Merge branch 'next' into for-linus

15 files changed, 396 insertions, 206 deletions

diff --git a/fs/ceph/Makefile b/fs/ceph/Makefile
index 9e6c4f2e8ff1..bd352125e829 100644
--- a/fs/ceph/Makefile
+++ b/fs/ceph/Makefile
@@ -2,31 +2,10 @@
 # Makefile for CEPH filesystem.
 #
 
-ifneq ($(KERNELRELEASE),)
-
 obj-$(CONFIG_CEPH_FS) += ceph.o
 
-ceph-objs := super.o inode.o dir.o file.o locks.o addr.o ioctl.o \
+ceph-y := super.o inode.o dir.o file.o locks.o addr.o ioctl.o \
 	export.o caps.o snap.o xattr.o \
 	mds_client.o mdsmap.o strings.o ceph_frag.o \
 	debugfs.o
 
-else
-#Otherwise we were called directly from the command
-# line; invoke the kernel build system.
-
-KERNELDIR ?= /lib/modules/$(shell uname -r)/build
-PWD := $(shell pwd)
-
-default: all
-
-all:
-	$(MAKE) -C $(KERNELDIR) M=$(PWD) CONFIG_CEPH_FS=m modules
-
-modules_install:
-	$(MAKE) -C $(KERNELDIR) M=$(PWD) CONFIG_CEPH_FS=m modules_install
-
-clean:
-	$(MAKE) -C $(KERNELDIR) M=$(PWD) clean
-
-endif
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index e9c874abc9e1..561438b6a50c 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -204,7 +204,7 @@ static int readpage_nounlock(struct file *filp, struct page *page)
 	err = ceph_osdc_readpages(osdc, ceph_vino(inode), &ci->i_layout,
 				  page->index << PAGE_CACHE_SHIFT, &len,
 				  ci->i_truncate_seq, ci->i_truncate_size,
-				  &page, 1);
+				  &page, 1, 0);
 	if (err == -ENOENT)
 		err = 0;
 	if (err < 0) {
@@ -287,7 +287,7 @@ static int ceph_readpages(struct file *file, struct address_space *mapping,
 	rc = ceph_osdc_readpages(osdc, ceph_vino(inode), &ci->i_layout,
 				 offset, &len,
 				 ci->i_truncate_seq, ci->i_truncate_size,
-				 pages, nr_pages);
+				 pages, nr_pages, 0);
 	if (rc == -ENOENT)
 		rc = 0;
 	if (rc < 0)
@@ -774,7 +774,7 @@ get_more_pages:
 					    snapc, do_sync,
 					    ci->i_truncate_seq,
 					    ci->i_truncate_size,
-					    &inode->i_mtime, true, 1);
+					    &inode->i_mtime, true, 1, 0);
 				max_pages = req->r_num_pages;
 
 				alloc_page_vec(fsc, req);
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index 98ab13e2b71d..6b61ded701e1 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -1430,8 +1430,8 @@ static int try_nonblocking_invalidate(struct inode *inode)
 	    invalidating_gen == ci->i_rdcache_gen) {
 		/* success. */
 		dout("try_nonblocking_invalidate %p success\n", inode);
-		ci->i_rdcache_gen = 0;
-		ci->i_rdcache_revoking = 0;
+		/* save any racing async invalidate some trouble */
+		ci->i_rdcache_revoking = ci->i_rdcache_gen - 1;
 		return 0;
 	}
 	dout("try_nonblocking_invalidate %p failed\n", inode);
@@ -1560,9 +1560,10 @@ retry_locked:
 		/* NOTE: no side-effects allowed, until we take s_mutex */
 
 		revoking = cap->implemented & ~cap->issued;
-		if (revoking)
-			dout(" mds%d revoking %s\n", cap->mds,
-			     ceph_cap_string(revoking));
+		dout(" mds%d cap %p issued %s implemented %s revoking %s\n",
+		     cap->mds, cap, ceph_cap_string(cap->issued),
+		     ceph_cap_string(cap->implemented),
+		     ceph_cap_string(revoking));
 
 		if (cap == ci->i_auth_cap &&
 		    (cap->issued & CEPH_CAP_FILE_WR)) {
@@ -1658,6 +1659,8 @@ ack:
 
 		if (cap == ci->i_auth_cap && ci->i_dirty_caps)
 			flushing = __mark_caps_flushing(inode, session);
+		else
+			flushing = 0;
 
 		mds = cap->mds;  /* remember mds, so we don't repeat */
 		sent++;
@@ -1940,6 +1943,35 @@ void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc,
 	}
 }
 
+static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc,
+				     struct ceph_mds_session *session,
+				     struct inode *inode)
+{
+	struct ceph_inode_info *ci = ceph_inode(inode);
+	struct ceph_cap *cap;
+	int delayed = 0;
+
+	spin_lock(&inode->i_lock);
+	cap = ci->i_auth_cap;
+	dout("kick_flushing_inode_caps %p flushing %s flush_seq %lld\n", inode,
+	     ceph_cap_string(ci->i_flushing_caps), ci->i_cap_flush_seq);
+	__ceph_flush_snaps(ci, &session, 1);
+	if (ci->i_flushing_caps) {
+		delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH,
+				     __ceph_caps_used(ci),
+				     __ceph_caps_wanted(ci),
+				     cap->issued | cap->implemented,
+				     ci->i_flushing_caps, NULL);
+		if (delayed) {
+			spin_lock(&inode->i_lock);
+			__cap_delay_requeue(mdsc, ci);
+			spin_unlock(&inode->i_lock);
+		}
+	} else {
+		spin_unlock(&inode->i_lock);
+	}
+}
+
 
 /*
  * Take references to capabilities we hold, so that we don't release
@@ -2273,8 +2305,7 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int mds = session->s_mds;
-	unsigned seq = le32_to_cpu(grant->seq);
-	unsigned issue_seq = le32_to_cpu(grant->issue_seq);
+	int seq = le32_to_cpu(grant->seq);
 	int newcaps = le32_to_cpu(grant->caps);
 	int issued, implemented, used, wanted, dirty;
 	u64 size = le64_to_cpu(grant->size);
@@ -2286,8 +2317,8 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 	int revoked_rdcache = 0;
 	int queue_invalidate = 0;
 
-	dout("handle_cap_grant inode %p cap %p mds%d seq %u/%u %s\n",
-	     inode, cap, mds, seq, issue_seq, ceph_cap_string(newcaps));
+	dout("handle_cap_grant inode %p cap %p mds%d seq %d %s\n",
+	     inode, cap, mds, seq, ceph_cap_string(newcaps));
 	dout(" size %llu max_size %llu, i_size %llu\n", size, max_size,
 		inode->i_size);
 
@@ -2383,7 +2414,6 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 	}
 
 	cap->seq = seq;
-	cap->issue_seq = issue_seq;
 
