Improve rustdoc-gui search-color test
Thanks to the add of "functions" in `browser-ui-test`, we can start to reduce the size of the scripts. It'll be very useful for all color checks.
r? `@notriddle`
Use BOLT in CI to optimize LLVM
This PR adds an optimization step in the Linux `dist` CI pipeline that uses [BOLT](https://github.com/llvm/llvm-project/tree/main/bolt) to optimize the `libLLVM.so` library built by boostrap.
Steps:
- [x] Use LLVM 15 as a bootstrap compiler and use it to build BOLT
- [x] Compile LLVM with support for relocations (`-DCMAKE_SHARED_LINKER_FLAGS="-Wl,-q"`)
- [x] Gather profile data using instrumented LLVM
- [x] Apply profile to LLVM that has already been PGOfied
- [x] Run with BOLT profiling on more benchmarks
- [x] Decide on the order of optimization (PGO -> BOLT?)
- [x] Decide how we should get `bolt` (currently we use the host `bolt`)
- [x] Clean up
The latest perf results can be found [here](https://github.com/rust-lang/rust/pull/94381#issuecomment-1258269440). The current CI build time with BOLT applied is around 1h 55 minutes.
Update CI to use Android NDK r25b
This commit updates the CI definitions to use the most recent Android LTS NDK release: r25b. Changes since the last NDK used by Rust negate the need to generate "standalone toolchains" and newer NDKs can be used in-place.
See https://developer.android.com/ndk/guides/other_build_systems#overview
Compressed debug is enabled by default for gas (assembly) on Linux/x86
targets, and we started building our own in #102530, but that made our
`compiler_builtins` incompatible with binutils < 2.32. Add an explicit
option to disable that in our crosstool-ng config. Fixes#102703.
This commit updates the CI definitions to use the most recent Android
LTS NDK release: r25b. Changes since the last NDK used by Rust negate
the need to generate "standalone toolchains" and newer NDKs can be used
in-place.
See https://developer.android.com/ndk/guides/other_build_systems#overview
The system GCC 5 in ubuntu:16.04 will be too old to compile LLVM 16, so
we need an upgrade. To avoid raising the minimum glibc requirements for
`i586-unknown-linux-gnu`, this target is converted to a crosstool-ng
toolchain, *relaxing* it to the same Linux 3.2 / glibc 2.17 minimum we
use elsewhere. The musl targets still use Ubuntu's system toolchain, but
this doesn't have the same compatibility concerns.
Make the `c` feature for `compiler-builtins` an explicit opt-in
Its build script doesn't support cross-compilation. I tried fixing it, but the cc crate itself doesn't appear to support cross-compiling to windows either unless you use the -gnu toolchain:
```
error occurred: Failed to find tool. Is `lib.exe` installed?
```
Fixes https://github.com/rust-lang/rust/issues/101172.
The build script for `compiler_builtins` doesn't support cross-compilation. I tried fixing it, but the cc crate itself
doesn't appear to support cross-compiling to windows either unless you use the -gnu toolchain:
```
error occurred: Failed to find tool. Is `lib.exe` installed?
```
Rather than trying to fix it or special-case the platforms without bugs,
make it opt-in instead of automatic.
Distribute bootstrap in CI
This pre-compiles bootstrap from source and adds it to the existing `rust-dev` component. There are two main goals here:
1. Make it faster to build rust from source, both the first time and incrementally
2. Make it easier to add non-python entrypoints, since they can call out to bootstrap directly rather than having to figure out the right flags to pre-compile it. This second part is still in a bit of flux, see the tracking issue below for more information.
There are also several changes to make bootstrap able to run on a machine other than the one it was built (particularly around `config.src` and `config.out` detection). I (`@jyn514)` am slightly concerned these will regress unless tested - maybe we should add an automated test that runs bootstrap in a chroot or something? Unclear whether the effort is worth the test coverage.
Helps with https://github.com/rust-lang/rust/issues/94829.
Improve rustdoc GUI tests
I finally finished the update so we can now store values in variables and use them. It improves things nicely.
r? ``@notriddle``
- Add a new `bootstrap` component
Originally, we planned to combine this with the `rust-dev` component.
However, I realized that would force LLVM to be redownloaded whenever bootstrap is modified.
LLVM is a much larger download, so split this to get better caching.
- Build bootstrap for all tier 1 and 2 targets
ci: Upgrade android containers from ubuntu:16.04 to 22.04
The main goal of updating to 22.04 is to get away from `llvm.allow-old-toolchain`.
These containers are not building LLVM for android, so only the host version matters.
A side benefit is that they can also use the system `cmake` instead of building one.
ci: Upgrade non-dist Linux testers from ubuntu:16.04 to 22.04
The main goal of updating to 22.04 is to get away from `llvm.allow-old-toolchain`.
A side benefit is that they can also use the system `cmake` instead of building one.
Rootless podman creates a separate user namespace, where an inner
`LOCAL_USER_ID` will map to a different subuid range on the host.
The "keep-id" mode maps the current UID directly into the container.
This makes `src/ci/docker/run.sh` work better for testing container
images on systems running podman, where "docker" is just a shim.
Download, rather than sccache-cache, LLVM in CI
My hope/expectation is that we can do better than sccache in CI for cached builds -- currently it looks like on macOS those still take upwards of 10-11 minutes, which is a significant amount of time that we could potentially cut.
This enables this mode for all non-dist builders; this should avoid any problems with the artifacts we distribute, while also providing for faster test builders (since they'll make use of PGO'd LLVM on the platforms we do that on, which is hopefully a nice win). It slightly increases the chance of test builders starting to fail only after a PR is merged (if PGO changes runtime behavior), but that should hopefully never happen, so I think this is worthwhile.
