libceph, rbd, ceph: convert to use the new mount API

Convert the ceph filesystem to the new internal mount API as the old
one will be obsoleted and removed.  This allows greater flexibility in
communication of mount parameters between userspace, the VFS and the
filesystem.

See Documentation/filesystems/mount_api.txt for more information.

[ Numerous string handling, leak and regression fixes; rbd conversion
  was particularly broken and had to be redone almost from scratch. ]

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

1 files changed, 324 insertions, 322 deletions

diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index b47f43fc2d68..9c9a7c68eea3 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -9,7 +9,8 @@
 #include <linux/in6.h>
 #include <linux/module.h>
 #include <linux/mount.h>
-#include <linux/parser.h>
+#include <linux/fs_context.h>
+#include <linux/fs_parser.h>
 #include <linux/sched.h>
 #include <linux/seq_file.h>
 #include <linux/slab.h>
@@ -138,280 +139,308 @@ enum {
 	Opt_readdir_max_entries,
 	Opt_readdir_max_bytes,
 	Opt_congestion_kb,
-	Opt_last_int,
 	/* int args above */
 	Opt_snapdirname,
 	Opt_mds_namespace,
-	Opt_fscache_uniq,
 	Opt_recover_session,
-	Opt_last_string,
+	Opt_source,
 	/* string args above */
 	Opt_dirstat,
-	Opt_nodirstat,
 	Opt_rbytes,
-	Opt_norbytes,
 	Opt_asyncreaddir,
-	Opt_noasyncreaddir,
 	Opt_dcache,
-	Opt_nodcache,
 	Opt_ino32,
-	Opt_noino32,
 	Opt_fscache,
-	Opt_nofscache,
 	Opt_poolperm,
-	Opt_nopoolperm,
 	Opt_require_active_mds,
-	Opt_norequire_active_mds,
-#ifdef CONFIG_CEPH_FS_POSIX_ACL
 	Opt_acl,
-#endif
-	Opt_noacl,
 	Opt_quotadf,
-	Opt_noquotadf,
 	Opt_copyfrom,
-	Opt_nocopyfrom,
 };
 
-static match_table_t fsopt_tokens = {
-	{Opt_wsize, "wsize=%d"},
-	{Opt_rsize, "rsize=%d"},
-	{Opt_rasize, "rasize=%d"},
-	{Opt_caps_wanted_delay_min, "caps_wanted_delay_min=%d"},
-	{Opt_caps_wanted_delay_max, "caps_wanted_delay_max=%d"},
-	{Opt_caps_max, "caps_max=%d"},
-	{Opt_readdir_max_entries, "readdir_max_entries=%d"},
-	{Opt_readdir_max_bytes, "readdir_max_bytes=%d"},
-	{Opt_congestion_kb, "write_congestion_kb=%d"},
-	/* int args above */
-	{Opt_snapdirname, "snapdirname=%s"},
-	{Opt_mds_namespace, "mds_namespace=%s"},
-	{Opt_recover_session, "recover_session=%s"},
-	{Opt_fscache_uniq, "fsc=%s"},
-	/* string args above */
-	{Opt_dirstat, "dirstat"},
-	{Opt_nodirstat, "nodirstat"},
-	{Opt_rbytes, "rbytes"},
-	{Opt_norbytes, "norbytes"},
-	{Opt_asyncreaddir, "asyncreaddir"},
-	{Opt_noasyncreaddir, "noasyncreaddir"},
-	{Opt_dcache, "dcache"},
-	{Opt_nodcache, "nodcache"},
-	{Opt_ino32, "ino32"},
-	{Opt_noino32, "noino32"},
-	{Opt_fscache, "fsc"},
-	{Opt_nofscache, "nofsc"},
-	{Opt_poolperm, "poolperm"},
-	{Opt_nopoolperm, "nopoolperm"},
-	{Opt_require_active_mds, "require_active_mds"},
-	{Opt_norequire_active_mds, "norequire_active_mds"},
-#ifdef CONFIG_CEPH_FS_POSIX_ACL
-	{Opt_acl, "acl"},
-#endif
-	{Opt_noacl, "noacl"},
-	{Opt_quotadf, "quotadf"},
-	{Opt_noquotadf, "noquotadf"},
-	{Opt_copyfrom, "copyfrom"},
-	{Opt_nocopyfrom, "nocopyfrom"},
-	{-1, NULL}
+enum ceph_recover_session_mode {
+	ceph_recover_session_no,
+	ceph_recover_session_clean
+};
+
+static const struct fs_parameter_enum ceph_mount_param_enums[] = {
+	{ Opt_recover_session,	"no",		ceph_recover_session_no },
+	{ Opt_recover_session,	"clean",	ceph_recover_session_clean },
+	{}
+};
+
+static const struct fs_parameter_spec ceph_mount_param_specs[] = {
+	fsparam_flag_no ("acl",				Opt_acl),
+	fsparam_flag_no ("asyncreaddir",		Opt_asyncreaddir),
+	fsparam_u32	("caps_max",			Opt_caps_max),
+	fsparam_u32	("caps_wanted_delay_max",	Opt_caps_wanted_delay_max),
+	fsparam_u32	("caps_wanted_delay_min",	Opt_caps_wanted_delay_min),
+	fsparam_s32	("write_congestion_kb",		Opt_congestion_kb),
+	fsparam_flag_no ("copyfrom",			Opt_copyfrom),
+	fsparam_flag_no ("dcache",			Opt_dcache),
+	fsparam_flag_no ("dirstat",			Opt_dirstat),
+	__fsparam	(fs_param_is_string, "fsc",	Opt_fscache,
+			 fs_param_neg_with_no | fs_param_v_optional),
+	fsparam_flag_no ("ino32",			Opt_ino32),
+	fsparam_string	("mds_namespace",		Opt_mds_namespace),
+	fsparam_flag_no ("poolperm",			Opt_poolperm),
+	fsparam_flag_no ("quotadf",			Opt_quotadf),
+	fsparam_u32	("rasize",			Opt_rasize),
+	fsparam_flag_no ("rbytes",			Opt_rbytes),
+	fsparam_s32	("readdir_max_bytes",		Opt_readdir_max_bytes),
+	fsparam_s32	("readdir_max_entries",		Opt_readdir_max_entries),
+	fsparam_enum	("recover_session",		Opt_recover_session),
+	fsparam_flag_no ("require_active_mds",		Opt_require_active_mds),
+	fsparam_u32	("rsize",			Opt_rsize),
+	fsparam_string	("snapdirname",			Opt_snapdirname),
+	fsparam_string	("source",			Opt_source),
+	fsparam_u32	("wsize",			Opt_wsize),
+	{}
+};
+
+static const struct fs_parameter_description ceph_mount_parameters = {
+	.name           = "ceph",
+	.specs          = ceph_mount_param_specs,
+	.enums		= ceph_mount_param_enums,
 };
 
