From 4f5c16d62f250024a111aaf827927c32bc88e4da Mon Sep 17 00:00:00 2001 From: Manuel Drehwald Date: Wed, 4 Sep 2024 15:25:43 -0400 Subject: [PATCH] Enzyme backend Co-authored-by: Lorenz Schmidt --- .gitmodules | 4 + config.example.toml | 3 + src/bootstrap/configure.py | 1 + src/bootstrap/src/core/build_steps/compile.rs | 22 +++++ src/bootstrap/src/core/build_steps/llvm.rs | 95 +++++++++++++++++++ src/bootstrap/src/core/builder.rs | 7 ++ src/bootstrap/src/core/config/config.rs | 8 ++ src/bootstrap/src/lib.rs | 10 ++ src/bootstrap/src/utils/change_tracker.rs | 5 + src/tools/enzyme | 1 + src/tools/tidy/config/black.toml | 1 + src/tools/tidy/config/ruff.toml | 2 + src/tools/tidy/src/walk.rs | 1 + 13 files changed, 160 insertions(+) create mode 160000 src/tools/enzyme diff --git a/.gitmodules b/.gitmodules index b5250d49386..926807336d7 100644 --- a/.gitmodules +++ b/.gitmodules @@ -47,3 +47,7 @@ path = src/tools/rustc-perf url = https://github.com/rust-lang/rustc-perf.git shallow = true +[submodule "src/tools/enzyme"] + path = src/tools/enzyme + url = https://github.com/EnzymeAD/Enzyme.git + shallow = true diff --git a/config.example.toml b/config.example.toml index 13f76933b16..e9433c9c9bd 100644 --- a/config.example.toml +++ b/config.example.toml @@ -78,6 +78,9 @@ # Indicates whether the LLVM plugin is enabled or not #plugins = false +# Wheter to build Enzyme as AutoDiff backend. +#enzyme = false + # Indicates whether ccache is used when building LLVM. Set to `true` to use the first `ccache` in # PATH, or set an absolute path to use a specific version. #ccache = false diff --git a/src/bootstrap/configure.py b/src/bootstrap/configure.py index 768aac912ce..49d564642bd 100755 --- a/src/bootstrap/configure.py +++ b/src/bootstrap/configure.py @@ -71,6 +71,7 @@ v("llvm-libunwind", "rust.llvm-libunwind", "use LLVM libunwind") # channel, etc. o("optimize-llvm", "llvm.optimize", "build optimized LLVM") o("llvm-assertions", "llvm.assertions", "build LLVM with assertions") +o("llvm-enzyme", "llvm.enzyme", "build LLVM with enzyme") o("llvm-plugins", "llvm.plugins", "build LLVM with plugin interface") o("debug-assertions", "rust.debug-assertions", "build with debugging assertions") o("debug-assertions-std", "rust.debug-assertions-std", "build the standard library with debugging assertions") diff --git a/src/bootstrap/src/core/build_steps/compile.rs b/src/bootstrap/src/core/build_steps/compile.rs index eef548033f1..8467f2e2111 100644 --- a/src/bootstrap/src/core/build_steps/compile.rs +++ b/src/bootstrap/src/core/build_steps/compile.rs @@ -1189,6 +1189,10 @@ pub fn rustc_cargo_env( cargo.env("RUSTC_VERIFY_LLVM_IR", "1"); } + if builder.config.llvm_enzyme { + cargo.rustflag("--cfg=llvm_enzyme"); + } + // Note that this is disabled if LLVM itself is disabled or we're in a check // build. If we are in a check build we still go ahead here presuming we've // detected that LLVM is already built and good to go which helps prevent @@ -1784,6 +1788,24 @@ impl Step for Assemble { // use that to bootstrap this compiler forward. let mut build_compiler = builder.compiler(target_compiler.stage - 1, builder.config.build); + // Build enzyme + let enzyme_install = if builder.config.llvm_enzyme { + Some(builder.ensure(llvm::Enzyme { target: build_compiler.host })) + } else { + None + }; + + if let Some(enzyme_install) = enzyme_install { + let lib_ext = std::env::consts::DLL_EXTENSION; + let src_lib = enzyme_install.join("build/Enzyme/libEnzyme-19").with_extension(lib_ext); + let libdir = builder.sysroot_libdir(build_compiler, build_compiler.host); + let target_libdir = builder.sysroot_libdir(target_compiler, target_compiler.host); + let dst_lib = libdir.join("libEnzyme-19").with_extension(lib_ext); + let target_dst_lib = target_libdir.join("libEnzyme-19").with_extension(lib_ext); + builder.copy_link(&src_lib, &dst_lib); + builder.copy_link(&src_lib, &target_dst_lib); + } + // Build the libraries for this compiler to link to (i.e., the libraries // it uses at runtime). NOTE: Crates the target compiler compiles don't // link to these. (FIXME: Is that correct? It seems to be correct most diff --git a/src/bootstrap/src/core/build_steps/llvm.rs b/src/bootstrap/src/core/build_steps/llvm.rs index e1eea31b3bb..442638d3203 100644 --- a/src/bootstrap/src/core/build_steps/llvm.rs +++ b/src/bootstrap/src/core/build_steps/llvm.rs @@ -529,6 +529,7 @@ impl Step for Llvm { } }; + // FIXME(ZuseZ4): Do we need that for Enzyme too? // When building LLVM with LLVM_LINK_LLVM_DYLIB for macOS, an unversioned // libLLVM.dylib will be built. However, llvm-config will still look // for a versioned path like libLLVM-14.dylib. Manually create a symbolic @@ -849,6 +850,100 @@ fn get_var(var_base: &str, host: &str, target: &str) -> Option { .or_else(|| env::var_os(var_base)) } +#[derive(Debug, Copy, Clone, Hash, PartialEq, Eq)] +pub struct Enzyme { + pub target: TargetSelection, +} + +impl Step for Enzyme { + type Output = PathBuf; + const ONLY_HOSTS: bool = true; + + fn should_run(run: ShouldRun<'_>) -> ShouldRun<'_> { + run.path("src/tools/enzyme/enzyme") + } + + fn make_run(run: RunConfig<'_>) { + run.builder.ensure(Enzyme { target: run.target }); + } + + /// Compile Enzyme for `target`. + fn run(self, builder: &Builder<'_>) -> PathBuf { + builder.require_submodule( + "src/tools/enzyme", + Some("The Enzyme sources are required for autodiff."), + ); + if builder.config.dry_run() { + let out_dir = builder.enzyme_out(self.target); + return out_dir; + } + let target = self.target; + + let LlvmResult { llvm_config, .. } = builder.ensure(Llvm { target: self.target }); + + static STAMP_HASH_MEMO: OnceLock = OnceLock::new(); + let smart_stamp_hash = STAMP_HASH_MEMO.get_or_init(|| { + generate_smart_stamp_hash( + builder, + &builder.config.src.join("src/tools/enzyme"), + builder.enzyme_info.sha().unwrap_or_default(), + ) + }); + + let out_dir = builder.enzyme_out(target); + let stamp = out_dir.join("enzyme-finished-building"); + let stamp = HashStamp::new(stamp, Some(smart_stamp_hash)); + + if stamp.is_done() { + if stamp.hash.is_none() { + builder.info( + "Could not determine the Enzyme submodule commit hash. \ + Assuming that an Enzyme rebuild is not necessary.", + ); + builder.info(&format!( + "To force Enzyme to rebuild, remove the file `{}`", + stamp.path.display() + )); + } + return out_dir; + } + + builder.info(&format!("Building Enzyme for {}", target)); + t!(stamp.remove()); + let _time = helpers::timeit(builder); + t!(fs::create_dir_all(&out_dir)); + + builder + .config + .update_submodule(Path::new("src").join("tools").join("enzyme").to_str().unwrap()); + let mut cfg = cmake::Config::new(builder.src.join("src/tools/enzyme/enzyme/")); + // FIXME(ZuseZ4): Find a nicer way to use Enzyme Debug builds + //cfg.profile("Debug"); + //cfg.define("CMAKE_BUILD_TYPE", "Debug"); + configure_cmake(builder, target, &mut cfg, true, LdFlags::default(), &[]); + + // Re-use the same flags as llvm to control the level of debug information + // generated for lld. + let profile = match (builder.config.llvm_optimize, builder.config.llvm_release_debuginfo) { + (false, _) => "Debug", + (true, false) => "Release", + (true, true) => "RelWithDebInfo", + }; + + cfg.out_dir(&out_dir) + .profile(profile) + .env("LLVM_CONFIG_REAL", &llvm_config) + .define("LLVM_ENABLE_ASSERTIONS", "ON") + .define("ENZYME_EXTERNAL_SHARED_LIB", "ON") + .define("LLVM_DIR", builder.llvm_out(target)); + + cfg.build(); + + t!