Merge patch series "efivarfs: support freeze/thaw"

Christian Brauner <[email protected]> says:

Allow efivarfs to partake to resync variable state during system
hibernation and suspend. Add freeze/thaw support.

This is a pretty straightforward implementation. We simply add regular
freeze/thaw support for both userspace and the kernel. This works
without any big issues and congrats afaict efivars is the first
pseudofilesystem that adds support for filesystem freezing and thawing.

The simplicity comes from the fact that we simply always resync variable
state after efivarfs has been frozen. It doesn't matter whether that's
because of suspend, userspace initiated freeze or hibernation. Efivars
is simple enough that it doesn't matter that we walk all dentries. There
are no directories and there aren't insane amounts of entries and both
freeze/thaw are already heavy-handed operations. If userspace initiated
a freeze/thaw cycle they would need CAP_SYS_ADMIN in the initial user
namespace (as that's where efivarfs is mounted) so it can't be triggered
by random userspace. IOW, we really really don't care.

* patches from https://lore.kernel.org/r/[email protected]:
  efivarfs: support freeze/thaw
  libfs: export find_next_child()

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Christian Brauner <[email protected]>

Author: brauner

fs/efivarfs/internal.h

@@ -17,7 +17,6 @@ struct efivarfs_fs_info {
 	struct efivarfs_mount_opts mount_opts;
 	struct super_block *sb;
 	struct notifier_block nb;
-	struct notifier_block pm_nb;
 };
 
 struct efi_variable {

fs/efivarfs/super.c

@@ -20,6 +20,7 @@
 #include <linux/printk.h>
 
 #include "internal.h"
+#include "../internal.h"
 
 static int efivarfs_ops_notifier(struct notifier_block *nb, unsigned long event,
 				 void *data)
@@ -119,12 +120,18 @@ static int efivarfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 
 	return 0;
 }
+
+static int efivarfs_freeze_fs(struct super_block *sb);
+static int efivarfs_unfreeze_fs(struct super_block *sb);
+
 static const struct super_operations efivarfs_ops = {
 	.statfs = efivarfs_statfs,
 	.drop_inode = generic_delete_inode,
 	.alloc_inode = efivarfs_alloc_inode,
 	.free_inode = efivarfs_free_inode,
 	.show_options = efivarfs_show_options,
+	.freeze_fs = efivarfs_freeze_fs,
+	.unfreeze_fs = efivarfs_unfreeze_fs,
 };
 
 /*
@@ -367,8 +374,6 @@ static int efivarfs_fill_super(struct super_block *sb, struct fs_context *fc)
 	if (err)
 		return err;
 
-	register_pm_notifier(&sfi->pm_nb);
-
 	return efivar_init(efivarfs_callback, sb, true);
 }
 
@@ -393,48 +398,6 @@ static const struct fs_context_operations efivarfs_context_ops = {
 	.reconfigure	= efivarfs_reconfigure,
 };
 
-struct efivarfs_ctx {
-	struct dir_context ctx;
-	struct super_block *sb;
-	struct dentry *dentry;
-};
-
-static bool efivarfs_actor(struct dir_context *ctx, const char *name, int len,
-			   loff_t offset, u64 ino, unsigned mode)
-{
-	unsigned long size;
-	struct efivarfs_ctx *ectx = container_of(ctx, struct efivarfs_ctx, ctx);
-	struct qstr qstr = { .name = name, .len = len };
-	struct dentry *dentry = d_hash_and_lookup(ectx->sb->s_root, &qstr);
-	struct inode *inode;
-	struct efivar_entry *entry;
-	int err;
-
-	if (IS_ERR_OR_NULL(dentry))
-		return true;
-
-	inode = d_inode(dentry);
-	entry = efivar_entry(inode);
-
-	err = efivar_entry_size(entry, &size);
-	size += sizeof(__u32);	/* attributes */
-	if (err)
-		size = 0;
-
-	inode_lock_nested(inode, I_MUTEX_CHILD);
-	i_size_write(inode, size);
-	inode_unlock(inode);
-
-	if (!size) {
-		ectx->dentry = dentry;
-		return false;
-	}
-
-	dput(dentry);
-
-	return true;
-}
-
 static int efivarfs_check_missing(efi_char16_t *name16, efi_guid_t vendor,
 				  unsigned long name_size, void *data)
 {
@@ -474,111 +437,59 @@ static int efivarfs_check_missing(efi_char16_t *name16, efi_guid_t vendor,
 	return err;
 }
 
-static void efivarfs_deactivate_super_work(struct work_struct *work)
-{
-	struct super_block *s = container_of(work, struct super_block,
-					     destroy_work);
-	/*
-	 * note: here s->destroy_work is free for reuse (which
-	 * will happen in deactivate_super)
-	 */
-	deactivate_super(s);
-}
-
 static struct file_system_type efivarfs_type;
 
