Files
ww/selfhost/cmd/wcc/cgen.ww
Hojun-Cho 404705b6fd selfhost: mirror defer; e2e tests for LIFO ordering
C cgen has carried defer for a while (defers[] global + reverse
walk on every return). Selfhost cgen now mirrors:

- cgen struct: deferbuf (**node, LIFO stack) + defertop counter.
- cgstmt N_DEFER: push n.lhs.
- cgreturn: rundefers() at entry — same as the C cgen pattern.
- cgfn fall-through return: rundefers() before zero-AX+RET.

DEFER_MAX = 16 matches C cgen.

Two new e2e rows: defer with an explicit `return acc;` (321 mod 256
= 65), and defer firing on an implicit void-fn fall-through (87).
Both rows verified via the wwstage cgen too.

Defer's semantics: queued exprs fire LIFO before the return expr
is evaluated, so a return that reads memory mutated by a deferred
call sees the post-defer state. Matches C cgen and Hare.
2026-05-12 03:11:13 +09:00

763 lines
20 KiB
Plaintext

// selfhost/cmd/wcc/cgen.ww — port of cmd/w6c/cgen.c.
//
// Status: GROWING. Each subsystem we add is verified by `wwdump_ww -c`
// producing byte-identical output to C-side `w6c` for the same source,
// then by assembling + linking + running the result.
//
// Current coverage:
// - decls: N_FILE, N_FNDECL (params, frame for locals, prologue
// + dual-epilogue suppression; FFI body-less fn skipped)
// - stmts: N_BLOCK, N_RETURN, N_EXPRSTMT, N_LET (no init),
// N_LET (int-literal / ident / call / N_BIN init),
// N_IF (with optional else), N_FOR (cond-only and full
// init/cond/post), N_BREAK, N_CONTINUE
// - exprs: N_INTLIT, N_IDENT (local/param), N_BIN with full op
// coverage (+/-/*/// %, &/|/^, <</>>, comparisons with
// signed-vs-unsigned dispatch, &&/||), N_UN (- ! ~ &amp; *),
// N_CALL (recursive R-to-L push, pop into argregs L-to-R),
// N_ASSIGN to local idents (plain and compound +=/-=)
//
// Type info is shallow — frame slots are 8 bytes per local, all loads
// /stores are MOVQ. Programs that mix i8/i32/i64 locals work but spill
// 8 bytes per local. Float, str, slice, struct, match, defer, alloc,
// tagged-union return — none of those are wired yet.
use os;
use mem;
use ast;
use tok;
use typ;
use sym;
use strconv;
// Split files. Bundler pulls these in transitively so consumers only
// need `use cgen;`. Order matters for the flat-bundle concat — utils
// first so cgenexpr/stmt/decl can reference helpers defined here.
use cgenutil;
use cgenexpr;
use cgenstmt;
use cgendecl;
// ---- typedef alias registry -----------------------------------------
//
// `type error = str;` makes `error` a struct-shape alias. We track
// alias→target so isstrtype / isslicetype / structlookup can
// resolve through the chain. Only direct N_TNAME aliases are mapped;
// `type p = struct {...}` is handled by collectstructs.
type aliasent = struct {
aname: str,
target: *node, // the rhs type expr
aanext: *aliasent,
};
fn collectaliases(c: *cgen, file: *node) void = {
c.aliases = nil;
let d: *node = file.list;
for (d != nil) {
if (d.kind == N_TYPEDECL) {
let body: *node = d.lhs;
if (body != nil) {
if (body.kind != N_TSTRUCT) {
let a: *aliasent = amalloc(c.a, 32u64): *aliasent;
a.aname = d.str;
a.target = body;
a.aanext = c.aliases;
c.aliases = a;
};
};
};
d = d.next;
};
};
fn aliaslookup(c: *cgen, name: str) *node = {
let a: *aliasent = c.aliases;
for (a != nil) {
let an: str = a.aname;
if (streq(an, name)) { return a.target; };
a = a.aanext;
};
return nil;
};
// resolvetype — follow typedef alias chains to a "canonical" type
// expr (str/slice/array/struct/...). Stops on cycles via depth limit.
fn resolvetype(c: *cgen, t: *node) *node = {
let cur: *node = t;
let depth: i32 = 0;
for (depth < 16) {
if (cur == nil) { return nil; };
if (cur.kind != N_TNAME) { return cur; };
let nm: str = cur.str;
let next: *node = aliaslookup(c, nm);
if (next == nil) { return cur; };
cur = next;
depth += 1;
};
return cur;
};
// ---- struct registry ------------------------------------------------
//
// Per-file map from struct name → list of fields with computed offsets
// and sizes. Built when cgfile walks N_TYPEDECL with N_TSTRUCT lhs.
