commit d37d6b3f8c61a24a3b2f259c22817b69525bf27d
parent c77e51615803de12a2838a0c3b72b9f850e4af86
Author: Ryan Sepassi <rsepassi@gmail.com>
Date: Sun, 24 May 2026 10:47:00 -0700
Integrate self-hosted bootstrap
Build bootstrap stages through the normal Makefile with cfree busybox-style tool symlinks, and remove the old stage2 link script.
Stage libcfree archive inputs with unique member names so cfree ar does not replace same-basename objects, and keep rt outputs under BUILD_DIR.
Tighten SSA slot promotion checks so mismatched load/store operand types do not get promoted.
Diffstat:
4 files changed, 75 insertions(+), 248 deletions(-)
diff --git a/Makefile b/Makefile
@@ -1,6 +1,9 @@
CC = clang
+AR = ar
+LD = ld
+BUILD_DIR ?= build
SYSROOT = $(shell xcrun --show-sdk-path)
-HOST_OPTFLAGS ?= -O2
+HOST_OPTFLAGS ?= -O1
# -isysroot lives in its own var so stage/bootstrap recipes can override
# host SDK handling when cfree is used as the compiler.
@@ -61,35 +64,36 @@ LANG_TOY_SRCS = $(wildcard lang/toy/*.c)
LANG_OBJS =
ifeq ($(CFREE_LANG_CPP_ENABLED),1)
-LANG_OBJS += $(patsubst lang/cpp/%.c,build/lang/cpp/%.o,$(LANG_CPP_SRCS))
+LANG_OBJS += $(patsubst lang/cpp/%.c,$(BUILD_DIR)/lang/cpp/%.o,$(LANG_CPP_SRCS))
endif
ifeq ($(CFREE_LANG_C_ENABLED),1)
-LANG_OBJS += $(patsubst lang/c/%.c,build/lang/c/%.o,$(LANG_C_SRCS))
+LANG_OBJS += $(patsubst lang/c/%.c,$(BUILD_DIR)/lang/c/%.o,$(LANG_C_SRCS))
endif
ifeq ($(CFREE_LANG_WASM_ENABLED),1)
-LANG_OBJS += $(patsubst lang/wasm/%.c,build/lang/wasm/%.o,$(LANG_WASM_SRCS))
+LANG_OBJS += $(patsubst lang/wasm/%.c,$(BUILD_DIR)/lang/wasm/%.o,$(LANG_WASM_SRCS))
endif
ifeq ($(CFREE_LANG_TOY_ENABLED),1)
-LANG_OBJS += $(patsubst lang/toy/%.c,build/lang/toy/%.o,$(LANG_TOY_SRCS))
+LANG_OBJS += $(patsubst lang/toy/%.c,$(BUILD_DIR)/lang/toy/%.o,$(LANG_TOY_SRCS))
endif
LIB_ASMS = $(shell find src -name '*.S')
-LIB_OBJS = $(patsubst src/%.c,build/lib/%.o,$(LIB_SRCS)) \
+LIB_OBJS = $(patsubst src/%.c,$(BUILD_DIR)/lib/%.o,$(LIB_SRCS)) \
$(LANG_OBJS) \
- $(patsubst src/%.S,build/lib/%.o,$(LIB_ASMS))
+ $(patsubst src/%.S,$(BUILD_DIR)/lib/%.o,$(LIB_ASMS))
LIB_DEPS = $(LIB_OBJS:.o=.d)
+LIB_AR_STAGING = $(BUILD_DIR)/ar/libcfree
DRIVER_SRCS = $(wildcard driver/*.c)
ifneq ($(CFREE_LANG_CPP_ENABLED),1)
DRIVER_SRCS := $(filter-out driver/cpp.c,$(DRIVER_SRCS))
endif
-DRIVER_OBJS = $(patsubst driver/%.c,build/driver/%.o,$(DRIVER_SRCS))
+DRIVER_OBJS = $(patsubst driver/%.c,$(BUILD_DIR)/driver/%.o,$(DRIVER_SRCS))
DRIVER_DEPS = $(DRIVER_OBJS:.o=.d)
-LIB_AR = build/libcfree.a
-BIN = build/cfree
+LIB_AR = $(BUILD_DIR)/libcfree.a
+BIN = $(BUILD_DIR)/cfree
-.PHONY: all lib bin format clean bootstrap bench-opt
