Hojun-Cho/ww
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

selfhost: mirror writable .data + R_X86_64_64 across the wwstage

Browse Source 
Bring the wwstage toolchain to parity with C-side DATAW / DATAR /
.data / .rela.data support. With this, w6c_ww + w6a_ww + w6l_ww can
compile, assemble and link `let g: str = "lit";` (and the scalar /
str / slice / struct globals that landed earlier) end-to-end, with
output that's byte-identical to the C-side pipeline.

w6a (types.ww / parse.ww / asm.ww / obj.ww):
  - A_DATAW + A_DATAR opcodes; parser learns `name+disp(SB)`;
    A_DATAR records an R_X86_64_64 reloc in .data via the new
    addrelocdata helper; areloc gains a `section` flag and asym
    an `isdata` flag; obj.ww splits relocs into .rela.text /
    .rela.data, emits .data PROGBITS + .rela.data conditionally,
    and shuffles section indices the same way cmd/w6a/obj.c does
    so byte output stays identical when no DATAW/DATAR are used.

w6l (sym.ww / obj.ww / pass.ww / out.ww / dynout.ww / main.ww):
  - lrel grows `section`; lsym grows `indata`; lobj tracks
    dataoff / datasize; lnk grows combined .data buffer;
  - obj.ww loads .data and .rela.data, registers data symbols
    with indata=1 and val shifted by the input's data_off, and
    the archive scanner includes both .text and .data globals;
  - pass.ww adds R_X86_64_64 (patch 8 bytes in .text or .data
    with sym_va + addend); relocate's signature becomes
    (textva, datava);
  - out.ww emits a second PT_LOAD (R+W) when datalen > 0, with
    .data at the page-aligned offset after .text;
  - dynout.ww refuses .data + -l/-L cleanly (matches the C-side
    error message);
  - main.ww computes text_va / data_va and passes both to
    relocate.

wcc cgen (cgen.ww / cgendecl.ww):
  - letpreintern walks top-level str-lets and interns the strlit
    BEFORE emitdatasection emits its DATA row, so emitletdataw
    can later look up the same label;
  - emitletdataw's 16B branch detects non-empty strlit init and
    emits the 8-zero + 8-LE-len DATAW plus a DATAR slot+0,strlit
    reloc, mirroring cmd/w6c/cgen.c.

Verified: `wwdump_ww -c` byte-matches `w6c` on a `let g: str =
"hello world\n";` fixture; `w6a_ww` and `w6l_ww` produce a
binary byte-identical to the C-side pipeline that runs and
prints "hello world". Bootstrap fixed-point holds (ww2 == ww3 ==
ww4), 26/26 tests green.
This commit is contained in:
Hojun-Cho
2026-05-12 13:03:03 +09:00
parent 003f707618
commit 1ac9980f7e
16 changed files with 1132 additions and 145 deletions
Download Patch File Download Diff File Expand all files Collapse all files

66
selfhost/cmd/w6a/asm.ww
View File

@@ -38,8 +38,9 @@ export fn emitu32(a: *asm_, v: u32) void = {
};
export fn addreloc(a: *asm_, off: u64, kind: i32, s: *asym, add: i64) void = {
let r: *areloc = amalloc(a.a, 48u64): *areloc;
let r: *areloc = amalloc(a.a, 64u64): *areloc;
r.off = off;
r.section = 0; // .text
r.kind = kind;
r.asy = s;
r.addend = add;
@@ -47,6 +48,37 @@ export fn addreloc(a: *asm_, off: u64, kind: i32, s: *asym, add: i64) void = {
a.relocs = r;
};
// Record a relocation that lives in the .data section. Used by
// DATAR to patch a 64-bit slot with a symbol's runtime VA. obj.ww
// separates these into .rela.data when emitting the .o.
export fn addrelocdata(a: *asm_, off: u64, kind: i32, s: *asym, add: i64) void = {
let r: *areloc = amalloc(a.a, 64u64): *areloc;
r.off = off;
r.section = 1; // .data
r.kind = kind;
r.asy = s;
r.addend = add;
r.rnext = a.relocs;
a.relocs = r;
};
// Append one byte to the writable .data buffer. Mirrors emitbyte
// but targets a.data instead of a.text.
export fn emitdatabyte(a: *asm_, b: u8) void = {
if (a.datalen + 1u64 > a.datacap) {
let nc: u64 = a.datacap;
if (nc == 0u64) { nc = 256u64; };
nc = nc * 2u64;
let nb: *u8 = os.alloc(nc): *u8;
let i: u64 = 0u64;
for (i < a.datalen) { nb[i] = a.data[i]; i += 1u64; };
a.data = nb;
a.datacap = nc;
};
a.data[a.datalen] = b;
a.datalen += 1u64;
};
// ---- register codes ----------------------------------------------------
// Low 3 bits of register encoding.
@@ -270,6 +302,38 @@ export fn encode(a: *asm_) i32 = {
for (i < p.nbytes) { emitbyte(a, p.bytes[i]); i += 1u64; };
p = p.link; continue;
};
if (op == A_DATAW) {
// Writable variant: bytes go into .data instead of
// .text. obj.ww emits the extra section conditionally
// on datalen > 0 so .o output stays byte-identical
// for inputs that don't use DATAW.
let s: *asym = intern(a, p.to.asym);
s.defined = 1;
s.isdata = 1;
s.isglobal = 1;
s.addr = a.datalen;
let i: u64 = 0u64;
for (i < p.nbytes) { emitdatabyte(a, p.bytes[i]); i += 1u64; };
p = p.link; continue;
};
if (op == A_DATAR) {
// DATAR slot+off(SB), target(SB) — record an
// R_X86_64_64 relocation at slot+off in .data
// pointing at target. Slot must already be defined
// by a prior DATAW.
let holder: *asym = intern(a, p.from.asym);
if (holder.defined == 0) {
p = p.link; continue;
};
if (holder.isdata == 0) {
p = p.link; continue;
};
let target: *asym = intern(a, p.to.asym);
let reloff: u64 = holder.addr + p.from.offset: u64;
addrelocdata(a, reloff, 1 /* R_X86_64_64 */,
target, 0i64);
p = p.link; continue;
};
if (op == A_RET) {
emitbyte(a, 195u8); // 0xC3
p = p.link; continue;

