Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

Pull ext4 updates from Ted Ts'o:
 "The big new feature added this time is supporting online resizing
  using the meta_bg feature.  This allows us to resize file systems
  which are greater than 16TB.  In addition, the speed of online
  resizing has been improved in general.

  We also fix a number of races, some of which could lead to deadlocks,
  in ext4's Asynchronous I/O and online defrag support, thanks to good
  work by Dmitry Monakhov.

  There are also a large number of more minor bug fixes and cleanups
  from a number of other ext4 contributors, quite of few of which have
  submitted fixes for the first time."

* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (69 commits)
  ext4: fix ext4_flush_completed_IO wait semantics
  ext4: fix mtime update in nodelalloc mode
  ext4: fix ext_remove_space for punch_hole case
  ext4: punch_hole should wait for DIO writers
  ext4: serialize truncate with owerwrite DIO workers
  ext4: endless truncate due to nonlocked dio readers
  ext4: serialize unlocked dio reads with truncate
  ext4: serialize dio nonlocked reads with defrag workers
  ext4: completed_io locking cleanup
  ext4: fix unwritten counter leakage
  ext4: give i_aiodio_unwritten a more appropriate name
  ext4: ext4_inode_info diet
  ext4: convert to use leXX_add_cpu()
  ext4: ext4_bread usage audit
  fs: reserve fallocate flag codepoint
  ext4: remove redundant offset check in mext_check_arguments()
  ext4: don't clear orphan list on ro mount with errors
  jbd2: fix assertion failure in commit code due to lacking transaction credits
  ext4: release donor reference when EXT4_IOC_MOVE_EXT ioctl fails
  ext4: enable FITRIM ioctl on bigalloc file system
  ...

Documentation/ABI/testing/sysfs-fs-ext4

@@ -96,3 +96,16 @@ Contact:	"Theodore Ts'o" <tytso@mit.edu>
 Description:
 		The maximum number of megabytes the writeback code will
 		try to write out before move on to another inode.
+
+What:		/sys/fs/ext4/<disk>/extent_max_zeroout_kb
+Date:		August 2012
+Contact:	"Theodore Ts'o" <tytso@mit.edu>
+Description:
+		The maximum number of kilobytes which will be zeroed
+		out in preference to creating a new uninitialized
+		extent when manipulating an inode's extent tree.  Note
+		that using a larger value will increase the
+		variability of time necessary to complete a random
+		write operation (since a 4k random write might turn
+		into a much larger write due to the zeroout
+		operation).

Documentation/filesystems/ext4.txt

@@ -375,6 +375,16 @@ dioread_nolock		locking. If the dioread_nolock option is specified
 			Because of the restrictions this options comprises
 			it is off by default (e.g. dioread_lock).
 
+max_dir_size_kb=n	This limits the size of directories so that any
+			attempt to expand them beyond the specified
+			limit in kilobytes will cause an ENOSPC error.
+			This is useful in memory constrained
+			environments, where a very large directory can
+			cause severe performance problems or even
+			provoke the Out Of Memory killer.  (For example,
+			if there is only 512mb memory available, a 176mb
+			directory may seriously cramp the system's style.)
+
 i_version		Enable 64-bit inode version support. This option is
 			off by default.
 

fs/buffer.c

@@ -2312,12 +2312,6 @@ int __block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf,
 	loff_t size;
 	int ret;
 
-	/*
-	 * Update file times before taking page lock. We may end up failing the
-	 * fault so this update may be superfluous but who really cares...
-	 */
-	file_update_time(vma->vm_file);
-
 	lock_page(page);
 	size = i_size_read(inode);
 	if ((page->mapping != inode->i_mapping) ||
@@ -2355,6 +2349,13 @@ int block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf,
 	struct super_block *sb = vma->vm_file->f_path.dentry->d_inode->i_sb;
 
 	sb_start_pagefault(sb);
+
+	/*
+	 * Update file times before taking page lock. We may end up failing the
+	 * fault so this update may be superfluous but who really cares...
+	 */
+	file_update_time(vma->vm_file);
+
 	ret = __block_page_mkwrite(vma, vmf, get_block);
 	sb_end_pagefault(sb);
 	return block_page_mkwrite_return(ret);

fs/ext4/ext4.h

@@ -186,7 +186,6 @@ struct mpage_da_data {
 #define EXT4_IO_END_ERROR	0x0002
 #define EXT4_IO_END_QUEUED	0x0004
 #define EXT4_IO_END_DIRECT	0x0008
-#define EXT4_IO_END_IN_FSYNC	0x0010
 
 struct ext4_io_page {
 	struct page	*p_page;
@@ -912,9 +911,7 @@ struct ext4_inode_info {
 	struct list_head i_completed_io_list;
 	spinlock_t i_completed_io_lock;
 	atomic_t i_ioend_count;	/* Number of outstanding io_end structs */
-	/* current io_end structure for async DIO write*/
-	ext4_io_end_t *cur_aio_dio;
-	atomic_t i_aiodio_unwritten; /* Nr. of inflight conversions pending */
+	atomic_t i_unwritten; /* Nr. of inflight conversions pending */
 
 	spinlock_t i_block_reservation_lock;
 
@@ -1233,6 +1230,7 @@ struct ext4_sb_info {
 	spinlock_t s_md_lock;
 	unsigned short *s_mb_offsets;
 	unsigned int *s_mb_maxs;
+	unsigned int s_group_info_size;
 
 	/* tunables */
 	unsigned long s_stripe;
@@ -1243,6 +1241,7 @@ struct ext4_sb_info {
 	unsigned int s_mb_order2_reqs;
 	unsigned int s_mb_group_prealloc;
 	unsigned int s_max_writeback_mb_bump;
+	unsigned int s_max_dir_size_kb;
 	/* where last allocation was done - for stream allocation */
 	unsigned long s_mb_last_group;
 	unsigned long s_mb_last_start;
@@ -1270,8 +1269,12 @@ struct ext4_sb_info {
 	unsigned long s_sectors_written_start;
 	u64 s_kbytes_written;
 
+	/* the size of zero-out chunk */
+	unsigned int s_extent_max_zeroout_kb;
+
 	unsigned int s_log_groups_per_flex;
 	struct flex_groups *s_flex_groups;
+	ext4_group_t s_flex_groups_allocated;
 
 	/* workqueue for dio unwritten */
 	struct workqueue_struct *dio_unwritten_wq;
@@ -1328,10 +1331,20 @@ static inline void ext4_set_io_unwritten_flag(struct inode *inode,
 {
 	if (!(io_end->flag & EXT4_IO_END_UNWRITTEN)) {
 		io_end->flag |= EXT4_IO_END_UNWRITTEN;
-		atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten);
+		atomic_inc(&EXT4_I(inode)->i_unwritten);
 	}
 }
 
+static inline ext4_io_end_t *ext4_inode_aio(struct inode *inode)
+{
+	return inode->i_private;
+}
+
+static inline void ext4_inode_aio_set(struct inode *inode, ext4_io_end_t *io)
+{
+	inode->i_private = io;
+}
+
 /*
  * Inode dynamic state flags
  */
@@ -1345,6 +1358,8 @@ enum {
 	EXT4_STATE_DIO_UNWRITTEN,	/* need convert on dio done*/
 	EXT4_STATE_NEWENTRY,		/* File just added to dir */
 	EXT4_STATE_DELALLOC_RESERVED,	/* blks already reserved for delalloc */
+	EXT4_STATE_DIOREAD_LOCK,	/* Disable support for dio read
+					   nolocking */
 };
 
 #define EXT4_INODE_BIT_FNS(name, field, offset)				\
@@ -1932,7 +1947,7 @@ extern void ext4_htree_free_dir_info(struct dir_private_info *p);
 
 /* fsync.c */
 extern int ext4_sync_file(struct file *, loff_t, loff_t, int);
-extern int ext4_flush_completed_IO(struct inode *);
+extern int ext4_flush_unwritten_io(struct inode *);
 
 /* hash.c */
 extern int ext4fs_dirhash(const char *name, int len, struct
@@ -1966,6 +1981,8 @@ extern void ext4_exit_mballoc(void);
 extern void ext4_free_blocks(handle_t *handle, struct inode *inode,
 			     struct buffer_head *bh, ext4_fsblk_t block,
 			     unsigned long count, int flags);
+extern int ext4_mb_alloc_groupinfo(struct super_block *sb,
+				   ext4_group_t ngroups);
 extern int ext4_mb_add_groupinfo(struct super_block *sb,
 		ext4_group_t i, struct ext4_group_desc *desc);
 extern int ext4_group_add_blocks(handle_t *handle, struct super_block *sb,
@@ -2051,6 +2068,8 @@ extern void ext4_superblock_csum_set(struct super_block *sb,
 extern void *ext4_kvmalloc(size_t size, gfp_t flags);
 extern void *ext4_kvzalloc(size_t size, gfp_t flags);
 extern void ext4_kvfree(void *ptr);
+extern int ext4_alloc_flex_bg_array(struct super_block *sb,
+				    ext4_group_t ngroup);
 extern __printf(4, 5)
 void __ext4_error(struct super_block *, const char *, unsigned int,
 		  const char *, ...);
@@ -2352,6 +2371,7 @@ extern const struct file_operations ext4_dir_operations;
 extern const struct inode_operations ext4_file_inode_operations;
 extern const struct file_operations ext4_file_operations;
 extern loff_t ext4_llseek(struct file *file, loff_t offset, int origin);
+extern void ext4_unwritten_wait(struct inode *inode);
 
 /* namei.c */
 extern const struct inode_operations ext4_dir_inode_operations;
@@ -2400,11 +2420,11 @@ extern int ext4_move_extents(struct file *o_filp, struct file *d_filp,
 
 /* page-io.c */
 extern int __init ext4_init_pageio(void);
+extern void ext4_add_complete_io(ext4_io_end_t *io_end);
 extern void ext4_exit_pageio(void);
 extern void ext4_ioend_wait(struct inode *);
 extern void ext4_free_io_end(ext4_io_end_t *io);
 extern ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags);
-extern int ext4_end_io_nolock(ext4_io_end_t *io);
 extern void ext4_io_submit(struct ext4_io_submit *io);
 extern int ext4_bio_write_page(struct ext4_io_submit *io,
 			       struct page *page,
@@ -2452,6 +2472,21 @@ static inline void set_bitmap_uptodate(struct buffer_head *bh)
 	set_bit(BH_BITMAP_UPTODATE, &(bh)->b_state);
 }
 
