commit da4ace5a31d7aba4dc1236405908b350ca25897e
parent ed9a6cddccc301aecb643dfdfd2f96a782dd63d9
Author: Ryan Sepassi <rsepassi@gmail.com>
Date: Tue, 19 May 2026 18:22:53 -0700
Preserve ordered link inputs
Diffstat:
12 files changed, 364 insertions(+), 200 deletions(-)
diff --git a/driver/ld.c b/driver/ld.c
@@ -90,6 +90,8 @@ typedef struct LdOptions {
* time → DT_NEEDED entries. */
LdDso* dsos;
uint32_t ndsos;
+ CfreeLinkInputOrder* order;
+ uint32_t norder;
const char** lib_dirs; /* -L */
uint32_t nlib_dirs;
@@ -206,10 +208,11 @@ static int ld_alloc_arrays(LdOptions* o, int argc) {
driver_alloc_zeroed(o->env, bound * sizeof(*o->object_files));
o->archives = driver_alloc_zeroed(o->env, bound * sizeof(*o->archives));
o->dsos = driver_alloc_zeroed(o->env, bound * sizeof(*o->dsos));
+ o->order = driver_alloc_zeroed(o->env, bound * sizeof(*o->order));
o->lib_dirs = driver_alloc_zeroed(o->env, bound * sizeof(*o->lib_dirs));
o->rpaths = driver_alloc_zeroed(o->env, bound * sizeof(*o->rpaths));
o->rpath_links = driver_alloc_zeroed(o->env, bound * sizeof(*o->rpath_links));
- if (!o->object_files || !o->archives || !o->dsos || !o->lib_dirs ||
+ if (!o->object_files || !o->archives || !o->dsos || !o->order || !o->lib_dirs ||
!o->rpaths || !o->rpath_links) {
driver_errf(LD_TOOL, "out of memory");
return 1;
@@ -220,6 +223,12 @@ static int ld_alloc_arrays(LdOptions* o, int argc) {
/* ---------- positional archive bookkeeping ---------- */
+static void ld_push_order(LdOptions* o, uint8_t kind, uint32_t index) {
+ CfreeLinkInputOrder* slot = &o->order[o->norder++];
+ slot->kind = kind;
+ slot->index = index;
+}
+
static void ld_push_archive(LdOptions* o, const char* path, int owned,
size_t owned_size) {
LdArchive* a = &o->archives[o->narchives++];
@@ -229,6 +238,7 @@ static void ld_push_archive(LdOptions* o, const char* path, int owned,
a->whole_archive = o->cur_whole_archive;
a->link_mode = o->cur_link_mode;
a->group_id = o->cur_group_id;
+ ld_push_order(o, CFREE_LINK_INPUT_ARCHIVE, o->narchives - 1u);
}
static void ld_push_dso(LdOptions* o, const char* path, int owned,
@@ -237,6 +247,7 @@ static void ld_push_dso(LdOptions* o, const char* path, int owned,
d->path = path;
d->owned = owned;
d->owned_size = owned_size;
+ ld_push_order(o, CFREE_LINK_INPUT_DSO, o->ndsos - 1u);
}
/* Filename ends in `.so` (with no further extension) or in `.so.N`
@@ -580,6 +591,7 @@ static int ld_parse(int argc, char** argv, LdOptions* o) {
ld_push_dso(o, a, 0, 0);
} else {
o->object_files[o->nobject_files++] = a;
+ ld_push_order(o, CFREE_LINK_INPUT_OBJ_BYTES, o->nobject_files - 1u);
}
}
@@ -622,6 +634,7 @@ static void ld_options_release(LdOptions* o) {
driver_free(o->env, o->object_files, bound * sizeof(*o->object_files));
driver_free(o->env, o->archives, bound * sizeof(*o->archives));
driver_free(o->env, o->dsos, bound * sizeof(*o->dsos));
+ driver_free(o->env, o->order, bound * sizeof(*o->order));
driver_free(o->env, o->lib_dirs, bound * sizeof(*o->lib_dirs));
driver_free(o->env, o->rpaths, bound * sizeof(*o->rpaths));
driver_free(o->env, o->rpath_links, bound * sizeof(*o->rpath_links));
@@ -801,6 +814,8 @@ static int ld_run_link(LdOptions* o) {
inputs.narchives = o->narchives;
inputs.dso_bytes = dso_in;
inputs.ndso_bytes = o->ndsos;
+ inputs.order = o->order;
+ inputs.norder = o->norder;
inputs.linker_script = script;
inputs.entry = o->entry;
inputs.build_id_mode = o->build_id_mode;
diff --git a/include/cfree/link.h b/include/cfree/link.h
@@ -131,6 +131,19 @@ typedef struct CfreeLinkArchiveInput {
uint8_t pad;
} CfreeLinkArchiveInput;
+typedef enum CfreeLinkInputOrderKind {
+ CFREE_LINK_INPUT_OBJ,
+ CFREE_LINK_INPUT_OBJ_BYTES,
+ CFREE_LINK_INPUT_ARCHIVE,
+ CFREE_LINK_INPUT_DSO,
+} CfreeLinkInputOrderKind;
+
+typedef struct CfreeLinkInputOrder {
+ uint8_t kind; /* CfreeLinkInputOrderKind */
+ uint8_t pad[3];
+ uint32_t index; /* index into the matching CfreeLinkInputs array */
+} CfreeLinkInputOrder;
+
typedef struct CfreeLinkInputs {
CfreeObjBuilder *const *objs;
uint32_t nobjs;
@@ -145,6 +158,8 @@ typedef struct CfreeLinkInputs {
uint8_t build_id_mode; /* CfreeBuildIdMode */
const uint8_t *build_id_bytes;
uint32_t build_id_len;
+ const CfreeLinkInputOrder *order; /* optional; NULL preserves legacy order */
+ uint32_t norder;
} CfreeLinkInputs;
typedef struct CfreeExeLinkOptions {
diff --git a/src/api/link.c b/src/api/link.c
@@ -23,8 +23,63 @@
CfreeJit* cfree_jit_from_image(LinkImage*);
+static SrcLoc no_loc(void) {
+ SrcLoc l = {0, 0, 0};
+ return l;
+}
+
static void load_inputs(Linker* l, const CfreeLinkInputs* in) {
uint32_t i;
+ if (in->order && in->norder) {
+ for (i = 0; i < in->norder; ++i) {
+ const CfreeLinkInputOrder* ord = &in->order[i];
+ switch ((CfreeLinkInputOrderKind)ord->kind) {
+ case CFREE_LINK_INPUT_OBJ:
+ if (ord->index >= in->nobjs)
+ compiler_panic(l->c, no_loc(),
+ "link: ordered obj input index out of range");
+ if (in->objs[ord->index]) link_add_obj(l, in->objs[ord->index]);
+ break;
+ case CFREE_LINK_INPUT_OBJ_BYTES: {
+ const CfreeBytes* b;
+ if (ord->index >= in->nobj_bytes)
+ compiler_panic(l->c, no_loc(),
+ "link: ordered object input index out of range");
+ b = &in->obj_bytes[ord->index];
+ link_add_obj_bytes(l, b->name, b->data, b->len);
+ break;
+ }
+ case CFREE_LINK_INPUT_ARCHIVE: {
+ const CfreeLinkArchiveInput* a;
+ if (ord->index >= in->narchives)
+ compiler_panic(l->c, no_loc(),
+ "link: ordered archive input index out of range");
+ a = &in->archives[ord->index];
+ link_add_archive_bytes(l, a->bytes.name, a->bytes.data,
+ a->bytes.len, a->whole_archive,
+ a->link_mode, a->group_id);
+ break;
+ }
+ case CFREE_LINK_INPUT_DSO: {
+ const CfreeBytes* b;
+ if (ord->index >= in->ndso_bytes)
+ compiler_panic(l->c, no_loc(),
+ "link: ordered DSO input index out of range");
+ b = &in->dso_bytes[ord->index];
+ link_add_dso_bytes(l, b->name, b->data, b->len);
+ break;
+ }
+ default:
+ compiler_panic(l->c, no_loc(), "link: invalid ordered input kind");
+ }
+ }
+ if (in->linker_script) link_set_script(l, in->linker_script);