+.PHONY: all lib bin format clean bootstrap bootstrap-test-toy bench-opt
all: lib bin
@@ -102,74 +106,98 @@ bin: $(BIN)
$(LIB_AR): $(LIB_OBJS)
@mkdir -p $(dir $@)
@rm -f $@
- ar rcs $@ $(LIB_OBJS)
+ @rm -rf $(LIB_AR_STAGING)
+ @mkdir -p $(LIB_AR_STAGING)
+ @set -e; \
+ for obj in $(LIB_OBJS); do \
+ rel=$${obj#$(BUILD_DIR)/}; \
+ name=$$(printf '%s' "$$rel" | tr '/' '_'); \
+ case "$$obj" in \
+ /*) target="$$obj" ;; \
+ *) target="$$PWD/$$obj" ;; \
+ esac; \
+ ln -sf "$$target" "$(LIB_AR_STAGING)/$$name"; \
+ done
+ $(AR) rcs $@ $(LIB_AR_STAGING)/*.o
$(BIN): $(DRIVER_OBJS) $(LIB_AR)
$(CC) $(HOST_SYSROOT_LDFLAGS) -o $@ $(DRIVER_OBJS) $(LIB_AR)
-build/lib/%.o: src/%.c Makefile
+$(BUILD_DIR)/lib/%.o: src/%.c Makefile
@mkdir -p $(dir $@)
$(CC) $(LIB_CFLAGS) $(DEPFLAGS) -c $< -o $@
# lang_registry.c is the one libcfree source that crosses into lang/*; it
# uses -Ilang so the frontend headers can be reached as "c/c.h" etc.
-build/lib/api/lang_registry.o: src/api/lang_registry.c Makefile
+$(BUILD_DIR)/lib/api/lang_registry.o: src/api/lang_registry.c Makefile
@mkdir -p $(dir $@)
$(CC) $(LIB_CFLAGS) -Ilang $(DEPFLAGS) -c $< -o $@
-build/lang/cpp/%.o: lang/cpp/%.c Makefile
+$(BUILD_DIR)/lang/cpp/%.o: lang/cpp/%.c Makefile
@mkdir -p $(dir $@)
$(CC) $(CFLAGS_COMMON) -ffreestanding -Iinclude -Ilang/cpp $(DEPFLAGS) -c $< -o $@
# The C frontend includes the lexer and preprocessor headers (pp/pp.h,
# lex/lex.h) which now live under lang/cpp/, and cpp_support.h is the
# shared substrate. So lang/c objects build with -Ilang/cpp -Ilang/c.
-build/lang/c/%.o: lang/c/%.c Makefile
+$(BUILD_DIR)/lang/c/%.o: lang/c/%.c Makefile
@mkdir -p $(dir $@)
$(CC) $(CFLAGS_COMMON) -ffreestanding -Iinclude -Ilang/cpp -Ilang/c $(DEPFLAGS) -c $< -o $@
-build/lang/wasm/%.o: lang/wasm/%.c Makefile
+$(BUILD_DIR)/lang/wasm/%.o: lang/wasm/%.c Makefile
@mkdir -p $(dir $@)
$(CC) $(CFLAGS_COMMON) -ffreestanding -Iinclude -Ilang/wasm $(DEPFLAGS) -c $< -o $@
-build/lib/%.o: src/%.S Makefile
+$(BUILD_DIR)/lib/%.o: src/%.S Makefile
@mkdir -p $(dir $@)
$(CC) $(LIB_CFLAGS) $(DEPFLAGS) -c $< -o $@
-build/driver/%.o: driver/%.c Makefile
+$(BUILD_DIR)/driver/%.o: driver/%.c Makefile
@mkdir -p $(dir $@)
$(CC) $(DRIVER_CFLAGS) $(DEPFLAGS) -c $< -o $@
-build/lang/toy/%.o: lang/toy/%.c Makefile
+$(BUILD_DIR)/lang/toy/%.o: lang/toy/%.c Makefile
@mkdir -p $(dir $@)
$(CC) $(CFLAGS_COMMON) -ffreestanding -Iinclude -Ilang/toy $(DEPFLAGS) -c $< -o $@
include rt/Makefile
-# Bootstrap: build cfree with the host compiler as stage 1, rebuild it with
-# stage 1 as stage 2, rebuild it again with stage 2 as stage 3, then require
-# stages 2 and 3 to be bitwise identical.