+/*
+ * Disable DIO read nolock optimization, so new dioreaders will be forced
+ * to grab i_mutex
+ */
+static inline void ext4_inode_block_unlocked_dio(struct inode *inode)
+{
+	ext4_set_inode_state(inode, EXT4_STATE_DIOREAD_LOCK);
+	smp_mb();
+}
+static inline void ext4_inode_resume_unlocked_dio(struct inode *inode)
+{
+	smp_mb();
+	ext4_clear_inode_state(inode, EXT4_STATE_DIOREAD_LOCK);
+}
+
 #define in_range(b, first, len)	((b) >= (first) && (b) <= (first) + (len) - 1)
 
 /* For ioend & aio unwritten conversion wait queues */

fs/ext4/file.c

@@ -55,11 +55,11 @@ static int ext4_release_file(struct inode *inode, struct file *filp)
 	return 0;
 }
 
-static void ext4_aiodio_wait(struct inode *inode)
+void ext4_unwritten_wait(struct inode *inode)
 {
 	wait_queue_head_t *wq = ext4_ioend_wq(inode);
 
-	wait_event(*wq, (atomic_read(&EXT4_I(inode)->i_aiodio_unwritten) == 0));
+	wait_event(*wq, (atomic_read(&EXT4_I(inode)->i_unwritten) == 0));
 }
 
 /*
@@ -116,7 +116,7 @@ ext4_file_dio_write(struct kiocb *iocb, const struct iovec *iov,
 				 "performance will be poor.",
 				 inode->i_ino, current->comm);
 		mutex_lock(ext4_aio_mutex(inode));
-		ext4_aiodio_wait(inode);
+		ext4_unwritten_wait(inode);
 	}
 
 	BUG_ON(iocb->ki_pos != pos);

fs/ext4/fsync.c

@@ -34,87 +34,6 @@
 
 #include <trace/events/ext4.h>
 
-static void dump_completed_IO(struct inode * inode)
-{
-#ifdef	EXT4FS_DEBUG
-	struct list_head *cur, *before, *after;
-	ext4_io_end_t *io, *io0, *io1;
-	unsigned long flags;
-
-	if (list_empty(&EXT4_I(inode)->i_completed_io_list)){
-		ext4_debug("inode %lu completed_io list is empty\n", inode->i_ino);
-		return;
-	}
-
-	ext4_debug("Dump inode %lu completed_io list \n", inode->i_ino);
-	spin_lock_irqsave(&EXT4_I(inode)->i_completed_io_lock, flags);
-	list_for_each_entry(io, &EXT4_I(inode)->i_completed_io_list, list){
-		cur = &io->list;
-		before = cur->prev;
-		io0 = container_of(before, ext4_io_end_t, list);
-		after = cur->next;
-		io1 = container_of(after, ext4_io_end_t, list);
-
-		ext4_debug("io 0x%p from inode %lu,prev 0x%p,next 0x%p\n",
-			    io, inode->i_ino, io0, io1);
-	}
-	spin_unlock_irqrestore(&EXT4_I(inode)->i_completed_io_lock, flags);
-#endif
-}
-
-/*
- * This function is called from ext4_sync_file().
- *
- * When IO is completed, the work to convert unwritten extents to
- * written is queued on workqueue but may not get immediately
- * scheduled. When fsync is called, we need to ensure the
- * conversion is complete before fsync returns.
- * The inode keeps track of a list of pending/completed IO that
- * might needs to do the conversion. This function walks through
- * the list and convert the related unwritten extents for completed IO
- * to written.
- * The function return the number of pending IOs on success.
- */
-int ext4_flush_completed_IO(struct inode *inode)
-{
-	ext4_io_end_t *io;
-	struct ext4_inode_info *ei = EXT4_I(inode);
-	unsigned long flags;
-	int ret = 0;
-	int ret2 = 0;
-
-	dump_completed_IO(inode);
-	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
-	while (!list_empty(&ei->i_completed_io_list)){
-		io = list_entry(ei->i_completed_io_list.next,
-				ext4_io_end_t, list);
-		list_del_init(&io->list);
-		io->flag |= EXT4_IO_END_IN_FSYNC;
-		/*
-		 * Calling ext4_end_io_nolock() to convert completed
-		 * IO to written.
-		 *
-		 * When ext4_sync_file() is called, run_queue() may already
-		 * about to flush the work corresponding to this io structure.
-		 * It will be upset if it founds the io structure related
-		 * to the work-to-be schedule is freed.
-		 *
-		 * Thus we need to keep the io structure still valid here after
-		 * conversion finished. The io structure has a flag to
-		 * avoid double converting from both fsync and background work
-		 * queue work.
-		 */
-		spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
-		ret = ext4_end_io_nolock(io);
-		if (ret < 0)
-			ret2 = ret;
-		spin_lock_irqsave(&ei->i_completed_io_lock, flags);
-		io->flag &= ~EXT4_IO_END_IN_FSYNC;
-	}
-	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
-	return (ret2 < 0) ? ret2 : 0;
-}
-
 /*
  * If we're not journaling and this is a just-created file, we have to
  * sync our parent directory (if it was freshly created) since
@@ -203,7 +122,7 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	struct inode *inode = file->f_mapping->host;
 	struct ext4_inode_info *ei = EXT4_I(inode);
 	journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
-	int ret;
+	int ret, err;
 	tid_t commit_tid;
 	bool needs_barrier = false;
 
@@ -219,7 +138,7 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	if (inode->i_sb->s_flags & MS_RDONLY)
 		goto out;
 
-	ret = ext4_flush_completed_IO(inode);
+	ret = ext4_flush_unwritten_io(inode);
 	if (ret < 0)
 		goto out;
 
@@ -255,8 +174,11 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 		needs_barrier = true;
 	jbd2_log_start_commit(journal, commit_tid);
 	ret = jbd2_log_wait_commit(journal, commit_tid);
-	if (needs_barrier)
-		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
+	if (needs_barrier) {
+		err = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
+		if (!ret)
+			ret = err;
+	}
  out:
 	mutex_unlock(&inode->i_mutex);
 	trace_ext4_sync_file_exit(inode, ret);

fs/ext4/ialloc.c

@@ -697,6 +697,15 @@ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, umode_t mode,
 		if (!gdp)
 			goto fail;
 
+		/*
+		 * Check free inodes count before loading bitmap.
+		 */
+		if (ext4_free_inodes_count(sb, gdp) == 0) {
+			if (++group == ngroups)
+				group = 0;
+			continue;
+		}
+
 		brelse(inode_bitmap_bh);
 		inode_bitmap_bh = ext4_read_inode_bitmap(sb, group);
 		if (!inode_bitmap_bh)

fs/ext4/indirect.c

@@ -807,16 +807,30 @@ ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb,
 
 retry:
 	if (rw == READ && ext4_should_dioread_nolock(inode)) {
-		if (unlikely(!list_empty(&ei->i_completed_io_list))) {
+		if (unlikely(atomic_read(&EXT4_I(inode)->i_unwritten))) {
 			mutex_lock(&inode->i_mutex);
-			ext4_flush_completed_IO(inode);
+			ext4_flush_unwritten_io(inode);
 			mutex_unlock(&inode->i_mutex);
 		}
+		/*
+		 * Nolock dioread optimization may be dynamically disabled
+		 * via ext4_inode_block_unlocked_dio(). Check inode's state
+		 * while holding extra i_dio_count ref.
+		 */
+		atomic_inc(&inode->i_dio_count);
+		smp_mb();
+		if (unlikely(ext4_test_inode_state(inode,
+						    EXT4_STATE_DIOREAD_LOCK))) {
+			inode_dio_done(inode);
+			goto locked;
+		}
 		ret = __blockdev_direct_IO(rw, iocb, inode,
 				 inode->i_sb->s_bdev, iov,
 				 offset, nr_segs,
 				 ext4_get_block, NULL, NULL, 0);
+		inode_dio_done(inode);
 	} else {
+locked:
 		ret = blockdev_direct_IO(rw, iocb, inode, iov,
 				 offset, nr_segs, ext4_get_block);
 

fs/ext4/inode.c

@@ -732,11 +732,13 @@ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode,
 	err = ext4_map_blocks(handle, inode, &map,
 			      create ? EXT4_GET_BLOCKS_CREATE : 0);
 
+	/* ensure we send some value back into *errp */
+	*errp = 0;
+
 	if (err < 0)
 		*errp = err;
 	if (err <= 0)
 		return NULL;
-	*errp = 0;
 
 	bh = sb_getblk(inode->i_sb, map.m_pblk);
 	if (!bh) {
@@ -1954,9 +1956,6 @@ static int __ext4_journalled_writepage(struct page *page,
 	return ret;
 }
 
-static int ext4_set_bh_endio(struct buffer_head *bh, struct inode *inode);
-static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate);
-
 /*
  * Note that we don't need to start a transaction unless we're journaling data
  * because we should have holes filled from ext4_page_mkwrite(). We even don't
@@ -2463,6 +2462,16 @@ static int ext4_nonda_switch(struct super_block *sb)
 	free_blocks  = EXT4_C2B(sbi,
 		percpu_counter_read_positive(&sbi->s_freeclusters_counter));
 	dirty_blocks = percpu_counter_read_positive(&sbi->s_dirtyclusters_counter);
+	/*
+	 * Start pushing delalloc when 1/2 of free blocks are dirty.
+	 */
+	if (dirty_blocks && (free_blocks < 2 * dirty_blocks) &&
+	    !writeback_in_progress(sb->s_bdi) &&
+	    down_read_trylock(&sb->s_umount)) {
+		writeback_inodes_sb(sb, WB_REASON_FS_FREE_SPACE);
+		up_read(&sb->s_umount);
+	}
+
 	if (2 * free_blocks < 3 * dirty_blocks ||
 		free_blocks < (dirty_blocks + EXT4_FREECLUSTERS_WATERMARK)) {
 		/*
@@ -2471,13 +2480,6 @@ static int ext4_nonda_switch(struct super_block *sb)
 		 */
 		return 1;
 	}
-	/*
-	 * Even if we don't switch but are nearing capacity,
-	 * start pushing delalloc when 1/2 of free blocks are dirty.
-	 */
-	if (free_blocks < 2 * dirty_blocks)
-		writeback_inodes_sb_if_idle(sb, WB_REASON_FS_FREE_SPACE);
-
 	return 0;
 }
 