-static int efivarfs_pm_notify(struct notifier_block *nb, unsigned long action,
-			      void *ptr)
+static int efivarfs_freeze_fs(struct super_block *sb)
 {
-	struct efivarfs_fs_info *sfi = container_of(nb, struct efivarfs_fs_info,
-						    pm_nb);
-	struct path path;
-	struct efivarfs_ctx ectx = {
-		.ctx = {
-			.actor	= efivarfs_actor,
-		},
-		.sb = sfi->sb,
-	};
-	struct file *file;
-	struct super_block *s = sfi->sb;
-	static bool rescan_done = true;
-
-	if (action == PM_HIBERNATION_PREPARE) {
-		rescan_done = false;
-		return NOTIFY_OK;
-	} else if (action != PM_POST_HIBERNATION) {
-		return NOTIFY_DONE;
-	}
-
-	if (rescan_done)
-		return NOTIFY_DONE;
-
-	/* ensure single superblock is alive and pin it */
-	if (!atomic_inc_not_zero(&s->s_active))
-		return NOTIFY_DONE;
-
-	pr_info("efivarfs: resyncing variable state\n");
-
-	path.dentry = sfi->sb->s_root;
-
-	/*
-	 * do not add SB_KERNMOUNT which a single superblock could
-	 * expose to userspace and which also causes MNT_INTERNAL, see
-	 * below
-	 */
-	path.mnt = vfs_kern_mount(&efivarfs_type, 0,
-				  efivarfs_type.name, NULL);
-	if (IS_ERR(path.mnt)) {
-		pr_err("efivarfs: internal mount failed\n");
-		/*
-		 * We may be the last pinner of the superblock but
-		 * calling efivarfs_kill_sb from within the notifier
-		 * here would deadlock trying to unregister it
-		 */
-		INIT_WORK(&s->destroy_work, efivarfs_deactivate_super_work);
-		schedule_work(&s->destroy_work);
-		return PTR_ERR(path.mnt);
-	}
-
-	/* path.mnt now has pin on superblock, so this must be above one */
-	atomic_dec(&s->s_active);
-
-	file = kernel_file_open(&path, O_RDONLY | O_DIRECTORY | O_NOATIME,
-				current_cred());
-	/*
-	 * safe even if last put because no MNT_INTERNAL means this
-	 * will do delayed deactivate_super and not deadlock
-	 */
-	mntput(path.mnt);
-	if (IS_ERR(file))
-		return NOTIFY_DONE;
+	/* Nothing for us to do. */
+	return 0;
+}
 
-	rescan_done = true;
+static int efivarfs_unfreeze_fs(struct super_block *sb)
+{
+	struct dentry *child = NULL;
 
 	/*
-	 * First loop over the directory and verify each entry exists,
-	 * removing it if it doesn't
+	 * Unconditionally resync the variable state on a thaw request.
+	 * Given the size of efivarfs it really doesn't matter to simply
+	 * iterate through all of the entries and resync. Freeze/thaw
+	 * requests are rare enough for that to not matter and the
+	 * number of entries is pretty low too. So we really don't care.
 	 */
-	file->f_pos = 2;	/* skip . and .. */
-	do {
-		ectx.dentry = NULL;
-		iterate_dir(file, &ectx.ctx);
-		if (ectx.dentry) {
-			pr_info("efivarfs: removing variable %pd\n",
-				ectx.dentry);
-			simple_recursive_removal(ectx.dentry, NULL);
-			dput(ectx.dentry);
+	pr_info("efivarfs: resyncing variable state\n");
+	for (;;) {
+		int err;
+		unsigned long size = 0;
+		struct inode *inode;
+		struct efivar_entry *entry;
+
+		child = find_next_child(sb->s_root, child);
+		if (!child)
+			break;
+
+		inode = d_inode(child);
+		entry = efivar_entry(inode);
+
+		err = efivar_entry_size(entry, &size);
+		if (err)
+			size = 0;
+		else
+			size += sizeof(__u32);
+
+		inode_lock(inode);
+		i_size_write(inode, size);
+		inode_unlock(inode);
+
+		/* The variable doesn't exist anymore, delete it. */
+		if (!size) {
+			pr_info("efivarfs: removing variable %pd\n", child);
+			simple_recursive_removal(child, NULL);
 		}
-	} while (ectx.dentry);
-	fput(file);
-
-	/*
-	 * then loop over variables, creating them if there's no matching
-	 * dentry
-	 */
-	efivar_init(efivarfs_check_missing, sfi->sb, false);
+	}
 
-	return NOTIFY_OK;
+	efivar_init(efivarfs_check_missing, sb, false);
+	pr_info("efivarfs: finished resyncing variable state\n");
+	return 0;
 }
 
 static int efivarfs_init_fs_context(struct fs_context *fc)
@@ -598,9 +509,6 @@ static int efivarfs_init_fs_context(struct fs_context *fc)
 	fc->s_fs_info = sfi;
 	fc->ops = &efivarfs_context_ops;
 
-	sfi->pm_nb.notifier_call = efivarfs_pm_notify;
-	sfi->pm_nb.priority = 0;
-
 	return 0;
 }
 
@@ -610,7 +518,6 @@ static void efivarfs_kill_sb(struct super_block *sb)
 
 	blocking_notifier_chain_unregister(&efivar_ops_nh, &sfi->nb);
 	kill_litter_super(sb);
-	unregister_pm_notifier(&sfi->pm_nb);
 
 	kfree(sfi);
 }

fs/internal.h

@@ -343,3 +343,4 @@ static inline bool path_mounted(const struct path *path)
 void file_f_owner_release(struct file *file);
 bool file_seek_cur_needs_f_lock(struct file *file);
 int statmount_mnt_idmap(struct mnt_idmap *idmap, struct seq_file *seq, bool uid_map);
+struct dentry *find_next_child(struct dentry *parent, struct dentry *prev);

fs/libfs.c

@@ -583,7 +583,7 @@ const struct file_operations simple_offset_dir_operations = {
 	.fsync		= noop_fsync,
 };
 
-static struct dentry *find_next_child(struct dentry *parent, struct dentry *prev)
+struct dentry *find_next_child(struct dentry *parent, struct dentry *prev)
 {
 	struct dentry *child = NULL, *d;
 
@@ -603,6 +603,7 @@ static struct dentry *find_next_child(struct dentry *parent, struct dentry *prev
 	dput(prev);
 	return child;
 }
+EXPORT_SYMBOL(find_next_child);
 
 void simple_recursive_removal(struct dentry *dentry,
                               void (*callback)(struct dentry *))