228
selfhost/cmd/w6a/main.combined.ww
View File

@@ -459,6 +459,13 @@ def A_JNZ: i32 = 58;
def A_SYSCALL: i32 = 59;
// Writable data + reloc-only data. Mirror cmd/w6c/6.out.h.
// A_DATAW: bytes land in .data (RW) instead of .text.
// A_DATAR: record an R_X86_64_64 reloc at a .data slot, patched
// to a target symbol's runtime VA at link time.
def A_DATAW: i32 = 60;
def A_DATAR: i32 = 61;
// ---- structs (mirror cmd/w6a/a.h) --------------------------------------
type aoperand = struct {
@@ -487,14 +494,16 @@ type asym = struct {
name: str,
defined: i32,
istext: i32,
isdata: i32, // mutually exclusive with istext; DATAW symbols
isglobal: i32,
addr: u64,
addr: u64, // offset within its section (.text or .data)
idx: i32,
snext: *asym,
};
type areloc = struct {
off: u64,
section: i32, // 0 = .text, 1 = .data
kind: i32,
asy: *asym,
addend: i64,
@@ -522,6 +531,14 @@ type asm_ = struct {
textcap: u64,
textlen: u64,
// Writable .data. Empty unless any DATAW directive was seen;
// obj.ww emits the extra section conditionally so .o output
// stays byte-identical for inputs that don't use DATAW (test
// 991 byte-diff invariant).
data: *u8,
datacap: u64,
datalen: u64,
syms: *asym,
relocs: *areloc,
fixups: *afixup,
@@ -686,6 +703,8 @@ fn opcodelookup(p: *u8, n: u64) i32 = {
if (streqlit(p, n, "SYSCALL")) { return A_SYSCALL; };
if (streqlit(p, n, "TEXT")) { return A_TEXT; };
if (streqlit(p, n, "DATA")) { return A_DATA; };
if (streqlit(p, n, "DATAW")) { return A_DATAW; };
if (streqlit(p, n, "DATAR")) { return A_DATAR; };
return A_NOP;
};
@@ -891,13 +910,25 @@ fn parseoperand(a: *asm_, p: *u8, offin: u64, n: u64, out: *aoperand) u64 = {
return after;
};
// IDENT — register, symbol(SB), or branch label
// IDENT — register, symbol(SB), symbol+disp(SB), or branch label
if (isidstart(c0: i32)) {
let istart: u64 = off;
for (off < n) {
if (isidcont(p[off]: i32)) { off += 1u64; continue; };
off = off; off += 0u64; // loop break
let in_: u64 = off - istart;
// Optional `+disp` between the ident and `(SB)`. Used
// by DATAR to address bytes within a previously-defined
// .data slot (e.g. `DATAR s+8(SB),...`).
let symdisp: i64 = 0i64;
if (off < n) { if (p[off] == 43u8) { // '+'
off += 1u64;
let v: i64;
let used: u64;
v, used = parsenum(p + off, n - off);
symdisp = v;
off += used;
};};
// IDENT(SB) — external
if (off < n) { if (p[off] == 40u8) { // '('
let rstart: u64 = off + 1u64;
@@ -910,6 +941,7 @@ fn parseoperand(a: *asm_, p: *u8, offin: u64, n: u64, out: *aoperand) u64 = {
if (r == D_PSB) {
out.atype = D_EXTERN;
out.asym = dupstr(a.a, p + istart, in_);
out.offset = symdisp;
} else {
out.atype = D_INDIR;
out.reg = r;
@@ -1059,8 +1091,9 @@ export fn parse(a: *asm_) i32 = {
a.line += 1; continue;
};
if (opc == A_DATA) {
// DATA name(SB),"escaped bytes" — find first '(' as end of name.
if (opc == A_DATA || opc == A_DATAW) {
// DATA / DATAW name(SB),"escaped bytes" — same syntax,
// different destination section (.text vs .data).
let q: u64 = r0;
let lparen: u64 = n;
let scand: bool = true;
@@ -1209,8 +1242,9 @@ export fn emitu32(a: *asm_, v: u32) void = {
};
export fn addreloc(a: *asm_, off: u64, kind: i32, s: *asym, add: i64) void = {
let r: *areloc = amalloc(a.a, 48u64): *areloc;
let r: *areloc = amalloc(a.a, 64u64): *areloc;
r.off = off;
r.section = 0; // .text
r.kind = kind;
r.asy = s;
r.addend = add;
@@ -1218,6 +1252,37 @@ export fn addreloc(a: *asm_, off: u64, kind: i32, s: *asym, add: i64) void = {
a.relocs = r;
};
// Record a relocation that lives in the .data section. Used by
// DATAR to patch a 64-bit slot with a symbol's runtime VA. obj.ww
// separates these into .rela.data when emitting the .o.
export fn addrelocdata(a: *asm_, off: u64, kind: i32, s: *asym, add: i64) void = {
let r: *areloc = amalloc(a.a, 64u64): *areloc;
r.off = off;
r.section = 1; // .data
r.kind = kind;
r.asy = s;
r.addend = add;
r.rnext = a.relocs;
a.relocs = r;
};
// Append one byte to the writable .data buffer. Mirrors emitbyte
// but targets a.data instead of a.text.
export fn emitdatabyte(a: *asm_, b: u8) void = {
if (a.datalen + 1u64 > a.datacap) {
let nc: u64 = a.datacap;
if (nc == 0u64) { nc = 256u64; };
nc = nc * 2u64;
let nb: *u8 = os.alloc(nc): *u8;
let i: u64 = 0u64;
for (i < a.datalen) { nb[i] = a.data[i]; i += 1u64; };
a.data = nb;
a.datacap = nc;
};
a.data[a.datalen] = b;
a.datalen += 1u64;
};
// ---- register codes ----------------------------------------------------
// Low 3 bits of register encoding.
@@ -1441,6 +1506,38 @@ export fn encode(a: *asm_) i32 = {
for (i < p.nbytes) { emitbyte(a, p.bytes[i]); i += 1u64; };
p = p.link; continue;
};
if (op == A_DATAW) {
// Writable variant: bytes go into .data instead of
// .text. obj.ww emits the extra section conditionally
// on datalen > 0 so .o output stays byte-identical
// for inputs that don't use DATAW.
let s: *asym = intern(a, p.to.asym);
s.defined = 1;
s.isdata = 1;
s.isglobal = 1;
s.addr = a.datalen;
let i: u64 = 0u64;
for (i < p.nbytes) { emitdatabyte(a, p.bytes[i]); i += 1u64; };
p = p.link; continue;
};
if (op == A_DATAR) {
// DATAR slot+off(SB), target(SB) — record an
// R_X86_64_64 relocation at slot+off in .data
// pointing at target. Slot must already be defined
// by a prior DATAW.
let holder: *asym = intern(a, p.from.asym);
if (holder.defined == 0) {
p = p.link; continue;
};
if (holder.isdata == 0) {
p = p.link; continue;
};
let target: *asym = intern(a, p.to.asym);
let reloff: u64 = holder.addr + p.from.offset: u64;
addrelocdata(a, reloff, 1 /* R_X86_64_64 */,
target, 0i64);
p = p.link; continue;
};
if (op == A_RET) {
emitbyte(a, 195u8); // 0xC3
p = p.link; continue;
@@ -1902,12 +1999,14 @@ def SHT_SYMTAB_C: u32 = 2u32;
def SHT_STRTAB_C: u32 = 3u32;
def SHT_RELA_C: u32 = 4u32;
def SHF_WRITE: u64 = 1u64;
def SHF_ALLOC: u64 = 2u64;
def SHF_EXECINSTR: u64 = 4u64;
def SHF_INFO_LINK: u64 = 64u64; // 0x40
def STB_GLOBAL: u8 = 1u8;
def STT_NOTYPE: u8 = 0u8;
def STT_OBJECT: u8 = 1u8;
def STT_FUNC: u8 = 2u8;
// Sizes of fixed structures.
@@ -1987,15 +2086,56 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let str_: buf; bufinit(&str_, a.a);
let sym: buf; bufinit(&sym, a.a);
let rela: buf; bufinit(&rela, a.a);
let relad: buf; bufinit(&relad, a.a);
// Index 0 = empty.
let zero: u8 = 0u8;
bufputb(&shstr, &zero, 1u64);
bufputb(&str_, &zero, 1u64);
// Section name offsets.
let hasdata: bool = a.datalen > 0u64;
let hasdatarelocs: bool = false;
let rscan: *areloc = a.relocs;
for (rscan != nil) {
if (rscan.section == 1) { hasdatarelocs = true; };
rscan = rscan.rnext;
};
// Section indices (mirror cmd/w6a/obj.c):
// without data, without data-relocs:
// 1=.text 2=.rela.text 3=.symtab 4=.strtab 5=.shstrtab
// with data, no data-relocs:
// 1=.text 2=.rela.text 3=.data 4=.symtab 5=.strtab 6=.shstrtab
// with data + data-relocs:
// 1=.text 2=.rela.text 3=.data 4=.rela.data 5=.symtab
// 6=.strtab 7=.shstrtab
let SH_TEXT: u16 = 1u16;
let SH_DATA: u16 = 0u16;
let SH_RELAD: u16 = 0u16;
let SH_SYMTAB: u16 = 3u16;
if (hasdata) {
SH_DATA = 3u16;
if (hasdatarelocs) {
SH_RELAD = 4u16;
SH_SYMTAB = 5u16;
} else {
SH_SYMTAB = 4u16;
};
};
let SH_STRTAB: u16 = SH_SYMTAB + 1u16;
let SH_SHSTR: u16 = SH_STRTAB + 1u16;
// Section name offsets. Append .data / .rela.data only when
// used so the .shstrtab buffer stays byte-identical for the
// no-DATAW case (test 991 byte-diff invariant).
let shntext: u32 = bufputcstr(&shstr, ".text");
let shnrela: u32 = bufputcstr(&shstr, ".rela.text");
let shndata: u32 = 0u32;
let shnrelad: u32 = 0u32;
if (hasdata) { shndata = bufputcstr(&shstr, ".data"); };
if (hasdata) { if (hasdatarelocs) {
shnrelad = bufputcstr(&shstr, ".rela.data");
};};
let shnsymtab: u32 = bufputcstr(&shstr, ".symtab");
let shnstrtab: u32 = bufputcstr(&shstr, ".strtab");
let shnshstrtab: u32 = bufputcstr(&shstr, ".shstrtab");
@@ -2006,8 +2146,6 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
for (zi < 24) { zsym[zi] = 0u8; zi += 1; };
bufputb(&sym, zsym.ptr, 24u64);
let SH_TEXT: u16 = 1u16;
// Build symbols.
let idx: i32 = 1;
let s: *asym = a.syms;
@@ -2018,8 +2156,13 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let stname: u32 = bufputcstr(&str_, s.name);
wru32(entry.ptr, 0u64, stname);
if (s.defined != 0) {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_FUNC));
wru16(entry.ptr, 6u64, SH_TEXT);
if (s.isdata != 0) {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_OBJECT));
wru16(entry.ptr, 6u64, SH_DATA);
} else {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_FUNC));
wru16(entry.ptr, 6u64, SH_TEXT);
};
wru64(entry.ptr, 8u64, s.addr);
} else {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_NOTYPE));
@@ -2031,7 +2174,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
s = s.snext;
};
// Build relocations.
// Build relocations — split into text vs data buffers.
let r: *areloc = a.relocs;
for (r != nil) {
let entry: [24]u8;
@@ -2039,7 +2182,11 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let rinfo: u64 = (r.asy.idx: u64 << 32u64) | (r.kind: u64 & 4294967295u64);
wru64(entry.ptr, 8u64, rinfo);
wru64(entry.ptr, 16u64, r.addend: u64);
bufputb(&rela, entry.ptr, 24u64);
if (r.section == 1) {
bufputb(&relad, entry.ptr, 24u64);
} else {
bufputb(&rela, entry.ptr, 24u64);
};
r = r.rnext;
};
@@ -2047,12 +2194,20 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let off: u64 = EHDR_SZ;
let offtext: u64 = off; off = off + a.textlen;
let offrela: u64 = off; off = off + rela.n;
let offdata: u64 = off; if (hasdata) { off = off + a.datalen; };
let offrelad: u64 = off; if (hasdata) { if (hasdatarelocs) {
off = off + relad.n;
};};
let offsym: u64 = off; off = off + sym.n;
let offstr: u64 = off; off = off + str_.n;
let offshstr: u64 = off; off = off + shstr.n;
for ((off & 7u64) != 0u64) { off += 1u64; };
let offshdr: u64 = off;
let NSECT: u16 = 6u16;
if (hasdata) {
if (hasdatarelocs) { NSECT = 8u16; }
else { NSECT = 7u16; };
};
// ---- Ehdr ----
let eh: [64]u8;
@@ -2077,7 +2232,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
wru16(eh.ptr, 56u64, 0u16); // e_phnum
wru16(eh.ptr, 58u64, 64u16); // e_shentsize
wru16(eh.ptr, 60u64, NSECT); // e_shnum
wru16(eh.ptr, 62u64, 5u16); // e_shstrndx
wru16(eh.ptr, 62u64, SH_SHSTR); // e_shstrndx
if (!wrn(fd, eh.ptr, 64u64, 64i64)) { return -1; };
if (a.textlen > 0u64) {
@@ -2086,6 +2241,16 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
if (rela.n > 0u64) {
if (!wrn(fd, rela.p, rela.n, rela.n: i64)) { return -1; };
};
if (hasdata) {
if (a.datalen > 0u64) {
if (!wrn(fd, a.data, a.datalen, a.datalen: i64)) { return -1; };
};
if (hasdatarelocs) {
if (relad.n > 0u64) {
if (!wrn(fd, relad.p, relad.n, relad.n: i64)) { return -1; };
};
};
};
if (sym.n > 0u64) {
if (!wrn(fd, sym.p, sym.n, sym.n: i64)) { return -1; };
};
@@ -2098,6 +2263,10 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
// Pad to 8 before shdrs.
let written: u64 = EHDR_SZ + a.textlen + rela.n + sym.n + str_.n + shstr.n;
if (hasdata) {
written += a.datalen;
if (hasdatarelocs) { written += relad.n; };
};
for ((written & 7u64) != 0u64) {
wrdrop(fd, &zero, 1u64);
written += 1u64;
@@ -2127,11 +2296,38 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
wru64(shbuf.ptr, 8u64, SHF_INFO_LINK);
wru64(shbuf.ptr, 24u64, offrela);
wru64(shbuf.ptr, 32u64, rela.n);
wru32(shbuf.ptr, 40u64, 3u32); // sh_link = symtab idx
wru32(shbuf.ptr, 44u64, 1u32); // sh_info = .text idx
wru32(shbuf.ptr, 40u64, SH_SYMTAB: u32); // sh_link
wru32(shbuf.ptr, 44u64, 1u32); // sh_info = .text idx
wru64(shbuf.ptr, 48u64, 8u64);
wru64(shbuf.ptr, 56u64, RELA_SZ);
wrdrop(fd, shbuf.ptr, 64u64);
if (hasdata) {
// .data
sn = 0;
for (sn < 64) { shbuf[sn] = 0u8; sn += 1; };
wru32(shbuf.ptr, 0u64, shndata);
wru32(shbuf.ptr, 4u64, SHT_PROGBITS_C);
wru64(shbuf.ptr, 8u64, SHF_ALLOC | SHF_WRITE);
wru64(shbuf.ptr, 24u64, offdata);
wru64(shbuf.ptr, 32u64, a.datalen);
wru64(shbuf.ptr, 48u64, 8u64); // sh_addralign
wrdrop(fd, shbuf.ptr, 64u64);
if (hasdatarelocs) {
// .rela.data
sn = 0;
for (sn < 64) { shbuf[sn] = 0u8; sn += 1; };
wru32(shbuf.ptr, 0u64, shnrelad);
wru32(shbuf.ptr, 4u64, SHT_RELA_C);
wru64(shbuf.ptr, 8u64, SHF_INFO_LINK);
wru64(shbuf.ptr, 24u64, offrelad);
wru64(shbuf.ptr, 32u64, relad.n);
wru32(shbuf.ptr, 40u64, SH_SYMTAB: u32);
wru32(shbuf.ptr, 44u64, SH_DATA: u32); // applies to .data
wru64(shbuf.ptr, 48u64, 8u64);
wru64(shbuf.ptr, 56u64, RELA_SZ);
wrdrop(fd, shbuf.ptr, 64u64);
};
};
// .symtab
sn = 0;
for (sn < 64) { shbuf[sn] = 0u8; sn += 1; };
@@ -2139,7 +2335,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
wru32(shbuf.ptr, 4u64, SHT_SYMTAB_C);
wru64(shbuf.ptr, 24u64, offsym);
wru64(shbuf.ptr, 32u64, sym.n);
wru32(shbuf.ptr, 40u64, 4u32); // sh_link = strtab
wru32(shbuf.ptr, 40u64, SH_STRTAB: u32); // sh_link
wru32(shbuf.ptr, 44u64, 1u32); // sh_info = one local (STN_UNDEF)
wru64(shbuf.ptr, 48u64, 8u64);
wru64(shbuf.ptr, 56u64, SYM_SZ);