+ if (in->entry) link_set_entry(l, in->entry);
+ (void)in->build_id_mode;
+ (void)in->build_id_bytes;
+ (void)in->build_id_len;
+ return;
+ }
for (i = 0; i < in->nobjs; ++i) {
if (in->objs[i]) link_add_obj(l, in->objs[i]);
}
diff --git a/src/link/link.c b/src/link/link.c
@@ -113,6 +113,7 @@ LinkInputId link_add_obj(Linker* l, ObjBuilder* ob) {
if (!l || !ob) return LINK_INPUT_NONE;
in = inputs_push(l, &id);
in->kind = LINK_INPUT_OBJ;
+ in->order = l->next_input_order++;
in->obj = ob;
return id;
}
@@ -150,6 +151,7 @@ LinkInputId link_add_obj_bytes(Linker* l, const char* name, const u8* data,
reader_name, name ? name : "(unnamed)");
in = inputs_push(l, &id);
in->kind = LINK_INPUT_OBJ_BYTES;
+ in->order = l->next_input_order++;
in->obj = ob; /* re-uses the ObjBuilder slot for ownership */
in->name = name ? pool_intern_cstr(l->c->global, name) : 0;
return id;
@@ -197,6 +199,7 @@ LinkInputId link_add_dso_bytes(Linker* l, const char* name, const u8* data,
reader_name, name ? name : "(unnamed)");
in = inputs_push(l, &id);
in->kind = LINK_INPUT_DSO_BYTES;
+ in->order = l->next_input_order++;
in->obj = ob;
in->name = name ? pool_intern_cstr(l->c->global, name) : 0;
/* DT_SONAME wins; fall back to the file's basename if the DSO has
@@ -247,6 +250,7 @@ LinkInputId link_add_archive_bytes(Linker* l, const char* name, const u8* data,
if (!ar)
compiler_panic(l->c, no_loc(), "link: out of memory growing archives");
ar->name = name ? pool_intern_cstr(l->c->global, name) : 0;
+ ar->order = l->next_input_order++;
ar->whole_archive = whole_archive;
ar->link_mode = link_mode;
ar->group_id = group_id;
diff --git a/src/link/link.h b/src/link/link.h
@@ -37,6 +37,7 @@ typedef struct LinkInput {
LinkInputId id;
u8 kind; /* LinkInputKind */
u8 pad[3];
+ u32 order;
ObjBuilder* obj; /* for LINK_INPUT_OBJ, otherwise NULL until read */
Sym name; /* diagnostic name for bytes inputs */
/* DSO-only: SONAME extracted from PT_DYNAMIC.DT_SONAME. 0 if absent.
diff --git a/src/link/link_internal.h b/src/link/link_internal.h
@@ -70,6 +70,7 @@ typedef struct LinkArchive {
Sym name;
LinkArchiveMember* members;
u32 nmembers;
+ u32 order;
u8 whole_archive;
u8 link_mode;
u8 group_id;
@@ -84,6 +85,7 @@ struct Linker {
Heap* heap;
LinkInputs inputs; /* LinkInputId = slot index + 1 */
LinkArchives archives;
+ u32 next_input_order;
Sym entry_name;
/* Set by link_set_script. NULL: layout takes the existing default
* bucket-based path. Non-NULL: layout_sections_scripted walks the
diff --git a/src/link/link_macho.c b/src/link/link_macho.c
@@ -155,10 +155,10 @@ static u32 mbuf_str(MByte* b, const char* s, u32 n) {
typedef struct MachImp {
LinkSymId sym;
Sym name;
- u32 dylib_ord; /* 1-based ordinal into LC_LOAD_DYLIB list */
- u32 stub_idx; /* 1-based index into __stubs (0 if data import) */
- u32 got_idx; /* 1-based index into __got */
- u32 imports_strx; /* offset into chained-fixups symbol pool */
+ u32 dylib_ord; /* 1-based ordinal into LC_LOAD_DYLIB list */
+ u32 stub_idx; /* 1-based index into __stubs (0 if data import) */
+ u32 got_idx; /* 1-based index into __got */
+ u32 imports_strx; /* offset into chained-fixups symbol pool */
u8 is_func;
u8 weak;
/* internal=1 means this entry is an in-image symbol that's referenced
@@ -169,7 +169,8 @@ typedef struct MachImp {
* resolving to NULL). No dylib_ord / stub_idx / chained-fixup bind. */
u8 internal;
u8 pad[1];
- u64 internal_vaddr; /* image-relative target vaddr; meaningful only when internal=1 */
+ u64 internal_vaddr; /* image-relative target vaddr; meaningful only when
+ internal=1 */
} MachImp;
typedef struct MachDylib {
@@ -183,11 +184,13 @@ typedef struct MachDylib {
* ones). The descriptor itself is laid out in __DATA,__thread_vars by
* either the input objects (internal) or the providing dylib (imported). */
typedef struct MachTlv {
- LinkSymId sym; /* canonical descriptor LinkSymId */
- u32 tlv_idx; /* 1-based slot index in __thread_ptrs */
- u8 imported; /* 1 == descriptor lives in a dylib (BIND), 0 == internal (REBASE) */
+ LinkSymId sym; /* canonical descriptor LinkSymId */
+ u32 tlv_idx; /* 1-based slot index in __thread_ptrs */
+ u8 imported; /* 1 == descriptor lives in a dylib (BIND), 0 == internal
+ (REBASE) */
u8 pad[3];
- u32 import_idx; /* 1-based MachImp index when imported (for chained-bind ordinal) */
+ u32 import_idx; /* 1-based MachImp index when imported (for chained-bind
+ ordinal) */
} MachTlv;
/* ---- planned section ---- */
@@ -211,10 +214,10 @@ typedef struct MSec {
u64 file_offset;
u64 size;
u32 align;
- u32 flags; /* S_TYPE | S_ATTR_* */
+ u32 flags; /* S_TYPE | S_ATTR_* */
u32 reserved1;
u32 reserved2;
- u8 segidx; /* 1=__TEXT, 2=__DATA_CONST, 3=__DATA */
+ u8 segidx; /* 1=__TEXT, 2=__DATA_CONST, 3=__DATA */
u8 is_zerofill;
u8 pad[6];
} MSec;
@@ -232,10 +235,10 @@ typedef struct MSeg {
u64 vmsize;
u64 fileoff;
u64 filesize;
- u32 nsects; /* MSec count in segment — internal layout */
- u32 first_sec; /* first index into MSec[] */
- u32 nouts; /* OutSec count in segment — what hits the file */
- u32 first_out; /* first index into OutSec[] */
+ u32 nsects; /* MSec count in segment — internal layout */
+ u32 first_sec; /* first index into MSec[] */
+ u32 nouts; /* OutSec count in segment — what hits the file */
+ u32 first_out; /* first index into OutSec[] */
} MSeg;
/* On-disk section view: one record per (segname, sectname) within a
@@ -270,12 +273,13 @@ typedef struct MCtx {
/* imports */
MachImp* imports;
u32 nimports;
- u32 nimports_real; /* count of imports with internal=0 (== prefix length;
- * collect_imports appends internal=1 entries last) */
+ u32 nimports_real; /* count of imports with internal=0 (== prefix length;
+ * collect_imports appends internal=1 entries last) */
u32 nimport_funcs;
MachDylib* dylibs;
u32 ndylibs;
- /* sym->import index, 1-based, 0 = not an import. Sized to LinkSymId space + 1. */
+ /* sym->import index, 1-based, 0 = not an import. Sized to LinkSymId space
+ * + 1. */
u32* sym_to_imp;
u32 sym_to_imp_size;
