Merge branch 'linus' into timers/hpet

10 files changed, 165 insertions, 183 deletions

diff --git a/fs/ocfs2/Kconfig b/fs/ocfs2/Kconfig
new file mode 100644
index 00000000000..701b7a3a872
--- /dev/null
+++ b/fs/ocfs2/Kconfig
@@ -0,0 +1,85 @@
+config OCFS2_FS
+	tristate "OCFS2 file system support"
+	depends on NET && SYSFS
+	select CONFIGFS_FS
+	select JBD2
+	select CRC32
+	select QUOTA
+	select QUOTA_TREE
+	help
+	  OCFS2 is a general purpose extent based shared disk cluster file
+	  system with many similarities to ext3. It supports 64 bit inode
+	  numbers, and has automatically extending metadata groups which may
+	  also make it attractive for non-clustered use.
+
+	  You'll want to install the ocfs2-tools package in order to at least
+	  get "mount.ocfs2".
+
+	  Project web page:    http://oss.oracle.com/projects/ocfs2
+	  Tools web page:      http://oss.oracle.com/projects/ocfs2-tools
+	  OCFS2 mailing lists: http://oss.oracle.com/projects/ocfs2/mailman/
+
+	  For more information on OCFS2, see the file
+	  <file:Documentation/filesystems/ocfs2.txt>.
+
+config OCFS2_FS_O2CB
+	tristate "O2CB Kernelspace Clustering"
+	depends on OCFS2_FS
+	default y
+	help
+	  OCFS2 includes a simple kernelspace clustering package, the OCFS2
+	  Cluster Base.  It only requires a very small userspace component
+	  to configure it. This comes with the standard ocfs2-tools package.
+	  O2CB is limited to maintaining a cluster for OCFS2 file systems.
+	  It cannot manage any other cluster applications.
+
+	  It is always safe to say Y here, as the clustering method is
+	  run-time selectable.
+
+config OCFS2_FS_USERSPACE_CLUSTER
+	tristate "OCFS2 Userspace Clustering"
+	depends on OCFS2_FS && DLM
+	default y
+	help
+	  This option will allow OCFS2 to use userspace clustering services
+	  in conjunction with the DLM in fs/dlm.  If you are using a
+	  userspace cluster manager, say Y here.
+
+	  It is safe to say Y, as the clustering method is run-time
+	  selectable.
+
+config OCFS2_FS_STATS
+	bool "OCFS2 statistics"
+	depends on OCFS2_FS
+	default y
+	help
+	  This option allows some fs statistics to be captured. Enabling
+	  this option may increase the memory consumption.
+
+config OCFS2_DEBUG_MASKLOG
+	bool "OCFS2 logging support"
+	depends on OCFS2_FS
+	default y
+	help
+	  The ocfs2 filesystem has an extensive logging system.  The system
+	  allows selection of events to log via files in /sys/o2cb/logmask/.
+	  This option will enlarge your kernel, but it allows debugging of
+	  ocfs2 filesystem issues.
+
+config OCFS2_DEBUG_FS
+	bool "OCFS2 expensive checks"
+	depends on OCFS2_FS
+	default n
+	help
+	  This option will enable expensive consistency checks. Enable
+	  this option for debugging only as it is likely to decrease
+	  performance of the filesystem.
+
+config OCFS2_FS_POSIX_ACL
+	bool "OCFS2 POSIX Access Control Lists"
+	depends on OCFS2_FS
+	select FS_POSIX_ACL
+	default n
+	help
+	  Posix Access Control Lists (ACLs) support permissions for users and
+	  groups beyond the owner/group/world scheme.
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index d861096c9d8..60fe74035db 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -5390,6 +5390,9 @@ int ocfs2_remove_btree_range(struct inode *inode,
 		goto out;
 	}
 
