Merge tag 'ceph-for-5.5-rc1' of git://github.com/ceph/ceph-client

Pull ceph updates from Ilya Dryomov:
 "The two highlights are a set of improvements to how rbd read-only
  mappings are handled and a conversion to the new mount API (slightly
  complicated by the fact that we had a common option parsing framework
  that called out into rbd and the filesystem instead of them calling
  into it).

  Also included a few scattered fixes and a MAINTAINERS update for rbd,
  adding Dongsheng as a reviewer"

* tag 'ceph-for-5.5-rc1' of git://github.com/ceph/ceph-client:
  libceph, rbd, ceph: convert to use the new mount API
  rbd: ask for a weaker incompat mask for read-only mappings
  rbd: don't query snapshot features
  rbd: remove snapshot existence validation code
  rbd: don't establish watch for read-only mappings
  rbd: don't acquire exclusive lock for read-only mappings
  rbd: disallow read-write partitions on images mapped read-only
  rbd: treat images mapped read-only seriously
  rbd: introduce RBD_DEV_FLAG_READONLY
  rbd: introduce rbd_is_snap()
  ceph: don't leave ino field in ceph_mds_request_head uninitialized
  ceph: tone down loglevel on ceph_mdsc_build_path warning
  rbd: update MAINTAINERS info
  ceph: fix geting random mds from mdsmap
  rbd: fix spelling mistake "requeueing" -> "requeuing"
  ceph: make several helper accessors take const pointers
  libceph: drop unnecessary check from dispatch() in mon_client.c

MAINTAINERS

@@ -13773,7 +13773,7 @@ F:	drivers/media/radio/radio-tea5777.c
 RADOS BLOCK DEVICE (RBD)
 M:	Ilya Dryomov <idryomov@gmail.com>
 M:	Sage Weil <sage@redhat.com>
-M:	Alex Elder <elder@kernel.org>
+R:	Dongsheng Yang <dongsheng.yang@easystack.cn>
 L:	ceph-devel@vger.kernel.org
 W:	http://ceph.com/
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client.git

drivers/block/rbd.c

@@ -34,7 +34,7 @@
 #include <linux/ceph/cls_lock_client.h>
 #include <linux/ceph/striper.h>
 #include <linux/ceph/decode.h>
-#include <linux/parser.h>
+#include <linux/fs_parser.h>
 #include <linux/bsearch.h>
 
 #include <linux/kernel.h>
@@ -377,7 +377,6 @@ struct rbd_client_id {
 
 struct rbd_mapping {
 	u64                     size;
-	u64                     features;
 };
 
 /*
@@ -462,8 +461,9 @@ struct rbd_device {
  *   by rbd_dev->lock
  */
 enum rbd_dev_flags {
-	RBD_DEV_FLAG_EXISTS,	/* mapped snapshot has not been deleted */
+	RBD_DEV_FLAG_EXISTS,	/* rbd_dev_device_setup() ran */
 	RBD_DEV_FLAG_REMOVING,	/* this mapping is being removed */
+	RBD_DEV_FLAG_READONLY,  /* -o ro or snapshot */
 };
 
 static DEFINE_MUTEX(client_mutex);	/* Serialize client creation */
@@ -514,6 +514,16 @@ static int minor_to_rbd_dev_id(int minor)
 	return minor >> RBD_SINGLE_MAJOR_PART_SHIFT;
 }
 
+static bool rbd_is_ro(struct rbd_device *rbd_dev)
+{
+	return test_bit(RBD_DEV_FLAG_READONLY, &rbd_dev->flags);
+}
+
+static bool rbd_is_snap(struct rbd_device *rbd_dev)
+{
+	return rbd_dev->spec->snap_id != CEPH_NOSNAP;
+}
+
 static bool __rbd_is_lock_owner(struct rbd_device *rbd_dev)
 {
 	lockdep_assert_held(&rbd_dev->lock_rwsem);
@@ -633,8 +643,6 @@ static const char *rbd_dev_v2_snap_name(struct rbd_device *rbd_dev,
 					u64 snap_id);
 static int _rbd_dev_v2_snap_size(struct rbd_device *rbd_dev, u64 snap_id,
 				u8 *order, u64 *snap_size);
-static int _rbd_dev_v2_snap_features(struct rbd_device *rbd_dev, u64 snap_id,
-		u64 *snap_features);
 static int rbd_dev_v2_get_flags(struct rbd_device *rbd_dev);
 
 static void rbd_obj_handle_request(struct rbd_obj_request *obj_req, int result);
@@ -695,9 +703,16 @@ static int rbd_ioctl_set_ro(struct rbd_device *rbd_dev, unsigned long arg)
 	if (get_user(ro, (int __user *)arg))
 		return -EFAULT;
 
-	/* Snapshots can't be marked read-write */
-	if (rbd_dev->spec->snap_id != CEPH_NOSNAP && !ro)
-		return -EROFS;
+	/*
+	 * Both images mapped read-only and snapshots can't be marked
+	 * read-write.
+	 */
+	if (!ro) {
+		if (rbd_is_ro(rbd_dev))
+			return -EROFS;
+
+		rbd_assert(!rbd_is_snap(rbd_dev));
+	}
 
 	/* Let blkdev_roset() handle it */
 	return -ENOTTY;
@@ -823,34 +838,34 @@ enum {
 	Opt_queue_depth,
 	Opt_alloc_size,
 	Opt_lock_timeout,
-	Opt_last_int,
 	/* int args above */
 	Opt_pool_ns,
-	Opt_last_string,
 	/* string args above */
 	Opt_read_only,
 	Opt_read_write,
 	Opt_lock_on_read,
 	Opt_exclusive,
 	Opt_notrim,
-	Opt_err
 };
 
-static match_table_t rbd_opts_tokens = {
-	{Opt_queue_depth, "queue_depth=%d"},
-	{Opt_alloc_size, "alloc_size=%d"},
-	{Opt_lock_timeout, "lock_timeout=%d"},
-	/* int args above */
-	{Opt_pool_ns, "_pool_ns=%s"},
-	/* string args above */
-	{Opt_read_only, "read_only"},
-	{Opt_read_only, "ro"},		/* Alternate spelling */
-	{Opt_read_write, "read_write"},
-	{Opt_read_write, "rw"},		/* Alternate spelling */
-	{Opt_lock_on_read, "lock_on_read"},
-	{Opt_exclusive, "exclusive"},
-	{Opt_notrim, "notrim"},
-	{Opt_err, NULL}
+static const struct fs_parameter_spec rbd_param_specs[] = {
+	fsparam_u32	("alloc_size",			Opt_alloc_size),
+	fsparam_flag	("exclusive",			Opt_exclusive),
+	fsparam_flag	("lock_on_read",		Opt_lock_on_read),
+	fsparam_u32	("lock_timeout",		Opt_lock_timeout),
+	fsparam_flag	("notrim",			Opt_notrim),
+	fsparam_string	("_pool_ns",			Opt_pool_ns),
+	fsparam_u32	("queue_depth",			Opt_queue_depth),
+	fsparam_flag	("read_only",			Opt_read_only),
+	fsparam_flag	("read_write",			Opt_read_write),
+	fsparam_flag	("ro",				Opt_read_only),
+	fsparam_flag	("rw",				Opt_read_write),
+	{}
+};
+
+static const struct fs_parameter_description rbd_parameters = {
+	.name		= "rbd",
+	.specs		= rbd_param_specs,
 };
 
 struct rbd_options {
@@ -871,87 +886,12 @@ struct rbd_options {
 #define RBD_EXCLUSIVE_DEFAULT	false
 #define RBD_TRIM_DEFAULT	true
 
