X-Git-Url: https://git.librecmc.org/?a=blobdiff_plain;ds=sidebyside;f=ldso%2Fdynlink.c;h=a2f059cbc9941513435aa847200767cd2b8f9ada;hb=2f1f51ae7b2d78247568e7fdb8462f3c19e469a4;hp=34775b83417bf0ab114cb2cd1d40822007284dcb;hpb=403555690775f7c8806372644f543518e6664e3b;p=oweals%2Fmusl.git

diff --git a/ldso/dynlink.c b/ldso/dynlink.c
index 34775b83..a2f059cb 100644
--- a/ldso/dynlink.c
+++ b/ldso/dynlink.c
@@ -73,8 +73,11 @@ struct dso {
 	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;
@@ -121,13 +124,16 @@ static int runtime;
 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 **main_ctor_queue;
 static struct fdpic_loadmap *app_loadmap;
 static struct fdpic_dummy_loadmap app_dummy_loadmap;
 
@@ -1107,6 +1113,7 @@ static struct dso *load_library(const char *name, struct dso *needed_by)
 	p->ino = st.st_ino;
 	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;
@@ -1144,16 +1151,23 @@ static struct dso *load_library(const char *name, struct dso *needed_by)
 
 static void load_direct_deps(struct dso *p)
 {
-	size_t i, cnt;
+	size_t i, cnt=0;
 	if (p->deps) return;
-	for (i=cnt=0; p->dynv[i]; i+=2)
+	/* 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++;
 	p->deps = calloc(cnt+1, sizeof *p->deps);
 	if (!p->deps) {
 		error("Error loading dependencies for %s", p->name);
 		if (runtime) longjmp(*rtld_fail, 1);
 	}
-	for (i=cnt=0; p->dynv[i]; i+=2) {
+	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) {
@@ -1163,8 +1177,8 @@ static void load_direct_deps(struct dso *p)
 			continue;
 		}
 		p->deps[cnt++] = dep;
-		p->deps[cnt] = 0;
 	}
+	p->deps[cnt] = 0;
 	p->ndeps_direct = cnt;
 }
 
@@ -1180,6 +1194,10 @@ 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. */
+	int no_realloc = __malloc_replaced && !p->runtime_loaded;
+
 	if (p->bfs_built) return;
 	ndeps_all = p->ndeps_direct;
 
@@ -1195,12 +1213,18 @@ static void extend_bfs_deps(struct dso *p)
 		struct dso *dep = p->deps[i];
 		for (j=cnt=0; j<dep->ndeps_direct; j++)
 			if (!dep->deps[j]->mark) cnt++;
-		tmp = realloc(p->deps, sizeof(*p->deps) * (ndeps_all+cnt+1));
+		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;
@@ -1327,7 +1351,18 @@ void __libc_exit_fini()
 {
 	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)) {
@@ -1342,22 +1377,87 @@ void __libc_exit_fini()
 	}
 }
 
-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) {
+	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 */
+	stack = 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... */
+	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;
 		if (p->constructed) continue;
-		p->constructed = 1;
+		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])();
@@ -1367,17 +1467,20 @@ static void do_init_fini(struct dso *p)
 			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) free(main_ctor_queue);
+	main_ctor_queue = 0;
 }
 
 static void dl_debug_state(void)
@@ -1746,6 +1849,14 @@ _Noreturn void __dls3(size_t *sp)
 		}
 	}
 
+	/* 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);
@@ -1833,6 +1944,7 @@ void *dlopen(const char *file, int mode)
 	size_t i;
 	int cs;
 	jmp_buf jb;
+	struct dso **volatile ctor_queue = 0;
 
 	if (!file) return head;
 
@@ -1841,6 +1953,10 @@ void *dlopen(const char *file, int mode)
 	__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;
@@ -1867,6 +1983,7 @@ void *dlopen(const char *file, int mode)
 			free(p->deps);
 			unmap_library(p);
 			free(p);
+			free(ctor_queue);
 		}
 		if (!orig_tls_tail) libc.tls_head = 0;
 		tls_tail = orig_tls_tail;
@@ -1892,6 +2009,9 @@ void *dlopen(const char *file, int mode)
 	/* First load handling */
 	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++)
@@ -1929,7 +2049,10 @@ 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;
 }