Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
|
|
|
|
|
|
So that we can actually use it anyplace that an OptionDictType could be
used. I've also done a bit optimizing/simplifying of the implementation.
This is needed for cuda, as it returns an OptionOverrideProxy where we
ask for an OptionDicType
|
|
|
|
|
|
So that every subclass doesn't have to reimplement it. Especially since
the Gnu implementation moved out of the CCompiler and into the
GnuLikeCompiler mixin
|
|
Only the GnuLikeCompiler overrides this, and it's implemented multiple
places
|
|
|
|
|
|
The get_variable method is able to do everything they do and more,
making it generally more useful. Let's tell people to stop using the old
ones.
|
|
|
|
|
|
Intel compilers on Windows (and the Microsoft C++ compiler) use /Od to disable optimisation, not /O0.
|
|
Extend the Curses dependency to use config tools and hand rolled search
|
|
with msys ncurses-config returns a unix style path (currently, though
it's been fixed upstream), which the compilers don't understand, so we
can't do that. Additionally, while the system search does work, there's
missing include directories that need to be added.
|
|
On win32 there is pdcurses, so we detect it first, because python
depends on ncursesw, so if we don't want to use ncursesw, we should make
sure pdcurses detect before ncursesw
|
|
|
|
This is mostly important for the system dependency where we need to roll
the version check ourselves.
|
|
That calls find_library and has_header in conjunction to look for curses
implementations that are baked into the system without any other find
method.
|
|
has_header returns a tuple of (found: bool, cached: bool), so `if
has_header` will always return true because the tuple is non-empty. We
need to check if the found value is true or not.
|
|
These are mostly duplicated with pkg-config, but maybe someone has one
but not another, and they're easy to turn on with the
ConfigToolDependency.
|
|
look for (in order): ncursesw, ncurses, curses.
|
|
|
|
|
|
dependencies/hdf5: Convert to a dependency_factory
|
|
Improve the output for meson wrapped commands
|
|
- Fixed using debug and optimization built-in options in MSVC.
- Fixed that VS backend does not create pdb files in release mode.
VS implicitly adds the debug fields if left out.
- Fix that it is possible to add debug info with ninja backend with
optimizations.
|
|
Deprecate meson.build_root() and meson.source_root()
|
|
Add a note about the portability of using shell constructs in the
custom_target() command.
|
|
typing: CMake module
|
|
strangely enough the syntax is:
/foo/* (files in foo, but not in foo/dir/)
/foo/ (files in foo and any subfolder of foo (recursively)
|
|
|
|
|
|
When building with vs2019 (not ninja), a path length error will be thrown
if the path to a resource file is even remotely deep within the tree.
This is largely because the target name includes the string "Windows
resource for file 'full path'", which is then expanded twice (once for
the .vcxproj itself, and once for IntDir) and added to the full path.
When combined with the tiny path limits on Windows, it is easy to exceed
path limits.
This error is largely avoided by the ninja back-end. Unlike the
vs back-end, the ninja back-end does not use target.get_id() as part of
the project file path, nor does it use target.get_id() as part of
get_target_private_dir().
Example error:
error MSB4184: The expression "[MSBuild]::NormalizePath(
C:\src\mesonbuild\Misc\FreeRDP-master\client\X11\xfreerdp\xfreerdp,
f3f7317@@Windows resource for file
'Misc_FreeRDP-master_client_X11_xfreerdp_xfreerdp_xfreerdp.rc'@cus\,
f3f7317@@Windows resource for file
'Misc_FreeRDP-master_client_X11_xfreerdp_xfreerdp_xfreerdp.rc'@cus.
vcxproj.CopyComplete)" cannot be evaluated. Path:
C:\src\mesonbuild\Misc\FreeRDP-master\client\X11\xfreerdp\xfreerdp\f3f7317
@@Windows resource for file
'Misc_FreeRDP-master_client_X11_xfreerdp_xfreerdp_xfreerdp.rc'@cus\f3f7317
@@Windows resource for file
'Misc_FreeRDP-master_client_X11_xfreerdp_xfreerdp_xfreerdp.rc'@cus.
vcxproj.CopyComplete exceeds the OS max path limit.
The fully qualified file name must be less than 260 characters.
|
|
|
|
|
|
|
|
|
|
|
|
```
2020-09-23T01:25:14.7849070Z
2020-09-23T01:25:14.7849592Z 1/5 Boost linktest TIMEOUT 30.16s
2020-09-23T01:25:14.7849811Z
2020-09-23T01:25:14.7850027Z --- command ---
2020-09-23T01:25:14.7850281Z 01:24:30 D:\a\1\s\b 125f976e40\linkedexe.exe
2020-09-23T01:25:14.7850561Z -------
2020-09-23T01:25:14.7850693Z
2020-09-23T01:25:14.7850947Z 2/5 Boost UTF test TIMEOUT 31.15s
2020-09-23T01:25:14.7851141Z
2020-09-23T01:25:14.7851347Z --- command ---
2020-09-23T01:25:14.7851580Z 01:24:30 D:\a\1\s\b 125f976e40\utf.exe
2020-09-23T01:25:14.7851862Z --- stdout ---
2020-09-23T01:25:14.7852065Z Running 1 test case...
2020-09-23T01:25:14.7852441Z
```
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
|
|
Add most missing zsh completions
|
|
|
|
|
|
|
|
|
|
There are two cases that won't work (these are taken from fortran/9
cpp), using gfortran with a non-gcc compiler on windows, or using
gfortran with apple clang. The latter is due to default search paths,
which we can fix, but is out of scope for this MR.
|
|
Instead of the default ones, this is especially important when cross
compiling or when using compilers that aren't compatible with the
default ones.
squash! dependencies/hdf5: Use the actual system compilers
|
|
Instead of a mega dependency that does everything, use a dependency
factory for config-tool and pkg-config
|