30 files changed, 447 insertions, 220 deletions

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 4545e2e2ad45..5699bbc23feb 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -931,7 +931,7 @@ static int check_async_write(struct inode *inode, unsigned long bio_flags)
 	if (bio_flags & EXTENT_BIO_TREE_LOG)
 		return 0;
 #ifdef CONFIG_X86
-	if (static_cpu_has_safe(X86_FEATURE_XMM4_2))
+	if (static_cpu_has(X86_FEATURE_XMM4_2))
 		return 0;
 #endif
 	return 1;
diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c
index 7cf8509deda7..2c849b08a91b 100644
--- a/fs/btrfs/root-tree.c
+++ b/fs/btrfs/root-tree.c
@@ -310,8 +310,16 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root)
 		set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state);
 
 		err = btrfs_insert_fs_root(root->fs_info, root);
+		/*
+		 * The root might have been inserted already, as before we look
+		 * for orphan roots, log replay might have happened, which
+		 * triggers a transaction commit and qgroup accounting, which
+		 * in turn reads and inserts fs roots while doing backref
+		 * walking.
+		 */
+		if (err == -EEXIST)
+			err = 0;
 		if (err) {
-			BUG_ON(err == -EEXIST);
 			btrfs_free_fs_root(root);
 			break;
 		}
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index c22213789090..19adeb0ef82a 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1756,6 +1756,10 @@ int ceph_pool_perm_check(struct ceph_inode_info *ci, int need)
 	u32 pool;
 	int ret, flags;
 
+	/* does not support pool namespace yet */
+	if (ci->i_pool_ns_len)
+		return -EIO;
+
 	if (ceph_test_mount_opt(ceph_inode_to_client(&ci->vfs_inode),
 				NOPOOLPERM))
 		return 0;
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index cdbf8cf3d52c..6fe0ad26a7df 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -2753,7 +2753,8 @@ static void handle_cap_grant(struct ceph_mds_client *mdsc,
 			     void *inline_data, int inline_len,
 			     struct ceph_buffer *xattr_buf,
 			     struct ceph_mds_session *session,
-			     struct ceph_cap *cap, int issued)
+			     struct ceph_cap *cap, int issued,
+			     u32 pool_ns_len)
 	__releases(ci->i_ceph_lock)
 	__releases(mdsc->snap_rwsem)
 {
@@ -2873,6 +2874,8 @@ static void handle_cap_grant(struct ceph_mds_client *mdsc,
 	if (newcaps & (CEPH_CAP_ANY_FILE_RD | CEPH_CAP_ANY_FILE_WR)) {
 		/* file layout may have changed */
 		ci->i_layout = grant->layout;
+		ci->i_pool_ns_len = pool_ns_len;
+
 		/* size/truncate_seq? */
 		queue_trunc = ceph_fill_file_size(inode, issued,
 					le32_to_cpu(grant->truncate_seq),
@@ -3411,6 +3414,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	u32  inline_len = 0;
 	void *snaptrace;
 	size_t snaptrace_len;
+	u32 pool_ns_len = 0;
 	void *p, *end;
 
 	dout("handle_caps from mds%d\n", mds);
@@ -3463,6 +3467,21 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 		p += inline_len;
 	}
 
+	if (le16_to_cpu(msg->hdr.version) >= 8) {
+		u64 flush_tid;
+		u32 caller_uid, caller_gid;
+		u32 osd_epoch_barrier;
+		/* version >= 5 */
+		ceph_decode_32_safe(&p, end, osd_epoch_barrier, bad);
+		/* version >= 6 */
+		ceph_decode_64_safe(&p, end, flush_tid, bad);
+		/* version >= 7 */
+		ceph_decode_32_safe(&p, end, caller_uid, bad);
+		ceph_decode_32_safe(&p, end, caller_gid, bad);
+		/* version >= 8 */
+		ceph_decode_32_safe(&p, end, pool_ns_len, bad);
+	}
+
 	/* lookup ino */
 	inode = ceph_find_inode(sb, vino);
 	ci = ceph_inode(inode);
@@ -3518,7 +3537,8 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 				  &cap, &issued);
 		handle_cap_grant(mdsc, inode, h,
 				 inline_version, inline_data, inline_len,
-				 msg->middle, session, cap, issued);
+				 msg->middle, session, cap, issued,
+				 pool_ns_len);
 		if (realm)
 			ceph_put_snap_realm(mdsc, realm);
 		goto done_unlocked;
@@ -3542,7 +3562,8 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 		issued |= __ceph_caps_dirty(ci);
 		handle_cap_grant(mdsc, inode, h,
 				 inline_version, inline_data, inline_len,
-				 msg->middle, session, cap, issued);
+				 msg->middle, session, cap, issued,
+				 pool_ns_len);
 		goto done_unlocked;
 
 	case CEPH_CAP_OP_FLUSH_ACK:
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index fb4ba2e4e2a5..5849b88bbed3 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -396,6 +396,7 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 	ci->i_symlink = NULL;
 
 	memset(&ci->i_dir_layout, 0, sizeof(ci->i_dir_layout));
+	ci->i_pool_ns_len = 0;
 
 	ci->i_fragtree = RB_ROOT;
 	mutex_init(&ci->i_fragtree_mutex);
@@ -756,6 +757,7 @@ static int fill_inode(struct inode *inode, struct page *locked_page,
 		if (ci->i_layout.fl_pg_pool != info->layout.fl_pg_pool)
 			ci->i_ceph_flags &= ~CEPH_I_POOL_PERM;
 		ci->i_layout = info->layout;
+		ci->i_pool_ns_len = iinfo->pool_ns_len;
 
 		queue_trunc = ceph_fill_file_size(inode, issued,
 					le32_to_cpu(info->truncate_seq),
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index e7b130a637f9..911d64d865f1 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -100,6 +100,14 @@ static int parse_reply_info_in(void **p, void *end,
 	} else
 		info->inline_version = CEPH_INLINE_NONE;
 
+	if (features & CEPH_FEATURE_FS_FILE_LAYOUT_V2) {
+		ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
+		ceph_decode_need(p, end, info->pool_ns_len, bad);
+		*p += info->pool_ns_len;
+	} else {
+		info->pool_ns_len = 0;
+	}
+
 	return 0;
 bad:
 	return err;
@@ -2298,6 +2306,14 @@ int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
 		ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
 				  CEPH_CAP_PIN);
 
+	/* deny access to directories with pool_ns layouts */
+	if (req->r_inode && S_ISDIR(req->r_inode->i_mode) &&
+	    ceph_inode(req->r_inode)->i_pool_ns_len)
+		return -EIO;
+	if (req->r_locked_dir &&
+	    ceph_inode(req->r_locked_dir)->i_pool_ns_len)
+		return -EIO;
+
 	/* issue */
 	mutex_lock(&mdsc->mutex);
 	__register_request(mdsc, req, dir);
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index ccf11ef0ca87..37712ccffcc6 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -44,6 +44,7 @@ struct ceph_mds_reply_info_in {
 	u64 inline_version;
 	u32 inline_len;
 	char *inline_data;
+	u32 pool_ns_len;
 };
 