121
selfhost/cmd/w6a/obj.ww
View File

@@ -48,12 +48,14 @@ def SHT_SYMTAB_C: u32 = 2u32;
def SHT_STRTAB_C: u32 = 3u32;
def SHT_RELA_C: u32 = 4u32;
def SHF_WRITE: u64 = 1u64;
def SHF_ALLOC: u64 = 2u64;
def SHF_EXECINSTR: u64 = 4u64;
def SHF_INFO_LINK: u64 = 64u64; // 0x40
def STB_GLOBAL: u8 = 1u8;
def STT_NOTYPE: u8 = 0u8;
def STT_OBJECT: u8 = 1u8;
def STT_FUNC: u8 = 2u8;
// Sizes of fixed structures.
@@ -133,15 +135,56 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let str_: buf; bufinit(&str_, a.a);
let sym: buf; bufinit(&sym, a.a);
let rela: buf; bufinit(&rela, a.a);
let relad: buf; bufinit(&relad, a.a);
// Index 0 = empty.
let zero: u8 = 0u8;
bufputb(&shstr, &zero, 1u64);
bufputb(&str_, &zero, 1u64);
// Section name offsets.
let hasdata: bool = a.datalen > 0u64;
let hasdatarelocs: bool = false;
let rscan: *areloc = a.relocs;
for (rscan != nil) {
if (rscan.section == 1) { hasdatarelocs = true; };
rscan = rscan.rnext;
};
// Section indices (mirror cmd/w6a/obj.c):
// without data, without data-relocs:
// 1=.text 2=.rela.text 3=.symtab 4=.strtab 5=.shstrtab
// with data, no data-relocs:
// 1=.text 2=.rela.text 3=.data 4=.symtab 5=.strtab 6=.shstrtab
// with data + data-relocs:
// 1=.text 2=.rela.text 3=.data 4=.rela.data 5=.symtab
// 6=.strtab 7=.shstrtab
let SH_TEXT: u16 = 1u16;
let SH_DATA: u16 = 0u16;
let SH_RELAD: u16 = 0u16;
let SH_SYMTAB: u16 = 3u16;
if (hasdata) {
SH_DATA = 3u16;
if (hasdatarelocs) {
SH_RELAD = 4u16;
SH_SYMTAB = 5u16;
} else {
SH_SYMTAB = 4u16;
};
};
let SH_STRTAB: u16 = SH_SYMTAB + 1u16;
let SH_SHSTR: u16 = SH_STRTAB + 1u16;
// Section name offsets. Append .data / .rela.data only when
// used so the .shstrtab buffer stays byte-identical for the
// no-DATAW case (test 991 byte-diff invariant).
let shntext: u32 = bufputcstr(&shstr, ".text");
let shnrela: u32 = bufputcstr(&shstr, ".rela.text");
let shndata: u32 = 0u32;
let shnrelad: u32 = 0u32;
if (hasdata) { shndata = bufputcstr(&shstr, ".data"); };
if (hasdata) { if (hasdatarelocs) {
shnrelad = bufputcstr(&shstr, ".rela.data");
};};
let shnsymtab: u32 = bufputcstr(&shstr, ".symtab");
let shnstrtab: u32 = bufputcstr(&shstr, ".strtab");
let shnshstrtab: u32 = bufputcstr(&shstr, ".shstrtab");
@@ -152,8 +195,6 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
for (zi < 24) { zsym[zi] = 0u8; zi += 1; };
bufputb(&sym, zsym.ptr, 24u64);
let SH_TEXT: u16 = 1u16;
// Build symbols.
let idx: i32 = 1;
let s: *asym = a.syms;
@@ -164,8 +205,13 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let stname: u32 = bufputcstr(&str_, s.name);
wru32(entry.ptr, 0u64, stname);
if (s.defined != 0) {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_FUNC));
wru16(entry.ptr, 6u64, SH_TEXT);
if (s.isdata != 0) {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_OBJECT));
wru16(entry.ptr, 6u64, SH_DATA);
} else {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_FUNC));
wru16(entry.ptr, 6u64, SH_TEXT);
};
wru64(entry.ptr, 8u64, s.addr);
} else {
wru8(entry.ptr, 4u64, ((STB_GLOBAL << 4u8) | STT_NOTYPE));
@@ -177,7 +223,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
s = s.snext;
};
// Build relocations.
// Build relocations — split into text vs data buffers.
let r: *areloc = a.relocs;
for (r != nil) {
let entry: [24]u8;
@@ -185,7 +231,11 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let rinfo: u64 = (r.asy.idx: u64 << 32u64) | (r.kind: u64 & 4294967295u64);
wru64(entry.ptr, 8u64, rinfo);
wru64(entry.ptr, 16u64, r.addend: u64);
bufputb(&rela, entry.ptr, 24u64);
if (r.section == 1) {
bufputb(&relad, entry.ptr, 24u64);
} else {
bufputb(&rela, entry.ptr, 24u64);
};
r = r.rnext;
};
@@ -193,12 +243,20 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
let off: u64 = EHDR_SZ;
let offtext: u64 = off; off = off + a.textlen;
let offrela: u64 = off; off = off + rela.n;
let offdata: u64 = off; if (hasdata) { off = off + a.datalen; };
let offrelad: u64 = off; if (hasdata) { if (hasdatarelocs) {
off = off + relad.n;
};};
let offsym: u64 = off; off = off + sym.n;
let offstr: u64 = off; off = off + str_.n;
let offshstr: u64 = off; off = off + shstr.n;
for ((off & 7u64) != 0u64) { off += 1u64; };
let offshdr: u64 = off;
let NSECT: u16 = 6u16;
if (hasdata) {
if (hasdatarelocs) { NSECT = 8u16; }
else { NSECT = 7u16; };
};
// ---- Ehdr ----
let eh: [64]u8;
@@ -223,7 +281,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
wru16(eh.ptr, 56u64, 0u16); // e_phnum
wru16(eh.ptr, 58u64, 64u16); // e_shentsize
wru16(eh.ptr, 60u64, NSECT); // e_shnum
wru16(eh.ptr, 62u64, 5u16); // e_shstrndx
wru16(eh.ptr, 62u64, SH_SHSTR); // e_shstrndx
if (!wrn(fd, eh.ptr, 64u64, 64i64)) { return -1; };
if (a.textlen > 0u64) {
@@ -232,6 +290,16 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
if (rela.n > 0u64) {
if (!wrn(fd, rela.p, rela.n, rela.n: i64)) { return -1; };
};
if (hasdata) {
if (a.datalen > 0u64) {
if (!wrn(fd, a.data, a.datalen, a.datalen: i64)) { return -1; };
};
if (hasdatarelocs) {
if (relad.n > 0u64) {
if (!wrn(fd, relad.p, relad.n, relad.n: i64)) { return -1; };
};
};
};
if (sym.n > 0u64) {
if (!wrn(fd, sym.p, sym.n, sym.n: i64)) { return -1; };
};
@@ -244,6 +312,10 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
// Pad to 8 before shdrs.
let written: u64 = EHDR_SZ + a.textlen + rela.n + sym.n + str_.n + shstr.n;
if (hasdata) {
written += a.datalen;
if (hasdatarelocs) { written += relad.n; };
};
for ((written & 7u64) != 0u64) {
wrdrop(fd, &zero, 1u64);
written += 1u64;
@@ -273,11 +345,38 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
wru64(shbuf.ptr, 8u64, SHF_INFO_LINK);
wru64(shbuf.ptr, 24u64, offrela);
wru64(shbuf.ptr, 32u64, rela.n);
wru32(shbuf.ptr, 40u64, 3u32); // sh_link = symtab idx
wru32(shbuf.ptr, 44u64, 1u32); // sh_info = .text idx
wru32(shbuf.ptr, 40u64, SH_SYMTAB: u32); // sh_link
wru32(shbuf.ptr, 44u64, 1u32); // sh_info = .text idx
wru64(shbuf.ptr, 48u64, 8u64);
wru64(shbuf.ptr, 56u64, RELA_SZ);
wrdrop(fd, shbuf.ptr, 64u64);
if (hasdata) {
// .data
sn = 0;
for (sn < 64) { shbuf[sn] = 0u8; sn += 1; };
wru32(shbuf.ptr, 0u64, shndata);
wru32(shbuf.ptr, 4u64, SHT_PROGBITS_C);
wru64(shbuf.ptr, 8u64, SHF_ALLOC | SHF_WRITE);
wru64(shbuf.ptr, 24u64, offdata);
wru64(shbuf.ptr, 32u64, a.datalen);
wru64(shbuf.ptr, 48u64, 8u64); // sh_addralign
wrdrop(fd, shbuf.ptr, 64u64);
if (hasdatarelocs) {
// .rela.data
sn = 0;
for (sn < 64) { shbuf[sn] = 0u8; sn += 1; };
wru32(shbuf.ptr, 0u64, shnrelad);
wru32(shbuf.ptr, 4u64, SHT_RELA_C);
wru64(shbuf.ptr, 8u64, SHF_INFO_LINK);
wru64(shbuf.ptr, 24u64, offrelad);
wru64(shbuf.ptr, 32u64, relad.n);
wru32(shbuf.ptr, 40u64, SH_SYMTAB: u32);
wru32(shbuf.ptr, 44u64, SH_DATA: u32); // applies to .data
wru64(shbuf.ptr, 48u64, 8u64);
wru64(shbuf.ptr, 56u64, RELA_SZ);
wrdrop(fd, shbuf.ptr, 64u64);
};
};
// .symtab
sn = 0;
for (sn < 64) { shbuf[sn] = 0u8; sn += 1; };
@@ -285,7 +384,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = {
wru32(shbuf.ptr, 4u64, SHT_SYMTAB_C);
wru64(shbuf.ptr, 24u64, offsym);
wru64(shbuf.ptr, 32u64, sym.n);
wru32(shbuf.ptr, 40u64, 4u32); // sh_link = strtab
wru32(shbuf.ptr, 40u64, SH_STRTAB: u32); // sh_link
wru32(shbuf.ptr, 44u64, 1u32); // sh_info = one local (STN_UNDEF)
wru64(shbuf.ptr, 48u64, 8u64);
wru64(shbuf.ptr, 56u64, SYM_SZ);