@@ -284,7 +288,7 @@ typedef struct MCtx {
u32 nsecs;
OutSec* outs;
u32 nouts;
- MSeg segs[5]; /* PAGEZERO, TEXT, DATA_CONST, DATA, LINKEDIT */
+ MSeg segs[5]; /* PAGEZERO, TEXT, DATA_CONST, DATA, LINKEDIT */
u32 nsegs;
/* Synthetic byte buffers, owned. */
@@ -324,16 +328,16 @@ typedef struct MCtx {
u64 data_memsz;
u64 linkedit_vaddr;
u64 linkedit_fileoff;
- u32 entry_offset; /* offset of entry within __TEXT segment */
+ u32 entry_offset; /* offset of entry within __TEXT segment */
- u64 headers_size; /* header + loadcmds */
+ u64 headers_size; /* header + loadcmds */
/* LINKEDIT contents */
MByte chained_fixups;
MByte exports_trie;
- MByte symtab; /* binary nlist_64 array */
+ MByte symtab; /* binary nlist_64 array */
MByte strtab;
- MByte indirect; /* u32 array */
+ MByte indirect; /* u32 array */
MByte fn_starts;
MByte data_in_code;
MByte codesig;
@@ -424,8 +428,10 @@ static void collect_imports(MCtx* x) {
LinkInputId dso_id = s ? s->dso_input_id : LINK_INPUT_NONE;
Sym install = 0;
if (dso_id != LINK_INPUT_NONE && x->linker &&
- dso_id - 1u < LinkInputs_count(&x->linker->inputs))
- install = LinkInputs_at(&x->linker->inputs, dso_id - 1u)->soname;
+ dso_id - 1u < LinkInputs_count(&x->linker->inputs)) {
+ LinkInput* in = LinkInputs_at(&x->linker->inputs, dso_id - 1u);
+ if (in->kind == LINK_INPUT_DSO_BYTES) install = in->soname;
+ }
if (install == 0)
install = pool_intern_cstr(x->c->global, "/usr/lib/libSystem.B.dylib");
u32 ord = dylib_ordinal_of(x, install);
@@ -609,6 +615,22 @@ static int sec_is_zerofill(const LinkSection* ls) {
return ls->sem == SSEM_NOBITS;
}
+static int section_has_abs64_reloc(const LinkImage* img, LinkSectionId id) {
+ for (u32 i = 0; i < LinkRelocs_count(&img->relocs); ++i) {
+ const LinkRelocApply* r = LinkRelocs_at(&img->relocs, i);
+ if (r->link_section_id == id && r->kind == R_ABS64) return 1;
+ }
+ return 0;
+}
+
+static int sec_needs_data_const(const LinkImage* img, const LinkSection* ls) {
+ if (!ls || !ls->size || sec_is_exec(ls) || sec_is_writable(ls) ||
+ sec_is_zerofill(ls)) {
+ return 0;
+ }
+ return section_has_abs64_reloc(img, ls->id);
+}
+
/* Pick (segname, sectname) for a LinkSection. Comma-form Mach-O names
* round-trip into MSec's inline 16-byte buffers; literal defaults point
* at .rodata strings. Caller passes the MSec for per-section storage —
@@ -660,12 +682,12 @@ static void plan_layout(MCtx* x) {
x->segs[0].first_sec = 0;
/* Segments 1..4 */
- seg_init(&x->segs[1], "__TEXT",
- VM_PROT_READ | VM_PROT_EXECUTE, VM_PROT_READ | VM_PROT_EXECUTE);
- seg_init(&x->segs[2], "__DATA_CONST",
- VM_PROT_READ | VM_PROT_WRITE, VM_PROT_READ | VM_PROT_WRITE);
- seg_init(&x->segs[3], "__DATA",
- VM_PROT_READ | VM_PROT_WRITE, VM_PROT_READ | VM_PROT_WRITE);
+ seg_init(&x->segs[1], "__TEXT", VM_PROT_READ | VM_PROT_EXECUTE,
+ VM_PROT_READ | VM_PROT_EXECUTE);
+ seg_init(&x->segs[2], "__DATA_CONST", VM_PROT_READ | VM_PROT_WRITE,
+ VM_PROT_READ | VM_PROT_WRITE);
+ seg_init(&x->segs[3], "__DATA", VM_PROT_READ | VM_PROT_WRITE,
+ VM_PROT_READ | VM_PROT_WRITE);
seg_init(&x->segs[4], "__LINKEDIT", VM_PROT_READ, VM_PROT_READ);
x->nsegs = 5;
@@ -703,7 +725,8 @@ static void plan_layout(MCtx* x) {
LinkSection* ls = &img->sections[i];
if (!ls->size) continue;
if (sec_is_writable(ls)) continue;
- if (sec_is_zerofill(ls)) continue; /* placed in __DATA */
+ if (sec_is_zerofill(ls)) continue; /* placed in __DATA */
+ if (sec_needs_data_const(img, ls)) continue;
MSec* m = &x->secs[x->nsecs++];
memset(m, 0, sizeof(*m));
m->link_sec_id = ls->id;
@@ -713,10 +736,9 @@ static void plan_layout(MCtx* x) {
m->align = ls->align ? ls->align : 1u;
m->size = ls->size;
m->segidx = 1;
- m->flags = sec_is_exec(ls)
- ? (0x80000000u /*S_ATTR_PURE_INSTRUCTIONS*/ |
- 0x00000400u /*S_ATTR_SOME_INSTRUCTIONS*/)
- : 0u;
+ m->flags = sec_is_exec(ls) ? (0x80000000u /*S_ATTR_PURE_INSTRUCTIONS*/ |
+ 0x00000400u /*S_ATTR_SOME_INSTRUCTIONS*/)
+ : 0u;
}
/* __stubs synthetic */
@@ -736,7 +758,7 @@ static void plan_layout(MCtx* x) {
m->size = x->stubs_size;
m->segidx = 1;
m->flags = 0x80000000u | 0x00000400u | 0x00000008u /*S_SYMBOL_STUBS*/;
- m->reserved1 = 0; /* fill in later: indirect-symtab base */
+ m->reserved1 = 0; /* fill in later: indirect-symtab base */
m->reserved2 = x->arch->macho_stub_size;
}
x->segs[1].nsects = x->nsecs - first_text_sec;
@@ -747,8 +769,7 @@ static void plan_layout(MCtx* x) {
if (x->nimports) {
x->got_size = x->nimports * MZ_GOT_SIZE;
x->got_bytes = (u8*)h->alloc(h, x->got_size, 8);
- if (!x->got_bytes)
- compiler_panic(x->c, no_loc(), "link_macho: oom on got");
+ if (!x->got_bytes) compiler_panic(x->c, no_loc(), "link_macho: oom on got");
memset(x->got_bytes, 0, x->got_size);
MSec* m = &x->secs[x->nsecs++];
memset(m, 0, sizeof(*m));
@@ -760,7 +781,20 @@ static void plan_layout(MCtx* x) {
m->size = x->got_size;
m->segidx = 2;
m->flags = 0x00000006u /*S_NON_LAZY_SYMBOL_POINTERS*/;
- m->reserved1 = 0; /* indirect-symtab base */
+ m->reserved1 = 0; /* indirect-symtab base */
+ }
+ for (u32 i = 0; i < img->nsections; ++i) {
+ LinkSection* ls = &img->sections[i];
+ if (!sec_needs_data_const(img, ls)) continue;
+ MSec* m = &x->secs[x->nsecs++];
+ memset(m, 0, sizeof(*m));
+ m->link_sec_id = ls->id;
+ pick_macho_names(ls, x->c, m);
+ m->segname = "__DATA_CONST";
+ m->align = ls->align ? ls->align : 1u;
+ m->size = ls->size;
+ m->segidx = 2;
+ m->flags = 0;
}
x->segs[2].nsects = x->nsecs - first_dc;
x->segs[2].first_sec = first_dc;
@@ -885,7 +919,7 @@ static void plan_layout(MCtx* x) {
if (!seen) ++cnt;
}
sg->nouts = cnt;
- sg->first_out = 0; /* assigned in Phase B */
+ sg->first_out = 0; /* assigned in Phase B */
}
/* Compute load-command count + sizeofcmds, then back-fill section
@@ -896,11 +930,11 @@ static void plan_layout(MCtx* x) {
if (i == 0) {
++nseg_real;
continue;
- } /* PAGEZERO */
+ } /* PAGEZERO */
if (i == 4) {
++nseg_real;
continue;
- } /* LINKEDIT always */
+ } /* LINKEDIT always */
if (x->segs[i].nsects > 0) ++nseg_real;
}
/* Each LC_SEGMENT_64 carries 72 + 80*nouts bytes (one section_64
