#define _GNU_SOURCE
+#define SYSCALL_NO_TLS 1
#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#define ALIGN(x,y) ((x)+(y)-1 & -(y))
#define container_of(p,t,m) ((t*)((char *)(p)-offsetof(t,m)))
+#define countof(a) ((sizeof (a))/(sizeof (a)[0]))
struct debug {
int ver;
size_t relro_start, relro_end;
uintptr_t *new_dtv;
unsigned char *new_tls;
- volatile int new_dtv_idx, new_tls_idx;
struct td_index *td_index;
struct dso *fini_next;
char *shortname;
static int ldd_mode;
static int ldso_fail;
static int noload;
+static int shutting_down;
static jmp_buf *rtld_fail;
static pthread_rwlock_t lock;
static struct debug debug;
static size_t static_tls_cnt;
static pthread_mutex_t init_fini_lock;
static pthread_cond_t ctor_cond;
+static struct dso *builtin_deps[2];
+static struct dso *const no_deps[1];
+static struct dso *builtin_ctor_queue[4];
static struct dso **main_ctor_queue;
static struct fdpic_loadmap *app_loadmap;
static struct fdpic_dummy_loadmap app_dummy_loadmap;
#define ARCH_SYM_REJECT_UND(s) 0
#endif
-static struct symdef find_sym(struct dso *dso, const char *s, int need_def)
+#if defined(__GNUC__)
+__attribute__((always_inline))
+#endif
+static inline struct symdef find_sym2(struct dso *dso, const char *s, int need_def, int use_deps)
{
uint32_t h = 0, gh = gnu_hash(s), gho = gh / (8*sizeof(size_t)), *ght;
size_t ghm = 1ul << gh % (8*sizeof(size_t));
struct symdef def = {0};
- for (; dso; dso=dso->syms_next) {
+ struct dso **deps = use_deps ? dso->deps : 0;
+ for (; dso; dso=use_deps ? *deps++ : dso->syms_next) {
Sym *sym;
if ((ght = dso->ghashtab)) {
sym = gnu_lookup_filtered(gh, ght, dso, s, gho, ghm);
return def;
}
+static struct symdef find_sym(struct dso *dso, const char *s, int need_def)
+{
+ return find_sym2(dso, s, need_def, 0);
+}
+
static void do_relocs(struct dso *dso, size_t *rel, size_t rel_size, size_t stride)
{
unsigned char *base = dso->base;
sym = syms + sym_index;
name = strings + sym->st_name;
ctx = type==REL_COPY ? head->syms_next : head;
- def = (sym->st_info&0xf) == STT_SECTION
+ def = (sym->st_info>>4) == STB_LOCAL
? (struct symdef){ .dso = dso, .sym = sym }
: find_sym(ctx, name, type==REL_PLT);
if (!def.sym && (sym->st_shndx != SHN_UNDEF
tls_val = def.sym ? def.sym->st_value : 0;
if ((type == REL_TPOFF || type == REL_TPOFF_NEG)
- && runtime && def.dso->tls_id > static_tls_cnt) {
+ && def.dso->tls_id > static_tls_cnt) {
error("Error relocating %s: %s: initial-exec TLS "
"resolves to dynamic definition in %s",
dso->name, name, def.dso->name);
case REL_PLT:
*reloc_addr = sym_val + addend;
break;
+ case REL_USYMBOLIC:
+ memcpy(reloc_addr, &(size_t){sym_val + addend}, sizeof(size_t));
+ break;
case REL_RELATIVE:
*reloc_addr = (size_t)base + addend;
break;
#endif
case REL_TLSDESC:
if (stride<3) addend = reloc_addr[1];
- if (runtime && def.dso->tls_id > static_tls_cnt) {
+ if (def.dso->tls_id > static_tls_cnt) {
struct td_index *new = malloc(sizeof *new);
if (!new) {
error(
#else
p = (void *)__syscall(SYS_mmap, 0, n, prot, flags, -1, 0);
#endif
- return p == MAP_FAILED ? 0 : p;
+ return (unsigned long)p > -4096UL ? 0 : p;
}
static void makefuncdescs(struct dso *p)
p->tls_id = ++tls_cnt;
tls_align = MAXP2(tls_align, p->tls.align);
#ifdef TLS_ABOVE_TP
- p->tls.offset = tls_offset + ( (tls_align-1) &
- -(tls_offset + (uintptr_t)p->tls.image) );
- tls_offset += p->tls.size;
+ p->tls.offset = tls_offset + ( (p->tls.align-1) &
+ (-tls_offset + (uintptr_t)p->tls.image) );
+ tls_offset = p->tls.offset + p->tls.size;
#else
tls_offset += p->tls.size + p->tls.align - 1;
tls_offset -= (tls_offset + (uintptr_t)p->tls.image)
static void load_direct_deps(struct dso *p)
{
size_t i, cnt=0;
+
if (p->deps) return;
/* For head, all preloads are direct pseudo-dependencies.