-STAGE1_BIN = build/cfree-stage1
-BOOTSTRAP_STAGE2_OUT = build/stage2-probe
-BOOTSTRAP_STAGE3_OUT = build/stage3-probe
-BOOTSTRAP_STAGE2_BIN = build/cfree-stage2
-BOOTSTRAP_STAGE3_BIN = build/cfree-stage3
+# Bootstrap: build cfree with the host compiler as stage 1, then rebuild it
+# twice through the normal Makefile using cfree's busybox-style cc/ar/ld
+# symlinks. Stages 2 and 3 must be bitwise identical.
+BOOTSTRAP_DIR = $(BUILD_DIR)/bootstrap
+BOOTSTRAP_STAGE1_DIR = $(BOOTSTRAP_DIR)/stage1
+BOOTSTRAP_STAGE2_DIR = $(BOOTSTRAP_DIR)/stage2
+BOOTSTRAP_STAGE3_DIR = $(BOOTSTRAP_DIR)/stage3
+BOOTSTRAP_STAGE1_BIN = $(BOOTSTRAP_STAGE1_DIR)/cfree
+BOOTSTRAP_STAGE2_BIN = $(BOOTSTRAP_STAGE2_DIR)/cfree
+BOOTSTRAP_STAGE3_BIN = $(BOOTSTRAP_STAGE3_DIR)/cfree
+BOOTSTRAP_TOOLS = cc ld ar ranlib as
bootstrap: $(BIN) rt
- cp $(BIN) $(STAGE1_BIN)
- rm -rf $(BOOTSTRAP_STAGE2_OUT) $(BOOTSTRAP_STAGE3_OUT) \
- $(BOOTSTRAP_STAGE2_BIN) $(BOOTSTRAP_STAGE3_BIN)
- CFREE_STAGE_BIN='$(abspath $(STAGE1_BIN))' \
- CFREE_STAGE_OUT='$(abspath $(BOOTSTRAP_STAGE2_OUT))' \
- CFREE_STAGE_OUTPUT='$(abspath $(BOOTSTRAP_STAGE2_BIN))' \
- scripts/stage2_link.sh
- CFREE_STAGE_BIN='$(abspath $(BOOTSTRAP_STAGE2_BIN))' \
- CFREE_STAGE_OUT='$(abspath $(BOOTSTRAP_STAGE3_OUT))' \
- CFREE_STAGE_OUTPUT='$(abspath $(BOOTSTRAP_STAGE3_BIN))' \
- scripts/stage2_link.sh
+ rm -rf $(BOOTSTRAP_DIR)
+ @mkdir -p $(BOOTSTRAP_STAGE1_DIR)
+ cp $(BIN) $(BOOTSTRAP_STAGE1_BIN)
+ @for tool in $(BOOTSTRAP_TOOLS); do ln -sf cfree "$(BOOTSTRAP_STAGE1_DIR)/$$tool"; done
+ $(MAKE) lib bin \
+ BUILD_DIR='$(abspath $(BOOTSTRAP_STAGE2_DIR))' \
+ HOST_OPTFLAGS='$(HOST_OPTFLAGS)' \
+ CC='$(abspath $(BOOTSTRAP_STAGE1_DIR))/cc' \
+ AR='$(abspath $(BOOTSTRAP_STAGE1_DIR))/ar' \
+ LD='$(abspath $(BOOTSTRAP_STAGE1_DIR))/ld'
+ @for tool in $(BOOTSTRAP_TOOLS); do ln -sf cfree "$(BOOTSTRAP_STAGE2_DIR)/$$tool"; done
+ $(MAKE) lib bin \
+ BUILD_DIR='$(abspath $(BOOTSTRAP_STAGE3_DIR))' \
+ HOST_OPTFLAGS='$(HOST_OPTFLAGS)' \
+ CC='$(abspath $(BOOTSTRAP_STAGE2_DIR))/cc' \
+ AR='$(abspath $(BOOTSTRAP_STAGE2_DIR))/ar' \
