Remove test_minor_version_does_not_reconfigure_wipe() because when run
during dev cycle that test reconfigure with .99 -> .100 which is
considered a major version change now. It is covered by a more efficient
internal test now anyway.
While at it, remove no-op `with Path(self.builddir):` statement, the
intention was clearly to set workdir.
Fixes: #9260
It is a commonly needed information to help debugging build issues. We
already were printing options with non-default value at the end of the
configure but outside of the summary.
Keeping the list of user defined options in the interpreter will also in
the future be useful to use new default value on reconfigure.
We have a lot of these. Some of them are harmless, if unidiomatic, such
as `if (condition)`, others are potentially dangerous `assert(...)`, as
`assert(condtion)` works as expected, but `assert(condition, message)`
will result in an assertion that never triggers, as what you're actually
asserting is `bool(tuple[2])`, which will always be true.
* backends: Add a Visual Studio 2013 backend
This is more-or-less a quick port from the VS2015 backend, except that
we update the Visual Studio version strings and toolset versions
accordingly. Also correct the generator string for Visual Studio 2015
in mesonbuild/cmake/common.py.
* backend: Add VS2012 backend
Similar to what we did for Visual Studio 2013, add a Visual Studio 2012
backend.
* vs2010backend.py: Implement `link_whole:` if needed
We actually need Visual Studio 2015 Update 2 to use `/WHOLEARCHIVE:`,
which is what we are currently using for `link_whole:` on Visual Studio.
For Visual Studio versions before that, we need to expand from the
static targets that were indicated by `link_whole:`, and any of the
sub-dependent targets that were pulled in via the dependent target's
`link_whole:`. This wil ensure `link_whole:` would actually work in
such cases.
* vs2010backend.py: Handle objects from generated sources
Unforunately, we can't use backends.determine_ext_objs() reliably, as
the Visual Studio backends handle this differently.
* vs2010backend.py: Fix generating VS2010 projects
Visual Studio 2010 (at least the Express Edition) does not set the envvar
%VisualStudioVersion% in its command prompt, so fix generating VS2010
projects by taking account into this, so that we can determine the location
of vcvarsall.bat correctly.
* whole archive test: Disable on vs2012/2013 backends too
The Visual Studio 2012/2013 IDE has problems handling the items that would be
generated from this test case, so skip this test when using
--backend=vs[2012|2013]. This test does work for the Ninja backend when
VS2012 or VS2013 is used, though.
Consolidate this error message with XCode along with the vs2010 backend.
* docs: Add the new vs2012 and vs2013 backends
Let people know that we have backends for vs2012 and 2013. Also let
people know that generating Visual Studio 2010 projects have been fixed
and the pre-vs2015 backends now handle the `link_whole:` project option.
We need to konw on rconfigure which options have already bee set not
just for the super project, but also for the subproject. However, using
first_invocation is not sufficient, as a reconfigure could add a new
subpproject that wasn't present before, and we need to initialize that
project's builtins.
These options often don't exist for the build machine, as they're
generally invalid (ie, prefix, libdir, etc). So we should always check
for their existance as host optiuons, since all we want to know in this
case is if the option is yeilding (valid per-subproject), and build/vs
host doesn't matter in that cases:
Fixes#8613
They are supposed to have different behavior. The environment variables
apply to both the compiler and linker when the compiler acts as a
linker, but the command line ones do not.
Fixes#8345
This fix issue when using --wipe and the machine file was passed as a
pipe and written locally, or when the file was resolved in XDG_DATA_HOME
or XDG_DATA_DIRS.
Fixes: #8560
When reverting from 0.57 to 0.56, one can see an error like this:
File /Users/pm215/src/qemu-for-merges/meson/mesonbuild/coredata.py,
line 1016, in load
obj = pickle.load(f)
ModuleNotFoundError: No module named 'mesonbuild.mesonlib.universal';
'mesonbuild.mesonlib' is not a package
FAILED: build.ninja
The reason is that the old version fails to resolve mesonbuild.mesonlib,
which is a similar situation to the existing AttributeError check. Raise
a MesonException for ModuleNotFoundError as well, so that reconfiguration
proceeds using cmd_line.txt.
All changes were created by running
"pyupgrade --py3-only --keep-percent-format"
and committing the results. I have not touched string formatting for
now.
- use set literals
- simplify .format() parameter naming
- remove __future__
- remove default "r" mode for open()
- use OSError rather than compatibility aliases
- remove stray parentheses in function(generator) scopes
This has a bunch of nice features. It obviously centralizes everything,
which is nice. It also means that env is only re-read at `meson --wipe`,
not `meson --reconfigure`. And it's going to allow more cleanups.