* Count and include them now to avoid realloc later. */
cnt++;
for (i=0; p->dynv[i]; i+=2)
if (p->dynv[i] == DT_NEEDED) cnt++;
- p->deps = calloc(cnt+1, sizeof *p->deps);
+ /* Use builtin buffer for apps with no external deps, to
+ * preserve property of no runtime failure paths. */
+ p->deps = (p==head && cnt<2) ? builtin_deps :
+ calloc(cnt+1, sizeof *p->deps);
if (!p->deps) {
error("Error loading dependencies for %s", p->name);
if (runtime) longjmp(*rtld_fail, 1);
struct dso **tmp;
/* Can't use realloc if the original p->deps was allocated at
- * program entry and malloc has been replaced. */
- int no_realloc = __malloc_replaced && !p->runtime_loaded;
+ * program entry and malloc has been replaced, or if it's
+ * the builtin non-allocated trivial main program deps array. */
+ int no_realloc = (__malloc_replaced && !p->runtime_loaded)
+ || p->deps == builtin_deps;
if (p->bfs_built) return;
ndeps_all = p->ndeps_direct;
{
struct dso *p;
size_t dyn[DYN_CNT];
+ int self = __pthread_self()->tid;
+
+ /* Take both locks before setting shutting_down, so that
+ * either lock is sufficient to read its value. The lock
+ * order matches that in dlopen to avoid deadlock. */
+ pthread_rwlock_wrlock(&lock);
+ pthread_mutex_lock(&init_fini_lock);
+ shutting_down = 1;
+ pthread_rwlock_unlock(&lock);
for (p=fini_head; p; p=p->fini_next) {
+ while (p->ctor_visitor && p->ctor_visitor!=self)
+ pthread_cond_wait(&ctor_cond, &init_fini_lock);
if (!p->constructed) continue;
decode_vec(p->dynv, dyn, DYN_CNT);
if (dyn[0] & (1<<DT_FINI_ARRAY)) {
p->mark = 0;
}
cnt++; /* termination slot */
- stack = queue = calloc(cnt, sizeof *queue);
+ if (dso==head && cnt <= countof(builtin_ctor_queue))
+ queue = builtin_ctor_queue;
+ else
+ queue = calloc(cnt, sizeof *queue);
if (!queue) {
error("Error allocating constructor queue: %m\n");
/* Opposite ends of the allocated buffer serve as an output queue
* and a working stack. Setup initial stack with just the argument
* dso and initial queue empty... */
+ stack = queue;
qpos = 0;
spos = cnt;
stack[--spos] = dso;
pthread_mutex_lock(&init_fini_lock);
for (i=0; (p=queue[i]); i++) {
- while (p->ctor_visitor && p->ctor_visitor!=self)
+ while ((p->ctor_visitor && p->ctor_visitor!=self) || shutting_down)
pthread_cond_wait(&ctor_cond, &init_fini_lock);
if (p->ctor_visitor || p->constructed)
continue;
- if (p->constructed) continue;
p->ctor_visitor = self;
decode_vec(p->dynv, dyn, DYN_CNT);
void __libc_start_init(void)
{
do_init_fini(main_ctor_queue);
- /* This is valid because the queue was allocated after redoing
- * relocations with any interposed malloc having taken effect. */
- free(main_ctor_queue);
+ if (!__malloc_replaced && main_ctor_queue != builtin_ctor_queue)
+ free(main_ctor_queue);
main_ctor_queue = 0;
}
* so that loads of the thread pointer and &errno can be pure/const and
* thereby hoistable. */
-_Noreturn void __dls2b(size_t *sp)
+void __dls2b(size_t *sp)
{
/* Setup early thread pointer in builtin_tls for ldso/libc itself to
* use during dynamic linking. If possible it will also serve as the
* process dependencies and relocations for the main application and