@@ -2879,9 +2881,6 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 {
 	struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
         ext4_io_end_t *io_end = iocb->private;
-	struct workqueue_struct *wq;
-	unsigned long flags;
-	struct ext4_inode_info *ei;
 
 	/* if not async direct IO or dio with 0 bytes write, just return */
 	if (!io_end || !size)
@@ -2910,24 +2909,14 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 		io_end->iocb = iocb;
 		io_end->result = ret;
 	}
-	wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq;
 
-	/* Add the io_end to per-inode completed aio dio list*/
-	ei = EXT4_I(io_end->inode);
-	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
-	list_add_tail(&io_end->list, &ei->i_completed_io_list);
-	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
-
-	/* queue the work to convert unwritten extents to written */
-	queue_work(wq, &io_end->work);
+	ext4_add_complete_io(io_end);
 }
 
 static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
 {
 	ext4_io_end_t *io_end = bh->b_private;
-	struct workqueue_struct *wq;
 	struct inode *inode;
-	unsigned long flags;
 
 	if (!test_clear_buffer_uninit(bh) || !io_end)
 		goto out;
@@ -2946,15 +2935,7 @@ static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
 	 */
 	inode = io_end->inode;
 	ext4_set_io_unwritten_flag(inode, io_end);
-
-	/* Add the io_end to per-inode completed io list*/
-	spin_lock_irqsave(&EXT4_I(inode)->i_completed_io_lock, flags);
-	list_add_tail(&io_end->list, &EXT4_I(inode)->i_completed_io_list);
-	spin_unlock_irqrestore(&EXT4_I(inode)->i_completed_io_lock, flags);
-
-	wq = EXT4_SB(inode->i_sb)->dio_unwritten_wq;
-	/* queue the work to convert unwritten extents to written */
-	queue_work(wq, &io_end->work);
+	ext4_add_complete_io(io_end);
 out:
 	bh->b_private = NULL;
 	bh->b_end_io = NULL;
@@ -3029,6 +3010,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 		overwrite = *((int *)iocb->private);
 
 		if (overwrite) {
+			atomic_inc(&inode->i_dio_count);
 			down_read(&EXT4_I(inode)->i_data_sem);
 			mutex_unlock(&inode->i_mutex);
 		}
@@ -3054,7 +3036,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 		 * hook to the iocb.
  		 */
 		iocb->private = NULL;
-		EXT4_I(inode)->cur_aio_dio = NULL;
+		ext4_inode_aio_set(inode, NULL);
 		if (!is_sync_kiocb(iocb)) {
 			ext4_io_end_t *io_end =
 				ext4_init_io_end(inode, GFP_NOFS);
@@ -3071,7 +3053,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 			 * is a unwritten extents needs to be converted
 			 * when IO is completed.
 			 */
-			EXT4_I(inode)->cur_aio_dio = iocb->private;
+			ext4_inode_aio_set(inode, io_end);
 		}
 
 		if (overwrite)
@@ -3091,7 +3073,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 						 NULL,
 						 DIO_LOCKING);
 		if (iocb->private)
-			EXT4_I(inode)->cur_aio_dio = NULL;
+			ext4_inode_aio_set(inode, NULL);
 		/*
 		 * The io_end structure takes a reference to the inode,
 		 * that structure needs to be destroyed and the
@@ -3126,6 +3108,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 	retake_lock:
 		/* take i_mutex locking again if we do a ovewrite dio */
 		if (overwrite) {
+			inode_dio_done(inode);
 			up_read(&EXT4_I(inode)->i_data_sem);
 			mutex_lock(&inode->i_mutex);
 		}
@@ -4052,6 +4035,7 @@ static int ext4_do_update_inode(handle_t *handle,
 	struct ext4_inode_info *ei = EXT4_I(inode);
 	struct buffer_head *bh = iloc->bh;
 	int err = 0, rc, block;
+	int need_datasync = 0;
 	uid_t i_uid;
 	gid_t i_gid;
 
@@ -4102,7 +4086,10 @@ static int ext4_do_update_inode(handle_t *handle,
 		raw_inode->i_file_acl_high =
 			cpu_to_le16(ei->i_file_acl >> 32);
 	raw_inode->i_file_acl_lo = cpu_to_le32(ei->i_file_acl);
-	ext4_isize_set(raw_inode, ei->i_disksize);
+	if (ei->i_disksize != ext4_isize(raw_inode)) {
+		ext4_isize_set(raw_inode, ei->i_disksize);
+		need_datasync = 1;
+	}
 	if (ei->i_disksize > 0x7fffffffULL) {
 		struct super_block *sb = inode->i_sb;
 		if (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
@@ -4155,7 +4142,7 @@ static int ext4_do_update_inode(handle_t *handle,
 		err = rc;
 	ext4_clear_inode_state(inode, EXT4_STATE_NEW);
 
-	ext4_update_inode_fsync_trans(handle, inode, 0);
+	ext4_update_inode_fsync_trans(handle, inode, need_datasync);
 out_brelse:
 	brelse(bh);
 	ext4_std_error(inode->i_sb, err);
@@ -4298,7 +4285,6 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
-		inode_dio_wait(inode);
 
 		if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 			struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
@@ -4347,8 +4333,17 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
-		if (attr->ia_size != i_size_read(inode))
+		if (attr->ia_size != i_size_read(inode)) {
 			truncate_setsize(inode, attr->ia_size);
+			/* Inode size will be reduced, wait for dio in flight.
+			 * Temporarily disable dioread_nolock to prevent
+			 * livelock. */
+			if (orphan) {
+				ext4_inode_block_unlocked_dio(inode);
+				inode_dio_wait(inode);
+				ext4_inode_resume_unlocked_dio(inode);
+			}
+		}
 		ext4_truncate(inode);
 	}
 
@@ -4727,6 +4722,10 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 			return err;
 	}
 
+	/* Wait for all existing dio workers */
+	ext4_inode_block_unlocked_dio(inode);
+	inode_dio_wait(inode);
+
 	jbd2_journal_lock_updates(journal);
 
 	/*
@@ -4746,6 +4745,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 	ext4_set_aops(inode);
 
 	jbd2_journal_unlock_updates(journal);
+	ext4_inode_resume_unlocked_dio(inode);
 
 	/* Finally we can mark the inode as dirty. */
 
@@ -4780,6 +4780,7 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	int retries = 0;
 
 	sb_start_pagefault(inode->i_sb);
+	file_update_time(vma->vm_file);
 	/* Delalloc case is easy... */
 	if (test_opt(inode->i_sb, DELALLOC) &&
 	    !ext4_should_journal_data(inode) &&

fs/ext4/ioctl.c

@@ -366,26 +366,11 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 			return -EOPNOTSUPP;
 		}
 
-		if (EXT4_HAS_INCOMPAT_FEATURE(sb,
-			       EXT4_FEATURE_INCOMPAT_META_BG)) {
-			ext4_msg(sb, KERN_ERR,
-				 "Online resizing not (yet) supported with meta_bg");
-			return -EOPNOTSUPP;
-		}
-
 		if (copy_from_user(&n_blocks_count, (__u64 __user *)arg,
 				   sizeof(__u64))) {
 			return -EFAULT;
 		}
 
-		if (n_blocks_count > MAX_32_NUM &&
-		    !EXT4_HAS_INCOMPAT_FEATURE(sb,
-					       EXT4_FEATURE_INCOMPAT_64BIT)) {
-			ext4_msg(sb, KERN_ERR,
-				 "File system only supports 32-bit block numbers");
-			return -EOPNOTSUPP;
-		}
-
 		err = ext4_resize_begin(sb);
 		if (err)
 			return err;
@@ -420,13 +405,6 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		if (!blk_queue_discard(q))
 			return -EOPNOTSUPP;
 
-		if (EXT4_HAS_RO_COMPAT_FEATURE(sb,
-			       EXT4_FEATURE_RO_COMPAT_BIGALLOC)) {
-			ext4_msg(sb, KERN_ERR,
-				 "FITRIM not supported with bigalloc");
-			return -EOPNOTSUPP;
-		}
-
 		if (copy_from_user(&range, (struct fstrim_range __user *)arg,
 		    sizeof(range)))
 			return -EFAULT;

fs/ext4/mballoc.c

@@ -24,6 +24,7 @@
 #include "ext4_jbd2.h"
 #include "mballoc.h"
 #include <linux/debugfs.h>
+#include <linux/log2.h>
 #include <linux/slab.h>
 #include <trace/events/ext4.h>
 
@@ -1338,17 +1339,17 @@ static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
 	mb_check_buddy(e4b);
 }
 
-static int mb_find_extent(struct ext4_buddy *e4b, int order, int block,
+static int mb_find_extent(struct ext4_buddy *e4b, int block,
 				int needed, struct ext4_free_extent *ex)
 {
 	int next = block;
-	int max;
+	int max, order;
 	void *buddy;
 
 	assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group));
 	BUG_ON(ex == NULL);
 
-	buddy = mb_find_buddy(e4b, order, &max);
+	buddy = mb_find_buddy(e4b, 0, &max);
 	BUG_ON(buddy == NULL);
 	BUG_ON(block >= max);
 	if (mb_test_bit(block, buddy)) {
@@ -1358,12 +1359,9 @@ static int mb_find_extent(struct ext4_buddy *e4b, int order, int block,
 		return 0;
 	}
 