 /*
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 75b7d125ce66..9c458eb52245 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -287,6 +287,7 @@ struct ceph_inode_info {
 
 	struct ceph_dir_layout i_dir_layout;
 	struct ceph_file_layout i_layout;
+	size_t i_pool_ns_len;
 	char *i_symlink;
 
 	/* for dirs */
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index c48ca13673e3..2eea40353e60 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -1013,7 +1013,6 @@ const struct file_operations cifs_file_strict_ops = {
 	.llseek = cifs_llseek,
 	.unlocked_ioctl	= cifs_ioctl,
 	.clone_file_range = cifs_clone_file_range,
-	.clone_file_range = cifs_clone_file_range,
 	.setlease = cifs_setlease,
 	.fallocate = cifs_fallocate,
 };
diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
index 68c4547528c4..83aac8ba50b0 100644
--- a/fs/cifs/cifsfs.h
+++ b/fs/cifs/cifsfs.h
@@ -31,19 +31,15 @@
  * so that it will fit. We use hash_64 to convert the value to 31 bits, and
  * then add 1, to ensure that we don't end up with a 0 as the value.
  */
-#if BITS_PER_LONG == 64
 static inline ino_t
 cifs_uniqueid_to_ino_t(u64 fileid)
 {
+	if ((sizeof(ino_t)) < (sizeof(u64)))
+		return (ino_t)hash_64(fileid, (sizeof(ino_t) * 8) - 1) + 1;
+
 	return (ino_t)fileid;
+
 }
-#else
-static inline ino_t
-cifs_uniqueid_to_ino_t(u64 fileid)
-{
-	return (ino_t)hash_64(fileid, (sizeof(ino_t) * 8) - 1) + 1;
-}
-#endif
 
 extern struct file_system_type cifs_fs_type;
 extern const struct address_space_operations cifs_addr_ops;
diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c
index 90b4f9f7de66..76fcb50295a3 100644
--- a/fs/cifs/cifssmb.c
+++ b/fs/cifs/cifssmb.c
@@ -1396,11 +1396,10 @@ openRetry:
  * current bigbuf.
  */
 static int
-cifs_readv_discard(struct TCP_Server_Info *server, struct mid_q_entry *mid)
+discard_remaining_data(struct TCP_Server_Info *server)
 {
 	unsigned int rfclen = get_rfc1002_length(server->smallbuf);
 	int remaining = rfclen + 4 - server->total_read;
-	struct cifs_readdata *rdata = mid->callback_data;
 
 	while (remaining > 0) {
 		int length;
@@ -1414,10 +1413,20 @@ cifs_readv_discard(struct TCP_Server_Info *server, struct mid_q_entry *mid)
 		remaining -= length;
 	}
 
-	dequeue_mid(mid, rdata->result);
 	return 0;
 }
 
+static int
+cifs_readv_discard(struct TCP_Server_Info *server, struct mid_q_entry *mid)
+{
+	int length;
+	struct cifs_readdata *rdata = mid->callback_data;
+
+	length = discard_remaining_data(server);
+	dequeue_mid(mid, rdata->result);
+	return length;
+}
+
 int
 cifs_readv_receive(struct TCP_Server_Info *server, struct mid_q_entry *mid)
 {
@@ -1446,6 +1455,12 @@ cifs_readv_receive(struct TCP_Server_Info *server, struct mid_q_entry *mid)
 		return length;
 	server->total_read += length;
 
+	if (server->ops->is_status_pending &&
+	    server->ops->is_status_pending(buf, server, 0)) {
+		discard_remaining_data(server);
+		return -1;
+	}
+
 	/* Was the SMB read successful? */
 	rdata->result = server->ops->map_error(buf, false);
 	if (rdata->result != 0) {
diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c
index 10f8d5cf5681..42e1f440eb1e 100644
--- a/fs/cifs/smb2pdu.c
+++ b/fs/cifs/smb2pdu.c
@@ -1106,21 +1106,25 @@ parse_lease_state(struct TCP_Server_Info *server, struct smb2_create_rsp *rsp,
 {
 	char *data_offset;
 	struct create_context *cc;
-	unsigned int next = 0;
+	unsigned int next;
+	unsigned int remaining;
 	char *name;
 
 	data_offset = (char *)rsp + 4 + le32_to_cpu(rsp->CreateContextsOffset);
+	remaining = le32_to_cpu(rsp->CreateContextsLength);
 	cc = (struct create_context *)data_offset;
-	do {
-		cc = (struct create_context *)((char *)cc + next);
+	while (remaining >= sizeof(struct create_context)) {
 		name = le16_to_cpu(cc->NameOffset) + (char *)cc;
-		if (le16_to_cpu(cc->NameLength) != 4 ||
-		    strncmp(name, "RqLs", 4)) {
-			next = le32_to_cpu(cc->Next);
-			continue;
-		}
-		return server->ops->parse_lease_buf(cc, epoch);
-	} while (next != 0);
+		if (le16_to_cpu(cc->NameLength) == 4 &&
+		    strncmp(name, "RqLs", 4) == 0)
+			return server->ops->parse_lease_buf(cc, epoch);
+
+		next = le32_to_cpu(cc->Next);
+		if (!next)
+			break;
+		remaining -= next;
+		cc = (struct create_context *)((char *)cc + next);
+	}
 
 	return 0;
 }
diff --git a/fs/dax.c b/fs/dax.c
index 711172450da6..bbb2ad783770 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -1056,6 +1056,7 @@ EXPORT_SYMBOL_GPL(dax_pmd_fault);
 int dax_pfn_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct file *file = vma->vm_file;
+	int error;
 
 	/*
 	 * We pass NO_SECTOR to dax_radix_entry() because we expect that a
@@ -1065,7 +1066,13 @@ int dax_pfn_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	 * saves us from having to make a call to get_block() here to look
 	 * up the sector.
 	 */
-	dax_radix_entry(file->f_mapping, vmf->pgoff, NO_SECTOR, false, true);
+	error = dax_radix_entry(file->f_mapping, vmf->pgoff, NO_SECTOR, false,
+			true);
+
+	if (error == -ENOMEM)
+		return VM_FAULT_OOM;
+	if (error)
+		return VM_FAULT_SIGBUS;
 	return VM_FAULT_NOPAGE;
 }
 EXPORT_SYMBOL_GPL(dax_pfn_mkwrite);
diff --git a/fs/dcache.c b/fs/dcache.c
index 92d5140de851..2398f9f94337 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -269,9 +269,6 @@ static inline int dname_external(const struct dentry *dentry)
 	return dentry->d_name.name != dentry->d_iname;
 }
 
