#define _GNU_SOURCE
+#define SYSCALL_NO_TLS 1
#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#include <pthread.h>
#include <ctype.h>
#include <dlfcn.h>
+#include <semaphore.h>
+#include <sys/membarrier.h>
#include "pthread_impl.h"
#include "libc.h"
#include "dynlink.h"
+#include "malloc_impl.h"
static void error(const char *, ...);
#define MAXP2(a,b) (-(-(a)&-(b)))
#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;
void *head;
Phdr *phdr;
int phnum;
size_t phentsize;
- int refcnt;
Sym *syms;
Elf_Symndx *hashtab;
uint32_t *ghashtab;
char relocated;
char constructed;
char kernel_mapped;
+ char mark;
+ char bfs_built;
+ char runtime_loaded;
struct dso **deps, *needed_by;
+ size_t ndeps_direct;
+ size_t next_dep;
+ int ctor_visitor;
char *rpath_orig, *rpath;
struct tls_module tls;
size_t tls_id;
size_t relro_start, relro_end;
- void **new_dtv;
+ 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;
struct dso *dso;
};
-int __init_tp(void *);
-void __init_libc(char **, char *);
-void *__copy_tls(unsigned char *);
-
-__attribute__((__visibility__("hidden")))
-const char *__libc_get_version(void);
-
static struct builtin_tls {
char c;
struct pthread pt;
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 struct tls_module *tls_tail;
static size_t tls_cnt, tls_offset, tls_align = MIN_TLS_ALIGN;
static size_t static_tls_cnt;
-static pthread_mutex_t init_fini_lock = { ._m_type = PTHREAD_MUTEX_RECURSIVE };
+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;
struct debug *_dl_debug_addr = &debug;
-__attribute__((__visibility__("hidden")))
-void (*const __init_array_start)(void)=0, (*const __fini_array_start)(void)=0;
+extern hidden int __malloc_replaced;
-__attribute__((__visibility__("hidden")))
-extern void (*const __init_array_end)(void), (*const __fini_array_end)(void);
+hidden void (*const __init_array_start)(void)=0, (*const __fini_array_start)(void)=0;
+
+extern hidden void (*const __init_array_end)(void), (*const __fini_array_end)(void);
weak_alias(__init_array_start, __init_array_end);
weak_alias(__fini_array_start, __fini_array_end);
for (j=0; v-p->loadmap->segs[j].p_vaddr >= p->loadmap->segs[j].p_memsz; j++);
return (void *)(v - p->loadmap->segs[j].p_vaddr + p->loadmap->segs[j].addr);
}
+static void *laddr_pg(const struct dso *p, size_t v)
+{
+ size_t j=0;
+ size_t pgsz = PAGE_SIZE;
+ if (!p->loadmap) return p->base + v;
+ for (j=0; ; j++) {
+ size_t a = p->loadmap->segs[j].p_vaddr;
+ size_t b = a + p->loadmap->segs[j].p_memsz;
+ a &= -pgsz;
+ b += pgsz-1;
+ b &= -pgsz;
+ if (v-a<b-a) break;
+ }
+ return (void *)(v - p->loadmap->segs[j].p_vaddr + p->loadmap->segs[j].addr);
+}
#define fpaddr(p, v) ((void (*)())&(struct funcdesc){ \
laddr(p, v), (p)->got })
#else
#define laddr(p, v) (void *)((p)->base + (v))
+#define laddr_pg(p, v) laddr(p, v)
#define fpaddr(p, v) ((void (*)())laddr(p, v))
#endif
static struct symdef find_sym(struct dso *dso, const char *s, int need_def)
{
- uint32_t h = 0, gh, gho, *ght;
- size_t ghm = 0;
+ 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) {
Sym *sym;
if ((ght = dso->ghashtab)) {
- if (!ghm) {
- gh = gnu_hash(s);
