| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
|
|
not to pass None as parameter
507cf47507cf47 broke the combination of clone-recursive without depth.
Because it passed depth (as None) to git.
To fix this the depth option is now generated once as a list when depth is set
and else set to an empty list and unpacked into the argument this when
needed. This reduces duplication and allows code to work the same for
both cases.
|
|
Previously, warnings genereated in CMake were not logged because
CMake sends them to stderr alongside the trace information. With
this PR, only real trace lines are send to the trace parser and
all other lines in stderr are logged as warnings.
0.52 branch version
|
|
Same fix as e7b25018c4715e538234d4ba51e32194b4757388, was accidentally
missed when this code was rewritten.
|
|
This reverts commit 7b9c348102792030859ed0001a51416506a0a092.
Closes #6027.
|
|
Fixes #5800
|
|
This means that when either cflags or libs lookup fails, this will be
logged always in the log, leading to much easier to debug messages
like:
Called `/usr/local/bin/pkg-config --cflags gnutls` -> 1
pkg-config error with 'gnutls': Could not generate cargs for gnutls:
Package nettle was not found in the pkg-config search path.
Perhaps you should add the directory containing `nettle.pc'
to the PKG_CONFIG_PATH environment variable
Package 'nettle', required by 'gnutls', not found
Run-time dependency gnutls found: NO (tried pkgconfig)
|
|
The stderr of pkg-config was never used, leading to no details about
failures at all in exception messages.
|
|
Previously even when meson failed to obtain cflags or libs for a
dependency it would lead to the dependency being found, causing very
confusing errors at build time.
|
|
|
|
When a static library link_whole to a bunch of other static libraries,
we have to extract all their objects recursively. But that could
introduce duplicated objects. ar is dumb enough to allow this without
error, but once the resulting static library is linked into an
executable or shared library, the linker will complain about duplicated
symbols.
|
|
Under Windows, the colon ':' is used after the drive letter.
So, the colon should not be used as a list separator or for splitting,
otherwise it could lead to paths in CMAKE_PREFIX_PATH with a
semicolon ';' between the drive letter and the rest of the path, e.g:
-DCMAKE_PREFIX_PATH=C;/foo/bar
instead of
-DCMAKE_PREFIX_PATH=C:/foo/bar
Use os.pathsep instead of ':' to split the environmental variable
CMAKE_PREFIX_PATH
|
|
|
|
The is_disabler() check in evaluate_foreach() was useless because we
already checked that items is a list.
|
|
Closes: #6010
|
|
|
|
The optlink linker is slowly getting phased out now since DMD ships with the LLVM linker, so it can be used when Visual Studio is not installed.
|
|
|
|
Previously it worked by accident because BasicLinkerIsCompilerMixin had
that method misspelled.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
get_allow_undefined_link_args is the compiler method
get_allow_undefined_args is the linker method
|
|
|
|
The lack of newlines in the print() statement for this error causes the result
to be jumbled and hard to read:
$ ./meson.py introspect
Current directory is not a meson build directory.Please specify a valid build dir or change the working directory to it.It is also possible that the build directory was generated with an oldmeson version. Please regenerate it in this case.
Add newlines so the message is more readable:
$ ./meson.py introspect
Current directory is not a meson build directory.
Please specify a valid build dir or change the working directory to it.
It is also possible that the build directory was generated with an old
meson version. Please regenerate it in this case.
|
|
fixes #6000
The idea is that end-users want to specify an array of directories to search by default
without an if/elif stack. It's obvious that Unix absolute paths are not absolute on
Windows, so silently discard Unix absolute paths here for Windows instead of raising
exception.
|
|
|
|
|
|
It is perfectly valid to pass the arguments separately `-L /some/dir/`,
however, meson later groups arguments by whether they start with -L or
not, which breaks passing the -L and the directory separately.
Fixes #6003
|
|
Since cuda is currently reimplementing all of the shared methods itself,
including a bug fixed by the next patch.
|
|
|
|
|
|
Add is_system to dependency
|
|
|
|
cmake: fix detection of libraries with .so version
|
|
Don't manually emulate a partial `git clone` when `depth` option is not
used. This keeps `git describe` working as before and generally supports
workflows that depend on tags and branches to exist in a wrap downloaded
subproject.
This also fixes downloading via git not working at all on CentOS 7
(git version 1.8.3.1).
For the `depth` case use `git clone --branch=... --depth=...` when
possible and only fall back to manual emulation wraps that specify a
full commit id, because for some reason that does not work with the
clone based workflow.
Fixes: #5991 (Regression in wrap support with git)
|
|
Correctly warn and document options to benchmarks
|
|
When using the '--gdb' argument of meson test the executed binary can
now be specified with '--gdb-path'.
Closing: #4373
|
|
This reverts commit 73f006149823a498f88e0131f963a0ff81043788.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
