This requires setting CMAKE_CXX_COMPILER_TARGET manually; the LLVM
build system infers the wrong architecture when CMake is an
x86_64 binary.
(The wrong inferred architecture doesn't affect building libcxx,
but the wrong architecture gets passed as --target= parameter
when running the tests.)
Since 9c67ae861ccb0312fc680414a8441892d169d14a in llvm-project
(May 9th), these tests should pass on aarch64 too.
Co-authored-by: Jeremy Drake <github@jdrake.com>
Currently, "macos-latest" gets "macos-14", while there also are
"macos-15" runners available. Switch back to the latest variant
in preparation for that being updated at some point.
This reverts 7bd4b284f9.
As the preinstalled CMake on the windows-11-arm runners currently
is an x86_64 build, we need to explicitly specify
CMAKE_SYSTEM_PROCESSOR, and CMAKE_SYSTEM_NAME to make CMake
consider it a cross compilation (otherwise CMAKE_SYSTEM_PROCESSOR
gets overridden by CMAKE_HOST_SYSTEM_PROCESSOR, which gets influenced
by the x86_64 as CMake is running emulated).
We also need to set LLVM_HOST_TRIPLE, as the autodetection of
it relies on CMAKE_HOST_SYSTEM_PROCESSOR. The host triple needs
to be in normalized form with "-windows-gnu", otherwise a couple
dozens of tests fail.
We also include the AArch64 target, to make sure the built compiler
can generate code for the host.
Since 07bc54bf4554398b199f4dc849e5193b98422f23 in llvm-project
(May 5th), these tests should pass on aarch64 too.
Since 3e56f5f41843204b6e612362542b77110f8f8188 in llvm-project,
this should pass for all 4 architectures. (This doesn't pass
in the pinned release version right now.)
This avoids relying on finding these dependencies in $PATH while
running the tests. Some tests may set PATH="" while executing them,
which previously caused those tests to fail in this build setup.
The same is not an issue with MSVC builds with
-DCMAKE_MSVC_RUNTIME_LIBRARY=MultiThreadedDLL, because those
DLLs are also available in global system directories (including
DLLs such as msvcp140.dll and vcruntime140.dll).
When removing them, also accept that they are missing to begin
with.
Since 655933070219f2b6f3a457c7e5af7edd4b5291b4 in llvm-project
(Apr 25th), asan only gets built for x86.
This matches what 28851edf164a337c334755ae33fd58f03cffd5a2 in
llvm-project did.
When setting up the built libcxx files for testing, the built
files are installed into a temporary test root, and the install of
them litters the build log with mostly uninteresting output.
This was not converted in 5fe2e05419,
probably because those (automake based) makefiles work fine with
BSD make.
Use $MAKE here anyway, for consistency.
As the releases are built (without a surrounding extra container)
on GitHub Actions these days, we don't need to hold the docker
images back to an old version for the sake of compatibility;
instead, updating to a newer version provides more modern tools
for users of the docker image.
The script release.sh, which earlier was used for building releases,
uses docker for doing clean isolated builds, and thus needs to be
updated with the distro version used in the dockerfiles.
CMake 4.0 removed compatibility with CMake versions below 3.5.
Therefore, it errors out on projects with a
cmake_minimum_required() less than 3.5.
(Since CMake 3.31, it already warns that compatibility with versions
below 3.10 is deprecated.)
By passing CMAKE_POLICY_VERSION_MINIMUM=3.5, we indicate that
we are ok with configuring this project with the policies of
CMake 3.5, even if the project files themselves don't state it.
This flag is only recognized by CMake 4.0; older versions will
only warn that the manually specified variable is unused.
Relevant changes in LLVM 20.1.x compared with LLVM 19.x are:
- Don't warn when doing LTO with mixed Rust/C object files, where
the target triples differ in the vendor field
(<arch>-w64-windows-gnu vs <arch>-pc-windows-gnu)
- Fix including LLVM/MSVC style import libraries in libraries linked
with --whole-archive
- Support for files over 2 GB in the libcxx fstream methods
tellg()/seekg()
- Lots of work on improved ARM64EC and ARM64X support in LLVM and LLD
- New linker flags --dependent-load-flag and --functionpadmin, similar
to the corresponding lld-link/MS link.exe flags
- Improvements in address sanitizer, supporting many more x86
instructions
- Changes to address sanitizer, that (unintentionally) makes it fail
harder than before when used together with msvcrt.dll. Using
address sanitizer with msvcrt.dll is unsupported though.
Since 6127724786d581d2653df7420d1f28709288a107 in llvm-project
(Nov 4 2024), this variable no longer is referenced, and it
probably hasn't been needed for much longer than that.
It's enough to check for the _UCRT define these days, as long
as we are including a header that pulls in the base mingw headers.
Since 8b2ece59fccf0a65257b1ba2fe8d67d7e8a2e8c1 in mingw-w64 (May 6th
2019), the mingw-w64 headers define _UCRT. Since
94a28397eec79e54ce5a8354512f43da88f815d2 (committed on the same date),
the mingw-w64 headers use 0xE00 instead of 0x1400 for UCRT.
This reverts commit 490dfffb32.
This is no longer necessary, as we now can build runtimes using
the latest LLVM 19 versions, that support building sanitizers
for all architectures.
Update to using clang-18 from the host system; version 16 or
newer is required for the implicitly selected config files
(which are used since 866d47cced);
that version changed whether the target triples are normalized
before locating config files.
This allows building runtimes with the current versions of LLVM
19.x, rather than needing to select an older version.
This is one of the packages which aren't needed by the toolchain
build itself, but is provided as a convenience to useres of the image.
Many projects that build standalone x86 assembly can use either
nasm or yasm, but lately most projects prefer nasm (which has
support for newer x86 simd extensions).
In Ubuntu 24.04, this package is no longer available.
However, this package probably never actually was needed anyway,
and was probably only provided as a convenience for users of the
docker image.
This is required for building libffi. With ubuntu-22.04 images,
the libltdl-dev package was available out of the box.
Also explicitly install swig, which we rely on - this also seems
to have been available implicitly.
This fixes building on Ubuntu 24.04 (the one currently used by
the github actions ubuntu-latest tag), by including a fix for texinfo
7.0.
Not bumping to ffmpeg 7.x, as some tests there can be flaky at times.
At this point, it is reasonably mature.
There still obviously are projects that won't build with llvm-mingw,
both due to unimplemented features that GCC/binutils have, or for
other reasons, but the toolchain has seen quite a lot of usage
since this paragraph originally was written.
The wrapper script no longer has any purpose on Unix.
Just do a plain symlink to the clang-scan-deps tool.
On Windows, the wrapper executable still is used to inject a
default -target option, if the executable name doesn't contain
a triple.
clang-scan-deps can pick it up automatically from the first argument
now.
For the windows wrapper, keep passing -target, to allow running with
a potentially foreign clang-scan-deps.exe.
Clang only picks up config files implicitly using the normalized
triple form, so use file names with "-windows-gnu" instead of
"-mingw32".
In theory, we could also want to pick up the config file implicitly
by having "<triple>-<mode>" be a symlink directly to the clang
executable. That works for regular mingw32 targets, but for the
mingw32uwp target, there's no distinction in the normalized
triple, so both would end up using the same "-windows-gnu" config
files. So as long as we want to support the mingw32uwp prefixed
tools, we need to use a wrapper.
Martin Storsjö