- int maskbits = 8 * sizeof ghm;
- gho = gh / maskbits;
- ghm = 1ul << gh % maskbits;
- }
sym = gnu_lookup_filtered(gh, ght, dso, s, gho, ghm);
} else {
if (!h) h = sysv_hash(s);
return def;
}
-__attribute__((__visibility__("hidden")))
-ptrdiff_t __tlsdesc_static(), __tlsdesc_dynamic();
-
static void do_relocs(struct dso *dso, size_t *rel, size_t rel_size, size_t stride)
{
unsigned char *base = dso->base;
sym_val = def.sym ? (size_t)laddr(def.dso, def.sym->st_value) : 0;
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) {
+ error("Error relocating %s: %s: initial-exec TLS "
+ "resolves to dynamic definition in %s",
+ dso->name, name, def.dso->name);
+ longjmp(*rtld_fail, 1);
+ }
+
switch(type) {
case REL_NONE:
break;
#endif
case REL_TLSDESC:
if (stride<3) addend = reloc_addr[1];
- if (runtime && def.dso->tls_id >= static_tls_cnt) {
+ if (runtime && def.dso->tls_id > static_tls_cnt) {
struct td_index *new = malloc(sizeof *new);
if (!new) {
error(
new->next = dso->td_index;
dso->td_index = new;
new->args[0] = def.dso->tls_id;
- new->args[1] = tls_val + addend;
+ new->args[1] = tls_val + addend - DTP_OFFSET;
reloc_addr[0] = (size_t)__tlsdesc_dynamic;
reloc_addr[1] = (size_t)new;
} else {
+ addend;
#endif
}
+#ifdef TLSDESC_BACKWARDS
+ /* Some archs (32-bit ARM at least) invert the order of
+ * the descriptor members. Fix them up here. */
+ size_t tmp = reloc_addr[0];
+ reloc_addr[0] = reloc_addr[1];
+ reloc_addr[1] = tmp;
+#endif
break;
default:
error("Error relocating %s: unsupported relocation type %d",
/* A huge hack: to make up for the wastefulness of shared libraries
* needing at least a page of dirty memory even if they have no global
* data, we reclaim the gaps at the beginning and end of writable maps
- * and "donate" them to the heap by setting up minimal malloc
- * structures and then freeing them. */
+ * and "donate" them to the heap. */
static void reclaim(struct dso *dso, size_t start, size_t end)
{
- size_t *a, *z;
if (start >= dso->relro_start && start < dso->relro_end) start = dso->relro_end;
if (end >= dso->relro_start && end < dso->relro_end) end = dso->relro_start;
- start = start + 6*sizeof(size_t)-1 & -4*sizeof(size_t);
- end = (end & -4*sizeof(size_t)) - 2*sizeof(size_t);
- if (start>end || end-start < 4*sizeof(size_t)) return;
- a = laddr(dso, start);
- z = laddr(dso, end);
- a[-2] = 1;
- a[-1] = z[0] = end-start + 2*sizeof(size_t) | 1;
- z[1] = 1;
- free(a);
+ if (start >= end) return;
+ char *base = laddr_pg(dso, start);
+ __malloc_donate(base, base+(end-start));
}
static void reclaim_gaps(struct dso *dso)
Phdr *ph = dso->phdr;
size_t phcnt = dso->phnum;
- if (DL_FDPIC) return; // FIXME
for (; phcnt--; ph=(void *)((char *)ph+dso->phentsize)) {
if (ph->p_type!=PT_LOAD) continue;
if ((ph->p_flags&(PF_R|PF_W))!=(PF_R|PF_W)) continue;
} else if (ph->p_type == PT_GNU_RELRO) {
dso->relro_start = ph->p_vaddr & -PAGE_SIZE;
dso->relro_end = (ph->p_vaddr + ph->p_memsz) & -PAGE_SIZE;
+ } else if (ph->p_type == PT_GNU_STACK) {
+ if (!runtime && ph->p_memsz > __default_stacksize) {
+ __default_stacksize =
+ ph->p_memsz < DEFAULT_STACK_MAX ?
