commit c9156e6d621e9cd054a69a81f36cfa93015afd81
parent 9efd799a8b6b7d1a9ab4a2cae8a6abc84e1e3019
Author: Ryan Sepassi <rsepassi@gmail.com>
Date: Wed, 20 May 2026 05:18:11 -0700
Fix stage2 aggregate and aa64 bootstrap bugs
Diffstat:
14 files changed, 231 insertions(+), 21 deletions(-)
diff --git a/driver/hosted.c b/driver/hosted.c
@@ -2,6 +2,7 @@
#include <stddef.h>
#include <stdint.h>
+#include <string.h>
static char* hosted_join2(DriverEnv* env, const char* a, const char* b,
size_t* out_size) {
@@ -614,8 +615,5 @@ void driver_hosted_plan_fini(DriverEnv* env, DriverHostedPlan* plan) {
driver_free(env, plan->owned_system_includes[i],
plan->owned_system_include_sizes[i]);
}
- {
- DriverHostedPlan zero = {0};
- *plan = zero;
- }
+ memset(plan, 0, sizeof(*plan));
}
diff --git a/lang/c/parse/cg_adapter.c b/lang/c/parse/cg_adapter.c
@@ -479,6 +479,8 @@ void pcg_convert(Parser* p, const Type* dst) {
if (emit) cfree_cg_sext(p->cg, id);
} else if (ds > ss) {
if (emit) cfree_cg_zext(p->cg, id);
+ } else if (type_is_ptr(src) != type_is_ptr(dst)) {
+ if (emit) cfree_cg_bitcast(p->cg, id);
}
} else if (type_is_int(src) && df) {
if (pcg_type_is_signed(src)) {
diff --git a/lang/c/parse/parse_expr.c b/lang/c/parse/parse_expr.c
@@ -1050,7 +1050,10 @@ void to_rvalue(Parser* p) {
cg_addr(p->cg);
return;
}
- if (t->kind == TY_STRUCT || t->kind == TY_UNION) return;
+ if (t->kind == TY_STRUCT || t->kind == TY_UNION) {
+ p->cg_type_stack[p->cg_type_sp - 1u] = type_unqual(p->pool, t);
+ return;
+ }
}
cg_load(p->cg);
}
diff --git a/lang/c/parse/parse_init.c b/lang/c/parse/parse_init.c
@@ -63,6 +63,98 @@ static void init_aggregate_remainder(Parser* p, FrameSlot slot,
const Type* arr_ty, u32 offset,
const Type* ty, u32 start_index);
+static void replay_recorded_initializer_expr(Parser* p) {
+ if (p->replay_len == 0)
+ perr(p, "internal: empty initializer expression replay");
+ p->cur = p->replay[0];
+ p->replay_pos = 1;
+ p->replay_active = 1;
+ p->has_next = 0;
+}
+
+static void record_initializer_expr_for_replay(Parser* p) {
+ u32 len = 0;
+ u32 cap = 0;
+ Tok* buf = NULL;
+ int paren_depth = 0;
+ int brack_depth = 0;
+ int brace_depth = 0;
+
+ for (;;) {
+ if (len == cap) {
+ u32 new_cap = cap ? cap * 2u : 16u;
+ Tok* nb = arena_array(p->pool->arena, Tok, new_cap);
+ if (!nb) perr(p, "out of memory recording initializer expression");
+ if (buf && len) memcpy(nb, buf, len * sizeof(*buf));
+ buf = nb;
+ cap = new_cap;
+ }
+ buf[len++] = p->cur;
+
+ if (p->cur.kind == TOK_EOF) break;
+ if (paren_depth == 0 && brack_depth == 0 && brace_depth == 0 &&
+ (is_punct(&p->cur, ',') || is_punct(&p->cur, '}'))) {
+ break;
+ }
+
+ if (is_punct(&p->cur, '('))
+ ++paren_depth;
+ else if (is_punct(&p->cur, ')'))
+ --paren_depth;
+ else if (is_punct(&p->cur, '['))
+ ++brack_depth;
+ else if (is_punct(&p->cur, ']'))
+ --brack_depth;
+ else if (is_punct(&p->cur, '{'))
+ ++brace_depth;
+ else if (is_punct(&p->cur, '}'))
+ --brace_depth;
+
+ advance(p);
+ }
+
+ p->replay = buf;
+ p->replay_cap = cap;
+ p->replay_len = len;
+ replay_recorded_initializer_expr(p);
+}
+
+static int try_init_aggregate_from_expr(Parser* p, FrameSlot slot,
+ const Type* arr_ty, u32 offset,
+ const Type* ty) {
+ const Type* expr_ty;
+ u32 save_sp;