-	/* FIXME dorp order completely ? */
-	if (likely(order == 0)) {
-		/* find actual order */
-		order = mb_find_order_for_block(e4b, block);
-		block = block >> order;
-	}
+	/* find actual order */
+	order = mb_find_order_for_block(e4b, block);
+	block = block >> order;
 
 	ex->fe_len = 1 << order;
 	ex->fe_start = block << order;
@@ -1549,7 +1547,7 @@ static void ext4_mb_check_limits(struct ext4_allocation_context *ac,
 		/* recheck chunk's availability - we don't know
 		 * when it was found (within this lock-unlock
 		 * period or not) */
-		max = mb_find_extent(e4b, 0, bex->fe_start, gex->fe_len, &ex);
+		max = mb_find_extent(e4b, bex->fe_start, gex->fe_len, &ex);
 		if (max >= gex->fe_len) {
 			ext4_mb_use_best_found(ac, e4b);
 			return;
@@ -1641,7 +1639,7 @@ int ext4_mb_try_best_found(struct ext4_allocation_context *ac,
 		return err;
 
 	ext4_lock_group(ac->ac_sb, group);
-	max = mb_find_extent(e4b, 0, ex.fe_start, ex.fe_len, &ex);
+	max = mb_find_extent(e4b, ex.fe_start, ex.fe_len, &ex);
 
 	if (max > 0) {
 		ac->ac_b_ex = ex;
@@ -1662,17 +1660,20 @@ int ext4_mb_find_by_goal(struct ext4_allocation_context *ac,
 	int max;
 	int err;
 	struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb);
+	struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group);
 	struct ext4_free_extent ex;
 
 	if (!(ac->ac_flags & EXT4_MB_HINT_TRY_GOAL))
 		return 0;
+	if (grp->bb_free == 0)
+		return 0;
 
 	err = ext4_mb_load_buddy(ac->ac_sb, group, e4b);
 	if (err)
 		return err;
 
 	ext4_lock_group(ac->ac_sb, group);
-	max = mb_find_extent(e4b, 0, ac->ac_g_ex.fe_start,
+	max = mb_find_extent(e4b, ac->ac_g_ex.fe_start,
 			     ac->ac_g_ex.fe_len, &ex);
 
 	if (max >= ac->ac_g_ex.fe_len && ac->ac_g_ex.fe_len == sbi->s_stripe) {
@@ -1788,7 +1789,7 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac,
 			break;
 		}
 
-		mb_find_extent(e4b, 0, i, ac->ac_g_ex.fe_len, &ex);
+		mb_find_extent(e4b, i, ac->ac_g_ex.fe_len, &ex);
 		BUG_ON(ex.fe_len <= 0);
 		if (free < ex.fe_len) {
 			ext4_grp_locked_error(sb, e4b->bd_group, 0, 0,
@@ -1840,7 +1841,7 @@ void ext4_mb_scan_aligned(struct ext4_allocation_context *ac,
 
 	while (i < EXT4_CLUSTERS_PER_GROUP(sb)) {
 		if (!mb_test_bit(i, bitmap)) {
-			max = mb_find_extent(e4b, 0, i, sbi->s_stripe, &ex);
+			max = mb_find_extent(e4b, i, sbi->s_stripe, &ex);
 			if (max >= sbi->s_stripe) {
 				ac->ac_found++;
 				ac->ac_b_ex = ex;
@@ -1862,6 +1863,12 @@ static int ext4_mb_good_group(struct ext4_allocation_context *ac,
 
 	BUG_ON(cr < 0 || cr >= 4);
 
+	free = grp->bb_free;
+	if (free == 0)
+		return 0;
+	if (cr <= 2 && free < ac->ac_g_ex.fe_len)
+		return 0;
+
 	/* We only do this if the grp has never been initialized */
 	if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) {
 		int ret = ext4_mb_init_group(ac->ac_sb, group);
@@ -1869,10 +1876,7 @@ static int ext4_mb_good_group(struct ext4_allocation_context *ac,
 			return 0;
 	}
 
-	free = grp->bb_free;
 	fragments = grp->bb_fragments;
-	if (free == 0)
-		return 0;
 	if (fragments == 0)
 		return 0;
 
@@ -2163,6 +2167,39 @@ static struct kmem_cache *get_groupinfo_cache(int blocksize_bits)
 	return cachep;
 }
 
+/*
+ * Allocate the top-level s_group_info array for the specified number
+ * of groups
+ */
+int ext4_mb_alloc_groupinfo(struct super_block *sb, ext4_group_t ngroups)
+{
+	struct ext4_sb_info *sbi = EXT4_SB(sb);
+	unsigned size;
+	struct ext4_group_info ***new_groupinfo;
+
+	size = (ngroups + EXT4_DESC_PER_BLOCK(sb) - 1) >>
+		EXT4_DESC_PER_BLOCK_BITS(sb);
+	if (size <= sbi->s_group_info_size)
+		return 0;
+
+	size = roundup_pow_of_two(sizeof(*sbi->s_group_info) * size);
+	new_groupinfo = ext4_kvzalloc(size, GFP_KERNEL);
+	if (!new_groupinfo) {
+		ext4_msg(sb, KERN_ERR, "can't allocate buddy meta group");
+		return -ENOMEM;
+	}
+	if (sbi->s_group_info) {
+		memcpy(new_groupinfo, sbi->s_group_info,
+		       sbi->s_group_info_size * sizeof(*sbi->s_group_info));
+		ext4_kvfree(sbi->s_group_info);
+	}
+	sbi->s_group_info = new_groupinfo;
+	sbi->s_group_info_size = size / sizeof(*sbi->s_group_info);
+	ext4_debug("allocated s_groupinfo array for %d meta_bg's\n", 
+		   sbi->s_group_info_size);
+	return 0;
+}
+
 /* Create and initialize ext4_group_info data for the given group. */
 int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
 			  struct ext4_group_desc *desc)
@@ -2195,12 +2232,11 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
 		sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)];
 	i = group & (EXT4_DESC_PER_BLOCK(sb) - 1);
 
-	meta_group_info[i] = kmem_cache_alloc(cachep, GFP_KERNEL);
+	meta_group_info[i] = kmem_cache_zalloc(cachep, GFP_KERNEL);
 	if (meta_group_info[i] == NULL) {
 		ext4_msg(sb, KERN_ERR, "can't allocate buddy mem");
 		goto exit_group_info;
 	}
-	memset(meta_group_info[i], 0, kmem_cache_size(cachep));
 	set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT,
 		&(meta_group_info[i]->bb_state));
 
@@ -2252,49 +2288,14 @@ static int ext4_mb_init_backend(struct super_block *sb)
 	ext4_group_t ngroups = ext4_get_groups_count(sb);
 	ext4_group_t i;
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
-	struct ext4_super_block *es = sbi->s_es;
-	int num_meta_group_infos;
-	int num_meta_group_infos_max;
-	int array_size;
+	int err;
 	struct ext4_group_desc *desc;
 	struct kmem_cache *cachep;
 
-	/* This is the number of blocks used by GDT */
-	num_meta_group_infos = (ngroups + EXT4_DESC_PER_BLOCK(sb) -
-				1) >> EXT4_DESC_PER_BLOCK_BITS(sb);
-
-	/*
-	 * This is the total number of blocks used by GDT including
-	 * the number of reserved blocks for GDT.
-	 * The s_group_info array is allocated with this value
-	 * to allow a clean online resize without a complex
-	 * manipulation of pointer.
-	 * The drawback is the unused memory when no resize
-	 * occurs but it's very low in terms of pages
-	 * (see comments below)
-	 * Need to handle this properly when META_BG resizing is allowed
-	 */
-	num_meta_group_infos_max = num_meta_group_infos +
-				le16_to_cpu(es->s_reserved_gdt_blocks);
+	err = ext4_mb_alloc_groupinfo(sb, ngroups);
+	if (err)
+		return err;
 
-	/*
-	 * array_size is the size of s_group_info array. We round it
-	 * to the next power of two because this approximation is done
-	 * internally by kmalloc so we can have some more memory
-	 * for free here (e.g. may be used for META_BG resize).
-	 */
-	array_size = 1;
-	while (array_size < sizeof(*sbi->s_group_info) *
-	       num_meta_group_infos_max)
-		array_size = array_size << 1;
-	/* An 8TB filesystem with 64-bit pointers requires a 4096 byte
-	 * kmalloc. A 128kb malloc should suffice for a 256TB filesystem.
-	 * So a two level scheme suffices for now. */
-	sbi->s_group_info = ext4_kvzalloc(array_size, GFP_KERNEL);
-	if (sbi->s_group_info == NULL) {
-		ext4_msg(sb, KERN_ERR, "can't allocate buddy meta group");
-		return -ENOMEM;
-	}
 	sbi->s_buddy_cache = new_inode(sb);
 	if (sbi->s_buddy_cache == NULL) {
 		ext4_msg(sb, KERN_ERR, "can't get new inode");
@@ -2322,7 +2323,7 @@ static int ext4_mb_init_backend(struct super_block *sb)
 	cachep = get_groupinfo_cache(sb->s_blocksize_bits);
 	while (i-- > 0)
 		kmem_cache_free(cachep, ext4_get_group_info(sb, i));
-	i = num_meta_group_infos;
+	i = sbi->s_group_info_size;
 	while (i-- > 0)
 		kfree(sbi->s_group_info[i]);
 	iput(sbi->s_buddy_cache);
@@ -4008,7 +4009,6 @@ ext4_mb_initialize_context(struct ext4_allocation_context *ac,
 	ext4_get_group_no_and_offset(sb, goal, &group, &block);
 
 	/* set up allocation goals */
-	memset(ac, 0, sizeof(struct ext4_allocation_context));
 	ac->ac_b_ex.fe_logical = ar->logical & ~(sbi->s_cluster_ratio - 1);
 	ac->ac_status = AC_STATUS_CONTINUE;
 	ac->ac_sb = sb;
@@ -4291,7 +4291,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
 		}
 	}
 
-	ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS);
+	ac = kmem_cache_zalloc(ext4_ac_cachep, GFP_NOFS);
 	if (!ac) {
 		ar->len = 0;
 		*errp = -ENOMEM;
@@ -4657,6 +4657,8 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode,
 		 * with group lock held. generate_buddy look at
 		 * them with group lock_held
 		 */
+		if (test_opt(sb, DISCARD))
+			ext4_issue_discard(sb, block_group, bit, count);
 		ext4_lock_group(sb, block_group);
 		mb_clear_bits(bitmap_bh->b_data, bit, count_clusters);
 		mb_free_blocks(inode, &e4b, bit, count_clusters);
@@ -4988,7 +4990,8 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range)
 
 	start = range->start >> sb->s_blocksize_bits;
 	end = start + (range->len >> sb->s_blocksize_bits) - 1;
-	minlen = range->minlen >> sb->s_blocksize_bits;
+	minlen = EXT4_NUM_B2C(EXT4_SB(sb),
+			      range->minlen >> sb->s_blocksize_bits);
 
 	if (unlikely(minlen > EXT4_CLUSTERS_PER_GROUP(sb)) ||
 	    unlikely(start >= max_blks))
@@ -5048,6 +5051,6 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range)
 		atomic_set(&EXT4_SB(sb)->s_last_trim_minblks, minlen);
 
 out:
-	range->len = trimmed * sb->s_blocksize;
+	range->len = EXT4_C2B(EXT4_SB(sb), trimmed) << sb->s_blocksize_bits;
 	return ret;
 }

fs/ext4/mballoc.h

@@ -64,11 +64,6 @@ extern u8 mb_enable_debug;
  */
 #define MB_DEFAULT_MIN_TO_SCAN		10
 
