9 files changed, 270 insertions, 177 deletions

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index f4e387452246..7dd9b50d5ebc 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -892,7 +892,16 @@ out:
         return err;
 }
 
-#define DIO_CREDITS (EXT4_RESERVE_TRANS_BLOCKS + 32)
+/* Maximum number of blocks we map for direct IO at once. */
+#define DIO_MAX_BLOCKS 4096
+/*
+ * Number of credits we need for writing DIO_MAX_BLOCKS:
+ * We need sb + group descriptor + bitmap + inode -> 4
+ * For B blocks with A block pointers per block we need:
+ * 1 (triple ind.) + (B/A/A + 2) (doubly ind.) + (B/A + 2) (indirect).
+ * If we plug in 4096 for B and 256 for A (for 1KB block size), we get 25.
+ */
+#define DIO_CREDITS 25
 
 int ext4_get_blocks_wrap(handle_t *handle, struct inode *inode, sector_t block,
                         unsigned long max_blocks, struct buffer_head *bh,
@@ -939,49 +948,31 @@ static int ext4_get_block(struct inode *inode, sector_t iblock,
                         struct buffer_head *bh_result, int create)
 {
         handle_t *handle = ext4_journal_current_handle();
-        int ret = 0;
+        int ret = 0, started = 0;
         unsigned max_blocks = bh_result->b_size >> inode->i_blkbits;
 
-        if (!create)
-                goto get_block;         /* A read */
-
-        if (max_blocks == 1)
-                goto get_block;         /* A single block get */
-
-        if (handle->h_transaction->t_state == T_LOCKED) {
-                /*
-                 * Huge direct-io writes can hold off commits for long
-                 * periods of time.  Let this commit run.
-                 */
-                ext4_journal_stop(handle);
-                handle = ext4_journal_start(inode, DIO_CREDITS);
-                if (IS_ERR(handle))
+        if (create && !handle) {
+                /* Direct IO write... */
+                if (max_blocks > DIO_MAX_BLOCKS)
+                        max_blocks = DIO_MAX_BLOCKS;
+                handle = ext4_journal_start(inode, DIO_CREDITS +
+                              2 * EXT4_QUOTA_TRANS_BLOCKS(inode->i_sb));
+                if (IS_ERR(handle)) {
                         ret = PTR_ERR(handle);
-                goto get_block;
-        }
-
-        if (handle->h_buffer_credits <= EXT4_RESERVE_TRANS_BLOCKS) {
-                /*
-                 * Getting low on buffer credits...
-                 */
-                ret = ext4_journal_extend(handle, DIO_CREDITS);
-                if (ret > 0) {
-                        /*
-                         * Couldn't extend the transaction.  Start a new one.
-                         */
-                        ret = ext4_journal_restart(handle, DIO_CREDITS);
+                        goto out;
                 }
+                started = 1;
         }
 
-get_block:
-        if (ret == 0) {
-                ret = ext4_get_blocks_wrap(handle, inode, iblock,
+        ret = ext4_get_blocks_wrap(handle, inode, iblock,
                                         max_blocks, bh_result, create, 0);
-                if (ret > 0) {
-                        bh_result->b_size = (ret << inode->i_blkbits);
-                        ret = 0;
-                }
+        if (ret > 0) {
+                bh_result->b_size = (ret << inode->i_blkbits);
+                ret = 0;
         }
+        if (started)
+                ext4_journal_stop(handle);
+out:
         return ret;
 }
 
@@ -1671,7 +1662,8 @@ static int ext4_releasepage(struct page *page, gfp_t wait)
  * if the machine crashes during the write.
  *
  * If the O_DIRECT write is intantiating holes inside i_size and the machine
- * crashes then stale disk data _may_ be exposed inside the file.
+ * crashes then stale disk data _may_ be exposed inside the file. But current
+ * VFS code falls back into buffered path in that case so we are safe.
  */
 static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
                         const struct iovec *iov, loff_t offset,
@@ -1680,7 +1672,7 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
         struct file *file = iocb->ki_filp;
         struct inode *inode = file->f_mapping->host;
         struct ext4_inode_info *ei = EXT4_I(inode);
-        handle_t *handle = NULL;
+        handle_t *handle;
         ssize_t ret;
         int orphan = 0;
         size_t count = iov_length(iov, nr_segs);
@@ -1688,17 +1680,21 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
         if (rw == WRITE) {
                 loff_t final_size = offset + count;
 
-                handle = ext4_journal_start(inode, DIO_CREDITS);
-                if (IS_ERR(handle)) {
-                        ret = PTR_ERR(handle);
-                        goto out;
-                }
                 if (final_size > inode->i_size) {
+                        /* Credits for sb + inode write */
+                        handle = ext4_journal_start(inode, 2);
+                        if (IS_ERR(handle)) {
+                                ret = PTR_ERR(handle);
+                                goto out;
+                        }
                         ret = ext4_orphan_add(handle, inode);
-                        if (ret)
-                                goto out_stop;
+                        if (ret) {
+                                ext4_journal_stop(handle);
+                                goto out;
+                        }
                         orphan = 1;
                         ei->i_disksize = inode->i_size;
+                        ext4_journal_stop(handle);
                 }
         }
 
@@ -1706,18 +1702,21 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
                                  offset, nr_segs,
                                  ext4_get_block, NULL);
 
-        /*
-         * Reacquire the handle: ext4_get_block() can restart the transaction
-         */
-        handle = ext4_journal_current_handle();
-
-out_stop:
-        if (handle) {
+        if (orphan) {
                 int err;
 
-                if (orphan && inode->i_nlink)
+                /* Credits for sb + inode write */
+                handle = ext4_journal_start(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
+                         * the write failed... */
+                        ret = PTR_ERR(handle);
+                        goto out;
+                }
+                if (inode->i_nlink)
                         ext4_orphan_del(handle, inode);
-                if (orphan && ret > 0) {
+                if (ret > 0) {
                         loff_t end = offset + ret;
                         if (end > inode->i_size) {
                                 ei->i_disksize = end;
@@ -2758,13 +2757,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
                 ei->i_data[block] = raw_inode->i_block[block];
         INIT_LIST_HEAD(&ei->i_orphan);
 
-        if (inode->i_ino >= EXT4_FIRST_INO(inode->i_sb) + 1 &&
-            EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {
-                /*
-                 * When mke2fs creates big inodes it does not zero out
-                 * the unused bytes above EXT4_GOOD_OLD_INODE_SIZE,
+        if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {