to be
printed after returning to the shell prompt. But this is an improvement
over potentially launching all the remaining tasks after having returned
to the shell, which is what used to happen rather often.
In practice, it appears that we cleanly exit with a pretty high
probability unless CTRL-C is hit very early after starting (presumably
before the thread pool has launched subprocesses on all its threads).
The following is valid meson:
```meson
a = '''This string can't be simplified'''
```
which cannot be simplified because of the `'` in it, as
```meson
a = 'This string can't be simplified'
```
Is invalid.
Potentially we could convert that with escapes, but it seems reasonable
to me to leave this, since it may be desirable to not have lots of
escapes in a string. `'''I can't believe it's her's!'''` is much more
readable than `'I can\'t believe it\'s her\'s!'`, for example.
Closes: #13564
Fixes#13508
- Fix indentation of comments in arrays
- Fix indentation of comments in dicts
- Fix indentation of comments in if clauses
- Fix indentation of comments in foreach clauses
Which happens when a .editorconfig is in a subdirectory, not the root.
In this case we need Set the fallback value to `False`, which is what
editorconfig expects.
Closes: #13568
Meson accidentally strips '-isystem' from C build args like ['-isystem',
'/path/to/headers'] if the compiler includes the current working directory
in the header search paths. The root cause is that '-isystem'[8:] evaluates
to an empty string and os.path.realpath('') returns the absolute path to
the current working directory, causing meson to think that '-isystem'
specifies a default include path.
Different compiler versions varies whether the current working directory is
in its search paths. For example, on Ubuntu 21.04:
# gcc -xc -v -E -
gcc version 10.3.0 (Ubuntu 10.3.0-1ubuntu1)
#include "..." search starts here:
#include <...> search starts here:
.
/usr/lib/gcc/x86_64-linux-gnu/10/include
/usr/local/include
/usr/include/x86_64-linux-gnu
/usr/include
End of search list.
While on Ubuntu 24.04:
# gcc -xc -v -E -
gcc version 13.2.0 (Ubuntu 13.2.0-23ubuntu4)
...
#include "..." search starts here:
#include <...> search starts here:
/usr/lib/gcc/x86_64-linux-gnu/13/include
/usr/local/include
/usr/include/x86_64-linux-gnu
/usr/include
End of search list.
Do not check the '-isystem=' and '-isystem/path/to/header' cases when the
option is '-isystem' but the path that follows is not a default search
path.
Signed-off-by: Junjie Mao <junjie.mao@hotmail.com>
- check version of DUB for compatibility with Meson
- use "cacheArtifactPath" to locate DUB libraries in the cache
- propose `dub build --deep` command to Meson users for missing DUB
packages
Depending on DUB version, it will look either in the old cache structure
or use this new `dub describe` entry.
Updating the log message generated when installing a symbolic link. The
new message format may be a bit more clear on the link being created.
Signed-off-by: James Knight <james.d.knight@live.com>
The command we use to heuristically parse whether it is dirty by
interpreting prose descriptions of the repository state, is vulnerable
to changes in locale resulting in failing to match the English word that
means it is clean.
Unfortunately, I am no mercurial expert so I am unaware if mercurial
supports scripting, like git does. Perhaps the technology simply does
not exist. A quick attempt at searching for the answer turned nothing
up. It appears that #4278 had good cause indeed for using this prose
parsing command.
So, we simply sanitize the environment due to lack of any better idea.
Bug: https://bugs.gentoo.org/936670
In particular, it is only ever constructed once, and all but two of its
arguments are always passed. So there's no reason or valid justification
to initialize them as None or pretend that you have to check them for
None values. Nor in fact does the command list need to be a
default_factory.
There are also two instance attributes which it is never instantiated
with, but which are tracked after the fact. Don't add this to the
signature for `__init__` as that is misleading.
create_target_linker_introspection is only ever called from one place,
which passes in CompilerArgs. This was implemented in commit
5eb55075ba which performed this conversion
as a preventative measure, since its type was not obvious (and thereby
modified the *type* of the variable in place).
