# -*- mode: makefile-gmake; indent-tabs-mode: t -*- # Put no definitions when ZJIT isn't configured ifneq ($(ZJIT_SUPPORT),no) ZJIT_SRC_FILES = $(wildcard \ $(top_srcdir)/zjit/Cargo.* \ $(top_srcdir)/zjit/src/*.rs \ $(top_srcdir)/zjit/src/*/*.rs \ $(top_srcdir)/zjit/src/*/*/*.rs \ $(top_srcdir)/zjit/src/*/*/*/*.rs \ $(top_srcdir)/jit/src/lib.rs \ ) $(RUST_LIB): $(ZJIT_SRC_FILES) # Absolute path to match RUST_LIB rules to avoid picking # the "target" dir in the source directory through VPATH. BUILD_ZJIT_LIBS = $(TOP_BUILD_DIR)/$(ZJIT_LIBS) # In a ZJIT-only build (no YJIT) ifneq ($(strip $(ZJIT_LIBS)),) $(BUILD_ZJIT_LIBS): $(ZJIT_SRC_FILES) $(ECHO) 'building Rust ZJIT (release mode)' $(gnumake_recursive)$(Q) $(RUSTC) $(ZJIT_RUSTC_ARGS) else ifneq ($(strip $(RLIB_DIR)),) # combo build # Absolute path to avoid VPATH ambiguity ZJIT_RLIB = $(TOP_BUILD_DIR)/$(RLIB_DIR)/libzjit.rlib $(ZJIT_RLIB): $(ZJIT_SRC_FILES) $(ECHO) 'building $(@F)' $(gnumake_recursive)$(Q) $(RUSTC) '-L$(@D)' --extern=jit $(ZJIT_RUSTC_ARGS) $(RUST_LIB): $(ZJIT_RLIB) endif # ifneq ($(strip $(ZJIT_LIBS)),) # By using ZJIT_BENCH_OPTS instead of RUN_OPTS, you can skip passing the options to `make install` ZJIT_BENCH_OPTS = $(RUN_OPTS) --enable-gems ZJIT_BENCH = benchmarks/railsbench/benchmark.rb # Run zjit-bench's ./run_once.sh for CI zjit-bench: install update-zjit-bench PHONY $(Q) cd $(srcdir)/zjit-bench && PATH=$(prefix)/bin:$$PATH \ ./run_once.sh $(ZJIT_BENCH_OPTS) $(ZJIT_BENCH) update-zjit-bench: $(Q) $(tooldir)/git-refresh -C $(srcdir) --branch main \ https://github.com/Shopify/zjit-bench zjit-bench $(GIT_OPTS) # Gives quick feedback about ZJIT. Not a replacement for a full test run. .PHONY: zjit-check zjit-check: $(MAKE) zjit-test $(MAKE) test-all TESTS='$(top_srcdir)/test/ruby/test_zjit.rb' ZJIT_BINDGEN_DIFF_OPTS = # Generate Rust bindings. See source for details. # Needs `./configure --enable-zjit=dev` and Clang. ifneq ($(strip $(CARGO)),) # if configure found Cargo .PHONY: zjit-bindgen zjit-bindgen-show-unused zjit-test zjit-test-update .PHONY: zjit-test-debug zjit-test-lldb zjit-test-gdb zjit-test-rr zjit-bindgen: zjit.$(OBJEXT) ZJIT_SRC_ROOT_PATH='$(top_srcdir)' BINDGEN_JIT_NAME=zjit $(CARGO) run --manifest-path '$(top_srcdir)/zjit/bindgen/Cargo.toml' -- $(CFLAGS) $(XCFLAGS) $(CPPFLAGS) $(Q) if [ 'x$(HAVE_GIT)' = xyes ]; then $(GIT) -C "$(top_srcdir)" diff $(ZJIT_BINDGEN_DIFF_OPTS) zjit/src/cruby_bindings.inc.rs; fi # Build env should roughly match what's used for miniruby to help with caching. ZJIT_NEXTEST_ENV := RUBY_BUILD_DIR='$(TOP_BUILD_DIR)' \ RUBY_LD_FLAGS='$(LDFLAGS) $(XLDFLAGS) $(MAINLIBS)' \ MACOSX_DEPLOYMENT_TARGET=11.0 \ CARGO_TARGET_DIR='$(CARGO_TARGET_DIR)' # We need `cargo