-/*
- * How many groups mballoc will scan looking for the best chunk
- */
-#define MB_DEFAULT_MAX_GROUPS_TO_SCAN	5
-
 /*
  * with 'ext4_mb_stats' allocator will collect stats that will be
  * shown at umount. The collecting costs though!

fs/ext4/namei.c

@@ -55,6 +55,13 @@ static struct buffer_head *ext4_append(handle_t *handle,
 {
 	struct buffer_head *bh;
 
+	if (unlikely(EXT4_SB(inode->i_sb)->s_max_dir_size_kb &&
+		     ((inode->i_size >> 10) >=
+		      EXT4_SB(inode->i_sb)->s_max_dir_size_kb))) {
+		*err = -ENOSPC;
+		return NULL;
+	}
+
 	*block = inode->i_size >> inode->i_sb->s_blocksize_bits;
 
 	bh = ext4_bread(handle, inode, *block, 1, err);
@@ -67,6 +74,12 @@ static struct buffer_head *ext4_append(handle_t *handle,
 			bh = NULL;
 		}
 	}
+	if (!bh && !(*err)) {
+		*err = -EIO;
+		ext4_error(inode->i_sb,
+			   "Directory hole detected on inode %lu\n",
+			   inode->i_ino);
+	}
 	return bh;
 }
 
@@ -594,8 +607,11 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 	u32 hash;
 
 	frame->bh = NULL;
-	if (!(bh = ext4_bread (NULL,dir, 0, 0, err)))
+	if (!(bh = ext4_bread(NULL, dir, 0, 0, err))) {
+		if (*err == 0)
+			*err = ERR_BAD_DX_DIR;
 		goto fail;
+	}
 	root = (struct dx_root *) bh->b_data;
 	if (root->info.hash_version != DX_HASH_TEA &&
 	    root->info.hash_version != DX_HASH_HALF_MD4 &&
@@ -696,8 +712,11 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 		frame->entries = entries;
 		frame->at = at;
 		if (!indirect--) return frame;
-		if (!(bh = ext4_bread (NULL,dir, dx_get_block(at), 0, err)))
+		if (!(bh = ext4_bread(NULL, dir, dx_get_block(at), 0, err))) {
+			if (!(*err))
+				*err = ERR_BAD_DX_DIR;
 			goto fail2;
+		}
 		at = entries = ((struct dx_node *) bh->b_data)->entries;
 
 		if (!buffer_verified(bh) &&
@@ -807,8 +826,15 @@ static int ext4_htree_next_block(struct inode *dir, __u32 hash,
 	 */
 	while (num_frames--) {
 		if (!(bh = ext4_bread(NULL, dir, dx_get_block(p->at),
-				      0, &err)))
+				      0, &err))) {
+			if (!err) {
+				ext4_error(dir->i_sb,
+					   "Directory hole detected on inode %lu\n",
+					   dir->i_ino);
+				return -EIO;
+			}
 			return err; /* Failure */
+		}
 
 		if (!buffer_verified(bh) &&
 		    !ext4_dx_csum_verify(dir,
@@ -839,12 +865,19 @@ static int htree_dirblock_to_tree(struct file *dir_file,
 {
 	struct buffer_head *bh;
 	struct ext4_dir_entry_2 *de, *top;
-	int err, count = 0;
+	int err = 0, count = 0;
 
 	dxtrace(printk(KERN_INFO "In htree dirblock_to_tree: block %lu\n",
 							(unsigned long)block));
-	if (!(bh = ext4_bread (NULL, dir, block, 0, &err)))
+	if (!(bh = ext4_bread(NULL, dir, block, 0, &err))) {
+		if (!err) {
+			err = -EIO;
+			ext4_error(dir->i_sb,
+				   "Directory hole detected on inode %lu\n",
+				   dir->i_ino);
+		}
 		return err;
+	}
 
 	if (!buffer_verified(bh) &&
 	    !ext4_dirent_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data))
@@ -1267,8 +1300,15 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q
 		return NULL;
 	do {
 		block = dx_get_block(frame->at);
-		if (!(bh = ext4_bread(NULL, dir, block, 0, err)))
+		if (!(bh = ext4_bread(NULL, dir, block, 0, err))) {
+			if (!(*err)) {
+				*err = -EIO;
+				ext4_error(dir->i_sb,
+					   "Directory hole detected on inode %lu\n",
+					   dir->i_ino);
+			}
 			goto errout;
+		}
 
 		if (!buffer_verified(bh) &&
 		    !ext4_dirent_csum_verify(dir,
@@ -1801,9 +1841,15 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry,
 	}
 	blocks = dir->i_size >> sb->s_blocksize_bits;
 	for (block = 0; block < blocks; block++) {
-		bh = ext4_bread(handle, dir, block, 0, &retval);
-		if(!bh)
+		if (!(bh = ext4_bread(handle, dir, block, 0, &retval))) {
+			if (!retval) {
+				retval = -EIO;
+				ext4_error(inode->i_sb,
+					   "Directory hole detected on inode %lu\n",
+					   inode->i_ino);
+			}
 			return retval;
+		}
 		if (!buffer_verified(bh) &&
 		    !ext4_dirent_csum_verify(dir,
 				(struct ext4_dir_entry *)bh->b_data))
@@ -1860,8 +1906,15 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry,
 	entries = frame->entries;
 	at = frame->at;
 
-	if (!(bh = ext4_bread(handle,dir, dx_get_block(frame->at), 0, &err)))
+	if (!(bh = ext4_bread(handle, dir, dx_get_block(frame->at), 0, &err))) {
+		if (!err) {
+			err = -EIO;
+			ext4_error(dir->i_sb,
+				   "Directory hole detected on inode %lu\n",
+				   dir->i_ino);
+		}
 		goto cleanup;
+	}
 
 	if (!buffer_verified(bh) &&
 	    !ext4_dirent_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data))
@@ -2149,9 +2202,7 @@ static int ext4_mknod(struct inode *dir, struct dentry *dentry,
 	err = PTR_ERR(inode);
 	if (!IS_ERR(inode)) {
 		init_special_inode(inode, inode->i_mode, rdev);
-#ifdef CONFIG_EXT4_FS_XATTR
 		inode->i_op = &ext4_special_inode_operations;
-#endif
 		err = ext4_add_nondir(handle, dentry, inode);
 	}
 	ext4_journal_stop(handle);
@@ -2199,9 +2250,15 @@ static int ext4_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 	inode->i_op = &ext4_dir_inode_operations;
 	inode->i_fop = &ext4_dir_operations;
 	inode->i_size = EXT4_I(inode)->i_disksize = inode->i_sb->s_blocksize;
-	dir_block = ext4_bread(handle, inode, 0, 1, &err);
-	if (!dir_block)
+	if (!(dir_block = ext4_bread(handle, inode, 0, 1, &err))) {
+		if (!err) {
+			err = -EIO;
+			ext4_error(inode->i_sb,
+				   "Directory hole detected on inode %lu\n",
+				   inode->i_ino);
+		}
 		goto out_clear_inode;
+	}
 	BUFFER_TRACE(dir_block, "get_write_access");
 	err = ext4_journal_get_write_access(handle, dir_block);
 	if (err)
@@ -2318,6 +2375,11 @@ static int empty_dir(struct inode *inode)
 					EXT4_ERROR_INODE(inode,
 						"error %d reading directory "
 						"lblock %u", err, lblock);
+				else
+					ext4_warning(inode->i_sb,
+						"bad directory (dir #%lu) - no data block",
+						inode->i_ino);
+
 				offset += sb->s_blocksize;
 				continue;
 			}
@@ -2362,7 +2424,7 @@ int ext4_orphan_add(handle_t *handle, struct inode *inode)
 	struct ext4_iloc iloc;
 	int err = 0, rc;
 
-	if (!ext4_handle_valid(handle))
+	if (!EXT4_SB(sb)->s_journal)
 		return 0;
 
 	mutex_lock(&EXT4_SB(sb)->s_orphan_lock);
@@ -2436,8 +2498,7 @@ int ext4_orphan_del(handle_t *handle, struct inode *inode)
 	struct ext4_iloc iloc;
 	int err = 0;
 
-	/* ext4_handle_valid() assumes a valid handle_t pointer */
-	if (handle && !ext4_handle_valid(handle))
+	if (!EXT4_SB(inode->i_sb)->s_journal)
 		return 0;
 
 	mutex_lock(&EXT4_SB(inode->i_sb)->s_orphan_lock);
@@ -2456,7 +2517,7 @@ int ext4_orphan_del(handle_t *handle, struct inode *inode)
 	 * transaction handle with which to update the orphan list on
 	 * disk, but we still need to remove the inode from the linked
 	 * list in memory. */
-	if (sbi->s_journal && !handle)
+	if (!handle)
 		goto out;
 
 	err = ext4_reserve_inode_write(handle, inode, &iloc);
@@ -2826,9 +2887,15 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 				goto end_rename;
 		}
 		retval = -EIO;
-		dir_bh = ext4_bread(handle, old_inode, 0, 0, &retval);
-		if (!dir_bh)
+		if (!(dir_bh = ext4_bread(handle, old_inode, 0, 0, &retval))) {
+			if (!retval) {
+				retval = -EIO;
+				ext4_error(old_inode->i_sb,
+					   "Directory hole detected on inode %lu\n",
+					   old_inode->i_ino);
+			}
 			goto end_rename;
+		}
 		if (!buffer_verified(dir_bh) &&
 		    !ext4_dirent_csum_verify(old_inode,
 				(struct ext4_dir_entry *)dir_bh->b_data))

fs/ext4/page-io.c

@@ -71,6 +71,9 @@ void ext4_free_io_end(ext4_io_end_t *io)
 	int i;
 
