Conversation
This reverts commit 59a507e.
Add support for tier 3 targets bpfeb-unknown-none and bpfel-unknown-none (see https://github.com/rust-lang/rust/blob/f5e2df7/src/doc/rustc/src/platform-support.md?plain=1#L311-L312). This is modeled after bazelbuild#3507 and should probably be updated if/when bazelbuild/platforms#131 is merged. (please use rebase merge when landing this as the proper commit message is in the commit, rather than the PR description) /cc @avrabe
…elbuild#3829)" This reverts commit f198dde.
…() (bazelbuild#3816)" This reverts commit 9586468.
…bazelbuild#3) * 0 * Add rust analyzer test coverage
…hollow rlibs: the RustcMetadata action runs rustc to completion with -Zno-codegen, emitting a .rlib archive. This approach mirrors the one used by buck2 and avoids needing to kill rustc mid-output in order to produce metadata. While not fixing problems with SVH mismatches when non-determinism, this does simplify the codepath and uses a production tested technique that doesn't have any of the dangers associated with killing the rustc process while it's still active.
Port the sharding wrapper feature from bazelbuild#3774 into the hermeticbuild fork. The implementation wraps rust_test executables when experimental_enable_sharding is set while keeping rustc_compile_action's existing provider-list API for internal and extension callers. rust_test now scans the returned providers to replace DefaultInfo for the wrapper, so extensions such as prost and wasm-bindgen continue to consume rustc_compile_action without API churn. Co-authored-by: Brian Duff <bduff@linkedin.com> Co-authored-by: Codex <noreply@openai.com>
Rustc emits GNU-like Windows staticlibs as lib<crate>.a, but rules_rust was stripping the lib prefix for all Windows non-rlib library outputs. Keep the prefix for staticlib outputs when the target ABI is gnu or gnullvm so declared outputs match rustc.
Add documentation extraction targets for the public cargo and rust Starlark packages. Declare the bazel_features, selects, and cc_debug_helper_bzl dependencies required by those targets.
Scan code-generating rustc outputs for the resolved ${pwd} value after successful compilation. Reject embedded CARGO_MANIFEST_DIR, OUT_DIR, and other sandbox paths while preserving compile-time include_str! and relative paths.
Inspect cargo_build_script OUT_DIR through test runfiles instead of retaining compile-action paths.
Assisted-by: OpenAI Codex
* rustc: stage cc dynamic libraries in test runfiles The cc dynamic libraries a target loads at runtime were collected into its runfiles only when the crate type was bin, cdylib or staticlib. A rust_test reports the type of the crate under test -- "lib" for rust_test(crate = ":foo") -- so tests were skipped entirely. The collection also walked only `deps`, never `crate`, so even a bin built that way would have missed them. A rust_test depending on a cc_import shared library therefore produced runfiles containing just the executable and the repo mapping. Its RUNPATH is $ORIGIN/../../_solib_<cpu>/..., so it runs only where that execroot directory happens to sit beside the binary. Locally it does, and the test passes. Under remote execution the input tree is exactly the declared runfiles, and the test dies at load time with `cannot open shared object file`. Gate on crate_info.is_test as well, and scan `crate` alongside `deps`. * test: cover imported dylibs in rust_test runfiles A cc_binary(linkshared = True) in `deps` already reached a consumer's runfiles through the generic deps traversal, so the existing check_runfiles cases passed regardless of the dynamic library collection. A cc_import contributes no runfiles of its own -- the dylib is only in its CcInfo linking context -- so it isolates that collection. Adds two cases over an imported dylib: a rust_binary (which reaches it through `deps`) and a rust_test (through `crate`). The rust_test case fails without the preceding commit and passes with it. * ci: bump llvm to 0.8.4 for web.archive.org id_ URL fix llvm 0.7.7 pins the macOS SDK archive at a web.archive.org URL without the id_ modifier (web/20260430051604/...), which now serves the Wayback HTML wrapper page instead of the raw pkg. The download checksum-mismatches and the build aborts during analysis of any target that transitively fetches @macos_sdk (e.g. //test/genquery:bar_deps), skipping every test. llvm 0.8.4+ uses the id_ modifier (web/20260430051604id_/...) which returns the raw file with the correct sha256. Bump the bazel_dep to unblock CI.