// N_DOT and N_ASSIGN consult this to resolve `s.field` for struct or
// *struct bases.
type fieldinfo = struct {
fname: str,
foff: i32,
fsz: i32,
tnode: *node, // the field type expr, for nested struct lookups
finext: *fieldinfo,
};
type structinfo = struct {
sname: str,
fields: *fieldinfo,
totsize: i32,
sinext: *structinfo,
};
// ---- locals / frame --------------------------------------------------
type local = struct {
name: str,
off: i32,
tnode: *node, // declared type expr (N_TNAME / N_TPTR / ...) or nil
lnext: *local,
};
// strlit — interned string literal record. Emitted as a DATA directive
// after all functions; cgexpr N_STRLIT loads (LEAQ ptr, MOVQ len).
type strlit = struct {
label: str, // "_S_<seq>"
bytes: str,
slnext: *strlit,
};
// ffi — `@symbol("name")` mapping. Body-less fn `foo` with this attr
// gets its CALL target rewritten to `name`.
type ffi = struct {
ident: str,
symbol: str,
fnext: *ffi,
};
def LOOP_MAX: i32 = 16;
def DEFER_MAX: i32 = 16;
type cgen = struct {
a: *arena,
locals: *local,
frame: i32,
lastwasreturn: i32,
labelseq: i32,
strlitseq: i32,
strlits: *strlit,
ffis: *ffi,
defs: *defent,
fnrets: *fnret,
aliases: *aliasent,
structs: *structinfo,
mods: *modent, // non-exported decls → originating module
fnname: str,
fnret: *node, // declared return type of current fn (or nil)
looptop: i32,
loopendbuf: *str, // stack of end labels for break
loopcontbuf: *str, // stack of cont labels for continue
yieldtop: i32,
yieldbuf: *str, // stack of match end labels for yield
defertop: i32,
deferbuf: **node, // stack of deferred exprs (LIFO at return)
};
fn cgeninit(c: *cgen, a: *arena) void = {
c.a = a;
c.locals = nil;
c.frame = 0;
c.lastwasreturn = 0;
c.labelseq = 0;
// Note: strlit_seq, strlits, ffis are *not* reset here; they
// persist across cgfn calls within one file. cgfile resets them
// at the start of each compilation unit.
c.looptop = 0;
c.loopendbuf = amalloc(a, (LOOP_MAX: u64) * 16u64): *str;
c.loopcontbuf = amalloc(a, (LOOP_MAX: u64) * 16u64): *str;
c.yieldtop = 0;
c.yieldbuf = amalloc(a, (LOOP_MAX: u64) * 16u64): *str;
c.defertop = 0;
c.deferbuf = amalloc(a, (DEFER_MAX: u64) * 8u64): **node;
};
// localalloc — append a slot for `name` without dedup. Used for
// match-arm bindings, which C cgen allocates via cgexpr's by-value
// `locals` list — so two separate matches each get fresh slots even
// when their bind names collide. scanlocals follows the same rule
// for N_MCASE.
fn localalloc(c: *cgen, name: str, sz: i32, tnode: *node) i32 = {
let asz: i32 = sz;
if (asz < 8) { asz = 8; };
if ((asz & 7) != 0) { asz = (asz + 7) & ~7; };
c.frame += asz;
let off: i32 = 0 - c.frame;
let l: *local = amalloc(c.a, 48u64): *local;
l.name = name;
l.off = off;
l.tnode = tnode;
l.lnext = c.locals;
c.locals = l;
return off;
};
fn localadd(c: *cgen, name: str, sz: i32, tnode: *node) i32 = {
// Name-based slot reuse for N_LETs and params: if `name` is
// already declared in this function, return its existing
// offset. Mirrors C cgen (cmd/w6c/cgen.c:localoff). Two
// disjoint scopes that declare the same name share one slot —
// so `escape` in wwdump (three `let cp: pos;` across separate
// branches) reserves one slot, not three. scanlocals does
// the matching dedup at prologue time so the SUBQ stays in
// sync.