 	/* file layout may have changed */
 	ci->i_layout = grant->layout;
@@ -2689,8 +2719,13 @@ static void handle_cap_import(struct ceph_mds_client *mdsc,
 	ceph_add_cap(inode, session, cap_id, -1,
 		     issued, wanted, seq, mseq, realmino, CEPH_CAP_FLAG_AUTH,
 		     NULL /* no caps context */);
-	try_flush_caps(inode, session, NULL);
+	kick_flushing_inode_caps(mdsc, session, inode);
 	up_read(&mdsc->snap_rwsem);
+
+	/* make sure we re-request max_size, if necessary */
+	spin_lock(&inode->i_lock);
+	ci->i_requested_max_size = 0;
+	spin_unlock(&inode->i_lock);
 }
 
 /*
@@ -2782,8 +2817,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	case CEPH_CAP_OP_IMPORT:
 		handle_cap_import(mdsc, inode, h, session,
 				  snaptrace, snaptrace_len);
-		ceph_check_caps(ceph_inode(inode), CHECK_CAPS_NODELAY,
-				session);
+		ceph_check_caps(ceph_inode(inode), 0, session);
 		goto done_unlocked;
 	}
 
diff --git a/fs/ceph/debugfs.c b/fs/ceph/debugfs.c
index 7ae1b3d55b58..08f65faac112 100644
--- a/fs/ceph/debugfs.c
+++ b/fs/ceph/debugfs.c
@@ -60,10 +60,13 @@ static int mdsc_show(struct seq_file *s, void *p)
 	for (rp = rb_first(&mdsc->request_tree); rp; rp = rb_next(rp)) {
 		req = rb_entry(rp, struct ceph_mds_request, r_node);
 
-		if (req->r_request)
-			seq_printf(s, "%lld\tmds%d\t", req->r_tid, req->r_mds);
-		else
+		if (req->r_request && req->r_session)
+			seq_printf(s, "%lld\tmds%d\t", req->r_tid,
+				   req->r_session->s_mds);
+		else if (!req->r_request)
 			seq_printf(s, "%lld\t(no request)\t", req->r_tid);
+		else
+			seq_printf(s, "%lld\t(no session)\t", req->r_tid);
 
 		seq_printf(s, "%s", ceph_mds_op_name(req->r_op));
 
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index e0a2dc6fcafc..0bc68de8edd7 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -40,12 +40,13 @@ int ceph_init_dentry(struct dentry *dentry)
 	if (dentry->d_fsdata)
 		return 0;
 
-	if (ceph_snap(dentry->d_parent->d_inode) == CEPH_NOSNAP)
-		dentry->d_op = &ceph_dentry_ops;
+	if (dentry->d_parent == NULL ||   /* nfs fh_to_dentry */
+	    ceph_snap(dentry->d_parent->d_inode) == CEPH_NOSNAP)
+		d_set_d_op(dentry, &ceph_dentry_ops);
 	else if (ceph_snap(dentry->d_parent->d_inode) == CEPH_SNAPDIR)
-		dentry->d_op = &ceph_snapdir_dentry_ops;
+		d_set_d_op(dentry, &ceph_snapdir_dentry_ops);
 	else
-		dentry->d_op = &ceph_snap_dentry_ops;
+		d_set_d_op(dentry, &ceph_snap_dentry_ops);
 
 	di = kmem_cache_alloc(ceph_dentry_cachep, GFP_NOFS | __GFP_ZERO);
 	if (!di)
@@ -111,11 +112,11 @@ static int __dcache_readdir(struct file *filp,
 	dout("__dcache_readdir %p at %llu (last %p)\n", dir, filp->f_pos,
 	     last);
 
-	spin_lock(&dcache_lock);
+	spin_lock(&parent->d_lock);
 
 	/* start at beginning? */
-	if (filp->f_pos == 2 || (last &&
-				 filp->f_pos < ceph_dentry(last)->offset)) {
+	if (filp->f_pos == 2 || last == NULL ||
+	    filp->f_pos < ceph_dentry(last)->offset) {
 		if (list_empty(&parent->d_subdirs))
 			goto out_unlock;
 		p = parent->d_subdirs.prev;
@@ -135,6 +136,7 @@ more:
 			fi->at_end = 1;
 			goto out_unlock;
 		}
+		spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED);
 		if (!d_unhashed(dentry) && dentry->d_inode &&
 		    ceph_snap(dentry->d_inode) != CEPH_SNAPDIR &&
 		    ceph_ino(dentry->d_inode) != CEPH_INO_CEPH &&
@@ -144,13 +146,15 @@ more:
 		     dentry->d_name.len, dentry->d_name.name, di->offset,
 		     filp->f_pos, d_unhashed(dentry) ? " unhashed" : "",
 		     !dentry->d_inode ? " null" : "");
+		spin_unlock(&dentry->d_lock);
 		p = p->prev;
 		dentry = list_entry(p, struct dentry, d_u.d_child);
 		di = ceph_dentry(dentry);
 	}
 
-	atomic_inc(&dentry->d_count);
-	spin_unlock(&dcache_lock);
+	dget_dlock(dentry);
+	spin_unlock(&dentry->d_lock);
+	spin_unlock(&parent->d_lock);
 
 	dout(" %llu (%llu) dentry %p %.*s %p\n", di->offset, filp->f_pos,
 	     dentry, dentry->d_name.len, dentry->d_name.name, dentry->d_inode);
@@ -176,19 +180,19 @@ more:
 
 	filp->f_pos++;
 
-	/* make sure a dentry wasn't dropped while we didn't have dcache_lock */
+	/* make sure a dentry wasn't dropped while we didn't have parent lock */
 	if (!ceph_i_test(dir, CEPH_I_COMPLETE)) {
 		dout(" lost I_COMPLETE on %p; falling back to mds\n", dir);
 		err = -EAGAIN;
 		goto out;
 	}
 