+ int compatible;
+
+ if (!ty || (ty->kind != TY_STRUCT && ty->kind != TY_UNION)) return 0;
+ if (is_punct(&p->cur, '{') || is_punct(&p->cur, '.') ||
+ is_punct(&p->cur, '[')) {
+ return 0;
+ }
+
+ record_initializer_expr_for_replay(p);
+
+ save_sp = p->cg_type_sp;
+ pcg_codegen_suppress_push(p);
+ parse_assign_expr(p);
+ expr_ty = cg_top_type(p->cg);
+ compatible =
+ type_compatible(type_unqual(p->pool, ty), type_unqual(p->pool, expr_ty));
+ p->cg_type_sp = save_sp;
+ pcg_codegen_suppress_pop(p);
+
+ replay_recorded_initializer_expr(p);
+ if (!compatible) return 0;
+
+ parse_assign_expr(p);
+ if (!type_compatible(type_unqual(p->pool, ty),
+ type_unqual(p->pool, cg_top_type(p->cg)))) {
+ perr(p, "incompatible aggregate initializer");
+ }
+ emit_struct_copy_into_slot(p, slot, arr_ty, offset, ty);
+ return 1;
+}
+
/* Push the lvalue of a sub-object at byte offset `offset` within the array
* local `slot` (whose type is `arr_ty`), with element type `elem_ty`. */
void push_subobject_lv(Parser* p, FrameSlot slot, const Type* arr_ty,
@@ -582,6 +674,7 @@ void init_at(Parser* p, FrameSlot slot, const Type* arr_ty, u32 offset,
}
if (ty->kind == TY_STRUCT) {
if (!is_punct(&p->cur, '{')) {
+ if (try_init_aggregate_from_expr(p, slot, arr_ty, offset, ty)) return;
init_aggregate_remainder(p, slot, arr_ty, offset, ty, 0);
return;
}
@@ -592,6 +685,10 @@ void init_at(Parser* p, FrameSlot slot, const Type* arr_ty, u32 offset,
}
if (ty->kind == TY_UNION) {
int had_brace = accept_punct(p, '{');
+ if (!had_brace &&
+ try_init_aggregate_from_expr(p, slot, arr_ty, offset, ty)) {
+ return;
+ }
if (ty->rec.nfields == 0) {
if (had_brace) expect_punct(p, '}', "'}'");
return;
diff --git a/scripts/stage2_link.sh b/scripts/stage2_link.sh
@@ -13,30 +13,37 @@ set -u
ROOT="$(cd "$(dirname "$0")/.." && pwd)"
cd "$ROOT"
-BIN="$ROOT/build/cfree"
+BIN="${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="$ROOT/build/stage2-probe"
+OUT="${CFREE_STAGE_OUT:-$ROOT/build/stage2-probe}"
LIB_OUT="$OUT/lib"
+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" "$DRV_OUT" "$LOG"
+mkdir -p "$LIB_OUT" "$LANG_C_OUT" "$LANG_WASM_OUT" "$LANG_TOY_OUT" "$DRV_OUT" "$LOG"
CFREE_FLAGS="-isystem $ROOT/rt/include -isystem $ROOT/rt/include/libc -Iinclude -Isrc"
-DRIVER_FLAGS="-isystem $ROOT/rt/include -isystem $ROOT/rt/include/libc -Iinclude"
+LANG_C_FLAGS="-isystem $ROOT/rt/include -isystem $ROOT/rt/include/libc -Iinclude -Ilang/c"
+LANG_WASM_FLAGS="-isystem $ROOT/rt/include -isystem $ROOT/rt/include/libc -Iinclude -Ilang/wasm"
+LANG_TOY_FLAGS="-isystem $ROOT/rt/include -isystem $ROOT/rt/include/libc -Iinclude"
+DRIVER_FLAGS="-isystem $ROOT/rt/include -isystem $ROOT/rt/include/libc -Iinclude -I."
# Driver files cfree still cannot parse (doc/STAGE2.md A2). Compile with
# clang in stage-2 flag style so the resulting objects are still arm64
# Mach-O at a matching SDK level.
-CLANG_FALLBACK=("env.c" "ld.c")
+CLANG_FALLBACK=("env.c")
cfree_objs=()
clang_objs=()
fail_src=()
+fail_lang=()
fail_driver=()
compile_with_cfree() {
@@ -62,6 +69,48 @@ while IFS= read -r src; do
done < <(find src -name '*.c' | sort)
echo
+echo "=== compiling lang/c with cfree ==="
+while IFS= read -r src; do
+ rel="${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")"