-/*
- * Make sure other CPUs see the inode attached before the type is set.
- */
 static inline void __d_set_inode_and_type(struct dentry *dentry,
 					  struct inode *inode,
 					  unsigned type_flags)
@@ -279,28 +276,18 @@ static inline void __d_set_inode_and_type(struct dentry *dentry,
 	unsigned flags;
 
 	dentry->d_inode = inode;
-	smp_wmb();
 	flags = READ_ONCE(dentry->d_flags);
 	flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU);
 	flags |= type_flags;
 	WRITE_ONCE(dentry->d_flags, flags);
 }
 
-/*
- * Ideally, we want to make sure that other CPUs see the flags cleared before
- * the inode is detached, but this is really a violation of RCU principles
- * since the ordering suggests we should always set inode before flags.
- *
- * We should instead replace or discard the entire dentry - but that sucks
- * performancewise on mass deletion/rename.
- */
 static inline void __d_clear_type_and_inode(struct dentry *dentry)
 {
 	unsigned flags = READ_ONCE(dentry->d_flags);
 
 	flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU);
 	WRITE_ONCE(dentry->d_flags, flags);
-	smp_wmb();
 	dentry->d_inode = NULL;
 }
 
@@ -370,9 +357,11 @@ static void dentry_unlink_inode(struct dentry * dentry)
 	__releases(dentry->d_inode->i_lock)
 {
 	struct inode *inode = dentry->d_inode;
+
+	raw_write_seqcount_begin(&dentry->d_seq);
 	__d_clear_type_and_inode(dentry);
 	hlist_del_init(&dentry->d_u.d_alias);
-	dentry_rcuwalk_invalidate(dentry);
+	raw_write_seqcount_end(&dentry->d_seq);
 	spin_unlock(&dentry->d_lock);
 	spin_unlock(&inode->i_lock);
 	if (!inode->i_nlink)
@@ -1758,8 +1747,9 @@ static void __d_instantiate(struct dentry *dentry, struct inode *inode)
 	spin_lock(&dentry->d_lock);
 	if (inode)
 		hlist_add_head(&dentry->d_u.d_alias, &inode->i_dentry);
+	raw_write_seqcount_begin(&dentry->d_seq);
 	__d_set_inode_and_type(dentry, inode, add_flags);
-	dentry_rcuwalk_invalidate(dentry);
+	raw_write_seqcount_end(&dentry->d_seq);
 	spin_unlock(&dentry->d_lock);
 	fsnotify_d_instantiate(dentry, inode);
 }
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index e032a0423e35..4098acc701c3 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -390,6 +390,7 @@ data_copy:
 		*err = ext4_get_block(orig_inode, orig_blk_offset + i, bh, 0);
 		if (*err < 0)
 			break;
+		bh = bh->b_this_page;
 	}
 	if (!*err)
 		*err = block_commit_write(pagep[0], from, from + replaced_size);
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 1f76d8950a57..5c46ed9f3e14 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -223,6 +223,9 @@ static void wb_wait_for_completion(struct backing_dev_info *bdi,
 #define WB_FRN_HIST_MAX_SLOTS	(WB_FRN_HIST_THR_SLOTS / 2 + 1)
 					/* one round can affect upto 5 slots */
 
+static atomic_t isw_nr_in_flight = ATOMIC_INIT(0);
+static struct workqueue_struct *isw_wq;
+
 void __inode_attach_wb(struct inode *inode, struct page *page)
 {
 	struct backing_dev_info *bdi = inode_to_bdi(inode);
@@ -317,7 +320,6 @@ static void inode_switch_wbs_work_fn(struct work_struct *work)
 	struct inode_switch_wbs_context *isw =
 		container_of(work, struct inode_switch_wbs_context, work);
 	struct inode *inode = isw->inode;
-	struct super_block *sb = inode->i_sb;
 	struct address_space *mapping = inode->i_mapping;
 	struct bdi_writeback *old_wb = inode->i_wb;
 	struct bdi_writeback *new_wb = isw->new_wb;
@@ -424,8 +426,9 @@ skip_switch:
 	wb_put(new_wb);
 
 	iput(inode);
-	deactivate_super(sb);
 	kfree(isw);
+
+	atomic_dec(&isw_nr_in_flight);
 }
 
 static void inode_switch_wbs_rcu_fn(struct rcu_head *rcu_head)
@@ -435,7 +438,7 @@ static void inode_switch_wbs_rcu_fn(struct rcu_head *rcu_head)
 
 	/* needs to grab bh-unsafe locks, bounce to work item */
 	INIT_WORK(&isw->work, inode_switch_wbs_work_fn);
-	schedule_work(&isw->work);
+	queue_work(isw_wq, &isw->work);
 }
 
 /**
@@ -471,20 +474,20 @@ static void inode_switch_wbs(struct inode *inode, int new_wb_id)
 
 	/* while holding I_WB_SWITCH, no one else can update the association */
 	spin_lock(&inode->i_lock);
-
-	if (inode->i_state & (I_WB_SWITCH | I_FREEING) ||
-	    inode_to_wb(inode) == isw->new_wb)
-		goto out_unlock;
-
-	if (!atomic_inc_not_zero(&inode->i_sb->s_active))
-		goto out_unlock;
-
+	if (!(inode->i_sb->s_flags & MS_ACTIVE) ||
+	    inode->i_state & (I_WB_SWITCH | I_FREEING) ||
+	    inode_to_wb(inode) == isw->new_wb) {
+		spin_unlock(&inode->i_lock);
+		goto out_free;
+	}
 	inode->i_state |= I_WB_SWITCH;
 	spin_unlock(&inode->i_lock);
 
 	ihold(inode);
 	isw->inode = inode;
 
+	atomic_inc(&isw_nr_in_flight);
+
 	/*
 	 * In addition to synchronizing among switchers, I_WB_SWITCH tells
 	 * the RCU protected stat update paths to grab the mapping's
@@ -494,8 +497,6 @@ static void inode_switch_wbs(struct inode *inode, int new_wb_id)
 	call_rcu(&isw->rcu_head, inode_switch_wbs_rcu_fn);
 	return;
 
-out_unlock:
-	spin_unlock(&inode->i_lock);
 out_free:
 	if (isw->new_wb)
 		wb_put(isw->new_wb);
@@ -847,6 +848,33 @@ restart:
 		wb_put(last_wb);
 }
 