-	spin_lock(&dcache_lock);
+	spin_lock(&parent->d_lock);
 	p = p->prev;	/* advance to next dentry */
 	goto more;
 
 out_unlock:
-	spin_unlock(&dcache_lock);
+	spin_unlock(&parent->d_lock);
 out:
 	if (last)
 		dput(last);
@@ -336,7 +340,10 @@ more:
 		if (req->r_reply_info.dir_end) {
 			kfree(fi->last_name);
 			fi->last_name = NULL;
-			fi->next_offset = 2;
+			if (ceph_frag_is_rightmost(frag))
+				fi->next_offset = 2;
+			else
+				fi->next_offset = 0;
 		} else {
 			rinfo = &req->r_reply_info;
 			err = note_last_dentry(fi,
@@ -355,18 +362,22 @@ more:
 		u64 pos = ceph_make_fpos(frag, off);
 		struct ceph_mds_reply_inode *in =
 			rinfo->dir_in[off - fi->offset].in;
+		struct ceph_vino vino;
+		ino_t ino;
+
 		dout("readdir off %d (%d/%d) -> %lld '%.*s' %p\n",
 		     off, off - fi->offset, rinfo->dir_nr, pos,
 		     rinfo->dir_dname_len[off - fi->offset],
 		     rinfo->dir_dname[off - fi->offset], in);
 		BUG_ON(!in);
 		ftype = le32_to_cpu(in->mode) >> 12;
+		vino.ino = le64_to_cpu(in->ino);
+		vino.snap = le64_to_cpu(in->snapid);
+		ino = ceph_vino_to_ino(vino);
 		if (filldir(dirent,
 			    rinfo->dir_dname[off - fi->offset],
 			    rinfo->dir_dname_len[off - fi->offset],
-			    pos,
-			    le64_to_cpu(in->ino),
-			    ftype) < 0) {
+			    pos, ino, ftype) < 0) {
 			dout("filldir stopping us...\n");
 			return 0;
 		}
@@ -414,6 +425,7 @@ static void reset_readdir(struct ceph_file_info *fi)
 		fi->last_readdir = NULL;
 	}
 	kfree(fi->last_name);
+	fi->last_name = NULL;
 	fi->next_offset = 2;  /* compensate for . and .. */
 	if (fi->dentry) {
 		dput(fi->dentry);
@@ -978,7 +990,12 @@ static int dir_lease_is_valid(struct inode *dir, struct dentry *dentry)
  */
 static int ceph_d_revalidate(struct dentry *dentry, struct nameidata *nd)
 {
-	struct inode *dir = dentry->d_parent->d_inode;
+	struct inode *dir;
+
+	if (nd->flags & LOOKUP_RCU)
+		return -ECHILD;
+
+	dir = dentry->d_parent->d_inode;
 
 	dout("d_revalidate %p '%.*s' inode %p offset %lld\n", dentry,
 	     dentry->d_name.len, dentry->d_name.name, dentry->d_inode,
@@ -1207,6 +1224,26 @@ 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/file.c b/fs/ceph/file.c
index e77c28cf3690..7d0e4a82d898 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -154,11 +154,13 @@ int ceph_open(struct inode *inode, struct file *file)
 	}
 
 	/*
-	 * No need to block if we have any caps.  Update wanted set
+	 * No need to block if we have caps on the auth MDS (for
+	 * write) or any MDS (for read).  Update wanted set
 	 * asynchronously.
 	 */
 	spin_lock(&inode->i_lock);
-	if (__ceph_is_any_real_caps(ci)) {
+	if (__ceph_is_any_real_caps(ci) &&
+	    (((fmode & CEPH_FILE_MODE_WR) == 0) || ci->i_auth_cap)) {
 		int mds_wanted = __ceph_caps_mds_wanted(ci);
 		int issued = __ceph_caps_issued(ci, NULL);
 
@@ -280,11 +282,13 @@ int ceph_release(struct inode *inode, struct file *file)
 static int striped_read(struct inode *inode,
 			u64 off, u64 len,
 			struct page **pages, int num_pages,
-			int *checkeof)
+			int *checkeof, bool align_to_pages,
+			unsigned long buf_align)
 {
 	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	u64 pos, this_len;
+	int io_align, page_align;
 	int page_off = off & ~PAGE_CACHE_MASK; /* first byte's offset in page */
 	int left, pages_left;
 	int read;
@@ -300,14 +304,19 @@ static int striped_read(struct inode *inode,
 	page_pos = pages;
 	pages_left = num_pages;
 	read = 0;
+	io_align = off & ~PAGE_MASK;
 
 more:
+	if (align_to_pages)
+		page_align = (pos - io_align + buf_align) & ~PAGE_MASK;
+	else
+		page_align = pos & ~PAGE_MASK;
 	this_len = left;
 	ret = ceph_osdc_readpages(&fsc->client->osdc, ceph_vino(inode),
 				  &ci->i_layout, pos, &this_len,
 				  ci->i_truncate_seq,
 				  ci->i_truncate_size,
-				  page_pos, pages_left);
+				  page_pos, pages_left, page_align);
 	hit_stripe = this_len < left;
 	was_short = ret >= 0 && ret < this_len;
 	if (ret == -ENOENT)
@@ -368,32 +377,34 @@ static ssize_t ceph_sync_read(struct file *file, char __user *data,
 	struct inode *inode = file->f_dentry->d_inode;
 	struct page **pages;
 	u64 off = *poff;
-	int num_pages = calc_pages_for(off, len);
-	int ret;
+	int num_pages, ret;
 
 	dout("sync_read on file %p %llu~%u %s\n", file, off, len,
 	     (file->f_flags & O_DIRECT) ? "O_DIRECT" : "");
 
 	if (file->f_flags & O_DIRECT) {
-		pages = ceph_get_direct_page_vector(data, num_pages, off, len);
-
-		/*
-		 * flush any page cache pages in this range.  this
-		 * will make concurrent normal and O_DIRECT io slow,
-		 * but it will at least behave sensibly when they are
-		 * in sequence.
-		 */
+		num_pages = calc_pages_for((unsigned long)data, len);
+		pages = ceph_get_direct_page_vector(data, num_pages, true);
 	} else {
+		num_pages = calc_pages_for(off, len);
 		pages = ceph_alloc_page_vector(num_pages, GFP_NOFS);
 	}
 	if (IS_ERR(pages))
 		return PTR_ERR(pages);
 
+	/*
+	 * flush any page cache pages in this range.  this
+	 * will make concurrent normal and sync io slow,
+	 * but it will at least behave sensibly when they are
+	 * in sequence.
+	 */
 	ret = filemap_write_and_wait(inode->i_mapping);
 	if (ret < 0)
 		goto done;
 
-	ret = striped_read(inode, off, len, pages, num_pages, checkeof);
+	ret = striped_read(inode, off, len, pages, num_pages, checkeof,
+			   file->f_flags & O_DIRECT,
+			   (unsigned long)data & ~PAGE_MASK);
 
 	if (ret >= 0 && (file->f_flags & O_DIRECT) == 0)
 		ret = ceph_copy_page_vector_to_user(pages, data, off, ret);
@@ -402,7 +413,7 @@ static ssize_t ceph_sync_read(struct file *file, char __user *data,
 
 done:
 	if (file->f_flags & O_DIRECT)
-		ceph_put_page_vector(pages, num_pages);
+		ceph_put_page_vector(pages, num_pages, true);
 	else
 		ceph_release_page_vector(pages, num_pages);
 	dout("sync_read result %d\n", ret);
@@ -448,6 +459,8 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
 	int flags;
 	int do_sync = 0;
 	int check_caps = 0;
+	int page_align, io_align;
+	unsigned long buf_align;
 	int ret;
 	struct timespec mtime = CURRENT_TIME;
 
