w6c+selfhost: codegen for top-level mutable let

Third step toward writable globals. The C cgen and its selfhost
mirror now:

  - emit DATAW <name>(SB),"<8 LE bytes>" for every top-level `let`
    whose type lands in the scalar set (i8..i64/u8..u64/bool/rune/
    int/uint/uintptr/ptr; floats and multi-word types deferred);
  - drop the "no writable .data" silent-drop guard at the N_IDENT
    store path, replacing it with a RIP-relative MOVQ for `=` and
    a load→combine→store sequence for the compound ops; and
  - route `&name` through LEAQ name(SB) instead of dropping it.

Type aliases resolve via aliaslookup so `type counter = i32; let c:
counter = 0;` still emits a DATAW slot. Non-literal initialisers
silently skip, which surfaces as a clean undefined-symbol error if
the binding is ever referenced.

The selfhost mirror lands in the same commit because test 990
diffs the C cgen against wwdump_ww -c on err.ww (which has
top-level `let nerrors: i32 = 0; ... nerrors += 1;`). Any drift
between the two cgens makes 990 fail. Bootstrap stays at a fixed
point: ww2 == ww3 == ww4 byte-identical.
This commit is contained in:
2026-05-12 11:56:51 +09:00
parent 38e0b6510a
commit 3c812faa08
9 changed files with 943 additions and 11 deletions

View File

@@ -214,7 +214,7 @@ TESTS = $(BIN)/test_smoke $(BIN)/test_lex $(BIN)/test_parse $(BIN)/test_check \
$(BIN)/test_data_link \ $(BIN)/test_data_link \
$(BIN)/test_arch \ $(BIN)/test_arch \
$(BIN)/test_e2e $(BIN)/test_ffi $(BIN)/test_dyn $(BIN)/test_stdlib \ $(BIN)/test_e2e $(BIN)/test_ffi $(BIN)/test_dyn $(BIN)/test_stdlib \
$(BIN)/test_at_test \ $(BIN)/test_at_test $(BIN)/test_let_global \
$(BIN)/test_selfhost $(BIN)/test_w6a_ww $(BIN)/test_w6l_ww \ $(BIN)/test_selfhost $(BIN)/test_w6a_ww $(BIN)/test_w6l_ww \
$(BIN)/test_w6c_ww $(BIN)/test_ww_ww $(BIN)/test_self_rebuild \ $(BIN)/test_w6c_ww $(BIN)/test_ww_ww $(BIN)/test_self_rebuild \
$(BIN)/test_dyn_ww $(BIN)/test_selfcheck $(BIN)/test_at_test_ww $(BIN)/test_dyn_ww $(BIN)/test_selfcheck $(BIN)/test_at_test_ww
@@ -268,6 +268,10 @@ $(BIN)/test_at_test: test/wcc/910_at_test.c $(BIN)/ww $(BIN)/w6c $(BIN)/w6a \
$(BIN)/w6l $(LIB)/libwwrt.a | $(BIN) $(BIN)/w6l $(LIB)/libwwrt.a | $(BIN)
$(CC) $(CFLAGS) -o $@ $< $(CC) $(CFLAGS) -o $@ $<
$(BIN)/test_let_global: test/wcc/630_let_global.c $(BIN)/ww $(BIN)/w6c \
$(BIN)/w6a $(BIN)/w6l $(LIB)/libwwrt.a | $(BIN)
$(CC) $(CFLAGS) -o $@ $<
$(BIN)/test_selfhost: test/wcc/990_selfhost.c $(BIN)/w6c $(BIN)/w6a $(BIN)/w6l \ $(BIN)/test_selfhost: test/wcc/990_selfhost.c $(BIN)/w6c $(BIN)/w6a $(BIN)/w6l \
$(BIN)/ww $(BIN)/wwdump $(BIN)/wwdump_ww $(LIB)/libwwrt.a | $(BIN) $(BIN)/ww $(BIN)/wwdump $(BIN)/wwdump_ww $(LIB)/libwwrt.a | $(BIN)
$(CC) $(CFLAGS) -o $@ $< $(CC) $(CFLAGS) -o $@ $<