+ ph->p_memsz : DEFAULT_STACK_MAX;
+ }
}
if (ph->p_type != PT_LOAD) continue;
nsegs++;
dso->phnum = eh->e_phnum;
dso->phentsize = eh->e_phentsize;
}
- /* Reuse the existing mapping for the lowest-address LOAD */
- if ((ph->p_vaddr & -PAGE_SIZE) == addr_min && !DL_NOMMU_SUPPORT)
- continue;
this_min = ph->p_vaddr & -PAGE_SIZE;
this_max = ph->p_vaddr+ph->p_memsz+PAGE_SIZE-1 & -PAGE_SIZE;
off_start = ph->p_offset & -PAGE_SIZE;
prot = (((ph->p_flags&PF_R) ? PROT_READ : 0) |
((ph->p_flags&PF_W) ? PROT_WRITE: 0) |
((ph->p_flags&PF_X) ? PROT_EXEC : 0));
- if (mmap_fixed(base+this_min, this_max-this_min, prot, MAP_PRIVATE|MAP_FIXED, fd, off_start) == MAP_FAILED)
- goto error;
- if (ph->p_memsz > ph->p_filesz) {
+ /* Reuse the existing mapping for the lowest-address LOAD */
+ if ((ph->p_vaddr & -PAGE_SIZE) != addr_min || DL_NOMMU_SUPPORT)
+ if (mmap_fixed(base+this_min, this_max-this_min, prot, MAP_PRIVATE|MAP_FIXED, fd, off_start) == MAP_FAILED)
+ goto error;
+ if (ph->p_memsz > ph->p_filesz && (ph->p_flags&PF_W)) {
size_t brk = (size_t)base+ph->p_vaddr+ph->p_filesz;
size_t pgbrk = brk+PAGE_SIZE-1 & -PAGE_SIZE;
memset((void *)brk, 0, pgbrk-brk & PAGE_SIZE-1);
dso->base = base;
dso->dynv = laddr(dso, dyn);
if (dso->tls.size) dso->tls.image = laddr(dso, tls_image);
- if (!runtime) reclaim_gaps(dso);
free(allocated_buf);
return map;
noexec:
origin = p->name;
}
t = strrchr(origin, '/');
- l = t ? t-origin : 0;
+ if (t) {
+ l = t-origin;
+ } else {
+ /* Normally p->name will always be an absolute or relative
+ * pathname containing at least one '/' character, but in the
+ * case where ldso was invoked as a command to execute a
+ * program in the working directory, app.name may not. Fix. */
+ origin = ".";
+ l = 1;
+ }
+ /* Disallow non-absolute origins for suid/sgid/AT_SECURE. */
+ if (libc.secure && *origin != '/')
+ return 0;
p->rpath = malloc(strlen(p->rpath_orig) + n*l + 1);
if (!p->rpath) return -1;
#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)
/* Search for the name to see if it's already loaded */
for (p=head->next; p; p=p->next) {
if (p->shortname && !strcmp(p->shortname, name)) {
- p->refcnt++;
return p;
}
}
if (!p->shortname && pathname != name)
p->shortname = strrchr(p->name, '/')+1;
close(fd);
- p->refcnt++;
return p;
}
}
unmap_library(&temp_dso);
return load_library("libc.so", needed_by);
}
+ /* Past this point, if we haven't reached runtime yet, ldso has
+ * committed either to use the mapped library or to abort execution.
+ * Unmapping is not possible, so we can safely reclaim gaps. */
+ if (!runtime) reclaim_gaps(&temp_dso);
/* Allocate storage for the new DSO. When there is TLS, this
* storage must include a reservation for all pre-existing
memcpy(p, &temp_dso, sizeof temp_dso);
p->dev = st.st_dev;
p->ino = st.st_ino;
- p->refcnt = 1;
p->needed_by = needed_by;
p->name = p->buf;
+ p->runtime_loaded = runtime;
strcpy(p->name, pathname);
/* Add a shortname only if name arg was not an explicit pathname. */
if (pathname != name) p->shortname = strrchr(p->name, '/')+1;
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)
return p;
}
+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. */
+ if (p==head) for (struct dso *q=p->next; q; q=q->next)
+ cnt++;
+ for (i=0; p->dynv[i]; i+=2)
+ if (p->dynv[i] == DT_NEEDED) cnt++;
+ /* 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);
+ }
+ cnt=0;
+ if (p==head) for (struct dso *q=p->next; q; q=q->next)
+ p->deps[cnt++] = q;
+ for (i=0; p->dynv[i]; i+=2) {
+ if (p->dynv[i] != DT_NEEDED) continue;
+ struct dso *dep = load_library(p->strings + p->dynv[i+1], p);
+ if (!dep) {
+ error("Error loading shared library %s: %m (needed by %s)",
+ p->strings + p->dynv[i+1], p->name);
+ if (runtime) longjmp(*rtld_fail, 1);
+ continue;
+ }
+ p->deps[cnt++] = dep;
+ }
+ p->deps[cnt] = 0;
+ p->ndeps_direct = cnt;
+}
+
static void load_deps(struct dso *p)
{
- size_t i, ndeps=0;
- struct dso ***deps = &p->deps, **tmp, *dep;
- for (; p; p=p->next) {
- for (i=0; p->dynv[i]; i+=2) {
- if (p->dynv[i] != DT_NEEDED) continue;
- dep = load_library(p->strings + p->dynv[i+1], p);
- if (!dep) {
- error("Error loading shared library %s: %m (needed by %s)",
- p->strings + p->dynv[i+1], p->name);
- if (runtime) longjmp(*rtld_fail, 1);
- continue;
- }
- if (runtime) {
- tmp = realloc(*deps, sizeof(*tmp)*(ndeps+2));
- if (!tmp) longjmp(*rtld_fail, 1);
- tmp[ndeps++] = dep;
- tmp[ndeps] = 0;
- *deps = tmp;
- }
+ if (p->deps) return;
+ for (; p; p=p->next)
+ load_direct_deps(p);
+}
+
+static void extend_bfs_deps(struct dso *p)
+{
+ size_t i, j, cnt, ndeps_all;
+ struct dso **tmp;
+
+ /* Can't use realloc if the original p->deps was allocated at
+ * 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;
+
+ /* Mark existing (direct) deps so they won't be duplicated. */
+ for (i=0; p->deps[i]; i++)
+ p->deps[i]->mark = 1;
+
+ /* For each dependency already in the list, copy its list of direct
+ * dependencies to the list, excluding any items already in the
+ * list. Note that the list this loop iterates over will grow during
+ * the loop, but since duplicates are excluded, growth is bounded. */
+ for (i=0; p->deps[i]; i++) {
+ struct dso *dep = p->deps[i];
+ for (j=cnt=0; j<dep->ndeps_direct; j++)
+ if (!dep->deps[j]->mark) cnt++;
+ tmp = no_realloc ?