@@ -74,7 +123,7 @@ for src in $(ls driver/*.c | sort); do
# -fno-{,asynchronous-}unwind-tables: cfree's macho_read doesn't yet
# ingest the section-relative UNSIGNED relocs that clang emits in
# __LD,__compact_unwind. Suppress the section entirely.
- if clang -arch arm64 -isysroot "$SDK" -Iinclude \
+ if clang -arch arm64 -isysroot "$SDK" -Iinclude -I. \
-fno-unwind-tables -fno-asynchronous-unwind-tables \
-c "$src" -o "$obj" >"$LOG/$base.log" 2>&1; then
clang_objs+=("$obj")
@@ -90,7 +139,7 @@ for src in $(ls driver/*.c | sort); do
printf ' ok %s\n' "$src"
else
fail_driver+=("$src (cfree)")
- head -1 "$LOG/$base.log" | sed "s|^| FAIL $src: |"
+ head -1 "$LOG/$(basename "$obj").log" | sed "s|^| FAIL $src: |"
fi
done
@@ -99,9 +148,11 @@ echo "=== compile summary ==="
echo " cfree objects: ${#cfree_objs[@]}"
echo " clang objects: ${#clang_objs[@]}"
echo " src failures: ${#fail_src[@]}"
+echo " lang failures: ${#fail_lang[@]}"
echo " driver failures: ${#fail_driver[@]}"
-if [ "${#fail_src[@]}" -gt 0 ] || [ "${#fail_driver[@]}" -gt 0 ]; then
+if [ "${#fail_src[@]}" -gt 0 ] || [ "${#fail_lang[@]}" -gt 0 ] || \
+ [ "${#fail_driver[@]}" -gt 0 ]; then
echo
echo "compile failures present; skipping link" >&2
exit 1
@@ -109,7 +160,7 @@ fi
echo
echo "=== linking with cfree ld ==="
-BIN_OUT="$OUT/cfree-stage2"
+BIN_OUT="${CFREE_STAGE_OUTPUT:-$OUT/cfree-stage2}"
LIBSYS_DIR="$SDK/usr/lib"
if [ ! -f "$LIBSYS_DIR/libSystem.B.tbd" ] && \
[ ! -f "$LIBSYS_DIR/libSystem.tbd" ]; then
diff --git a/src/arch/aa64/ops.c b/src/arch/aa64/ops.c
@@ -379,7 +379,7 @@ static u32 agg_addr_reg(CGTarget* t, Operand op, u32 scratch) {
AAImpl* a = impl_of(t);
AASlot* s = aa64_slot_get(a, op.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 agg: bad slot");
- aa64_emit32(t->mc, aa64_sub_imm(1, scratch, 29, s->off, 0));
+ aa64_emit_addr_adjust(t->mc, scratch, 29, -(i32)s->off);
return scratch;
}
compiler_panic(t->c, impl_of(t)->loc,
@@ -912,7 +912,7 @@ static void emit_arg_value(CGTarget* t, const ABIFuncInfo* fi,
if (av->storage.kind == OPK_LOCAL) {
AASlot* s = aa64_slot_get(a, av->storage.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 call: bad byval slot");
- aa64_emit32(t->mc, aa64_sub_imm(1, dst_reg, 29, s->off, 0));
+ aa64_emit_addr_adjust(t->mc, dst_reg, 29, -(i32)s->off);
} else if (av->storage.kind == OPK_INDIRECT) {
aa64_emit_addr_adjust(t->mc, dst_reg, av->storage.v.ind.base & 0x1f,
av->storage.v.ind.ofs);
@@ -1196,7 +1196,7 @@ static void aa_call(CGTarget* t, const CGCallDesc* d) {
}
AASlot* s = aa64_slot_get(a, d->ret.storage.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 call: bad sret slot");
- aa64_emit32(mc, aa64_sub_imm(1, 8, 29, s->off, 0));
+ aa64_emit_addr_adjust(mc, 8, 29, -(i32)s->off);
}
for (u32 i = 0; i < d->nargs; ++i) {
@@ -1778,7 +1778,7 @@ static void aa_atomic_load(CGTarget* t, Operand dst, Operand addr, MemAccess ma,
AASlot* s = aa64_slot_get(a, addr.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 atomic_load: bad slot");
base = AA_TMP0;
- aa64_emit32(mc, aa64_sub_imm(1, base, 29, s->off, 0));
+ aa64_emit_addr_adjust(mc, base, 29, -(i32)s->off);
} else {
compiler_panic(t->c, a->loc,
"aarch64 atomic_load: addr kind %d unsupported",
@@ -1816,7 +1816,7 @@ static void aa_atomic_store(CGTarget* t, Operand addr, Operand src,