View File

@@ -386,6 +386,40 @@ struct Mod {
}; };
static Mod *mod_map; static Mod *mod_map;
/* Top-level `let` map. Populated alongside mod_map; consulted by the
* N_IDENT store path and the &-of path to route reads/writes through
* a RIP-relative reference rather than dropping them as the (pre-
* writable-.data) compiler did. emit_lets emits a DATAW for each. */
typedef struct LetVar LetVar;
struct LetVar {
const char *name;
LetVar *next;
};
static LetVar *letvars;
/* Subset of types we know how to store in 8 bytes of .data and load
* back with a plain MOVQ. Floats need MOVSS/MOVSD; str/slice/struct/
* tagged unions are multi-word; enums route through their storage
* type. Keep this tight — extending it requires the matching load/
* store code below. */
static int
let_scalar_ok(Type *t)
{
if (t == NULL) return 0;
Type *u = (t->kind == TY_NAMED) ? t->under : t;
if (u == NULL) return 0;
switch (u->kind) {
case TY_BOOL: case TY_RUNE:
case TY_I8: case TY_I16: case TY_I32: case TY_I64:
case TY_U8: case TY_U16: case TY_U32: case TY_U64:
case TY_INT: case TY_UINT: case TY_UINTPTR:
case TY_PTR:
return 1;
default:
return 0;
}
}
static int static int
decl_has_ffisym(Node *d) decl_has_ffisym(Node *d)
{ {
@@ -430,6 +464,35 @@ mod_lookup(const char *name)
return NULL; return NULL;
} }
/* Collect every top-level `let` whose declared type we can store
* in a single .data slot. Names not in this map fall through to
* the old "drop assignment" path; with a clear link-time
* undefined-symbol error on any read. */
static void
let_collect(Cg *c, Node *file)
{
letvars = NULL;
if (file == NULL) return;
for (Node *d = file->list; d; d = d->next) {
if (d->kind != N_LET) continue;
if (d->str == NULL || d->str[0] == '\0') continue;
if (!let_scalar_ok(d->type)) continue;
LetVar *lv = amalloc(c->a, sizeof *lv);
lv->name = d->str;
lv->next = letvars;
letvars = lv;
}
}
static int
let_islet(const char *name)
{
if (name == NULL) return 0;
for (LetVar *lv = letvars; lv; lv = lv->next)
if (strcmp(lv->name, name) == 0) return 1;
return 0;
}
/* Mangle an AST identifier into its asm linker symbol: /* Mangle an AST identifier into its asm linker symbol:
* - @symbol("...") binding wins (return mapped name). * - @symbol("...") binding wins (return mapped name).
* - module-private decl → <module>.<name>. * - module-private decl → <module>.<name>.
@@ -733,10 +796,17 @@ cgexpr(Cg *c, Node *n, Local *locals)
break; break;
} }
case TK_AMP: { case TK_AMP: {
/* address-of for an N_IDENT local */ /* address-of for an N_IDENT: local frame slot first,
* else a top-level mutable let (RIP-relative LEAQ).
* Anything else (e.g. & on an undefined name) silently
* drops, matching the pre-existing behaviour. */
if (n->lhs->kind == N_IDENT) { if (n->lhs->kind == N_IDENT) {
int off = localfind(locals, n->lhs->str); int off = localfind(locals, n->lhs->str);
ins2(c, A_LEAQ, amem(D_BP, off), areg(D_AX)); if (off != 0) {
ins2(c, A_LEAQ, amem(D_BP, off), areg(D_AX));
} else if (let_islet(n->lhs->str)) {
ins2(c, A_LEAQ, masym(c, n->lhs->str), areg(D_AX));
}
} }
break; break;
} }
@@ -1367,11 +1437,56 @@ cgexpr(Cg *c, Node *n, Local *locals)
} }
if (n->lhs->kind == N_IDENT) { if (n->lhs->kind == N_IDENT) {
int off = localfind(locals, n->lhs->str); int off = localfind(locals, n->lhs->str);
/* No local match: drop the whole assignment, including if (off == 0) {
* the RHS evaluation. Top-level let mutation isn't /* Top-level let target — RIP-relative store
* supported (no writable .data) and we don't want to * (or load→combine→store for compound). Names
* leak a dead `MOVQ $rhs, AX` into the output. */ * we don't recognise as scalar lets fall through
if (off == 0) break; * to the existing drop behaviour, which produces
* a clean link-time undefined-symbol error if
* the binding was ever supposed to exist. */
if (!let_islet(n->lhs->str)) break;
cgexpr(c, n->rhs, locals);
if (n->op == TK_ASSIGN) {
ins2(c, A_MOVQ, areg(D_AX),
masym(c, n->lhs->str));
break;
}
/* Compound: BX = load; combine with AX; store
* BX. The asm has no RIP-relative ADDQ/SUBQ
* mem-form, so we use the explicit load→
* combine→store sequence uniformly. */
ins2(c, A_MOVQ, masym(c, n->lhs->str),
areg(D_BX));
int did_compound = 1;
switch (n->op) {
case TK_PLUSEQ: ins2(c, A_ADDQ, areg(D_AX), areg(D_BX)); break;
case TK_MINUSEQ: ins2(c, A_SUBQ, areg(D_AX), areg(D_BX)); break;
case TK_STAREQ: ins2(c, A_IMULQ, areg(D_AX), areg(D_BX)); break;
case TK_AMPEQ: ins2(c, A_ANDQ, areg(D_AX), areg(D_BX)); break;
case TK_PIPEEQ: ins2(c, A_ORQ, areg(D_AX), areg(D_BX)); break;
case TK_CARETEQ: ins2(c, A_XORQ, areg(D_AX), areg(D_BX)); break;
case TK_LSHIFTEQ:
ins2(c, A_MOVQ, areg(D_AX), areg(D_CX));
ins2(c, A_SHLQ, areg(D_CX), areg(D_BX));
break;
case TK_RSHIFTEQ:
ins2(c, A_MOVQ, areg(D_AX), areg(D_CX));
ins2(c, A_SHRQ, areg(D_CX), areg(D_BX));
break;
default:
/* Unsupported compound op: store rhs
* directly. Mirrors the local path's
* fallback for TK_SLASHEQ etc. */
did_compound = 0;
ins2(c, A_MOVQ, areg(D_AX),
masym(c, n->lhs->str));
break;
}
if (did_compound)
ins2(c, A_MOVQ, areg(D_BX),
masym(c, n->lhs->str));
break;
}
cgexpr(c, n->rhs, locals); cgexpr(c, n->rhs, locals);
if (n->op == TK_ASSIGN) { if (n->op == TK_ASSIGN) {
ins2(c, A_MOVQ, areg(D_AX), amem(D_BP, off)); ins2(c, A_MOVQ, areg(D_AX), amem(D_BP, off));
@@ -3455,6 +3570,53 @@ cgfn(Cg *c, FILE *out, Node *fn)
txt_emit(out, c->head); txt_emit(out, c->head);
} }
/* Emit a single 8-byte DATA/DATAW row for a scalar value. Shares
* the escape rules with emit_defs/emit_data so the .o bytes stay
* stable. */
static void
emit_data_row(FILE *out, const char *dir, const char *name, u64 v)
{
fprintf(out, "%s %s(SB),\"", dir, name);
for (int i = 0; i < 8; i++) {
unsigned b = (unsigned)((v >> (i * 8)) & 0xff);
if (b == '"' || b == '\\')
fprintf(out, "\\%c", b);
else if (b < 0x20 || b >= 0x7f)
fprintf(out, "\\x%02x", b);
else
fputc(b, out);
}
fputs("\"\n", out);
}
/* Emit DATAW directives for top-level mutable `let` decls. Only scalar
* types in `let_scalar_ok` are supported; the rest are silently
* skipped at cgen and link with an undefined-symbol error if used.
* Initialisers must be integer/rune/bool/nil literals (or a `let`
* with no init, which zero-initialises). */
static void
emit_lets(Cg *c, FILE *out, Node *file)
{
for (Node *d = file->list; d; d = d->next) {
if (d->kind != N_LET) continue;
if (d->str == NULL || d->str[0] == '\0') continue;
if (!let_scalar_ok(d->type)) continue;
u64 v = 0;
if (d->rhs != NULL) {
Node *r = d->rhs;
while (r != NULL && r->kind == N_CAST) r = r->lhs;
if (r == NULL) continue;
if (r->kind == N_INTLIT) v = r->uval;
else if (r->kind == N_RUNELIT) v = r->uval;
else if (r->kind == N_TRUE) v = 1;
else if (r->kind == N_FALSE) v = 0;
else if (r->kind == N_NIL) v = 0;
else continue; /* non-literal init: skip */
}
emit_data_row(out, "DATAW", mod_mangle(c, d->str), v);
}
}
/* Emit DATA directives for top-level `def` constants whose value is /* Emit DATA directives for top-level `def` constants whose value is
* an integer/rune literal. The w6a side stores the bytes inside .text * an integer/rune literal. The w6a side stores the bytes inside .text
* and accesses are RIP-relative. * and accesses are RIP-relative.
@@ -3516,6 +3678,7 @@ cg_file(Cg *c, FILE *out, Node *file)
ffi_collect(c, file); ffi_collect(c, file);
mod_collect(c, file); mod_collect(c, file);
sdef_collect(c, file); sdef_collect(c, file);
let_collect(c, file);
strlits = NULL; strlits = NULL;
strlit_seq = 0; strlit_seq = 0;
for (Node *d = file->list; d; d = d->next) { for (Node *d = file->list; d; d = d->next) {
@@ -3524,6 +3687,7 @@ cg_file(Cg *c, FILE *out, Node *file)
} }
emit_data(c, out); emit_data(c, out);
emit_defs(c, out, file); emit_defs(c, out, file);
emit_lets(c, out, file);
} }
void peephole(Cg *c) { (void)c; } void peephole(Cg *c) { (void)c; }