-static int parse_fsopt_token(char *c, void *private)
+struct ceph_parse_opts_ctx {
+	struct ceph_options		*copts;
+	struct ceph_mount_options	*opts;
+};
+
+/*
+ * Parse the source parameter.  Distinguish the server list from the path.
+ * Internally we do not include the leading '/' in the path.
+ *
+ * The source will look like:
+ *     <server_spec>[,<server_spec>...]:[<path>]
+ * where
+ *     <server_spec> is <ip>[:<port>]
+ *     <path> is optional, but if present must begin with '/'
+ */
+static int ceph_parse_source(struct fs_parameter *param, struct fs_context *fc)
 {
-	struct ceph_mount_options *fsopt = private;
-	substring_t argstr[MAX_OPT_ARGS];
-	int token, intval, ret;
+	struct ceph_parse_opts_ctx *pctx = fc->fs_private;
+	struct ceph_mount_options *fsopt = pctx->opts;
+	char *dev_name = param->string, *dev_name_end;
+	int ret;
 
-	token = match_token((char *)c, fsopt_tokens, argstr);
-	if (token < 0)
-		return -EINVAL;
+	dout("%s '%s'\n", __func__, dev_name);
+	if (!dev_name || !*dev_name)
+		return invalf(fc, "ceph: Empty source");
 
-	if (token < Opt_last_int) {
-		ret = match_int(&argstr[0], &intval);
-		if (ret < 0) {
-			pr_err("bad option arg (not int) at '%s'\n", c);
-			return ret;
+	dev_name_end = strchr(dev_name, '/');
+	if (dev_name_end) {
+		if (strlen(dev_name_end) > 1) {
+			kfree(fsopt->server_path);
+			fsopt->server_path = kstrdup(dev_name_end, GFP_KERNEL);
+			if (!fsopt->server_path)
+				return -ENOMEM;
 		}
-		dout("got int token %d val %d\n", token, intval);
-	} else if (token > Opt_last_int && token < Opt_last_string) {
-		dout("got string token %d val %s\n", token,
-		     argstr[0].from);
 	} else {
-		dout("got token %d\n", token);
+		dev_name_end = dev_name + strlen(dev_name);
 	}
 