22
selfhost/cmd/w6a/parse.ww
View File

@@ -86,6 +86,8 @@ fn opcodelookup(p: *u8, n: u64) i32 = {
if (streqlit(p, n, "SYSCALL")) { return A_SYSCALL; };
if (streqlit(p, n, "TEXT")) { return A_TEXT; };
if (streqlit(p, n, "DATA")) { return A_DATA; };
if (streqlit(p, n, "DATAW")) { return A_DATAW; };
if (streqlit(p, n, "DATAR")) { return A_DATAR; };
return A_NOP;
};
@@ -291,13 +293,25 @@ fn parseoperand(a: *asm_, p: *u8, offin: u64, n: u64, out: *aoperand) u64 = {
return after;
};
// IDENT — register, symbol(SB), or branch label
// IDENT — register, symbol(SB), symbol+disp(SB), or branch label
if (isidstart(c0: i32)) {
let istart: u64 = off;
for (off < n) {
if (isidcont(p[off]: i32)) { off += 1u64; continue; };
off = off; off += 0u64; // loop break
let in_: u64 = off - istart;
// Optional `+disp` between the ident and `(SB)`. Used
// by DATAR to address bytes within a previously-defined
// .data slot (e.g. `DATAR s+8(SB),...`).
let symdisp: i64 = 0i64;
if (off < n) { if (p[off] == 43u8) { // '+'
off += 1u64;
let v: i64;
let used: u64;
v, used = parsenum(p + off, n - off);
symdisp = v;
off += used;
};};
// IDENT(SB) — external
if (off < n) { if (p[off] == 40u8) { // '('
let rstart: u64 = off + 1u64;
@@ -310,6 +324,7 @@ fn parseoperand(a: *asm_, p: *u8, offin: u64, n: u64, out: *aoperand) u64 = {
if (r == D_PSB) {
out.atype = D_EXTERN;
out.asym = dupstr(a.a, p + istart, in_);
out.offset = symdisp;
} else {
out.atype = D_INDIR;
out.reg = r;
@@ -459,8 +474,9 @@ export fn parse(a: *asm_) i32 = {
a.line += 1; continue;
};
if (opc == A_DATA) {
// DATA name(SB),"escaped bytes" — find first '(' as end of name.
if (opc == A_DATA || opc == A_DATAW) {
// DATA / DATAW name(SB),"escaped bytes" — same syntax,
// different destination section (.text vs .data).
let q: u64 = r0;
let lparen: u64 = n;
let scand: bool = true;

19
selfhost/cmd/w6a/types.ww
View File

@@ -121,6 +121,13 @@ def A_JNZ: i32 = 58;
def A_SYSCALL: i32 = 59;
// Writable data + reloc-only data. Mirror cmd/w6c/6.out.h.
// A_DATAW: bytes land in .data (RW) instead of .text.
// A_DATAR: record an R_X86_64_64 reloc at a .data slot, patched
// to a target symbol's runtime VA at link time.
def A_DATAW: i32 = 60;
def A_DATAR: i32 = 61;
// ---- structs (mirror cmd/w6a/a.h) --------------------------------------
type aoperand = struct {
@@ -149,14 +156,16 @@ type asym = struct {
name: str,
defined: i32,
istext: i32,
isdata: i32, // mutually exclusive with istext; DATAW symbols
isglobal: i32,
addr: u64,
addr: u64, // offset within its section (.text or .data)
idx: i32,
snext: *asym,
};
type areloc = struct {
off: u64,
section: i32, // 0 = .text, 1 = .data
kind: i32,
asy: *asym,
addend: i64,
@@ -184,6 +193,14 @@ type asm_ = struct {
textcap: u64,
textlen: u64,
// Writable .data. Empty unless any DATAW directive was seen;
// obj.ww emits the extra section conditionally so .o output
// stays byte-identical for inputs that don't use DATAW (test
// 991 byte-diff invariant).
data: *u8,
datacap: u64,
datalen: u64,
syms: *asym,
relocs: *areloc,
fixups: *afixup,

111
selfhost/cmd/w6c/main.combined.ww
View File

@@ -8576,6 +8576,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
};
d = d.next;
};
letpreintern(c, file);
emitdatasection(c);
emitdefconstants(c, file);
emitletdataw(c, file);
@@ -9345,13 +9346,45 @@ fn emitdatawbyte(b: u8) void = {
os.write(1, bb.ptr, 1u64);
};
// letpreintern — intern strlits referenced from top-level str-let
// initialisers BEFORE emitdatasection runs. Mirrors cmd/w6c/cgen.c
// let_pre_intern: emitletdataw later looks up the same label, and
// emitdatasection emits the DATA row in the same .s file. Running
// emitletdataw after emitdatasection would flip the (DATA strlits,
// DATAW lets) section order and break byte-identity.
export fn letpreintern(c: *cgen, file: *node) void = {
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let sz: i32 = letemitsize(c, d);
if (sz == 16) {
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
if (r != nil) {
if (r.kind == nkind.N_STRLIT) {
if (r.str.len > 0) {
internstrlit(c, r.str);
};
};
};
};
};
d = d.next;
};
};
// emitletdataw — DATAW directive per top-level `let` global.
// 8B scalar with int/rune/bool/nil literal init (or no init).
// 16B str with no init (or `nil` / `""`) — zero header; the
// program must assign a real strlit at runtime before
// using .ptr / .len.
// Non-literal scalar inits and non-empty strlit inits are skipped
// so the link surfaces an undefined-symbol error.
// 16B str — no init / `nil` / `""` → 16 zero bytes; or non-empty
// strlit init → 8 zero placeholder + 8 LE len bytes plus a
// DATAR slot+0,strlit reloc that the linker patches at load.
// sz struct — zero only.
// Non-literal scalar inits and unsupported shapes are skipped so the
// link surfaces an undefined-symbol error if the binding is used.
fn emitletdataw(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
@@ -9394,31 +9427,67 @@ fn emitletdataw(c: *cgen, file: *node) void = {
};
};
if (sz == 16 && !issg) {
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_NIL) { ok = true; };
if (r.kind == nkind.N_STRLIT) {
if (r.str.len == 0) { ok = true; };
};
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
// str-literal init (non-empty): emit
// the 16B payload as 8 placeholder zero
// bytes + 8 LE bytes of length, then a
// DATAR reloc to patch the ptr half with
// the strlit's runtime VA.
let strlitinit: bool = false;
if (r != nil) {
if (r.kind == nkind.N_STRLIT) {
if (r.str.len > 0) { strlitinit = true; };
};
};
if (ok) {
if (strlitinit) {
let lab: str = internstrlit(c, r.str);
let v: u64 = r.str.len: u64;
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
for (i < 16) {
emitdatawbyte(0u8);
for (i < 8) { emitdatawbyte(0u8); i += 1; };
i = 0;
let nv: u64 = v;
for (i < 8) {
emitdatawbyte((nv & 255u64): u8);
nv = nv >> 8u64;
i += 1;
};
emitline("\"\n");
emitline("DATAR ");
emitsymname(c, nm);
emitline("+0(SB),");
os.write(1, lab.ptr, lab.len: u64);
emitline("(SB)\n");
} else {
// zero-init: accept no rhs, nil,
// or empty strlit.
let ok: bool = true;
if (d.rhs != nil) {
ok = false;
if (r != nil) {
if (r.kind == nkind.N_NIL) { ok = true; };
if (r.kind == nkind.N_STRLIT) {
if (r.str.len == 0) { ok = true; };
};
};
};
if (ok) {
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
for (i < 16) {
emitdatawbyte(0u8);
i += 1;
};
emitline("\"\n");
};
};
};
if (sz == 24 && !issg) {

8
selfhost/cmd/w6l/dynout.ww
View File

@@ -163,6 +163,14 @@ fn streqd(a: str, b: str) bool = {
// ---- main entry --------------------------------------------------------
export fn emitdynelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
// Writable globals on the dynamic-link path need their own R+W
// segment that the existing gotplt + .dynamic layout doesn't
// account for. Refuse rather than silently miscompile; folding
// .data into the existing R+W block is a follow-up.
if (l.datalen > 0u64) {
os.write(2, "w6l: top-level mutable globals (.data) not yet supported with -l/-L\n".ptr, 67u64);
return 1;
};
let a: *arena = l.a;
let n: i32 = l.dynn;
let nu: u64 = n: u64;

244
selfhost/cmd/w6l/main.combined.ww
View File

@@ -345,8 +345,10 @@ use mem;
type lsym = struct {
name: str,
val: u64, // offset within combined .text once linked
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
@@ -361,7 +363,8 @@ type lsym = struct {
};
type lrel = struct {
off: u64, // offset within combined .text
off: u64, // offset within the relocation's section
section: i32, // 0 = .text, 1 = .data
kind: i32, // R_X86_64_*
sym: *lsym,
addend: i64,
@@ -374,6 +377,8 @@ type lobj = struct {
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,
};
@@ -402,6 +407,11 @@ type lnk = struct {
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
};
@@ -564,6 +574,28 @@ fn emittext(l: *lnk, src: *u8, n: u64) void = {
l.textlen += n;
};
fn emitdata(l: *lnk, src: *u8, n: u64) void = {
if (l.datalen + n > l.datacap) {
let nc: u64 = l.datacap;
if (nc == 0u64) { nc = 256u64; };
for (nc < l.datalen + n) { nc = nc * 2u64; };
let nb: *u8 = os.alloc(nc): *u8;
let i: u64 = 0u64;
for (i < l.datalen) {
nb[i] = l.data[i];
i += 1u64;
};
l.data = nb;
l.datacap = nc;
};
let i: u64 = 0u64;
for (i < n) {
l.data[l.datalen + i] = src[i];
i += 1u64;
};
l.datalen += n;
};
// ---- C-string helpers --------------------------------------------------
fn cstrlen(p: *u8) u64 = {
@@ -654,6 +686,7 @@ fn elfglobals(a: *arena, buf: *u8, len: u64) *defent = {
let shstr: *u8 = buf + shstrshoff;
let idxtext: i32 = -1;
let idxdata: i32 = -1;
let idxsymtab: i32 = -1;
let i: u32 = 0u32;
for (i < shnum) {
@@ -663,6 +696,7 @@ fn elfglobals(a: *arena, buf: *u8, len: u64) *defent = {
let nm: *u8 = shstr + (shname: u64);
if (shtype == SHT_PROGBITS: u32) {
if (cstreq(nm, ".text")) { idxtext = i: i32; };
if (cstreq(nm, ".data")) { idxdata = i: i32; };
};
if (shtype == SHT_SYMTAB: u32) { idxsymtab = i: i32; };
i += 1u32;
@@ -688,10 +722,17 @@ fn elfglobals(a: *arena, buf: *u8, len: u64) *defent = {
let stinfo: u8 = buf[symp + SYM_INFO];
let stshndx: u16 = rdu16(buf, symp + SYM_SHNDX);
let bind: u32 = (stinfo: u32) >> 4u32;
// STB_GLOBAL = 1; defined in .text.
// STB_GLOBAL = 1; defined in .text or .data. Both are
// included so an archive member that owns a data global
// gets pulled in when something references it.
if (bind == 1u32) {
if (stshndx != 0u16) {
if ((stshndx: i32) == idxtext) {
let intext: bool = (stshndx: i32) == idxtext;
let indt: bool = false;
if (idxdata >= 0) {
indt = (stshndx: i32) == idxdata;
};
if (intext || indt) {
let nmp: *u8 = strtab + (stname: u64);
if (nmp[0u64] != 0u8) {
let nm: str = cstrtostr(a, nmp);
@@ -813,10 +854,12 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let shstrshoff: u64 = rdu64(buf, shoff + (shstrndx: u64) * SHDR_SIZE + SHDR_OFFSET);
let shstr: *u8 = buf + shstrshoff;
// find .text, .symtab, .rela.text
// find .text, .data, .symtab, .rela.text, .rela.data
let idxtext: i32 = -1;
let idxdata: i32 = -1;
let idxsymtab: i32 = -1;
let idxrela: i32 = -1;
let idxrelad: i32 = -1;
let i: u32 = 0u32;
for (i < shnum) {
let secoff: u64 = shoff + (i: u64) * SHDR_SIZE;
@@ -825,10 +868,12 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let nm: *u8 = shstr + (shname: u64);
if (shtype == SHT_PROGBITS: u32) {
if (cstreq(nm, ".text")) { idxtext = i: i32; };
if (cstreq(nm, ".data")) { idxdata = i: i32; };
};
if (shtype == SHT_SYMTAB: u32) { idxsymtab = i: i32; };
if (shtype == SHT_RELA: u32) {
if (cstreq(nm, ".rela.text")) { idxrela = i: i32; };
if (cstreq(nm, ".rela.data")) { idxrelad = i: i32; };
};
i += 1u32;
};
@@ -855,18 +900,34 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let stroff: u64 = rdu64(buf, strsh + SHDR_OFFSET);
let strtab: *u8 = buf + stroff;
let datasize: u64 = 0u64;
let dataoff: u64 = 0u64;
if (idxdata >= 0) {
let datash: u64 = shoff + (idxdata: u64) * SHDR_SIZE;
dataoff = rdu64(buf, datash + SHDR_OFFSET);
datasize = rdu64(buf, datash + SHDR_SIZE_F);
};
// Track this object.
let ob: *lobj = amalloc(l.a, 64u64): *lobj;
let ob: *lobj = amalloc(l.a, 96u64): *lobj;
ob.path = cstrtostr(l.a, path);
ob.buf = buf;
ob.len = len;
ob.textoff = l.textlen;
ob.textsize = textsize;
ob.dataoff = l.datalen;
ob.datasize = datasize;
ob.onext = l.objs;
l.objs = ob;
// Append .text bytes to the combined image.
emittext(l, buf + textoff, textsize);
// Append .data bytes (if present) to the combined .data buffer.
if (idxdata >= 0) {
if (datasize > 0u64) {
emitdata(l, buf + dataoff, datasize);
};
};
// Walk symbols. We don't keep a per-object map[] of *lsym. Instead
// the reloc loop re-walks symtab and re-interns by name. Simpler
@@ -882,7 +943,13 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let nm: str = cstrtostr(l.a, nmp);
let gs: *lsym = intern(l, nm);
if (stshndx != 0u16) {
if ((stshndx: i32) == idxtext) {
let intext: bool = (stshndx: i32) == idxtext;
let indt: bool = false;
if (idxdata >= 0) {
indt = (stshndx: i32) == idxdata;
};
if (intext) { if (indt) { indt = false; }; };
if (intext) {
if (gs.defined != 0) {
os.write(2, "w6l: duplicate symbol\n".ptr, 21u64);
l.errs += 1;
@@ -893,6 +960,18 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
gs.val = ob.textoff + stvalue;
};
};
if (indt) {
if (gs.defined != 0) {
os.write(2, "w6l: duplicate symbol\n".ptr, 21u64);
l.errs += 1;
} else {
gs.defined = 1;
gs.indata = 1;
gs.owner = ob;
gs.idxinowner = si: i32;
gs.val = ob.dataoff + stvalue;
};
};
};
};
si += 1u64;
@@ -912,8 +991,9 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let raddend: u64 = rdu64(buf, rp + RELA_ADDEND);
let rsymidx: u32 = (rinfo >> 32u64): u32;
let rkind: i32 = ((rinfo & 4294967295u64): u32): i32;
let nr: *lrel = amalloc(l.a, 48u64): *lrel;
let nr: *lrel = amalloc(l.a, 64u64): *lrel;
nr.off = ob.textoff + roff;
nr.section = 0;
nr.kind = rkind;
nr.addend = raddend: i64;
// Look up the referenced sym by name (re-walk symtab).
@@ -932,6 +1012,41 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
};
};
// Data-reloc collection. Offsets land in .data, shifted by
// this object's data_off so they index the combined buffer.
if (idxrelad >= 0) {
let relash: u64 = shoff + (idxrelad: u64) * SHDR_SIZE;
let relaoff: u64 = rdu64(buf, relash + SHDR_OFFSET);
let relasize: u64 = rdu64(buf, relash + SHDR_SIZE_F);
let nrel: u64 = relasize / RELA_SIZE;
let ri: u64 = 0u64;
for (ri < nrel) {
let rp: u64 = relaoff + ri * RELA_SIZE;
let roff: u64 = rdu64(buf, rp + RELA_OFFSET);
let rinfo: u64 = rdu64(buf, rp + RELA_INFO);
let raddend: u64 = rdu64(buf, rp + RELA_ADDEND);
let rsymidx: u32 = (rinfo >> 32u64): u32;
let rkind: i32 = ((rinfo & 4294967295u64): u32): i32;
let nr: *lrel = amalloc(l.a, 64u64): *lrel;
nr.off = ob.dataoff + roff;
nr.section = 1;
nr.kind = rkind;
nr.addend = raddend: i64;
if ((rsymidx: u64) < nsyms) {
let sp: u64 = symoff + (rsymidx: u64) * SYM_SIZE;
let sname: u32 = rdu32(buf, sp + SYM_NAME);
let snm: *u8 = strtab + (sname: u64);
if (snm[0u64] != 0u8) {
let nm: str = cstrtostr(l.a, snm);
nr.sym = intern(l, nm);
};
};
nr.rnext = l.rels;
l.rels = nr;
ri += 1u64;
};
};
return 0;
};
@@ -1361,6 +1476,7 @@ use os;
use sym;
use dyn;
def R_X86_64_64: i32 = 1;
def R_X86_64_PC32: i32 = 2;
def R_X86_64_PLT32: i32 = 4;
@@ -1427,7 +1543,15 @@ fn patchu32(p: *u8, v: u32) void = {
p[3] = ((v >> 24u32) & 255u32): u8;
};
export fn relocate(l: *lnk, base: u64) i32 = {
fn patchu64(p: *u8, v: u64) void = {
let i: i32 = 0;
for (i < 8) {
p[i] = ((v >> (i: u64 * 8u64)) & 255u64): u8;
i += 1;
};
};
export fn relocate(l: *lnk, textva: u64, datava: u64) i32 = {
let r: *lrel = l.rels;
for (r != nil) {
if (r.sym != nil) {
@@ -1438,21 +1562,30 @@ export fn relocate(l: *lnk, base: u64) i32 = {
continue;
};
if (r.sym.defined != 0) {
let symva: u64 = textva + r.sym.val;
if (r.sym.indata != 0) { symva = datava + r.sym.val; };
let k: i32 = r.kind;
if (k == R_X86_64_PC32) {
let site: u64 = base + r.off;
let target: i64 = (base + r.sym.val): i64;
let rel: i64 = (target - site: i64) + r.addend;
let site: u64 = textva + r.off;
let rel: i64 = (symva: i64 - site: i64) + r.addend;
patchu32(l.text + r.off, rel: u32);
} else { if (k == R_X86_64_PLT32) {
let site: u64 = base + r.off;
let target: i64 = (base + r.sym.val): i64;
let rel: i64 = (target - site: i64) + r.addend;
let site: u64 = textva + r.off;
let rel: i64 = (symva: i64 - site: i64) + r.addend;
patchu32(l.text + r.off, rel: u32);
} else { if (k == R_X86_64_64) {
// Absolute 64-bit. Currently used only
// for DATAR slots in .data.
let v: u64 = (symva: i64 + r.addend): u64;
if (r.section == 1) {
patchu64(l.data + r.off, v);
} else {
patchu64(l.text + r.off, v);
};
} else {
os.write(2, "w6l: unsupported reloc kind\n".ptr, 27u64);
l.errs += 1;
};};
};};};
};
};
r = r.rnext;
@@ -1626,6 +1759,14 @@ fn streqd(a: str, b: str) bool = {
// ---- main entry --------------------------------------------------------
export fn emitdynelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
// Writable globals on the dynamic-link path need their own R+W
// segment that the existing gotplt + .dynamic layout doesn't
// account for. Refuse rather than silently miscompile; folding
// .data into the existing R+W block is a follow-up.
if (l.datalen > 0u64) {
os.write(2, "w6l: top-level mutable globals (.data) not yet supported with -l/-L\n".ptr, 67u64);
return 1;
};
let a: *arena = l.a;
let n: i32 = l.dynn;
let nu: u64 = n: u64;
@@ -2212,9 +2353,11 @@ def ELFCLASS64: u8 = 2u8;
def ELFDATA2LSB: u8 = 1u8;
def PT_LOAD: u32 = 1u32;
def PF_X: u32 = 1u32;
def PF_W: u32 = 2u32;
def PF_R: u32 = 4u32;
def TEXT_OFF: u64 = 4096u64; // 0x1000
def PAGE_SZ: u64 = 4096u64;
// ---- little-endian byte writers ----------------------------------------
@@ -2246,7 +2389,14 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
};
};
let filesz: u64 = TEXT_OFF + l.textlen;
let hasdata: bool = l.datalen > 0u64;
let rxend: u64 = TEXT_OFF + l.textlen;
// .data lands at the next page boundary so the loader can give
// it fresh R+W permissions without overlapping the R+X mapping.
let dataoff: u64 = 0u64;
if (hasdata) {
dataoff = (rxend + PAGE_SZ - 1u64) & ~(PAGE_SZ - 1u64);
};
// One contiguous header buffer covering [0..0x1000), then .text.
let hdr: *u8 = os.alloc(TEXT_OFF): *u8; // zero-initialised by mmap
@@ -2268,21 +2418,34 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
wru32(hdr, 48u64, 0u32); // e_flags
wru16(hdr, 52u64, 64u16); // e_ehsize
wru16(hdr, 54u64, 56u16); // e_phentsize
wru16(hdr, 56u64, 1u16); // e_phnum
if (hasdata) { wru16(hdr, 56u64, 2u16); }
else { wru16(hdr, 56u64, 1u16); };
wru16(hdr, 58u64, 0u16); // e_shentsize
wru16(hdr, 60u64, 0u16); // e_shnum
wru16(hdr, 62u64, 0u16); // e_shstrndx
// --- Phdr (56 bytes) at offset 64 ---
wru32(hdr, 64u64, PT_LOAD); // p_type
wru32(hdr, 68u64, PF_R | PF_X); // p_flags
// --- Phdr #1 (R+X) at offset 64 ---
wru32(hdr, 64u64, PT_LOAD);
wru32(hdr, 68u64, PF_R | PF_X);
wru64(hdr, 72u64, 0u64); // p_offset
wru64(hdr, 80u64, base); // p_vaddr
wru64(hdr, 88u64, base); // p_paddr
wru64(hdr, 96u64, filesz); // p_filesz
wru64(hdr, 104u64, filesz); // p_memsz
wru64(hdr, 96u64, rxend); // p_filesz
wru64(hdr, 104u64, rxend); // p_memsz
wru64(hdr, 112u64, TEXT_OFF); // p_align
if (hasdata) {
// --- Phdr #2 (R+W) at offset 64+56=120 ---
wru32(hdr, 120u64, PT_LOAD);
wru32(hdr, 124u64, PF_R | PF_W);
wru64(hdr, 128u64, dataoff); // p_offset
wru64(hdr, 136u64, base + dataoff); // p_vaddr
wru64(hdr, 144u64, base + dataoff); // p_paddr
wru64(hdr, 152u64, l.datalen); // p_filesz
wru64(hdr, 160u64, l.datalen); // p_memsz
wru64(hdr, 168u64, PAGE_SZ); // p_align
};
// Write [0..0x1000) then .text.
let r1: (i64 | os.oserror) = os.writeall(fd, hdr, TEXT_OFF);
let n1: i64 = 0i64;
@@ -2300,6 +2463,26 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
};
if (n2 != l.textlen: i64) { return -1; };
};
if (hasdata) {
// Pad to the page-aligned data offset, then write .data.
let here: u64 = TEXT_OFF + l.textlen;
let zero: u8 = 0u8;
for (here < dataoff) {
let r3: (i64 | os.oserror) = os.writeall(fd, &zero, 1u64);
match (r3) {
case let v: i64 => { };
case let e: os.oserror => return -1;
};
here += 1u64;
};
let r4: (i64 | os.oserror) = os.writeall(fd, l.data, l.datalen);
let n4: i64 = 0i64;
match (r4) {
case let v: i64 => n4 = v;
case let e: os.oserror => return -1;
};
if (n4 != l.datalen: i64) { return -1; };
};
return 0;
};
@@ -2324,7 +2507,7 @@ def BASE: u64 = 4194304u64; // 0x400000
def CODE_VA_OFF: u64 = 4096u64; // .text starts at base + 0x1000
fn mklnk(a: *arena) *lnk = {
let l: *lnk = amalloc(a, 112u64): *lnk;
let l: *lnk = amalloc(a, 144u64): *lnk;
l.a = a;
return l;
};
@@ -2604,7 +2787,18 @@ export fn main(argc: i32, argv: **u8) i32 = {
};
if (resolve(l) != 0) { return 1; };
if (relocate(l, BASE + CODE_VA_OFF) != 0) { return 1; };
// Static layout for relocation purposes. .data lands at the
// next page after .text; the dyn path overrides if its layout
// is different (text→text PC32 cancels the absolute VA so
// only the data delta matters here).
let textva: u64 = BASE + CODE_VA_OFF;
let datava: u64 = 0u64;
if (l.datalen > 0u64) {
let page: u64 = 4096u64;
let dataoff: u64 = (CODE_VA_OFF + l.textlen + page - 1u64) & ~(page - 1u64);
datava = BASE + dataoff;
};
if (relocate(l, textva, datava) != 0) { return 1; };
let entrysym: *lsym = lookup(l, "_start");
if (entrysym == nil) { entrysym = lookup(l, "main"); }