View File

@@ -25,6 +25,7 @@ anames(int op)
case A_NOP: return "NOP"; case A_NOP: return "NOP";
case A_TEXT: return "TEXT"; case A_TEXT: return "TEXT";
case A_DATA: return "DATA"; case A_DATA: return "DATA";
case A_DATAW: return "DATAW";
case A_GLOBL: return "GLOBL"; case A_GLOBL: return "GLOBL";
case A_END: return "END"; case A_END: return "END";
case A_MOVQ: return "MOVQ"; case A_MOVQ: return "MOVQ";

View File

@@ -6678,6 +6678,13 @@ fn cgun(c: *cgen, n: *node) void = {
emitline("(BP), AX\n"); emitline("(BP), AX\n");
return; return;
}; };
// Top-level mutable let — RIP-relative LEAQ.
if (isletvar(c, nm)) {
emitline("\tLEAQ\t");
emitsymname(c, nm);
emitline("(SB), AX\n");
return;
};
}; };
}; };
return; return;
@@ -7253,7 +7260,52 @@ fn cgassign(c: *cgen, n: *node) void = {
if (lhs.kind == nkind.N_IDENT) { if (lhs.kind == nkind.N_IDENT) {
let nm: str = lhs.str; let nm: str = lhs.str;
let off: i32 = localfind(c, nm); let off: i32 = localfind(c, nm);
if (off == 0) { return; }; if (off == 0) {
// Top-level let target: RIP-relative store
// for `=`, or load→combine→store for the
// compound forms.
if (!isletvar(c, nm)) { return; };
cgexpr(c, n.rhs);
if (n.op == tkind.TK_ASSIGN) {
emitline("\tMOVQ\tAX, ");
emitsymname(c, nm);
emitline("(SB)\n");
return;
};
emitline("\tMOVQ\t");
emitsymname(c, nm);
emitline("(SB), BX\n");
let didcompound: bool = true;
if (n.op == tkind.TK_PLUSEQ) { emitline("\tADDQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_MINUSEQ) { emitline("\tSUBQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_STAREQ) { emitline("\tIMULQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_AMPEQ) { emitline("\tANDQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_PIPEEQ) { emitline("\tORQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_CARETEQ) { emitline("\tXORQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_LSHIFTEQ) {
emitline("\tMOVQ\tAX, CX\n");
emitline("\tSHLQ\tCX, BX\n");
}
else { if (n.op == tkind.TK_RSHIFTEQ) {
emitline("\tMOVQ\tAX, CX\n");
emitline("\tSHRQ\tCX, BX\n");
}
else {
// Unsupported compound: store rhs
// directly. Mirrors the local path's
// fallback for TK_SLASHEQ etc.
didcompound = false;
emitline("\tMOVQ\tAX, ");
emitsymname(c, nm);
emitline("(SB)\n");
};};};};};};};};
if (didcompound) {
emitline("\tMOVQ\tBX, ");
emitsymname(c, nm);
emitline("(SB)\n");
};
return;
};
// Detect str-typed local — assignment must store both // Detect str-typed local — assignment must store both
// halves (AX=ptr at +0, BX=len at +8). // halves (AX=ptr at +0, BX=len at +8).
let lcstr: bool = false; let lcstr: bool = false;
@@ -8326,6 +8378,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
collectfnrets(c, file); collectfnrets(c, file);
fficollect(c, file); fficollect(c, file);
collectmods(c, file); collectmods(c, file);
collectlets(c, file);
let d: *node = file.list; let d: *node = file.list;
for (d != nil) { for (d != nil) {
if (d.kind == nkind.N_FNDECL) { if (d.kind == nkind.N_FNDECL) {
@@ -8337,6 +8390,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
}; };
emitdatasection(c); emitdatasection(c);
emitdefconstants(c, file); emitdefconstants(c, file);
emitletdataw(c, file);
}; };
// MODULE: wcc // MODULE: wcc
@@ -8661,6 +8715,7 @@ type cgen = struct {
structs: *structinfo, structs: *structinfo,
enums: *enumtype, enums: *enumtype,
mods: *modent, // non-exported decls → originating module mods: *modent, // non-exported decls → originating module
lets: *letvar, // top-level mutable scalar `let` bindings
fnname: str, fnname: str,
fnret: *node, // declared return type of current fn (or nil) fnret: *node, // declared return type of current fn (or nil)
looptop: i32, looptop: i32,
@@ -8672,6 +8727,15 @@ type cgen = struct {
deferbuf: **node, // stack of deferred exprs (LIFO at return) deferbuf: **node, // stack of deferred exprs (LIFO at return)
}; };
// Top-level mutable `let` registry. Mirrors cmd/w6c/cgen.c LetVar.
// Populated alongside modents; consulted by cgassign and the TK_AMP
// path so reads/writes hit a RIP-relative DATAW slot instead of being
// silently dropped. Only scalar (≤8B) types make the list.
type letvar = struct {
name: str,
lvnext: *letvar,
};
fn cgeninit(c: *cgen, a: *arena) void = { fn cgeninit(c: *cgen, a: *arena) void = {
c.a = a; c.a = a;