+/**
+ * cgroup_writeback_umount - flush inode wb switches for umount
+ *
+ * This function is called when a super_block is about to be destroyed and
+ * flushes in-flight inode wb switches.  An inode wb switch goes through
+ * RCU and then workqueue, so the two need to be flushed in order to ensure
+ * that all previously scheduled switches are finished.  As wb switches are
+ * rare occurrences and synchronize_rcu() can take a while, perform
+ * flushing iff wb switches are in flight.
+ */
+void cgroup_writeback_umount(void)
+{
+	if (atomic_read(&isw_nr_in_flight)) {
+		synchronize_rcu();
+		flush_workqueue(isw_wq);
+	}
+}
+
+static int __init cgroup_writeback_init(void)
+{
+	isw_wq = alloc_workqueue("inode_switch_wbs", 0, 0);
+	if (!isw_wq)
+		return -ENOMEM;
+	return 0;
+}
+fs_initcall(cgroup_writeback_init);
+
 #else	/* CONFIG_CGROUP_WRITEBACK */
 
 static struct bdi_writeback *
diff --git a/fs/jffs2/README.Locking b/fs/jffs2/README.Locking
index 3ea36554107f..8918ac905a3b 100644
--- a/fs/jffs2/README.Locking
+++ b/fs/jffs2/README.Locking
@@ -2,10 +2,6 @@
 	JFFS2 LOCKING DOCUMENTATION
 	---------------------------
 
-At least theoretically, JFFS2 does not require the Big Kernel Lock
-(BKL), which was always helpfully obtained for it by Linux 2.4 VFS
-code. It has its own locking, as described below.
-
 This document attempts to describe the existing locking rules for
 JFFS2. It is not expected to remain perfectly up to date, but ought to
 be fairly close.
@@ -69,6 +65,7 @@ Ordering constraints:
 	   any f->sem held.
 	2. Never attempt to lock two file mutexes in one thread.
 	   No ordering rules have been made for doing so.
+	3. Never lock a page cache page with f->sem held.
 
 
 	erase_completion_lock spinlock
diff --git a/fs/jffs2/build.c b/fs/jffs2/build.c
index 0ae91ad6df2d..b288c8ae1236 100644
--- a/fs/jffs2/build.c
+++ b/fs/jffs2/build.c
@@ -50,7 +50,8 @@ next_inode(int *i, struct jffs2_inode_cache *ic, struct jffs2_sb_info *c)
 
 
 static void jffs2_build_inode_pass1(struct jffs2_sb_info *c,
-				    struct jffs2_inode_cache *ic)
+				    struct jffs2_inode_cache *ic,
+				    int *dir_hardlinks)
 {
 	struct jffs2_full_dirent *fd;
 
@@ -69,19 +70,21 @@ static void jffs2_build_inode_pass1(struct jffs2_sb_info *c,
 			dbg_fsbuild("child \"%s\" (ino #%u) of dir ino #%u doesn't exist!\n",
 				  fd->name, fd->ino, ic->ino);
 			jffs2_mark_node_obsolete(c, fd->raw);
+			/* Clear the ic/raw union so it doesn't cause problems later. */
+			fd->ic = NULL;
 			continue;
 		}
 
+		/* From this point, fd->raw is no longer used so we can set fd->ic */
+		fd->ic = child_ic;
+		child_ic->pino_nlink++;
+		/* If we appear (at this stage) to have hard-linked directories,
+		 * set a flag to trigger a scan later */
 		if (fd->type == DT_DIR) {
-			if (child_ic->pino_nlink) {
-				JFFS2_ERROR("child dir \"%s\" (ino #%u) of dir ino #%u appears to be a hard link\n",
-					    fd->name, fd->ino, ic->ino);
-				/* TODO: What do we do about it? */
-			} else {
-				child_ic->pino_nlink = ic->ino;
-			}
-		} else
-			child_ic->pino_nlink++;
+			child_ic->flags |= INO_FLAGS_IS_DIR;
+			if (child_ic->pino_nlink > 1)
+				*dir_hardlinks = 1;
+		}
 
 		dbg_fsbuild("increased nlink for child \"%s\" (ino #%u)\n", fd->name, fd->ino);
 		/* Can't free scan_dents so far. We might need them in pass 2 */
@@ -95,8 +98,7 @@ static void jffs2_build_inode_pass1(struct jffs2_sb_info *c,
 */
 static int jffs2_build_filesystem(struct jffs2_sb_info *c)
 {
-	int ret;
-	int i;
+	int ret, i, dir_hardlinks = 0;
 	struct jffs2_inode_cache *ic;
 	struct jffs2_full_dirent *fd;
 	struct jffs2_full_dirent *dead_fds = NULL;
@@ -120,7 +122,7 @@ static int jffs2_build_filesystem(struct jffs2_sb_info *c)
 	/* Now scan the directory tree, increasing nlink according to every dirent found. */
 	for_each_inode(i, c, ic) {
 		if (ic->scan_dents) {
-			jffs2_build_inode_pass1(c, ic);
+			jffs2_build_inode_pass1(c, ic, &dir_hardlinks);
 			cond_resched();
 		}
 	}
@@ -156,6 +158,20 @@ static int jffs2_build_filesystem(struct jffs2_sb_info *c)
 	}
 
 	dbg_fsbuild("pass 2a complete\n");
+
+	if (dir_hardlinks) {
+		/* If we detected directory hardlinks earlier, *hopefully*
+		 * they are gone now because some of the links were from
+		 * dead directories which still had some old dirents lying
+		 * around and not yet garbage-collected, but which have
+		 * been discarded above. So clear the pino_nlink field
+		 * in each directory, so that the final scan below can
+		 * print appropriate warnings. */
+		for_each_inode(i, c, ic) {
+			if (ic->flags & INO_FLAGS_IS_DIR)
+				ic->pino_nlink = 0;
+		}
+	}
 	dbg_fsbuild("freeing temporary data structures\n");
 
 	/* Finally, we can scan again and free the dirent structs */
@@ -163,6 +179,33 @@ static int jffs2_build_filesystem(struct jffs2_sb_info *c)
 		while(ic->scan_dents) {
 			fd = ic->scan_dents;
 			ic->scan_dents = fd->next;
+			/* We do use the pino_nlink field to count nlink of
+			 * directories during fs build, so set it to the
+			 * parent ino# now. Now that there's hopefully only
+			 * one. */
+			if (fd->type == DT_DIR) {
+				if (!fd->ic) {
+					/* We'll have complained about it and marked the coresponding
+					   raw node obsolete already. Just skip it. */
+					continue;
+				}
+
+				/* We *have* to have set this in jffs2_build_inode_pass1() */
+				BUG_ON(!(fd->ic->flags & INO_FLAGS_IS_DIR));
+
+				/* We clear ic->pino_nlink ∀ directories' ic *only* if dir_hardlinks
+				 * is set. Otherwise, we know this should never trigger anyway, so
+				 * we don't do the check. And ic->pino_nlink still contains the nlink
+				 * value (which is 1). */
+				if (dir_hardlinks && fd->ic->pino_nlink) {
+					JFFS2_ERROR("child dir \"%s\" (ino #%u) of dir ino #%u is also hard linked from dir ino #%u\n",
+						    fd->name, fd->ino, ic->ino, fd->ic->pino_nlink);
+					/* Should we unlink it from its previous parent? */
+				}
+
+				/* For directories, ic->pino_nlink holds that parent inode # */
+				fd->ic->pino_nlink = ic->ino;
+			}
 			jffs2_free_full_dirent(fd);
 		}
 		ic->scan_dents = NULL;
@@ -241,11 +284,7 @@ static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *c,
 