We have a function that wraps two others because it first checks whether
the input is static or shared. By the same token, it has to return types
valid for either. We already *know* that we are a shared library, so we
can and should use the real function directly, which is both a
micro-optimization of function call overhead, and fixes a mypy
"union-attr" error.
It's always true because the func is always a real, truthy func object.
In the cmake case, the logic seems to be broken because if a path is not
a file, then that includes the case where it does not exist. It also
clearly meant "or" instead of "and".
What actually ended up happening was that this check never fired at all.
Because "if not func and not ..." would always fail, because "not func"
is always false. Maybe we don't need this logic at all...
On Windows, the output read from the stream has '\r\n', which in .txt,
.json and console logger (when captured to a file) translates to '\n\n'.
This results in every log line being separated by an empty line.
This is a bit of a hack, since the rule is added outside of the
`__init__` method, and that's probably bad. But at least we can get some
additional help by telling type checkers what it will be
URLError is a subclass of OSError and intermittent server errors can
manifest as OSError while reading instead of a URLError while
establishing a connection, which will cause the fallback url to be
ignored:
```
Looking for a fallback subproject for the dependency gudev-1.0
Downloading libgudev source from https://gitlab.gnome.org/GNOME/libgudev/-/archive/238/libgudev-238.tar.bz2
HTTP Error 404: Not Found
WARNING: failed to download with error: could not get https://gitlab.gnome.org/GNOME/libgudev/-/archive/238/libgudev-238.tar.bz2 is the internet available?. Trying after a delay...
HTTP Error 404: Not Found
WARNING: failed to download with error: could not get https://gitlab.gnome.org/GNOME/libgudev/-/archive/238/libgudev-238.tar.bz2 is the internet available?. Trying after a delay...
HTTP Error 404: Not Found
WARNING: failed to download with error: could not get https://gitlab.gnome.org/GNOME/libgudev/-/archive/238/libgudev-238.tar.bz2 is the internet available?. Trying after a delay...
WARNING: failed to download with error: The read operation timed out. Trying after a delay...
WARNING: failed to download with error: The read operation timed out. Trying after a delay...
ERROR: Unhandled python OSError. This is probably not a Meson bug, but an issue with your build environment.
```
At an OS level, Unix-like OSes usually have very large or even
unlimited sized command line limits. In practice, however, many
applications do not handle this (intentionally or otherwise). Notably
Wine has the same limits Windows does, 32,768 characters. Because we
previously double counted most characters, we papered over most
situations that we would need an RSP file on Unix-like OSes with Wine.
To fix this issue I have set the command line limit to 32k, this is
still a massive command line to pass without an RSP file, and will only
cause the use of an RSP file where it is not strictly necessary in a
small number of cases, but will fix Wine applications. Projects who wish
to not use an RSP file can still set the MESON_RSP_THRESHOLD environment
variable to a very large number instead.
Fixes: #13414
Fixes: cf0fecfce ("backend/ninja: Fix bug in NinjaRule.length_estimate")
This causes us to not count the spaces between arguments, thereby
undercounting the number of elements. This is extra important because we
previously double counted all actual characters, covering this issue up.
Fixes: cf0fecfce ("backend/ninja: Fix bug in NinjaRule.length_estimate")
Eli Schwartz
Marek Pikuła
David Gibson
Dylan Baker
L. E. Segovia
Mark A. Tsuchida
Jussi Pakkanen
slimeuniverse
Wolfgang Walther
Charles Brunet
Xavier Claessens
Junjie Mao
Benoit Pierre
Jonathan Schleifer
Kacper Michajłow
Remi Thebault
Eli Schwartz
James Knight
Nirbheek Chauhan
H. Vetinari
Mateusz Patyk
Jannik Glückert
Monson Shao