c.locals = nil; c.locals = nil;
@@ -8889,6 +8953,157 @@ fn internstrlit(c: *cgen, bytes: str) str = {
return lab; return lab;
}; };
// letscalarprim — recognise the bare type-name keywords whose values
// fit in an 8-byte .data slot and load back with a plain MOVQ. Float
// types deliberately excluded; they need MOVSS/MOVSD. Mirrors C
// `let_scalar_ok` for the unwrapped TY_* enumeration. Pointer types
// (N_TPTR) handle separately at the callsite.
fn letscalarprim(nm: str) bool = {
if (streq(nm, "bool")) { return true; };
if (streq(nm, "rune")) { return true; };
if (streq(nm, "i8")) { return true; };
if (streq(nm, "i16")) { return true; };
if (streq(nm, "i32")) { return true; };
if (streq(nm, "i64")) { return true; };
if (streq(nm, "u8")) { return true; };
if (streq(nm, "u16")) { return true; };
if (streq(nm, "u32")) { return true; };
if (streq(nm, "u64")) { return true; };
if (streq(nm, "int")) { return true; };
if (streq(nm, "uint")) { return true; };
if (streq(nm, "uintptr")) { return true; };
return false;
};
// letscalarok — true iff the N_LET decl's declared type lands in the
// scalar set, walking type aliases. Pointer types are always ok.
fn letscalarok(c: *cgen, d: *node) bool = {
if (d == nil) { return false; };
let t: *node = d.lhs;
for (t != nil) {
if (t.kind == nkind.N_TPTR) { return true; };
if (t.kind != nkind.N_TNAME) { return false; };
let nm: str = t.str;
if (letscalarprim(nm)) { return true; };
// Resolve a `type x = y;` alias and look again. C cgen
// works on the resolved Type, so byte output diverges
// here without the walk.
let next: *node = aliaslookup(c, nm);
if (next == nil) { return false; };
t = next;
};
return false;
};
fn collectlets(c: *cgen, file: *node) void = {
c.lets = nil;
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let nm: str = d.str;
if (nm.len > 0) {
if (letscalarok(c, d)) {
let lv: *letvar = amalloc(c.a, 32u64): *letvar;
lv.name = nm;
lv.lvnext = c.lets;
c.lets = lv;
};
};
};
d = d.next;
};
};
fn isletvar(c: *cgen, name: str) bool = {
let lv: *letvar = c.lets;
for (lv != nil) {
if (streq(lv.name, name)) { return true; };
lv = lv.lvnext;
};
return false;
};
// emitletdataw — DATAW directive per top-level scalar `let`. Same
// 8-byte LE byte encoding as emitdefconstants; only the directive
// keyword differs ("DATAW" vs "DATA"). Mirrors cmd/w6c/cgen.c
// emit_lets — w6a routes DATAW into a writable .data section, and
// w6l covers it with a second R+W PT_LOAD.
fn emitletdataw(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let nm: str = d.str;
if (nm.len > 0) {
if (letscalarok(c, d)) {
let v: u64 = 0u64;
let ok: bool = true;
if (d.rhs != nil) {
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
ok = false;
if (r != nil) {
if (r.kind == nkind.N_INTLIT) { v = r.uval; ok = true; };
if (r.kind == nkind.N_RUNELIT) { v = r.uval; ok = true; };
if (r.kind == nkind.N_TRUE) { v = 1u64; ok = true; };
if (r.kind == nkind.N_FALSE) { v = 0u64; ok = true; };
if (r.kind == nkind.N_NIL) { v = 0u64; ok = true; };
};
};
if (ok) {
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
let n: u64 = v;
for (i < 8) {
let b: u8 = (n & 255u64): u8;
n = n >> 8u64;
if (b == 34u8) { emitline("\\\""); }
else { if (b == 92u8) { emitline("\\\\"); }
else {
if (b < 32u8) {
emitline("\\x");
let hi: u8 = b >> 4u8;
let lo: u8 = b & 15u8;
let bb: [2]u8;
if (hi < 10u8) { bb[0] = hi + 48u8; }
else { bb[0] = (hi - 10u8) + 97u8; };
if (lo < 10u8) { bb[1] = lo + 48u8; }
else { bb[1] = (lo - 10u8) + 97u8; };
os.write(1, bb.ptr, 2u64);
} else {
if (b >= 127u8) {
emitline("\\x");
let hi: u8 = b >> 4u8;
let lo: u8 = b & 15u8;
let bb: [2]u8;
if (hi < 10u8) { bb[0] = hi + 48u8; }
else { bb[0] = (hi - 10u8) + 97u8; };
if (lo < 10u8) { bb[1] = lo + 48u8; }
else { bb[1] = (lo - 10u8) + 97u8; };
os.write(1, bb.ptr, 2u64);
} else {
let bb: [1]u8;
bb[0] = b;
os.write(1, bb.ptr, 1u64);
};
};
};};
i += 1;
};
emitline("\"\n");
};
};
};
};
d = d.next;
};
};
// emitdefconstants — DATA directive per top-level int-literal `def`. // emitdefconstants — DATA directive per top-level int-literal `def`.
// 8 bytes little-endian to match what the C cgen emits. // 8 bytes little-endian to match what the C cgen emits.
fn emitdefconstants(c: *cgen, file: *node) void = { fn emitdefconstants(c: *cgen, file: *node) void = {