+	vfs_dq_free_space_nodirty(inode,
+				  ocfs2_clusters_to_bytes(inode->i_sb, len));
+
 	ret = ocfs2_remove_extent(inode, et, cpos, len, handle, meta_ac,
 				  dealloc);
 	if (ret) {
diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c
index b1cc7c381e8..e9d7c2038c0 100644
--- a/fs/ocfs2/dcache.c
+++ b/fs/ocfs2/dcache.c
@@ -38,6 +38,7 @@
 #include "dlmglue.h"
 #include "file.h"
 #include "inode.h"
+#include "super.h"
 
 
 static int ocfs2_dentry_revalidate(struct dentry *dentry,
@@ -294,6 +295,34 @@ out_attach:
 	return ret;
 }
 
+static DEFINE_SPINLOCK(dentry_list_lock);
+
+/* We limit the number of dentry locks to drop in one go. We have
+ * this limit so that we don't starve other users of ocfs2_wq. */
+#define DL_INODE_DROP_COUNT 64
+
+/* Drop inode references from dentry locks */
+void ocfs2_drop_dl_inodes(struct work_struct *work)
+{
+	struct ocfs2_super *osb = container_of(work, struct ocfs2_super,
+					       dentry_lock_work);
+	struct ocfs2_dentry_lock *dl;
+	int drop_count = DL_INODE_DROP_COUNT;
+
+	spin_lock(&dentry_list_lock);
+	while (osb->dentry_lock_list && drop_count--) {
+		dl = osb->dentry_lock_list;
+		osb->dentry_lock_list = dl->dl_next;
+		spin_unlock(&dentry_list_lock);
+		iput(dl->dl_inode);
+		kfree(dl);
+		spin_lock(&dentry_list_lock);
+	}
+	if (osb->dentry_lock_list)
+		queue_work(ocfs2_wq, &osb->dentry_lock_work);
+	spin_unlock(&dentry_list_lock);
+}
+
 /*
  * ocfs2_dentry_iput() and friends.
  *
@@ -318,16 +347,23 @@ out_attach:
 static void ocfs2_drop_dentry_lock(struct ocfs2_super *osb,
 				   struct ocfs2_dentry_lock *dl)
 {
-	iput(dl->dl_inode);
 	ocfs2_simple_drop_lockres(osb, &dl->dl_lockres);
 	ocfs2_lock_res_free(&dl->dl_lockres);
-	kfree(dl);
+
+	/* We leave dropping of inode reference to ocfs2_wq as that can
+	 * possibly lead to inode deletion which gets tricky */
+	spin_lock(&dentry_list_lock);
+	if (!osb->dentry_lock_list)
+		queue_work(ocfs2_wq, &osb->dentry_lock_work);
+	dl->dl_next = osb->dentry_lock_list;
+	osb->dentry_lock_list = dl;
+	spin_unlock(&dentry_list_lock);
 }
 
 void ocfs2_dentry_lock_put(struct ocfs2_super *osb,
 			   struct ocfs2_dentry_lock *dl)
 {
-	int unlock = 0;
+	int unlock;
 
 	BUG_ON(dl->dl_count == 0);
 
diff --git a/fs/ocfs2/dcache.h b/fs/ocfs2/dcache.h
index c091c34d988..d06e16c0664 100644
--- a/fs/ocfs2/dcache.h
+++ b/fs/ocfs2/dcache.h
@@ -29,8 +29,13 @@
 extern struct dentry_operations ocfs2_dentry_ops;
 
 struct ocfs2_dentry_lock {
+	/* Use count of dentry lock */
 	unsigned int		dl_count;
-	u64			dl_parent_blkno;
+	union {
+		/* Linked list of dentry locks to release */
+		struct ocfs2_dentry_lock *dl_next;
+		u64			dl_parent_blkno;
+	};
 
 	/*
 	 * The ocfs2_dentry_lock keeps an inode reference until
@@ -47,6 +52,8 @@ int ocfs2_dentry_attach_lock(struct dentry *dentry, struct inode *inode,
 void ocfs2_dentry_lock_put(struct ocfs2_super *osb,
 			   struct ocfs2_dentry_lock *dl);
 