 	BUG_ON(!io);
+	BUG_ON(!list_empty(&io->list));
+	BUG_ON(io->flag & EXT4_IO_END_UNWRITTEN);
+
 	if (io->page)
 		put_page(io->page);
 	for (i = 0; i < io->num_io_pages; i++)
@@ -81,13 +84,8 @@ void ext4_free_io_end(ext4_io_end_t *io)
 	kmem_cache_free(io_end_cachep, io);
 }
 
-/*
- * check a range of space and convert unwritten extents to written.
- *
- * Called with inode->i_mutex; we depend on this when we manipulate
- * io->flag, since we could otherwise race with ext4_flush_completed_IO()
- */
-int ext4_end_io_nolock(ext4_io_end_t *io)
+/* check a range of space and convert unwritten extents to written. */
+static int ext4_end_io(ext4_io_end_t *io)
 {
 	struct inode *inode = io->inode;
 	loff_t offset = io->offset;
@@ -106,63 +104,136 @@ int ext4_end_io_nolock(ext4_io_end_t *io)
 			 "(inode %lu, offset %llu, size %zd, error %d)",
 			 inode->i_ino, offset, size, ret);
 	}
-
 	if (io->iocb)
 		aio_complete(io->iocb, io->result, 0);
 
 	if (io->flag & EXT4_IO_END_DIRECT)
 		inode_dio_done(inode);
 	/* Wake up anyone waiting on unwritten extent conversion */
-	if (atomic_dec_and_test(&EXT4_I(inode)->i_aiodio_unwritten))
+	if (atomic_dec_and_test(&EXT4_I(inode)->i_unwritten))
 		wake_up_all(ext4_ioend_wq(io->inode));
 	return ret;
 }
 
-/*
- * work on completed aio dio IO, to convert unwritten extents to extents
- */
-static void ext4_end_io_work(struct work_struct *work)
+static void dump_completed_IO(struct inode *inode)
+{
+#ifdef	EXT4FS_DEBUG
+	struct list_head *cur, *before, *after;
+	ext4_io_end_t *io, *io0, *io1;
+	unsigned long flags;
+
+	if (list_empty(&EXT4_I(inode)->i_completed_io_list)) {
+		ext4_debug("inode %lu completed_io list is empty\n",
+			   inode->i_ino);
+		return;
+	}
+
+	ext4_debug("Dump inode %lu completed_io list\n", inode->i_ino);
+	list_for_each_entry(io, &EXT4_I(inode)->i_completed_io_list, list) {
+		cur = &io->list;
+		before = cur->prev;
+		io0 = container_of(before, ext4_io_end_t, list);
+		after = cur->next;
+		io1 = container_of(after, ext4_io_end_t, list);
+
+		ext4_debug("io 0x%p from inode %lu,prev 0x%p,next 0x%p\n",
+			    io, inode->i_ino, io0, io1);
+	}
+#endif
+}
+
+/* Add the io_end to per-inode completed end_io list. */
+void ext4_add_complete_io(ext4_io_end_t *io_end)
 {
-	ext4_io_end_t		*io = container_of(work, ext4_io_end_t, work);
-	struct inode		*inode = io->inode;
-	struct ext4_inode_info	*ei = EXT4_I(inode);
-	unsigned long		flags;
+	struct ext4_inode_info *ei = EXT4_I(io_end->inode);
+	struct workqueue_struct *wq;
+	unsigned long flags;
+
+	BUG_ON(!(io_end->flag & EXT4_IO_END_UNWRITTEN));
+	wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq;
 
 	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
-	if (io->flag & EXT4_IO_END_IN_FSYNC)
-		goto requeue;
-	if (list_empty(&io->list)) {
-		spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
-		goto free;
+	if (list_empty(&ei->i_completed_io_list)) {
+		io_end->flag |= EXT4_IO_END_QUEUED;
+		queue_work(wq, &io_end->work);
 	}
+	list_add_tail(&io_end->list, &ei->i_completed_io_list);
+	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
+}
 
-	if (!mutex_trylock(&inode->i_mutex)) {
-		bool was_queued;
-requeue:
-		was_queued = !!(io->flag & EXT4_IO_END_QUEUED);
-		io->flag |= EXT4_IO_END_QUEUED;
-		spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
-		/*
-		 * Requeue the work instead of waiting so that the work
-		 * items queued after this can be processed.
-		 */
-		queue_work(EXT4_SB(inode->i_sb)->dio_unwritten_wq, &io->work);
-		/*
-		 * To prevent the ext4-dio-unwritten thread from keeping
-		 * requeueing end_io requests and occupying cpu for too long,
-		 * yield the cpu if it sees an end_io request that has already
-		 * been requeued.
-		 */
-		if (was_queued)
-			yield();
-		return;
+static int ext4_do_flush_completed_IO(struct inode *inode,
+				      ext4_io_end_t *work_io)
+{
+	ext4_io_end_t *io;
+	struct list_head unwritten, complete, to_free;
+	unsigned long flags;
+	struct ext4_inode_info *ei = EXT4_I(inode);
+	int err, ret = 0;
+
+	INIT_LIST_HEAD(&complete);
+	INIT_LIST_HEAD(&to_free);
+
+	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
+	dump_completed_IO(inode);
+	list_replace_init(&ei->i_completed_io_list, &unwritten);
+	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
+
+	while (!list_empty(&unwritten)) {
+		io = list_entry(unwritten.next, ext4_io_end_t, list);
+		BUG_ON(!(io->flag & EXT4_IO_END_UNWRITTEN));
+		list_del_init(&io->list);
+
+		err = ext4_end_io(io);
+		if (unlikely(!ret && err))
+			ret = err;
+
+		list_add_tail(&io->list, &complete);
+	}
+	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
+	while (!list_empty(&complete)) {
+		io = list_entry(complete.next, ext4_io_end_t, list);
+		io->flag &= ~EXT4_IO_END_UNWRITTEN;
+		/* end_io context can not be destroyed now because it still
+		 * used by queued worker. Worker thread will destroy it later */
+		if (io->flag & EXT4_IO_END_QUEUED)
+			list_del_init(&io->list);
+		else
+			list_move(&io->list, &to_free);
+	}
+	/* If we are called from worker context, it is time to clear queued
+	 * flag, and destroy it's end_io if it was converted already */
+	if (work_io) {
+		work_io->flag &= ~EXT4_IO_END_QUEUED;
+		if (!(work_io->flag & EXT4_IO_END_UNWRITTEN))
+			list_add_tail(&work_io->list, &to_free);
 	}
-	list_del_init(&io->list);
 	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
-	(void) ext4_end_io_nolock(io);
-	mutex_unlock(&inode->i_mutex);
-free:
-	ext4_free_io_end(io);
+
+	while (!list_empty(&to_free)) {
+		io = list_entry(to_free.next, ext4_io_end_t, list);
+		list_del_init(&io->list);
+		ext4_free_io_end(io);
+	}
+	return ret;
+}
+
+/*
+ * work on completed aio dio IO, to convert unwritten extents to extents
+ */
+static void ext4_end_io_work(struct work_struct *work)
+{
+	ext4_io_end_t *io = container_of(work, ext4_io_end_t, work);
+	ext4_do_flush_completed_IO(io->inode, io);
+}
+
+int ext4_flush_unwritten_io(struct inode *inode)
+{
+	int ret;
+	WARN_ON_ONCE(!mutex_is_locked(&inode->i_mutex) &&
+		     !(inode->i_state & I_FREEING));
+	ret = ext4_do_flush_completed_IO(inode, NULL);
+	ext4_unwritten_wait(inode);
+	return ret;
 }
 
 ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags)
@@ -195,9 +266,7 @@ static void buffer_io_error(struct buffer_head *bh)
 static void ext4_end_bio(struct bio *bio, int error)
 {
 	ext4_io_end_t *io_end = bio->bi_private;
-	struct workqueue_struct *wq;
 	struct inode *inode;
-	unsigned long flags;
 	int i;
 	sector_t bi_sector = bio->bi_sector;
 
@@ -255,14 +324,7 @@ static void ext4_end_bio(struct bio *bio, int error)
 		return;
 	}
 
-	/* Add the io_end to per-inode completed io list*/
-	spin_lock_irqsave(&EXT4_I(inode)->i_completed_io_lock, flags);
-	list_add_tail(&io_end->list, &EXT4_I(inode)->i_completed_io_list);
-	spin_unlock_irqrestore(&EXT4_I(inode)->i_completed_io_lock, flags);
-
-	wq = EXT4_SB(inode->i_sb)->dio_unwritten_wq;
-	/* queue the work to convert unwritten extents to written */
-	queue_work(wq, &io_end->work);
+	ext4_add_complete_io(io_end);
 }
 
 void ext4_io_submit(struct ext4_io_submit *io)

fs/ext4/super.c

@@ -420,7 +420,7 @@ static void __save_error_info(struct super_block *sb, const char *func,
 	 */
 	if (!es->s_error_count)
 		mod_timer(&EXT4_SB(sb)->s_err_report, jiffies + 24*60*60*HZ);
-	es->s_error_count = cpu_to_le32(le32_to_cpu(es->s_error_count) + 1);
+	le32_add_cpu(&es->s_error_count, 1);
 }
 
 static void save_error_info(struct super_block *sb, const char *func,
@@ -850,7 +850,6 @@ static void ext4_put_super(struct super_block *sb)
 	flush_workqueue(sbi->dio_unwritten_wq);
 	destroy_workqueue(sbi->dio_unwritten_wq);
 