View File

@@ -319,6 +319,7 @@ type cgen = struct {
structs: *structinfo, structs: *structinfo,
enums: *enumtype, enums: *enumtype,
mods: *modent, // non-exported decls → originating module mods: *modent, // non-exported decls → originating module
lets: *letvar, // top-level mutable scalar `let` bindings
fnname: str, fnname: str,
fnret: *node, // declared return type of current fn (or nil) fnret: *node, // declared return type of current fn (or nil)
looptop: i32, looptop: i32,
@@ -330,6 +331,15 @@ type cgen = struct {
deferbuf: **node, // stack of deferred exprs (LIFO at return) deferbuf: **node, // stack of deferred exprs (LIFO at return)
}; };
// Top-level mutable `let` registry. Mirrors cmd/w6c/cgen.c LetVar.
// Populated alongside modents; consulted by cgassign and the TK_AMP
// path so reads/writes hit a RIP-relative DATAW slot instead of being
// silently dropped. Only scalar (≤8B) types make the list.
type letvar = struct {
name: str,
lvnext: *letvar,
};
fn cgeninit(c: *cgen, a: *arena) void = { fn cgeninit(c: *cgen, a: *arena) void = {
c.a = a; c.a = a;
c.locals = nil; c.locals = nil;
@@ -547,6 +557,157 @@ fn internstrlit(c: *cgen, bytes: str) str = {
return lab; return lab;
}; };
// letscalarprim — recognise the bare type-name keywords whose values
// fit in an 8-byte .data slot and load back with a plain MOVQ. Float
// types deliberately excluded; they need MOVSS/MOVSD. Mirrors C
// `let_scalar_ok` for the unwrapped TY_* enumeration. Pointer types
// (N_TPTR) handle separately at the callsite.
fn letscalarprim(nm: str) bool = {
if (streq(nm, "bool")) { return true; };
if (streq(nm, "rune")) { return true; };
if (streq(nm, "i8")) { return true; };
if (streq(nm, "i16")) { return true; };
if (streq(nm, "i32")) { return true; };
if (streq(nm, "i64")) { return true; };
if (streq(nm, "u8")) { return true; };
if (streq(nm, "u16")) { return true; };
if (streq(nm, "u32")) { return true; };
if (streq(nm, "u64")) { return true; };
if (streq(nm, "int")) { return true; };
if (streq(nm, "uint")) { return true; };
if (streq(nm, "uintptr")) { return true; };
return false;
};
// letscalarok — true iff the N_LET decl's declared type lands in the
// scalar set, walking type aliases. Pointer types are always ok.
fn letscalarok(c: *cgen, d: *node) bool = {
if (d == nil) { return false; };
let t: *node = d.lhs;
for (t != nil) {
if (t.kind == nkind.N_TPTR) { return true; };
if (t.kind != nkind.N_TNAME) { return false; };
let nm: str = t.str;
if (letscalarprim(nm)) { return true; };
// Resolve a `type x = y;` alias and look again. C cgen
// works on the resolved Type, so byte output diverges
// here without the walk.
let next: *node = aliaslookup(c, nm);
if (next == nil) { return false; };
t = next;
};
return false;
};
fn collectlets(c: *cgen, file: *node) void = {
c.lets = nil;
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let nm: str = d.str;
if (nm.len > 0) {
if (letscalarok(c, d)) {
let lv: *letvar = amalloc(c.a, 32u64): *letvar;
lv.name = nm;
lv.lvnext = c.lets;
c.lets = lv;
};
};
};
d = d.next;
};
};
fn isletvar(c: *cgen, name: str) bool = {
let lv: *letvar = c.lets;
for (lv != nil) {
if (streq(lv.name, name)) { return true; };
lv = lv.lvnext;
};
return false;
};
// emitletdataw — DATAW directive per top-level scalar `let`. Same
// 8-byte LE byte encoding as emitdefconstants; only the directive
// keyword differs ("DATAW" vs "DATA"). Mirrors cmd/w6c/cgen.c
// emit_lets — w6a routes DATAW into a writable .data section, and
// w6l covers it with a second R+W PT_LOAD.
fn emitletdataw(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let nm: str = d.str;
if (nm.len > 0) {
if (letscalarok(c, d)) {
let v: u64 = 0u64;
let ok: bool = true;
if (d.rhs != nil) {
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
ok = false;
if (r != nil) {
if (r.kind == nkind.N_INTLIT) { v = r.uval; ok = true; };
if (r.kind == nkind.N_RUNELIT) { v = r.uval; ok = true; };
if (r.kind == nkind.N_TRUE) { v = 1u64; ok = true; };
if (r.kind == nkind.N_FALSE) { v = 0u64; ok = true; };
if (r.kind == nkind.N_NIL) { v = 0u64; ok = true; };
};
};
if (ok) {
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
let n: u64 = v;
for (i < 8) {
let b: u8 = (n & 255u64): u8;
n = n >> 8u64;
if (b == 34u8) { emitline("\\\""); }
else { if (b == 92u8) { emitline("\\\\"); }
else {
if (b < 32u8) {
emitline("\\x");
let hi: u8 = b >> 4u8;
let lo: u8 = b & 15u8;
let bb: [2]u8;
if (hi < 10u8) { bb[0] = hi + 48u8; }
else { bb[0] = (hi - 10u8) + 97u8; };
if (lo < 10u8) { bb[1] = lo + 48u8; }
else { bb[1] = (lo - 10u8) + 97u8; };
os.write(1, bb.ptr, 2u64);
} else {
if (b >= 127u8) {
emitline("\\x");
let hi: u8 = b >> 4u8;
let lo: u8 = b & 15u8;
let bb: [2]u8;
if (hi < 10u8) { bb[0] = hi + 48u8; }
else { bb[0] = (hi - 10u8) + 97u8; };
if (lo < 10u8) { bb[1] = lo + 48u8; }
else { bb[1] = (lo - 10u8) + 97u8; };
os.write(1, bb.ptr, 2u64);
} else {
let bb: [1]u8;
bb[0] = b;
os.write(1, bb.ptr, 1u64);
};
};
};};
i += 1;
};
emitline("\"\n");
};
};
};
};
d = d.next;
};
};
// emitdefconstants — DATA directive per top-level int-literal `def`. // emitdefconstants — DATA directive per top-level int-literal `def`.
// 8 bytes little-endian to match what the C cgen emits. // 8 bytes little-endian to match what the C cgen emits.
fn emitdefconstants(c: *cgen, file: *node) void = { fn emitdefconstants(c: *cgen, file: *node) void = {