@@ -462,6 +475,9 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
 	else
 		pos = *offset;
 
+	io_align = pos & ~PAGE_MASK;
+	buf_align = (unsigned long)data & ~PAGE_MASK;
+
 	ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + left);
 	if (ret < 0)
 		return ret;
@@ -486,20 +502,27 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
 	 */
 more:
 	len = left;
+	if (file->f_flags & O_DIRECT) {
+		/* write from beginning of first page, regardless of
+		   io alignment */
+		page_align = (pos - io_align + buf_align) & ~PAGE_MASK;
+		num_pages = calc_pages_for((unsigned long)data, len);
+	} else {
+		page_align = pos & ~PAGE_MASK;
+		num_pages = calc_pages_for(pos, len);
+	}
 	req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout,
 				    ceph_vino(inode), pos, &len,
 				    CEPH_OSD_OP_WRITE, flags,
 				    ci->i_snap_realm->cached_context,
 				    do_sync,
 				    ci->i_truncate_seq, ci->i_truncate_size,
-				    &mtime, false, 2);
+				    &mtime, false, 2, page_align);
 	if (!req)
 		return -ENOMEM;
 
-	num_pages = calc_pages_for(pos, len);
-
 	if (file->f_flags & O_DIRECT) {
-		pages = ceph_get_direct_page_vector(data, num_pages, pos, len);
+		pages = ceph_get_direct_page_vector(data, num_pages, false);
 		if (IS_ERR(pages)) {
 			ret = PTR_ERR(pages);
 			goto out;
@@ -549,7 +572,7 @@ more:
 	}
 
 	if (file->f_flags & O_DIRECT)
-		ceph_put_page_vector(pages, num_pages);
+		ceph_put_page_vector(pages, num_pages, false);
 	else if (file->f_flags & O_SYNC)
 		ceph_release_page_vector(pages, num_pages);
 
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 1d6a45b5a04c..5625463aa479 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -2,7 +2,6 @@
 
 #include <linux/module.h>
 #include <linux/fs.h>
-#include <linux/smp_lock.h>
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/uaccess.h>
@@ -298,6 +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);
 
@@ -369,6 +370,15 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 	return &ci->vfs_inode;
 }
 
+static void ceph_i_callback(struct rcu_head *head)
+{
+	struct inode *inode = container_of(head, struct inode, i_rcu);
+	struct ceph_inode_info *ci = ceph_inode(inode);
+
+	INIT_LIST_HEAD(&inode->i_dentry);
+	kmem_cache_free(ceph_inode_cachep, ci);
+}
+
 void ceph_destroy_inode(struct inode *inode)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
@@ -408,7 +418,7 @@ void ceph_destroy_inode(struct inode *inode)
 	if (ci->i_xattrs.prealloc_blob)
 		ceph_buffer_put(ci->i_xattrs.prealloc_blob);
 
-	kmem_cache_free(ceph_inode_cachep, ci);
+	call_rcu(&inode->i_rcu, ceph_i_callback);
 }
 
 
@@ -471,7 +481,9 @@ void ceph_fill_file_time(struct inode *inode, int issued,
 
 	if (issued & (CEPH_CAP_FILE_EXCL|
 		      CEPH_CAP_FILE_WR|
-		      CEPH_CAP_FILE_BUFFER)) {
+		      CEPH_CAP_FILE_BUFFER|
+		      CEPH_CAP_AUTH_EXCL|
+		      CEPH_CAP_XATTR_EXCL)) {
 		if (timespec_compare(ctime, &inode->i_ctime) > 0) {
 			dout("ctime %ld.%09ld -> %ld.%09ld inc w/ cap\n",
 			     inode->i_ctime.tv_sec, inode->i_ctime.tv_nsec,
@@ -511,7 +523,7 @@ void ceph_fill_file_time(struct inode *inode, int issued,
 			warn = 1;
 		}
 	} else {
-		/* we have no write caps; whatever the MDS says is true */
+		/* we have no write|excl caps; whatever the MDS says is true */
 		if (ceph_seq_cmp(time_warp_seq, ci->i_time_warp_seq) >= 0) {
 			inode->i_ctime = *ctime;
 			inode->i_mtime = *mtime;
@@ -567,12 +579,17 @@ static int fill_inode(struct inode *inode,
 
 	/*
 	 * provided version will be odd if inode value is projected,
-	 * even if stable.  skip the update if we have a newer info
-	 * (e.g., due to inode info racing form multiple MDSs), or if
-	 * we are getting projected (unstable) inode info.
+	 * even if stable.  skip the update if we have newer stable
+	 * info (ours>=theirs, e.g. due to racing mds replies), unless
+	 * we are getting projected (unstable) info (in which case the
+	 * version is odd, and we want ours>theirs).
+	 *   us   them
+	 *   2    2     skip
+	 *   3    2     skip
+	 *   3    3     update
 	 */
 	if (le64_to_cpu(info->version) > 0 &&
-	    (ci->i_version & ~1) > le64_to_cpu(info->version))
+	    (ci->i_version & ~1) >= le64_to_cpu(info->version))
 		goto no_change;
 
 	issued = __ceph_caps_issued(ci, &implemented);
@@ -606,7 +623,14 @@ static int fill_inode(struct inode *inode,
 			    le32_to_cpu(info->time_warp_seq),
 			    &ctime, &mtime, &atime);
 
-	ci->i_max_size = le64_to_cpu(info->max_size);
+	/* only update max_size on auth cap */
+	if ((info->cap.flags & CEPH_CAP_FLAG_AUTH) &&
+	    ci->i_max_size != le64_to_cpu(info->max_size)) {
+		dout("max_size %lld -> %llu\n", ci->i_max_size,
+		     le64_to_cpu(info->max_size));
+		ci->i_max_size = le64_to_cpu(info->max_size);
+	}
+
 	ci->i_layout = info->layout;
 	inode->i_blkbits = fls(le32_to_cpu(info->layout.fl_stripe_unit)) - 1;
 
@@ -667,6 +691,8 @@ 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);
@@ -684,10 +710,6 @@ static int fill_inode(struct inode *inode,
 			ci->i_ceph_flags |= CEPH_I_COMPLETE;
 			ci->i_max_offset = 2;
 		}
-
-		/* it may be better to set st_size in getattr instead? */
-		if (ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), RBYTES))
-			inode->i_size = ci->i_rbytes;
 		break;
 	default:
 		pr_err("fill_inode %llx.%llx BAD mode 0%o\n",
@@ -828,13 +850,13 @@ static void ceph_set_dentry_offset(struct dentry *dn)
 	di->offset = ceph_inode(inode)->i_max_offset++;
 	spin_unlock(&inode->i_lock);
 