-	lock_super(sb);
 	if (sbi->s_journal) {
 		err = jbd2_journal_destroy(sbi->s_journal);
 		sbi->s_journal = NULL;
@@ -917,7 +916,6 @@ static void ext4_put_super(struct super_block *sb)
 	 * Now that we are completely done shutting down the
 	 * superblock, we need to actually destroy the kobject.
 	 */
-	unlock_super(sb);
 	kobject_put(&sbi->s_kobj);
 	wait_for_completion(&sbi->s_kobj_unregister);
 	if (sbi->s_chksum_driver)
@@ -956,11 +954,10 @@ static struct inode *ext4_alloc_inode(struct super_block *sb)
 	ei->jinode = NULL;
 	INIT_LIST_HEAD(&ei->i_completed_io_list);
 	spin_lock_init(&ei->i_completed_io_lock);
-	ei->cur_aio_dio = NULL;
 	ei->i_sync_tid = 0;
 	ei->i_datasync_tid = 0;
 	atomic_set(&ei->i_ioend_count, 0);
-	atomic_set(&ei->i_aiodio_unwritten, 0);
+	atomic_set(&ei->i_unwritten, 0);
 
 	return &ei->vfs_inode;
 }
@@ -1224,6 +1221,7 @@ enum {
 	Opt_inode_readahead_blks, Opt_journal_ioprio,
 	Opt_dioread_nolock, Opt_dioread_lock,
 	Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
+	Opt_max_dir_size_kb,
 };
 
 static const match_table_t tokens = {
@@ -1297,6 +1295,7 @@ static const match_table_t tokens = {
 	{Opt_init_itable, "init_itable=%u"},
 	{Opt_init_itable, "init_itable"},
 	{Opt_noinit_itable, "noinit_itable"},
+	{Opt_max_dir_size_kb, "max_dir_size_kb=%u"},
 	{Opt_removed, "check=none"},	/* mount option from ext2/3 */
 	{Opt_removed, "nocheck"},	/* mount option from ext2/3 */
 	{Opt_removed, "reservation"},	/* mount option from ext2/3 */
@@ -1477,6 +1476,7 @@ static const struct mount_opts {
 	{Opt_jqfmt_vfsold, QFMT_VFS_OLD, MOPT_QFMT},
 	{Opt_jqfmt_vfsv0, QFMT_VFS_V0, MOPT_QFMT},
 	{Opt_jqfmt_vfsv1, QFMT_VFS_V1, MOPT_QFMT},
+	{Opt_max_dir_size_kb, 0, MOPT_GTE0},
 	{Opt_err, 0, 0}
 };
 
@@ -1592,6 +1592,8 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
 			if (!args->from)
 				arg = EXT4_DEF_LI_WAIT_MULT;
 			sbi->s_li_wait_mult = arg;
+		} else if (token == Opt_max_dir_size_kb) {
+			sbi->s_max_dir_size_kb = arg;
 		} else if (token == Opt_stripe) {
 			sbi->s_stripe = arg;
 		} else if (m->flags & MOPT_DATAJ) {
@@ -1664,7 +1666,7 @@ static int parse_options(char *options, struct super_block *sb,
 		 * Initialize args struct so we know whether arg was
 		 * found; some options take optional arguments.
 		 */
-		args[0].to = args[0].from = 0;
+		args[0].to = args[0].from = NULL;
 		token = match_token(p, tokens, args);
 		if (handle_mount_opt(sb, p, token, args, journal_devnum,
 				     journal_ioprio, is_remount) < 0)
@@ -1740,7 +1742,7 @@ static inline void ext4_show_quota_options(struct seq_file *seq,
 
 static const char *token2str(int token)
 {
-	static const struct match_token *t;
+	const struct match_token *t;
 
 	for (t = tokens; t->token != Opt_err; t++)
 		if (t->token == token && !strchr(t->pattern, '='))
@@ -1823,6 +1825,8 @@ static int _ext4_show_options(struct seq_file *seq, struct super_block *sb,
 	if (nodefs || (test_opt(sb, INIT_INODE_TABLE) &&
 		       (sbi->s_li_wait_mult != EXT4_DEF_LI_WAIT_MULT)))
 		SEQ_OPTS_PRINT("init_itable=%u", sbi->s_li_wait_mult);
+	if (nodefs || sbi->s_max_dir_size_kb)
+		SEQ_OPTS_PRINT("max_dir_size_kb=%u", sbi->s_max_dir_size_kb);
 
 	ext4_show_quota_options(seq, sb);
 	return 0;
@@ -1914,15 +1918,45 @@ static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es,
 	return res;
 }
 
+int ext4_alloc_flex_bg_array(struct super_block *sb, ext4_group_t ngroup)
+{
+	struct ext4_sb_info *sbi = EXT4_SB(sb);
+	struct flex_groups *new_groups;
+	int size;
+
+	if (!sbi->s_log_groups_per_flex)
+		return 0;
+
+	size = ext4_flex_group(sbi, ngroup - 1) + 1;
+	if (size <= sbi->s_flex_groups_allocated)
+		return 0;
+
+	size = roundup_pow_of_two(size * sizeof(struct flex_groups));
+	new_groups = ext4_kvzalloc(size, GFP_KERNEL);
+	if (!new_groups) {
+		ext4_msg(sb, KERN_ERR, "not enough memory for %d flex groups",
+			 size / (int) sizeof(struct flex_groups));
+		return -ENOMEM;
+	}
+
+	if (sbi->s_flex_groups) {
+		memcpy(new_groups, sbi->s_flex_groups,
+		       (sbi->s_flex_groups_allocated *
+			sizeof(struct flex_groups)));
+		ext4_kvfree(sbi->s_flex_groups);
+	}
+	sbi->s_flex_groups = new_groups;
+	sbi->s_flex_groups_allocated = size / sizeof(struct flex_groups);
+	return 0;
+}
+
 static int ext4_fill_flex_info(struct super_block *sb)
 {
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 	struct ext4_group_desc *gdp = NULL;
-	ext4_group_t flex_group_count;
 	ext4_group_t flex_group;
 	unsigned int groups_per_flex = 0;
-	size_t size;
-	int i;
+	int i, err;
 
 	sbi->s_log_groups_per_flex = sbi->s_es->s_log_groups_per_flex;
 	if (sbi->s_log_groups_per_flex < 1 || sbi->s_log_groups_per_flex > 31) {
@@ -1931,17 +1965,9 @@ static int ext4_fill_flex_info(struct super_block *sb)
 	}
 	groups_per_flex = 1 << sbi->s_log_groups_per_flex;
 
-	/* We allocate both existing and potentially added groups */
-	flex_group_count = ((sbi->s_groups_count + groups_per_flex - 1) +
-			((le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) + 1) <<
-			      EXT4_DESC_PER_BLOCK_BITS(sb))) / groups_per_flex;
-	size = flex_group_count * sizeof(struct flex_groups);
-	sbi->s_flex_groups = ext4_kvzalloc(size, GFP_KERNEL);
-	if (sbi->s_flex_groups == NULL) {
-		ext4_msg(sb, KERN_ERR, "not enough memory for %u flex groups",
-			 flex_group_count);
+	err = ext4_alloc_flex_bg_array(sb, sbi->s_groups_count);
+	if (err)
 		goto failed;
-	}
 
 	for (i = 0; i < sbi->s_groups_count; i++) {
 		gdp = ext4_get_group_desc(sb, i, NULL);
@@ -2144,10 +2170,12 @@ static void ext4_orphan_cleanup(struct super_block *sb,
 	}
 
 	if (EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS) {
-		if (es->s_last_orphan)
+		/* don't clear list on RO mount w/ errors */
+		if (es->s_last_orphan && !(s_flags & MS_RDONLY)) {
 			jbd_debug(1, "Errors on filesystem, "
 				  "clearing orphan list.\n");
-		es->s_last_orphan = 0;
+			es->s_last_orphan = 0;
+		}
 		jbd_debug(1, "Skipping orphan recovery on fs with errors.\n");
 		return;
 	}
@@ -2528,6 +2556,7 @@ EXT4_RW_ATTR_SBI_UI(mb_order2_req, s_mb_order2_reqs);
 EXT4_RW_ATTR_SBI_UI(mb_stream_req, s_mb_stream_request);
 EXT4_RW_ATTR_SBI_UI(mb_group_prealloc, s_mb_group_prealloc);
 EXT4_RW_ATTR_SBI_UI(max_writeback_mb_bump, s_max_writeback_mb_bump);
+EXT4_RW_ATTR_SBI_UI(extent_max_zeroout_kb, s_extent_max_zeroout_kb);
 EXT4_ATTR(trigger_fs_error, 0200, NULL, trigger_test_error);
 
 static struct attribute *ext4_attrs[] = {
@@ -2543,6 +2572,7 @@ static struct attribute *ext4_attrs[] = {
 	ATTR_LIST(mb_stream_req),
 	ATTR_LIST(mb_group_prealloc),
 	ATTR_LIST(max_writeback_mb_bump),
+	ATTR_LIST(extent_max_zeroout_kb),
 	ATTR_LIST(trigger_fs_error),
 	NULL,
 };
@@ -2550,10 +2580,12 @@ static struct attribute *ext4_attrs[] = {
 /* Features this copy of ext4 supports */
 EXT4_INFO_ATTR(lazy_itable_init);
 EXT4_INFO_ATTR(batched_discard);
+EXT4_INFO_ATTR(meta_bg_resize);
 
 static struct attribute *ext4_feat_attrs[] = {
 	ATTR_LIST(lazy_itable_init),
 	ATTR_LIST(batched_discard),
+	ATTR_LIST(meta_bg_resize),
 	NULL,
 };
 
@@ -3374,7 +3406,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	 * enable delayed allocation by default
 	 * Use -o nodelalloc to turn it off
 	 */
-	if (!IS_EXT3_SB(sb) &&
+	if (!IS_EXT3_SB(sb) && !IS_EXT2_SB(sb) &&
 	    ((def_mount_opts & EXT4_DEFM_NODELALLOC) == 0))
 		set_opt(sb, DELALLOC);
 
@@ -3743,6 +3775,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 
 	sbi->s_stripe = ext4_get_stripe_size(sbi);
 	sbi->s_max_writeback_mb_bump = 128;
+	sbi->s_extent_max_zeroout_kb = 32;
 
 	/*
 	 * set up enough so that it can read an inode
@@ -4519,11 +4552,9 @@ static int ext4_unfreeze(struct super_block *sb)
 	if (sb->s_flags & MS_RDONLY)
 		return 0;
 
-	lock_super(sb);
 	/* Reset the needs_recovery flag before the fs is unlocked. */
 	EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
 	ext4_commit_super(sb, 1);
-	unlock_super(sb);
 	return 0;
 }
 
@@ -4559,7 +4590,6 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 	char *orig_data = kstrdup(data, GFP_KERNEL);
 
 	/* Store the original options */
-	lock_super(sb);
 	old_sb_flags = sb->s_flags;
 	old_opts.s_mount_opt = sbi->s_mount_opt;
 	old_opts.s_mount_opt2 = sbi->s_mount_opt2;
@@ -4701,7 +4731,6 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 	if (sbi->s_journal == NULL)
 		ext4_commit_super(sb, 1);
 
