convert alloc_super() to struct fs_context

IMO it's premature by several cycles - David might hope to get
rid of sget_userns() and friends, but it will take a while and
until that's done...  Let's just say that on-stack instances
of struct fs_context, skipping the normal constructors and not
being fed to destructor are Not Nice(tm) in general.  Yes, in
this particular case we are safe, but it really doesn't buy
us anything worthwhile.

*IF* all callers of alloc_super() end up with handy fs_context -
sure, this will be the obvious change.  Until then... I'm taking
that to the very end of the series lest it gets lost, but it's
not going to be in any pull requests this cycle.  And probably
the next cycle or two as well.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

fs/super.c

diff --git a/fs/super.c b/fs/super.c
index aef58be..254291b 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -188,16 +188,13 @@ static void destroy_unused_super(struct super_block *s)
 }
 
 /**
- *	alloc_super	-	create new superblock
- *	@type:	filesystem type superblock should belong to
- *	@flags: the mount flags
- *	@user_ns: User namespace for the super_block
+ *	alloc_super - Create new superblock
+ *	@fc: The filesystem configuration context
  *
  *	Allocates and initializes a new &struct super_block.  alloc_super()
  *	returns a pointer new superblock or %NULL if allocation had failed.
  */
-static struct super_block *alloc_super(struct file_system_type *type, int flags,
-				       struct user_namespace *user_ns)
+static struct super_block *alloc_super(struct fs_context *fc)
 {
 	struct super_block *s = kzalloc(sizeof(struct super_block),  GFP_USER);
 	static const struct super_operations default_op;
@@ -207,9 +204,9 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags,
 		return NULL;
 
 	INIT_LIST_HEAD(&s->s_mounts);
-	s->s_user_ns = get_user_ns(user_ns);
+	s->s_user_ns = get_user_ns(fc->user_ns);
 	init_rwsem(&s->s_umount);
-	lockdep_set_class(&s->s_umount, &type->s_umount_key);
+	lockdep_set_class(&s->s_umount, &fc->fs_type->s_umount_key);
 	/*
 	 * sget() can have s_umount recursion.
 	 *
@@ -233,12 +230,12 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags,
 	for (i = 0; i < SB_FREEZE_LEVELS; i++) {
 		if (__percpu_init_rwsem(&s->s_writers.rw_sem[i],
 					sb_writers_name[i],
-					&type->s_writers_key[i]))
+					&fc->fs_type->s_writers_key[i]))
 			goto fail;
 	}
 	init_waitqueue_head(&s->s_writers.wait_unfrozen);
 	s->s_bdi = &noop_backing_dev_info;
-	s->s_flags = flags;
+	s->s_flags = fc->sb_flags;
 	if (s->s_user_ns != &init_user_ns)
 		s->s_iflags |= SB_I_NODEV;
 	INIT_HLIST_NODE(&s->s_instances);
@@ -252,7 +249,7 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags,
 	s->s_count = 1;
 	atomic_set(&s->s_active, 1);
 	mutex_init(&s->s_vfs_rename_mutex);
-	lockdep_set_class(&s->s_vfs_rename_mutex, &type->s_vfs_rename_key);
+	lockdep_set_class(&s->s_vfs_rename_mutex, &fc->fs_type->s_vfs_rename_key);
 	init_rwsem(&s->s_dquot.dqio_sem);
 	s->s_maxbytes = MAX_NON_LFS;
 	s->s_op = &default_op;
@@ -524,7 +521,7 @@ struct super_block *sget_fc(struct fs_context *fc,
 	}
 	if (!s) {
 		spin_unlock(&sb_lock);
-		s = alloc_super(fc->fs_type, fc->sb_flags, fc->user_ns);
+		s = alloc_super(fc);
 		if (!s)
 			return ERR_PTR(-ENOMEM);
 		goto retry;
@@ -603,7 +600,14 @@ struct super_block *sget_userns(struct file_system_type *type,
 	}
 	if (!s) {
 		spin_unlock(&sb_lock);
-		s = alloc_super(type, (flags & ~SB_SUBMOUNT), user_ns);
+		{
+			struct fs_context fc = {
+				.fs_type	= type,
+				.sb_flags	= flags & ~SB_SUBMOUNT,
+				.user_ns	= user_ns,
+			};
+			s = alloc_super(&fc);
+		}
 		if (!s)
 			return ERR_PTR(-ENOMEM);
 		goto retry;