1ab2304fd2Accepting request 1256093 from devel:languages:rustAna Guerrero2025-03-26 20:20:11 +00:00
d2e718c8d8- Update to version 0.9.1~22: * Add experimental concurrent cache support * Add tests for distributed cargo building * Fix tests for SELinux * strip CARGO_MAKEFLAGS * Allow using S3 virtual host style endpoints (#2329) * Apply clippy fixes and bump ghac from v2 to v4 bump (#2341) * prepare release 0.10.0 * feat: Bump opendal to 0.52 to support ghac v2 * Fix formatting * Add scheduler connect message * Separate sccache_heartbeat * Separate sccache_http * lru: do not fail removing file if already removed * build and run integration tests on ubuntu-24.04 * Fix and update sccache for CUDA 12.8 * cache (and dist-compile) cudafe++ invocations to ensure the original module_id file is restored and used for all PTX compilations * hash the module_id file (if it exists), --gen_module_id_file arg, and --module_id_file_name arg so PTX generated by nvcc -c and nvcc -ptx yield different hashes * remove --gen_module_id_file from internal cicc calls when using CTK<12.8 nvcc -c * update GHA workflows to CUDA 12.8 * Add bind_address config field (#2307) * Don't cache dep file (#2322) * CI: Move the the current LLVM stable (#2323) * Scheduler: oversubscribe cores just a little to make up for slack (#2315) * Pass through --remap-path-prefix argument. (#2270)William Brown2025-03-26 01:51:29 +00:00
520eb2915fAccepting request 1242944 from devel:languages:rustAna Guerrero2025-02-04 17:12:56 +00:00
9ca37e2c81- Update to version 0.9.0~13: * Add experimental concurrent cache support * chore: replace is-terminal with standard library (#2311) * prepare release 0.9.1 * Add support for caching target json specs for Rust compilation. (#2269) * Don't use CARGO_REGISTRIES_* configuration in hash keys (#2308) * fix: bring "cargo build -Zprofile" support back (#2306) * deps: Update cc * add test for #2299 * fix test_rlib_dep_reader_call failure when $CARGO_HOME isn't in the user's $HOME * Never cache the outer CUDA compilation (because nvcc -E can't be trusted). Always decompose via nvcc --dryrun, then cache and report the host compiler call as a CUDA compilation * Always use sccache's own jobserver * Adjust placement of crate dependencies and a comment Crates memmap2, object and shlex are not dist-only requirements. * Skip tests when needed features (gcs & s3) are disabledWilliam Brown2025-02-04 00:32:32 +00:00
ca497f4e6aAccepting request 1230229 from devel:languages:rustAna Guerrero2024-12-12 20:18:23 +00:00
07ef8af23d- Update to version 0.9.0~1: * Add experimental concurrent cache support * prepare release 0.9.0 (#2293) * build(deps): bump codecov/codecov-action from 4 to 5 (#2289) * feat: set user agent in webdav requests (#2284) * test CTK 11.1 in CI * add test to ensure -v|--verbose are never dist-compiled * don't generate a dist-compile command for clang-cuda * don't use leading digit in renamed file names * include the output file name in nvcc trace logs * revert adding --gen_module_id_file when --module_id_file_name is present * read NVCC_{PREPEND,APPEND}_FLAGS from the compile environment, not the server environment * add --default-stream arg, fix parsing concatenated form of nvcc -t1 * always add --gen_module_id_file if --module_id_file_name is specified * Normalize nvcc subcommand order for CTK <12.0, ensuring the DAG is parsed by inputs/outputs even if the preprocessor, cicc, and ptxas commands are out of order. * hash --gen_module_id_file and --module_id_file_name arguments * add more clang flags * compiler invocations with -v or --verbose must not be dist-compiled, since tools like CMake parse the output and expect to see client paths not dist-server paths * report total compilation count and compile times for uncached and failed compilations * ensure dist_type is reported for failed and uncached compilations * check for more host-compiler nvcc defines to accommodate older nvcc versions * Pass correct argument to rustc -Z ls * Make start_compile_task spawn a tokio task again * ensure SCCACHE_NO_CACHE calls aren't tracked as an uncacheable compilations in the stats * Attempt to fix intermittent config_from_env() test failures by ensuring envvars are reset before unwrapping/asserting * Test nvcc and clang-cuda in workflows/ci.yml * Ensure compatibility with MSVC compatibility on Windows * fix exit_status propagation, handle long nvcc language option (--x) * Detect both nvcc and host compiler version so we can safely revert to caching outer nvcc invocations * Update Nvcc compiler to support distributed compilationWilliam Brown2024-12-12 01:15:56 +00:00
5716f01e36Accepting request 1206191 from devel:languages:rustAna Guerrero2024-10-08 15:24:27 +00:00
2f8ad18f4b- Update to version 0.8.2~1: * Add experimental concurrent cache support * sccache: prepare new release * Update FreeBSD docs to refer to 14.1 (#2263) * build(deps): bump jobserver from 0.1.31 to 0.1.32 (#2255)
xiaoguang wang
2024-10-08 00:56:54 +00:00
ff3cb7653aAccepting request 1200214 from devel:languages:rustAna Guerrero2024-09-12 14:56:51 +00:00
280b72c6f4- explicitly depend on cargo to pull in latest compiler revision - Update to version 0.8.1~31: * Add experimental concurrent cache support * build(deps): bump quinn-proto from 0.11.3 to 0.11.8 (#2254) * build(deps): bump which from 6.0.0 to 6.0.3 (#2252) * Disable "warnings as errors" flag for CL preprocessing (#2251) * Add windows arm64 to the CI (#2243) * build(deps): bump log from 0.4.20 to 0.4.22 (#2218) * feat: Bump opendal to 0.48.0 (#2240) * deps: update winapi-util (#2236) * refactor: use windows-sys instead of winapi (#2233) * build(deps): bump vmactions/freebsd-vm from 1.0.7 to 1.0.8 (#2219) * Add cache hit percentage to stats (#2211) * Add custom zstd compression level (#2199) * Remove the heapsize feature * Add lost feature status * Update OpenDAL for fixing #2220 * build(deps): bump reqwest from 0.12.4 to 0.12.5 (#2215) * build(deps): bump uuid from 1.7.0 to 1.9.0 (#2213) * Update CI version of macOS * build(deps): bump zstd from 0.13.0 to 0.13.1 * build(deps): bump async-trait from 0.1.77 to 0.1.80 * build(deps): bump anyhow from 1.0.81 to 1.0.86 * Format toml * refactor: Bump OpenDAL to 0.47 * Remove ref_env * Fix clippy * Fix clippy * refactor: Bump MSRV to 1.75 for async in trait * Log when preprocessor cache cannot be used (#2197)William Brown2024-09-12 00:41:32 +00:00
Dominique Leuenberger
William Brown
Ana Guerrero
xiaoguang wang