Merge tag 'modules-next-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux

Pull module updates from Rusty Russell:
 "Main excitement here is Peter Zijlstra's lockless rbtree optimization
  to speed module address lookup.  He found some abusers of the module
  lock doing that too.

  A little bit of parameter work here too; including Dan Streetman's
  breaking up the big param mutex so writing a parameter can load
  another module (yeah, really).  Unfortunately that broke the usual
  suspects, !CONFIG_MODULES and !CONFIG_SYSFS, so those fixes were
  appended too"

* tag 'modules-next-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux: (26 commits)
  modules: only use mod->param_lock if CONFIG_MODULES
  param: fix module param locks when !CONFIG_SYSFS.
  rcu: merge fix for Convert ACCESS_ONCE() to READ_ONCE() and WRITE_ONCE()
  module: add per-module param_lock
  module: make perm const
  params: suppress unused variable error, warn once just in case code changes.
  modules: clarify CONFIG_MODULE_COMPRESS help, suggest 'N'.
  kernel/module.c: avoid ifdefs for sig_enforce declaration
  kernel/workqueue.c: remove ifdefs over wq_power_efficient
  kernel/params.c: export param_ops_bool_enable_only
  kernel/params.c: generalize bool_enable_only
  kernel/module.c: use generic module param operaters for sig_enforce
  kernel/params: constify struct kernel_param_ops uses
  sysfs: tightened sysfs permission checks
  module: Rework module_addr_{min,max}
  module: Use __module_address() for module_address_lookup()
  module: Make the mod_tree stuff conditional on PERF_EVENTS || TRACING
  module: Optimize __module_address() using a latched RB-tree
  rbtree: Implement generic latch_tree
  seqlock: Introduce raw_read_seqcount_latch()
  ...

1 files changed, 233 insertions, 76 deletions

diff --git a/kernel/module.c b/kernel/module.c
index f80a97f..3e0e197 100644
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -101,48 +101,201 @@
 DEFINE_MUTEX(module_mutex);
 EXPORT_SYMBOL_GPL(module_mutex);
 static LIST_HEAD(modules);
-#ifdef CONFIG_KGDB_KDB
-struct list_head *kdb_modules = &modules; /* kdb needs the list of modules */
-#endif /* CONFIG_KGDB_KDB */
 
-#ifdef CONFIG_MODULE_SIG
-#ifdef CONFIG_MODULE_SIG_FORCE
-static bool sig_enforce = true;
-#else
-static bool sig_enforce = false;
+#ifdef CONFIG_MODULES_TREE_LOOKUP
+
+/*
+ * Use a latched RB-tree for __module_address(); this allows us to use
+ * RCU-sched lookups of the address from any context.
+ *
+ * Because modules have two address ranges: init and core, we need two
+ * latch_tree_nodes entries. Therefore we need the back-pointer from
+ * mod_tree_node.
+ *
+ * Because init ranges are short lived we mark them unlikely and have placed
+ * them outside the critical cacheline in struct module.
+ *
+ * This is conditional on PERF_EVENTS || TRACING because those can really hit
+ * __module_address() hard by doing a lot of stack unwinding; potentially from
+ * NMI context.
+ */
 
-static int param_set_bool_enable_only(const char *val,
-				      const struct kernel_param *kp)
+static __always_inline unsigned long __mod_tree_val(struct latch_tree_node *n)
 {
-	int err;
-	bool test;
-	struct kernel_param dummy_kp = *kp;
+	struct mod_tree_node *mtn = container_of(n, struct mod_tree_node, node);
+	struct module *mod = mtn->mod;
 
-	dummy_kp.arg = &test;
+	if (unlikely(mtn == &mod->mtn_init))
+		return (unsigned long)mod->module_init;
 
-	err = param_set_bool(val, &dummy_kp);
-	if (err)
-		return err;
+	return (unsigned long)mod->module_core;
+}
+
+static __always_inline unsigned long __mod_tree_size(struct latch_tree_node *n)
+{
+	struct mod_tree_node *mtn = container_of(n, struct mod_tree_node, node);
+	struct module *mod = mtn->mod;
 
-	/* Don't let them unset it once it's set! */
-	if (!test && sig_enforce)
-		return -EROFS;
+	if (unlikely(mtn == &mod->mtn_init))
+		return (unsigned long)mod->init_size;
+
+	return (unsigned long)mod->core_size;
+}
+
+static __always_inline bool
+mod_tree_less(struct latch_tree_node *a, struct latch_tree_node *b)
+{
+	return __mod_tree_val(a) < __mod_tree_val(b);
+}
+
+static __always_inline int
+mod_tree_comp(void *key, struct latch_tree_node *n)
+{
+	unsigned long val = (unsigned long)key;
+	unsigned long start, end;
+
+	start = __mod_tree_val(n);
+	if (val < start)
+		return -1;
+
+	end = start + __mod_tree_size(n);
+	if (val >= end)
+		return 1;
 
