ext4: pass context information to jbd2__journal_start()

So we can better understand what bits of ext4 are responsible for
long-running jbd2 handles, use jbd2__journal_start() so we can pass
context information for logging purposes.

The recommended way for finding the longer-running handles is:

   T=/sys/kernel/debug/tracing
   EVENT=$T/events/jbd2/jbd2_handle_stats
   echo "interval > 5" > $EVENT/filter
   echo 1 > $EVENT/enable

   ./run-my-fs-benchmark

   cat $T/trace > /tmp/problem-handles

This will list handles that were active for longer than 20ms.  Having
longer-running handles is bad, because a commit started at the wrong
time could stall for those 20+ milliseconds, which could delay an
fsync() or an O_SYNC operation.  Here is an example line from the
trace file describing a handle which lived on for 311 jiffies, or over
1.2 seconds:

postmark-2917  [000] ....   196.435786: jbd2_handle_stats: dev 254,32 
   tid 570 type 2 line_no 2541 interval 311 sync 0 requested_blocks 1
   dirtied_blocks 0

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

fs/ext4/acl.c

@@ -324,7 +324,7 @@ ext4_acl_chmod(struct inode *inode)
 	if (error)
 		return error;
 retry:
-	handle = ext4_journal_start(inode,
+	handle = ext4_journal_start(inode, EXT4_HT_XATTR,
 			EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
 	if (IS_ERR(handle)) {
 		error = PTR_ERR(handle);
@@ -422,7 +422,8 @@ ext4_xattr_set_acl(struct dentry *dentry, const char *name, const void *value,
 		acl = NULL;
 
 retry:
-	handle = ext4_journal_start(inode, EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
+	handle = ext4_journal_start(inode, EXT4_HT_XATTR,
+				    EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
 	if (IS_ERR(handle)) {
 		error = PTR_ERR(handle);
 		goto release_and_out;

fs/ext4/ext4_jbd2.c

@@ -38,7 +38,8 @@ static void ext4_put_nojournal(handle_t *handle)
 /*
  * Wrappers for jbd2_journal_start/end.
  */
-handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
+handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line,
+				  int type, int nblocks)
 {
 	journal_t *journal;
 
@@ -59,7 +60,7 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
 		ext4_abort(sb, "Detected aborted journal");
 		return ERR_PTR(-EROFS);
 	}
-	return jbd2_journal_start(journal, nblocks);
+	return jbd2__journal_start(journal, nblocks, GFP_NOFS, type, line);
 }
 
 int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle)

fs/ext4/ext4_jbd2.h

@@ -110,6 +110,22 @@
 #define EXT4_MAXQUOTAS_INIT_BLOCKS(sb) (MAXQUOTAS*EXT4_QUOTA_INIT_BLOCKS(sb))
 #define EXT4_MAXQUOTAS_DEL_BLOCKS(sb) (MAXQUOTAS*EXT4_QUOTA_DEL_BLOCKS(sb))
 
+/*
+ * Ext4 handle operation types -- for logging purposes
+ */
+#define EXT4_HT_MISC             0
+#define EXT4_HT_INODE            1
+#define EXT4_HT_WRITE_PAGE       2
+#define EXT4_HT_MAP_BLOCKS       3
+#define EXT4_HT_DIR              4
+#define EXT4_HT_TRUNCATE         5
+#define EXT4_HT_QUOTA            6
+#define EXT4_HT_RESIZE           7
+#define EXT4_HT_MIGRATE          8
+#define EXT4_HT_MOVE_EXTENTS     9
+#define EXT4_HT_XATTR           10
+#define EXT4_HT_MAX             11
+
 /**
  *   struct ext4_journal_cb_entry - Base structure for callback information.
  *
@@ -234,7 +250,8 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line,
 #define ext4_handle_dirty_super(handle, sb) \
 	__ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb))
 
-handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks);
+handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line,
+				  int type, int nblocks);
 int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle);
 
 #define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096)
@@ -268,9 +285,17 @@ static inline int ext4_handle_has_enough_credits(handle_t *handle, int needed)
 	return 1;
 }
 