+	dev_name_end--;		/* back up to ':' separator */
+	if (dev_name_end < dev_name || *dev_name_end != ':')
+		return invalf(fc, "ceph: No path or : separator in source");
+
+	dout("device name '%.*s'\n", (int)(dev_name_end - dev_name), dev_name);
+	if (fsopt->server_path)
+		dout("server path '%s'\n", fsopt->server_path);
+
+	ret = ceph_parse_mon_ips(param->string, dev_name_end - dev_name,
+				 pctx->copts, fc);
+	if (ret)
+		return ret;
+
+	fc->source = param->string;
+	param->string = NULL;
+	return 0;
+}
+
+static int ceph_parse_mount_param(struct fs_context *fc,
+				  struct fs_parameter *param)
+{
+	struct ceph_parse_opts_ctx *pctx = fc->fs_private;
+	struct ceph_mount_options *fsopt = pctx->opts;
+	struct fs_parse_result result;
+	unsigned int mode;
+	int token, ret;
+
+	ret = ceph_parse_param(param, pctx->copts, fc);
+	if (ret != -ENOPARAM)
+		return ret;
+
+	token = fs_parse(fc, &ceph_mount_parameters, param, &result);
+	dout("%s fs_parse '%s' token %d\n", __func__, param->key, token);
+	if (token < 0)
+		return token;
+
 	switch (token) {
 	case Opt_snapdirname:
 		kfree(fsopt->snapdir_name);
-		fsopt->snapdir_name = kstrndup(argstr[0].from,
-					       argstr[0].to-argstr[0].from,
-					       GFP_KERNEL);
-		if (!fsopt->snapdir_name)
-			return -ENOMEM;
+		fsopt->snapdir_name = param->string;
+		param->string = NULL;
 		break;
 	case Opt_mds_namespace:
 		kfree(fsopt->mds_namespace);
-		fsopt->mds_namespace = kstrndup(argstr[0].from,
-						argstr[0].to-argstr[0].from,
-						GFP_KERNEL);
-		if (!fsopt->mds_namespace)
-			return -ENOMEM;
+		fsopt->mds_namespace = param->string;
+		param->string = NULL;
 		break;
 	case Opt_recover_session:
-		if (!strncmp(argstr[0].from, "no",
-			     argstr[0].to - argstr[0].from)) {
+		mode = result.uint_32;
+		if (mode == ceph_recover_session_no)
 			fsopt->flags &= ~CEPH_MOUNT_OPT_CLEANRECOVER;
-		} else if (!strncmp(argstr[0].from, "clean",
-				    argstr[0].to - argstr[0].from)) {
+		else if (mode == ceph_recover_session_clean)
 			fsopt->flags |= CEPH_MOUNT_OPT_CLEANRECOVER;
-		} else {
-			return -EINVAL;
-		}
-		break;
-	case Opt_fscache_uniq:
-#ifdef CONFIG_CEPH_FSCACHE
-		kfree(fsopt->fscache_uniq);
-		fsopt->fscache_uniq = kstrndup(argstr[0].from,
-					       argstr[0].to-argstr[0].from,
-					       GFP_KERNEL);
-		if (!fsopt->fscache_uniq)
-			return -ENOMEM;
-		fsopt->flags |= CEPH_MOUNT_OPT_FSCACHE;
+		else
+			BUG();
 		break;
-#else
-		pr_err("fscache support is disabled\n");
-		return -EINVAL;
-#endif
+	case Opt_source:
+		if (fc->source)
+			return invalf(fc, "ceph: Multiple sources specified");
+		return ceph_parse_source(param, fc);
 	case Opt_wsize:
-		if (intval < (int)PAGE_SIZE || intval > CEPH_MAX_WRITE_SIZE)
-			return -EINVAL;
-		fsopt->wsize = ALIGN(intval, PAGE_SIZE);
+		if (result.uint_32 < PAGE_SIZE ||
+		    result.uint_32 > CEPH_MAX_WRITE_SIZE)
+			goto out_of_range;
+		fsopt->wsize = ALIGN(result.uint_32, PAGE_SIZE);
 		break;
 	case Opt_rsize:
-		if (intval < (int)PAGE_SIZE || intval > CEPH_MAX_READ_SIZE)
-			return -EINVAL;
-		fsopt->rsize = ALIGN(intval, PAGE_SIZE);
+		if (result.uint_32 < PAGE_SIZE ||
+		    result.uint_32 > CEPH_MAX_READ_SIZE)
+			goto out_of_range;
+		fsopt->rsize = ALIGN(result.uint_32, PAGE_SIZE);
 		break;
 	case Opt_rasize:
-		if (intval < 0)
-			return -EINVAL;
-		fsopt->rasize = ALIGN(intval, PAGE_SIZE);
+		fsopt->rasize = ALIGN(result.uint_32, PAGE_SIZE);
 		break;
 	case Opt_caps_wanted_delay_min:
-		if (intval < 1)
-			return -EINVAL;
-		fsopt->caps_wanted_delay_min = intval;
+		if (result.uint_32 < 1)
+			goto out_of_range;
+		fsopt->caps_wanted_delay_min = result.uint_32;
 		break;
 	case Opt_caps_wanted_delay_max:
-		if (intval < 1)
-			return -EINVAL;
-		fsopt->caps_wanted_delay_max = intval;
+		if (result.uint_32 < 1)
+			goto out_of_range;
+		fsopt->caps_wanted_delay_max = result.uint_32;
 		break;
 	case Opt_caps_max:
-		if (intval < 0)
-			return -EINVAL;
-		fsopt->caps_max = intval;
+		fsopt->caps_max = result.uint_32;
 		break;
 	case Opt_readdir_max_entries:
-		if (intval < 1)
-			return -EINVAL;
-		fsopt->max_readdir = intval;
+		if (result.uint_32 < 1)
+			goto out_of_range;
+		fsopt->max_readdir = result.uint_32;
 		break;
 	case Opt_readdir_max_bytes:
-		if (intval < (int)PAGE_SIZE && intval != 0)