-	if (test)
-		sig_enforce = true;
 	return 0;
 }
 
-static const struct kernel_param_ops param_ops_bool_enable_only = {
-	.flags = KERNEL_PARAM_OPS_FL_NOARG,
-	.set = param_set_bool_enable_only,
-	.get = param_get_bool,
+static const struct latch_tree_ops mod_tree_ops = {
+	.less = mod_tree_less,
+	.comp = mod_tree_comp,
 };
-#define param_check_bool_enable_only param_check_bool
 
+static struct mod_tree_root {
+	struct latch_tree_root root;
+	unsigned long addr_min;
+	unsigned long addr_max;
+} mod_tree __cacheline_aligned = {
+	.addr_min = -1UL,
+};
+
+#define module_addr_min mod_tree.addr_min
+#define module_addr_max mod_tree.addr_max
+
+static noinline void __mod_tree_insert(struct mod_tree_node *node)
+{
+	latch_tree_insert(&node->node, &mod_tree.root, &mod_tree_ops);
+}
+
+static void __mod_tree_remove(struct mod_tree_node *node)
+{
+	latch_tree_erase(&node->node, &mod_tree.root, &mod_tree_ops);
+}
+
+/*
+ * These modifications: insert, remove_init and remove; are serialized by the
+ * module_mutex.
+ */
+static void mod_tree_insert(struct module *mod)
+{
+	mod->mtn_core.mod = mod;
+	mod->mtn_init.mod = mod;
+
+	__mod_tree_insert(&mod->mtn_core);
+	if (mod->init_size)
+		__mod_tree_insert(&mod->mtn_init);
+}
+
+static void mod_tree_remove_init(struct module *mod)
+{
+	if (mod->init_size)
+		__mod_tree_remove(&mod->mtn_init);
+}
+
+static void mod_tree_remove(struct module *mod)
+{
+	__mod_tree_remove(&mod->mtn_core);
+	mod_tree_remove_init(mod);
+}
+
+static struct module *mod_find(unsigned long addr)
+{
+	struct latch_tree_node *ltn;
+
+	ltn = latch_tree_find((void *)addr, &mod_tree.root, &mod_tree_ops);
+	if (!ltn)
+		return NULL;
+
+	return container_of(ltn, struct mod_tree_node, node)->mod;
+}
+
+#else /* MODULES_TREE_LOOKUP */
+
+static unsigned long module_addr_min = -1UL, module_addr_max = 0;
+
+static void mod_tree_insert(struct module *mod) { }
+static void mod_tree_remove_init(struct module *mod) { }
+static void mod_tree_remove(struct module *mod) { }
+
+static struct module *mod_find(unsigned long addr)
+{
+	struct module *mod;
+
+	list_for_each_entry_rcu(mod, &modules, list) {
+		if (within_module(addr, mod))
+			return mod;
+	}
+
+	return NULL;
+}
+
+#endif /* MODULES_TREE_LOOKUP */
+
+/*
+ * Bounds of module text, for speeding up __module_address.
+ * Protected by module_mutex.
+ */
+static void __mod_update_bounds(void *base, unsigned int size)
+{
+	unsigned long min = (unsigned long)base;
+	unsigned long max = min + size;
+
+	if (min < module_addr_min)
+		module_addr_min = min;
+	if (max > module_addr_max)
+		module_addr_max = max;
+}
+
+static void mod_update_bounds(struct module *mod)
+{
+	__mod_update_bounds(mod->module_core, mod->core_size);
+	if (mod->init_size)
+		__mod_update_bounds(mod->module_init, mod->init_size);
+}
+
+#ifdef CONFIG_KGDB_KDB
+struct list_head *kdb_modules = &modules; /* kdb needs the list of modules */
+#endif /* CONFIG_KGDB_KDB */
+
+static void module_assert_mutex(void)
+{
+	lockdep_assert_held(&module_mutex);
+}
+
+static void module_assert_mutex_or_preempt(void)
+{
+#ifdef CONFIG_LOCKDEP
+	if (unlikely(!debug_locks))
+		return;
+
+	WARN_ON(!rcu_read_lock_sched_held() &&
+		!lockdep_is_held(&module_mutex));
+#endif
+}
+
+static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE);
+#ifndef CONFIG_MODULE_SIG_FORCE
 module_param(sig_enforce, bool_enable_only, 0644);
 #endif /* !CONFIG_MODULE_SIG_FORCE */
-#endif /* CONFIG_MODULE_SIG */
 