15
selfhost/cmd/w6l/main.ww
View File

@@ -18,7 +18,7 @@ def BASE: u64 = 4194304u64; // 0x400000
def CODE_VA_OFF: u64 = 4096u64; // .text starts at base + 0x1000
fn mklnk(a: *arena) *lnk = {
let l: *lnk = amalloc(a, 112u64): *lnk;
let l: *lnk = amalloc(a, 144u64): *lnk;
l.a = a;
return l;
};
@@ -298,7 +298,18 @@ export fn main(argc: i32, argv: **u8) i32 = {
};
if (resolve(l) != 0) { return 1; };
if (relocate(l, BASE + CODE_VA_OFF) != 0) { return 1; };
// Static layout for relocation purposes. .data lands at the
// next page after .text; the dyn path overrides if its layout
// is different (text→text PC32 cancels the absolute VA so
// only the data delta matters here).
let textva: u64 = BASE + CODE_VA_OFF;
let datava: u64 = 0u64;
if (l.datalen > 0u64) {
let page: u64 = 4096u64;
let dataoff: u64 = (CODE_VA_OFF + l.textlen + page - 1u64) & ~(page - 1u64);
datava = BASE + dataoff;
};
if (relocate(l, textva, datava) != 0) { return 1; };
let entrysym: *lsym = lookup(l, "_start");
if (entrysym == nil) { entrysym = lookup(l, "main"); }