-			return -EINVAL;
-		fsopt->max_readdir_bytes = intval;
+		if (result.uint_32 < PAGE_SIZE && result.uint_32 != 0)
+			goto out_of_range;
+		fsopt->max_readdir_bytes = result.uint_32;
 		break;
 	case Opt_congestion_kb:
-		if (intval < 1024) /* at least 1M */
-			return -EINVAL;
-		fsopt->congestion_kb = intval;
+		if (result.uint_32 < 1024) /* at least 1M */
+			goto out_of_range;
+		fsopt->congestion_kb = result.uint_32;
 		break;
 	case Opt_dirstat:
-		fsopt->flags |= CEPH_MOUNT_OPT_DIRSTAT;
-		break;
-	case Opt_nodirstat:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_DIRSTAT;
+		if (!result.negated)
+			fsopt->flags |= CEPH_MOUNT_OPT_DIRSTAT;
+		else
+			fsopt->flags &= ~CEPH_MOUNT_OPT_DIRSTAT;
 		break;
 	case Opt_rbytes:
-		fsopt->flags |= CEPH_MOUNT_OPT_RBYTES;
-		break;
-	case Opt_norbytes:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_RBYTES;
+		if (!result.negated)
+			fsopt->flags |= CEPH_MOUNT_OPT_RBYTES;
+		else
+			fsopt->flags &= ~CEPH_MOUNT_OPT_RBYTES;
 		break;
 	case Opt_asyncreaddir:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_NOASYNCREADDIR;
-		break;
-	case Opt_noasyncreaddir:
-		fsopt->flags |= CEPH_MOUNT_OPT_NOASYNCREADDIR;
+		if (!result.negated)
+			fsopt->flags &= ~CEPH_MOUNT_OPT_NOASYNCREADDIR;
+		else
+			fsopt->flags |= CEPH_MOUNT_OPT_NOASYNCREADDIR;
 		break;
 	case Opt_dcache:
-		fsopt->flags |= CEPH_MOUNT_OPT_DCACHE;
-		break;
-	case Opt_nodcache:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_DCACHE;
+		if (!result.negated)
+			fsopt->flags |= CEPH_MOUNT_OPT_DCACHE;
+		else
+			fsopt->flags &= ~CEPH_MOUNT_OPT_DCACHE;
 		break;
 	case Opt_ino32:
-		fsopt->flags |= CEPH_MOUNT_OPT_INO32;
-		break;
-	case Opt_noino32:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_INO32;
+		if (!result.negated)
+			fsopt->flags |= CEPH_MOUNT_OPT_INO32;
+		else
+			fsopt->flags &= ~CEPH_MOUNT_OPT_INO32;
 		break;
+
 	case Opt_fscache:
 #ifdef CONFIG_CEPH_FSCACHE
-		fsopt->flags |= CEPH_MOUNT_OPT_FSCACHE;
 		kfree(fsopt->fscache_uniq);
 		fsopt->fscache_uniq = NULL;
+		if (result.negated) {
+			fsopt->flags &= ~CEPH_MOUNT_OPT_FSCACHE;
+		} else {
+			fsopt->flags |= CEPH_MOUNT_OPT_FSCACHE;
+			fsopt->fscache_uniq = param->string;
+			param->string = NULL;
+		}
 		break;
 #else
-		pr_err("fscache support is disabled\n");
-		return -EINVAL;
+		return invalf(fc, "ceph: fscache support is disabled");
 #endif
-	case Opt_nofscache:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_FSCACHE;
-		kfree(fsopt->fscache_uniq);
-		fsopt->fscache_uniq = NULL;
-		break;
 	case Opt_poolperm:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_NOPOOLPERM;
-		break;
-	case Opt_nopoolperm:
-		fsopt->flags |= CEPH_MOUNT_OPT_NOPOOLPERM;
+		if (!result.negated)
+			fsopt->flags &= ~CEPH_MOUNT_OPT_NOPOOLPERM;
+		else
+			fsopt->flags |= CEPH_MOUNT_OPT_NOPOOLPERM;
 		break;
 	case Opt_require_active_mds:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_MOUNTWAIT;
-		break;
-	case Opt_norequire_active_mds:
-		fsopt->flags |= CEPH_MOUNT_OPT_MOUNTWAIT;
+		if (!result.negated)
+			fsopt->flags &= ~CEPH_MOUNT_OPT_MOUNTWAIT;
+		else
+			fsopt->flags |= CEPH_MOUNT_OPT_MOUNTWAIT;
 		break;
 	case Opt_quotadf:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_NOQUOTADF;
-		break;
-	case Opt_noquotadf:
-		fsopt->flags |= CEPH_MOUNT_OPT_NOQUOTADF;
+		if (!result.negated)
+			fsopt->flags &= ~CEPH_MOUNT_OPT_NOQUOTADF;
+		else
+			fsopt->flags |= CEPH_MOUNT_OPT_NOQUOTADF;
 		break;
 	case Opt_copyfrom:
-		fsopt->flags &= ~CEPH_MOUNT_OPT_NOCOPYFROM;
-		break;
-	case Opt_nocopyfrom:
-		fsopt->flags |= CEPH_MOUNT_OPT_NOCOPYFROM;
+		if (!result.negated)
+			fsopt->flags &= ~CEPH_MOUNT_OPT_NOCOPYFROM;
+		else
+			fsopt->flags |= CEPH_MOUNT_OPT_NOCOPYFROM;
 		break;
-#ifdef CONFIG_CEPH_FS_POSIX_ACL
 	case Opt_acl:
-		fsopt->sb_flags |= SB_POSIXACL;
-		break;
+		if (!result.negated) {
+#ifdef CONFIG_CEPH_FS_POSIX_ACL
+			fc->sb_flags |= SB_POSIXACL;
+#else
+			return invalf(fc, "ceph: POSIX ACL support is disabled");
 #endif
-	case Opt_noacl:
-		fsopt->sb_flags &= ~SB_POSIXACL;
+		} else {
+			fc->sb_flags &= ~SB_POSIXACL;
+		}
 		break;
 	default:
-		BUG_ON(token);
+		BUG();
 	}
 	return 0;
+
+out_of_range:
+	return invalf(fc, "ceph: %s out of range", param->key);
 }
 
