Merge remote branch 'ceph/for-next'

9 files changed, 67 insertions, 16 deletions

diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 7d447af84ec4..149fb8bf4313 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -1215,6 +1215,26 @@ void ceph_dentry_lru_del(struct dentry *dn)
 	}
 }
 
+/*
+ * Return name hash for a given dentry.  This is dependent on
+ * the parent directory's hash function.
+ */
+unsigned ceph_dentry_hash(struct dentry *dn)
+{
+	struct inode *dir = dn->d_parent->d_inode;
+	struct ceph_inode_info *dci = ceph_inode(dir);
+
+	switch (dci->i_dir_layout.dl_dir_hash) {
+	case 0:	/* for backward compat */
+	case CEPH_STR_HASH_LINUX:
+		return dn->d_name.hash;
+
+	default:
+		return ceph_str_hash(dci->i_dir_layout.dl_dir_hash,
+				     dn->d_name.name, dn->d_name.len);
+	}
+}
+
 const struct file_operations ceph_dir_fops = {
 	.read = ceph_read_dir,
 	.readdir = ceph_readdir,
diff --git a/fs/ceph/export.c b/fs/ceph/export.c
index 2297d9426992..e41056174bf8 100644
--- a/fs/ceph/export.c
+++ b/fs/ceph/export.c
@@ -59,7 +59,7 @@ static int ceph_encode_fh(struct dentry *dentry, u32 *rawfh, int *max_len,
 		dout("encode_fh %p connectable\n", dentry);
 		cfh->ino = ceph_ino(dentry->d_inode);
 		cfh->parent_ino = ceph_ino(parent->d_inode);
-		cfh->parent_name_hash = parent->d_name.hash;
+		cfh->parent_name_hash = ceph_dentry_hash(parent);
 		*max_len = connected_handle_length;
 		type = 2;
 	} else if (*max_len >= handle_length) {
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index bf1286588f26..e791fa34b23d 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -297,6 +297,8 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 	ci->i_release_count = 0;
 	ci->i_symlink = NULL;
 
+	memset(&ci->i_dir_layout, 0, sizeof(ci->i_dir_layout));
+
 	ci->i_fragtree = RB_ROOT;
 	mutex_init(&ci->i_fragtree_mutex);
 
@@ -680,6 +682,8 @@ static int fill_inode(struct inode *inode,
 		inode->i_op = &ceph_dir_iops;
 		inode->i_fop = &ceph_dir_fops;
 
+		ci->i_dir_layout = iinfo->dir_layout;
+
 		ci->i_files = le64_to_cpu(info->files);
 		ci->i_subdirs = le64_to_cpu(info->subdirs);
 		ci->i_rbytes = le64_to_cpu(info->rbytes);
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 098b18508479..6316da3d12f1 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -60,7 +60,8 @@ static const struct ceph_connection_operations mds_con_ops;
  * parse individual inode info
  */
 static int parse_reply_info_in(void **p, void *end,
-			       struct ceph_mds_reply_info_in *info)
+			       struct ceph_mds_reply_info_in *info,
+			       int features)
 {
 	int err = -EIO;
 
@@ -74,6 +75,12 @@ static int parse_reply_info_in(void **p, void *end,
 	info->symlink = *p;
 	*p += info->symlink_len;
 
+	if (features & CEPH_FEATURE_DIRLAYOUTHASH)
+		ceph_decode_copy_safe(p, end, &info->dir_layout,
+				      sizeof(info->dir_layout), bad);
+	else
+		memset(&info->dir_layout, 0, sizeof(info->dir_layout));
+
 	ceph_decode_32_safe(p, end, info->xattr_len, bad);
 	ceph_decode_need(p, end, info->xattr_len, bad);
 	info->xattr_data = *p;
@@ -88,12 +95,13 @@ bad:
  * target inode.
  */
 static int parse_reply_info_trace(void **p, void *end,
-				  struct ceph_mds_reply_info_parsed *info)
+				  struct ceph_mds_reply_info_parsed *info,
+				  int features)
 {
 	int err;
 
 	if (info->head->is_dentry) {
-		err = parse_reply_info_in(p, end, &info->diri);
+		err = parse_reply_info_in(p, end, &info->diri, features);
 		if (err < 0)
 			goto out_bad;
 