(stamp.write()); + out_dir + } +} + #[derive(Debug, Clone, Hash, PartialEq, Eq)] pub struct Lld { pub target: TargetSelection, diff --git a/src/bootstrap/src/core/builder.rs b/src/bootstrap/src/core/builder.rs index 304fe8da2bf..a28b19d33a6 100644 --- a/src/bootstrap/src/core/builder.rs +++ b/src/bootstrap/src/core/builder.rs @@ -798,6 +798,7 @@ impl<'a> Builder<'a> { tool::Miri, tool::CargoMiri, llvm::Lld, + llvm::Enzyme, llvm::CrtBeginEnd, tool::RustdocGUITest, tool::OptimizedDist, @@ -1588,6 +1589,12 @@ impl<'a> Builder<'a> { rustflags.arg(sysroot_str); } + // https://rust-lang.zulipchat.com/#narrow/stream/182449-t-compiler.2Fhelp/topic/.E2.9C.94.20link.20new.20library.20into.20stage1.2Frustc + if self.config.llvm_enzyme { + rustflags.arg("-l"); + rustflags.arg("Enzyme-19"); + } + let use_new_symbol_mangling = match self.config.rust_new_symbol_mangling { Some(setting) => { // If an explicit setting is given, use that diff --git a/src/bootstrap/src/core/config/config.rs b/src/bootstrap/src/core/config/config.rs index f7cc9ee7219..52c1c462788 100644 --- a/src/bootstrap/src/core/config/config.rs +++ b/src/bootstrap/src/core/config/config.rs @@ -221,6 +221,7 @@ pub struct Config { // llvm codegen options pub llvm_assertions: bool, pub llvm_tests: bool, + pub llvm_enzyme: bool, pub llvm_plugins: bool, pub llvm_optimize: bool, pub llvm_thin_lto: bool, @@ -898,6 +899,7 @@ define_config! { release_debuginfo: Option = "release-debuginfo", assertions: Option = "assertions", tests: Option = "tests", + enzyme: Option = "enzyme", plugins: Option = "plugins", ccache: Option = "ccache", static_libstdcpp: Option = "static-libstdcpp", @@ -1603,6 +1605,7 @@ impl Config { // we'll infer default values for them later let mut llvm_assertions = None; let mut llvm_tests = None; + let mut llvm_enzyme = None; let mut llvm_plugins = None; let mut debug = None; let mut debug_assertions = None; @@ -1722,6 +1725,8 @@ impl Config { config.llvm_tools_enabled = llvm_tools.unwrap_or(true); config.rustc_parallel = parallel_compiler.unwrap_or(config.channel == "dev" || config.channel == "nightly"); + config.llvm_enzyme = + llvm_enzyme.unwrap_or(config.channel == "dev" || config.channel == "nightly"); config.rustc_default_linker = default_linker; config.musl_root = musl_root.map(PathBuf::from); config.save_toolstates = save_toolstates.map(PathBuf::from); @@ -1806,6 +1811,7 @@ impl Config { release_debuginfo, assertions, tests, + enzyme, plugins, ccache, static_libstdcpp, @@ -1839,6 +1845,7 @@ impl Config { set(&mut config.ninja_in_file, ninja); llvm_assertions = assertions; llvm_tests = tests; + llvm_enzyme = enzyme; llvm_plugins = plugins; set(&mut config.llvm_optimize, optimize_toml); set(&mut config.llvm_thin_lto, thin_lto); @@ -2055,6 +2062,7 @@ impl Config { config.llvm_assertions = llvm_assertions.unwrap_or(false); config.llvm_tests = llvm_tests.unwrap_or(false); + config.llvm_enzyme = llvm_enzyme.unwrap_or(false); config.llvm_plugins = llvm_plugins.unwrap_or(false); config.rust_optimize = optimize.unwrap_or(RustOptimize::Bool(true)); diff --git a/src/bootstrap/src/lib.rs b/src/bootstrap/src/lib.rs index 5751c398f30..624246479f7 100644 --- a/src/bootstrap/src/lib.rs +++ b/src/bootstrap/src/lib.rs @@ -77,6 +77,9 @@ const LLD_FILE_NAMES: &[&str] = &["ld.lld", "ld64.lld", "lld-link", "wasm-ld"]; #[allow(clippy::type_complexity)] // It's fine for hard-coded list and type is explained above. const