 static void destroy_mount_options(struct ceph_mount_options *args)
 {
 	dout("destroy_mount_options %p\n", args);
+	if (!args)
+		return;
+
 	kfree(args->snapdir_name);
 	kfree(args->mds_namespace);
 	kfree(args->server_path);
@@ -459,91 +488,6 @@ static int compare_mount_options(struct ceph_mount_options *new_fsopt,
 	return ceph_compare_options(new_opt, fsc->client);
 }
 
-static int parse_mount_options(struct ceph_mount_options **pfsopt,
-			       struct ceph_options **popt,
-			       int flags, char *options,
-			       const char *dev_name)
-{
-	struct ceph_mount_options *fsopt;
-	const char *dev_name_end;
-	int err;
-
-	if (!dev_name || !*dev_name)
-		return -EINVAL;
-
-	fsopt = kzalloc(sizeof(*fsopt), GFP_KERNEL);
-	if (!fsopt)
-		return -ENOMEM;
-
-	dout("parse_mount_options %p, dev_name '%s'\n", fsopt, dev_name);
-
-	fsopt->sb_flags = flags;
-	fsopt->flags = CEPH_MOUNT_OPT_DEFAULT;
-
-	fsopt->wsize = CEPH_MAX_WRITE_SIZE;
-	fsopt->rsize = CEPH_MAX_READ_SIZE;
-	fsopt->rasize = CEPH_RASIZE_DEFAULT;
-	fsopt->snapdir_name = kstrdup(CEPH_SNAPDIRNAME_DEFAULT, GFP_KERNEL);
-	if (!fsopt->snapdir_name) {
-		err = -ENOMEM;
-		goto out;
-	}
-
-	fsopt->caps_wanted_delay_min = CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT;
-	fsopt->caps_wanted_delay_max = CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT;
-	fsopt->max_readdir = CEPH_MAX_READDIR_DEFAULT;
-	fsopt->max_readdir_bytes = CEPH_MAX_READDIR_BYTES_DEFAULT;
-	fsopt->congestion_kb = default_congestion_kb();
-
-	/*
-	 * Distinguish the server list from the path in "dev_name".
-	 * Internally we do not include the leading '/' in the path.
-	 *
-	 * "dev_name" will look like:
-	 *     <server_spec>[,<server_spec>...]:[<path>]
-	 * where
-	 *     <server_spec> is <ip>[:<port>]
-	 *     <path> is optional, but if present must begin with '/'
-	 */
-	dev_name_end = strchr(dev_name, '/');
-	if (dev_name_end) {
-		if (strlen(dev_name_end) > 1) {
-			fsopt->server_path = kstrdup(dev_name_end, GFP_KERNEL);
-			if (!fsopt->server_path) {
-				err = -ENOMEM;
-				goto out;
-			}
-		}
-	} else {
-		dev_name_end = dev_name + strlen(dev_name);
-	}
-	err = -EINVAL;
-	dev_name_end--;		/* back up to ':' separator */
-	if (dev_name_end < dev_name || *dev_name_end != ':') {
-		pr_err("device name is missing path (no : separator in %s)\n",
-				dev_name);
-		goto out;
-	}
-	dout("device name '%.*s'\n", (int)(dev_name_end - dev_name), dev_name);
-	if (fsopt->server_path)
-		dout("server path '%s'\n", fsopt->server_path);
-
-	*popt = ceph_parse_options(options, dev_name, dev_name_end,
-				 parse_fsopt_token, (void *)fsopt);
-	if (IS_ERR(*popt)) {
-		err = PTR_ERR(*popt);
-		goto out;
-	}
-
-	/* success */
-	*pfsopt = fsopt;
-	return 0;
-
-out:
-	destroy_mount_options(fsopt);
-	return err;
-}
-
 /**
  * ceph_show_options - Show mount options in /proc/mounts
  * @m: seq_file to write to
@@ -587,7 +531,7 @@ static int ceph_show_options(struct seq_file *m, struct dentry *root)
 		seq_puts(m, ",noquotadf");
 
 #ifdef CONFIG_CEPH_FS_POSIX_ACL
-	if (fsopt->sb_flags & SB_POSIXACL)
+	if (root->d_sb->s_flags & SB_POSIXACL)
 		seq_puts(m, ",acl");
 	else
 		seq_puts(m, ",noacl");
@@ -860,12 +804,6 @@ static void ceph_umount_begin(struct super_block *sb)
 	fsc->filp_gen++; // invalidate open files
 }
 
-static int ceph_remount(struct super_block *sb, int *flags, char *data)
-{
-	sync_filesystem(sb);
-	return 0;
-}
-
 static const struct super_operations ceph_super_ops = {
 	.alloc_inode	= ceph_alloc_inode,
 	.free_inode	= ceph_free_inode,
@@ -874,7 +812,6 @@ static const struct super_operations ceph_super_ops = {
 	.evict_inode	= ceph_evict_inode,
 	.sync_fs        = ceph_sync_fs,
 	.put_super	= ceph_put_super,
-	.remount_fs	= ceph_remount,
 	.show_options   = ceph_show_options,
 	.statfs		= ceph_statfs,
 	.umount_begin   = ceph_umount_begin,
@@ -935,7 +872,8 @@ out:
 /*
  * mount: join the ceph cluster, and open root directory.
  */
-static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc)
+static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc,
+				      struct fs_context *fc)
 {
 	int err;
 	unsigned long started = jiffies;  /* note the start time */
@@ -952,7 +890,7 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc)
 
 		/* setup fscache */
 		if (fsc->mount_options->flags & CEPH_MOUNT_OPT_FSCACHE) {
-			err = ceph_fscache_register_fs(fsc);
+			err = ceph_fscache_register_fs(fsc, fc);
 			if (err < 0)
 				goto out;
 		}
@@ -987,18 +925,16 @@ out:
 	return ERR_PTR(err);
 }
 