This patches takes the options work to it's logical conclusion: A single
flat dictionary of OptionKey: UserOptions. This allows us to simplify a
large number of cases, as we don't need to check if an option is in this
dict or that one (or any of 5 or 6, actually).
I would have prefered to do these seperatately, but they are combined in
some cases, so it was much easier to convert them together.
this eliminates the builtins_per_machine dict, as it's duplicated with
the OptionKey's machine parameter.
This is useful for figuring out what kind of option this is. My hope is
that in the longterm this is less useful, but we'll still want it for
the configuration summary printing.
This is a complex key that can store multiple bits of data in a single
place. It can be generated from a command line formatted string, and
it's str method returns it to that form.
It's intentionally immutable, use the evolve() method to create
variations of an existing key.
This is consistent with c_args in machine file overriding CFLAGS from
env. This also spotted an issue where in a native build this resulted
in pkg_config_path being /bar instead of /foo:
`-Dpkg_config_path=/foo -Dbuild.pkg_config_path=/bar`
Fixes: #7573
It is much easier to not try to parse options into complicated
structures until we actually collected all options: machine files,
command line, project()'s default_options, environment.
Currently if you change the `choices` field in the meson_options.txt
file, no update will be done until `meson setup --wipe` is called. Now
if the choices change then the options will be properly merged.
If the currently select value is still valid it is guaranteed to be
kept, if it is now invalid the new default value will be used and a
warning will be printed.
Fixes#7386
This is required to make the various keys in the [user options] section
work the same as they do in the meson_options.txt file, where we don't
have any rules about case sensitivity.
There is some risk here. Someone may be relying on this lower by default
behavior, and this could break their machine files.
Fixes#7731
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
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
Currently we do some crazy hackery where we add extra properties to a
Popen object and return that. That's crazy. Especially since some of our
hackery is to delete attributes off of the Popen we don't want. Instead,
let's just have a discrete type that has exactly the properties we want.
There is nothing to "warn" about, this is a completely routine
occurence. OTOH, when something is corrupted, we should warn. Keep
the red color and "WARNING:" prefix in that case.
Example output:
$ ninja -C build
Regenerating configuration from scratch: Build directory has been generated with Meson version 0.55.999, which is incompatible with current version 0.56.0.
The Meson build system
Version: 0.56.0
...
This creates a full set of option in environment that mirror those in
coredata, this mirroring of the coredata structure is convenient because
lookups int env (such as when initializing compilers) becomes a straight
dict lookup, with no list iteration. It also means that all of the
command line and machine files are read and stored in the correct order
before they're ever accessed, simplifying the logic of using them.
This puts all of them together, in the next patch they'll be pulled back
out, but it's convenient to start that refactor by moving them all
there, then moving them into env as a whole.
This is like the project options, but for meson builtin options.
The only real differences here have to do with the differences between
meson builtin options and project options. Some meson options can be set
on a per-machine basis (build.pkg_config_path vs pkg_config_path) others
can be set on a per-subproject basis, but should inherit the parent
setting.
This allows adding a `[project options]` section to a cross or native file
that contains the options defined for a project in it's meson_option.txt
file.
Some warnings are out of the user's control, such as the RCC QT bug,
or the GNU windres bug, or our informational warning about
auto-disabling of options when -Db_bitcode is enabled.
Such warnings should not be fatal when --fatal-meson-warnings is
passed because there's no action that the user can take to fix it. The
only purpose it serves is to prevent people who use those features
from using --fatal-meson-warnings.
Machine files already supports `+` operator as an implementation detail,
since it's using eval(). Now make it an officially supported feature and
add a way to define constants that are used while evaluating an entry
value.
Generally, we'd want to use str() rather than repr() in error messages
anyhow, as that explicitly gives something designed to be read by
humans.
Sometimes {!r} is being used as a shortcut to avoid writing the quotes
in '{!s}'.
Unfortunately, these things aren't quite the same, as the repr of a
string containing '\' (the path separator on Windows) will have those
escaped.
We don't have a good string representation to use for the arbitrary
internal object used as an argument for install_data() when it's neither
a string nor file (which doesn't lead to a good error message), so drop
that for the moment.
A current rather untyped storage of options is one of the things that
contributes to the options code being so complex. This takes a small
step in synching down by storing the compiler options in dicts per
language.