-struct parse_rbd_opts_ctx {
+struct rbd_parse_opts_ctx {
 	struct rbd_spec		*spec;
+	struct ceph_options	*copts;
 	struct rbd_options	*opts;
 };
 
-static int parse_rbd_opts_token(char *c, void *private)
-{
-	struct parse_rbd_opts_ctx *pctx = private;
-	substring_t argstr[MAX_OPT_ARGS];
-	int token, intval, ret;
-
-	token = match_token(c, rbd_opts_tokens, argstr);
-	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;
-		}
-		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);
-	}
-
-	switch (token) {
-	case Opt_queue_depth:
-		if (intval < 1) {
-			pr_err("queue_depth out of range\n");
-			return -EINVAL;
-		}
-		pctx->opts->queue_depth = intval;
-		break;
-	case Opt_alloc_size:
-		if (intval < SECTOR_SIZE) {
-			pr_err("alloc_size out of range\n");
-			return -EINVAL;
-		}
-		if (!is_power_of_2(intval)) {
-			pr_err("alloc_size must be a power of 2\n");
-			return -EINVAL;
-		}
-		pctx->opts->alloc_size = intval;
-		break;
-	case Opt_lock_timeout:
-		/* 0 is "wait forever" (i.e. infinite timeout) */
-		if (intval < 0 || intval > INT_MAX / 1000) {
-			pr_err("lock_timeout out of range\n");
-			return -EINVAL;
-		}
-		pctx->opts->lock_timeout = msecs_to_jiffies(intval * 1000);
-		break;
-	case Opt_pool_ns:
-		kfree(pctx->spec->pool_ns);
-		pctx->spec->pool_ns = match_strdup(argstr);
-		if (!pctx->spec->pool_ns)
-			return -ENOMEM;
-		break;
-	case Opt_read_only:
-		pctx->opts->read_only = true;
-		break;
-	case Opt_read_write:
-		pctx->opts->read_only = false;
-		break;
-	case Opt_lock_on_read:
-		pctx->opts->lock_on_read = true;
-		break;
-	case Opt_exclusive:
-		pctx->opts->exclusive = true;
-		break;
-	case Opt_notrim:
-		pctx->opts->trim = false;
-		break;
-	default:
-		/* libceph prints "bad option" msg */
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
 static char* obj_op_name(enum obj_operation_type op_type)
 {
 	switch (op_type) {
@@ -1302,51 +1242,23 @@ static int rbd_snap_size(struct rbd_device *rbd_dev, u64 snap_id,
 	return 0;
 }
 
-static int rbd_snap_features(struct rbd_device *rbd_dev, u64 snap_id,
-			u64 *snap_features)
-{
-	rbd_assert(rbd_image_format_valid(rbd_dev->image_format));
-	if (snap_id == CEPH_NOSNAP) {
-		*snap_features = rbd_dev->header.features;
-	} else if (rbd_dev->image_format == 1) {
-		*snap_features = 0;	/* No features for format 1 */
-	} else {
-		u64 features = 0;
-		int ret;
-
-		ret = _rbd_dev_v2_snap_features(rbd_dev, snap_id, &features);
-		if (ret)
-			return ret;
-
-		*snap_features = features;
-	}
-	return 0;
-}
-
 static int rbd_dev_mapping_set(struct rbd_device *rbd_dev)
 {
 	u64 snap_id = rbd_dev->spec->snap_id;
 	u64 size = 0;
-	u64 features = 0;
 	int ret;
 
 	ret = rbd_snap_size(rbd_dev, snap_id, &size);
-	if (ret)
-		return ret;
-	ret = rbd_snap_features(rbd_dev, snap_id, &features);
 	if (ret)
 		return ret;
 
 	rbd_dev->mapping.size = size;
-	rbd_dev->mapping.features = features;
-
 	return 0;
 }
 
 static void rbd_dev_mapping_clear(struct rbd_device *rbd_dev)
 {
 	rbd_dev->mapping.size = 0;
-	rbd_dev->mapping.features = 0;
 }
 
 static void zero_bvec(struct bio_vec *bv)
@@ -1832,6 +1744,17 @@ static u8 rbd_object_map_get(struct rbd_device *rbd_dev, u64 objno)
 
 static bool use_object_map(struct rbd_device *rbd_dev)
 {
+	/*
+	 * An image mapped read-only can't use the object map -- it isn't
+	 * loaded because the header lock isn't acquired.  Someone else can
+	 * write to the image and update the object map behind our back.
+	 *
+	 * A snapshot can't be written to, so using the object map is always
+	 * safe.
+	 */
+	if (!rbd_is_snap(rbd_dev) && rbd_is_ro(rbd_dev))
+		return false;
+
 	return ((rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP) &&
 		!(rbd_dev->object_map_flags & RBD_FLAG_OBJECT_MAP_INVALID));
 }
@@ -3555,7 +3478,7 @@ static bool need_exclusive_lock(struct rbd_img_request *img_req)
 	if (!(rbd_dev->header.features & RBD_FEATURE_EXCLUSIVE_LOCK))
 		return false;
 
-	if (rbd_dev->spec->snap_id != CEPH_NOSNAP)
+	if (rbd_is_ro(rbd_dev))
 		return false;
 
 	rbd_assert(!test_bit(IMG_REQ_CHILD, &img_req->flags));
@@ -4230,7 +4153,7 @@ static void rbd_acquire_lock(struct work_struct *work)
 		 * lock owner acked, but resend if we don't see them
 		 * release the lock
 		 */
-		dout("%s rbd_dev %p requeueing lock_dwork\n", __func__,
+		dout("%s rbd_dev %p requeuing lock_dwork\n", __func__,
 		     rbd_dev);
 		mod_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork,
 		    msecs_to_jiffies(2 * RBD_NOTIFY_TIMEOUT * MSEC_PER_SEC));
@@ -4826,24 +4749,14 @@ static void rbd_queue_workfn(struct work_struct *work)
 		goto err_rq;
 	}
 
-	if (op_type != OBJ_OP_READ && rbd_dev->spec->snap_id != CEPH_NOSNAP) {
-		rbd_warn(rbd_dev, "%s on read-only snapshot",
-			 obj_op_name(op_type));
-		result = -EIO;
-		goto err;
-	}
-
-	/*
-	 * Quit early if the mapped snapshot no longer exists.  It's
-	 * still possible the snapshot will have disappeared by the
-	 * time our request arrives at the osd, but there's no sense in
-	 * sending it if we already know.
-	 */
-	if (!test_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags)) {
-		dout("request for non-existent snapshot");
-		rbd_assert(rbd_dev->spec->snap_id != CEPH_NOSNAP);
-		result = -ENXIO;
-		goto err_rq;
+	if (op_type != OBJ_OP_READ) {
+		if (rbd_is_ro(rbd_dev)) {
+			rbd_warn(rbd_dev, "%s on read-only mapping",
+				 obj_op_name(op_type));
+			result = -EIO;
+			goto err;
+		}
+		rbd_assert(!rbd_is_snap(rbd_dev));
 	}
 
 	if (offset && length > U64_MAX - offset + 1) {
@@ -5025,25 +4938,6 @@ static int rbd_dev_v1_header_info(struct rbd_device *rbd_dev)
 	return ret;
 }
 
