[libcxx-commits] [libcxx] [libc++] Guard additional headers with _LIBCPP_HAS_LOCALIZATION (PR #131921)

[David Spickett via libcxx-commits]

DavidSpickett wrote:

The instructions I posted were done on a real machine that, if I had thought about this first, I would have realised had built qemu before. Apologies for that.

These commands work on a `ubuntu:noble` image via Docker. It uses the AArch64 release of llvm, otherwise it should work on any host:
```
apt update && \
  DEBIAN_FRONTEND=noninteractive apt install curl unzip cmake python3-psutil wget xz-utils gcc g++ python3-venv ninja-build bzip2 libglib2.0-dev git make -y --no-install-recommends && \
  wget https://download.qemu.org/qemu-9.0.2.tar.xz && \
  tar -xvf qemu-9.0.2.tar.xz && \
  cd qemu-9.0.2 && \
  ./configure --prefix=/usr/local --target-list=arm-softmmu && \
  make -j$(nproc) && \
  make install && \
  cd .. && \
  wget https://github.com/llvm/llvm-project/releases/download/llvmorg-19.1.7/clang+llvm-19.1.7-aarch64-linux-gnu.tar.xz && \
  tar -xvf clang+llvm-19.1.7-aarch64-linux-gnu.tar.xz && \
  export CC=/clang+llvm-19.1.7-aarch64-linux-gnu/bin/clang && \
  export CXX=/clang+llvm-19.1.7-aarch64-linux-gnu/bin/clang++ && \
  export PATH=$PATH:/clang+llvm-19.1.7-aarch64-linux-gnu/bin/ && \
  git clone https://github.com/llvm/llvm-project.git && \
  cd llvm-project && \
  ./libcxx/utils/ci/run-buildbot armv7m-picolibc
```

Btw, given this is run in simulation, if it would help you to have this done using the "normal" CI machines and Docker image, I can look into porting the commands into that.

(Linaro remains committed to supporting the configuration regardless of where it runs)

If there is a follow up to this you'd like to see done, or a change in picolib that would help, we can handle that. I just can't promise to do it right now, as I need time to understand the issue fully.

https://github.com/llvm/llvm-project/pull/131921