 			/* Reduce nlink of the child. If it's now zero, stick it on the
 			   dead_fds list to be cleaned up later. Else just free the fd */
-
-			if (fd->type == DT_DIR)
-				child_ic->pino_nlink = 0;
-			else
-				child_ic->pino_nlink--;
+			child_ic->pino_nlink--;
 
 			if (!child_ic->pino_nlink) {
 				dbg_fsbuild("inode #%u (\"%s\") now has no links; adding to dead_fds list.\n",
diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c
index d211b8e18566..30c4c9ebb693 100644
--- a/fs/jffs2/dir.c
+++ b/fs/jffs2/dir.c
@@ -843,9 +843,14 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry,
 
 		pr_notice("%s(): Link succeeded, unlink failed (err %d). You now have a hard link\n",
 			  __func__, ret);
-		/* Might as well let the VFS know */
-		d_instantiate(new_dentry, d_inode(old_dentry));
-		ihold(d_inode(old_dentry));
+		/*
+		 * We can't keep the target in dcache after that.
+		 * For one thing, we can't afford dentry aliases for directories.
+		 * For another, if there was a victim, we _can't_ set new inode
+		 * for that sucker and we have to trigger mount eviction - the
+		 * caller won't do it on its own since we are returning an error.
+		 */
+		d_invalidate(new_dentry);
 		new_dir_i->i_mtime = new_dir_i->i_ctime = ITIME(now);
 		return ret;
 	}
diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index c5ac5944bc1b..cad86bac3453 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -137,39 +137,33 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 	struct page *pg;
 	struct inode *inode = mapping->host;
 	struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode);
-	struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb);
-	struct jffs2_raw_inode ri;
-	uint32_t alloc_len = 0;
 	pgoff_t index = pos >> PAGE_CACHE_SHIFT;
 	uint32_t pageofs = index << PAGE_CACHE_SHIFT;
 	int ret = 0;
 
-	jffs2_dbg(1, "%s()\n", __func__);
-
-	if (pageofs > inode->i_size) {
-		ret = jffs2_reserve_space(c, sizeof(ri), &alloc_len,
-					  ALLOC_NORMAL, JFFS2_SUMMARY_INODE_SIZE);
-		if (ret)
-			return ret;
-	}
-
-	mutex_lock(&f->sem);
 	pg = grab_cache_page_write_begin(mapping, index, flags);
-	if (!pg) {
-		if (alloc_len)
-			jffs2_complete_reservation(c);
-		mutex_unlock(&f->sem);
+	if (!pg)
 		return -ENOMEM;
-	}
 	*pagep = pg;
 
-	if (alloc_len) {
+	jffs2_dbg(1, "%s()\n", __func__);
+
+	if (pageofs > inode->i_size) {
 		/* Make new hole frag from old EOF to new page */
+		struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb);
+		struct jffs2_raw_inode ri;
 		struct jffs2_full_dnode *fn;
+		uint32_t alloc_len;
 
 		jffs2_dbg(1, "Writing new hole frag 0x%x-0x%x between current EOF and new page\n",
 			  (unsigned int)inode->i_size, pageofs);
 
+		ret = jffs2_reserve_space(c, sizeof(ri), &alloc_len,
+					  ALLOC_NORMAL, JFFS2_SUMMARY_INODE_SIZE);
+		if (ret)
+			goto out_page;
+
+		mutex_lock(&f->sem);
 		memset(&ri, 0, sizeof(ri));
 
 		ri.magic = cpu_to_je16(JFFS2_MAGIC_BITMASK);
@@ -196,6 +190,7 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 		if (IS_ERR(fn)) {
 			ret = PTR_ERR(fn);
 			jffs2_complete_reservation(c);
+			mutex_unlock(&f->sem);
 			goto out_page;
 		}
 		ret = jffs2_add_full_dnode_to_inode(c, f, fn);
@@ -210,10 +205,12 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 			jffs2_mark_node_obsolete(c, fn->raw);
 			jffs2_free_full_dnode(fn);
 			jffs2_complete_reservation(c);
+			mutex_unlock(&f->sem);
 			goto out_page;
 		}
 		jffs2_complete_reservation(c);
 		inode->i_size = pageofs;
+		mutex_unlock(&f->sem);
 	}
 
 	/*
@@ -222,18 +219,18 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 	 * case of a short-copy.
 	 */
 	if (!PageUptodate(pg)) {
+		mutex_lock(&f->sem);
 		ret = jffs2_do_readpage_nolock(inode, pg);
+		mutex_unlock(&f->sem);
 		if (ret)
 			goto out_page;
 	}
-	mutex_unlock(&f->sem);
 	jffs2_dbg(1, "end write_begin(). pg->flags %lx\n", pg->flags);
 	return ret;
 
 out_page:
 	unlock_page(pg);
 	page_cache_release(pg);
-	mutex_unlock(&f->sem);
 	return ret;
 }
 
diff --git a/fs/jffs2/gc.c b/fs/jffs2/gc.c
index 5a2dec2b064c..95d5880a63ee 100644
--- a/fs/jffs2/gc.c
+++ b/fs/jffs2/gc.c
@@ -1296,14 +1296,17 @@ static int jffs2_garbage_collect_dnode(struct jffs2_sb_info *c, struct jffs2_era
 		BUG_ON(start > orig_start);
 	}
 
-	/* First, use readpage() to read the appropriate page into the page cache */
-	/* Q: What happens if we actually try to GC the _same_ page for which commit_write()
-	 *    triggered garbage collection in the first place?
-	 * A: I _think_ it's OK. read_cache_page shouldn't deadlock, we'll write out the
-	 *    page OK. We'll actually write it out again in commit_write, which is a little
-	 *    suboptimal, but at least we're correct.
-	 */
+	/* The rules state that we must obtain the page lock *before* f->sem, so
+	 * drop f->sem temporarily. Since we also hold c->alloc_sem, nothing's
+	 * actually going to *change* so we're safe; we only allow reading.
+	 *
+	 * It is important to note that jffs2_write_begin() will ensure that its
+	 * page is marked Uptodate before allocating space. That means that if we
+	 * end up here trying to GC the *same* page that jffs2_write_begin() is
+	 * trying to write out, read_cache_page() will not deadlock. */
+	mutex_unlock(&f->sem);
 	pg_ptr = jffs2_gc_fetch_page(c, f, start, &pg);
+	mutex_lock(&f->sem);
 
 	if (IS_ERR(pg_ptr)) {
 		pr_warn("read_cache_page() returned error: %ld\n",
diff --git a/fs/jffs2/nodelist.h b/fs/jffs2/nodelist.h
index fa35ff79ab35..0637271f3770 100644
--- a/fs/jffs2/nodelist.h
+++ b/fs/jffs2/nodelist.h
@@ -194,6 +194,7 @@ struct jffs2_inode_cache {
 #define INO_STATE_CLEARING	6	/* In clear_inode() */
 