-static int ceph_set_super(struct super_block *s, void *data)
+static int ceph_set_super(struct super_block *s, struct fs_context *fc)
 {
-	struct ceph_fs_client *fsc = data;
+	struct ceph_fs_client *fsc = s->s_fs_info;
 	int ret;
 
-	dout("set_super %p data %p\n", s, data);
+	dout("set_super %p\n", s);
 
-	s->s_flags = fsc->mount_options->sb_flags;
 	s->s_maxbytes = MAX_LFS_FILESIZE;
 
 	s->s_xattr = ceph_xattr_handlers;
-	s->s_fs_info = fsc;
 	fsc->sb = s;
 	fsc->max_file_size = 1ULL << 40; /* temp value until we get mdsmap */
 
@@ -1010,24 +946,18 @@ static int ceph_set_super(struct super_block *s, void *data)
 	s->s_time_min = 0;
 	s->s_time_max = U32_MAX;
 
-	ret = set_anon_super(s, NULL);  /* what is that second arg for? */
+	ret = set_anon_super_fc(s, fc);
 	if (ret != 0)
-		goto fail;
-
-	return ret;
-
-fail:
-	s->s_fs_info = NULL;
-	fsc->sb = NULL;
+		fsc->sb = NULL;
 	return ret;
 }
 
 /*
  * share superblock if same fs AND options
  */
-static int ceph_compare_super(struct super_block *sb, void *data)
+static int ceph_compare_super(struct super_block *sb, struct fs_context *fc)
 {
-	struct ceph_fs_client *new = data;
+	struct ceph_fs_client *new = fc->s_fs_info;
 	struct ceph_mount_options *fsopt = new->mount_options;
 	struct ceph_options *opt = new->client->options;
 	struct ceph_fs_client *other = ceph_sb_to_client(sb);
@@ -1043,7 +973,7 @@ static int ceph_compare_super(struct super_block *sb, void *data)
 		dout("fsid doesn't match\n");
 		return 0;
 	}
-	if (fsopt->sb_flags != other->mount_options->sb_flags) {
+	if (fc->sb_flags != (sb->s_flags & ~SB_BORN)) {
 		dout("flags differ\n");
 		return 0;
 	}
@@ -1073,46 +1003,46 @@ static int ceph_setup_bdi(struct super_block *sb, struct ceph_fs_client *fsc)
 	return 0;
 }
 
-static struct dentry *ceph_mount(struct file_system_type *fs_type,
-		       int flags, const char *dev_name, void *data)
+static int ceph_get_tree(struct fs_context *fc)
 {
+	struct ceph_parse_opts_ctx *pctx = fc->fs_private;
 	struct super_block *sb;
 	struct ceph_fs_client *fsc;
 	struct dentry *res;
+	int (*compare_super)(struct super_block *, struct fs_context *) =
+		ceph_compare_super;
 	int err;
-	int (*compare_super)(struct super_block *, void *) = ceph_compare_super;
-	struct ceph_mount_options *fsopt = NULL;
-	struct ceph_options *opt = NULL;
 
-	dout("ceph_mount\n");
+	dout("ceph_get_tree\n");
+
+	if (!fc->source)
+		return invalf(fc, "ceph: No source");
 
 #ifdef CONFIG_CEPH_FS_POSIX_ACL
-	flags |= SB_POSIXACL;
+	fc->sb_flags |= SB_POSIXACL;
 #endif
-	err = parse_mount_options(&fsopt, &opt, flags, data, dev_name);
-	if (err < 0) {
-		res = ERR_PTR(err);
-		goto out_final;
-	}
 
 	/* create client (which we may/may not use) */
-	fsc = create_fs_client(fsopt, opt);
+	fsc = create_fs_client(pctx->opts, pctx->copts);
+	pctx->opts = NULL;
+	pctx->copts = NULL;
 	if (IS_ERR(fsc)) {
-		res = ERR_CAST(fsc);
+		err = PTR_ERR(fsc);
 		goto out_final;
 	}
 
 	err = ceph_mdsc_init(fsc);
-	if (err < 0) {
-		res = ERR_PTR(err);
+	if (err < 0)
 		goto out;
-	}
 
 	if (ceph_test_opt(fsc->client, NOSHARE))
 		compare_super = NULL;
-	sb = sget(fs_type, compare_super, ceph_set_super, flags, fsc);
+
+	fc->s_fs_info = fsc;
+	sb = sget_fc(fc, compare_super, ceph_set_super);
+	fc->s_fs_info = NULL;
 	if (IS_ERR(sb)) {
-		res = ERR_CAST(sb);
+		err = PTR_ERR(sb);
 		goto out;
 	}
 