EXTRA_CHECK_CFGS: &[(Option, &str, Option<&[&'static str]>)] = &[ (None, "bootstrap", None), + (Some(Mode::Rustc), "llvm_enzyme", None), + (Some(Mode::Codegen), "llvm_enzyme", None), + (Some(Mode::ToolRustc), "llvm_enzyme", None), (Some(Mode::Rustc), "parallel_compiler", None), (Some(Mode::ToolRustc), "parallel_compiler", None), (Some(Mode::ToolRustc), "rust_analyzer", None), @@ -140,6 +143,7 @@ pub struct Build { clippy_info: GitInfo, miri_info: GitInfo, rustfmt_info: GitInfo, + enzyme_info: GitInfo, in_tree_llvm_info: GitInfo, local_rebuild: bool, fail_fast: bool, @@ -306,6 +310,7 @@ impl Build { let clippy_info = GitInfo::new(omit_git_hash, &src.join("src/tools/clippy")); let miri_info = GitInfo::new(omit_git_hash, &src.join("src/tools/miri")); let rustfmt_info = GitInfo::new(omit_git_hash, &src.join("src/tools/rustfmt")); + let enzyme_info = GitInfo::new(omit_git_hash, &src.join("src/tools/enzyme")); // we always try to use git for LLVM builds let in_tree_llvm_info = GitInfo::new(false, &src.join("src/llvm-project")); @@ -393,6 +398,7 @@ impl Build { clippy_info, miri_info, rustfmt_info, + enzyme_info, in_tree_llvm_info, cc: RefCell::new(HashMap::new()), cxx: RefCell::new(HashMap::new()), @@ -740,6 +746,10 @@ impl Build { } } + fn enzyme_out(&self, target: TargetSelection) -> PathBuf { + self.out.join(&*target.triple).join("enzyme") + } + fn lld_out(&self, target: TargetSelection) -> PathBuf { self.out.join(target).join("lld") } diff --git a/src/bootstrap/src/utils/change_tracker.rs b/src/bootstrap/src/utils/change_tracker.rs index 3fec2d953d9..80ab09881fe 100644 --- a/src/bootstrap/src/utils/change_tracker.rs +++ b/src/bootstrap/src/utils/change_tracker.rs @@ -245,4 +245,9 @@ pub const CONFIG_CHANGE_HISTORY: &[ChangeInfo] = &[ severity: ChangeSeverity::Warning, summary: "Removed `rust.split-debuginfo` as it was deprecated long time ago.", }, + ChangeInfo { + change_id: 129176, + severity: ChangeSeverity::Info, + summary: "New option `llvm.enzyme` to control whether the llvm based autodiff tool (Enzyme) is built.", + }, ]; diff --git a/src/tools/enzyme b/src/tools/enzyme new file mode 160000 index 00000000000..2fe5164a242 --- /dev/null +++ b/src/tools/enzyme @@ -0,0 +1 @@ +Subproject commit 2fe5164a2423dd67ef25e2c4fb204fd06362494b diff --git a/src/tools/tidy/config/black.toml b/src/tools/tidy/config/black.toml index e73847a93ba..a1e1eef987a 100644 --- a/src/tools/tidy/config/black.toml +++ b/src/tools/tidy/config/black.toml @@ -12,5 +12,6 @@ extend-exclude = """(\ src/llvm-project/|\ src/doc/embedded-book/|\ src/tools/rustc-perf/|\ + src/tools/enzyme/|\ library/backtrace/ )""" diff --git a/src/tools/tidy/config/ruff.toml b/src/tools/tidy/config/ruff.toml index c87c17f7833..2be109010a0 100644 --- a/src/tools/tidy/config/ruff.toml +++ b/src/tools/tidy/config/ruff.toml @@ -16,6 +16,7 @@ extend-exclude = [ "src/llvm-project/", "src/doc/embedded-book/", "library/backtrace/", + "src/tools/enzyme/", "src/tools/rustc-perf/", # Hack: CI runs from a subdirectory under the main checkout "../src/doc/nomicon/", @@ -29,6 +30,7 @@ extend-exclude = [ "../src/llvm-project/", "../src/doc/embedded-book/", "../library/backtrace/", + "../src/tools/enzyme/", "../src/tools/rustc-perf/", ] diff --git a/src/tools/tidy/src/walk.rs b/src/tools/tidy/src/walk.rs index d4721b5a5fd..4a69022ebdc 100644 --- a/src/tools/tidy/src/walk.rs +++ b/src/tools/tidy/src/walk.rs @@ -24,6 +24,7 @@ pub fn filter_dirs(path: &Path) -> bool { "src/tools/rust-analyzer", "src/tools/rustc-perf", "src/tools/rustfmt", + "src/tools/enzyme", "src/doc/book", "src/doc/edition-guide", "src/doc/embedded-book",