 #define INO_FLAGS_XATTR_CHECKED	0x01	/* has no duplicate xattr_ref */
+#define INO_FLAGS_IS_DIR	0x02	/* is a directory */
 
 #define RAWNODE_CLASS_INODE_CACHE	0
 #define RAWNODE_CLASS_XATTR_DATUM	1
@@ -249,7 +250,10 @@ struct jffs2_readinode_info
 
 struct jffs2_full_dirent
 {
-	struct jffs2_raw_node_ref *raw;
+	union {
+		struct jffs2_raw_node_ref *raw;
+		struct jffs2_inode_cache *ic; /* Just during part of build */
+	};
 	struct jffs2_full_dirent *next;
 	uint32_t version;
 	uint32_t ino; /* == zero for unlink */
diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c
index 26c2de2de13f..b7f8eaeea5d8 100644
--- a/fs/ncpfs/dir.c
+++ b/fs/ncpfs/dir.c
@@ -633,7 +633,7 @@ ncp_fill_cache(struct file *file, struct dir_context *ctx,
 				d_rehash(newdent);
 		} else {
 			spin_lock(&dentry->d_lock);
-			NCP_FINFO(inode)->flags &= ~NCPI_DIR_CACHE;
+			NCP_FINFO(dir)->flags &= ~NCPI_DIR_CACHE;
 			spin_unlock(&dentry->d_lock);
 		}
 	} else {
diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c
index 9581d190f6e1..77ebc2bc1cca 100644
--- a/fs/ocfs2/mmap.c
+++ b/fs/ocfs2/mmap.c
@@ -147,6 +147,10 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	ret = ocfs2_inode_lock(inode, &di_bh, 1);
 	if (ret < 0) {
 		mlog_errno(ret);
+		if (ret == -ENOMEM)
+			ret = VM_FAULT_OOM;
+		else
+			ret = VM_FAULT_SIGBUS;
 		goto out;
 	}
 
diff --git a/fs/overlayfs/dir.c b/fs/overlayfs/dir.c
index ed95272d57a6..52f6de5d40a9 100644
--- a/fs/overlayfs/dir.c
+++ b/fs/overlayfs/dir.c
@@ -618,7 +618,8 @@ static int ovl_remove_upper(struct dentry *dentry, bool is_dir)
 	 * sole user of this dentry.  Too tricky...  Just unhash for
 	 * now.
 	 */
-	d_drop(dentry);
+	if (!err)
+		d_drop(dentry);
 	inode_unlock(dir);
 
 	return err;
@@ -903,6 +904,13 @@ static int ovl_rename2(struct inode *olddir, struct dentry *old,
 	if (!overwrite && new_is_dir && !old_opaque && new_opaque)
 		ovl_remove_opaque(newdentry);
 
+	/*
+	 * Old dentry now lives in different location. Dentries in
+	 * lowerstack are stale. We cannot drop them here because
+	 * access to them is lockless. This could be only pure upper
+	 * or opaque directory - numlower is zero. Or upper non-dir
+	 * entry - its pureness is tracked by flag opaque.
+	 */
 	if (old_opaque != new_opaque) {
 		ovl_dentry_set_opaque(old, new_opaque);
 		if (!overwrite)
diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c
index 49e204560655..a4ff5d0d7db9 100644
--- a/fs/overlayfs/inode.c
+++ b/fs/overlayfs/inode.c
@@ -65,6 +65,8 @@ int ovl_setattr(struct dentry *dentry, struct iattr *attr)
 
 		inode_lock(upperdentry->d_inode);
 		err = notify_change(upperdentry, attr, NULL);
+		if (!err)
+			ovl_copyattr(upperdentry->d_inode, dentry->d_inode);
 		inode_unlock(upperdentry->d_inode);
 	}
 	ovl_drop_write(dentry);
diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index 8d826bd56b26..619ad4b016d2 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -76,12 +76,14 @@ enum ovl_path_type ovl_path_type(struct dentry *dentry)
 	if (oe->__upperdentry) {
 		type = __OVL_PATH_UPPER;
 
-		if (oe->numlower) {
-			if (S_ISDIR(dentry->d_inode->i_mode))
-				type |= __OVL_PATH_MERGE;
-		} else if (!oe->opaque) {
+		/*
+		 * Non-dir dentry can hold lower dentry from previous
+		 * location. Its purity depends only on opaque flag.
+		 */
+		if (oe->numlower && S_ISDIR(dentry->d_inode->i_mode))
+			type |= __OVL_PATH_MERGE;
+		else if (!oe->opaque)
 			type |= __OVL_PATH_PURE;
-		}
 	} else {
 		if (oe->numlower > 1)
 			type |= __OVL_PATH_MERGE;
@@ -341,6 +343,7 @@ static const struct dentry_operations ovl_dentry_operations = {
 
 static const struct dentry_operations ovl_reval_dentry_operations = {
 	.d_release = ovl_dentry_release,
+	.d_select_inode = ovl_d_select_inode,
 	.d_revalidate = ovl_dentry_revalidate,
 	.d_weak_revalidate = ovl_dentry_weak_revalidate,
 };
diff --git a/fs/super.c b/fs/super.c
index 1182af8fd5ff..74914b1bae70 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -415,6 +415,7 @@ void generic_shutdown_super(struct super_block *sb)
 		sb->s_flags &= ~MS_ACTIVE;
 
 		fsnotify_unmount_inodes(sb);
+		cgroup_writeback_umount();
 
 		evict_inodes(sb);
 
diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
index 50311703135b..66cdb44616d5 100644
--- a/fs/userfaultfd.c
+++ b/fs/userfaultfd.c
@@ -287,6 +287,12 @@ int handle_userfault(struct vm_area_struct *vma, unsigned long address,
 		goto out;
 
 	/*
+	 * We don't do userfault handling for the final child pid update.
+	 */
+	if (current->flags & PF_EXITING)
+		goto out;
+
+	/*
 	 * Check that we can return VM_FAULT_RETRY.
 	 *
 	 * NOTE: it should become possible to return VM_FAULT_RETRY
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 594f7e63b432..be5568839442 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1109,27 +1109,10 @@ xlog_verify_head(
 	bool			tmp_wrapped;
 