Future work might be replacing the langauge strings with an enum, and
defaultdict with a custom struct, just like `PerMachine` and
`MachineChoice`.
It may not be obvious to users that these two ways to set build-types
override each other and specifying both is redundant, and conflicts
are resolved based on whichever is specified later.
Closes https://github.com/mesonbuild/meson/issues/6742
With the current logic passing `--debug` will actually be parsed as
`-Ddebug=false`, which is absolutely not what is expected.
There is no case in which the presence of a boolean option in `--foo`
form will mean 'I want feature foo disabled', regardless of the
*default* value of that option.
Also includes a test.
Closes https://github.com/mesonbuild/meson/issues/4686
The previous code was assuming that options do not depend on each
other, and that you can set defaults using `dict.setdefault()`. This
is not true for `buildtype` + `optimization`/`debug`, so we add
defaults + overrides in the right order and use the options parsing
code later to compute the values.
Includes a test.
Closes https://github.com/mesonbuild/meson/issues/6752
This ensures that options are always parsed in the order in which they
were specified on the command-line, even with Python 3.5, and
non-CPython implementations compatible with CPython 3.5 and 3.6.
Closes https://github.com/mesonbuild/meson/issues/6742
Currently it's just like if all builtin/base/compiler options are
yielding. This patch makes possible to have non-yielding builtin
options. The value in is overriden in this order:
- Value from parent project
- Value from subproject's default_options if set
- Value from subproject() default_options if set
- Value from command line if set
* Extend test_prefix_dependent_defaults unit test to cover default case
Extend test_prefix_dependent_defaults unit test to cover the default
case, when the default prefix is '/usr/local'. (On Windows, the default
prefix is 'c:/')
* Restore adjusting option defaults depending on the default prefix
Restore adjusting option defaults, depending on the default prefix.
Droppped in d778a371
If a user passes -fuse-ld=gold to gcc or clang, they expect that they'll
get ld.gold, not whatever the default is. Meson currently doesn't do
that, because it doesn't pass these arguments to the linker detection
logic. This patch fixes that. Another case that this is needed is with
clang's --target option
This is a bad solution, honestly, and it would be better to use $LD or a
cross/native file but this is needed for backwards compatability.
Fixes#6057
Previously if a user tried to pass a command line build
option that contained a '%' character the command line
parser assumed that there was string interpolation to be
done. As there is no sense in such a scenario no code
provides any input for the interpolation. This then leads to
a failure.
In this commit we specifically override the defaults in
ConfigParser and set interpolation to None, which disables
command line build option interpolation.
Fixes#6157
Otherwise having an output directory matching the name of the cross file
in the current working directory will cause an error. This patch instead
collects any invalid names and prints them along with the error message
when no valid file can be found.
Fixes#5640
This is used in a lot of places, having a single way to reference it is
convenient. It's placed under a typing.TYPE_CHECKING guard to mimimize
runtime impact
* coredata: Correctly handle receiving a pipe for native/cross files
In some cases a cross/native file may be a pipe, such as when using bash
process replacement `meson --native-file
<([binaries]llvm-config='/opt/bin/llvm-config')`, for example. In this
case we copy the contents of the pipe into a file in the meson-private
directory so we can create a proper ninja dependency, and be able to
reload the file on --wipe/--reconfigure. This requires some extra
negotiation to preserve these native/cross files.
Fixes#5505
* run_unitests: Add a unit test for native files that are pipes
Using mkfifo.
In most cases instead pass `for_machine`, the name of the relevant
machines (what compilers target, what targets run on, etc). This allows
us to use the cross code path in the native case, deduplicating the
code.
As one can see, environment got bigger as more information is kept
structured there, while ninjabackend got a smaller. Overall a few amount
of lines were added, but the hope is what's added is a lot simpler than
what's removed.
Handling the PKG_CONFIG_PATH variable in meson introduces a new problem
for caching dependencies. We want to encode the pkg_config_path (or
cross_pkg_config_path if we're cross compiling) to be part of the key,
but we don't want to put that into the key for non-pkg-config
dependencies to avoid spurious cache misses (since pkg_config_path isn't
relevant to cmake, for example). However, on a cache lookup we can't
know that a dependency is a pkg-config dependency until we've looked in
the cache.
My solution is a two layer cache, the first layer remains the same as
before, the second layer is a dict-like object that encapsulates the
dependency type information and uses pkg_config_path and
cross_pkg_config_path as a sub key (and could be extended easily for
other types). A new object type is introduced to encapsulate this so
that callers don't need to be aware of the implementation details.
Xavier Claessens