-/*
- * Clear the rbd device's EXISTS flag if the snapshot it's mapped to
- * has disappeared from the (just updated) snapshot context.
- */
-static void rbd_exists_validate(struct rbd_device *rbd_dev)
-{
-	u64 snap_id;
-
-	if (!test_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags))
-		return;
-
-	snap_id = rbd_dev->spec->snap_id;
-	if (snap_id == CEPH_NOSNAP)
-		return;
-
-	if (rbd_dev_snap_index(rbd_dev, snap_id) == BAD_SNAP_INDEX)
-		clear_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags);
-}
-
 static void rbd_dev_update_size(struct rbd_device *rbd_dev)
 {
 	sector_t size;
@@ -5084,12 +4978,8 @@ static int rbd_dev_refresh(struct rbd_device *rbd_dev)
 			goto out;
 	}
 
-	if (rbd_dev->spec->snap_id == CEPH_NOSNAP) {
-		rbd_dev->mapping.size = rbd_dev->header.image_size;
-	} else {
-		/* validate mapped snapshot's EXISTS flag */
-		rbd_exists_validate(rbd_dev);
-	}
+	rbd_assert(!rbd_is_snap(rbd_dev));
+	rbd_dev->mapping.size = rbd_dev->header.image_size;
 
 out:
 	up_write(&rbd_dev->header_rwsem);
@@ -5211,17 +5101,12 @@ static ssize_t rbd_size_show(struct device *dev,
 		(unsigned long long)rbd_dev->mapping.size);
 }
 
-/*
- * Note this shows the features for whatever's mapped, which is not
- * necessarily the base image.
- */
 static ssize_t rbd_features_show(struct device *dev,
 			     struct device_attribute *attr, char *buf)
 {
 	struct rbd_device *rbd_dev = dev_to_rbd_dev(dev);
 
-	return sprintf(buf, "0x%016llx\n",
-			(unsigned long long)rbd_dev->mapping.features);
+	return sprintf(buf, "0x%016llx\n", rbd_dev->header.features);
 }
 
 static ssize_t rbd_major_show(struct device *dev,
@@ -5709,9 +5594,12 @@ static int rbd_dev_v2_object_prefix(struct rbd_device *rbd_dev)
 }
 
 static int _rbd_dev_v2_snap_features(struct rbd_device *rbd_dev, u64 snap_id,
-		u64 *snap_features)
+				     bool read_only, u64 *snap_features)
 {
-	__le64 snapid = cpu_to_le64(snap_id);
+	struct {
+		__le64 snap_id;
+		u8 read_only;
+	} features_in;
 	struct {
 		__le64 features;
 		__le64 incompat;
@@ -5719,9 +5607,12 @@ static int _rbd_dev_v2_snap_features(struct rbd_device *rbd_dev, u64 snap_id,
 	u64 unsup;
 	int ret;
 
+	features_in.snap_id = cpu_to_le64(snap_id);
+	features_in.read_only = read_only;
+
 	ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid,
 				  &rbd_dev->header_oloc, "get_features",
-				  &snapid, sizeof(snapid),
+				  &features_in, sizeof(features_in),
 				  &features_buf, sizeof(features_buf));
 	dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret);
 	if (ret < 0)
@@ -5749,7 +5640,8 @@ static int _rbd_dev_v2_snap_features(struct rbd_device *rbd_dev, u64 snap_id,
 static int rbd_dev_v2_features(struct rbd_device *rbd_dev)
 {
 	return _rbd_dev_v2_snap_features(rbd_dev, CEPH_NOSNAP,
-						&rbd_dev->header.features);
+					 rbd_is_ro(rbd_dev),
+					 &rbd_dev->header.features);
 }
 
 /*
@@ -6456,6 +6348,122 @@ static inline char *dup_token(const char **buf, size_t *lenp)
 	return dup;
 }
 
+static int rbd_parse_param(struct fs_parameter *param,
+			    struct rbd_parse_opts_ctx *pctx)
+{
+	struct rbd_options *opt = pctx->opts;
+	struct fs_parse_result result;
+	int token, ret;
+
+	ret = ceph_parse_param(param, pctx->copts, NULL);
+	if (ret != -ENOPARAM)
+		return ret;
+
+	token = fs_parse(NULL, &rbd_parameters, param, &result);
+	dout("%s fs_parse '%s' token %d\n", __func__, param->key, token);
+	if (token < 0) {
+		if (token == -ENOPARAM) {
+			return invalf(NULL, "rbd: Unknown parameter '%s'",
+				      param->key);
+		}
+		return token;
+	}
+
+	switch (token) {
+	case Opt_queue_depth:
+		if (result.uint_32 < 1)
+			goto out_of_range;
+		opt->queue_depth = result.uint_32;
+		break;
+	case Opt_alloc_size:
+		if (result.uint_32 < SECTOR_SIZE)
+			goto out_of_range;
+		if (!is_power_of_2(result.uint_32)) {
+			return invalf(NULL, "rbd: alloc_size must be a power of 2");
+		}
+		opt->alloc_size = result.uint_32;
+		break;
+	case Opt_lock_timeout:
+		/* 0 is "wait forever" (i.e. infinite timeout) */
+		if (result.uint_32 > INT_MAX / 1000)
+			goto out_of_range;
+		opt->lock_timeout = msecs_to_jiffies(result.uint_32 * 1000);
+		break;
+	case Opt_pool_ns:
+		kfree(pctx->spec->pool_ns);
+		pctx->spec->pool_ns = param->string;
+		param->string = NULL;
+		break;
+	case Opt_read_only:
+		opt->read_only = true;
+		break;
+	case Opt_read_write:
+		opt->read_only = false;
+		break;
+	case Opt_lock_on_read:
+		opt->lock_on_read = true;
+		break;
+	case Opt_exclusive:
+		opt->exclusive = true;
+		break;
+	case Opt_notrim:
+		opt->trim = false;
+		break;
+	default:
+		BUG();
+	}
+
+	return 0;
+
+out_of_range:
+	return invalf(NULL, "rbd: %s out of range", param->key);
+}
+
+/*
+ * This duplicates most of generic_parse_monolithic(), untying it from
+ * fs_context and skipping standard superblock and security options.
+ */
+static int rbd_parse_options(char *options, struct rbd_parse_opts_ctx *pctx)
+{
+	char *key;
+	int ret = 0;
+
+	dout("%s '%s'\n", __func__, options);
+	while ((key = strsep(&options, ",")) != NULL) {
+		if (*key) {
+			struct fs_parameter param = {
+				.key	= key,
+				.type	= fs_value_is_string,
+			};
+			char *value = strchr(key, '=');
+			size_t v_len = 0;
+
+			if (value) {
+				if (value == key)
+					continue;
+				*value++ = 0;
+				v_len = strlen(value);
+			}
+
+
+			if (v_len > 0) {
+				param.string = kmemdup_nul(value, v_len,
+							   GFP_KERNEL);
+				if (!param.string)
+					return -ENOMEM;
+			}
+			param.size = v_len;
+
+			ret = rbd_parse_param(&param, pctx);
+			kfree(param.string);
+			if (ret)
+				break;
+		}
+	}
+
+	return ret;
+}
+
 /*
  * Parse the options provided for an "rbd add" (i.e., rbd image
  * mapping) request.  These arrive via a write to /sys/bus/rbd/add,
@@ -6507,8 +6515,7 @@ static int rbd_add_parse_args(const char *buf,
 	const char *mon_addrs;
 	char *snap_name;
 	size_t mon_addrs_size;
-	struct parse_rbd_opts_ctx pctx = { 0 };
-	struct ceph_options *copts;
+	struct rbd_parse_opts_ctx pctx = { 0 };
 	int ret;
 