 	/*
-	 * Search backwards through the log looking for the log record header
-	 * block. This wraps all the way back around to the head so something is
-	 * seriously wrong if we can't find it.
-	 */
-	found = xlog_rseek_logrec_hdr(log, *head_blk, *head_blk, 1, bp, rhead_blk,
-				      rhead, wrapped);
-	if (found < 0)
-		return found;
-	if (!found) {
-		xfs_warn(log->l_mp, "%s: couldn't find sync record", __func__);
-		return -EIO;
-	}
-
-	*tail_blk = BLOCK_LSN(be64_to_cpu((*rhead)->h_tail_lsn));
-
-	/*
-	 * Now that we have a tail block, check the head of the log for torn
-	 * writes. Search again until we hit the tail or the maximum number of
-	 * log record I/Os that could have been in flight at one time. Use a
-	 * temporary buffer so we don't trash the rhead/bp pointer from the
-	 * call above.
+	 * Check the head of the log for torn writes. Search backwards from the
+	 * head until we hit the tail or the maximum number of log record I/Os
+	 * that could have been in flight at one time. Use a temporary buffer so
+	 * we don't trash the rhead/bp pointers from the caller.
 	 */
 	tmp_bp = xlog_get_bp(log, 1);
 	if (!tmp_bp)
@@ -1216,6 +1199,115 @@ xlog_verify_head(
 }
 
 /*
+ * Check whether the head of the log points to an unmount record. In other
+ * words, determine whether the log is clean. If so, update the in-core state
+ * appropriately.
+ */
+static int
+xlog_check_unmount_rec(
+	struct xlog		*log,
+	xfs_daddr_t		*head_blk,
+	xfs_daddr_t		*tail_blk,
+	struct xlog_rec_header	*rhead,
+	xfs_daddr_t		rhead_blk,
+	struct xfs_buf		*bp,
+	bool			*clean)
+{
+	struct xlog_op_header	*op_head;
+	xfs_daddr_t		umount_data_blk;
+	xfs_daddr_t		after_umount_blk;
+	int			hblks;
+	int			error;
+	char			*offset;
+
+	*clean = false;
+
+	/*
+	 * Look for unmount record. If we find it, then we know there was a
+	 * clean unmount. Since 'i' could be the last block in the physical
+	 * log, we convert to a log block before comparing to the head_blk.
+	 *
+	 * Save the current tail lsn to use to pass to xlog_clear_stale_blocks()
+	 * below. We won't want to clear the unmount record if there is one, so
+	 * we pass the lsn of the unmount record rather than the block after it.
+	 */
+	if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) {
+		int	h_size = be32_to_cpu(rhead->h_size);
+		int	h_version = be32_to_cpu(rhead->h_version);
+
+		if ((h_version & XLOG_VERSION_2) &&
+		    (h_size > XLOG_HEADER_CYCLE_SIZE)) {
+			hblks = h_size / XLOG_HEADER_CYCLE_SIZE;
+			if (h_size % XLOG_HEADER_CYCLE_SIZE)
+				hblks++;
+		} else {
+			hblks = 1;
+		}
+	} else {
+		hblks = 1;
+	}
+	after_umount_blk = rhead_blk + hblks + BTOBB(be32_to_cpu(rhead->h_len));
+	after_umount_blk = do_mod(after_umount_blk, log->l_logBBsize);
+	if (*head_blk == after_umount_blk &&
+	    be32_to_cpu(rhead->h_num_logops) == 1) {
+		umount_data_blk = rhead_blk + hblks;
+		umount_data_blk = do_mod(umount_data_blk, log->l_logBBsize);
+		error = xlog_bread(log, umount_data_blk, 1, bp, &offset);
+		if (error)
+			return error;
+
+		op_head = (struct xlog_op_header *)offset;
+		if (op_head->oh_flags & XLOG_UNMOUNT_TRANS) {
+			/*
+			 * Set tail and last sync so that newly written log
+			 * records will point recovery to after the current
+			 * unmount record.
+			 */
+			xlog_assign_atomic_lsn(&log->l_tail_lsn,
+					log->l_curr_cycle, after_umount_blk);
+			xlog_assign_atomic_lsn(&log->l_last_sync_lsn,
+					log->l_curr_cycle, after_umount_blk);
+			*tail_blk = after_umount_blk;
+
+			*clean = true;
+		}
+	}
+
+	return 0;
+}
+
+static void
+xlog_set_state(
+	struct xlog		*log,
+	xfs_daddr_t		head_blk,
+	struct xlog_rec_header	*rhead,
+	xfs_daddr_t		rhead_blk,
+	bool			bump_cycle)
+{
+	/*
+	 * Reset log values according to the state of the log when we
+	 * crashed.  In the case where head_blk == 0, we bump curr_cycle
+	 * one because the next write starts a new cycle rather than
+	 * continuing the cycle of the last good log record.  At this
+	 * point we have guaranteed that all partial log records have been
+	 * accounted for.  Therefore, we know that the last good log record
+	 * written was complete and ended exactly on the end boundary
+	 * of the physical log.
+	 */
+	log->l_prev_block = rhead_blk;
+	log->l_curr_block = (int)head_blk;
+	log->l_curr_cycle = be32_to_cpu(rhead->h_cycle);
+	if (bump_cycle)
+		log->l_curr_cycle++;
+	atomic64_set(&log->l_tail_lsn, be64_to_cpu(rhead->h_tail_lsn));
+	atomic64_set(&log->l_last_sync_lsn, be64_to_cpu(rhead->h_lsn));
+	xlog_assign_grant_head(&log->l_reserve_head.grant, log->l_curr_cycle,
+					BBTOB(log->l_curr_block));
+	xlog_assign_grant_head(&log->l_write_head.grant, log->l_curr_cycle,
+					BBTOB(log->l_curr_block));
+}
+
+/*
  * Find the sync block number or the tail of the log.
  *
  * This will be the block number of the last record to have its
@@ -1238,22 +1330,20 @@ xlog_find_tail(
 	xfs_daddr_t		*tail_blk)
 {
 	xlog_rec_header_t	*rhead;
-	xlog_op_header_t	*op_head;
 	char			*offset = NULL;
 	xfs_buf_t		*bp;
 	int			error;
-	xfs_daddr_t		umount_data_blk;
-	xfs_daddr_t		after_umount_blk;
 	xfs_daddr_t		rhead_blk;
 	xfs_lsn_t		tail_lsn;
-	int			hblks;
 	bool			wrapped = false;
+	bool			clean = false;
 
 	/*
 	 * Find previous log record
 	 */
 	if ((error = xlog_find_head(log, head_blk)))
 		return error;
+	ASSERT(*head_blk < INT_MAX);
 
 	bp = xlog_get_bp(log, 1);
 	if (!bp)
@@ -1271,100 +1361,75 @@ xlog_find_tail(
 	}
 