@@ -114,7 +122,7 @@ static int parse_reply_info_trace(void **p, void *end,
 	}
 
 	if (info->head->is_target) {
-		err = parse_reply_info_in(p, end, &info->targeti);
+		err = parse_reply_info_in(p, end, &info->targeti, features);
 		if (err < 0)
 			goto out_bad;
 	}
@@ -134,7 +142,8 @@ out_bad:
  * parse readdir results
  */
 static int parse_reply_info_dir(void **p, void *end,
-				struct ceph_mds_reply_info_parsed *info)
+				struct ceph_mds_reply_info_parsed *info,
+				int features)
 {
 	u32 num, i = 0;
 	int err;
@@ -182,7 +191,7 @@ static int parse_reply_info_dir(void **p, void *end,
 		*p += sizeof(struct ceph_mds_reply_lease);
 
 		/* inode */
-		err = parse_reply_info_in(p, end, &info->dir_in[i]);
+		err = parse_reply_info_in(p, end, &info->dir_in[i], features);
 		if (err < 0)
 			goto out_bad;
 		i++;
@@ -205,7 +214,8 @@ out_bad:
  * parse entire mds reply
  */
 static int parse_reply_info(struct ceph_msg *msg,
-			    struct ceph_mds_reply_info_parsed *info)
+			    struct ceph_mds_reply_info_parsed *info,
+			    int features)
 {
 	void *p, *end;
 	u32 len;
@@ -218,7 +228,7 @@ static int parse_reply_info(struct ceph_msg *msg,
 	/* trace */
 	ceph_decode_32_safe(&p, end, len, bad);
 	if (len > 0) {
-		err = parse_reply_info_trace(&p, p+len, info);
+		err = parse_reply_info_trace(&p, p+len, info, features);
 		if (err < 0)
 			goto out_bad;
 	}
@@ -226,7 +236,7 @@ static int parse_reply_info(struct ceph_msg *msg,
 	/* dir content */
 	ceph_decode_32_safe(&p, end, len, bad);
 	if (len > 0) {
-		err = parse_reply_info_dir(&p, p+len, info);
+		err = parse_reply_info_dir(&p, p+len, info, features);
 		if (err < 0)
 			goto out_bad;
 	}
@@ -622,7 +632,7 @@ static int __choose_mds(struct ceph_mds_client *mdsc,
 		} else {
 			/* dir + name */
 			inode = dir;
-			hash = req->r_dentry->d_name.hash;
+			hash = ceph_dentry_hash(req->r_dentry);
 			is_hash = true;
 		}
 	}
@@ -2069,7 +2079,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
 
 	dout("handle_reply tid %lld result %d\n", tid, result);
 	rinfo = &req->r_reply_info;
-	err = parse_reply_info(msg, rinfo);
+	err = parse_reply_info(msg, rinfo, session->s_con.peer_features);
 	mutex_unlock(&mdsc->mutex);
 
 	mutex_lock(&session->s_mutex);
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index 9341fd4f1432..21e02a59abaf 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -35,6 +35,7 @@ struct ceph_cap;
  */
 struct ceph_mds_reply_info_in {
 	struct ceph_mds_reply_inode *in;
+	struct ceph_dir_layout dir_layout;
 	u32 symlink_len;
 	char *symlink;
 	u32 xattr_len;
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index 08b460ae0539..1417f3f3e246 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -428,7 +428,8 @@ struct ceph_fs_client *create_fs_client(struct ceph_mount_options *fsopt,
 		goto fail;
 	}
 	fsc->client->extra_mon_dispatch = extra_mon_dispatch;
-	fsc->client->supported_features |= CEPH_FEATURE_FLOCK;
+	fsc->client->supported_features |= CEPH_FEATURE_FLOCK |
+		CEPH_FEATURE_DIRLAYOUTHASH;
 	fsc->client->monc.want_mdsmap = 1;
 
 	fsc->mount_options = fsopt;
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 7f01728a4657..6e0826695112 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -239,6 +239,7 @@ struct ceph_inode_info {
 	unsigned i_ceph_flags;
 	unsigned long i_release_count;
 