-static inline handle_t *ext4_journal_start(struct inode *inode, int nblocks)
+#define ext4_journal_start_sb(sb, type, nblocks)			\
+	__ext4_journal_start_sb((sb), __LINE__, (type), (nblocks))
+
+#define ext4_journal_start(inode, type, nblocks)			\
+	__ext4_journal_start((inode), __LINE__, (type), (nblocks))
+
+static inline handle_t *__ext4_journal_start(struct inode *inode,
+					     unsigned int line, int type,
+					     int nblocks)
 {
-	return ext4_journal_start_sb(inode->i_sb, nblocks);
+	return __ext4_journal_start_sb(inode->i_sb, line, type, nblocks);
 }
 
 #define ext4_journal_stop(handle) \

fs/ext4/extents.c

@@ -2656,7 +2656,7 @@ static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
 	ext_debug("truncate since %u to %u\n", start, end);
 
 	/* probably first extent we're gonna free will be last in block */
-	handle = ext4_journal_start(inode, depth + 1);
+	handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, depth + 1);
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
 
@@ -4287,7 +4287,7 @@ void ext4_ext_truncate(struct inode *inode)
 	 * probably first extent we're gonna free will be last in block
 	 */
 	err = ext4_writepage_trans_blocks(inode);
-	handle = ext4_journal_start(inode, err);
+	handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, err);
 	if (IS_ERR(handle))
 		return;
 
@@ -4454,7 +4454,8 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
 	while (ret >= 0 && ret < max_blocks) {
 		map.m_lblk = map.m_lblk + ret;
 		map.m_len = max_blocks = max_blocks - ret;
-		handle = ext4_journal_start(inode, credits);
+		handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS,
+					    credits);
 		if (IS_ERR(handle)) {
 			ret = PTR_ERR(handle);
 			break;
@@ -4532,7 +4533,7 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
 	while (ret >= 0 && ret < max_blocks) {
 		map.m_lblk += ret;
 		map.m_len = (max_blocks -= ret);
-		handle = ext4_journal_start(inode, credits);
+		handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, credits);
 		if (IS_ERR(handle)) {
 			ret = PTR_ERR(handle);
 			break;
@@ -4710,7 +4711,7 @@ int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length)
 	inode_dio_wait(inode);
 
 	credits = ext4_writepage_trans_blocks(inode);
-	handle = ext4_journal_start(inode, credits);
+	handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits);
 	if (IS_ERR(handle)) {
 		err = PTR_ERR(handle);
 		goto out_dio;

fs/ext4/file.c

@@ -240,7 +240,7 @@ static int ext4_file_open(struct inode * inode, struct file * filp)
 			handle_t *handle;
 			int err;
 
-			handle = ext4_journal_start_sb(sb, 1);
+			handle = ext4_journal_start_sb(sb, EXT4_HT_MISC, 1);
 			if (IS_ERR(handle))
 				return PTR_ERR(handle);
 			err = ext4_journal_get_write_access(handle, sbi->s_sbh);

fs/ext4/ialloc.c

@@ -1137,7 +1137,7 @@ int ext4_init_inode_table(struct super_block *sb, ext4_group_t group,
 	if (gdp->bg_flags & cpu_to_le16(EXT4_BG_INODE_ZEROED))
 		goto out;
 
-	handle = ext4_journal_start_sb(sb, 1);
+	handle = ext4_journal_start_sb(sb, EXT4_HT_MISC, 1);
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		goto out;

fs/ext4/indirect.c

@@ -791,7 +791,7 @@ ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb,
 
 		if (final_size > inode->i_size) {
 			/* Credits for sb + inode write */
-			handle = ext4_journal_start(inode, 2);
+			handle = ext4_journal_start(inode, EXT4_HT_INODE, 2);
 			if (IS_ERR(handle)) {
 				ret = PTR_ERR(handle);
 				goto out;
@@ -851,7 +851,7 @@ ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb,
 		int err;
 
 		/* Credits for sb + inode write */
-		handle = ext4_journal_start(inode, 2);
+		handle = ext4_journal_start(inode, EXT4_HT_INODE, 2);
 		if (IS_ERR(handle)) {
 			/* This is really bad luck. We've written the data
 			 * but cannot extend i_size. Bail out and pretend
@@ -950,7 +950,8 @@ static handle_t *start_transaction(struct inode *inode)
 {
 	handle_t *result;
 