@@ -908,7 +942,7 @@ static void plan_layout(MCtx* x) {
u32 sizeofcmds = 0;
for (u32 i = 0; i < x->nsegs; ++i) {
if (i == 0 || i == 4) {
- sizeofcmds += MACHO_SEGCMD64_SIZE; /* no sections */
+ sizeofcmds += MACHO_SEGCMD64_SIZE; /* no sections */
continue;
}
if (x->segs[i].nsects == 0) continue;
@@ -982,10 +1016,10 @@ static void plan_layout(MCtx* x) {
cur_v += m->size;
seen_zerofill = 1;
}
- sg->filesize = (i == 1) ? (cur_f - seg_start_f)
- : (first_zerofill_v
- ? (first_zerofill_v - seg_start_v)
- : (cur_v - seg_start_v));
+ sg->filesize = (i == 1)
+ ? (cur_f - seg_start_f)
+ : (first_zerofill_v ? (first_zerofill_v - seg_start_v)
+ : (cur_v - seg_start_v));
sg->vmsize = ALIGN_UP(cur_v - seg_start_v, MZ_PAGE);
if (sg->vmsize == 0 && sg->nsects > 0) sg->vmsize = MZ_PAGE;
if (i == 1) {
@@ -1025,7 +1059,14 @@ static void plan_layout(MCtx* x) {
x->data_memsz = sg->vmsize;
}
vaddr = sg->vmaddr + sg->vmsize;
- fileoff = sg->fileoff + sg->filesize;
+ /* Mach-O segments are mapped in page units. If a segment's memory
+ * image extends past its initialized file bytes (for example
+ * __DATA,__bss), the following segment's fileoff must not reuse those
+ * pages or the kernel can map later file contents into the zero-fill
+ * tail. */
+ fileoff = sg->fileoff + ((sg->vmsize > ALIGN_UP(sg->filesize, MZ_PAGE))
+ ? sg->vmsize
+ : sg->filesize);
(void)seen_zerofill;
}
/* LINKEDIT placeholder; size is filled after blob assembly. */
@@ -1076,8 +1117,7 @@ static void plan_layout(MCtx* x) {
}
u32 cap = x->nsecs + 1u;
x->outs = (OutSec*)h->alloc(h, sizeof(OutSec) * cap, _Alignof(OutSec));
- if (!x->outs)
- compiler_panic(x->c, no_loc(), "link_macho: oom on OutSec");
+ if (!x->outs) compiler_panic(x->c, no_loc(), "link_macho: oom on OutSec");
memset(x->outs, 0, sizeof(OutSec) * cap);
x->nouts = 0;
for (u32 i = 0; i < x->nsecs; ++i) {
@@ -1197,12 +1237,12 @@ static void shift_sections(MCtx* x) {
*/
typedef struct FixSite {
- u8 segidx; /* 2 = __DATA_CONST, 3 = __DATA */
- u8 is_bind; /* 0 = rebase, 1 = bind */
+ u8 segidx; /* 2 = __DATA_CONST, 3 = __DATA */
+ u8 is_bind; /* 0 = rebase, 1 = bind */
u8 pad[2];
- u32 import_idx; /* 1-based import index for binds, 0 for rebases */
- u64 vaddr; /* absolute VA of the slot */
- u64 rebase_target; /* unslid target VA; only used for rebases */
+ u32 import_idx; /* 1-based import index for binds, 0 for rebases */
+ u64 vaddr; /* absolute VA of the slot */
+ u64 rebase_target; /* unslid target VA; only used for rebases */
} FixSite;
typedef struct FixList {
@@ -1445,7 +1485,7 @@ static void apply_relocs(MCtx* x, FixList* fl) {
u64 tgt_v = s ? s->vaddr : 0;
u8* slot = x->got_bytes + (mi->got_idx - 1u) * MZ_GOT_SIZE;
wr_u64_le(slot, tgt_v);
- if (tgt_v == 0) continue; /* weak-undef → NULL */
+ if (tgt_v == 0) continue; /* weak-undef → NULL */
FixSite fs = {2u, 0, {0}, 0, slot_v, tgt_v};
fix_push(fl, &fs);
} else {
@@ -1476,7 +1516,7 @@ static void apply_relocs(MCtx* x, FixList* fl) {
LinkSymbol* s = sym_at(img, ts->sym);
u64 tgt_v = s ? s->vaddr : 0;
wr_u64_le(slot, tgt_v);
- if (tgt_v == 0) continue; /* weak-undef descriptor → NULL */
+ if (tgt_v == 0) continue; /* weak-undef descriptor → NULL */
FixSite fs = {3u, 0, {0}, 0, slot_v, tgt_v};
fix_push(fl, &fs);
}
@@ -1492,9 +1532,9 @@ static void apply_relocs(MCtx* x, FixList* fl) {
*/
typedef struct PageChain {
- u32 first_offset_in_page; /* relative to page start */
+ u32 first_offset_in_page; /* relative to page start */
u32 nsites;
- u32 first_site_idx; /* into a per-segment site array */
+ u32 first_site_idx; /* into a per-segment site array */
} PageChain;
static int site_cmp_by_vaddr(const void* a, const void* b) {
@@ -1526,16 +1566,15 @@ static void emit_pointer(u8* slot, int is_bind, u32 ord_or_target_lo,
*/
u64 v = 0;
if (is_bind) {
- u64 ordinal = (u64)ord_or_target_lo & 0xffffffull; /* 24 bits */
+ u64 ordinal = (u64)ord_or_target_lo & 0xffffffull; /* 24 bits */
u64 addend = 0;
u64 next = (u64)next4 & 0xfffull;
v = ordinal | (addend << 24) | (0ull /* reserved */ << 32) | (next << 51) |
((u64)1 << 63);
} else {
/* rebase: target is full vmaddr; we get hi:lo split. */
- u64 target =
- ((u64)high_or_target_hi << 32) | (u64)ord_or_target_lo;
- target &= ((u64)1 << 36) - 1u; /* 36 bits */
+ u64 target = ((u64)high_or_target_hi << 32) | (u64)ord_or_target_lo;
+ target &= ((u64)1 << 36) - 1u; /* 36 bits */
u64 high8 = 0;
u64 next = (u64)next4 & 0xfffull;
v = target | (high8 << 36) | (0ull /* reserved */ << 44) | (next << 51) |
@@ -1558,14 +1597,14 @@ static void build_chained_fixups(MCtx* x, FixList* fl) {
* uint32 imports_format (=1)
* uint32 symbols_format (=0)
*/
- u32 hdr_pos = mbuf_u32(out, 0); /* fixups_version */
+ u32 hdr_pos = mbuf_u32(out, 0); /* fixups_version */
(void)hdr_pos;
u32 starts_offset_pos = mbuf_u32(out, 0);
u32 imports_offset_pos = mbuf_u32(out, 0);
u32 symbols_offset_pos = mbuf_u32(out, 0);
mbuf_u32(out, x->nimports_real);
mbuf_u32(out, DYLD_CHAINED_IMPORT);
- mbuf_u32(out, 0); /* symbols uncompressed */
+ mbuf_u32(out, 0); /* symbols uncompressed */
/* dyld expects 8-byte alignment of the starts table. */
mbuf_align(out, 4);
@@ -1618,10 +1657,10 @@ static void build_chained_fixups(MCtx* x, FixList* fl) {
* uint16 page_count
* uint16 page_start[page_count] (0xFFFF = no fixups in page)
*/
- u32 sis_size_pos = mbuf_u32(out, 0); /* fill below */
+ u32 sis_size_pos = mbuf_u32(out, 0); /* fill below */
mbuf_u16(out, (u16)MZ_PAGE);
mbuf_u16(out, (u16)DYLD_CHAINED_PTR_64);
- mbuf_u64(out, (u64)x->segs[si].fileoff); /* segment file offset */
+ mbuf_u64(out, (u64)x->segs[si].fileoff); /* segment file offset */
mbuf_u32(out, 0);
mbuf_u16(out, (u16)page_count);
u32 page_starts_pos = out->len;
@@ -1639,8 +1678,7 @@ static void build_chained_fixups(MCtx* x, FixList* fl) {
/* Sites are sorted globally; collect contiguous run for this seg. */
u32 cur = first;
while (cur < first + count) {
- u32 page_idx =
- (u32)((fl->a[cur].vaddr - seg_va) / MZ_PAGE);
+ u32 page_idx = (u32)((fl->a[cur].vaddr - seg_va) / MZ_PAGE);