Allow disabling the default hermetic macOS SDKROOT globally. Preserve explicit SDKROOT values and C/C++ toolchain SDKROOT.
Bootstrap rustc actions force flags into response files, but the minimal wrapper only substituted direct arguments. Rewrite response files with the same substitutions before invoking rustc and cover success and failure behavior. Refs hermeticbuild/rules_rs#233 Assisted-by: OpenAI Codex
The MSVC spawn API splits embedded spaces unless the caller supplies quotes. Quote generated response-file arguments when needed so the child receives the expanded path as one argv entry. Assisted-by: OpenAI Codex
dzbarsky
added a commit
that referenced
this pull request
Sep 5, 2026
Bootstrap actions pass rustc flags through @params, but the C++ bootstrap process wrapper substitutes only direct arguments. The literal remap placeholders leave execution directories in tinyjson crate metadata. Use -Zremap-cwd-prefix for bootstrap compilation so rustc resolves its own working directory. Default RUSTC_BOOTSTRAP to the bootstrap crate's name, preserving explicit environment settings. Keep the existing response files and SDKROOT expansion without changing the C++ wrapper. Add a regression test that reads the actual bootstrap tinyjson archive. The test fails before the fix. Sandboxed and local builds in different output bases now produce identical archives without execution directories. Validation: Bazel 9.1.0, Rust 1.98.0, macOS arm64; bazel test //... passed 621 tests and skipped 37 platform-incompatible tests. Refs hermeticbuild/rules_rs#233 and #46. Co-authored-by: Codex <noreply@openai.com>
Member
dzbarsky
added a commit
that referenced
this pull request
Sep 5, 2026
Bootstrap actions pass rustc flags through @params, but the C++ bootstrap process wrapper substitutes only direct arguments. The literal remap placeholders leave execution directories in tinyjson crate metadata. Use -Zremap-cwd-prefix for bootstrap compilation so rustc resolves its own working directory. Default RUSTC_BOOTSTRAP to the bootstrap crate's name, preserving explicit environment settings. Keep the existing response files and SDKROOT expansion without changing the C++ wrapper. Add a regression test that reads the actual bootstrap tinyjson archive. The test fails before the fix. Sandboxed and local builds in different output bases now produce identical archives without execution directories. Validation: Bazel 9.1.0, Rust 1.98.0, macOS arm64; bazel test //... passed 621 tests and skipped 37 platform-incompatible tests. Refs hermeticbuild/rules_rs#233 and #46. Co-authored-by: Codex <noreply@openai.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
11 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
${pwd},${output_base}, and${exec_root}inside bootstrap@responsefiles before invoking rustc..expandedresponse file while leaving Bazel's generated input untouched.Problem
Bootstrap rustc actions force compiler flags through an
@paramsfile.bootstrap_process_wrappercurrently substitutes placeholders only in direct arguments, so remap flags inside the response file remain literal. rustc can then embed output-base or sandbox paths intinyjsonmetadata, making clean builds produce different.rlibbytes.Refs hermeticbuild/rules_rs#233.
Validation
bazelisk test //test/process_wrapper_bootstrap:bootstrap_process_wrapper_test --test_output=errors --nocache_test_resultsbazelisk test //util/process_wrapper:process_wrapper_test --test_output=errors --nocache_test_resultsbazelisk test //... --test_output=errors(600 passed, 30 platform-specific tests skipped on Linux)tinyjsonartifact from two fresh output bases under both local andlinux-sandboxexecution. Both pairs had identical SHA-256 hashes, and neither artifact contained an output-base or sandbox path.pre-commit run --files util/process_wrapper/private/bootstrap_process_wrapper.cc test/process_wrapper_bootstrap/bootstrap_process_wrapper_test.rs