+ malloc(sizeof(*tmp) * (ndeps_all+cnt+1)) :
+ realloc(p->deps, sizeof(*tmp) * (ndeps_all+cnt+1));
+ if (!tmp) {
+ error("Error recording dependencies for %s", p->name);
+ if (runtime) longjmp(*rtld_fail, 1);
+ continue;
+ }
+ if (no_realloc) {
+ memcpy(tmp, p->deps, sizeof(*tmp) * (ndeps_all+1));
+ no_realloc = 0;
+ }
+ p->deps = tmp;
+ for (j=0; j<dep->ndeps_direct; j++) {
+ if (dep->deps[j]->mark) continue;
+ dep->deps[j]->mark = 1;
+ p->deps[ndeps_all++] = dep->deps[j];
}
+ p->deps[ndeps_all] = 0;
}
+ p->bfs_built = 1;
+ for (p=head; p; p=p->next)
+ p->mark = 0;
}
static void load_preload(char *s)
} else if (ph->p_type == PT_GNU_RELRO) {
p->relro_start = ph->p_vaddr & -PAGE_SIZE;
p->relro_end = (ph->p_vaddr + ph->p_memsz) & -PAGE_SIZE;
+ } else if (ph->p_type == PT_GNU_STACK) {
+ if (!runtime && ph->p_memsz > __default_stacksize) {
+ __default_stacksize =
+ ph->p_memsz < DEFAULT_STACK_MAX ?
+ ph->p_memsz : DEFAULT_STACK_MAX;
+ }
}
if (ph->p_type != PT_LOAD) continue;
if (ph->p_vaddr < min_addr)
{
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)) {
}
}
-static void do_init_fini(struct dso *p)
+static struct dso **queue_ctors(struct dso *dso)
{
- size_t dyn[DYN_CNT];
- int need_locking = libc.threads_minus_1;
- /* Allow recursive calls that arise when a library calls
- * dlopen from one of its constructors, but block any
- * other threads until all ctors have finished. */
- if (need_locking) pthread_mutex_lock(&init_fini_lock);
- for (; p; p=p->prev) {
- if (p->constructed) continue;
- p->constructed = 1;
+ size_t cnt, qpos, spos, i;
+ struct dso *p, **queue, **stack;
+
+ if (ldd_mode) return 0;
+
+ /* Bound on queue size is the total number of indirect deps.