//
// On a dedup hit we also overwrite the stored tnode to match
// the new declaration's type. C reads `n->lhs->type` (filled
// by the checker) at every N_DOT/N_CAST site; we read
// `lc.tnode`, so it must follow source order. Without this,
// a later `let m: *node` inside a branch keeps an earlier
// `let m: i32`'s tnode and `m.next` falls into the SB fallback.
let cur: *local = c.locals;
for (cur != nil) {
let cn: str = cur.name;
if (streq(cn, name)) {
cur.tnode = tnode;
return cur.off;
};
cur = cur.lnext;
};
return localalloc(c, name, sz, tnode);
};
// scanseenmark — called by scanlocals on every let / match-bind
// site. Returns true if `name` is already tracked in c.locals (so
// the slot will be shared at emission time — no new frame bump).
// Otherwise appends a name-only stub and returns false. Stubs are
// thrown away when cgfn resets c.locals before emission.
fn scanseenmark(c: *cgen, name: str) bool = {
if (localfindnode(c, name) != nil) { return true; };
let l: *local = amalloc(c.a, 48u64): *local;
l.name = name;
l.off = 0;
l.tnode = nil;
l.lnext = c.locals;
c.locals = l;
return false;
};
fn localfindnode(c: *cgen, name: str) *local = {
let l: *local = c.locals;
for (l != nil) {
let ln: str = l.name;
if (streq(ln, name)) { return l; };
l = l.lnext;
};
return nil;
};
fn localfind(c: *cgen, name: str) i32 = {
let l: *local = c.locals;
for (l != nil) {
let ln: str = l.name;
if (ln.len == name.len) {
let i: i32 = 0;
let eq: bool = true;
for (i < name.len) {
if (ln[i] != name[i]) { eq = false; i = name.len; }
else { i += 1; };
};
if (eq) { return l.off; };
};
l = l.lnext;
};
return 0;
};
// ---- emit helpers ---------------------------------------------------
fn emitline(s: str) void = { os.write(1, s.ptr, s.len: u64); };
fn emitint(v: i64) void = {
let buf: [32]u8;
let n: i32 = strconv.i64tos(buf[0:32], v);
os.write(1, buf.ptr, n: u64);
};
fn emituint(v: u64) void = {
let buf: [32]u8;
let n: i32 = strconv.u64tos(buf[0:32], v);
os.write(1, buf.ptr, n: u64);
};
// emitdispreg — print "disp(reg)" or "(reg)" when disp == 0, the
// way Plan 9 6c/6a do.
fn emitdispreg(off: i64, reg: str) void = {
if (off != 0i64) { emitint(off); };
emitline("(");
emitline(reg);
emitline(")");
};
// emitoff — print an integer offset, suppressing it entirely when 0.
// Use before any emitline("(BP)...") or emitline("(SB)...") sequence.
// Plan 9 cc convention: "(BP)" not "0(BP)".
fn emitoff(v: i64) void = {
if (v != 0i64) { emitint(v); };
};
// mklabel — fresh label "<fnname>_<base>_<seq>". Returns an
// arena-owned str. Mirrors C cgen's mklabel so diffs match.
fn mklabel(c: *cgen, base: str) str = {
let buf: [128]u8;
let i: i32 = 0;
let fname: str = c.fnname;
let j: i32 = 0;
for (j < fname.len) {
buf[i] = fname[j];
i += 1; j += 1;
};
buf[i] = 95u8; i += 1; // '_'
j = 0;
for (j < base.len) {
buf[i] = base[j];
i += 1; j += 1;
};
buf[i] = 95u8; i += 1; // '_'
let n: i32 = strconv.i64tos(buf[i:128], c.labelseq: i64);
c.labelseq += 1;
let total: i32 = i + n;
let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8;
let k: i32 = 0;
for (k < total) {
p[k] = buf[k];
k += 1;
};
p[total] = 0u8;
let r: str;
r.ptr = p;
r.len = total;
return r;
};
fn emitlabel(s: str) void = {
os.write(1, s.ptr, s.len: u64);
emitline(":\n");
};
// ---- string interning ------------------------------------------------
//
// streq is provided by sym.ww and reused here.
// internstrlit — return a stable label for `bytes`. Dedups by content
// so identical literals share storage.
fn internstrlit(c: *cgen, bytes: str) str = {
let s: *strlit = c.strlits;
for (s != nil) {
let bs: str = s.bytes;
if (streq(bs, bytes)) {
return s.label;
};
s = s.slnext;
};
// New label "_S_<seq>".
let buf: [32]u8;
buf[0] = 95u8; buf[1] = 83u8; buf[2] = 95u8; // "_S_"
let n: i32 = strconv.i64tos(buf[3:32], c.strlitseq: i64);
c.strlitseq += 1;
let total: i32 = 3 + n;
let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8;
let i: i32 = 0;
for (i < total) { p[i] = buf[i]; i += 1; };
p[total] = 0u8;
let lab: str;
lab.ptr = p;
lab.len = total;
let nw: *strlit = amalloc(c.a, 48u64): *strlit;
nw.label = lab;
nw.bytes = bytes;
nw.slnext = c.strlits;
c.strlits = nw;
return lab;
};
// emitdefconstants — DATA directive per top-level int-literal `def`.