-	spin_lock(&dcache_lock);
-	spin_lock(&dn->d_lock);
+	spin_lock(&dir->d_lock);
+	spin_lock_nested(&dn->d_lock, DENTRY_D_LOCK_NESTED);
 	list_move(&dn->d_u.d_child, &dir->d_subdirs);
 	dout("set_dentry_offset %p %lld (%p %p)\n", dn, di->offset,
 	     dn->d_u.d_child.prev, dn->d_u.d_child.next);
 	spin_unlock(&dn->d_lock);
-	spin_unlock(&dcache_lock);
+	spin_unlock(&dir->d_lock);
 }
 
 /*
@@ -866,8 +888,8 @@ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in,
 	} else if (realdn) {
 		dout("dn %p (%d) spliced with %p (%d) "
 		     "inode %p ino %llx.%llx\n",
-		     dn, atomic_read(&dn->d_count),
-		     realdn, atomic_read(&realdn->d_count),
+		     dn, dn->d_count,
+		     realdn, realdn->d_count,
 		     realdn->d_inode, ceph_vinop(realdn->d_inode));
 		dput(dn);
 		dn = realdn;
@@ -1055,7 +1077,8 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req,
 		ininfo = rinfo->targeti.in;
 		vino.ino = le64_to_cpu(ininfo->ino);
 		vino.snap = le64_to_cpu(ininfo->snapid);
-		if (!dn->d_inode) {
+		in = dn->d_inode;
+		if (!in) {
 			in = ceph_get_inode(sb, vino);
 			if (IS_ERR(in)) {
 				pr_err("fill_trace bad get_inode "
@@ -1217,11 +1240,11 @@ retry_lookup:
 			goto retry_lookup;
 		} else {
 			/* reorder parent's d_subdirs */
-			spin_lock(&dcache_lock);
-			spin_lock(&dn->d_lock);
+			spin_lock(&parent->d_lock);
+			spin_lock_nested(&dn->d_lock, DENTRY_D_LOCK_NESTED);
 			list_move(&dn->d_u.d_child, &parent->d_subdirs);
 			spin_unlock(&dn->d_lock);
-			spin_unlock(&dcache_lock);
+			spin_unlock(&parent->d_lock);
 		}
 
 		di = dn->d_fsdata;
@@ -1386,11 +1409,8 @@ static void ceph_invalidate_work(struct work_struct *work)
 	spin_lock(&inode->i_lock);
 	dout("invalidate_pages %p gen %d revoking %d\n", inode,
 	     ci->i_rdcache_gen, ci->i_rdcache_revoking);
-	if (ci->i_rdcache_gen == 0 ||
-	    ci->i_rdcache_revoking != ci->i_rdcache_gen) {
-		BUG_ON(ci->i_rdcache_revoking > ci->i_rdcache_gen);
+	if (ci->i_rdcache_revoking != ci->i_rdcache_gen) {
 		/* nevermind! */
-		ci->i_rdcache_revoking = 0;
 		spin_unlock(&inode->i_lock);
 		goto out;
 	}
@@ -1400,15 +1420,16 @@ static void ceph_invalidate_work(struct work_struct *work)
 	ceph_invalidate_nondirty_pages(inode->i_mapping);
 
 	spin_lock(&inode->i_lock);
-	if (orig_gen == ci->i_rdcache_gen) {
+	if (orig_gen == ci->i_rdcache_gen &&
+	    orig_gen == ci->i_rdcache_revoking) {
 		dout("invalidate_pages %p gen %d successful\n", inode,
 		     ci->i_rdcache_gen);
-		ci->i_rdcache_gen = 0;
-		ci->i_rdcache_revoking = 0;
+		ci->i_rdcache_revoking--;
 		check = 1;
 	} else {
-		dout("invalidate_pages %p gen %d raced, gen now %d\n",
-		     inode, orig_gen, ci->i_rdcache_gen);
+		dout("invalidate_pages %p gen %d raced, now %d revoking %d\n",
+		     inode, orig_gen, ci->i_rdcache_gen,
+		     ci->i_rdcache_revoking);
 	}
 	spin_unlock(&inode->i_lock);
 
@@ -1739,7 +1760,7 @@ int ceph_do_getattr(struct inode *inode, int mask)
 		return 0;
 	}
 
-	dout("do_getattr inode %p mask %s\n", inode, ceph_cap_string(mask));
+	dout("do_getattr inode %p mask %s mode 0%o\n", inode, ceph_cap_string(mask), inode->i_mode);
 	if (ceph_caps_issued_mask(ceph_inode(inode), mask, 1))
 		return 0;
 
@@ -1760,12 +1781,17 @@ int ceph_do_getattr(struct inode *inode, int mask)
  * Check inode permissions.  We verify we have a valid value for
  * the AUTH cap, then call the generic handler.
  */
-int ceph_permission(struct inode *inode, int mask)
+int ceph_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	int err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED);
+	int err;
+
+	if (flags & IPERM_FLAG_RCU)
+		return -ECHILD;
+
+	err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED);
 
 	if (!err)
-		err = generic_permission(inode, mask, NULL);
+		err = generic_permission(inode, mask, flags, NULL);
 	return err;
 }
 
@@ -1789,7 +1815,11 @@ int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry,
 		else
 			stat->dev = 0;
 		if (S_ISDIR(inode->i_mode)) {
-			stat->size = ci->i_rbytes;
+			if (ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb),
+						RBYTES))
+				stat->size = ci->i_rbytes;
+			else
+				stat->size = ci->i_files + ci->i_subdirs;
 			stat->blocks = 0;
 			stat->blksize = 65536;
 		}
diff --git a/fs/ceph/ioctl.h b/fs/ceph/ioctl.h
index a6ce54e94eb5..52e8fd74d450 100644
--- a/fs/ceph/ioctl.h
+++ b/fs/ceph/ioctl.h
@@ -4,7 +4,7 @@
 #include <linux/ioctl.h>
 #include <linux/types.h>
 
-#define CEPH_IOCTL_MAGIC 0x98
+#define CEPH_IOCTL_MAGIC 0x97
 
 /* just use u64 to align sanely on all archs */
 struct ceph_ioctl_layout {
diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c
index 40abde93c345..476b329867d4 100644
--- a/fs/ceph/locks.c
+++ b/fs/ceph/locks.c
@@ -11,40 +11,68 @@
  * Implement fcntl and flock locking functions.
  */
 static int ceph_lock_message(u8 lock_type, u16 operation, struct file *file,
-			     u64 pid, u64 pid_ns,
-			     int cmd, u64 start, u64 length, u8 wait)
+			     int cmd, u8 wait, struct file_lock *fl)
 {
 	struct inode *inode = file->f_dentry->d_inode;
 	struct ceph_mds_client *mdsc =
 		ceph_sb_to_client(inode->i_sb)->mdsc;
 	struct ceph_mds_request *req;
 	int err;
+	u64 length = 0;
 
 	req = ceph_mdsc_create_request(mdsc, operation, USE_AUTH_MDS);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 	req->r_inode = igrab(inode);
 