117
selfhost/cmd/w6l/obj.ww
View File

@@ -123,6 +123,28 @@ fn emittext(l: *lnk, src: *u8, n: u64) void = {
l.textlen += n;
};
fn emitdata(l: *lnk, src: *u8, n: u64) void = {
if (l.datalen + n > l.datacap) {
let nc: u64 = l.datacap;
if (nc == 0u64) { nc = 256u64; };
for (nc < l.datalen + n) { nc = nc * 2u64; };
let nb: *u8 = os.alloc(nc): *u8;
let i: u64 = 0u64;
for (i < l.datalen) {
nb[i] = l.data[i];
i += 1u64;
};
l.data = nb;
l.datacap = nc;
};
let i: u64 = 0u64;
for (i < n) {
l.data[l.datalen + i] = src[i];
i += 1u64;
};
l.datalen += n;
};
// ---- C-string helpers --------------------------------------------------
fn cstrlen(p: *u8) u64 = {
@@ -213,6 +235,7 @@ fn elfglobals(a: *arena, buf: *u8, len: u64) *defent = {
let shstr: *u8 = buf + shstrshoff;
let idxtext: i32 = -1;
let idxdata: i32 = -1;
let idxsymtab: i32 = -1;
let i: u32 = 0u32;
for (i < shnum) {
@@ -222,6 +245,7 @@ fn elfglobals(a: *arena, buf: *u8, len: u64) *defent = {
let nm: *u8 = shstr + (shname: u64);
if (shtype == SHT_PROGBITS: u32) {
if (cstreq(nm, ".text")) { idxtext = i: i32; };
if (cstreq(nm, ".data")) { idxdata = i: i32; };
};
if (shtype == SHT_SYMTAB: u32) { idxsymtab = i: i32; };
i += 1u32;
@@ -247,10 +271,17 @@ fn elfglobals(a: *arena, buf: *u8, len: u64) *defent = {
let stinfo: u8 = buf[symp + SYM_INFO];
let stshndx: u16 = rdu16(buf, symp + SYM_SHNDX);
let bind: u32 = (stinfo: u32) >> 4u32;
// STB_GLOBAL = 1; defined in .text.
// STB_GLOBAL = 1; defined in .text or .data. Both are
// included so an archive member that owns a data global
// gets pulled in when something references it.
if (bind == 1u32) {
if (stshndx != 0u16) {
if ((stshndx: i32) == idxtext) {
let intext: bool = (stshndx: i32) == idxtext;
let indt: bool = false;
if (idxdata >= 0) {
indt = (stshndx: i32) == idxdata;
};
if (intext || indt) {
let nmp: *u8 = strtab + (stname: u64);
if (nmp[0u64] != 0u8) {
let nm: str = cstrtostr(a, nmp);
@@ -372,10 +403,12 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let shstrshoff: u64 = rdu64(buf, shoff + (shstrndx: u64) * SHDR_SIZE + SHDR_OFFSET);
let shstr: *u8 = buf + shstrshoff;
// find .text, .symtab, .rela.text
// find .text, .data, .symtab, .rela.text, .rela.data
let idxtext: i32 = -1;
let idxdata: i32 = -1;
let idxsymtab: i32 = -1;
let idxrela: i32 = -1;
let idxrelad: i32 = -1;
let i: u32 = 0u32;
for (i < shnum) {
let secoff: u64 = shoff + (i: u64) * SHDR_SIZE;
@@ -384,10 +417,12 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let nm: *u8 = shstr + (shname: u64);
if (shtype == SHT_PROGBITS: u32) {
if (cstreq(nm, ".text")) { idxtext = i: i32; };
if (cstreq(nm, ".data")) { idxdata = i: i32; };
};
if (shtype == SHT_SYMTAB: u32) { idxsymtab = i: i32; };
if (shtype == SHT_RELA: u32) {
if (cstreq(nm, ".rela.text")) { idxrela = i: i32; };
if (cstreq(nm, ".rela.data")) { idxrelad = i: i32; };
};
i += 1u32;
};
@@ -414,18 +449,34 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let stroff: u64 = rdu64(buf, strsh + SHDR_OFFSET);
let strtab: *u8 = buf + stroff;
let datasize: u64 = 0u64;
let dataoff: u64 = 0u64;
if (idxdata >= 0) {
let datash: u64 = shoff + (idxdata: u64) * SHDR_SIZE;
dataoff = rdu64(buf, datash + SHDR_OFFSET);
datasize = rdu64(buf, datash + SHDR_SIZE_F);
};
// Track this object.
let ob: *lobj = amalloc(l.a, 64u64): *lobj;
let ob: *lobj = amalloc(l.a, 96u64): *lobj;
ob.path = cstrtostr(l.a, path);
ob.buf = buf;
ob.len = len;
ob.textoff = l.textlen;
ob.textsize = textsize;
ob.dataoff = l.datalen;
ob.datasize = datasize;
ob.onext = l.objs;
l.objs = ob;
// Append .text bytes to the combined image.
emittext(l, buf + textoff, textsize);
// Append .data bytes (if present) to the combined .data buffer.
if (idxdata >= 0) {
if (datasize > 0u64) {
emitdata(l, buf + dataoff, datasize);
};
};
// Walk symbols. We don't keep a per-object map[] of *lsym. Instead
// the reloc loop re-walks symtab and re-interns by name. Simpler
@@ -441,7 +492,13 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let nm: str = cstrtostr(l.a, nmp);
let gs: *lsym = intern(l, nm);
if (stshndx != 0u16) {
if ((stshndx: i32) == idxtext) {
let intext: bool = (stshndx: i32) == idxtext;
let indt: bool = false;
if (idxdata >= 0) {
indt = (stshndx: i32) == idxdata;
};
if (intext) { if (indt) { indt = false; }; };
if (intext) {
if (gs.defined != 0) {
os.write(2, "w6l: duplicate symbol\n".ptr, 21u64);
l.errs += 1;
@@ -452,6 +509,18 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
gs.val = ob.textoff + stvalue;
};
};
if (indt) {
if (gs.defined != 0) {
os.write(2, "w6l: duplicate symbol\n".ptr, 21u64);
l.errs += 1;
} else {
gs.defined = 1;
gs.indata = 1;
gs.owner = ob;
gs.idxinowner = si: i32;
gs.val = ob.dataoff + stvalue;
};
};
};
};
si += 1u64;
@@ -471,8 +540,9 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
let raddend: u64 = rdu64(buf, rp + RELA_ADDEND);
let rsymidx: u32 = (rinfo >> 32u64): u32;
let rkind: i32 = ((rinfo & 4294967295u64): u32): i32;
let nr: *lrel = amalloc(l.a, 48u64): *lrel;
let nr: *lrel = amalloc(l.a, 64u64): *lrel;
nr.off = ob.textoff + roff;
nr.section = 0;
nr.kind = rkind;
nr.addend = raddend: i64;
// Look up the referenced sym by name (re-walk symtab).
@@ -491,5 +561,40 @@ fn loadimage(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = {
};
};
// Data-reloc collection. Offsets land in .data, shifted by
// this object's data_off so they index the combined buffer.
if (idxrelad >= 0) {
let relash: u64 = shoff + (idxrelad: u64) * SHDR_SIZE;
let relaoff: u64 = rdu64(buf, relash + SHDR_OFFSET);
let relasize: u64 = rdu64(buf, relash + SHDR_SIZE_F);
let nrel: u64 = relasize / RELA_SIZE;
let ri: u64 = 0u64;
for (ri < nrel) {
let rp: u64 = relaoff + ri * RELA_SIZE;
let roff: u64 = rdu64(buf, rp + RELA_OFFSET);
let rinfo: u64 = rdu64(buf, rp + RELA_INFO);
let raddend: u64 = rdu64(buf, rp + RELA_ADDEND);
let rsymidx: u32 = (rinfo >> 32u64): u32;
let rkind: i32 = ((rinfo & 4294967295u64): u32): i32;
let nr: *lrel = amalloc(l.a, 64u64): *lrel;
nr.off = ob.dataoff + roff;
nr.section = 1;
nr.kind = rkind;
nr.addend = raddend: i64;
if ((rsymidx: u64) < nsyms) {
let sp: u64 = symoff + (rsymidx: u64) * SYM_SIZE;
let sname: u32 = rdu32(buf, sp + SYM_NAME);
let snm: *u8 = strtab + (sname: u64);
if (snm[0u64] != 0u8) {
let nm: str = cstrtostr(l.a, snm);
nr.sym = intern(l, nm);
};
};
nr.rnext = l.rels;
l.rels = nr;
ri += 1u64;
};
};
return 0;
};

56
selfhost/cmd/w6l/out.ww
View File

@@ -18,9 +18,11 @@ def ELFCLASS64: u8 = 2u8;
def ELFDATA2LSB: u8 = 1u8;
def PT_LOAD: u32 = 1u32;
def PF_X: u32 = 1u32;
def PF_W: u32 = 2u32;
def PF_R: u32 = 4u32;
def TEXT_OFF: u64 = 4096u64; // 0x1000
def PAGE_SZ: u64 = 4096u64;
// ---- little-endian byte writers ----------------------------------------
@@ -52,7 +54,14 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
};
};
let filesz: u64 = TEXT_OFF + l.textlen;
let hasdata: bool = l.datalen > 0u64;
let rxend: u64 = TEXT_OFF + l.textlen;
// .data lands at the next page boundary so the loader can give
// it fresh R+W permissions without overlapping the R+X mapping.
let dataoff: u64 = 0u64;
if (hasdata) {
dataoff = (rxend + PAGE_SZ - 1u64) & ~(PAGE_SZ - 1u64);
};
// One contiguous header buffer covering [0..0x1000), then .text.
let hdr: *u8 = os.alloc(TEXT_OFF): *u8; // zero-initialised by mmap
@@ -74,21 +83,34 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
wru32(hdr, 48u64, 0u32); // e_flags
wru16(hdr, 52u64, 64u16); // e_ehsize
wru16(hdr, 54u64, 56u16); // e_phentsize
wru16(hdr, 56u64, 1u16); // e_phnum
if (hasdata) { wru16(hdr, 56u64, 2u16); }
else { wru16(hdr, 56u64, 1u16); };
wru16(hdr, 58u64, 0u16); // e_shentsize
wru16(hdr, 60u64, 0u16); // e_shnum
wru16(hdr, 62u64, 0u16); // e_shstrndx
// --- Phdr (56 bytes) at offset 64 ---
wru32(hdr, 64u64, PT_LOAD); // p_type
wru32(hdr, 68u64, PF_R | PF_X); // p_flags
// --- Phdr #1 (R+X) at offset 64 ---
wru32(hdr, 64u64, PT_LOAD);
wru32(hdr, 68u64, PF_R | PF_X);
wru64(hdr, 72u64, 0u64); // p_offset
wru64(hdr, 80u64, base); // p_vaddr
wru64(hdr, 88u64, base); // p_paddr
wru64(hdr, 96u64, filesz); // p_filesz
wru64(hdr, 104u64, filesz); // p_memsz
wru64(hdr, 96u64, rxend); // p_filesz
wru64(hdr, 104u64, rxend); // p_memsz
wru64(hdr, 112u64, TEXT_OFF); // p_align
if (hasdata) {
// --- Phdr #2 (R+W) at offset 64+56=120 ---
wru32(hdr, 120u64, PT_LOAD);
wru32(hdr, 124u64, PF_R | PF_W);
wru64(hdr, 128u64, dataoff); // p_offset
wru64(hdr, 136u64, base + dataoff); // p_vaddr
wru64(hdr, 144u64, base + dataoff); // p_paddr
wru64(hdr, 152u64, l.datalen); // p_filesz
wru64(hdr, 160u64, l.datalen); // p_memsz
wru64(hdr, 168u64, PAGE_SZ); // p_align
};
// Write [0..0x1000) then .text.
let r1: (i64 | os.oserror) = os.writeall(fd, hdr, TEXT_OFF);
let n1: i64 = 0i64;
@@ -106,5 +128,25 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = {
};
if (n2 != l.textlen: i64) { return -1; };
};
if (hasdata) {
// Pad to the page-aligned data offset, then write .data.
let here: u64 = TEXT_OFF + l.textlen;
let zero: u8 = 0u8;
for (here < dataoff) {
let r3: (i64 | os.oserror) = os.writeall(fd, &zero, 1u64);
match (r3) {
case let v: i64 => { };
case let e: os.oserror => return -1;
};
here += 1u64;
};
let r4: (i64 | os.oserror) = os.writeall(fd, l.data, l.datalen);
let n4: i64 = 0i64;
match (r4) {
case let v: i64 => n4 = v;
case let e: os.oserror => return -1;
};
if (n4 != l.datalen: i64) { return -1; };
};
return 0;
};