+ LD='$(abspath $(BOOTSTRAP_STAGE2_DIR))/ld'
cmp $(BOOTSTRAP_STAGE2_BIN) $(BOOTSTRAP_STAGE3_BIN)
shasum -a 256 $(BOOTSTRAP_STAGE2_BIN) $(BOOTSTRAP_STAGE3_BIN)
+bootstrap-test-toy: bootstrap
+ @CFREE='$(abspath $(BOOTSTRAP_STAGE3_BIN))' test/toy/run.sh
+
bench-opt: bin
@bash scripts/opt_bench.sh
@@ -177,7 +205,7 @@ format:
find src include driver lang test rt -path test/pp -prune -o \( -name '*.c' -o -name '*.h' \) -print | xargs clang-format -i --style=google
clean:
- rm -rf build
+ rm -rf $(BUILD_DIR)
-include $(LIB_DEPS)
-include $(DRIVER_DEPS)
diff --git a/rt/Makefile b/rt/Makefile
@@ -5,7 +5,7 @@ RT_AR ?= $(BIN) ar
RT_AS ?= $(BIN) as
RT_AS_COMPILE_FLAGS ?=
-RT_BUILD_DIR = build/rt
+RT_BUILD_DIR ?= $(BUILD_DIR)/rt
RT_COMMON_CFLAGS = -Werror
RT_LIB_INCS = -Irt/lib/include/common -Irt/lib/impl
diff --git a/scripts/stage2_link.sh b/scripts/stage2_link.sh
@@ -1,205 +0,0 @@
-#!/usr/bin/env bash
-# Stage-2 standalone link probe.
-#
-# 1. Compile every src/**/*.c with cfree-stage1.
-# 2. Compile every driver/*.c with cfree-stage1.
-# 3. Link all the resulting objects with `cfree ld` against libSystem.B.tbd.
-#
-# Runs out-of-tree under build/stage2-probe/ so the Makefile's build/ tree
-# is left alone.
-set -u
-
-ROOT="$(cd "$(dirname "$0")/.." && pwd)"
-cd "$ROOT"
-
-BIN="${CFREE_STAGE_BIN:-$ROOT/build/cfree}"
-if [ ! -x "$BIN" ]; then
- echo "missing $BIN — run \`make\` first" >&2
- exit 2
-fi
-
-SDK="$(xcrun --show-sdk-path)"
-OUT="${CFREE_STAGE_OUT:-$ROOT/build/stage2-probe}"
-LIB_OUT="$OUT/lib"
-LANG_CPP_OUT="$OUT/lang/cpp"
-LANG_C_OUT="$OUT/lang/c"
-LANG_WASM_OUT="$OUT/lang/wasm"
-LANG_TOY_OUT="$OUT/lang/toy"
-DRV_OUT="$OUT/driver"
-LOG="$OUT/log"
-mkdir -p "$LIB_OUT" "$LANG_CPP_OUT" "$LANG_C_OUT" "$LANG_WASM_OUT" \
- "$LANG_TOY_OUT" "$DRV_OUT" "$LOG"
-
-CFREE_FLAGS="--support-dir $ROOT -Iinclude -Isrc"
-# The lexer/preprocessor substrate (lang/cpp) is part of libcfree.
-LANG_CPP_FLAGS="--support-dir $ROOT -Iinclude -Ilang/cpp"
-# lang/c sources reach the lexer/preprocessor and cpp_support.h substrate under
-# lang/cpp, mirroring the Makefile's `-Ilang/cpp -Ilang/c`.