+	struct ceph_dir_layout i_dir_layout;
 	struct ceph_file_layout i_layout;
 	char *i_symlink;
 
@@ -768,6 +769,7 @@ extern void ceph_dentry_lru_add(struct dentry *dn);
 extern void ceph_dentry_lru_touch(struct dentry *dn);
 extern void ceph_dentry_lru_del(struct dentry *dn);
 extern void ceph_invalidate_dentry_lease(struct dentry *dentry);
+extern unsigned ceph_dentry_hash(struct dentry *dn);
 
 /*
  * our d_ops vary depending on whether the inode is live,
diff --git a/include/linux/ceph/ceph_fs.h b/include/linux/ceph/ceph_fs.h
index c3c74aef289d..09dcc0c2ffd5 100644
--- a/include/linux/ceph/ceph_fs.h
+++ b/include/linux/ceph/ceph_fs.h
@@ -43,6 +43,10 @@
 #define CEPH_FEATURE_NOSRCADDR      (1<<1)
 #define CEPH_FEATURE_MONCLOCKCHECK  (1<<2)
 #define CEPH_FEATURE_FLOCK          (1<<3)
+#define CEPH_FEATURE_SUBSCRIBE2     (1<<4)
+#define CEPH_FEATURE_MONNAMES       (1<<5)
+#define CEPH_FEATURE_RECONNECT_SEQ  (1<<6)
+#define CEPH_FEATURE_DIRLAYOUTHASH  (1<<7)
 
 
 /*
@@ -55,10 +59,10 @@ struct ceph_file_layout {
 	__le32 fl_stripe_count;    /* over this many objects */
 	__le32 fl_object_size;     /* until objects are this big, then move to
 				      new objects */
-	__le32 fl_cas_hash;        /* 0 = none; 1 = sha256 */
+	__le32 fl_cas_hash;        /* UNUSED.  0 = none; 1 = sha256 */
 
 	/* pg -> disk layout */
-	__le32 fl_object_stripe_unit;  /* for per-object parity, if any */
+	__le32 fl_object_stripe_unit;  /* UNUSED.  for per-object parity, if any */
 
 	/* object -> pg layout */
 	__le32 fl_pg_preferred; /* preferred primary for pg (-1 for none) */
@@ -69,6 +73,12 @@ struct ceph_file_layout {
 
 int ceph_file_layout_is_valid(const struct ceph_file_layout *layout);
 
+struct ceph_dir_layout {
+	__u8   dl_dir_hash;   /* see ceph_hash.h for ids */
+	__u8   dl_unused1;
+	__u16  dl_unused2;
+	__u32  dl_unused3;
+} __attribute__ ((packed));
 
 /* crypto algorithms */
 #define CEPH_CRYPTO_NONE 0x0
@@ -457,7 +467,7 @@ struct ceph_mds_reply_inode {
 	struct ceph_timespec rctime;
 	struct ceph_frag_tree_head fragtree;  /* (must be at end of struct) */
 } __attribute__ ((packed));
-/* followed by frag array, then symlink string, then xattr blob */
+/* followed by frag array, symlink string, dir layout, xattr blob */
 
 /* reply_lease follows dname, and reply_inode */
 struct ceph_mds_reply_lease {
diff --git a/net/ceph/ceph_hash.c b/net/ceph/ceph_hash.c
index 815ef8826796..0a1b53bce76d 100644
--- a/net/ceph/ceph_hash.c
+++ b/net/ceph/ceph_hash.c
@@ -1,5 +1,6 @@
 
 #include <linux/ceph/types.h>
+#include <linux/module.h>
 
 /*
  * Robert Jenkin's hash function.
@@ -104,6 +105,7 @@ unsigned ceph_str_hash(int type, const char *s, unsigned len)
 		return -1;
 	}
 }
+EXPORT_SYMBOL(ceph_str_hash);
 
 const char *ceph_str_hash_name(int type)
 {
@@ -116,3 +118,4 @@ const char *ceph_str_hash_name(int type)
 		return "unknown";
 	}
 }
+EXPORT_SYMBOL(ceph_str_hash_name);