 	/* The first four tokens are required */
@@ -6519,7 +6526,7 @@ static int rbd_add_parse_args(const char *buf,
 		return -EINVAL;
 	}
 	mon_addrs = buf;
-	mon_addrs_size = len + 1;
+	mon_addrs_size = len;
 	buf += len;
 
 	ret = -EINVAL;
@@ -6569,6 +6576,10 @@ static int rbd_add_parse_args(const char *buf,
 	*(snap_name + len) = '\0';
 	pctx.spec->snap_name = snap_name;
 
+	pctx.copts = ceph_alloc_options();
+	if (!pctx.copts)
+		goto out_mem;
+
 	/* Initialize all rbd options to the defaults */
 
 	pctx.opts = kzalloc(sizeof(*pctx.opts), GFP_KERNEL);
@@ -6583,27 +6594,27 @@ static int rbd_add_parse_args(const char *buf,
 	pctx.opts->exclusive = RBD_EXCLUSIVE_DEFAULT;
 	pctx.opts->trim = RBD_TRIM_DEFAULT;
 
-	copts = ceph_parse_options(options, mon_addrs,
-				   mon_addrs + mon_addrs_size - 1,
-				   parse_rbd_opts_token, &pctx);
-	if (IS_ERR(copts)) {
-		ret = PTR_ERR(copts);
+	ret = ceph_parse_mon_ips(mon_addrs, mon_addrs_size, pctx.copts, NULL);
+	if (ret)
+		goto out_err;
+
+	ret = rbd_parse_options(options, &pctx);
+	if (ret)
 		goto out_err;
-	}
-	kfree(options);
 
-	*ceph_opts = copts;
+	*ceph_opts = pctx.copts;
 	*opts = pctx.opts;
 	*rbd_spec = pctx.spec;
-
+	kfree(options);
 	return 0;
+
 out_mem:
 	ret = -ENOMEM;
 out_err:
 	kfree(pctx.opts);
+	ceph_destroy_options(pctx.copts);
 	rbd_spec_put(pctx.spec);
 	kfree(options);
-
 	return ret;
 }
 
@@ -6632,7 +6643,7 @@ static int rbd_add_acquire_lock(struct rbd_device *rbd_dev)
 		return -EINVAL;
 	}
 
-	if (rbd_dev->spec->snap_id != CEPH_NOSNAP)
+	if (rbd_is_ro(rbd_dev))
 		return 0;
 
 	rbd_assert(!rbd_is_lock_owner(rbd_dev));
@@ -6838,6 +6849,8 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev, int depth)
 	__rbd_get_client(rbd_dev->rbd_client);
 	rbd_spec_get(rbd_dev->parent_spec);
 
+	__set_bit(RBD_DEV_FLAG_READONLY, &parent->flags);
+
 	ret = rbd_dev_image_probe(parent, depth);
 	if (ret < 0)
 		goto out_err;
@@ -6889,7 +6902,7 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev)
 		goto err_out_blkdev;
 
 	set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE);
-	set_disk_ro(rbd_dev->disk, rbd_dev->opts->read_only);
+	set_disk_ro(rbd_dev->disk, rbd_is_ro(rbd_dev));
 
 	ret = dev_set_name(&rbd_dev->dev, "%d", rbd_dev->dev_id);
 	if (ret)
@@ -6927,6 +6940,24 @@ static int rbd_dev_header_name(struct rbd_device *rbd_dev)
 	return ret;
 }
 
+static void rbd_print_dne(struct rbd_device *rbd_dev, bool is_snap)
+{
+	if (!is_snap) {
+		pr_info("image %s/%s%s%s does not exist\n",
+			rbd_dev->spec->pool_name,
+			rbd_dev->spec->pool_ns ?: "",
+			rbd_dev->spec->pool_ns ? "/" : "",
+			rbd_dev->spec->image_name);
+	} else {
+		pr_info("snap %s/%s%s%s@%s does not exist\n",
+			rbd_dev->spec->pool_name,
+			rbd_dev->spec->pool_ns ?: "",
+			rbd_dev->spec->pool_ns ? "/" : "",
+			rbd_dev->spec->image_name,
+			rbd_dev->spec->snap_name);
+	}
+}
+
 static void rbd_dev_image_release(struct rbd_device *rbd_dev)
 {
 	rbd_dev_unprobe(rbd_dev);
@@ -6945,6 +6976,7 @@ static void rbd_dev_image_release(struct rbd_device *rbd_dev)
  */
 static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth)
 {
+	bool need_watch = !rbd_is_ro(rbd_dev);
 	int ret;
 
 	/*
@@ -6961,22 +6993,21 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth)
 	if (ret)
 		goto err_out_format;
 
-	if (!depth) {
+	if (need_watch) {
 		ret = rbd_register_watch(rbd_dev);
 		if (ret) {
 			if (ret == -ENOENT)
-				pr_info("image %s/%s%s%s does not exist\n",
-					rbd_dev->spec->pool_name,
-					rbd_dev->spec->pool_ns ?: "",
-					rbd_dev->spec->pool_ns ? "/" : "",
-					rbd_dev->spec->image_name);
+				rbd_print_dne(rbd_dev, false);
 			goto err_out_format;
 		}
 	}
 
 	ret = rbd_dev_header_info(rbd_dev);
-	if (ret)
+	if (ret) {
+		if (ret == -ENOENT && !need_watch)
+			rbd_print_dne(rbd_dev, false);
 		goto err_out_watch;
+	}
 
 	/*
 	 * If this image is the one being mapped, we have pool name and
@@ -6990,12 +7021,7 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth)
 		ret = rbd_spec_fill_names(rbd_dev);
 	if (ret) {
 		if (ret == -ENOENT)
-			pr_info("snap %s/%s%s%s@%s does not exist\n",
-				rbd_dev->spec->pool_name,
-				rbd_dev->spec->pool_ns ?: "",
-				rbd_dev->spec->pool_ns ? "/" : "",
-				rbd_dev->spec->image_name,
-				rbd_dev->spec->snap_name);
+			rbd_print_dne(rbd_dev, true);
 		goto err_out_probe;
 	}
 