u32 offset_in_page = (u32)((fl->a[cur].vaddr - seg_va) % MZ_PAGE);
wr_u16_le(out->data + page_starts_pos + page_idx * 2u,
(u16)offset_in_page);
@@ -1659,27 +1697,25 @@ static void build_chained_fixups(MCtx* x, FixList* fl) {
u64 dist = fl->a[k + 1].vaddr - s->vaddr;
next4 = (u32)(dist / 4u);
}
- /* Find segment bytes; for slot in __DATA_CONST __got use
- * x->got_bytes; for __DATA, special-case the synthetic
- * __thread_ptrs region and otherwise walk LinkSections. */
+ /* Find segment bytes. Synthetic pointer sections have private
+ * buffers; file-backed sections can live in any segment, including
+ * pointer-bearing read-only constants in __TEXT. */
u8* slot = NULL;
- if (s->segidx == 2) {
+ if (s->segidx == 2 && x->got_bytes && s->vaddr >= x->got_vaddr &&
+ s->vaddr < x->got_vaddr + x->got_size) {
/* __DATA_CONST: __got slot. */
slot = x->got_bytes + (s->vaddr - x->got_vaddr);
- } else if (s->segidx == 3) {
- if (x->tlv_ptrs_bytes && s->vaddr >= x->tlv_ptrs_vaddr &&
- s->vaddr < x->tlv_ptrs_vaddr + x->tlv_ptrs_size) {
- slot = x->tlv_ptrs_bytes + (s->vaddr - x->tlv_ptrs_vaddr);
- } else {
- /* __DATA: walk MSecs to find the matching one. */
- MSec* m = msec_for_vaddr(x, s->vaddr);
- if (m && m->link_sec_id) {
- u8* base = bytes_for_section(x, m, x->img);
- if (base) {
- LinkSection* ls = &x->img->sections[m->link_sec_id - 1u];
- u32 in_off = (u32)(ls->input_offset + (s->vaddr - m->vaddr));
- slot = base + in_off;
- }
+ } else if (x->tlv_ptrs_bytes && s->vaddr >= x->tlv_ptrs_vaddr &&
+ s->vaddr < x->tlv_ptrs_vaddr + x->tlv_ptrs_size) {
+ slot = x->tlv_ptrs_bytes + (s->vaddr - x->tlv_ptrs_vaddr);
+ } else {
+ MSec* m = msec_for_vaddr(x, s->vaddr);
+ if (m && m->link_sec_id) {
+ u8* base = bytes_for_section(x, m, x->img);
+ if (base) {
+ LinkSection* ls = &x->img->sections[m->link_sec_id - 1u];
+ u32 in_off = (u32)(ls->input_offset + (s->vaddr - m->vaddr));
+ slot = base + in_off;
}
}
}
@@ -1691,6 +1727,14 @@ static void build_chained_fixups(MCtx* x, FixList* fl) {
if (s->is_bind) {
/* ordinal is import index (1-based) - 1; chained-import format
* uses 0-based. */
+ if (s->import_idx == 0 || s->import_idx > x->nimports_real) {
+ compiler_panic(
+ x->c, no_loc(),
+ "link_macho: chained bind for vaddr 0x%llx uses import index "
+ "%u outside real import table size %u",
+ (unsigned long long)s->vaddr, (unsigned)s->import_idx,
+ (unsigned)x->nimports_real);
+ }
u32 ord = s->import_idx - 1u;
emit_pointer(slot, 1, ord, 0, next4);
} else {
@@ -1722,13 +1766,18 @@ static void build_chained_fixups(MCtx* x, FixList* fl) {
MachImp* mi = &x->imports[i];
size_t nl;
const char* nm = pool_str(x->c->global, mi->name, &nl);
+ if (!nm || !nl || mi->dylib_ord == 0 || mi->dylib_ord > x->ndylibs) {
+ compiler_panic(x->c, no_loc(),
+ "link_macho: invalid chained import %u "
+ "(name=%u dylib_ord=%u ndylibs=%u)",
+ (unsigned)i, (unsigned)mi->name,
+ (unsigned)mi->dylib_ord, (unsigned)x->ndylibs);
+ }
u32 off = out->len - symbols_off;
- if (nm && nl) mbuf_str(out, nm, (u32)nl);
- else mbuf_u8(out, 0);
+ mbuf_str(out, nm, (u32)nl);
/* Patch the import slot. */
u32 packed = ((u32)mi->dylib_ord & 0xffu) |
- ((u32)(mi->weak ? 1u : 0u) << 8) |
- ((off & 0x7fffffu) << 9);
+ ((u32)(mi->weak ? 1u : 0u) << 8) | ((off & 0x7fffffu) << 9);
wr_u32_le(out->data + imports_off + i * 4u, packed);
}
(void)symbols_off;
@@ -1767,8 +1816,8 @@ static void build_exports_trie(MCtx* x) {
LinkSymbol* esym = sym_at(img, img->entry_sym);
if (!esym || !esym->defined) {
/* No entry — emit a single empty terminal trie. */
- mbuf_u8(out, 0); /* terminal_size 0 */
- mbuf_u8(out, 0); /* children 0 */
+ mbuf_u8(out, 0); /* terminal_size 0 */
+ mbuf_u8(out, 0); /* children 0 */
return;
}
size_t nl;
@@ -1803,8 +1852,8 @@ static void build_exports_trie(MCtx* x) {
/* root: terminal_size=0, children_count=1, "_main"\0, child_offset=
* (leaf-position uleb). */
/* We'll back-patch child_offset after we know the leaf position. */
- mbuf_u8(out, 0); /* root terminal size */
- mbuf_u8(out, 1); /* children_count */
+ mbuf_u8(out, 0); /* root terminal size */
+ mbuf_u8(out, 1); /* children_count */
mbuf_str(out, nm, (u32)nl);
/* child offset: 5 bytes max for uleb128(u32). Reserve and patch. */
u32 child_off_pos = out->len;
@@ -1816,7 +1865,7 @@ static void build_exports_trie(MCtx* x) {
mbuf_u8(out, (u8)leaf_payload_len);
uleb128(out, flags);
uleb128(out, entry_off);
- mbuf_u8(out, 0); /* children_count */
+ mbuf_u8(out, 0); /* children_count */
/* Patch child_offset uleb. */
u32 v = leaf_pos;
@@ -1834,8 +1883,7 @@ static void build_exports_trie(MCtx* x) {
* non-zero byte and also forcing remaining bytes to be 0x80
* extension or trim. Simpler: set last byte explicitly. */
out->data[child_off_pos + i] = (u8)(out->data[child_off_pos + i] & 0x7fu);
- for (u32 j = i + 1; j < 5; ++j)
- out->data[child_off_pos + j] = 0x80;
+ for (u32 j = i + 1; j < 5; ++j) out->data[child_off_pos + j] = 0x80;
out->data[child_off_pos + 4] = 0x00;
break;
}
@@ -1849,7 +1897,7 @@ static void build_exports_trie(MCtx* x) {
typedef struct NlistRec {
u32 strx;
u8 type;
- u8 sect; /* 1-based section index (Mach-O) */
+ u8 sect; /* 1-based section index (Mach-O) */
u16 desc;
u64 value;
} NlistRec;
@@ -1885,7 +1933,7 @@ static void build_symtab(MCtx* x) {
if (!s->defined) continue;
if (s->bind != SB_GLOBAL && s->bind != SB_WEAK) continue;
if (s->name == 0) continue;
- if (s->kind == SK_ABS) continue; /* skip abs externs */
+ if (s->kind == SK_ABS) continue; /* skip abs externs */
/* Locate which OutSec contains this vaddr to figure out n_sect.
* n_sect is the 1-based index into the flat section_64 table the
* file actually contains (post-coalesce), matching what we emit
@@ -1935,9 +1983,9 @@ static void build_symtab(MCtx* x) {
wr_u32_le(t + 0, strx);
t[4] = N_UNDF | N_EXT;
t[5] = 0;
- /* n_desc carries dylib ordinal in high byte (REFERENCED_DYNAMICALLY etc.) */
- u16 desc =
- (u16)(((u16)mi->dylib_ord & 0xff) << 8);
+ /* n_desc carries dylib ordinal in high byte (REFERENCED_DYNAMICALLY etc.)