34
selfhost/cmd/w6l/pass.ww
View File

@@ -14,6 +14,7 @@ use os;
use sym;
use dyn;
def R_X86_64_64: i32 = 1;
def R_X86_64_PC32: i32 = 2;
def R_X86_64_PLT32: i32 = 4;
@@ -80,7 +81,15 @@ fn patchu32(p: *u8, v: u32) void = {
p[3] = ((v >> 24u32) & 255u32): u8;
};
export fn relocate(l: *lnk, base: u64) i32 = {
fn patchu64(p: *u8, v: u64) void = {
let i: i32 = 0;
for (i < 8) {
p[i] = ((v >> (i: u64 * 8u64)) & 255u64): u8;
i += 1;
};
};
export fn relocate(l: *lnk, textva: u64, datava: u64) i32 = {
let r: *lrel = l.rels;
for (r != nil) {
if (r.sym != nil) {
@@ -91,21 +100,30 @@ export fn relocate(l: *lnk, base: u64) i32 = {
continue;
};
if (r.sym.defined != 0) {
let symva: u64 = textva + r.sym.val;
if (r.sym.indata != 0) { symva = datava + r.sym.val; };
let k: i32 = r.kind;
if (k == R_X86_64_PC32) {
let site: u64 = base + r.off;
let target: i64 = (base + r.sym.val): i64;
let rel: i64 = (target - site: i64) + r.addend;
let site: u64 = textva + r.off;
let rel: i64 = (symva: i64 - site: i64) + r.addend;
patchu32(l.text + r.off, rel: u32);
} else { if (k == R_X86_64_PLT32) {
let site: u64 = base + r.off;
let target: i64 = (base + r.sym.val): i64;
let rel: i64 = (target - site: i64) + r.addend;
let site: u64 = textva + r.off;
let rel: i64 = (symva: i64 - site: i64) + r.addend;
patchu32(l.text + r.off, rel: u32);
} else { if (k == R_X86_64_64) {
// Absolute 64-bit. Currently used only
// for DATAR slots in .data.
let v: u64 = (symva: i64 + r.addend): u64;
if (r.section == 1) {
patchu64(l.data + r.off, v);
} else {
patchu64(l.text + r.off, v);
};
} else {
os.write(2, "w6l: unsupported reloc kind\n".ptr, 27u64);
l.errs += 1;
};};
};};};
};
};
r = r.rnext;

14
selfhost/cmd/w6l/sym.ww
View File

@@ -7,8 +7,10 @@ use mem;
type lsym = struct {
name: str,
val: u64, // offset within combined .text once linked
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
@@ -23,7 +25,8 @@ type lsym = struct {
};
type lrel = struct {
off: u64, // offset within combined .text
off: u64, // offset within the relocation's section
section: i32, // 0 = .text, 1 = .data
kind: i32, // R_X86_64_*
sym: *lsym,
addend: i64,
@@ -36,6 +39,8 @@ type lobj = struct {
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,
};
@@ -64,6 +69,11 @@ type lnk = struct {
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
};

110
selfhost/cmd/wcc/cgen.ww
View File

@@ -761,13 +761,45 @@ fn emitdatawbyte(b: u8) void = {
os.write(1, bb.ptr, 1u64);
};
// letpreintern — intern strlits referenced from top-level str-let
// initialisers BEFORE emitdatasection runs. Mirrors cmd/w6c/cgen.c
// let_pre_intern: emitletdataw later looks up the same label, and
// emitdatasection emits the DATA row in the same .s file. Running
// emitletdataw after emitdatasection would flip the (DATA strlits,
// DATAW lets) section order and break byte-identity.
export fn letpreintern(c: *cgen, file: *node) void = {
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let sz: i32 = letemitsize(c, d);
if (sz == 16) {
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
if (r != nil) {
if (r.kind == nkind.N_STRLIT) {
if (r.str.len > 0) {
internstrlit(c, r.str);
};
};
};
};
};
d = d.next;
};
};
// emitletdataw — DATAW directive per top-level `let` global.
// 8B scalar with int/rune/bool/nil literal init (or no init).
// 16B str with no init (or `nil` / `""`) — zero header; the
// program must assign a real strlit at runtime before
// using .ptr / .len.
// Non-literal scalar inits and non-empty strlit inits are skipped
// so the link surfaces an undefined-symbol error.
// 16B str — no init / `nil` / `""` → 16 zero bytes; or non-empty
// strlit init → 8 zero placeholder + 8 LE len bytes plus a
// DATAR slot+0,strlit reloc that the linker patches at load.
// sz struct — zero only.
// Non-literal scalar inits and unsupported shapes are skipped so the
// link surfaces an undefined-symbol error if the binding is used.
fn emitletdataw(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
@@ -810,31 +842,67 @@ fn emitletdataw(c: *cgen, file: *node) void = {
};
};
if (sz == 16 && !issg) {
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_NIL) { ok = true; };
if (r.kind == nkind.N_STRLIT) {
if (r.str.len == 0) { ok = true; };
};
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
// str-literal init (non-empty): emit
// the 16B payload as 8 placeholder zero
// bytes + 8 LE bytes of length, then a
// DATAR reloc to patch the ptr half with
// the strlit's runtime VA.
let strlitinit: bool = false;
if (r != nil) {
if (r.kind == nkind.N_STRLIT) {
if (r.str.len > 0) { strlitinit = true; };
};
};
if (ok) {
if (strlitinit) {
let lab: str = internstrlit(c, r.str);
let v: u64 = r.str.len: u64;
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
for (i < 16) {
emitdatawbyte(0u8);
for (i < 8) { emitdatawbyte(0u8); i += 1; };
i = 0;
let nv: u64 = v;
for (i < 8) {
emitdatawbyte((nv & 255u64): u8);
nv = nv >> 8u64;
i += 1;
};
emitline("\"\n");
emitline("DATAR ");
emitsymname(c, nm);
emitline("+0(SB),");
os.write(1, lab.ptr, lab.len: u64);
emitline("(SB)\n");
} else {
// zero-init: accept no rhs, nil,
// or empty strlit.
let ok: bool = true;
if (d.rhs != nil) {
ok = false;
if (r != nil) {
if (r.kind == nkind.N_NIL) { ok = true; };
if (r.kind == nkind.N_STRLIT) {
if (r.str.len == 0) { ok = true; };
};
};
};
if (ok) {
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
for (i < 16) {
emitdatawbyte(0u8);
i += 1;
};
emitline("\"\n");
};
};
};
if (sz == 24 && !issg) {

1
selfhost/cmd/wcc/cgendecl.ww
View File

@@ -297,6 +297,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
};
d = d.next;
};
letpreintern(c, file);
emitdatasection(c);
emitdefconstants(c, file);
emitletdataw(c, file);

111
selfhost/cmd/wwdump/main.combined.ww
View File

@@ -8576,6 +8576,7 @@ export fn cgfile(c: *cgen, file: *node) void = {
};
d = d.next;
};
letpreintern(c, file);
emitdatasection(c);
emitdefconstants(c, file);
emitletdataw(c, file);
@@ -9345,13 +9346,45 @@ fn emitdatawbyte(b: u8) void = {
os.write(1, bb.ptr, 1u64);
};
// letpreintern — intern strlits referenced from top-level str-let
// initialisers BEFORE emitdatasection runs. Mirrors cmd/w6c/cgen.c
// let_pre_intern: emitletdataw later looks up the same label, and
// emitdatasection emits the DATA row in the same .s file. Running
// emitletdataw after emitdatasection would flip the (DATA strlits,
// DATAW lets) section order and break byte-identity.
export fn letpreintern(c: *cgen, file: *node) void = {
if (file == nil) { return; };
let d: *node = file.list;
for (d != nil) {
if (d.kind == nkind.N_LET) {
let sz: i32 = letemitsize(c, d);
if (sz == 16) {
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
if (r != nil) {
if (r.kind == nkind.N_STRLIT) {
if (r.str.len > 0) {
internstrlit(c, r.str);
};
};
};
};
};
d = d.next;
};
};
// emitletdataw — DATAW directive per top-level `let` global.
// 8B scalar with int/rune/bool/nil literal init (or no init).
// 16B str with no init (or `nil` / `""`) — zero header; the
// program must assign a real strlit at runtime before
// using .ptr / .len.
// Non-literal scalar inits and non-empty strlit inits are skipped
// so the link surfaces an undefined-symbol error.
// 16B str — no init / `nil` / `""` → 16 zero bytes; or non-empty
// strlit init → 8 zero placeholder + 8 LE len bytes plus a
// DATAR slot+0,strlit reloc that the linker patches at load.
// sz struct — zero only.
// Non-literal scalar inits and unsupported shapes are skipped so the
// link surfaces an undefined-symbol error if the binding is used.
fn emitletdataw(c: *cgen, file: *node) void = {
let d: *node = file.list;
for (d != nil) {
@@ -9394,31 +9427,67 @@ fn emitletdataw(c: *cgen, file: *node) void = {
};
};
if (sz == 16 && !issg) {
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_NIL) { ok = true; };
if (r.kind == nkind.N_STRLIT) {
if (r.str.len == 0) { ok = true; };
};
let r: *node = d.rhs;
for (r != nil) {
if (r.kind != nkind.N_CAST) { break; };
r = r.lhs;
};
// str-literal init (non-empty): emit
// the 16B payload as 8 placeholder zero
// bytes + 8 LE bytes of length, then a
// DATAR reloc to patch the ptr half with
// the strlit's runtime VA.
let strlitinit: bool = false;
if (r != nil) {
if (r.kind == nkind.N_STRLIT) {
if (r.str.len > 0) { strlitinit = true; };
};
};
if (ok) {
if (strlitinit) {
let lab: str = internstrlit(c, r.str);
let v: u64 = r.str.len: u64;
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
for (i < 16) {
emitdatawbyte(0u8);
for (i < 8) { emitdatawbyte(0u8); i += 1; };
i = 0;
let nv: u64 = v;
for (i < 8) {
emitdatawbyte((nv & 255u64): u8);
nv = nv >> 8u64;
i += 1;
};
emitline("\"\n");
emitline("DATAR ");
emitsymname(c, nm);
emitline("+0(SB),");
os.write(1, lab.ptr, lab.len: u64);
emitline("(SB)\n");
} else {
// zero-init: accept no rhs, nil,
// or empty strlit.
let ok: bool = true;
if (d.rhs != nil) {
ok = false;
if (r != nil) {
if (r.kind == nkind.N_NIL) { ok = true; };
if (r.kind == nkind.N_STRLIT) {
if (r.str.len == 0) { ok = true; };
};
};
};
if (ok) {
emitline("DATAW ");
emitsymname(c, nm);
emitline("(SB),\"");
let i: i32 = 0;
for (i < 16) {
emitdatawbyte(0u8);
i += 1;
};
emitline("\"\n");
};
};
};
if (sz == 24 && !issg) {