 	/*
-	 * Trim the head block back to skip over torn records. We can have
-	 * multiple log I/Os in flight at any time, so we assume CRC failures
-	 * back through the previous several records are torn writes and skip
-	 * them.
+	 * Search backwards through the log looking for the log record header
+	 * block. This wraps all the way back around to the head so something is
+	 * seriously wrong if we can't find it.
 	 */
-	ASSERT(*head_blk < INT_MAX);
-	error = xlog_verify_head(log, head_blk, tail_blk, bp, &rhead_blk,
-				 &rhead, &wrapped);
-	if (error)
-		goto done;
+	error = xlog_rseek_logrec_hdr(log, *head_blk, *head_blk, 1, bp,
+				      &rhead_blk, &rhead, &wrapped);
+	if (error < 0)
+		return error;
+	if (!error) {
+		xfs_warn(log->l_mp, "%s: couldn't find sync record", __func__);
+		return -EIO;
+	}
+	*tail_blk = BLOCK_LSN(be64_to_cpu(rhead->h_tail_lsn));
 
 	/*
-	 * Reset log values according to the state of the log when we
-	 * crashed.  In the case where head_blk == 0, we bump curr_cycle
-	 * one because the next write starts a new cycle rather than
-	 * continuing the cycle of the last good log record.  At this
-	 * point we have guaranteed that all partial log records have been
-	 * accounted for.  Therefore, we know that the last good log record
-	 * written was complete and ended exactly on the end boundary
-	 * of the physical log.
+	 * Set the log state based on the current head record.
 	 */
-	log->l_prev_block = rhead_blk;
-	log->l_curr_block = (int)*head_blk;
-	log->l_curr_cycle = be32_to_cpu(rhead->h_cycle);
-	if (wrapped)
-		log->l_curr_cycle++;
-	atomic64_set(&log->l_tail_lsn, be64_to_cpu(rhead->h_tail_lsn));
-	atomic64_set(&log->l_last_sync_lsn, be64_to_cpu(rhead->h_lsn));
-	xlog_assign_grant_head(&log->l_reserve_head.grant, log->l_curr_cycle,
-					BBTOB(log->l_curr_block));
-	xlog_assign_grant_head(&log->l_write_head.grant, log->l_curr_cycle,
-					BBTOB(log->l_curr_block));
+	xlog_set_state(log, *head_blk, rhead, rhead_blk, wrapped);
+	tail_lsn = atomic64_read(&log->l_tail_lsn);
 
 	/*
-	 * Look for unmount record.  If we find it, then we know there
-	 * was a clean unmount.  Since 'i' could be the last block in
-	 * the physical log, we convert to a log block before comparing
-	 * to the head_blk.
+	 * Look for an unmount record at the head of the log. This sets the log
+	 * state to determine whether recovery is necessary.
+	 */
+	error = xlog_check_unmount_rec(log, head_blk, tail_blk, rhead,
+				       rhead_blk, bp, &clean);
+	if (error)
+		goto done;
+
+	/*
+	 * Verify the log head if the log is not clean (e.g., we have anything
+	 * but an unmount record at the head). This uses CRC verification to
+	 * detect and trim torn writes. If discovered, CRC failures are
+	 * considered torn writes and the log head is trimmed accordingly.
 	 *
-	 * Save the current tail lsn to use to pass to
-	 * xlog_clear_stale_blocks() below.  We won't want to clear the
-	 * unmount record if there is one, so we pass the lsn of the
-	 * unmount record rather than the block after it.
+	 * Note that we can only run CRC verification when the log is dirty
+	 * because there's no guarantee that the log data behind an unmount
+	 * record is compatible with the current architecture.
 	 */
-	if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) {
-		int	h_size = be32_to_cpu(rhead->h_size);
-		int	h_version = be32_to_cpu(rhead->h_version);
+	if (!clean) {
+		xfs_daddr_t	orig_head = *head_blk;
 
-		if ((h_version & XLOG_VERSION_2) &&
-		    (h_size > XLOG_HEADER_CYCLE_SIZE)) {
-			hblks = h_size / XLOG_HEADER_CYCLE_SIZE;
-			if (h_size % XLOG_HEADER_CYCLE_SIZE)
-				hblks++;
-		} else {
-			hblks = 1;
-		}
-	} else {
-		hblks = 1;
-	}
-	after_umount_blk = rhead_blk + hblks + BTOBB(be32_to_cpu(rhead->h_len));
-	after_umount_blk = do_mod(after_umount_blk, log->l_logBBsize);
-	tail_lsn = atomic64_read(&log->l_tail_lsn);
-	if (*head_blk == after_umount_blk &&
-	    be32_to_cpu(rhead->h_num_logops) == 1) {
-		umount_data_blk = rhead_blk + hblks;
-		umount_data_blk = do_mod(umount_data_blk, log->l_logBBsize);
-		error = xlog_bread(log, umount_data_blk, 1, bp, &offset);
+		error = xlog_verify_head(log, head_blk, tail_blk, bp,
+					 &rhead_blk, &rhead, &wrapped);
 		if (error)
 			goto done;
 
-		op_head = (xlog_op_header_t *)offset;
-		if (op_head->oh_flags & XLOG_UNMOUNT_TRANS) {
-			/*
-			 * Set tail and last sync so that newly written
-			 * log records will point recovery to after the
-			 * current unmount record.
-			 */
-			xlog_assign_atomic_lsn(&log->l_tail_lsn,
-					log->l_curr_cycle, after_umount_blk);
-			xlog_assign_atomic_lsn(&log->l_last_sync_lsn,
-					log->l_curr_cycle, after_umount_blk);
-			*tail_blk = after_umount_blk;
-
-			/*
-			 * Note that the unmount was clean. If the unmount
-			 * was not clean, we need to know this to rebuild the
-			 * superblock counters from the perag headers if we
-			 * have a filesystem using non-persistent counters.
-			 */
-			log->l_mp->m_flags |= XFS_MOUNT_WAS_CLEAN;
+		/* update in-core state again if the head changed */
+		if (*head_blk != orig_head) {
+			xlog_set_state(log, *head_blk, rhead, rhead_blk,
+				       wrapped);
+			tail_lsn = atomic64_read(&log->l_tail_lsn);
+			error = xlog_check_unmount_rec(log, head_blk, tail_blk,
+						       rhead, rhead_blk, bp,
+						       &clean);
+			if (error)
+				goto done;
 		}
 	}
 
 	/*
+	 * Note that the unmount was clean. If the unmount was not clean, we
+	 * need to know this to rebuild the superblock counters from the perag
+	 * headers if we have a filesystem using non-persistent counters.
+	 */
+	if (clean)
+		log->l_mp->m_flags |= XFS_MOUNT_WAS_CLEAN;
+
+	/*
 	 * Make sure that there are no blocks in front of the head
 	 * with the same cycle number as the head.  This can happen
 	 * because we allow multiple outstanding log writes concurrently,