Measurements on the PR for apple-1 don't show any noticeable improvement in CI times, but those can be pretty noisy -- I'm inclined to land this since it *should* pretty much always be better and we can reconsider if that ever turns out not to be the case.
Initial implementation of REUSE
This PR implements the first two steps of #99414 by:
* Adding some scaffolding for REUSE. The `.reuse/dep5` file now marks every file as the custom "TODO" license, which I'll remove in a future PR once Debian imports their metadata. The TODO license is needed so that `reuse lint` works.
* Runs `reuse lint` in CI, in the `mingw-check` builder. REUSE currently has a bug when parsing some files in the LLVM source code. This means REUSE will fail when running it in source tarballs of rustc, and that bug prevents us from passing the `--include-submodules` flag in CI. I opened https://github.com/fsfe/reuse-tool/pull/560 upstream with a fix, and as soon as it's merged/released I planned to bump the pinned version to include the fix we need.
r? `@Mark-Simulacrum`
See comment added for details on the test builder restriction. This is primarily
intended for macOS CI, but is likely to be a slight win on other builders too.
Update to LLVM 15
For preliminary testing. Some LLVM 15 compatibility fixes were applied separately in #99512.
Release timeline:
* LLVM 15 branched on Jul 26.
* The final LLVM 15.0.0 release is scheduled for Sep 6.
* Current nightly (1.65.0) is scheduled for Nov 3.
Changes in this PR (apart from the LLVM update):
* Pass `--set llvm.allow-old-toolchain` for many Docker images. LLVM 16 will require GCC >= 7.1, while LLVM 15 still allows older compilers with an option. Specify the option for builders still using GCC 5.4. #95026 updated some of the used toolchains, but not all.
* Use the `+atomics-32` target feature for thumbv6m.
* Explicitly link libatomic when cross-compiling LLVM to 32-bit target.
* Explicitly disable zstd support, to avoid libzstd.so dependency.
New LLVM patches ([commits](https://github.com/rust-lang/llvm-project/commits/rustc/15.0-2022-08-09)):
* [rust-only] Fix ICE with GCC 5.4 (15be58d7f0)
* [rust-only] Fix build with GCC 5.4 (774edc10fa)
* ~~[rust-only] Fix build with GCC 5.2 (1a6069a7bb)~~
* ~~[rust-only] Fix ICE with GCC 5.2 (493081f290)~~
* ~~[rust-only] Fix build with GCC 5.2 (0fc5979d73)~~
* [backported] Addition of `+atomics` target feature (57bdd9892d).
* [backported] Revert compiler-rt change that broke powerpc (9c68b43915)
* [awaiting backport] Fix RelLookupTableConverter on gnux32 (639388a05f / https://github.com/llvm/llvm-project/issues/57021)
Tested images: dist-x86_64-linux, armhf-gnu, arm-android, dist-s390x-linux, dist-x86_64-illumos, dist-x86_64-freebsd, wasm32, dist-x86_64-musl, dist-various-1, dist-riscv64-linux, dist-mips-linux, dist-mipsel-linux, dist-powerpc-linux, dist-aarch64-linux, dist-x86_64-apple, x86_64-msvc-1, x86_64-msvc-2, dist-various-2, dist-arm-linux
Tested up to the usual ipv6 error: test-various, i686-gnu, x86_64-gnu-nopt
r? `@ghost`
Add `x.sh` and `x.ps1` shell scripts
This is a more ambitious version of https://github.com/rust-lang/rust/pull/98716.
It still changes the x.py shebang back to python3, for compatibility with non-Unix systems,
but also adds alternative entrypoints for systems without `python3` installed.
These scripts will be necessary for the rust entrypoint (#94829), so I see
little downside in adding them early.
I'll update the dev-guide to suggest using these instead of x.py once this is merged.
Fixes https://github.com/rust-lang/rust/issues/98650
r? `@Mark-Simulacrum` cc `@dtolnay` `@CAD97` `@yoshuawuyts`
This is a more ambitious version of https://github.com/rust-lang/rust/pull/98716.
It still changes the shebang back to python3, for compatibility with non-Unix systems,
but also adds alternative entrypoints for systems without `python3` installed.
These scripts will be necessary for the rust entrypoint (#94829), so I see
little downside in adding them early.
Fuchsia CI zircon lib improvement
Removing Zircon build process, instead pulling `sysroot` and related libs directly from Fuchsia SDK
cc. `@tmandry` `@djkoloski`
With binutils 2.32, we were getting errors like this:
relocation truncated to fit: R_PPC_PLTREL24 against symbol
`__cxa_atexit@@GLIBC_2.1.3' defined in .plt section in
/x-tools/powerpc-unknown-linux-gnu/powerpc-unknown-linux-gnu/sysroot/usr/lib/crt1.o
but it builds okay with binutils 2.30.
This is implementing the MCP from rust-lang/compiler-team#493. It is
increasing the minimum requirements of a couple Tier 1 targets, and
others at lower tiers, so this should go through FCP sign-offs for both
`T-compiler` and `T-release`.
The new `linux-gnu` baseline is kernel 3.2 and glibc 2.17. We will also
take that kernel as the minimum floor for _all_ `*-linux-*` targets, so
it may be broadly assumed in the implementation of the standard library.
That does not preclude specific targets from having greater requirements
where it makes sense, like a new arch needing something newer, or a
platform like `linux-android` choosing a newer baseline.