-	result = ext4_journal_start(inode, ext4_blocks_for_truncate(inode));
+	result = ext4_journal_start(inode, EXT4_HT_TRUNCATE,
+				    ext4_blocks_for_truncate(inode));
 	if (!IS_ERR(result))
 		return result;
 

fs/ext4/inline.c

@@ -545,7 +545,7 @@ static int ext4_convert_inline_data_to_extent(struct address_space *mapping,
 		return ret;
 
 retry:
-	handle = ext4_journal_start(inode, needed_blocks);
+	handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks);
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		handle = NULL;
@@ -657,7 +657,7 @@ int ext4_try_to_write_inline_data(struct address_space *mapping,
 	 * The possible write could happen in the inode,
 	 * so try to reserve the space in inode first.
 	 */
-	handle = ext4_journal_start(inode, 1);
+	handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		handle = NULL;
@@ -853,7 +853,7 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
 	if (ret)
 		return ret;
 
-	handle = ext4_journal_start(inode, 1);
+	handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		handle = NULL;
@@ -1770,7 +1770,7 @@ void ext4_inline_data_truncate(struct inode *inode, int *has_inline)
 
 
 	needed_blocks = ext4_writepage_trans_blocks(inode);
-	handle = ext4_journal_start(inode, needed_blocks);
+	handle = ext4_journal_start(inode, EXT4_HT_INODE, needed_blocks);
 	if (IS_ERR(handle))
 		return;
 
@@ -1862,7 +1862,7 @@ int ext4_convert_inline_data(struct inode *inode)
 	if (error)
 		return error;
 
-	handle = ext4_journal_start(inode, needed_blocks);
+	handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks);
 	if (IS_ERR(handle)) {
 		error = PTR_ERR(handle);
 		goto out_free;

fs/ext4/inode.c

@@ -234,7 +234,8 @@ void ext4_evict_inode(struct inode *inode)
 	 * protection against it
 	 */
 	sb_start_intwrite(inode->i_sb);
-	handle = ext4_journal_start(inode, ext4_blocks_for_truncate(inode)+3);
+	handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE,
+				    ext4_blocks_for_truncate(inode)+3);
 	if (IS_ERR(handle)) {
 		ext4_std_error(inode->i_sb, PTR_ERR(handle));
 		/*
@@ -656,7 +657,8 @@ static int _ext4_get_block(struct inode *inode, sector_t iblock,
 		if (map.m_len > DIO_MAX_BLOCKS)
 			map.m_len = DIO_MAX_BLOCKS;
 		dio_credits = ext4_chunk_trans_blocks(inode, map.m_len);
-		handle = ext4_journal_start(inode, dio_credits);
+		handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS,
+					    dio_credits);
 		if (IS_ERR(handle)) {
 			ret = PTR_ERR(handle);
 			return ret;
@@ -881,7 +883,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping,
 	}
 
 retry:
-	handle = ext4_journal_start(inode, needed_blocks);
+	handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks);
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		goto out;
@@ -1881,7 +1883,8 @@ static int __ext4_journalled_writepage(struct page *page,
 	 * references to buffers so we are safe */
 	unlock_page(page);
 
-	handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode));
+	handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE,
+				    ext4_writepage_trans_blocks(inode));
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		goto out;
@@ -2312,7 +2315,8 @@ static int ext4_da_writepages(struct address_space *mapping,
 		needed_blocks = ext4_da_writepages_trans_blocks(inode);
 
 		/* start a new transaction*/
-		handle = ext4_journal_start(inode, needed_blocks);
+		handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE,
+					    needed_blocks);
 		if (IS_ERR(handle)) {
 			ret = PTR_ERR(handle);
 			ext4_msg(inode->i_sb, KERN_CRIT, "%s: jbd2_start: "
@@ -2468,7 +2472,7 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 	 * to journalling the i_disksize update if writes to the end
 	 * of file which has an already mapped buffer.
 	 */
-	handle = ext4_journal_start(inode, 1);
+	handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, 1);
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		goto out;
@@ -4215,8 +4219,9 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 
 		/* (user+group)*(old+new) structure, inode write (sb,
 		 * inode block, ? - but truncate inode update has it) */
-		handle = ext4_journal_start(inode, (EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb)+
-					EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb))+3);
+		handle = ext4_journal_start(inode, EXT4_HT_QUOTA,
+			(EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb) +
+			 EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb)) + 3);
 		if (IS_ERR(handle)) {
 			error = PTR_ERR(handle);
 			goto err_out;
@@ -4251,7 +4256,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	    (attr->ia_size < inode->i_size)) {
 		handle_t *handle;
 
-		handle = ext4_journal_start(inode, 3);
+		handle = ext4_journal_start(inode, EXT4_HT_INODE, 3);
 		if (IS_ERR(handle)) {
 			error = PTR_ERR(handle);
 			goto err_out;
@@ -4271,7 +4276,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 							    attr->ia_size);
 			if (error) {
 				/* Do as much error cleanup as possible */
-				handle = ext4_journal_start(inode, 3);
+				handle = ext4_journal_start(inode,
+							    EXT4_HT_INODE, 3);
 				if (IS_ERR(handle)) {
 					ext4_orphan_del(NULL, inode);
 					goto err_out;
@@ -4612,7 +4618,7 @@ void ext4_dirty_inode(struct inode *inode, int flags)
 {
 	handle_t *handle;
 
-	handle = ext4_journal_start(inode, 2);
+	handle = ext4_journal_start(inode, EXT4_HT_INODE, 2);
 	if (IS_ERR(handle))
 		goto out;
 
@@ -4713,7 +4719,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 
 	/* Finally we can mark the inode as dirty. */
 
-	handle = ext4_journal_start(inode, 1);
+	handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
 
@@ -4791,7 +4797,8 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	else
 		get_block = ext4_get_block;
 retry_alloc:
-	handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode));
+	handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE,
+				    ext4_writepage_trans_blocks(inode));
 	if (IS_ERR(handle)) {
 		ret = VM_FAULT_SIGBUS;
 		goto out;

fs/ext4/ioctl.c

@@ -104,7 +104,7 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		} else if (oldflags & EXT4_EOFBLOCKS_FL)
 			ext4_truncate(inode);
 
-		handle = ext4_journal_start(inode, 1);
+		handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
 		if (IS_ERR(handle)) {
 			err = PTR_ERR(handle);
 			goto flags_out;
@@ -173,7 +173,7 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		}
 
 		mutex_lock(&inode->i_mutex);
-		handle = ext4_journal_start(inode, 1);
+		handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
 		if (IS_ERR(handle)) {
 			err = PTR_ERR(handle);
 			goto unlock_out;

fs/ext4/migrate.c

@@ -456,7 +456,7 @@ int ext4_ext_migrate(struct inode *inode)
 		 */
 		return retval;
 
-	handle = ext4_journal_start(inode,
+	handle = ext4_journal_start(inode, EXT4_HT_MIGRATE,
 					EXT4_DATA_TRANS_BLOCKS(inode->i_sb) +
 					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 +
 					EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb)
@@ -507,7 +507,7 @@ int ext4_ext_migrate(struct inode *inode)
 	ext4_set_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
 	up_read((&EXT4_I(inode)->i_data_sem));
 
-	handle = ext4_journal_start(inode, 1);
+	handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1);
 	if (IS_ERR(handle)) {
 		/*
 		 * It is impossible to update on-disk structures without

fs/ext4/move_extent.c

@@ -923,7 +923,7 @@ move_extent_per_page(struct file *o_filp, struct inode *donor_inode,
 again:
 	*err = 0;
 	jblocks = ext4_writepage_trans_blocks(orig_inode) * 2;
-	handle = ext4_journal_start(orig_inode, jblocks);
+	handle = ext4_journal_start(orig_inode, EXT4_HT_MOVE_EXTENTS, jblocks);
 	if (IS_ERR(handle)) {
 		*err = PTR_ERR(handle);
 		return 0;