+void ocfs2_drop_dl_inodes(struct work_struct *work);
+
 struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno,
 				      int skip_unhashed);
 
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index b0c4cadd4c4..206a2370876 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -2860,6 +2860,10 @@ static void ocfs2_unlock_ast(void *opaque, int error)
 	case OCFS2_UNLOCK_CANCEL_CONVERT:
 		mlog(0, "Cancel convert success for %s\n", lockres->l_name);
 		lockres->l_action = OCFS2_AST_INVALID;
+		/* Downconvert thread may have requeued this lock, we
+		 * need to wake it. */
+		if (lockres->l_flags & OCFS2_LOCK_BLOCKED)
+			ocfs2_wake_downconvert_thread(ocfs2_get_lockres_osb(lockres));
 		break;
 	case OCFS2_UNLOCK_DROP_LOCK:
 		lockres->l_level = DLM_LOCK_IV;
diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h
index 3c3532e1307..172850a9a12 100644
--- a/fs/ocfs2/journal.h
+++ b/fs/ocfs2/journal.h
@@ -513,8 +513,10 @@ static inline int ocfs2_jbd2_file_inode(handle_t *handle, struct inode *inode)
 static inline int ocfs2_begin_ordered_truncate(struct inode *inode,
 					       loff_t new_size)
 {
-	return jbd2_journal_begin_ordered_truncate(&OCFS2_I(inode)->ip_jinode,
-						   new_size);
+	return jbd2_journal_begin_ordered_truncate(
+				OCFS2_SB(inode->i_sb)->journal->j_journal,
+				&OCFS2_I(inode)->ip_jinode,
+				new_size);
 }
 
 #endif /* OCFS2_JOURNAL_H */
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index ad5c24a29ed..077384135f4 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -210,6 +210,7 @@ struct ocfs2_journal;
 struct ocfs2_slot_info;
 struct ocfs2_recovery_map;
 struct ocfs2_quota_recovery;
+struct ocfs2_dentry_lock;
 struct ocfs2_super
 {
 	struct task_struct *commit_task;
@@ -325,6 +326,11 @@ struct ocfs2_super
 	struct list_head blocked_lock_list;
 	unsigned long blocked_lock_count;
 
+	/* List of dentry locks to release. Anyone can add locks to
+	 * the list, ocfs2_wq processes the list  */
+	struct ocfs2_dentry_lock *dentry_lock_list;
+	struct work_struct dentry_lock_work;
+
 	wait_queue_head_t		osb_mount_event;
 
 	/* Truncate log info */
diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c
index 6aff8f2d3e4..1ed0f7c8686 100644
--- a/fs/ocfs2/quota_global.c
+++ b/fs/ocfs2/quota_global.c
@@ -754,7 +754,9 @@ static int ocfs2_mark_dquot_dirty(struct dquot *dquot)
 	if (dquot->dq_flags & mask)
 		sync = 1;
 	spin_unlock(&dq_data_lock);
-	if (!sync) {
+	/* This is a slight hack but we can't afford getting global quota
+	 * lock if we already have a transaction started. */
+	if (!sync || journal_current_handle()) {
 		status = ocfs2_write_dquot(dquot);
 		goto out;
 	}
@@ -810,171 +812,6 @@ out:
 	return status;
 }
 