View File

@@ -287,6 +287,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
collectfnrets(c, file); collectfnrets(c, file);
fficollect(c, file); fficollect(c, file);
collectmods(c, file); collectmods(c, file);
collectlets(c, file);
let d: *node = file.list; let d: *node = file.list;
for (d != nil) { for (d != nil) {
if (d.kind == nkind.N_FNDECL) { if (d.kind == nkind.N_FNDECL) {
@@ -298,4 +299,5 @@ export fn cgfile(c: *cgen, file: *node) void = {
}; };
emitdatasection(c); emitdatasection(c);
emitdefconstants(c, file); emitdefconstants(c, file);
emitletdataw(c, file);
}; };

View File

@@ -950,6 +950,13 @@ fn cgun(c: *cgen, n: *node) void = {
emitline("(BP), AX\n"); emitline("(BP), AX\n");
return; return;
}; };
// Top-level mutable let — RIP-relative LEAQ.
if (isletvar(c, nm)) {
emitline("\tLEAQ\t");
emitsymname(c, nm);
emitline("(SB), AX\n");
return;
};
}; };
}; };
return; return;
@@ -1525,7 +1532,52 @@ fn cgassign(c: *cgen, n: *node) void = {
if (lhs.kind == nkind.N_IDENT) { if (lhs.kind == nkind.N_IDENT) {
let nm: str = lhs.str; let nm: str = lhs.str;
let off: i32 = localfind(c, nm); let off: i32 = localfind(c, nm);
if (off == 0) { return; }; if (off == 0) {
// Top-level let target: RIP-relative store
// for `=`, or load→combine→store for the
// compound forms.
if (!isletvar(c, nm)) { return; };
cgexpr(c, n.rhs);
if (n.op == tkind.TK_ASSIGN) {
emitline("\tMOVQ\tAX, ");
emitsymname(c, nm);
emitline("(SB)\n");
return;
};
emitline("\tMOVQ\t");
emitsymname(c, nm);
emitline("(SB), BX\n");
let didcompound: bool = true;
if (n.op == tkind.TK_PLUSEQ) { emitline("\tADDQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_MINUSEQ) { emitline("\tSUBQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_STAREQ) { emitline("\tIMULQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_AMPEQ) { emitline("\tANDQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_PIPEEQ) { emitline("\tORQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_CARETEQ) { emitline("\tXORQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_LSHIFTEQ) {
emitline("\tMOVQ\tAX, CX\n");
emitline("\tSHLQ\tCX, BX\n");
}
else { if (n.op == tkind.TK_RSHIFTEQ) {
emitline("\tMOVQ\tAX, CX\n");
emitline("\tSHRQ\tCX, BX\n");
}
else {
// Unsupported compound: store rhs
// directly. Mirrors the local path's
// fallback for TK_SLASHEQ etc.
didcompound = false;
emitline("\tMOVQ\tAX, ");
emitsymname(c, nm);
emitline("(SB)\n");
};};};};};};};};
if (didcompound) {
emitline("\tMOVQ\tBX, ");
emitsymname(c, nm);
emitline("(SB)\n");
};
return;
};
// Detect str-typed local — assignment must store both // Detect str-typed local — assignment must store both
// halves (AX=ptr at +0, BX=len at +8). // halves (AX=ptr at +0, BX=len at +8).
let lcstr: bool = false; let lcstr: bool = false;