 /* Block module loading/unloading? */
 int modules_disabled = 0;
@@ -153,10 +306,6 @@ static DECLARE_WAIT_QUEUE_HEAD(module_wq);
 
 static BLOCKING_NOTIFIER_HEAD(module_notify_list);
 
-/* Bounds of module allocation, for speeding __module_address.
- * Protected by module_mutex. */
-static unsigned long module_addr_min = -1UL, module_addr_max = 0;
-
 int register_module_notifier(struct notifier_block *nb)
 {
 	return blocking_notifier_chain_register(&module_notify_list, nb);
@@ -318,6 +467,8 @@ bool each_symbol_section(bool (*fn)(const struct symsearch *arr,
 #endif
 	};
 
+	module_assert_mutex_or_preempt();
+
 	if (each_symbol_in_section(arr, ARRAY_SIZE(arr), NULL, fn, data))
 		return true;
 
@@ -457,6 +608,8 @@ static struct module *find_module_all(const char *name, size_t len,
 {
 	struct module *mod;
 
+	module_assert_mutex();
+
 	list_for_each_entry(mod, &modules, list) {
 		if (!even_unformed && mod->state == MODULE_STATE_UNFORMED)
 			continue;
@@ -1169,11 +1322,17 @@ static inline int check_modstruct_version(Elf_Shdr *sechdrs,
 {
 	const unsigned long *crc;
 
-	/* Since this should be found in kernel (which can't be removed),
-	 * no locking is necessary. */
+	/*
+	 * Since this should be found in kernel (which can't be removed), no
+	 * locking is necessary -- use preempt_disable() to placate lockdep.
+	 */
+	preempt_disable();
 	if (!find_symbol(VMLINUX_SYMBOL_STR(module_layout), NULL,
-			 &crc, true, false))
+			 &crc, true, false)) {
+		preempt_enable();
 		BUG();
+	}
+	preempt_enable();
 	return check_version(sechdrs, versindex,
 			     VMLINUX_SYMBOL_STR(module_layout), mod, crc,
 			     NULL);
@@ -1661,6 +1820,10 @@ static void mod_sysfs_fini(struct module *mod)
 	mod_kobject_put(mod);
 }
 
+static void init_param_lock(struct module *mod)
+{
+	mutex_init(&mod->param_lock);
+}
 #else /* !CONFIG_SYSFS */
 
 static int mod_sysfs_setup(struct module *mod,
@@ -1683,6 +1846,9 @@ static void del_usage_links(struct module *mod)
 {
 }
 
+static void init_param_lock(struct module *mod)
+{
+}
 #endif /* CONFIG_SYSFS */
 
 static void mod_sysfs_teardown(struct module *mod)
@@ -1852,10 +2018,11 @@ static void free_module(struct module *mod)
 	mutex_lock(&module_mutex);
 	/* Unlink carefully: kallsyms could be walking list. */
 	list_del_rcu(&mod->list);
+	mod_tree_remove(mod);
 	/* Remove this module from bug list, this uses list_del_rcu */
 	module_bug_cleanup(mod);
-	/* Wait for RCU synchronizing before releasing mod->list and buglist. */
-	synchronize_rcu();
+	/* Wait for RCU-sched synchronizing before releasing mod->list and buglist. */
+	synchronize_sched();
 	mutex_unlock(&module_mutex);
 
 	/* This may be NULL, but that's OK */
@@ -2384,22 +2551,6 @@ void * __weak module_alloc(unsigned long size)
 	return vmalloc_exec(size);
 }
 
-static void *module_alloc_update_bounds(unsigned long size)
-{
-	void *ret = module_alloc(size);
-
-	if (ret) {
-		mutex_lock(&module_mutex);
-		/* Update module bounds. */
-		if ((unsigned long)ret < module_addr_min)
-			module_addr_min = (unsigned long)ret;
-		if ((unsigned long)ret + size > module_addr_max)
-			module_addr_max = (unsigned long)ret + size;
-		mutex_unlock(&module_mutex);
-	}
-	return ret;
-}
-
 #ifdef CONFIG_DEBUG_KMEMLEAK
 static void kmemleak_load_module(const struct module *mod,
 				 const struct load_info *info)
@@ -2805,7 +2956,7 @@ static int move_module(struct module *mod, struct load_info *info)
 	void *ptr;
 