@@ -7003,7 +7029,7 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth)
 	if (ret)
 		goto err_out_probe;
 
-	if (rbd_dev->spec->snap_id != CEPH_NOSNAP &&
+	if (rbd_is_snap(rbd_dev) &&
 	    (rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP)) {
 		ret = rbd_object_map_load(rbd_dev);
 		if (ret)
@@ -7027,7 +7053,7 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth)
 err_out_probe:
 	rbd_dev_unprobe(rbd_dev);
 err_out_watch:
-	if (!depth)
+	if (need_watch)
 		rbd_unregister_watch(rbd_dev);
 err_out_format:
 	rbd_dev->image_format = 0;
@@ -7079,6 +7105,11 @@ static ssize_t do_rbd_add(struct bus_type *bus,
 	spec = NULL;		/* rbd_dev now owns this */
 	rbd_opts = NULL;	/* rbd_dev now owns this */
 
+	/* if we are mapping a snapshot it will be a read-only mapping */
+	if (rbd_dev->opts->read_only ||
+	    strcmp(rbd_dev->spec->snap_name, RBD_SNAP_HEAD_NAME))
+		__set_bit(RBD_DEV_FLAG_READONLY, &rbd_dev->flags);
+
 	rbd_dev->config_info = kstrdup(buf, GFP_KERNEL);
 	if (!rbd_dev->config_info) {
 		rc = -ENOMEM;
@@ -7092,10 +7123,6 @@ static ssize_t do_rbd_add(struct bus_type *bus,
 		goto err_out_rbd_dev;
 	}
 
-	/* If we are mapping a snapshot it must be marked read-only */
-	if (rbd_dev->spec->snap_id != CEPH_NOSNAP)
-		rbd_dev->opts->read_only = true;
-
 	if (rbd_dev->opts->alloc_size > rbd_dev->layout.object_size) {
 		rbd_warn(rbd_dev, "alloc_size adjusted to %u",
 			 rbd_dev->layout.object_size);

fs/ceph/cache.c

@@ -8,6 +8,7 @@
 
 #include <linux/ceph/ceph_debug.h>
 
+#include <linux/fs_context.h>
 #include "super.h"
 #include "cache.h"
 
@@ -49,7 +50,7 @@ void ceph_fscache_unregister(void)
 	fscache_unregister_netfs(&ceph_cache_netfs);
 }
 
-int ceph_fscache_register_fs(struct ceph_fs_client* fsc)
+int ceph_fscache_register_fs(struct ceph_fs_client* fsc, struct fs_context *fc)
 {
 	const struct ceph_fsid *fsid = &fsc->client->fsid;
 	const char *fscache_uniq = fsc->mount_options->fscache_uniq;
@@ -66,8 +67,8 @@ int ceph_fscache_register_fs(struct ceph_fs_client* fsc)
 		if (uniq_len && memcmp(ent->uniquifier, fscache_uniq, uniq_len))
 			continue;
 
-		pr_err("fscache cookie already registered for fsid %pU\n", fsid);
-		pr_err("  use fsc=%%s mount option to specify a uniquifier\n");
+		errorf(fc, "ceph: fscache cookie already registered for fsid %pU, use fsc=<uniquifier> option",
+		       fsid);
 		err = -EBUSY;
 		goto out_unlock;
 	}
@@ -95,7 +96,7 @@ int ceph_fscache_register_fs(struct ceph_fs_client* fsc)
 		list_add_tail(&ent->list, &ceph_fscache_list);
 	} else {
 		kfree(ent);
-		pr_err("unable to register fscache cookie for fsid %pU\n",
+		errorf(fc, "ceph: unable to register fscache cookie for fsid %pU",
 		       fsid);
 		/* all other fs ignore this error */
 	}

fs/ceph/cache.h

@@ -16,7 +16,7 @@ extern struct fscache_netfs ceph_cache_netfs;
 int ceph_fscache_register(void);
 void ceph_fscache_unregister(void);
 
-int ceph_fscache_register_fs(struct ceph_fs_client* fsc);
+int ceph_fscache_register_fs(struct ceph_fs_client* fsc, struct fs_context *fc);
 void ceph_fscache_unregister_fs(struct ceph_fs_client* fsc);
 
 void ceph_fscache_register_inode_cookie(struct inode *inode);
@@ -88,7 +88,8 @@ static inline void ceph_fscache_unregister(void)
 {
 }
 
-static inline int ceph_fscache_register_fs(struct ceph_fs_client* fsc)
+static inline int ceph_fscache_register_fs(struct ceph_fs_client* fsc,
+					   struct fs_context *fc)
 {
 	return 0;
 }

fs/ceph/mds_client.c

@@ -2182,13 +2182,17 @@ char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *pbase,
 	}
 	base = ceph_ino(d_inode(temp));
 	rcu_read_unlock();
-	if (pos < 0 || read_seqretry(&rename_lock, seq)) {
-		pr_err("build_path did not end path lookup where "
-		       "expected, pos is %d\n", pos);
-		/* presumably this is only possible if racing with a
-		   rename of one of the parent directories (we can not
-		   lock the dentries above us to prevent this, but
-		   retrying should be harmless) */
+
+	if (read_seqretry(&rename_lock, seq))
+		goto retry;
+
+	if (pos < 0) {
+		/*
+		 * A rename didn't occur, but somehow we didn't end up where
+		 * we thought we would. Throw a warning and try again.
+		 */
+		pr_warn("build_path did not end path lookup where "
+			"expected, pos is %d\n", pos);
 		goto retry;
 	}
 
@@ -2345,6 +2349,7 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc,
 	head->op = cpu_to_le32(req->r_op);
 	head->caller_uid = cpu_to_le32(from_kuid(&init_user_ns, req->r_uid));
 	head->caller_gid = cpu_to_le32(from_kgid(&init_user_ns, req->r_gid));
+	head->ino = 0;
 	head->args = req->r_args;
 
 	ceph_encode_filepath(&p, end, ino1, path1);

fs/ceph/mdsmap.c

@@ -20,7 +20,7 @@
 int ceph_mdsmap_get_random_mds(struct ceph_mdsmap *m)
 {
 	int n = 0;
-	int i;
+	int i, j;
 
 	/* special case for one mds */
 	if (1 == m->m_num_mds && m->m_info[0].state > 0)
@@ -35,9 +35,12 @@ int ceph_mdsmap_get_random_mds(struct ceph_mdsmap *m)
 
 	/* pick */
 	n = prandom_u32() % n;
-	for (i = 0; n > 0; i++, n--)
-		while (m->m_info[i].state <= 0)
-			i++;
+	for (j = 0, i = 0; i < m->m_num_mds; i++) {
+		if (m->m_info[i].state > 0)
+			j++;
+		if (j > n)
+			break;
+	}
 
 	return i;
 }

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 @@ static struct dentry *open_root_dentry(struct ceph_fs_client *fsc,
 /*
  * 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 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc)
 	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 @@ static struct dentry *ceph_mount(struct file_system_type *fs_type,
 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,
 };

fs/ceph/super.h

@@ -74,7 +74,6 @@
 
 struct ceph_mount_options {
 	int flags;
-	int sb_flags;
 
 	int wsize;            /* max write size */
 	int rsize;            /* max read size */
@@ -407,22 +406,26 @@ struct ceph_inode_info {
 	struct inode vfs_inode; /* at end */
 };
 
-static inline struct ceph_inode_info *ceph_inode(struct inode *inode)
+static inline struct ceph_inode_info *
+ceph_inode(const struct inode *inode)
 {
 	return container_of(inode, struct ceph_inode_info, vfs_inode);
 }
 
-static inline struct ceph_fs_client *ceph_inode_to_client(struct inode *inode)
+static inline struct ceph_fs_client *
+ceph_inode_to_client(const struct inode *inode)
 {
 	return (struct ceph_fs_client *)inode->i_sb->s_fs_info;
 }
 
-static inline struct ceph_fs_client *ceph_sb_to_client(struct super_block *sb)
+static inline struct ceph_fs_client *
+ceph_sb_to_client(const struct super_block *sb)
 {
 	return (struct ceph_fs_client *)sb->s_fs_info;
 }
 
-static inline struct ceph_vino ceph_vino(struct inode *inode)
+static inline struct ceph_vino
+ceph_vino(const struct inode *inode)
 {
 	return ceph_inode(inode)->i_vino;
 }

include/linux/ceph/libceph.h

@@ -280,10 +280,12 @@ extern const char *ceph_msg_type_name(int type);
 extern int ceph_check_fsid(struct ceph_client *client, struct ceph_fsid *fsid);
 extern void *ceph_kvmalloc(size_t size, gfp_t flags);
 
-extern struct ceph_options *ceph_parse_options(char *options,
-			      const char *dev_name, const char *dev_name_end,
-			      int (*parse_extra_token)(char *c, void *private),
-			      void *private);
+struct fs_parameter;
+struct ceph_options *ceph_alloc_options(void);
+int ceph_parse_mon_ips(const char *buf, size_t len, struct ceph_options *opt,
+		       struct fs_context *fc);
+int ceph_parse_param(struct fs_parameter *param, struct ceph_options *opt,
+		     struct fs_context *fc);
 int ceph_print_client_options(struct seq_file *m, struct ceph_client *client,
 			      bool show_all);
 extern void ceph_destroy_options(struct ceph_options *opt);

net/ceph/ceph_common.c

@@ -11,7 +11,7 @@
 #include <linux/module.h>
 #include <linux/mount.h>
 #include <linux/nsproxy.h>
-#include <linux/parser.h>
+#include <linux/fs_parser.h>
 #include <linux/sched.h>
 #include <linux/sched/mm.h>
 #include <linux/seq_file.h>
@@ -254,58 +254,77 @@ enum {
 	Opt_mount_timeout,
 	Opt_osd_idle_ttl,
 	Opt_osd_request_timeout,
-	Opt_last_int,
 	/* int args above */
 	Opt_fsid,
 	Opt_name,
 	Opt_secret,
 	Opt_key,
 	Opt_ip,
-	Opt_last_string,
 	/* string args above */
 	Opt_share,
-	Opt_noshare,
 	Opt_crc,
-	Opt_nocrc,
 	Opt_cephx_require_signatures,
-	Opt_nocephx_require_signatures,
 	Opt_cephx_sign_messages,
-	Opt_nocephx_sign_messages,
 	Opt_tcp_nodelay,
-	Opt_notcp_nodelay,
 	Opt_abort_on_full,
 };
 
-static match_table_t opt_tokens = {
-	{Opt_osdtimeout, "osdtimeout=%d"},
-	{Opt_osdkeepalivetimeout, "osdkeepalive=%d"},
-	{Opt_mount_timeout, "mount_timeout=%d"},
-	{Opt_osd_idle_ttl, "osd_idle_ttl=%d"},
-	{Opt_osd_request_timeout, "osd_request_timeout=%d"},
-	/* int args above */
-	{Opt_fsid, "fsid=%s"},
-	{Opt_name, "name=%s"},
-	{Opt_secret, "secret=%s"},
-	{Opt_key, "key=%s"},
-	{Opt_ip, "ip=%s"},
-	/* string args above */
-	{Opt_share, "share"},
-	{Opt_noshare, "noshare"},
-	{Opt_crc, "crc"},
-	{Opt_nocrc, "nocrc"},
-	{Opt_cephx_require_signatures, "cephx_require_signatures"},
-	{Opt_nocephx_require_signatures, "nocephx_require_signatures"},
-	{Opt_cephx_sign_messages, "cephx_sign_messages"},
-	{Opt_nocephx_sign_messages, "nocephx_sign_messages"},
-	{Opt_tcp_nodelay, "tcp_nodelay"},
-	{Opt_notcp_nodelay, "notcp_nodelay"},
-	{Opt_abort_on_full, "abort_on_full"},
-	{-1, NULL}
+static const struct fs_parameter_spec ceph_param_specs[] = {
+	fsparam_flag	("abort_on_full",		Opt_abort_on_full),
+	fsparam_flag_no ("cephx_require_signatures",	Opt_cephx_require_signatures),
+	fsparam_flag_no ("cephx_sign_messages",		Opt_cephx_sign_messages),
+	fsparam_flag_no ("crc",				Opt_crc),
+	fsparam_string	("fsid",			Opt_fsid),
+	fsparam_string	("ip",				Opt_ip),
+	fsparam_string	("key",				Opt_key),
+	fsparam_u32	("mount_timeout",		Opt_mount_timeout),
+	fsparam_string	("name",			Opt_name),
+	fsparam_u32	("osd_idle_ttl",		Opt_osd_idle_ttl),
+	fsparam_u32	("osd_request_timeout",		Opt_osd_request_timeout),
+	fsparam_u32	("osdkeepalive",		Opt_osdkeepalivetimeout),
+	__fsparam	(fs_param_is_s32, "osdtimeout", Opt_osdtimeout,
+			 fs_param_deprecated),
+	fsparam_string	("secret",			Opt_secret),
+	fsparam_flag_no ("share",			Opt_share),
+	fsparam_flag_no ("tcp_nodelay",			Opt_tcp_nodelay),
+	{}
+};
+
+static const struct fs_parameter_description ceph_parameters = {
+        .name           = "libceph",
+        .specs          = ceph_param_specs,
 };
 
+struct ceph_options *ceph_alloc_options(void)
+{
+	struct ceph_options *opt;
+
+	opt = kzalloc(sizeof(*opt), GFP_KERNEL);
+	if (!opt)
+		return NULL;
+
+	opt->mon_addr = kcalloc(CEPH_MAX_MON, sizeof(*opt->mon_addr),
+				GFP_KERNEL);
+	if (!opt->mon_addr) {
+		kfree(opt);
+		return NULL;
+	}
+
+	opt->flags = CEPH_OPT_DEFAULT;
+	opt->osd_keepalive_timeout = CEPH_OSD_KEEPALIVE_DEFAULT;
+	opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT;
+	opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT;
+	opt->osd_request_timeout = CEPH_OSD_REQUEST_TIMEOUT_DEFAULT;
+	return opt;
+}
+EXPORT_SYMBOL(ceph_alloc_options);
+
 void ceph_destroy_options(struct ceph_options *opt)
 {
 	dout("destroy_options %p\n", opt);
+	if (!opt)
+		return;
+
 	kfree(opt->name);
 	if (opt->key) {
 		ceph_crypto_key_destroy(opt->key);
@@ -317,7 +336,9 @@ void ceph_destroy_options(struct ceph_options *opt)
 EXPORT_SYMBOL(ceph_destroy_options);
 
 /* get secret from key store */
-static int get_secret(struct ceph_crypto_key *dst, const char *name) {
+static int get_secret(struct ceph_crypto_key *dst, const char *name,
+		      struct fs_context *fc)
+{
 	struct key *ukey;
 	int key_err;
 	int err = 0;
@@ -330,20 +351,20 @@ static int get_secret(struct ceph_crypto_key *dst, const char *name) {
 		key_err = PTR_ERR(ukey);
 		switch (key_err) {
 		case -ENOKEY:
-			pr_warn("ceph: Mount failed due to key not found: %s\n",
-				name);
+			errorf(fc, "libceph: Failed due to key not found: %s",
+			       name);
 			break;
 		case -EKEYEXPIRED:
-			pr_warn("ceph: Mount failed due to expired key: %s\n",
-				name);
+			errorf(fc, "libceph: Failed due to expired key: %s",
+			       name);
 			break;
 		case -EKEYREVOKED:
-			pr_warn("ceph: Mount failed due to revoked key: %s\n",
-				name);
+			errorf(fc, "libceph: Failed due to revoked key: %s",
+			       name);
 			break;
 		default:
-			pr_warn("ceph: Mount failed due to unknown key error %d: %s\n",
-				key_err, name);
+			errorf(fc, "libceph: Failed due to key error %d: %s",
+			       key_err, name);
 		}
 		err = -EPERM;
 		goto out;
@@ -361,217 +382,157 @@ static int get_secret(struct ceph_crypto_key *dst, const char *name) {
 	return err;
 }
 
-struct ceph_options *
-ceph_parse_options(char *options, const char *dev_name,
-			const char *dev_name_end,
-			int (*parse_extra_token)(char *c, void *private),
-			void *private)
+int ceph_parse_mon_ips(const char *buf, size_t len, struct ceph_options *opt,
+		       struct fs_context *fc)
 {
-	struct ceph_options *opt;
-	const char *c;
-	int err = -ENOMEM;
-	substring_t argstr[MAX_OPT_ARGS];
-
-	opt = kzalloc(sizeof(*opt), GFP_KERNEL);
-	if (!opt)
-		return ERR_PTR(-ENOMEM);
-	opt->mon_addr = kcalloc(CEPH_MAX_MON, sizeof(*opt->mon_addr),
-				GFP_KERNEL);
-	if (!opt->mon_addr)
-		goto out;
-
-	dout("parse_options %p options '%s' dev_name '%s'\n", opt, options,
-	     dev_name);
-
-	/* start with defaults */
-	opt->flags = CEPH_OPT_DEFAULT;
-	opt->osd_keepalive_timeout = CEPH_OSD_KEEPALIVE_DEFAULT;
-	opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT;
-	opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT;
-	opt->osd_request_timeout = CEPH_OSD_REQUEST_TIMEOUT_DEFAULT;
+	int ret;
 
-	/* get mon ip(s) */
 	/* ip1[:port1][,ip2[:port2]...] */
-	err = ceph_parse_ips(dev_name, dev_name_end, opt->mon_addr,
-			     CEPH_MAX_MON, &opt->num_mon);
-	if (err < 0)
-		goto out;
+	ret = ceph_parse_ips(buf, buf + len, opt->mon_addr, CEPH_MAX_MON,
+			     &opt->num_mon);
+	if (ret) {
+		errorf(fc, "libceph: Failed to parse monitor IPs: %d", ret);
+		return ret;
+	}
 
-	/* parse mount options */
-	while ((c = strsep(&options, ",")) != NULL) {
-		int token, intval;
-		if (!*c)
-			continue;
-		err = -EINVAL;
-		token = match_token((char *)c, opt_tokens, argstr);
-		if (token < 0 && parse_extra_token) {
-			/* extra? */
-			err = parse_extra_token((char *)c, private);
-			if (err < 0) {
-				pr_err("bad option at '%s'\n", c);
-				goto out;
-			}
-			continue;
-		}
-		if (token < Opt_last_int) {
-			err = match_int(&argstr[0], &intval);
-			if (err < 0) {
-				pr_err("bad option arg (not int) at '%s'\n", c);
-				goto out;
-			}
-			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);
+	return 0;
+}
+EXPORT_SYMBOL(ceph_parse_mon_ips);
+
+int ceph_parse_param(struct fs_parameter *param, struct ceph_options *opt,
+		     struct fs_context *fc)
+{
+	struct fs_parse_result result;
+	int token, err;
+
+	token = fs_parse(fc, &ceph_parameters, param, &result);
+	dout("%s fs_parse '%s' token %d\n", __func__, param->key, token);
+	if (token < 0)
+		return token;
+
+	switch (token) {
+	case Opt_ip:
+		err = ceph_parse_ips(param->string,
+				     param->string + param->size,
+				     &opt->my_addr,
+				     1, NULL);
+		if (err) {
+			errorf(fc, "libceph: Failed to parse ip: %d", err);
+			return err;
 		}
-		switch (token) {
-		case Opt_ip:
-			err = ceph_parse_ips(argstr[0].from,
-					     argstr[0].to,
-					     &opt->my_addr,
-					     1, NULL);
-			if (err < 0)
-				goto out;
-			opt->flags |= CEPH_OPT_MYIP;
-			break;
+		opt->flags |= CEPH_OPT_MYIP;
+		break;
 
-		case Opt_fsid:
-			err = parse_fsid(argstr[0].from, &opt->fsid);
-			if (err == 0)
-				opt->flags |= CEPH_OPT_FSID;
-			break;
-		case Opt_name:
-			kfree(opt->name);
-			opt->name = kstrndup(argstr[0].from,
-					      argstr[0].to-argstr[0].from,
-					      GFP_KERNEL);
-			if (!opt->name) {
-				err = -ENOMEM;
-				goto out;
-			}
-			break;
-		case Opt_secret:
-			ceph_crypto_key_destroy(opt->key);
-			kfree(opt->key);
-
-		        opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
-			if (!opt->key) {
-				err = -ENOMEM;
-				goto out;
-			}
-			err = ceph_crypto_key_unarmor(opt->key, argstr[0].from);
-			if (err < 0)
-				goto out;
-			break;
-		case Opt_key:
-			ceph_crypto_key_destroy(opt->key);
-			kfree(opt->key);
-
-		        opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
-			if (!opt->key) {
-				err = -ENOMEM;
-				goto out;
-			}
-			err = get_secret(opt->key, argstr[0].from);
-			if (err < 0)
-				goto out;
-			break;
+	case Opt_fsid:
+		err = parse_fsid(param->string, &opt->fsid);
+		if (err) {
+			errorf(fc, "libceph: Failed to parse fsid: %d", err);
+			return err;
+		}
+		opt->flags |= CEPH_OPT_FSID;
+		break;
+	case Opt_name:
+		kfree(opt->name);
+		opt->name = param->string;
+		param->string = NULL;
+		break;
+	case Opt_secret:
+		ceph_crypto_key_destroy(opt->key);
+		kfree(opt->key);
 
-			/* misc */
-		case Opt_osdtimeout:
-			pr_warn("ignoring deprecated osdtimeout option\n");
-			break;
-		case Opt_osdkeepalivetimeout:
-			/* 0 isn't well defined right now, reject it */
-			if (intval < 1 || intval > INT_MAX / 1000) {
-				pr_err("osdkeepalive out of range\n");
-				err = -EINVAL;
-				goto out;
-			}
-			opt->osd_keepalive_timeout =
-					msecs_to_jiffies(intval * 1000);
-			break;
-		case Opt_osd_idle_ttl:
-			/* 0 isn't well defined right now, reject it */
-			if (intval < 1 || intval > INT_MAX / 1000) {
-				pr_err("osd_idle_ttl out of range\n");
-				err = -EINVAL;
-				goto out;
-			}
-			opt->osd_idle_ttl = msecs_to_jiffies(intval * 1000);
-			break;
-		case Opt_mount_timeout:
-			/* 0 is "wait forever" (i.e. infinite timeout) */
-			if (intval < 0 || intval > INT_MAX / 1000) {
-				pr_err("mount_timeout out of range\n");
-				err = -EINVAL;
-				goto out;
-			}
-			opt->mount_timeout = msecs_to_jiffies(intval * 1000);
-			break;
-		case Opt_osd_request_timeout:
-			/* 0 is "wait forever" (i.e. infinite timeout) */
-			if (intval < 0 || intval > INT_MAX / 1000) {
-				pr_err("osd_request_timeout out of range\n");
-				err = -EINVAL;
-				goto out;
-			}
-			opt->osd_request_timeout = msecs_to_jiffies(intval * 1000);
-			break;
+		opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
+		if (!opt->key)
+			return -ENOMEM;
+		err = ceph_crypto_key_unarmor(opt->key, param->string);
+		if (err) {
+			errorf(fc, "libceph: Failed to parse secret: %d", err);
+			return err;
+		}
+		break;
+	case Opt_key:
+		ceph_crypto_key_destroy(opt->key);
+		kfree(opt->key);
 
-		case Opt_share:
+		opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
+		if (!opt->key)
+			return -ENOMEM;
+		return get_secret(opt->key, param->string, fc);
+
+	case Opt_osdtimeout:
+		warnf(fc, "libceph: Ignoring osdtimeout");
+		break;
+	case Opt_osdkeepalivetimeout:
+		/* 0 isn't well defined right now, reject it */
+		if (result.uint_32 < 1 || result.uint_32 > INT_MAX / 1000)
+			goto out_of_range;
+		opt->osd_keepalive_timeout =
+		    msecs_to_jiffies(result.uint_32 * 1000);
+		break;
+	case Opt_osd_idle_ttl:
+		/* 0 isn't well defined right now, reject it */
+		if (result.uint_32 < 1 || result.uint_32 > INT_MAX / 1000)
+			goto out_of_range;
+		opt->osd_idle_ttl = msecs_to_jiffies(result.uint_32 * 1000);
+		break;
+	case Opt_mount_timeout:
+		/* 0 is "wait forever" (i.e. infinite timeout) */
+		if (result.uint_32 > INT_MAX / 1000)
+			goto out_of_range;
+		opt->mount_timeout = msecs_to_jiffies(result.uint_32 * 1000);
+		break;
+	case Opt_osd_request_timeout:
+		/* 0 is "wait forever" (i.e. infinite timeout) */
+		if (result.uint_32 > INT_MAX / 1000)
+			goto out_of_range;
+		opt->osd_request_timeout =
+		    msecs_to_jiffies(result.uint_32 * 1000);
+		break;
+
+	case Opt_share:
+		if (!result.negated)
 			opt->flags &= ~CEPH_OPT_NOSHARE;
-			break;
-		case Opt_noshare:
+		else
 			opt->flags |= CEPH_OPT_NOSHARE;
-			break;
-
-		case Opt_crc:
+		break;
+	case Opt_crc:
+		if (!result.negated)
 			opt->flags &= ~CEPH_OPT_NOCRC;
-			break;
-		case Opt_nocrc:
+		else
 			opt->flags |= CEPH_OPT_NOCRC;
-			break;
-
-		case Opt_cephx_require_signatures:
+		break;
+	case Opt_cephx_require_signatures:
+		if (!result.negated)
 			opt->flags &= ~CEPH_OPT_NOMSGAUTH;
-			break;
-		case Opt_nocephx_require_signatures:
+		else
 			opt->flags |= CEPH_OPT_NOMSGAUTH;
-			break;
-		case Opt_cephx_sign_messages:
+		break;
+	case Opt_cephx_sign_messages:
+		if (!result.negated)
 			opt->flags &= ~CEPH_OPT_NOMSGSIGN;
-			break;
-		case Opt_nocephx_sign_messages:
+		else
 			opt->flags |= CEPH_OPT_NOMSGSIGN;
-			break;
-
-		case Opt_tcp_nodelay:
+		break;
+	case Opt_tcp_nodelay:
+		if (!result.negated)
 			opt->flags |= CEPH_OPT_TCP_NODELAY;
-			break;
-		case Opt_notcp_nodelay:
+		else
 			opt->flags &= ~CEPH_OPT_TCP_NODELAY;
-			break;
+		break;
 
-		case Opt_abort_on_full:
-			opt->flags |= CEPH_OPT_ABORT_ON_FULL;
-			break;
+	case Opt_abort_on_full:
+		opt->flags |= CEPH_OPT_ABORT_ON_FULL;
+		break;
 
-		default:
-			BUG_ON(token);
-		}
+	default:
+		BUG();
 	}
 
-	/* success */
-	return opt;
+	return 0;
 
-out:
-	ceph_destroy_options(opt);
-	return ERR_PTR(err);
+out_of_range:
+	return invalf(fc, "libceph: %s out of range", param->key);
 }
-EXPORT_SYMBOL(ceph_parse_options);
+EXPORT_SYMBOL(ceph_parse_param);
 
 int ceph_print_client_options(struct seq_file *m, struct ceph_client *client,
 			      bool show_all)

net/ceph/messenger.c

@@ -2004,10 +2004,8 @@ int ceph_parse_ips(const char *c, const char *end,
 	return 0;
 
 bad:
-	pr_err("parse_ips bad ip '%.*s'\n", (int)(end - c), c);
 	return ret;
 }
-EXPORT_SYMBOL(ceph_parse_ips);
 
 static int process_banner(struct ceph_connection *con)
 {

net/ceph/mon_client.c

@@ -1233,9 +1233,6 @@ static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
 	struct ceph_mon_client *monc = con->private;
 	int type = le16_to_cpu(msg->hdr.type);
 
-	if (!monc)
-		return;
-
 	switch (type) {
 	case CEPH_MSG_AUTH_REPLY:
 		handle_auth_reply(monc, msg);