AASlot* s = aa64_slot_get(a, addr.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 atomic_store: bad slot");
base = AA_TMP0;
- aa64_emit32(mc, aa64_sub_imm(1, base, 29, s->off, 0));
+ aa64_emit_addr_adjust(mc, base, 29, -(i32)s->off);
} else {
compiler_panic(t->c, a->loc,
"aarch64 atomic_store: addr kind %d unsupported",
@@ -1861,7 +1861,7 @@ static void aa_atomic_rmw(CGTarget* t, AtomicOp op, Operand dst, Operand addr,
} else if (addr.kind == OPK_LOCAL) {
AASlot* s = aa64_slot_get(a, addr.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 atomic_rmw: bad slot");
- aa64_emit32(mc, aa64_sub_imm(1, AA_TMP0, 29, s->off, 0));
+ aa64_emit_addr_adjust(mc, AA_TMP0, 29, -(i32)s->off);
} else {
compiler_panic(t->c, a->loc, "aarch64 atomic_rmw: addr kind %d unsupported",
(int)addr.kind);
@@ -1914,7 +1914,7 @@ static void aa_atomic_cas(CGTarget* t, Operand prior, Operand ok, Operand addr,
else if (addr.kind == OPK_LOCAL) {
AASlot* s = aa64_slot_get(a, addr.v.frame_slot);
if (!s) compiler_panic(t->c, a->loc, "aarch64 atomic_cas: bad slot");
- aa64_emit32(mc, aa64_sub_imm(1, AA_TMP0, 29, s->off, 0));
+ aa64_emit_addr_adjust(mc, AA_TMP0, 29, -(i32)s->off);
} else {
compiler_panic(t->c, a->loc, "aarch64 atomic_cas: addr kind %d unsupported",
(int)addr.kind);
diff --git a/test/parse/cases/6_5_15_05_cond_const_aggregate_compound.c b/test/parse/cases/6_5_15_05_cond_const_aggregate_compound.c
@@ -0,0 +1,11 @@
+typedef struct Pair {
+ int a;
+ int b;
+} Pair;
+
+int test_main(void) {
+ const Pair p = {7, 11};
+ const Pair* pp = &p;
+ Pair r = pp ? *pp : (Pair){1, 2};
+ return r.a + r.b;
+}
diff --git a/test/parse/cases/6_5_15_05_cond_const_aggregate_compound.expected b/test/parse/cases/6_5_15_05_cond_const_aggregate_compound.expected
@@ -0,0 +1 @@
+18
diff --git a/test/parse/cases/6_5_2_2_09_large_frame_struct_byval.c b/test/parse/cases/6_5_2_2_09_large_frame_struct_byval.c
@@ -0,0 +1,19 @@
+typedef struct Big {
+ int a;
+ int b;
+ int c;
+ int d;
+ int e;
+} Big;
+
+int take_big(Big b) {
+ return b.a + b.e;
+}
+
+int test_main(void) {
+ char pad[8192];
+ Big b = {3, 4, 5, 6, 7};
+ pad[0] = 1;
+ pad[8191] = 2;
+ return take_big(b) + pad[0] + pad[8191];
+}
diff --git a/test/parse/cases/6_5_2_2_09_large_frame_struct_byval.expected b/test/parse/cases/6_5_2_2_09_large_frame_struct_byval.expected
@@ -0,0 +1 @@
+13
diff --git a/test/parse/cases/6_5_4_01_int_to_ptr_deref.c b/test/parse/cases/6_5_4_01_int_to_ptr_deref.c
@@ -0,0 +1,6 @@
+int test_main(void) {
+ int x = 3;
+ unsigned long raw = (unsigned long)&x;
+ *(int*)raw = 9;
+ return x;
+}
diff --git a/test/parse/cases/6_5_4_01_int_to_ptr_deref.expected b/test/parse/cases/6_5_4_01_int_to_ptr_deref.expected
@@ -0,0 +1 @@
+9
diff --git a/test/parse/cases/6_7_9_36_designated_compound_literal_subobject.c b/test/parse/cases/6_7_9_36_designated_compound_literal_subobject.c
@@ -0,0 +1,19 @@
+typedef struct Inner {
+ int a;
+ int b;
+} Inner;
+
+typedef struct Outer {
+ int x;
+ Inner inner;
+ int y;
+} Outer;
+
+int test_main(void) {
+ Outer o = {
+ .x = 3,
+ .inner = (Inner){4, 5},
+ .y = 6,
+ };
+ return o.x + o.inner.a * 10 + o.inner.b * 20 + o.y;
+}
diff --git a/test/parse/cases/6_7_9_36_designated_compound_literal_subobject.expected b/test/parse/cases/6_7_9_36_designated_compound_literal_subobject.expected
@@ -0,0 +1 @@
+149