-	unlock_super(sb);
 #ifdef CONFIG_QUOTA
 	/* Release old quota file names */
 	for (i = 0; i < MAXQUOTAS; i++)
@@ -4714,10 +4743,8 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 		else if (EXT4_HAS_RO_COMPAT_FEATURE(sb,
 					EXT4_FEATURE_RO_COMPAT_QUOTA)) {
 			err = ext4_enable_quotas(sb);
-			if (err) {
-				lock_super(sb);
+			if (err)
 				goto restore_opts;
-			}
 		}
 	}
 #endif
@@ -4744,7 +4771,6 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 		sbi->s_qf_names[i] = old_opts.s_qf_names[i];
 	}
 #endif
-	unlock_super(sb);
 	kfree(orig_data);
 	return err;
 }
@@ -5269,8 +5295,10 @@ static int __init ext4_init_fs(void)
 	if (err)
 		goto out6;
 	ext4_kset = kset_create_and_add("ext4", NULL, fs_kobj);
-	if (!ext4_kset)
+	if (!ext4_kset) {
+		err = -ENOMEM;
 		goto out5;
+	}
 	ext4_proc_root = proc_mkdir("fs/ext4", NULL);
 
 	err = ext4_init_feat_adverts();

fs/fs-writeback.c

@@ -63,6 +63,7 @@ int writeback_in_progress(struct backing_dev_info *bdi)
 {
 	return test_bit(BDI_writeback_running, &bdi->state);
 }
+EXPORT_SYMBOL(writeback_in_progress);
 
 static inline struct backing_dev_info *inode_to_bdi(struct inode *inode)
 {

fs/jbd2/commit.c

@@ -1014,17 +1014,35 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 		 * there's no point in keeping a checkpoint record for
 		 * it. */
 
-		/* A buffer which has been freed while still being
-		 * journaled by a previous transaction may end up still
-		 * being dirty here, but we want to avoid writing back
-		 * that buffer in the future after the "add to orphan"
-		 * operation been committed,  That's not only a performance
-		 * gain, it also stops aliasing problems if the buffer is
-		 * left behind for writeback and gets reallocated for another
-		 * use in a different page. */
-		if (buffer_freed(bh) && !jh->b_next_transaction) {
-			clear_buffer_freed(bh);
-			clear_buffer_jbddirty(bh);
+		/*
+		* A buffer which has been freed while still being journaled by
+		* a previous transaction.
+		*/
+		if (buffer_freed(bh)) {
+			/*
+			 * If the running transaction is the one containing
+			 * "add to orphan" operation (b_next_transaction !=
+			 * NULL), we have to wait for that transaction to
+			 * commit before we can really get rid of the buffer.
+			 * So just clear b_modified to not confuse transaction
+			 * credit accounting and refile the buffer to
+			 * BJ_Forget of the running transaction. If the just
+			 * committed transaction contains "add to orphan"
+			 * operation, we can completely invalidate the buffer
+			 * now. We are rather through in that since the
+			 * buffer may be still accessible when blocksize <
+			 * pagesize and it is attached to the last partial
+			 * page.
+			 */
+			jh->b_modified = 0;
+			if (!jh->b_next_transaction) {
+				clear_buffer_freed(bh);
+				clear_buffer_jbddirty(bh);
+				clear_buffer_mapped(bh);
+				clear_buffer_new(bh);
+				clear_buffer_req(bh);
+				bh->b_bdev = NULL;
+			}
 		}
 
 		if (buffer_jbddirty(bh)) {

fs/jbd2/journal.c

@@ -1354,6 +1354,11 @@ static void jbd2_mark_journal_empty(journal_t *journal)
 
 	BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex));
 	read_lock(&journal->j_state_lock);
+	/* Is it already empty? */
+	if (sb->s_start == 0) {
+		read_unlock(&journal->j_state_lock);
+		return;
+	}
 	jbd_debug(1, "JBD2: Marking journal as empty (seq %d)\n",
 		  journal->j_tail_sequence);
 

fs/jbd2/recovery.c

@@ -289,8 +289,11 @@ int jbd2_journal_recover(journal_t *journal)
 	if (!err)
 		err = err2;
 	/* Make sure all replayed data is on permanent storage */
-	if (journal->j_flags & JBD2_BARRIER)
-		blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL);
+	if (journal->j_flags & JBD2_BARRIER) {
+		err2 = blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL);
+		if (!err)
+			err = err2;
+	}
 	return err;
 }
 

fs/jbd2/transaction.c

@@ -1841,15 +1841,16 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction)
  * We're outside-transaction here.  Either or both of j_running_transaction
  * and j_committing_transaction may be NULL.
  */
-static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
+static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh,
+				int partial_page)
 {
 	transaction_t *transaction;
 	struct journal_head *jh;
 	int may_free = 1;
-	int ret;
 
 	BUFFER_TRACE(bh, "entry");
 
+retry:
 	/*
 	 * It is safe to proceed here without the j_list_lock because the
 	 * buffers cannot be stolen by try_to_free_buffers as long as we are
@@ -1878,10 +1879,18 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 	 * clear the buffer dirty bit at latest at the moment when the
 	 * transaction marking the buffer as freed in the filesystem
 	 * structures is committed because from that moment on the
-	 * buffer can be reallocated and used by a different page.
+	 * block can be reallocated and used by a different page.
 	 * Since the block hasn't been freed yet but the inode has
 	 * already been added to orphan list, it is safe for us to add
 	 * the buffer to BJ_Forget list of the newest transaction.
+	 *
+	 * Also we have to clear buffer_mapped flag of a truncated buffer
+	 * because the buffer_head may be attached to the page straddling
+	 * i_size (can happen only when blocksize < pagesize) and thus the
+	 * buffer_head can be reused when the file is extended again. So we end
+	 * up keeping around invalidated buffers attached to transactions'
+	 * BJ_Forget list just to stop checkpointing code from cleaning up
+	 * the transaction this buffer was modified in.
 	 */
 	transaction = jh->b_transaction;
 	if (transaction == NULL) {
@@ -1908,13 +1917,9 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 			 * committed, the buffer won't be needed any
 			 * longer. */
 			JBUFFER_TRACE(jh, "checkpointed: add to BJ_Forget");
-			ret = __dispose_buffer(jh,
+			may_free = __dispose_buffer(jh,
 					journal->j_running_transaction);
-			jbd2_journal_put_journal_head(jh);
-			spin_unlock(&journal->j_list_lock);
-			jbd_unlock_bh_state(bh);
-			write_unlock(&journal->j_state_lock);
-			return ret;
+			goto zap_buffer;
 		} else {
 			/* There is no currently-running transaction. So the
 			 * orphan record which we wrote for this file must have
@@ -1922,13 +1927,9 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 			 * the committing transaction, if it exists. */
 			if (journal->j_committing_transaction) {
 				JBUFFER_TRACE(jh, "give to committing trans");
-				ret = __dispose_buffer(jh,
+				may_free = __dispose_buffer(jh,
 					journal->j_committing_transaction);
-				jbd2_journal_put_journal_head(jh);
-				spin_unlock(&journal->j_list_lock);
-				jbd_unlock_bh_state(bh);
-				write_unlock(&journal->j_state_lock);
-				return ret;
+				goto zap_buffer;
 			} else {
 				/* The orphan record's transaction has
 				 * committed.  We can cleanse this buffer */
@@ -1940,10 +1941,24 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 		JBUFFER_TRACE(jh, "on committing transaction");
 		/*
 		 * The buffer is committing, we simply cannot touch
-		 * it. So we just set j_next_transaction to the
-		 * running transaction (if there is one) and mark
-		 * buffer as freed so that commit code knows it should
-		 * clear dirty bits when it is done with the buffer.
+		 * it. If the page is straddling i_size we have to wait
+		 * for commit and try again.
+		 */
+		if (partial_page) {
+			tid_t tid = journal->j_committing_transaction->t_tid;
+
+			jbd2_journal_put_journal_head(jh);
+			spin_unlock(&journal->j_list_lock);
+			jbd_unlock_bh_state(bh);
+			write_unlock(&journal->j_state_lock);
+			jbd2_log_wait_commit(journal, tid);
+			goto retry;
+		}
+		/*
+		 * OK, buffer won't be reachable after truncate. We just set
+		 * j_next_transaction to the running transaction (if there is
+		 * one) and mark buffer as freed so that commit code knows it
+		 * should clear dirty bits when it is done with the buffer.
 		 */
 		set_buffer_freed(bh);
 		if (journal->j_running_transaction && buffer_jbddirty(bh))
@@ -1966,6 +1981,15 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 	}
 
 zap_buffer:
+	/*
+	 * This is tricky. Although the buffer is truncated, it may be reused
+	 * if blocksize < pagesize and it is attached to the page straddling
+	 * EOF. Since the buffer might have been added to BJ_Forget list of the
+	 * running transaction, journal_get_write_access() won't clear
+	 * b_modified and credit accounting gets confused. So clear b_modified
+	 * here.
+	 */
+	jh->b_modified = 0;
 	jbd2_journal_put_journal_head(jh);
 zap_buffer_no_jh:
 	spin_unlock(&journal->j_list_lock);
@@ -2017,7 +2041,8 @@ void jbd2_journal_invalidatepage(journal_t *journal,
 		if (offset <= curr_off) {
 			/* This block is wholly outside the truncation point */
 			lock_buffer(bh);
-			may_free &= journal_unmap_buffer(journal, bh);
+			may_free &= journal_unmap_buffer(journal, bh,
+							 offset > 0);
 			unlock_buffer(bh);
 		}
 		curr_off = next_off;

fs/nilfs2/file.c

@@ -116,6 +116,7 @@ static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	if (unlikely(ret))
 		goto out;
 
+	file_update_time(vma->vm_file);
 	ret = __block_page_mkwrite(vma, vmf, nilfs_get_block);
 	if (ret) {
 		nilfs_transaction_abort(inode->i_sb);

include/linux/falloc.h

@@ -3,6 +3,7 @@
 
 #define FALLOC_FL_KEEP_SIZE	0x01 /* default is extend size */
 #define FALLOC_FL_PUNCH_HOLE	0x02 /* de-allocates range */
+#define FALLOC_FL_NO_HIDE_STALE	0x04 /* reserved codepoint */
 
 #ifdef __KERNEL__