-LANG_C_FLAGS="--support-dir $ROOT -Iinclude -Ilang/cpp -Ilang/c"
-LANG_WASM_FLAGS="--support-dir $ROOT -Iinclude -Ilang/wasm"
-LANG_TOY_FLAGS="--support-dir $ROOT -Iinclude"
-# `cc` reaches the C frontend header as "c/c.h"; -Ilang mirrors DRIVER_CFLAGS.
-DRIVER_FLAGS="--support-dir $ROOT -Iinclude -I. -Ilang"
-
-cfree_objs=()
-fail_src=()
-fail_lang=()
-fail_driver=()
-
-compile_with_cfree() {
- local src="$1" obj="$2" flags="$3"
- mkdir -p "$(dirname "$obj")"
- if "$BIN" cc $flags -c "$src" -o "$obj" >"$LOG/$(basename "$obj").log" 2>&1; then
- return 0
- fi
- return 1
-}
-
-echo "=== compiling src/ with cfree ==="
-while IFS= read -r src; do
- rel="${src#src/}"
- obj="$LIB_OUT/${rel%.*}.o"
- src_flags="$CFREE_FLAGS"
- # lang_registry.c is the one libcfree source that reaches into lang/* and
- # includes frontend headers as "c/c.h" — matches the Makefile's -Ilang.
- if [ "$src" = "src/api/lang_registry.c" ]; then src_flags="$src_flags -Ilang"; fi
- # emu/cpu.c pulls in the system <math.h> for sqrt; cfree's bundled freestanding
- # header set has no math.h, so reach for the SDK's like the host build does.
- if [ "$src" = "src/emu/cpu.c" ]; then src_flags="$src_flags -isystem $SDK/usr/include"; fi
- if compile_with_cfree "$src" "$obj" "$src_flags"; then
- cfree_objs+=("$obj")
- printf ' ok %s\n' "$src"
- else
- fail_src+=("$src")
- head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
- fi
-done < <(find src \( -name '*.c' -o -name '*.S' \) | sort)
-
-echo
-echo "=== compiling lang/cpp with cfree ==="
-while IFS= read -r src; do
- rel="${src#lang/cpp/}"
- obj="$LANG_CPP_OUT/${rel%.c}.o"
- if compile_with_cfree "$src" "$obj" "$LANG_CPP_FLAGS"; then
- cfree_objs+=("$obj")
- printf ' ok %s\n' "$src"
- else
- fail_lang+=("$src")
- head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
- fi
-done < <(find lang/cpp -name '*.c' | sort)
-
-echo
-echo "=== compiling lang/c with cfree ==="
-while IFS= read -r src; do
- rel="${src#lang/c/}"
- obj="$LANG_C_OUT/${rel%.c}.o"
- if compile_with_cfree "$src" "$obj" "$LANG_C_FLAGS"; then
- cfree_objs+=("$obj")
- printf ' ok %s\n' "$src"
- else
- fail_lang+=("$src")
- head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
- fi
-done < <(find lang/c -name '*.c' | sort)
-
-echo
-echo "=== compiling lang/wasm with cfree ==="
-while IFS= read -r src; do
- rel="${src#lang/wasm/}"
- obj="$LANG_WASM_OUT/${rel%.c}.o"
- if compile_with_cfree "$src" "$obj" "$LANG_WASM_FLAGS"; then
- cfree_objs+=("$obj")
- printf ' ok %s\n' "$src"
- else
- fail_lang+=("$src")
- head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
- fi
-done < <(find lang/wasm -name '*.c' 2>/dev/null | sort)
-
-echo
-echo "=== compiling lang/toy with cfree ==="
-while IFS= read -r src; do
- rel="${src#lang/toy/}"
- obj="$LANG_TOY_OUT/${rel%.c}.o"
- if compile_with_cfree "$src" "$obj" "$LANG_TOY_FLAGS"; then
- cfree_objs+=("$obj")
- printf ' ok %s\n' "$src"
- else
- fail_lang+=("$src")
- head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
- fi