+	/* mds requires start and length rather than start and end */
+	if (LLONG_MAX == fl->fl_end)
+		length = 0;
+	else
+		length = fl->fl_end - fl->fl_start + 1;
+
 	dout("ceph_lock_message: rule: %d, op: %d, pid: %llu, start: %llu, "
 	     "length: %llu, wait: %d, type`: %d", (int)lock_type,
-	     (int)operation, pid, start, length, wait, cmd);
+	     (int)operation, (u64)fl->fl_pid, fl->fl_start,
+	     length, wait, fl->fl_type);
+
 
 	req->r_args.filelock_change.rule = lock_type;
 	req->r_args.filelock_change.type = cmd;
-	req->r_args.filelock_change.pid = cpu_to_le64(pid);
+	req->r_args.filelock_change.pid = cpu_to_le64((u64)fl->fl_pid);
 	/* This should be adjusted, but I'm not sure if
 	   namespaces actually get id numbers*/
 	req->r_args.filelock_change.pid_namespace =
-		cpu_to_le64((u64)pid_ns);
-	req->r_args.filelock_change.start = cpu_to_le64(start);
+		cpu_to_le64((u64)(unsigned long)fl->fl_nspid);
+	req->r_args.filelock_change.start = cpu_to_le64(fl->fl_start);
 	req->r_args.filelock_change.length = cpu_to_le64(length);
 	req->r_args.filelock_change.wait = wait;
 
 	err = ceph_mdsc_do_request(mdsc, inode, req);
+
+	if ( operation == CEPH_MDS_OP_GETFILELOCK){
+		fl->fl_pid = le64_to_cpu(req->r_reply_info.filelock_reply->pid);
+		if (CEPH_LOCK_SHARED == req->r_reply_info.filelock_reply->type)
+			fl->fl_type = F_RDLCK;
+		else if (CEPH_LOCK_EXCL == req->r_reply_info.filelock_reply->type)
+			fl->fl_type = F_WRLCK;
+		else
+			fl->fl_type = F_UNLCK;
+
+		fl->fl_start = le64_to_cpu(req->r_reply_info.filelock_reply->start);
+		length = le64_to_cpu(req->r_reply_info.filelock_reply->start) +
+						 le64_to_cpu(req->r_reply_info.filelock_reply->length);
+		if (length >= 1)
+			fl->fl_end = length -1;
+		else
+			fl->fl_end = 0;
+
+	}
 	ceph_mdsc_put_request(req);
 	dout("ceph_lock_message: rule: %d, op: %d, pid: %llu, start: %llu, "
-	     "length: %llu, wait: %d, type`: %d err code %d", (int)lock_type,
-	     (int)operation, pid, start, length, wait, cmd, err);
+	     "length: %llu, wait: %d, type`: %d, err code %d", (int)lock_type,
+	     (int)operation, (u64)fl->fl_pid, fl->fl_start,
+	     length, wait, fl->fl_type, err);
 	return err;
 }
 
@@ -54,7 +82,6 @@ static int ceph_lock_message(u8 lock_type, u16 operation, struct file *file,
  */
 int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
 {
-	u64 length;
 	u8 lock_cmd;
 	int err;
 	u8 wait = 0;
@@ -76,29 +103,20 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
 	else
 		lock_cmd = CEPH_LOCK_UNLOCK;
 
-	if (LLONG_MAX == fl->fl_end)
-		length = 0;
-	else
-		length = fl->fl_end - fl->fl_start + 1;
-
-	err = ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
-				(u64)fl->fl_pid,
-				(u64)(unsigned long)fl->fl_nspid,
-				lock_cmd, fl->fl_start,
-				length, wait);
+	err = ceph_lock_message(CEPH_LOCK_FCNTL, op, file, lock_cmd, wait, fl);
 	if (!err) {
-		dout("mds locked, locking locally");
-		err = posix_lock_file(file, fl, NULL);
-		if (err && (CEPH_MDS_OP_SETFILELOCK == op)) {
-			/* undo! This should only happen if the kernel detects
-			 * local deadlock. */
-			ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
-					  (u64)fl->fl_pid,
-					  (u64)(unsigned long)fl->fl_nspid,
-					  CEPH_LOCK_UNLOCK, fl->fl_start,
-					  length, 0);
-			dout("got %d on posix_lock_file, undid lock", err);
+		if ( op != CEPH_MDS_OP_GETFILELOCK ){
+			dout("mds locked, locking locally");
+			err = posix_lock_file(file, fl, NULL);
+			if (err && (CEPH_MDS_OP_SETFILELOCK == op)) {
+				/* undo! This should only happen if the kernel detects
+				 * local deadlock. */
+				ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
+						  CEPH_LOCK_UNLOCK, 0, fl);
+				dout("got %d on posix_lock_file, undid lock", err);
+			}
 		}
+
 	} else {
 		dout("mds returned error code %d", err);
 	}
@@ -107,7 +125,6 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
 
 int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
 {
-	u64 length;
 	u8 lock_cmd;
 	int err;
 	u8 wait = 1;
@@ -127,26 +144,15 @@ int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
 		lock_cmd = CEPH_LOCK_EXCL;
 	else
 		lock_cmd = CEPH_LOCK_UNLOCK;
-	/* mds requires start and length rather than start and end */
-	if (LLONG_MAX == fl->fl_end)
-		length = 0;
-	else
-		length = fl->fl_end - fl->fl_start + 1;
 
 	err = ceph_lock_message(CEPH_LOCK_FLOCK, CEPH_MDS_OP_SETFILELOCK,
-				file, (u64)fl->fl_pid,
-				(u64)(unsigned long)fl->fl_nspid,
-				lock_cmd, fl->fl_start,
-				length, wait);
+				file, lock_cmd, wait, fl);
 	if (!err) {
 		err = flock_lock_file_wait(file, fl);
 		if (err) {
 			ceph_lock_message(CEPH_LOCK_FLOCK,
 					  CEPH_MDS_OP_SETFILELOCK,
-					  file, (u64)fl->fl_pid,
-					  (u64)(unsigned long)fl->fl_nspid,
-					  CEPH_LOCK_UNLOCK, fl->fl_start,
-					  length, 0);
+					  file, CEPH_LOCK_UNLOCK, 0, fl);
 			dout("got %d on flock_lock_file_wait, undid lock", err);
 		}
 	} else {
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 3142b15940c2..a1ee8fa3a8e7 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -6,7 +6,6 @@
 #include <linux/sched.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 
 #include "super.h"
 #include "mds_client.h"
@@ -61,7 +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;
 
@@ -75,6 +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;
@@ -89,12 +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;
 