@@ -1123,18 +1053,19 @@ static struct dentry *ceph_mount(struct file_system_type *fs_type,
 	} else {
 		dout("get_sb using new client %p\n", fsc);
 		err = ceph_setup_bdi(sb, fsc);
-		if (err < 0) {
-			res = ERR_PTR(err);
+		if (err < 0)
 			goto out_splat;
-		}
 	}
 
-	res = ceph_real_mount(fsc);
-	if (IS_ERR(res))
+	res = ceph_real_mount(fsc, fc);
+	if (IS_ERR(res)) {
+		err = PTR_ERR(res);
 		goto out_splat;
+	}
 	dout("root %p inode %p ino %llx.%llx\n", res,
 	     d_inode(res), ceph_vinop(d_inode(res)));
-	return res;
+	fc->root = fsc->sb->s_root;
+	return 0;
 
 out_splat:
 	ceph_mdsc_close_sessions(fsc->mdsc);
@@ -1144,8 +1075,79 @@ out_splat:
 out:
 	destroy_fs_client(fsc);
 out_final:
-	dout("ceph_mount fail %ld\n", PTR_ERR(res));
-	return res;
+	dout("ceph_get_tree fail %d\n", err);
+	return err;
+}
+
+static void ceph_free_fc(struct fs_context *fc)
+{
+	struct ceph_parse_opts_ctx *pctx = fc->fs_private;
+
+	if (pctx) {
+		destroy_mount_options(pctx->opts);
+		ceph_destroy_options(pctx->copts);
+		kfree(pctx);
+	}
+}
+
+static int ceph_reconfigure_fc(struct fs_context *fc)
+{
+	sync_filesystem(fc->root->d_sb);
+	return 0;
+}
+
+static const struct fs_context_operations ceph_context_ops = {
+	.free		= ceph_free_fc,
+	.parse_param	= ceph_parse_mount_param,
+	.get_tree	= ceph_get_tree,
+	.reconfigure	= ceph_reconfigure_fc,
+};
+
+/*
+ * Set up the filesystem mount context.
+ */
+static int ceph_init_fs_context(struct fs_context *fc)
+{
+	struct ceph_parse_opts_ctx *pctx;
+	struct ceph_mount_options *fsopt;
+
+	pctx = kzalloc(sizeof(*pctx), GFP_KERNEL);
+	if (!pctx)
+		return -ENOMEM;
+
+	pctx->copts = ceph_alloc_options();
+	if (!pctx->copts)
+		goto nomem;
+
+	pctx->opts = kzalloc(sizeof(*pctx->opts), GFP_KERNEL);
+	if (!pctx->opts)
+		goto nomem;
+
+	fsopt = pctx->opts;
+	fsopt->flags = CEPH_MOUNT_OPT_DEFAULT;
+
+	fsopt->wsize = CEPH_MAX_WRITE_SIZE;
+	fsopt->rsize = CEPH_MAX_READ_SIZE;
+	fsopt->rasize = CEPH_RASIZE_DEFAULT;
+	fsopt->snapdir_name = kstrdup(CEPH_SNAPDIRNAME_DEFAULT, GFP_KERNEL);
+	if (!fsopt->snapdir_name)
+		goto nomem;
+
+	fsopt->caps_wanted_delay_min = CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT;
+	fsopt->caps_wanted_delay_max = CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT;
+	fsopt->max_readdir = CEPH_MAX_READDIR_DEFAULT;
+	fsopt->max_readdir_bytes = CEPH_MAX_READDIR_BYTES_DEFAULT;
+	fsopt->congestion_kb = default_congestion_kb();
+
+	fc->fs_private = pctx;
+	fc->ops = &ceph_context_ops;
+	return 0;
+
+nomem:
+	destroy_mount_options(pctx->opts);
+	ceph_destroy_options(pctx->copts);
+	kfree(pctx);
+	return -ENOMEM;
 }
 
 static void ceph_kill_sb(struct super_block *s)
@@ -1172,7 +1174,7 @@ static void ceph_kill_sb(struct super_block *s)
 static struct file_system_type ceph_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "ceph",
-	.mount		= ceph_mount,
+	.init_fs_context = ceph_init_fs_context,
 	.kill_sb	= ceph_kill_sb,
 	.fs_flags	= FS_RENAME_DOES_D_MOVE,
 };