// 8 bytes little-endian to match what the C cgen emits.
fn emitdefconstants(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
if (d.kind == N_DEF) {
let r: *node = d.rhs;
let v: u64 = 0u64;
let ok: bool = false;
if (r != nil) {
if (r.kind == N_INTLIT) { v = r.uval; ok = true; };
if (r.kind == N_RUNELIT) { v = r.uval; ok = true; };
if (r.kind == N_TRUE) { v = 1u64; ok = true; };
if (r.kind == N_FALSE) { v = 0u64; ok = true; };
if (r.kind == N_NIL) { v = 0u64; ok = true; };
};
if (ok) {
emitline("DATA ");
if (d.exported == 0) {
if (d.module.len > 0) {
os.write(1, d.module.ptr, d.module.len: u64);
os.write(1, ".".ptr, 1u64);
};
};
let nm: str = d.str;
os.write(1, nm.ptr, nm.len: u64);
emitline("(SB),\"");
let i: i32 = 0;
let n: u64 = v;
for (i < 8) {
let b: u8 = (n & 255u64): u8;
n = n >> 8u64;
// C emit_defs only special-cases " and \;
// every other non-printable goes as \xHH.
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;
};
};
// emitdatasection — DATA directives for every interned strlit.
// Trailing NUL appended so .ptr can be used as a C string by syscalls.
fn emitdatasection(c: *cgen) void = {
let s: *strlit = c.strlits;
for (s != nil) {
emitline("DATA ");
let lab: str = s.label;
os.write(1, lab.ptr, lab.len: u64);
emitline("(SB),\"");
let bs: str = s.bytes;
let i: i32 = 0;
for (i < bs.len) {
let b: u8 = bs[i];
if (b == 34u8) { emitline("\\\""); } // "
else { if (b == 92u8) { emitline("\\\\"); } // \
else { if (b == 10u8) { emitline("\\n"); }
else { if (b == 9u8) { emitline("\\t"); }
else { if (b == 13u8) { emitline("\\r"); }
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("\\x00\"\n");
s = s.slnext;
};
};
// ---- fn return-type map ---------------------------------------------
//
// Per-file: ident → ret-type-node. Used to decide whether to shuffle
// (AX, DX) → (AX, BX) after a CALL — needed for str-returning fns so
// the value flows through cgen as the canonical (AX, BX) str pair.
type fnret = struct {
fname: str,
rtype: *node,
frnext: *fnret,
};
fn collectfnrets(c: *cgen, file: *node) void = {
c.fnrets = nil;
let d: *node = file.list;
for (d != nil) {
if (d.kind == N_FNDECL) {
let f: *fnret = amalloc(c.a, 32u64): *fnret;
f.fname = d.str;
f.rtype = d.lhs;
f.frnext = c.fnrets;
c.fnrets = f;
};
d = d.next;
};
};
fn fnretlookup(c: *cgen, name: str) *node = {
let f: *fnret = c.fnrets;
for (f != nil) {
let fn_: str = f.fname;
if (streq(fn_, name)) { return f.rtype; };
f = f.frnext;
};
return nil;
};
// ---- def-constant registry ------------------------------------------
//
// `def NAME: T = LIT;` becomes a DATA symbol the C-side w6c emits; an
// ident reference loads it via `MOVQ NAME(SB), AX`. We collect them at
// file load and consult on N_IDENT lookup.
type defent = struct {
dname: str,
drhs: *node,
dnext: *defent,
};
fn collectdefs(c: *cgen, file: *node) void = {
c.defs = nil;
let d: *node = file.list;
for (d != nil) {
if (d.kind == N_DEF) {
let e: *defent = amalloc(c.a, 32u64): *defent;
e.dname = d.str;
e.drhs = d.rhs;
e.dnext = c.defs;
c.defs = e;
};
d = d.next;
};
};
fn deflookup(c: *cgen, name: str) bool = {
let e: *defent = c.defs;
for (e != nil) {
let dn: str = e.dname;
if (streq(dn, name)) { return true; };
e = e.dnext;
};
return false;
};
// Returns the rhs init node for a top-level `def`, or nil if `name`
// doesn't name a def. Used by cgdot to inline `.ptr`/`.len` on
// `def NAME: str = "..."` — those aren't laid out in memory.