View File

@@ -6678,6 +6678,13 @@ fn cgun(c: *cgen, n: *node) void = {
emitline("(BP), AX\n"); emitline("(BP), AX\n");
return; return;
}; };
// Top-level mutable let — RIP-relative LEAQ.
if (isletvar(c, nm)) {
emitline("\tLEAQ\t");
emitsymname(c, nm);
emitline("(SB), AX\n");
return;
};
}; };
}; };
return; return;
@@ -7253,7 +7260,52 @@ fn cgassign(c: *cgen, n: *node) void = {
if (lhs.kind == nkind.N_IDENT) { if (lhs.kind == nkind.N_IDENT) {
let nm: str = lhs.str; let nm: str = lhs.str;
let off: i32 = localfind(c, nm); let off: i32 = localfind(c, nm);
if (off == 0) { return; }; if (off == 0) {
// Top-level let target: RIP-relative store
// for `=`, or load→combine→store for the
// compound forms.
if (!isletvar(c, nm)) { return; };
cgexpr(c, n.rhs);
if (n.op == tkind.TK_ASSIGN) {
emitline("\tMOVQ\tAX, ");
emitsymname(c, nm);
emitline("(SB)\n");
return;
};
emitline("\tMOVQ\t");
emitsymname(c, nm);
emitline("(SB), BX\n");
let didcompound: bool = true;
if (n.op == tkind.TK_PLUSEQ) { emitline("\tADDQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_MINUSEQ) { emitline("\tSUBQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_STAREQ) { emitline("\tIMULQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_AMPEQ) { emitline("\tANDQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_PIPEEQ) { emitline("\tORQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_CARETEQ) { emitline("\tXORQ\tAX, BX\n"); }
else { if (n.op == tkind.TK_LSHIFTEQ) {
emitline("\tMOVQ\tAX, CX\n");
emitline("\tSHLQ\tCX, BX\n");
}
else { if (n.op == tkind.TK_RSHIFTEQ) {
emitline("\tMOVQ\tAX, CX\n");
emitline("\tSHRQ\tCX, BX\n");
}
else {
// Unsupported compound: store rhs
// directly. Mirrors the local path's
// fallback for TK_SLASHEQ etc.
didcompound = false;
emitline("\tMOVQ\tAX, ");
emitsymname(c, nm);
emitline("(SB)\n");
};};};};};};};};
if (didcompound) {
emitline("\tMOVQ\tBX, ");
emitsymname(c, nm);
emitline("(SB)\n");
};
return;
};
// Detect str-typed local — assignment must store both // Detect str-typed local — assignment must store both
// halves (AX=ptr at +0, BX=len at +8). // halves (AX=ptr at +0, BX=len at +8).
let lcstr: bool = false; let lcstr: bool = false;
@@ -8326,6 +8378,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
collectfnrets(c, file); collectfnrets(c, file);
fficollect(c, file); fficollect(c, file);
collectmods(c, file); collectmods(c, file);
collectlets(c, file);
let d: *node = file.list; let d: *node = file.list;
for (d != nil) { for (d != nil) {
if (d.kind == nkind.N_FNDECL) { if (d.kind == nkind.N_FNDECL) {
@@ -8337,6 +8390,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
}; };
emitdatasection(c); emitdatasection(c);
emitdefconstants(c, file); emitdefconstants(c, file);
emitletdataw(c, file);
}; };
// MODULE: wcc // MODULE: wcc
@@ -8661,6 +8715,7 @@ type cgen = struct {
structs: *structinfo, structs: *structinfo,
enums: *enumtype, enums: *enumtype,
mods: *modent, // non-exported decls → originating module mods: *modent, // non-exported decls → originating module
lets: *letvar, // top-level mutable scalar `let` bindings
fnname: str, fnname: str,
fnret: *node, // declared return type of current fn (or nil) fnret: *node, // declared return type of current fn (or nil)
looptop: i32, looptop: i32,
@@ -8672,6 +8727,15 @@ type cgen = struct {
deferbuf: **node, // stack of deferred exprs (LIFO at return) deferbuf: **node, // stack of deferred exprs (LIFO at return)
}; };
// Top-level mutable `let` registry. Mirrors cmd/w6c/cgen.c LetVar.
// Populated alongside modents; consulted by cgassign and the TK_AMP
// path so reads/writes hit a RIP-relative DATAW slot instead of being
// silently dropped. Only scalar (≤8B) types make the list.
type letvar = struct {
name: str,
lvnext: *letvar,
};
fn cgeninit(c: *cgen, a: *arena) void = { fn cgeninit(c: *cgen, a: *arena) void = {
c.a = a; c.a = a;
c.locals = nil; c.locals = nil;
@@ -8889,6 +8953,157 @@ fn internstrlit(c: *cgen, bytes: str) str = {
return lab; return lab;
}; };
// letscalarprim — recognise the bare type-name keywords whose values
// fit in an 8-byte .data slot and load back with a plain MOVQ. Float
// types deliberately excluded; they need MOVSS/MOVSD. Mirrors C
// `let_scalar_ok` for the unwrapped TY_* enumeration. Pointer types
// (N_TPTR) handle separately at the callsite.
fn letscalarprim(nm: str) bool = {
if (streq(nm, "bool")) { return true; };
if (streq(nm, "rune")) { return true; };
if (streq(nm, "i8")) { return true; };
if (streq(nm, "i16")) { return true; };
if (streq(nm, "i32")) { return true; };
if (streq(nm, "i64")) { return true; };
if (streq(nm, "u8")) { return true; };
if (streq(nm, "u16")) { return true; };
if (streq(nm, "u32")) { return true; };
if (streq(nm, "u64")) { return true; };
if (streq(nm, "int")) { return true; };
if (streq(nm, "uint")) { return true; };
if (streq(nm, "uintptr")) { return true; };
return false;
};
// letscalarok — true iff the N_LET decl's declared type lands in the
// scalar set, walking type aliases. Pointer types are always ok.
fn letscalarok(c: *cgen, d: *node) bool = {
if (d == nil) { return false; };
let t: *node = d.lhs;
for (t != nil) {
if (t.kind == nkind.N_TPTR) { return true; };
if (t.kind != nkind.N_TNAME) { return false; };
let nm: str = t.str;
if (letscalarprim(nm)) { return true; };
// Resolve a `type x = y;` alias and look again. C cgen
// works on the resolved Type, so byte output diverges
// here without the walk.
let next: *node = aliaslookup(c, nm);
if (next == nil) { return false; };
t = next;
};
return false;
};
fn collectlets(c: *cgen, file: *node) void = {
c.lets = nil;
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let nm: str = d.str;
if (nm.len > 0) {
if (letscalarok(c, d)) {
let lv: *letvar = amalloc(c.a, 32u64): *letvar;
lv.name = nm;
lv.lvnext = c.lets;
c.lets = lv;
};
};
};
d = d.next;
};
};
fn isletvar(c: *cgen, name: str) bool = {
let lv: *letvar = c.lets;
for (lv != nil) {
if (streq(lv.name, name)) { return true; };
lv = lv.lvnext;
};
return false;
};
// emitletdataw — DATAW directive per top-level scalar `let`. Same
// 8-byte LE byte encoding as emitdefconstants; only the directive
// keyword differs ("DATAW" vs "DATA"). Mirrors cmd/w6c/cgen.c
// emit_lets — w6a routes DATAW into a writable .data section, and
// w6l covers it with a second R+W PT_LOAD.
fn emitletdataw(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let nm: str = d.str;
if (nm.len > 0) {
if (letscalarok(c, d)) {
let v: u64 = 0u64;
let ok: bool = true;
if (d.rhs != nil) {
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
ok = false;
if (r != nil) {
if (r.kind == nkind.N_INTLIT) { v = r.uval; ok = true; };
if (r.kind == nkind.N_RUNELIT) { v = r.uval; ok = true; };
if (r.kind == nkind.N_TRUE) { v = 1u64; ok = true; };
if (r.kind == nkind.N_FALSE) { v = 0u64; ok = true; };
if (r.kind == nkind.N_NIL) { v = 0u64; ok = true; };
};
};
if (ok) {
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
let n: u64 = v;
for (i < 8) {
let b: u8 = (n & 255u64): u8;
n = n >> 8u64;
if (b == 34u8) { emitline("\\\""); }
else { if (b == 92u8) { emitline("\\\\"); }
else {
if (b < 32u8) {
emitline("\\x");
let hi: u8 = b >> 4u8;
let lo: u8 = b & 15u8;
let bb: [2]u8;
if (hi < 10u8) { bb[0] = hi + 48u8; }
else { bb[0] = (hi - 10u8) + 97u8; };
if (lo < 10u8) { bb[1] = lo + 48u8; }
else { bb[1] = (lo - 10u8) + 97u8; };
os.write(1, bb.ptr, 2u64);
} else {
if (b >= 127u8) {
emitline("\\x");
let hi: u8 = b >> 4u8;
let lo: u8 = b & 15u8;
let bb: [2]u8;
if (hi < 10u8) { bb[0] = hi + 48u8; }
else { bb[0] = (hi - 10u8) + 97u8; };
if (lo < 10u8) { bb[1] = lo + 48u8; }
else { bb[1] = (lo - 10u8) + 97u8; };
os.write(1, bb.ptr, 2u64);
} else {
let bb: [1]u8;
bb[0] = b;
os.write(1, bb.ptr, 1u64);
};
};
};};
i += 1;
};
emitline("\"\n");
};
};
};
};
d = d.next;
};
};
// emitdefconstants — DATA directive per top-level int-literal `def`. // emitdefconstants — DATA directive per top-level int-literal `def`.
// 8 bytes little-endian to match what the C cgen emits. // 8 bytes little-endian to match what the C cgen emits.
fn emitdefconstants(c: *cgen, file: *node) void = { fn emitdefconstants(c: *cgen, file: *node) void = {