nextest` for its one-process-per execution execution model # since we can only boot the VM once per process. Normal `cargo test` # runs tests in threads and can't handle this. # # On darwin, it's available through `brew install cargo-nextest`. See # https://nexte.st/docs/installation/pre-built-binaries/ otherwise. zjit-test: libminiruby.a @set +e; \ $(ZJIT_NEXTEST_ENV) $(CARGO) nextest run \ --manifest-path '$(top_srcdir)/zjit/Cargo.toml' \ --no-fail-fast \ '--features=$(ZJIT_TEST_FEATURES)' \ $(ZJIT_TESTS); \ exit_code=$$?; \ if [ -f '$(top_srcdir)/zjit/src/.hir.rs.pending-snap' ]; then \ echo ""; \ echo "Pending snapshots found. Accept with: make zjit-test-update"; \ fi; \ exit $$exit_code # Accept all pending snapshots (requires cargo-insta) # Install with: cargo install cargo-insta zjit-test-update: @$(CARGO) insta --version >/dev/null 2>&1 || { echo "Error: cargo-insta is not installed. Install with: cargo install cargo-insta"; exit 1; } @$(CARGO) insta accept --manifest-path '$(top_srcdir)/zjit/Cargo.toml' ZJIT_DEBUGGER = ZJIT_DEBUGGER_OPTS = # Run a ZJIT test written with Rust #[test] under $(ZJIT_DEBUGGER) zjit-test-debug: libminiruby.a $(Q)set -eu; \ if [ -z '$(ZJIT_TESTS)' ]; then \ echo "Please pass a ZJIT_TESTS=... filter to make."; \ echo "Many tests only work when it's the only test in the process."; \ exit 1; \ fi; \ exe_path=`$(ZJIT_NEXTEST_ENV) \ $(CARGO) nextest list --manifest-path '$(top_srcdir)/zjit/Cargo.toml' --message-format json --list-type=binaries-only | \ $(BASERUBY) -rjson -e 'puts JSON.load(STDIN.read).dig("rust-binaries", "zjit", "binary-path")'`; \ exec $(ZJIT_DEBUGGER) $$exe_path $(ZJIT_DEBUGGER_OPTS) --test-threads=1 $(ZJIT_TESTS) # Run a ZJIT test written with Rust #[test] under LLDB zjit-test-lldb: $(Q) $(MAKE) zjit-test-debug ZJIT_DEBUGGER=lldb ZJIT_DEBUGGER_OPTS=-- # Run a ZJIT test written with Rust #[test] under GDB zjit-test-gdb: libminiruby.a $(Q) $(MAKE) zjit-test-debug ZJIT_DEBUGGER="gdb --args" # Run a ZJIT test written with Rust #[test] under rr-debugger zjit-test-rr: libminiruby.a $(Q) $(MAKE) zjit-test-debug ZJIT_DEBUGGER="rr record" # A library for booting miniruby in tests. # Why not use libruby-static.a for this? # - Initialization of the full ruby involves dynamic linking for e.g. transcoding implementations # our tests don't need these functionalities so good to avoid their complexity. # - By being mini, it's faster to build # - Less likely to break since later stages of the build process also rely on miniruby. libminiruby.a: miniruby$(EXEEXT) $(ECHO) linking static-library $@ $(Q) $(AR) $(ARFLAGS) $@ $(MINIOBJS) $(COMMONOBJS) libminiruby: libminiruby.a endif # ifneq ($(strip $(CARGO)), endif # ifneq ($(ZJIT_SUPPORT),no)