@@ -115,7 +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;
 	}
@@ -135,7 +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;
@@ -183,7 +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++;
@@ -203,10 +211,45 @@ out_bad:
 }
 
 /*
+ * parse fcntl F_GETLK results
+ */
+static int parse_reply_info_filelock(void **p, void *end,
+				     struct ceph_mds_reply_info_parsed *info,
+				     int features)
+{
+	if (*p + sizeof(*info->filelock_reply) > end)
+		goto bad;
+
+	info->filelock_reply = *p;
+	*p += sizeof(*info->filelock_reply);
+
+	if (unlikely(*p != end))
+		goto bad;
+	return 0;
+
+bad:
+	return -EIO;
+}
+
+/*
+ * parse extra results
+ */
+static int parse_reply_info_extra(void **p, void *end,
+				  struct ceph_mds_reply_info_parsed *info,
+				  int features)
+{
+	if (info->head->op == CEPH_MDS_OP_GETFILELOCK)
+		return parse_reply_info_filelock(p, end, info, features);
+	else
+		return parse_reply_info_dir(p, end, info, features);
+}
+
+/*
  * 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;
@@ -219,15 +262,15 @@ 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;
 	}
 
-	/* dir content */
+	/* extra */
 	ceph_decode_32_safe(&p, end, len, bad);
 	if (len > 0) {
-		err = parse_reply_info_dir(&p, p+len, info);
+		err = parse_reply_info_extra(&p, p+len, info, features);
 		if (err < 0)
 			goto out_bad;
 	}
@@ -529,6 +572,9 @@ static void __register_request(struct ceph_mds_client *mdsc,
 	ceph_mdsc_get_request(req);
 	__insert_request(mdsc, req);
 
+	req->r_uid = current_fsuid();
+	req->r_gid = current_fsgid();
+
 	if (dir) {
 		struct ceph_inode_info *ci = ceph_inode(dir);
 
@@ -620,7 +666,7 @@ 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;
 		}
 	}
@@ -647,9 +693,11 @@ static int __choose_mds(struct ceph_mds_client *mdsc,
 				dout("choose_mds %p %llx.%llx "
 				     "frag %u mds%d (%d/%d)\n",
 				     inode, ceph_vinop(inode),
-				     frag.frag, frag.mds,
+				     frag.frag, mds,
 				     (int)r, frag.ndist);
-				return mds;
+				if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
+				    CEPH_MDS_STATE_ACTIVE)
+					return mds;
 			}
 
 			/* since this file/dir wasn't known to be
@@ -662,7 +710,9 @@ static int __choose_mds(struct ceph_mds_client *mdsc,
 				dout("choose_mds %p %llx.%llx "
 				     "frag %u mds%d (auth)\n",
 				     inode, ceph_vinop(inode), frag.frag, mds);
-				return mds;
+				if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
+				    CEPH_MDS_STATE_ACTIVE)
+					return mds;
 			}
 		}
 	}
@@ -1452,7 +1502,7 @@ retry:
 	*base = ceph_ino(temp->d_inode);
 	*plen = len;
 	dout("build_path on %p %d built %llx '%.*s'\n",
-	     dentry, atomic_read(&dentry->d_count), *base, len, path);
+	     dentry, dentry->d_count, *base, len, path);
 	return path;
 }
 
@@ -1588,8 +1638,8 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc,
 
 	head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
 	head->op = cpu_to_le32(req->r_op);
-	head->caller_uid = cpu_to_le32(current_fsuid());
-	head->caller_gid = cpu_to_le32(current_fsgid());
+	head->caller_uid = cpu_to_le32(req->r_uid);
+	head->caller_gid = cpu_to_le32(req->r_gid);
 	head->args = req->r_args;
 
 	ceph_encode_filepath(&p, end, ino1, path1);
@@ -1659,7 +1709,6 @@ static int __prepare_send_request(struct ceph_mds_client *mdsc,
 	struct ceph_msg *msg;
 	int flags = 0;
 
-	req->r_mds = mds;
 	req->r_attempts++;
 	if (req->r_inode) {
 		struct ceph_cap *cap =
@@ -1746,6 +1795,8 @@ static int __do_request(struct ceph_mds_client *mdsc,
 		goto finish;
 	}
 
+	put_request_session(req);
+
 	mds = __choose_mds(mdsc, req);
 	if (mds < 0 ||
 	    ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
@@ -1763,6 +1814,8 @@ static int __do_request(struct ceph_mds_client *mdsc,
 			goto finish;
 		}
 	}
+	req->r_session = get_session(session);
+
 	dout("do_request mds%d session %p state %s\n", mds, session,
 	     session_state_name(session->s_state));
 	if (session->s_state != CEPH_MDS_SESSION_OPEN &&
@@ -1775,7 +1828,6 @@ static int __do_request(struct ceph_mds_client *mdsc,
 	}
 
 	/* send request */
-	req->r_session = get_session(session);
 	req->r_resend_mds = -1;   /* forget any previous mds hint */
 
 	if (req->r_request_started == 0)   /* note request start time */
@@ -1829,7 +1881,6 @@ static void kick_requests(struct ceph_mds_client *mdsc, int mds)
 		if (req->r_session &&
 		    req->r_session->s_mds == mds) {
 			dout(" kicking tid %llu\n", req->r_tid);
-			put_request_session(req);
 			__do_request(mdsc, req);
 		}
 	}
@@ -2022,8 +2073,11 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
 			goto out;
 		} else  {
 			struct ceph_inode_info *ci = ceph_inode(req->r_inode);
-			struct ceph_cap *cap =
-				ceph_get_cap_for_mds(ci, req->r_mds);;
+			struct ceph_cap *cap = NULL;
+
+			if (req->r_session)
+				cap = ceph_get_cap_for_mds(ci,
+						   req->r_session->s_mds);
 
 			dout("already using auth");
 			if ((!cap || cap != ci->i_auth_cap) ||
@@ -2067,12 +2121,12 @@ 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);
 	if (err < 0) {
-		pr_err("mdsc_handle_reply got corrupt reply mds%d\n", mds);
+		pr_err("mdsc_handle_reply got corrupt reply mds%d(tid:%lld)\n", mds, tid);
 		ceph_msg_dump(msg);
 		goto out_err;
 	}
@@ -2092,7 +2146,8 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
 	mutex_lock(&req->r_fill_mutex);
 	err = ceph_fill_trace(mdsc->fsc->sb, req, req->r_session);
 	if (err == 0) {
-		if (result == 0 && rinfo->dir_nr)
+		if (result == 0 && req->r_op != CEPH_MDS_OP_GETFILELOCK &&
+		    rinfo->dir_nr)
 			ceph_readdir_prepopulate(req, req->r_session);
 		ceph_unreserve_caps(mdsc, &req->r_caps_reservation);
 	}
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index d66d63c72355..4e3a9cc0bba6 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;
@@ -42,26 +43,37 @@ struct ceph_mds_reply_info_in {
 };
 