* transfer control to its entry point. */
-_Noreturn void __dls3(size_t *sp)
+void __dls3(size_t *sp)
{
static struct dso app, vdso;
size_t aux[AUX_CNT], *auxv;
libc.auxv = auxv = (void *)(argv+i+1);
decode_vec(auxv, aux, AUX_CNT);
__hwcap = aux[AT_HWCAP];
+ search_vec(auxv, &__sysinfo, AT_SYSINFO);
+ __pthread_self()->sysinfo = __sysinfo;
libc.page_size = aux[AT_PAGESZ];
libc.secure = ((aux[0]&0x7800)!=0x7800 || aux[AT_UID]!=aux[AT_EUID]
|| aux[AT_GID]!=aux[AT_EGID] || aux[AT_SECURE]);
app.tls_id = tls_cnt = 1;
#ifdef TLS_ABOVE_TP
app.tls.offset = GAP_ABOVE_TP;
- app.tls.offset += -GAP_ABOVE_TP & (app.tls.align-1);
- tls_offset = app.tls.offset + app.tls.size
- + ( -((uintptr_t)app.tls.image + app.tls.size)
- & (app.tls.align-1) );
+ app.tls.offset += (-GAP_ABOVE_TP + (uintptr_t)app.tls.image)
+ & (app.tls.align-1);
+ tls_offset = app.tls.offset + app.tls.size;
#else
tls_offset = app.tls.offset = app.tls.size
+ ( -((uintptr_t)app.tls.image + app.tls.size)
reclaim_gaps(&ldso);
/* Load preload/needed libraries, add symbols to global namespace. */
+ ldso.deps = (struct dso **)no_deps;
if (env_preload) load_preload(env_preload);
load_deps(&app);
for (struct dso *p=head; p; p=p->next)
vdso.name = "";
vdso.shortname = "linux-gate.so.1";
vdso.relocated = 1;
+ vdso.deps = (struct dso **)no_deps;
decode_dyn(&vdso);
vdso.prev = tail;
tail->next = &vdso;
}
}
- /* The main program must be relocated LAST since it may contin
- * copy relocations which depend on libraries' relocations. */
- reloc_all(app.next);
- reloc_all(&app);
+ /* This must be done before final relocations, since it calls
+ * malloc, which may be provided by the application. Calling any
+ * application code prior to the jump to its entry point is not
+ * valid in our model and does not work with FDPIC, where there
+ * are additional relocation-like fixups that only the entry point
+ * code can see to perform. */
+ main_ctor_queue = queue_ctors(&app);
+ /* Initial TLS must also be allocated before final relocations
+ * might result in calloc being a call to application code. */
update_tls_size();
+ void *initial_tls = builtin_tls;
if (libc.tls_size > sizeof builtin_tls || tls_align > MIN_TLS_ALIGN) {
- void *initial_tls = calloc(libc.tls_size, 1);
+ initial_tls = calloc(libc.tls_size, 1);
if (!initial_tls) {
dprintf(2, "%s: Error getting %zu bytes thread-local storage: %m\n",
argv[0], libc.tls_size);
_exit(127);
}
+ }
+ static_tls_cnt = tls_cnt;
+
+ /* The main program must be relocated LAST since it may contain
+ * copy relocations which depend on libraries' relocations. */
+ reloc_all(app.next);
+ reloc_all(&app);
+
+ /* Actual copying to new TLS needs to happen after relocations,
+ * since the TLS images might have contained relocated addresses. */
+ if (initial_tls != builtin_tls) {
if (__init_tp(__copy_tls(initial_tls)) < 0) {
a_crash();
}
if (__copy_tls((void*)builtin_tls) != self) a_crash();
libc.tls_size = tmp_tls_size;
}
- static_tls_cnt = tls_cnt;
-
- main_ctor_queue = queue_ctors(&app);
if (ldso_fail) _exit(127);
if (ldd_mode) _exit(0);