+ */
+ u16 desc = (u16)(((u16)mi->dylib_ord & 0xff) << 8);
if (mi->weak) desc |= N_WEAK_REF;
wr_u16_le(t + 6, desc);
wr_u64_le(t + 8, 0);
@@ -2056,20 +2104,12 @@ static void layout_linkedit(MCtx* x) {
* u32 magic (0xfade0c02)
* u32 length (bytes including all hashes)
* u32 version (>=0x20400 for execSeg fields)
- * u32 flags (=0 ad-hoc — actually flags must include 0x2 (kSecCodeSignatureAdhoc))
- * u32 hashOffset (offset of first slot hash)
- * u32 identOffset (offset of identifier string)
- * u32 nSpecialSlots (=0)
- * u32 nCodeSlots
- * u32 codeLimit (file bytes covered)
- * u8 hashSize (=32)
- * u8 hashType (=2 sha256)
- * u8 platform (=0)
- * u8 pageSize (=12 for 4096)
- * u32 spare2 (=0)
- * u32 scatterOffset (=0)
- * u32 teamOffset (=0)
- * u32 spare3 (=0)
+ * u32 flags (=0 ad-hoc — actually flags must include 0x2
+ * (kSecCodeSignatureAdhoc)) u32 hashOffset (offset of first slot hash) u32
+ * identOffset (offset of identifier string) u32 nSpecialSlots (=0) u32
+ * nCodeSlots u32 codeLimit (file bytes covered) u8 hashSize (=32) u8
+ * hashType (=2 sha256) u8 platform (=0) u8 pageSize (=12 for 4096) u32
+ * spare2 (=0) u32 scatterOffset (=0) u32 teamOffset (=0) u32 spare3 (=0)
* u64 codeLimit64 (=0)
* u64 execSegBase (=__TEXT.fileoff)
* u64 execSegLimit (=__TEXT.filesize)
@@ -2088,7 +2128,7 @@ static void wr_u64_be(u8* p, u64 v) {
/* Build the codesig blob with placeholder hashes; size is precise so
* file layout is final after this. */
static void build_codesig_skeleton(MCtx* x, u32 code_limit, const char* ident) {
- u32 code_page = 1u << CS_PAGE_SIZE_LOG2; /* 4096 */
+ u32 code_page = 1u << CS_PAGE_SIZE_LOG2; /* 4096 */
u32 nslots = (code_limit + code_page - 1u) / code_page;
/* CodeDirectory size:
@@ -2112,7 +2152,7 @@ static void build_codesig_skeleton(MCtx* x, u32 code_limit, const char* ident) {
/* SuperBlob header */
wr_u32_be(sb + 0, CS_MAGIC_EMBEDDED_SIGNATURE);
wr_u32_be(sb + 4, sb_size);
- wr_u32_be(sb + 8, 1); /* count */
+ wr_u32_be(sb + 8, 1); /* count */
/* slot 0: type=CSSLOT_CODEDIRECTORY, offset=20 */
wr_u32_be(sb + 12, CSSLOT_CODEDIRECTORY);
wr_u32_be(sb + 16, 20u);
@@ -2121,24 +2161,24 @@ static void build_codesig_skeleton(MCtx* x, u32 code_limit, const char* ident) {
u8* cd = sb + 20;
wr_u32_be(cd + 0, CS_MAGIC_CODEDIRECTORY);
wr_u32_be(cd + 4, cd_size);
- wr_u32_be(cd + 8, 0x20400u); /* version with execSeg */
- wr_u32_be(cd + 12, 0x2u); /* flags = adhoc */
- wr_u32_be(cd + 16, cd_hdr + ident_len); /* hashOffset */
- wr_u32_be(cd + 20, cd_hdr); /* identOffset */
- wr_u32_be(cd + 24, 0); /* nSpecialSlots */
+ wr_u32_be(cd + 8, 0x20400u); /* version with execSeg */
+ wr_u32_be(cd + 12, 0x2u); /* flags = adhoc */
+ wr_u32_be(cd + 16, cd_hdr + ident_len); /* hashOffset */
+ wr_u32_be(cd + 20, cd_hdr); /* identOffset */
+ wr_u32_be(cd + 24, 0); /* nSpecialSlots */
wr_u32_be(cd + 28, nslots);
wr_u32_be(cd + 32, code_limit);
cd[36] = (u8)CS_SHA256_LEN;
cd[37] = (u8)CS_HASHTYPE_SHA256;
- cd[38] = 0; /* platform */
+ cd[38] = 0; /* platform */
cd[39] = (u8)CS_PAGE_SIZE_LOG2;
- wr_u32_be(cd + 40, 0); /* spare2 */
- wr_u32_be(cd + 44, 0); /* scatterOffset */
- wr_u32_be(cd + 48, 0); /* teamOffset */
- wr_u32_be(cd + 52, 0); /* spare3 */
- wr_u64_be(cd + 56, 0); /* codeLimit64 */
- wr_u64_be(cd + 64, x->segs[1].fileoff); /* execSegBase */
- wr_u64_be(cd + 72, x->segs[1].filesize); /* execSegLimit */
+ wr_u32_be(cd + 40, 0); /* spare2 */
+ wr_u32_be(cd + 44, 0); /* scatterOffset */
+ wr_u32_be(cd + 48, 0); /* teamOffset */
+ wr_u32_be(cd + 52, 0); /* spare3 */
+ wr_u64_be(cd + 56, 0); /* codeLimit64 */
+ wr_u64_be(cd + 64, x->segs[1].fileoff); /* execSegBase */
+ wr_u64_be(cd + 72, x->segs[1].filesize); /* execSegLimit */
wr_u64_be(cd + 80, CS_EXECSEG_MAIN_BINARY);
/* identifier */
@@ -2173,8 +2213,7 @@ static void compute_codesig(MCtx* x, const u8* full_file, u32 file_len_excl_cs,
static void emit_load_command_segment(MByte* lc, MCtx* x, u32 segidx) {
MSeg* sg = &x->segs[segidx];
- u32 seg_cmd_size =
- MACHO_SEGCMD64_SIZE + sg->nouts * MACHO_SECT64_SIZE;
+ u32 seg_cmd_size = MACHO_SEGCMD64_SIZE + sg->nouts * MACHO_SECT64_SIZE;
u32 base = lc->len;
mbuf_u32(lc, LC_SEGMENT_64);
mbuf_u32(lc, seg_cmd_size);
@@ -2192,7 +2231,7 @@ static void emit_load_command_segment(MByte* lc, MCtx* x, u32 segidx) {
mbuf_u32(lc, sg->maxprot);
mbuf_u32(lc, sg->initprot);
mbuf_u32(lc, sg->nouts);
- mbuf_u32(lc, 0); /* flags */
+ mbuf_u32(lc, 0); /* flags */
for (u32 j = 0; j < sg->nouts; ++j) {
OutSec* o = &x->outs[sg->first_out + j];
@@ -2202,7 +2241,7 @@ static void emit_load_command_segment(MByte* lc, MCtx* x, u32 segidx) {
size_t sl = o->sectname ? strlen(o->sectname) : 0;
if (sl > 16) sl = 16;
if (sl) memcpy(sname, o->sectname, sl);
- size_t gl = strlen(sg->name); /* segname must match */
+ size_t gl = strlen(sg->name); /* segname must match */
if (gl > 16) gl = 16;
memcpy(gname, sg->name, gl);
mbuf_append(lc, sname, 16);
@@ -2215,12 +2254,12 @@ static void emit_load_command_segment(MByte* lc, MCtx* x, u32 segidx) {
u32 al = 0;
while ((1u << al) < a) ++al;
mbuf_u32(lc, al);
- mbuf_u32(lc, 0); /* reloff */
- mbuf_u32(lc, 0); /* nreloc */
+ mbuf_u32(lc, 0); /* reloff */
+ mbuf_u32(lc, 0); /* nreloc */
mbuf_u32(lc, o->flags);
mbuf_u32(lc, o->reserved1);
mbuf_u32(lc, o->reserved2);
- mbuf_u32(lc, 0); /* reserved3 */
+ mbuf_u32(lc, 0); /* reserved3 */
}
(void)base;
}
@@ -2287,13 +2326,12 @@ void link_emit_macho(LinkImage* img, Writer* w) {
mbuf_init(&lc, x.h);
/* LC_SEGMENT_64 for each segment with sections (and PAGEZERO/LINKEDIT). */