-/* This is difficult. We have to lock quota inode and start transaction
- * in this function but we don't want to take the penalty of exlusive
- * quota file lock when we are just going to use cached structures. So
- * we just take read lock check whether we have dquot cached and if so,
- * we don't have to take the write lock... */
-static int ocfs2_dquot_initialize(struct inode *inode, int type)
-{
-	handle_t *handle = NULL;
-	int status = 0;
-	struct super_block *sb = inode->i_sb;
-	struct ocfs2_mem_dqinfo *oinfo;
-	int exclusive = 0;
-	int cnt;
-	qid_t id;
-
-	mlog_entry_void();
-
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
-		if (type != -1 && cnt != type)
-			continue;
-		if (!sb_has_quota_active(sb, cnt))
-			continue;
-		oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
-		status = ocfs2_lock_global_qf(oinfo, 0);
-		if (status < 0)
-			goto out;
-		/* This is just a performance optimization not a reliable test.
-		 * Since we hold an inode lock, noone can actually release
-		 * the structure until we are finished with initialization. */
-		if (inode->i_dquot[cnt] != NODQUOT) {
-			ocfs2_unlock_global_qf(oinfo, 0);
-			continue;
-		}
-		/* When we have inode lock, we know that no dquot_release() can
-		 * run and thus we can safely check whether we need to
-		 * read+modify global file to get quota information or whether
-		 * our node already has it. */
-		if (cnt == USRQUOTA)
-			id = inode->i_uid;
-		else if (cnt == GRPQUOTA)
-			id = inode->i_gid;
-		else
-			BUG();
-		/* Obtain exclusion from quota off... */
-		down_write(&sb_dqopt(sb)->dqptr_sem);
-		exclusive = !dquot_is_cached(sb, id, cnt);
-		up_write(&sb_dqopt(sb)->dqptr_sem);
-		if (exclusive) {
-			status = ocfs2_lock_global_qf(oinfo, 1);
-			if (status < 0) {
-				exclusive = 0;
-				mlog_errno(status);
-				goto out_ilock;
-			}
-			handle = ocfs2_start_trans(OCFS2_SB(sb),
-					ocfs2_calc_qinit_credits(sb, cnt));
-			if (IS_ERR(handle)) {
-				status = PTR_ERR(handle);
-				mlog_errno(status);
-				goto out_ilock;
-			}
-		}
-		dquot_initialize(inode, cnt);
-		if (exclusive) {
-			ocfs2_commit_trans(OCFS2_SB(sb), handle);
-			ocfs2_unlock_global_qf(oinfo, 1);
-		}
-		ocfs2_unlock_global_qf(oinfo, 0);
-	}
-	mlog_exit(0);
-	return 0;
-out_ilock:
-	if (exclusive)
-		ocfs2_unlock_global_qf(oinfo, 1);
-	ocfs2_unlock_global_qf(oinfo, 0);
-out:
-	mlog_exit(status);
-	return status;
-}
-
-static int ocfs2_dquot_drop_slow(struct inode *inode)
-{
-	int status = 0;
-	int cnt;
-	int got_lock[MAXQUOTAS] = {0, 0};
-	handle_t *handle;
-	struct super_block *sb = inode->i_sb;
-	struct ocfs2_mem_dqinfo *oinfo;
-
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
-		if (!sb_has_quota_active(sb, cnt))
-			continue;
-		oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
-		status = ocfs2_lock_global_qf(oinfo, 1);
-		if (status < 0)
-			goto out;
-		got_lock[cnt] = 1;
-	}
-	handle = ocfs2_start_trans(OCFS2_SB(sb),
-			ocfs2_calc_qinit_credits(sb, USRQUOTA) +
-			ocfs2_calc_qinit_credits(sb, GRPQUOTA));
-	if (IS_ERR(handle)) {
-		status = PTR_ERR(handle);
-		mlog_errno(status);
-		goto out;
-	}
-	dquot_drop(inode);
-	ocfs2_commit_trans(OCFS2_SB(sb), handle);
-out:
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
-		if (got_lock[cnt]) {
-			oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
-			ocfs2_unlock_global_qf(oinfo, 1);
-		}
-	return status;
-}
-
-/* See the comment before ocfs2_dquot_initialize. */
-static int ocfs2_dquot_drop(struct inode *inode)
-{
-	int status = 0;
-	struct super_block *sb = inode->i_sb;
-	struct ocfs2_mem_dqinfo *oinfo;
-	int exclusive = 0;
-	int cnt;
-	int got_lock[MAXQUOTAS] = {0, 0};
-
-	mlog_entry_void();
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
-		if (!sb_has_quota_active(sb, cnt))
-			continue;
-		oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
-		status = ocfs2_lock_global_qf(oinfo, 0);
-		if (status < 0)
-			goto out;
-		got_lock[cnt] = 1;
-	}
-	/* Lock against anyone releasing references so that when when we check
-	 * we know we are not going to be last ones to release dquot */
-	down_write(&sb_dqopt(sb)->dqptr_sem);
-	/* Urgh, this is a terrible hack :( */
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
-		if (inode->i_dquot[cnt] != NODQUOT &&
-		    atomic_read(&inode->i_dquot[cnt]->dq_count) > 1) {
-			exclusive = 1;
-			break;
-		}
-	}
-	if (!exclusive)
-		dquot_drop_locked(inode);
-	up_write(&sb_dqopt(sb)->dqptr_sem);
-out:
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
-		if (got_lock[cnt]) {
-			oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
-			ocfs2_unlock_global_qf(oinfo, 0);
-		}
-	/* In case we bailed out because we had to do expensive locking
-	 * do it now... */
-	if (exclusive)
-		status = ocfs2_dquot_drop_slow(inode);
-	mlog_exit(status);
-	return status;
-}
-
 static struct dquot *ocfs2_alloc_dquot(struct super_block *sb, int type)
 {
 	struct ocfs2_dquot *dquot =
@@ -991,8 +828,8 @@ static void ocfs2_destroy_dquot(struct dquot *dquot)
 }
 