__inhibit_ptc();
p = 0;
+ if (shutting_down) {
+ error("Cannot dlopen while program is exiting.");
+ goto end;
+ }
orig_tls_tail = tls_tail;
orig_tls_cnt = tls_cnt;
orig_tls_offset = tls_offset;
free(p->deps);
unmap_library(p);
free(p);
- free(ctor_queue);
}
+ free(ctor_queue);
+ ctor_queue = 0;
if (!orig_tls_tail) libc.tls_head = 0;
tls_tail = orig_tls_tail;
if (tls_tail) tls_tail->next = 0;
static void *do_dlsym(struct dso *p, const char *s, void *ra)
{
- size_t i;
- uint32_t h = 0, gh = 0, *ght;
- Sym *sym;
- if (p == head || p == RTLD_DEFAULT || p == RTLD_NEXT) {
- if (p == RTLD_DEFAULT) {
- p = head;
- } else if (p == RTLD_NEXT) {
- p = addr2dso((size_t)ra);
- if (!p) p=head;
- p = p->next;
- }
- struct symdef def = find_sym(p, s, 0);
- if (!def.sym) goto failed;
- if ((def.sym->st_info&0xf) == STT_TLS)
- return __tls_get_addr((tls_mod_off_t []){def.dso->tls_id, def.sym->st_value-DTP_OFFSET});
- if (DL_FDPIC && (def.sym->st_info&0xf) == STT_FUNC)
- return def.dso->funcdescs + (def.sym - def.dso->syms);
- return laddr(def.dso, def.sym->st_value);
- }
- if (__dl_invalid_handle(p))
+ int use_deps = 0;
+ if (p == head || p == RTLD_DEFAULT) {
+ p = head;
+ } else if (p == RTLD_NEXT) {
+ p = addr2dso((size_t)ra);
+ if (!p) p=head;
+ p = p->next;
+ } else if (__dl_invalid_handle(p)) {
return 0;
- if ((ght = p->ghashtab)) {
- gh = gnu_hash(s);
- sym = gnu_lookup(gh, ght, p, s);
- } else {
- h = sysv_hash(s);
- sym = sysv_lookup(s, h, p);
- }
- if (sym && (sym->st_info&0xf) == STT_TLS)
- return __tls_get_addr((tls_mod_off_t []){p->tls_id, sym->st_value-DTP_OFFSET});
- if (DL_FDPIC && sym && sym->st_shndx && (sym->st_info&0xf) == STT_FUNC)
- return p->funcdescs + (sym - p->syms);
- if (sym && sym->st_value && (1<<(sym->st_info&0xf) & OK_TYPES))
- return laddr(p, sym->st_value);
- for (i=0; p->deps[i]; i++) {
- if ((ght = p->deps[i]->ghashtab)) {
- if (!gh) gh = gnu_hash(s);
- sym = gnu_lookup(gh, ght, p->deps[i], s);
- } else {
- if (!h) h = sysv_hash(s);
- sym = sysv_lookup(s, h, p->deps[i]);
- }
- if (sym && (sym->st_info&0xf) == STT_TLS)
- return __tls_get_addr((tls_mod_off_t []){p->deps[i]->tls_id, sym->st_value-DTP_OFFSET});
- if (DL_FDPIC && sym && sym->st_shndx && (sym->st_info&0xf) == STT_FUNC)
- return p->deps[i]->funcdescs + (sym - p->deps[i]->syms);
- if (sym && sym->st_value && (1<<(sym->st_info&0xf) & OK_TYPES))
- return laddr(p->deps[i], sym->st_value);
- }
-failed:
- error("Symbol not found: %s", s);
- return 0;
+ } else
+ use_deps = 1;
+ struct symdef def = find_sym2(p, s, 0, use_deps);
+ if (!def.sym) {
+ error("Symbol not found: %s", s);
+ return 0;
+ }
+ if ((def.sym->st_info&0xf) == STT_TLS)
+ return __tls_get_addr((tls_mod_off_t []){def.dso->tls_id, def.sym->st_value-DTP_OFFSET});
+ if (DL_FDPIC && (def.sym->st_info&0xf) == STT_FUNC)
+ return def.dso->funcdescs + (def.sym - def.dso->syms);
+ return laddr(def.dso, def.sym->st_value);
}
int dladdr(const void *addr_arg, Dl_info *info)
}
}
- if (bestsym && besterr > bestsym->st_size-1) {
+ if (best && besterr > bestsym->st_size-1) {
best = 0;
bestsym = 0;
}
projects / oweals / musl.git / blobdiff