- emit_load_command_segment(&lc, &x, 0); /* PAGEZERO */
- emit_load_command_segment(&lc, &x, 1); /* TEXT */
+ emit_load_command_segment(&lc, &x, 0); /* PAGEZERO */
+ emit_load_command_segment(&lc, &x, 1); /* TEXT */
if (x.segs[2].nsects > 0)
- emit_load_command_segment(&lc, &x, 2); /* DATA_CONST */
- if (x.segs[3].nsects > 0)
- emit_load_command_segment(&lc, &x, 3); /* DATA */
- emit_load_command_segment(&lc, &x, 4); /* LINKEDIT */
+ emit_load_command_segment(&lc, &x, 2); /* DATA_CONST */
+ if (x.segs[3].nsects > 0) emit_load_command_segment(&lc, &x, 3); /* DATA */
+ emit_load_command_segment(&lc, &x, 4); /* LINKEDIT */
/* LC_DYLD_CHAINED_FIXUPS (linkedit_data_command: 16B) */
mbuf_u32(&lc, LC_DYLD_CHAINED_FIXUPS);
@@ -2324,19 +2362,24 @@ void link_emit_macho(LinkImage* img, Writer* w) {
u32 nextdef = (x.nsyms > nundef) ? x.nsyms - nundef - nlocal : 0;
mbuf_u32(&lc, LC_DYSYMTAB);
mbuf_u32(&lc, MACHO_DYSYMTAB_CMD_SIZE);
- mbuf_u32(&lc, 0); /* ilocalsym */
+ mbuf_u32(&lc, 0); /* ilocalsym */
mbuf_u32(&lc, nlocal);
mbuf_u32(&lc, nlocal);
mbuf_u32(&lc, nextdef);
mbuf_u32(&lc, nlocal + nextdef);
mbuf_u32(&lc, nundef);
- mbuf_u32(&lc, 0); mbuf_u32(&lc, 0); /* tocoff, ntoc */
- mbuf_u32(&lc, 0); mbuf_u32(&lc, 0); /* modtaboff, nmodtab */
- mbuf_u32(&lc, 0); mbuf_u32(&lc, 0); /* extrefsymoff, nextrefsyms */
+ mbuf_u32(&lc, 0);
+ mbuf_u32(&lc, 0); /* tocoff, ntoc */
+ mbuf_u32(&lc, 0);
+ mbuf_u32(&lc, 0); /* modtaboff, nmodtab */
+ mbuf_u32(&lc, 0);
+ mbuf_u32(&lc, 0); /* extrefsymoff, nextrefsyms */
mbuf_u32(&lc, x.indirect_off);
mbuf_u32(&lc, x.indirect.len / 4u);
- mbuf_u32(&lc, 0); mbuf_u32(&lc, 0); /* extreloff, nextrel */
- mbuf_u32(&lc, 0); mbuf_u32(&lc, 0); /* locreloff, nlocrel */
+ mbuf_u32(&lc, 0);
+ mbuf_u32(&lc, 0); /* extreloff, nextrel */
+ mbuf_u32(&lc, 0);
+ mbuf_u32(&lc, 0); /* locreloff, nlocrel */
/* LC_LOAD_DYLINKER */
{
@@ -2344,7 +2387,7 @@ void link_emit_macho(LinkImage* img, Writer* w) {
u32 cmd_size = (u32)ALIGN_UP((u64)(12u + (u32)strlen(dyld) + 1u), 8u);
mbuf_u32(&lc, LC_LOAD_DYLINKER);
mbuf_u32(&lc, cmd_size);
- mbuf_u32(&lc, 12u); /* name offset within cmd */
+ mbuf_u32(&lc, 12u); /* name offset within cmd */
u32 wrote = mbuf_str(&lc, dyld, (u32)strlen(dyld));
(void)wrote;
/* Pad to cmd_size. */
@@ -2369,17 +2412,17 @@ void link_emit_macho(LinkImage* img, Writer* w) {
/* LC_BUILD_VERSION */
mbuf_u32(&lc, LC_BUILD_VERSION);
mbuf_u32(&lc, 24);
- mbuf_u32(&lc, 1); /* PLATFORM_MACOS */
- mbuf_u32(&lc, (12u << 16) | 0); /* minos 12.0.0 */
- mbuf_u32(&lc, (12u << 16) | 0); /* sdk 12.0.0 */
- mbuf_u32(&lc, 0); /* ntools */
+ mbuf_u32(&lc, 1); /* PLATFORM_MACOS */
+ mbuf_u32(&lc, (12u << 16) | 0); /* minos 12.0.0 */
+ mbuf_u32(&lc, (12u << 16) | 0); /* sdk 12.0.0 */
+ mbuf_u32(&lc, 0); /* ntools */
/* LC_MAIN — entryoff is offset within __TEXT segment from its file
* start (0). */
mbuf_u32(&lc, LC_MAIN);
mbuf_u32(&lc, 24);
- mbuf_u64(&lc, (u64)x.entry_offset); /* entryoff = vaddr - __TEXT.vmaddr */
- mbuf_u64(&lc, 0); /* stacksize */
+ mbuf_u64(&lc, (u64)x.entry_offset); /* entryoff = vaddr - __TEXT.vmaddr */
+ mbuf_u64(&lc, 0); /* stacksize */
/* LC_LOAD_DYLIB per dylib. */
for (u32 i = 0; i < x.ndylibs; ++i) {
@@ -2389,10 +2432,10 @@ void link_emit_macho(LinkImage* img, Writer* w) {
u32 cmd_start = lc.len;
mbuf_u32(&lc, LC_LOAD_DYLIB);
mbuf_u32(&lc, cmd_size);
- mbuf_u32(&lc, 24u); /* name offset */
- mbuf_u32(&lc, 0); /* timestamp */
- mbuf_u32(&lc, (1u << 16)); /* current_version 1.0 */
- mbuf_u32(&lc, (1u << 16)); /* compat_version 1.0 */
+ mbuf_u32(&lc, 24u); /* name offset */
+ mbuf_u32(&lc, 0); /* timestamp */
+ mbuf_u32(&lc, (1u << 16)); /* current_version 1.0 */
+ mbuf_u32(&lc, (1u << 16)); /* compat_version 1.0 */
mbuf_str(&lc, nm ? nm : "", (u32)nl);
while (lc.len - cmd_start < cmd_size) mbuf_u8(&lc, 0);
}
@@ -2435,10 +2478,10 @@ void link_emit_macho(LinkImage* img, Writer* w) {
* + chained + exports_trie + symtab + dysymtab + dyld + uuid +
* build_version + main + nDylibs + fn_starts + data_in_code +
* codesig. */
- ncmds += 2; /* PAGEZERO + TEXT */
+ ncmds += 2; /* PAGEZERO + TEXT */
if (x.segs[2].nsects > 0) ncmds++;
if (x.segs[3].nsects > 0) ncmds++;
- ncmds++; /* LINKEDIT */
+ ncmds++; /* LINKEDIT */
ncmds += 11 + x.ndylibs;
/* (chained, exports_trie, symtab, dysymtab, dyld, uuid, build_version,
* main, fn_starts, data_in_code, codesig) = 11 */
@@ -2458,7 +2501,7 @@ void link_emit_macho(LinkImage* img, Writer* w) {
if (x.ntlv) mh_flags |= MH_HAS_TLV_DESCRIPTORS;
mbuf_u32(&file, mh_flags);
}
- mbuf_u32(&file, 0); /* reserved */
+ mbuf_u32(&file, 0); /* reserved */
mbuf_append(&file, lc.data, lc.len);
/* Pad to first section's file offset. */
@@ -2477,8 +2520,8 @@ void link_emit_macho(LinkImage* img, Writer* w) {
} else {
LinkSection* ls = &img->sections[m->link_sec_id - 1u];
u32 segid = ls->segment_id;
- u8* base = (segid != LINK_SEG_NONE) ? img->segment_bytes[segid - 1u]
- : NULL;
+ u8* base =
+ (segid != LINK_SEG_NONE) ? img->segment_bytes[segid - 1u] : NULL;
if (base && ls->size) {
mbuf_append(&file, base + ls->input_offset, (u32)ls->size);
} else if (ls->size) {
@@ -2529,7 +2572,7 @@ void link_emit_macho(LinkImage* img, Writer* w) {
mbuf_fini(&x.fn_starts);
mbuf_fini(&x.data_in_code);
mbuf_fini(&x.codesig);
- if (x.imports) x.h->free(x.h, x.imports, 0); /* VEC_GROW: cap unknown */
+ if (x.imports) x.h->free(x.h, x.imports, 0); /* VEC_GROW: cap unknown */
if (x.dylibs) x.h->free(x.h, x.dylibs, 0);
if (x.sym_to_imp)