fn deflookuprhs(c: *cgen, name: str) *node = {
let e: *defent = c.defs;
for (e != nil) {
let dn: str = e.dname;
if (streq(dn, name)) { return e.drhs; };
e = e.dnext;
};
return nil;
};
// ---- module-private symbol map --------------------------------------
//
// Non-exported top-level decls live in their originating module's
// namespace. cgen mangles those names to `<module>.<name>` at emission
// time, both at the def site (TEXT/DATA) and at every call/load site,
// so two modules can each privately define `cstrlen` without colliding
// at link time. Exported decls and FFI-bound decls keep their bare name.
type modent = struct {
mname: str, // the bare ident as it appears in source
module: str, // the originating module (`// MODULE: foo`)
mnext: *modent,
};
fn collectmods(c: *cgen, file: *node) void = {
c.mods = nil;
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
// Mirror collectfnrets' shape exactly (plain prepend in one
// branch). Earlier nested-if/early-return variants tickled a
// wwstage cgen bug that dropped most prepends.
if (d.kind == N_FNDECL) {
if (d.exported == 0) {
if (d.module.len > 0) {
if (!streq(d.str, "main")) {
let m: *modent = amalloc(c.a, 48u64): *modent;
m.mname = d.str;
m.module = d.module;
m.mnext = c.mods;
c.mods = m;
};
};
};
};
if (d.kind == N_DEF) {
if (d.exported == 0) {
if (d.module.len > 0) {
let m: *modent = amalloc(c.a, 48u64): *modent;
m.mname = d.str;
m.module = d.module;
m.mnext = c.mods;
c.mods = m;
};
};
};
if (d.kind == N_TYPEDECL) {
if (d.exported == 0) {
if (d.module.len > 0) {
let m: *modent = amalloc(c.a, 48u64): *modent;
m.mname = d.str;
m.module = d.module;
m.mnext = c.mods;
c.mods = m;
};
};
};
if (d.kind == N_LET) {
if (d.exported == 0) {
if (d.module.len > 0) {
let m: *modent = amalloc(c.a, 48u64): *modent;
m.mname = d.str;
m.module = d.module;
m.mnext = c.mods;
c.mods = m;
};
};
};
d = d.next;
};
};
fn modlookup(c: *cgen, name: str) str = {
let m: *modent = c.mods;
for (m != nil) {
if (streq(m.mname, name)) { return m.module; };
m = m.mnext;
};
let empty: str;
empty.ptr = nil;
empty.len = 0;
return empty;
};
// emitsymname — write the asm symbol name for `ident`. Honours, in
// order: FFI mapping (@symbol), module mangling (private decls), bare
// name. Use everywhere a top-level name is emitted before `(SB)` or in
// a `TEXT name,$N` header.
fn emitsymname(c: *cgen, ident: str) void = {
let resolved: str = ffiresolve(c, ident);
if (resolved.ptr != ident.ptr) {
// FFI hit — emit the mapped linker symbol verbatim.
os.write(1, resolved.ptr, resolved.len: u64);
return;
};
let mod: str = modlookup(c, ident);
if (mod.len > 0) {
os.write(1, mod.ptr, mod.len: u64);
os.write(1, ".".ptr, 1u64);
};
os.write(1, ident.ptr, ident.len: u64);
};
// ---- FFI map ---------------------------------------------------------
fn fficollect(c: *cgen, file: *node) void = {
c.ffis = nil;
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == N_FNDECL) {
let a: *node = d.attr;
for (a != nil) {
if (a.kind == N_ATTR) {
let aname: str = a.str;
if (streq(aname, "symbol")) {
let symnode: *node = a.list;
if (symnode != nil) {
if (symnode.kind == N_STRLIT) {
let f: *ffi = amalloc(c.a, 48u64): *ffi;
f.ident = d.str;
f.symbol = symnode.str;
f.fnext = c.ffis;
c.ffis = f;
};
};
};
};
a = a.next;
};
};
d = d.next;
};
};
fn ffiresolve(c: *cgen, ident: str) str = {
let f: *ffi = c.ffis;
for (f != nil) {
let id: str = f.ident;
if (streq(id, ident)) { return f.symbol; };
f = f.fnext;
};
return ident;
};
// ---- ABI argreg helpers ---------------------------------------------
fn argregname(i: i32) str = {
if (i == 0) { return "DI"; };
if (i == 1) { return "SI"; };
if (i == 2) { return "DX"; };
if (i == 3) { return "CX"; };
if (i == 4) { return "R8"; };
if (i == 5) { return "R9"; };
return "?";
};