+ * If a bfs deps list was built, we can use it. Otherwise,
+ * bound by the total number of DSOs, which is always safe and
+ * is reasonable we use it (for main app at startup). */
+ if (dso->bfs_built) {
+ for (cnt=0; dso->deps[cnt]; cnt++)
+ dso->deps[cnt]->mark = 0;
+ cnt++; /* self, not included in deps */
+ } else {
+ for (cnt=0, p=head; p; cnt++, p=p->next)
+ p->mark = 0;
+ }
+ cnt++; /* termination slot */
+ 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");
+ if (runtime) longjmp(*rtld_fail, 1);
+ return 0;
+ }
+
+ /* 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;
+ dso->next_dep = 0;
+ dso->mark = 1;
+
+ /* Then perform pseudo-DFS sort, but ignoring circular deps. */
+ while (spos<cnt) {
+ p = stack[spos++];
+ while (p->next_dep < p->ndeps_direct) {
+ if (p->deps[p->next_dep]->mark) {
+ p->next_dep++;
+ } else {
+ stack[--spos] = p;
+ p = p->deps[p->next_dep];
+ p->next_dep = 0;
+ p->mark = 1;
+ }
+ }
+ queue[qpos++] = p;
+ }
+ queue[qpos] = 0;
+ for (i=0; i<qpos; i++) queue[i]->mark = 0;
+
+ return queue;
+}
+
+static void do_init_fini(struct dso **queue)
+{
+ struct dso *p;
+ size_t dyn[DYN_CNT], i;
+ int self = __pthread_self()->tid;
+
+ pthread_mutex_lock(&init_fini_lock);
+ for (i=0; (p=queue[i]); i++) {
+ 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;
+ p->ctor_visitor = self;
+
decode_vec(p->dynv, dyn, DYN_CNT);
if (dyn[0] & ((1<<DT_FINI) | (1<<DT_FINI_ARRAY))) {
p->fini_next = fini_head;
fini_head = p;
}
+
+ pthread_mutex_unlock(&init_fini_lock);
+
#ifndef NO_LEGACY_INITFINI
if ((dyn[0] & (1<<DT_INIT)) && dyn[DT_INIT])
fpaddr(p, dyn[DT_INIT])();
size_t *fn = laddr(p, dyn[DT_INIT_ARRAY]);
while (n--) ((void (*)(void))*fn++)();
}
- if (!need_locking && libc.threads_minus_1) {
- need_locking = 1;
- pthread_mutex_lock(&init_fini_lock);
- }
+
+ pthread_mutex_lock(&init_fini_lock);
+ p->ctor_visitor = 0;
+ p->constructed = 1;
+ pthread_cond_broadcast(&ctor_cond);
}
- if (need_locking) pthread_mutex_unlock(&init_fini_lock);
+ pthread_mutex_unlock(&init_fini_lock);
}
void __libc_start_init(void)
{
- do_init_fini(tail);
+ do_init_fini(main_ctor_queue);
+ if (!__malloc_replaced && main_ctor_queue != builtin_ctor_queue)
+ free(main_ctor_queue);
+ main_ctor_queue = 0;
}
static void dl_debug_state(void)
{
}
-__attribute__((__visibility__("hidden")))
-void *__tls_get_new(tls_mod_off_t *v)
-{
- pthread_t self = __pthread_self();
-
- /* Block signals to make accessing new TLS async-signal-safe */
- sigset_t set;
- __block_all_sigs(&set);
- if (v[0]<=(size_t)self->dtv[0]) {
- __restore_sigs(&set);
- return (char *)self->dtv[v[0]]+v[1]+DTP_OFFSET;
- }
-
- /* This is safe without any locks held because, if the caller
- * is able to request the Nth entry of the DTV, the DSO list
- * must be valid at least that far out and it was synchronized
- * at program startup or by an already-completed call to dlopen. */
- struct dso *p;
- for (p=head; p->tls_id != v[0]; p=p->next);
-
- /* Get new DTV space from new DSO if needed */
- if (v[0] > (size_t)self->dtv[0]) {
- void **newdtv = p->new_dtv +
- (v[0]+1)*a_fetch_add(&p->new_dtv_idx,1);
- memcpy(newdtv, self->dtv,
- ((size_t)self->dtv[0]+1) * sizeof(void *));
- newdtv[0] = (void *)v[0];
- self->dtv = self->dtv_copy = newdtv;
- }
-
- /* Get new TLS memory from all new DSOs up to the requested one */
- unsigned char *mem;
- for (p=head; ; p=p->next) {