x.h->free(x.h, x.sym_to_imp, sizeof(u32) * x.sym_to_imp_size);
diff --git a/src/link/link_resolve.c b/src/link/link_resolve.c
@@ -59,6 +59,11 @@ static int bind_strength(u8 bind) {
}
}
+static int obj_sym_is_logical_undef(const ObjSym* s) {
+ return s && s->section_id == OBJ_SEC_NONE && s->kind != SK_ABS &&
+ s->kind != SK_COMMON;
+}
+
void link_resolve_symbols(Linker* l, LinkImage* img) {
u32 ii;
for (ii = 0; ii < LinkInputs_count(&l->inputs); ++ii) {
@@ -86,8 +91,7 @@ void link_resolve_symbols(Linker* l, LinkImage* img) {
LinkSymbol rec;
LinkSymId existing;
{
- int is_logical_undef = (s->section_id == OBJ_SEC_NONE) &&
- (s->kind != SK_ABS) && (s->kind != SK_COMMON);
+ int is_logical_undef = obj_sym_is_logical_undef(s);
if (is_logical_undef && !s->referenced &&
(s->bind == SB_GLOBAL || s->bind == SB_WEAK)) {
continue;
@@ -494,7 +498,8 @@ void link_gc_drop_dead_globals(Linker* l, LinkImage* img, const GcLive* g) {
/* ---- archive ingestion ---- */
-static void include_archive_member(Linker* l, LinkArchiveMember* mem) {
+static void include_archive_member(Linker* l, const LinkArchive* ar,
+ LinkArchiveMember* mem) {
LinkInput* in;
LinkInputId id;
u32 idx;
@@ -505,24 +510,28 @@ static void include_archive_member(Linker* l, LinkArchiveMember* mem) {
id = (LinkInputId)(idx + 1u);
in->id = id;
in->kind = LINK_INPUT_OBJ_BYTES;
+ in->order = ar->order;
in->obj = mem->obj;
in->name = mem->name;
mem->included = 1;
mem->obj = NULL;
}
-static void scan_presence(Linker* l, SymHash* defined, SymHash* undefs) {
+static void scan_presence_before(Linker* l, u32 max_order, SymHash* defined,
+ SymHash* undefs) {
u32 ii;
ObjSymIter* it;
ObjSymEntry e;
for (ii = 0; ii < LinkInputs_count(&l->inputs); ++ii) {
- ObjBuilder* ob = LinkInputs_at(&l->inputs, ii)->obj;
+ LinkInput* in = LinkInputs_at(&l->inputs, ii);
+ ObjBuilder* ob = in->obj;
+ if (!ob || in->order > max_order) continue;
it = obj_symiter_new(ob);
while (obj_symiter_next(it, &e)) {
const ObjSym* s = e.sym;
if (s->name == 0) continue;
if (s->bind == SB_LOCAL) continue;
- if (s->kind == SK_UNDEF)
+ if (obj_sym_is_logical_undef(s))
symhash_set(undefs, s->name, 1u);
else
symhash_set(defined, s->name, 1u);
@@ -531,12 +540,14 @@ static void scan_presence(Linker* l, SymHash* defined, SymHash* undefs) {
}
}
-static int inputs_have_defined_ifunc(Linker* l) {
+static int inputs_have_defined_ifunc_before(Linker* l, u32 max_order) {
u32 ii;
ObjSymIter* it;
ObjSymEntry e;
for (ii = 0; ii < LinkInputs_count(&l->inputs); ++ii) {
- ObjBuilder* ob = LinkInputs_at(&l->inputs, ii)->obj;
+ LinkInput* in = LinkInputs_at(&l->inputs, ii);
+ ObjBuilder* ob = in->obj;
+ if (!ob || in->order > max_order) continue;
it = obj_symiter_new(ob);
while (obj_symiter_next(it, &e)) {
const ObjSym* s = e.sym;
@@ -578,37 +589,35 @@ void link_ingest_archives(Linker* l) {
LinkArchive* ar = LinkArchives_at(&l->archives, a);
if (!ar->whole_archive) continue;
for (m = 0; m < ar->nmembers; ++m)
- include_archive_member(l, &ar->members[m]);
+ include_archive_member(l, ar, &ar->members[m]);
}
- Sym want_ifunc_init = 0;
- if (l->emit_static_exe && inputs_have_defined_ifunc(l)) {
- want_ifunc_init = pool_intern_cstr(l->c->global, "__cfree_ifunc_init");
- }
+ for (a = 0; a < LinkArchives_count(&l->archives); ++a) {
+ LinkArchive* ar = LinkArchives_at(&l->archives, a);
+ Sym want_ifunc_init = 0;
+ if (ar->whole_archive) continue;
+ if (l->emit_static_exe && inputs_have_defined_ifunc_before(l, ar->order))
+ want_ifunc_init = pool_intern_cstr(l->c->global, "__cfree_ifunc_init");
+ for (;;) {
+ SymHash defined, undefs;
+ int changed = 0;
+ symhash_init(&defined, l->heap);
+ symhash_init(&undefs, l->heap);
+ scan_presence_before(l, ar->order, &defined, &undefs);
+ if (want_ifunc_init != 0 &&
+ symhash_get(&defined, want_ifunc_init) == LINK_SYM_NONE)
+ symhash_set(&undefs, want_ifunc_init, 1u);
- for (;;) {
- SymHash defined, undefs;
- int changed = 0;
- symhash_init(&defined, l->heap);
- symhash_init(&undefs, l->heap);
- scan_presence(l, &defined, &undefs);
- if (want_ifunc_init != 0 &&
- symhash_get(&defined, want_ifunc_init) == LINK_SYM_NONE)
- symhash_set(&undefs, want_ifunc_init, 1u);
-
- for (a = 0; a < LinkArchives_count(&l->archives); ++a) {
- LinkArchive* ar = LinkArchives_at(&l->archives, a);
- if (ar->whole_archive) continue;
for (m = 0; m < ar->nmembers; ++m) {
LinkArchiveMember* mem = &ar->members[m];
if (mem->included) continue;
if (!member_satisfies(mem, &defined, &undefs)) continue;
- include_archive_member(l, mem);
+ include_archive_member(l, ar, mem);
changed = 1;
}
+ symhash_fini(&defined);
+ symhash_fini(&undefs);
+ if (!changed) break;
}
- symhash_fini(&defined);
- symhash_fini(&undefs);
- if (!changed) break;
}
}
diff --git a/test/link/cases/37_macho_ro_const_abs64/a.aa64.S b/test/link/cases/37_macho_ro_const_abs64/a.aa64.S
@@ -0,0 +1,17 @@
+.section __TEXT,__const
+.p2align 3
+_ro_const_ptr:
+ .quad _ro_const_value
+
+.p2align 2
+_ro_const_value:
+ .long 7
+
+.text
+.globl _test_main
+.p2align 2
+_test_main:
+ adrp x8, _ro_const_ptr@PAGE
+ ldr x8, [x8, _ro_const_ptr@PAGEOFF]
+ ldr w0, [x8]
+ ret
diff --git a/test/link/cases/37_macho_ro_const_abs64/expected b/test/link/cases/37_macho_ro_const_abs64/expected
@@ -0,0 +1 @@
+7
diff --git a/test/link/cases/37_macho_ro_const_abs64/j_targets b/test/link/cases/37_macho_ro_const_abs64/j_targets
@@ -0,0 +1 @@
+
diff --git a/test/link/cases/37_macho_ro_const_abs64/targets b/test/link/cases/37_macho_ro_const_abs64/targets
@@ -0,0 +1 @@
+aa64-macho