 /*
- * parsed info about an mds reply, including information about the
- * target inode and/or its parent directory and dentry, and directory
- * contents (for readdir results).
+ * parsed info about an mds reply, including information about
+ * either: 1) the target inode and/or its parent directory and dentry,
+ * and directory contents (for readdir results), or
+ * 2) the file range lock info (for fcntl F_GETLK results).
  */
 struct ceph_mds_reply_info_parsed {
 	struct ceph_mds_reply_head    *head;
 
+	/* trace */
 	struct ceph_mds_reply_info_in diri, targeti;
 	struct ceph_mds_reply_dirfrag *dirfrag;
 	char                          *dname;
 	u32                           dname_len;
 	struct ceph_mds_reply_lease   *dlease;
 
-	struct ceph_mds_reply_dirfrag *dir_dir;
-	int                           dir_nr;
-	char                          **dir_dname;
-	u32                           *dir_dname_len;
-	struct ceph_mds_reply_lease   **dir_dlease;
-	struct ceph_mds_reply_info_in *dir_in;
-	u8                            dir_complete, dir_end;
+	/* extra */
+	union {
+		/* for fcntl F_GETLK results */
+		struct ceph_filelock *filelock_reply;
+
+		/* for readdir results */
+		struct {
+			struct ceph_mds_reply_dirfrag *dir_dir;
+			int                           dir_nr;
+			char                          **dir_dname;
+			u32                           *dir_dname_len;
+			struct ceph_mds_reply_lease   **dir_dlease;
+			struct ceph_mds_reply_info_in *dir_in;
+			u8                            dir_complete, dir_end;
+		};
+	};
 
 	/* encoded blob describing snapshot contexts for certain
 	   operations (e.g., open) */
@@ -154,7 +166,6 @@ struct ceph_mds_request {
 	struct ceph_mds_client *r_mdsc;
 
 	int r_op;                    /* mds op code */
-	int r_mds;
 
 	/* operation on what? */
 	struct inode *r_inode;              /* arg1 */
@@ -170,6 +181,8 @@ struct ceph_mds_request {
 
 	union ceph_mds_request_args r_args;
 	int r_fmode;        /* file mode, if expecting cap */
+	uid_t r_uid;
+	gid_t r_gid;
 
 	/* for choosing which mds to send this request to */
 	int r_direct_mode;
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index 08b460ae0539..9c5085465a63 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -290,6 +290,8 @@ static int parse_mount_options(struct ceph_mount_options **pfsopt,
 
         fsopt->rsize = CEPH_MOUNT_RSIZE_DEFAULT;
         fsopt->snapdir_name = kstrdup(CEPH_SNAPDIRNAME_DEFAULT, GFP_KERNEL);
+	fsopt->caps_wanted_delay_min = CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT;
+	fsopt->caps_wanted_delay_max = CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT;
         fsopt->cap_release_safety = CEPH_CAP_RELEASE_SAFETY_DEFAULT;
         fsopt->max_readdir = CEPH_MAX_READDIR_DEFAULT;
         fsopt->max_readdir_bytes = CEPH_MAX_READDIR_BYTES_DEFAULT;
@@ -428,7 +430,8 @@ 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;
@@ -443,13 +446,17 @@ struct ceph_fs_client *create_fs_client(struct ceph_mount_options *fsopt,
 		goto fail_client;
 
 	err = -ENOMEM;
-	fsc->wb_wq = create_workqueue("ceph-writeback");
+	/*
+	 * The number of concurrent works can be high but they don't need
+	 * to be processed in parallel, limit concurrency.
+	 */
+	fsc->wb_wq = alloc_workqueue("ceph-writeback", 0, 1);
 	if (fsc->wb_wq == NULL)
 		goto fail_bdi;
-	fsc->pg_inv_wq = create_singlethread_workqueue("ceph-pg-invalid");
+	fsc->pg_inv_wq = alloc_workqueue("ceph-pg-invalid", 0, 1);
 	if (fsc->pg_inv_wq == NULL)
 		goto fail_wb_wq;
-	fsc->trunc_wq = create_singlethread_workqueue("ceph-trunc");
+	fsc->trunc_wq = alloc_workqueue("ceph-trunc", 0, 1);
 	if (fsc->trunc_wq == NULL)
 		goto fail_pg_inv_wq;
 
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 1886294e12f7..20b907d76ae2 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;
 
@@ -293,9 +294,7 @@ struct ceph_inode_info {
 	int i_rd_ref, i_rdcache_ref, i_wr_ref;
 	int i_wrbuffer_ref, i_wrbuffer_ref_head;
 	u32 i_shared_gen;       /* increment each time we get FILE_SHARED */
-	u32 i_rdcache_gen;      /* we increment this each time we get
-				   FILE_CACHE.  If it's non-zero, we
-				   _may_ have cached pages. */
+	u32 i_rdcache_gen;      /* incremented each time we get FILE_CACHE. */
 	u32 i_rdcache_revoking; /* RDCACHE gen to async invalidate, if any */
 
 	struct list_head i_unsafe_writes; /* uncommitted sync writes */
@@ -667,7 +666,7 @@ extern void ceph_queue_invalidate(struct inode *inode);
 extern void ceph_queue_writeback(struct inode *inode);
 
 extern int ceph_do_getattr(struct inode *inode, int mask);
-extern int ceph_permission(struct inode *inode, int mask);
+extern int ceph_permission(struct inode *inode, int mask, unsigned int flags);
 extern int ceph_setattr(struct dentry *dentry, struct iattr *attr);
 extern int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry,
 			struct kstat *stat);
@@ -770,6 +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/fs/ceph/xattr.c b/fs/ceph/xattr.c
index 6e12a6ba5f79..8c9eba6ef9df 100644
--- a/fs/ceph/xattr.c
+++ b/fs/ceph/xattr.c
@@ -219,6 +219,7 @@ static struct ceph_inode_xattr *__get_xattr(struct ceph_inode_info *ci,
 	struct rb_node **p;
 	struct rb_node *parent = NULL;
 	struct ceph_inode_xattr *xattr = NULL;
+	int name_len = strlen(name);
 	int c;
 
 	p = &ci->i_xattrs.index.rb_node;
@@ -226,6 +227,8 @@ static struct ceph_inode_xattr *__get_xattr(struct ceph_inode_info *ci,
 		parent = *p;
 		xattr = rb_entry(parent, struct ceph_inode_xattr, node);
 		c = strncmp(name, xattr->name, xattr->name_len);
+		if (c == 0 && name_len > xattr->name_len)
+			c = 1;
 		if (c < 0)
 			p = &(*p)->rb_left;
 		else if (c > 0)