- if (!p->tls_id || self->dtv[p->tls_id]) continue;
- mem = p->new_tls + (p->tls.size + p->tls.align)
- * a_fetch_add(&p->new_tls_idx,1);
- mem += ((uintptr_t)p->tls.image - (uintptr_t)mem)
- & (p->tls.align-1);
- self->dtv[p->tls_id] = mem;
- memcpy(mem, p->tls.image, p->tls.len);
- if (p->tls_id == v[0]) break;
- }
- __restore_sigs(&set);
- return mem + v[1] + DTP_OFFSET;
-}
-
static void update_tls_size()
{
libc.tls_cnt = tls_cnt;
tls_align);
}
+static void install_new_tls(void)
+{
+ sigset_t set;
+ pthread_t self = __pthread_self(), td;
+ struct dso *dtv_provider = container_of(tls_tail, struct dso, tls);
+ uintptr_t (*newdtv)[tls_cnt+1] = (void *)dtv_provider->new_dtv;
+ struct dso *p;
+ size_t i, j;
+ size_t old_cnt = self->dtv[0];
+
+ __block_app_sigs(&set);
+ __tl_lock();
+ /* Copy existing dtv contents from all existing threads. */
+ for (i=0, td=self; !i || td!=self; i++, td=td->next) {
+ memcpy(newdtv+i, td->dtv,
+ (old_cnt+1)*sizeof(uintptr_t));
+ newdtv[i][0] = tls_cnt;
+ }
+ /* Install new dtls into the enlarged, uninstalled dtv copies. */
+ for (p=head; ; p=p->next) {
+ if (p->tls_id <= old_cnt) continue;
+ unsigned char *mem = p->new_tls;
+ for (j=0; j<i; j++) {
+ unsigned char *new = mem;
+ new += ((uintptr_t)p->tls.image - (uintptr_t)mem)
+ & (p->tls.align-1);
+ memcpy(new, p->tls.image, p->tls.len);
+ newdtv[j][p->tls_id] =
+ (uintptr_t)new + DTP_OFFSET;
+ mem += p->tls.size + p->tls.align;
+ }
+ if (p->tls_id == tls_cnt) break;
+ }
+
+ /* Broadcast barrier to ensure contents of new dtv is visible
+ * if the new dtv pointer is. The __membarrier function has a
+ * fallback emulation using signals for kernels that lack the
+ * feature at the syscall level. */
+
+ __membarrier(MEMBARRIER_CMD_PRIVATE_EXPEDITED, 0);
+
+ /* Install new dtv for each thread. */
+ for (j=0, td=self; !j || td!=self; j++, td=td->next) {
+ td->dtv = td->dtv_copy = newdtv[j];
+ }
+
+ __tl_unlock();
+ __restore_sigs(&set);
+}
+
/* Stage 1 of the dynamic linker is defined in dlstart.c. It calls the
* following stage 2 and stage 3 functions via primitive symbolic lookup
* since it does not have access to their addresses to begin with. */
* linker itself, but some of the relocations performed may need to be
* replaced later due to copy relocations in the main program. */
-__attribute__((__visibility__("hidden")))
-void __dls2(unsigned char *base, size_t *sp)
+hidden void __dls2(unsigned char *base, size_t *sp)
{
if (DL_FDPIC) {
void *p1 = (void *)sp[-2];
void *p2 = (void *)sp[-1];
if (!p1) {
size_t *auxv, aux[AUX_CNT];
- for (auxv=sp+1+*sp+1; *auxv; auxv++); auxv++;
+ for (auxv=sp+1+*sp+1; *auxv; auxv++);
+ auxv++;
decode_vec(auxv, aux, AUX_CNT);
if (aux[AT_BASE]) ldso.base = (void *)aux[AT_BASE];
else ldso.base = (void *)(aux[AT_PHDR] & -4096);
ldso.relocated = 0;
- /* Call dynamic linker stage-3, __dls3, looking it up
+ /* Call dynamic linker stage-2b, __dls2b, looking it up
* symbolically as a barrier against moving the address
* load across the above relocation processing. */
+ struct symdef dls2b_def = find_sym(&ldso, "__dls2b", 0);
+ if (DL_FDPIC) ((stage3_func)&ldso.funcdescs[dls2b_def.sym-ldso.syms])(sp);
+ else ((stage3_func)laddr(&ldso, dls2b_def.sym->st_value))(sp);
+}
+
+/* Stage 2b sets up a valid thread pointer, which requires relocations
+ * completed in stage 2, and on which stage 3 is permitted to depend.