 	/* Do the allocs. */
-	ptr = module_alloc_update_bounds(mod->core_size);
+	ptr = module_alloc(mod->core_size);
 	/*
 	 * The pointer to this block is stored in the module structure
 	 * which is inside the block. Just mark it as not being a
@@ -2819,7 +2970,7 @@ static int move_module(struct module *mod, struct load_info *info)
 	mod->module_core = ptr;
 
 	if (mod->init_size) {
-		ptr = module_alloc_update_bounds(mod->init_size);
+		ptr = module_alloc(mod->init_size);
 		/*
 		 * The pointer to this block is stored in the module structure
 		 * which is inside the block. This block doesn't need to be
@@ -3119,6 +3270,7 @@ static noinline int do_init_module(struct module *mod)
 	mod->symtab = mod->core_symtab;
 	mod->strtab = mod->core_strtab;
 #endif
+	mod_tree_remove_init(mod);
 	unset_module_init_ro_nx(mod);
 	module_arch_freeing_init(mod);
 	mod->module_init = NULL;
@@ -3127,11 +3279,11 @@ static noinline int do_init_module(struct module *mod)
 	mod->init_text_size = 0;
 	/*
 	 * We want to free module_init, but be aware that kallsyms may be
-	 * walking this with preempt disabled.  In all the failure paths,
-	 * we call synchronize_rcu/synchronize_sched, but we don't want
-	 * to slow down the success path, so use actual RCU here.
+	 * walking this with preempt disabled.  In all the failure paths, we
+	 * call synchronize_sched(), but we don't want to slow down the success
+	 * path, so use actual RCU here.
 	 */
-	call_rcu(&freeinit->rcu, do_free_init);
+	call_rcu_sched(&freeinit->rcu, do_free_init);
 	mutex_unlock(&module_mutex);
 	wake_up_all(&module_wq);
 
@@ -3188,7 +3340,9 @@ again:
 		err = -EEXIST;
 		goto out;
 	}
+	mod_update_bounds(mod);
 	list_add_rcu(&mod->list, &modules);
+	mod_tree_insert(mod);
 	err = 0;
 
 out:
@@ -3304,6 +3458,8 @@ static int load_module(struct load_info *info, const char __user *uargs,
 	if (err)
 		goto unlink_mod;
 
+	init_param_lock(mod);
+
 	/* Now we've got everything in the final locations, we can
 	 * find optional sections. */
 	err = find_module_sections(mod, info);
@@ -3402,8 +3558,8 @@ static int load_module(struct load_info *info, const char __user *uargs,
 	/* Unlink carefully: kallsyms could be walking list. */
 	list_del_rcu(&mod->list);
 	wake_up_all(&module_wq);
-	/* Wait for RCU synchronizing before releasing mod->list. */
-	synchronize_rcu();
+	/* Wait for RCU-sched synchronizing before releasing mod->list. */
+	synchronize_sched();
 	mutex_unlock(&module_mutex);
  free_module:
 	/* Free lock-classes; relies on the preceding sync_rcu() */
@@ -3527,19 +3683,15 @@ const char *module_address_lookup(unsigned long addr,
 			    char **modname,
 			    char *namebuf)
 {
-	struct module *mod;
 	const char *ret = NULL;
+	struct module *mod;
 
 	preempt_disable();
-	list_for_each_entry_rcu(mod, &modules, list) {
-		if (mod->state == MODULE_STATE_UNFORMED)
-			continue;
-		if (within_module(addr, mod)) {
-			if (modname)
-				*modname = mod->name;
-			ret = get_ksymbol(mod, addr, size, offset);
-			break;
-		}
+	mod = __module_address(addr);
+	if (mod) {
+		if (modname)
+			*modname = mod->name;
+		ret = get_ksymbol(mod, addr, size, offset);
 	}
 	/* Make a copy in here where it's safe */
 	if (ret) {
@@ -3547,6 +3699,7 @@ const char *module_address_lookup(unsigned long addr,
 		ret = namebuf;
 	}
 	preempt_enable();
+
 	return ret;
 }
 
@@ -3670,6 +3823,8 @@ int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *,
 	unsigned int i;
 	int ret;
 
+	module_assert_mutex();
+
 	list_for_each_entry(mod, &modules, list) {
 		if (mod->state == MODULE_STATE_UNFORMED)
 			continue;
@@ -3844,13 +3999,15 @@ struct module *__module_address(unsigned long addr)
 	if (addr < module_addr_min || addr > module_addr_max)
 		return NULL;
 
-	list_for_each_entry_rcu(mod, &modules, list) {
+	module_assert_mutex_or_preempt();
+
+	mod = mod_find(addr);
+	if (mod) {
+		BUG_ON(!within_module(addr, mod));
 		if (mod->state == MODULE_STATE_UNFORMED)
-			continue;
-		if (within_module(addr, mod))
-			return mod;
+			mod = NULL;
 	}
-	return NULL;
+	return mod;
 }
 EXPORT_SYMBOL_GPL(__module_address);