-done < <(find lang/toy -name '*.c' | sort)
-
-echo
-echo "=== compiling driver/ ==="
-for src in $(ls driver/*.c | sort); do
- base="$(basename "$src")"
- obj="$DRV_OUT/${base%.c}.o"
- extra_flags=""
- if [ "$base" = "env.c" ]; then extra_flags="--sysroot $SDK -lc"; fi
- if compile_with_cfree "$src" "$obj" "$DRIVER_FLAGS $extra_flags"; then
- cfree_objs+=("$obj")
- printf ' ok %s\n' "$src"
- else
- fail_driver+=("$src (cfree)")
- head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
- fi
-done
-
-echo
-echo "=== compile summary ==="
-echo " cfree objects: ${#cfree_objs[@]}"
-echo " src failures: ${#fail_src[@]}"
-echo " lang failures: ${#fail_lang[@]}"
-echo " driver failures: ${#fail_driver[@]}"
-
-if [ "${#fail_src[@]}" -gt 0 ] || [ "${#fail_lang[@]}" -gt 0 ] || \
- [ "${#fail_driver[@]}" -gt 0 ]; then
- echo
- echo "compile failures present; skipping link" >&2
- exit 1
-fi
-
-echo
-echo "=== linking with cfree ld ==="
-BIN_OUT="${CFREE_STAGE_OUTPUT:-$OUT/cfree-stage2}"
-LIBSYS_DIR="$SDK/usr/lib"
-if [ ! -f "$LIBSYS_DIR/libSystem.B.tbd" ] && \
- [ ! -f "$LIBSYS_DIR/libSystem.tbd" ]; then
- echo "libSystem stub not found under $LIBSYS_DIR" >&2
- exit 2
-fi
-
-# cfree lowers some operations (e.g. 128-bit multiply -> __multi3) to
-# compiler-rt-style builtins that libSystem does not provide. The host build
-# resolves these from clang's compiler-rt; here we link cfree's own runtime for
-# the host architecture (built by `make rt`).
-case "$(uname -m)" in
- arm64|aarch64) RT_TRIPLE=aarch64-apple-darwin ;;
- x86_64) RT_TRIPLE=x86_64-apple-darwin ;;
- *) echo "unknown host arch $(uname -m) for runtime selection" >&2; exit 2 ;;
-esac
-RT_LIB="$ROOT/build/rt/$RT_TRIPLE/libcfree_rt.a"
-if [ ! -f "$RT_LIB" ]; then
- echo "missing $RT_LIB — run \`make rt\` first" >&2
- exit 2
-fi
-
-set -x
-"$BIN" ld -o "$BIN_OUT" -pie \
- "${cfree_objs[@]}" \
- "$RT_LIB" \
- -L "$LIBSYS_DIR" -lSystem
-status=$?
-set +x
-
-if [ "$status" -ne 0 ]; then
- echo "cfree ld failed (exit $status)" >&2
- exit "$status"
-fi
-
-echo
-echo "=== link succeeded ==="
-file "$BIN_OUT"
-ls -l "$BIN_OUT"
diff --git a/src/opt/pass_ssa.c b/src/opt/pass_ssa.c
@@ -131,7 +131,9 @@ static int slot_access_promotable(const Func* f, const Inst* in,
* slot's declared type. */
const IRFrameSlot* s = &f->frame_slots[slot_id - 1u];
CfreeCgTypeId at = in->extra.mem.type;
- return !at || at == s->type;
+ if (at && at != s->type) return 0;
+ if (in->opnds[0].type && in->opnds[0].type != s->type) return 0;
+ return 1;
}
if ((IROp)in->op == IR_STORE) {
if (in->nopnds < 2 || opnd_slot_id(&in->opnds[0]) != slot_id) return 1;
@@ -140,7 +142,9 @@ static int slot_access_promotable(const Func* f, const Inst* in,
return 0;
const IRFrameSlot* s = &f->frame_slots[slot_id - 1u];
CfreeCgTypeId at = in->extra.mem.type;
- return !at || at == s->type;
+ if (at && at != s->type) return 0;
+ if (in->opnds[1].type && in->opnds[1].type != s->type) return 0;
+ return 1;
}
for (u32 i = 0; i < in->nopnds; ++i)
if (opnd_slot_id(&in->opnds[i]) == slot_id) return 0;
