| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
|
|
the use of /LINKWHOLE with MSVC and the behaviour of re-exporting symbols of individual objects in a static library.
|
|
|
|
This reverts commit a2def550c586aeba4269588e79a1a308467f2582.
This results in a 2k line file being unconditionally imported at
startup, and transitively loading two more (for a total cost of 2759
lines of code), and it's not clear it was ever needed to begin with...
|
|
This saves on a 1500-line import at startup and may be skipped entirely
if no compiled languages are used. In exchange, we move the
implementation to a new file that is imported instead.
Followup to commit ab20eb5bbc21ae855bcd211131132d2778602bcf.
|
|
Linkers that aren't actually ar, were refactored into a base class in
commit 253ff71e6b801786290b6ecb23aea61dcfeebeca, which however didn't
take into account that we were doing checks for this.
|
|
We should try to figure out what is a cross-submodule object and what
isn't.
|
|
We don't need a CMakeInterpreter until and unless we actually attempt to
use a cmake subproject via the cmake module.
Minus 10 files and 3679 lines of code imported at startup.
|
|
Simply store the module it is expected to be found in. That module then
appends to the packages dict, which guarantees mypy can verify that
it's got the right type -- there is no casting needed.
|
|
This lessens the amount of code imported at Meson startup by mapping
each dependency to a dictionary entry and using a programmable import to
dynamically return it.
Minus 16 files and 6399 lines of code imported at startup.
|
|
We expose detect.py as the mesonbuild.dependencies entrypoint and import
it upfront everywhere. But unless the `dependency()` function is
actually invoked, we don't need *any* of the private implementations for
this.
Avoid doing so until, as part of actual dependency lookup, we attempt
that specific dependency method. This avoids importing big modules if
`method:` is specified, and in most cases hopefully pkg-config works and
we can avoid importing the cmake implementation particularly.
Actually avoiding most of these imports requires more refactoring. But
even so, the garden path no longer needs to import the dub dependency
impl.
|
|
We need to extend the candidates the same way per method, but we handle
each method twice: once in explicit method checks, and once for auto. We
can just handle auto as a special list of methods, though.
|
|
Do not import it and expose it at the package scope, it's never used
elsewhere except inside the dub module.
|
|
We can check something's subtype using properties, without importing the
module up front and doing isinstance checks on specific subclasses of
the interface -- or worse, solving cyclic imports by doing the import
inside the function. ;)
|
|
msetup.py doesn't need to know the gory details of PkgConfigDependency,
or directly import it at program startup. It's also slightly wasteful to
generate a devenv for the -uninstalled directory when a project doesn't
even, in the end, use the pkgconfig module anyway.
|
|
|
|
We have the same platform startup logic here for all platforms and I do
not believe it's important to test this on the slow CI machines.
|
|
|
|
|
|
|
|
And avoid passing variables around several functions just to finally get
where they need to be. These function signatures were kind of ugly...
Also the use of dataclasses makes a big chunk of this file now typed
properly.
|
|
Most of the dist handling is either git-specific or hg-specific.
Tangling it all together makes it much harder to analyze what is going
on.
|
|
|
|
|
|
Setting options or clearing cache is also an error without a valid build
directory.
|
|
Fixes: #11899
|
|
I noticed when building a project that uses a proc macro that Meson
passed -C prefer-dynamic for the executable, and not the proc macro,
while cargo passed -C prefer-dynamic for the proc macro, but not for
the executable. Meson's behavior broke setting -C panic=abort on the
executable.
As far as we can tell, because we explicitly pass each library path to
rustc, the only thing -C prefer-dynamic affects in Meson is how the
standard libraries are linked. Generally, one does not want the
standard libraries to be dynamically linked, because if the Rust
compiler is ever updated, anything linked against the old standard
libraries will likely break, due to the lack of a stable Rust ABI.
Therefore, I've reorganised Meson's behavior around the principle that
the standard libraries should only be dynamically linked when Rust
dynamic linking has already been opted into in some other way. The
details of how this manifests are now explained in the documentation.
|
|
The object is not used only for tests, but also for `custom_target` etc.
|
|
This is a side effect of requiring Python >= 3.7 which itself guarantees
dictionary order. This is now a Meson language guarantee as well which
is required for passing default_options as dict and is generally
expected by users.
|
|
|
|
|
|
We went from 192 mypy errors down to 144.
|
|
This function has a pretty unique name, and a simple grep shows that it
is only ever called as:
```
add_comment(PbxComment('...........'))
```
It doesn't need to include logic such as handling str. Moreover it looks
like that handling was broken anyway... it handled the case where
comment is type str, by constructing a new PbxComment(str) instead of
PbxComment(comment), a condition that cannot ever be valid (and crashed
due to other assertions).
Fixes:
mesonbuild/backend/xcodebackend.py:148:42: error: Argument 1 to "PbxComment" has incompatible type "type[str]"; expected "str" [arg-type]
|
|
|
|
I slapped the obvious type annotations onto a bunch of places, and got
the errors down from 668 to 112.
|
|
|
|
We are going to truncate it down below with int() anyway, no need to
carry extra precision. And mypy complains if the type changes between
float and int.
|
|
|
|
The new splitlines method on str is intended to replace usage of
fs.read('whatever').strip().split('\n').
The problem with the .strip().split() approach is that it doesn't have a
way to represent empty lists (an empty string becomes a list with one
empty string, not an empty list), and it doesn't handle Windows-style
line endings.
|
|
RUST_LD was deprecated in Meson 0.54 in favor of RUSTC_LD.
|
|
|
|
|
|
When meson is installed as editable, setuptools adds
`__editable___meson_1_0_99_finder` to the list. This contains the string
"meson" which isn't really accurate to what we want, which is modules
that are part of the `mesonbuild.*` namespace.
|
|
|
|
|
|
Clang, clang-cl, and MSVC all rely on RC.EXE to build resource files.
Fixes #11845
|
|
bug
Fixes: #11382
|
|
Although they could be moved to annotations, the truth is that they are
unneeded because they get inherited from the parent class.
|
|
We now warn everyone equally that totally ignored sources are bad and
should not be used.
|
|
an int only accepts operations on other ints, just like other primitive
types only accept operations on values of the same type.
But due to using isinstance in baseobjects "operator_call", an int
primitive allowed operations on a bool, even though reversing the
operator and having a bool perform operations on an int, would fail with
a type error.
Really, we should fail with a type error in both directions. But for
stability reasons, make this a loud warning and break --fatal-meson-warnings
builds.
|