+ * This is done as a separate stage, with symbolic lookup as a barrier,
+ * so that loads of the thread pointer and &errno can be pure/const and
+ * thereby hoistable. */
+
+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
+ * thread pointer at runtime. */
+ libc.tls_size = sizeof builtin_tls;
+ libc.tls_align = tls_align;
+ if (__init_tp(__copy_tls((void *)builtin_tls)) < 0) {
+ a_crash();
+ }
+
struct symdef dls3_def = find_sym(&ldso, "__dls3", 0);
if (DL_FDPIC) ((stage3_func)&ldso.funcdescs[dls3_def.sym-ldso.syms])(sp);
else ((stage3_func)laddr(&ldso, dls3_def.sym->st_value))(sp);
* 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;
size_t i;
char *env_preload=0;
+ char *replace_argv0=0;
size_t vdso_base;
int argc = *sp;
char **argv = (void *)(sp+1);
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]);
- /* Setup early thread pointer in builtin_tls for ldso/libc itself to
- * use during dynamic linking. If possible it will also serve as the
- * thread pointer at runtime. */
- libc.tls_size = sizeof builtin_tls;
- libc.tls_align = tls_align;
- if (__init_tp(__copy_tls((void *)builtin_tls)) < 0) {
- a_crash();
- }
-
/* Only trust user/env if kernel says we're not suid/sgid */
if (!libc.secure) {
env_path = getenv("LD_LIBRARY_PATH");
if (opt[7]=='=') env_preload = opt+8;
else if (opt[7]) *argv = 0;
else if (*argv) env_preload = *argv++;
+ } else if (!memcmp(opt, "argv0", 5)) {
+ if (opt[5]=='=') replace_argv0 = opt+6;
+ else if (opt[5]) *argv = 0;
+ else if (*argv) replace_argv0 = *argv++;
} else {
argv[0] = 0;
}
"Version %s\n"
"Dynamic Program Loader\n"
"Usage: %s [options] [--] pathname%s\n",
- __libc_get_version(), ldname,
+ __libc_version, ldname,
ldd_mode ? "" : " [args]");
_exit(1);
}
dprintf(2, "%s: cannot load %s: %s\n", ldname, argv[0], strerror(errno));
_exit(1);
}
- runtime = 1;
Ehdr *ehdr = (void *)map_library(fd, &app);
if (!ehdr) {
dprintf(2, "%s: %s: Not a valid dynamic program\n", ldname, argv[0]);
_exit(1);
}
- runtime = 0;
close(fd);
ldso.name = ldname;
app.name = argv[0];
libc.tls_head = tls_tail = &app.tls;
app.tls_id = tls_cnt = 1;
#ifdef TLS_ABOVE_TP
- app.tls.offset = 0;
- tls_offset = app.tls.size
- + ( -((uintptr_t)app.tls.image + app.tls.size)
- & (app.tls.align-1) );
+ app.tls.offset = GAP_ABOVE_TP;
+ 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;
}
}
+ /* 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);
+
/* The main program must be relocated LAST since it may contin
* copy relocations which depend on libraries' relocations. */
reloc_all(app.next);
if (ldso_fail) _exit(127);
if (ldd_mode) _exit(0);
+ /* Determine if malloc was interposed by a replacement implementation
+ * so that calloc and the memalign family can harden against the
+ * possibility of incomplete replacement. */
+ if (find_sym(head, "malloc", 1).dso != &ldso)
+ __malloc_replaced = 1;
+
/* Switch to runtime mode: any further failures in the dynamic
* linker are a reportable failure rather than a fatal startup
* error. */
debug.state = 0;
_dl_debug_state();
+ if (replace_argv0) argv[0] = replace_argv0;
+
errno = 0;
CRTJMP((void *)aux[AT_ENTRY], argv-1);
size_t i;
int cs;
jmp_buf jb;
+ struct dso **volatile ctor_queue = 0;
if (!file) return head;
__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;
unmap_library(p);
free(p);
}
+ 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;
}
/* First load handling */
- if (!p->relocated) {
- load_deps(p);
- if ((mode & RTLD_LAZY)) {
- prepare_lazy(p);
- if (p->deps) for (i=0; p->deps[i]; i++)
- if (!p->deps[i]->relocated)
- prepare_lazy(p->deps[i]);
- }
+ load_deps(p);
+ extend_bfs_deps(p);
+ pthread_mutex_lock(&init_fini_lock);
+ if (!p->constructed) ctor_queue = queue_ctors(p);
+ pthread_mutex_unlock(&init_fini_lock);
+ if (!p->relocated && (mode & RTLD_LAZY)) {
+ prepare_lazy(p);
+ for (i=0; p->deps[i]; i++)
+ if (!p->deps[i]->relocated)