 struct dquot_operations ocfs2_quota_operations = {
-	.initialize	= ocfs2_dquot_initialize,
-	.drop		= ocfs2_dquot_drop,
+	.initialize	= dquot_initialize,
+	.drop		= dquot_drop,
 	.alloc_space	= dquot_alloc_space,
 	.alloc_inode	= dquot_alloc_inode,
 	.free_space	= dquot_free_space,
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 43ed11345b5..b1cb38fbe80 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -1887,6 +1887,9 @@ static int ocfs2_initialize_super(struct super_block *sb,
 	INIT_WORK(&journal->j_recovery_work, ocfs2_complete_recovery);
 	journal->j_state = OCFS2_JOURNAL_FREE;
 
+	INIT_WORK(&osb->dentry_lock_work, ocfs2_drop_dl_inodes);
+	osb->dentry_lock_list = NULL;
+
 	/* get some pseudo constants for clustersize bits */
 	osb->s_clustersize_bits =
 		le32_to_cpu(di->id2.i_super.s_clustersize_bits);
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index e1d638af6ac..915039fffe6 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -4729,13 +4729,6 @@ static int ocfs2_xattr_bucket_value_truncate(struct inode *inode,
 	vb.vb_xv = (struct ocfs2_xattr_value_root *)
 		(vb.vb_bh->b_data + offset % blocksize);
 
-	ret = ocfs2_xattr_bucket_journal_access(ctxt->handle, bucket,
-						OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
-
 	/*
 	 * From here on out we have to dirty the bucket.  The generic
 	 * value calls only modify one of the bucket's bhs, but we need
@@ -4748,12 +4741,18 @@ static int ocfs2_xattr_bucket_value_truncate(struct inode *inode,
 	ret = ocfs2_xattr_value_truncate(inode, &vb, len, ctxt);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_dirty;
+		goto out;
+	}
+
+	ret = ocfs2_xattr_bucket_journal_access(ctxt->handle, bucket,
+						OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
 	}
 
 	xe->xe_value_size = cpu_to_le64(len);
 
-out_dirty:
 	ocfs2_xattr_bucket_journal_dirty(ctxt->handle, bucket);
 
 out: