Files
ww/selfhost/cmd/w6l/sym.ww
Hojun-Cho 469ec85551 selfhost/cmd/{w6a,w6l}: migrate 6 amalloc sites to alloc(T{...})!
Phase 0 batch 1. Six typed-struct allocations switch from
amalloc(arena, NNu64): *T over-sized byte counts to alloc(T{...})!
with partial struct literal initialization. MAP_ANON-zero from
rt_alloc covers any field the literal omits — same contract the
amalloc bump arena provided via its explicit zero loop, but
without the rule-13 size literal at the call site.

Converted:
 - w6a/asm.ww  addreloc, addrelocdata, addfixup (areloc, afixup)
 - w6l/sym.ww  intern (lsym)
 - w6l/dyn.ww  loadso (lso), lexport

lexport's `if (vernamecs == nil) { e.version.ptr = nil;
e.version.len = 0i32; }` branch dropped — MAP_ANON-zero provides
the empty-version slot for free; the inverted `if (vernamecs !=
nil)` only takes the dcstrtostr path.

w6l/sym.ww:20 comment updated from "amalloc-zeroing" to
"alloc-zeroing gives 0, not -1" so the documented mechanism
matches the call.

w6a/obj.ww's 2 remaining amalloc sites (bufinit + bufgrow) are
runtime-N *u8 byte buffers, deferred to #8 (runtime-N alloc API).

Verified via 991_w6a_ww + 992_w6l_ww + 995_self_rebuild +
996_dyn_ww byte-identity. make test 132/132.
2026-05-20 18:07:32 +09:00

112 lines
2.8 KiB
Plaintext

// selfhost/cmd/w6l/sym.ww — port of cmd/w6l/sym.c.
//
// Linker symbol table. Singly-linked list, usually a few hundred
// entries; hashing isn't worth it yet.
package w6l;
import mem;
type lsym = struct {
name: str,
val: u64, // offset within combined .text (or .data when
// indata=1) once linked
defined: i32, // 1 if some lobj defines this symbol
indata: i32, // 1 if defined in .data (writable globals)
owner: *lobj,
idxinowner: i32,
// Dynamic-linking fields. Set by resolve when an undefined sym
// is provided by some loaded lso. pltidx and dynsymidx default
// to -1 (set explicitly by resolve; alloc-zeroing gives 0, not -1).
isdyn: i32,
dynlib: *lso,
dynversion: str, // matched export's version; len 0 if none
pltidx: i32,
dynsymidx: i32,
snext: *lsym,
};
type lrel = struct {
off: u64, // offset within the relocation's section
section: i32, // 0 = .text, 1 = .data
kind: i32, // R_X86_64_*
sym: *lsym,
addend: i64,
rnext: *lrel,
};
type lobj = struct {
path: str,
buf: *u8, // object bytes
len: u64,
textoff: u64, // offset of .text in combined output
textsize: u64,
dataoff: u64, // offset of .data in combined output
datasize: u64, // bytes contributed to combined .data (0 if none)
onext: *lobj,
};
// lexport — one entry per GLOBAL/WEAK symbol exported by a loaded .so.
// Stored as a chain in the order the .so's dynsym presents them, so
// soprovides_v's first-match semantics agree with the C version.
type lexport = struct {
name: str,
version: str, // len 0 for unversioned globals
enext: *lexport,
};
type lso = struct {
path: str, // full filesystem path used to load
soname: str, // DT_SONAME, or basename if missing
exports: *lexport, // dynsym-order chain of exported names
sonext: *lso,
};
type lnk = struct {
a: *arena,
objs: *lobj,
sos: *lso,
syms: *lsym,
rels: *lrel,
text: *u8, // combined .text
textcap: u64,
textlen: u64,
// Combined .data (writable). Empty unless any input .o has a
// .data PROGBITS section.
data: *u8,
datacap: u64,
datalen: u64,
errs: i32,
dynn: i32, // number of syms routed through PLT
};
fn streq(a: str, b: str) bool = {
if (a.len != b.len) { return false; };
let i: i32 = 0;
for (i < a.len) {
if (a[i] != b[i]) { return false; };
i += 1;
};
return true;
};
export fn intern(l: *lnk, name: str) *lsym = {
let s: *lsym = l.syms;
for (s != nil) {
if (streq(s.name, name)) { return s; };
s = s.snext;
};
let n: *lsym = alloc(lsym { name = name, snext = l.syms })!;
l.syms = n;
return n;
};
export fn lookup(l: *lnk, name: str) *lsym = {
let s: *lsym = l.syms;
for (s != nil) {
if (streq(s.name, name)) { return s; };
s = s.snext;
};
return nil;
};