118
test/wcc/630_let_global.c Normal file
View File

@@ -0,0 +1,118 @@
/*
* 630_let_global — top-level mutable `let` end-to-end through the
* full Cstage pipeline (w6c → w6a → w6l). Each fixture is a tiny
* .ww program that exercises one path: literal-init read, plain
* assignment, compound assignment, and address-of.
*
* Exit code = the value the runtime feeds to `exit(2)`. main's
* return value flows into DI via rt/start.s and ends up as the
* shell's $?.
*/
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <sys/wait.h>
static int
run_exit(const char *cmd)
{
int rc = system(cmd);
if (rc == -1) return -1;
if (WIFEXITED(rc)) return WEXITSTATUS(rc);
return -1;
}
static int
build_and_run(const char *bin, const char *prog)
{
/* `ww run` builds with the driver (w6c → w6a → w6l + libwwrt.a)
* and executes the result. The exit code propagates back as the
* shell's $? so we just compare against the fixture's want. */
char src[64], cmd[1024];
snprintf(src, sizeof src, "/tmp/wwt_lg_%d.ww", getpid());
FILE *f = fopen(src, "wb");
if (!f) return -1;
fputs(prog, f);
fclose(f);
snprintf(cmd, sizeof cmd, "%s/ww run %s", bin, src);
int rc = run_exit(cmd);
unlink(src);
return rc;
}
struct fixture {
const char *label;
const char *prog;
int want;
};
static const struct fixture fixtures[] = {
{
"read",
"let counter: i32 = 42;\n"
"fn main() i32 = { return counter; };\n",
42,
},
{
"assign",
"let counter: i32 = 0;\n"
"fn main() i32 = { counter = 99; return counter; };\n",
99,
},
{
"compound",
"let counter: i32 = 1;\n"
"fn inc() void = { counter += 1; };\n"
"fn main() i32 = { inc(); inc(); inc(); return counter; };\n",
4,
},
{
"zero-init",
/* No initialiser → zero-filled DATAW slot. */
"let counter: i32;\n"
"fn main() i32 = { counter = 7; return counter; };\n",
7,
},
{
"addr-of",
"let counter: i32 = 11;\n"
"fn main() i32 = {\n"
"\tlet p: *i32 = &counter;\n"
"\t*p = 55;\n"
"\treturn counter;\n"
"};\n",
55,
},
{
"u64-read",
"let val: u64 = 123u64;\n"
"fn main() i32 = { return val: i32; };\n",
123,
},
{ NULL, NULL, 0 }
};
int
main(void)
{
const char *bin = getenv("BIN");
if (!bin) bin = "out/bin";
int fail = 0, ran = 0;
for (int i = 0; fixtures[i].label; i++, ran++) {
int got = build_and_run(bin, fixtures[i].prog);
if (got != fixtures[i].want) {
fprintf(stderr, "let_global[%s]: exit=%d, want %d\n",
fixtures[i].label, got, fixtures[i].want);
fail++;
}
}
if (fail) {
fprintf(stderr, "let_global: %d/%d fixtures failed\n", fail, ran);
return 1;
}
printf("let_global: %d/%d ok\n", ran, ran);
return 0;
}