+ prepare_lazy(p->deps[i]);
+ }
+ if (!p->relocated || (mode & RTLD_GLOBAL)) {
/* Make new symbols global, at least temporarily, so we can do
* relocations. If not RTLD_GLOBAL, this is reverted below. */
add_syms(p);
- if (p->deps) for (i=0; p->deps[i]; i++)
+ for (i=0; p->deps[i]; i++)
add_syms(p->deps[i]);
+ }
+ if (!p->relocated) {
reloc_all(p);
}
redo_lazy_relocs();
update_tls_size();
+ if (tls_cnt != orig_tls_cnt)
+ install_new_tls();
_dl_debug_state();
orig_tail = tail;
end:
__release_ptc();
if (p) gencnt++;
pthread_rwlock_unlock(&lock);
- if (p) do_init_fini(orig_tail);
+ if (ctor_queue) {
+ do_init_fini(ctor_queue);
+ free(ctor_queue);
+ }
pthread_setcancelstate(cs, 0);
return p;
}
-__attribute__((__visibility__("hidden")))
-int __dl_invalid_handle(void *h)
+hidden int __dl_invalid_handle(void *h)
{
struct dso *p;
for (p=head; p; p=p->next) if (h==p) return 0;
return p;
}
} else {
+ Phdr *ph = p->phdr;
+ size_t phcnt = p->phnum;
+ size_t entsz = p->phentsize;
+ size_t base = (size_t)p->base;
+ for (; phcnt--; ph=(void *)((char *)ph+entsz)) {
+ if (ph->p_type != PT_LOAD) continue;
+ if (a-base-ph->p_vaddr < ph->p_memsz)
+ return p;
+ }
if (a-(size_t)p->map < p->map_len)
- return p;
+ return 0;
}
}
return 0;
}
-void *__tls_get_addr(tls_mod_off_t *);
-
static void *do_dlsym(struct dso *p, const char *s, void *ra)
{
size_t i;
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});
+ 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);
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});
+ 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);
- if (p->deps) for (i=0; p->deps[i]; i++) {
+ 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);
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});
+ 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 0;
}
-int dladdr(const void *addr, Dl_info *info)
+int dladdr(const void *addr_arg, Dl_info *info)
{
+ size_t addr = (size_t)addr_arg;
struct dso *p;
Sym *sym, *bestsym;
uint32_t nsym;
char *strings;
- void *best = 0;
+ size_t best = 0;
+ size_t besterr = -1;
pthread_rwlock_rdlock(&lock);
- p = addr2dso((size_t)addr);
+ p = addr2dso(addr);
pthread_rwlock_unlock(&lock);
if (!p) return 0;
nsym = count_syms(p);
if (DL_FDPIC) {
- size_t idx = ((size_t)addr-(size_t)p->funcdescs)
+ size_t idx = (addr-(size_t)p->funcdescs)
/ sizeof(*p->funcdescs);
if (idx < nsym && (sym[idx].st_info&0xf) == STT_FUNC) {
- best = p->funcdescs + idx;
+ best = (size_t)(p->funcdescs + idx);
bestsym = sym + idx;
+ besterr = 0;
}
}
if (sym->st_value
&& (1<<(sym->st_info&0xf) & OK_TYPES)
&& (1<<(sym->st_info>>4) & OK_BINDS)) {
- void *symaddr = laddr(p, sym->st_value);
- if (symaddr > addr || symaddr < best)
+ size_t symaddr = (size_t)laddr(p, sym->st_value);
+ if (symaddr > addr || symaddr <= best)
continue;
best = symaddr;
bestsym = sym;
+ besterr = addr - symaddr;
if (addr == symaddr)
break;
}
}
- if (!best) return 0;
-
- if (DL_FDPIC && (bestsym->st_info&0xf) == STT_FUNC)
- best = p->funcdescs + (bestsym - p->syms);
+ if (best && besterr > bestsym->st_size-1) {
+ best = 0;
+ bestsym = 0;
+ }
info->dli_fname = p->name;
- info->dli_fbase = p->base;
+ info->dli_fbase = p->map;
+
+ if (!best) {
+ info->dli_sname = 0;
+ info->dli_saddr = 0;
+ return 1;
+ }
+
+ if (DL_FDPIC && (bestsym->st_info&0xf) == STT_FUNC)
+ best = (size_t)(p->funcdescs + (bestsym - p->syms));
info->dli_sname = strings + bestsym->st_name;
- info->dli_saddr = best;
+ info->dli_saddr = (void *)best;
return 1;
}
-__attribute__((__visibility__("hidden")))
-void *__dlsym(void *restrict p, const char *restrict s, void *restrict ra)
+hidden void *__dlsym(void *restrict p, const char *restrict s, void *restrict ra)
{
void *res;
pthread_rwlock_rdlock(&lock);
return ret;
}
-__attribute__((__visibility__("hidden")))
-void __dl_vseterr(const char *, va_list);
-
static void error(const char *fmt, ...)
{
va_list ap;
OSDN Git Service
