2 files changed, 58 insertions, 10 deletions

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 7923336ecf94..24518b57733e 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2043,11 +2043,12 @@ static int ext4_bh_unmapped_or_delay(handle_t *handle, struct buffer_head *bh)
         return !buffer_mapped(bh) || buffer_delay(bh);
 }
 
-/* FIXME!! only support data=writeback mode */
 /*
  * get called vi ext4_da_writepages after taking page lock
  * We may end up doing block allocation here in case
  * mpage_da_map_blocks failed to allocate blocks.
+ *
+ * We also get called via journal_submit_inode_data_buffers
  */
 static int ext4_da_writepage(struct page *page,
                                 struct writeback_control *wbc)
@@ -2066,6 +2067,7 @@ static int ext4_da_writepage(struct page *page,
                  * ext4_da_writepages() but directly (shrink_page_list).
                  * We cannot easily start a transaction here so we just skip
                  * writing the page in case we would have to do so.
+                 * We reach here also via journal_submit_inode_data_buffers
                  */
                 size = i_size_read(inode);
 
@@ -2081,8 +2083,11 @@ static int ext4_da_writepage(struct page *page,
                          * We can't do block allocation under
                          * page lock without a handle . So redirty
                          * the page and return
+                         * We may reach here when we do a journal commit
+                         * via journal_submit_inode_data_buffers.
+                         * If we don't have mapping block we just ignore
+                         * them
                          */
-                        BUG_ON(wbc->sync_mode != WB_SYNC_NONE);
                         redirty_page_for_writepage(wbc, page);
                         unlock_page(page);
                         return 0;
@@ -2097,7 +2102,6 @@ static int ext4_da_writepage(struct page *page,
         return ret;
 }
 
-
 /*
  * For now just follow the DIO way to estimate the max credits
  * needed to write out EXT4_MAX_WRITEBACK_PAGES.
@@ -2130,7 +2134,7 @@ static int ext4_da_writepages(struct address_space *mapping,
                 return 0;
 
         /*
-         *  Estimate the worse case needed credits to write out
+         * Estimate the worse case needed credits to write out
          * EXT4_MAX_BUF_BLOCKS pages
          */
         needed_blocks = EXT4_MAX_WRITEBACK_CREDITS;
@@ -2152,6 +2156,19 @@ static int ext4_da_writepages(struct address_space *mapping,
                         ret = PTR_ERR(handle);
                         goto out_writepages;
                 }
+                if (ext4_should_order_data(inode)) {
+                        /*
+                         * With ordered mode we need to add
+                         * the inode to the journal handle
+                         * when we do block allocation.
+                         */
+                        ret = ext4_jbd2_file_inode(handle, inode);
+                        if (ret) {
+                                ext4_journal_stop(handle);
+                                goto out_writepages;
+                        }
+
+                }
                 /*
                  * set the max dirty pages could be write at a time
                  * to fit into the reserved transaction credits
@@ -2735,7 +2752,10 @@ static const struct address_space_operations ext4_da_aops = {
 
 void ext4_set_aops(struct inode *inode)
 {
-        if (ext4_should_order_data(inode))
+        if (ext4_should_order_data(inode) &&
+                test_opt(inode->i_sb, DELALLOC))
+                inode->i_mapping->a_ops = &ext4_da_aops;
+        else if (ext4_should_order_data(inode))
                 inode->i_mapping->a_ops = &ext4_ordered_aops;
         else if (ext4_should_writeback_data(inode) &&
                  test_opt(inode->i_sb, DELALLOC))
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 483183d15ed5..f8b3be873226 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -22,6 +22,8 @@
 #include <linux/pagemap.h>
 #include <linux/jiffies.h>
 #include <linux/crc32.h>
+#include <linux/writeback.h>
+#include <linux/backing-dev.h>
 
 /*
  * Default IO end handler for temporary BJ_IO buffer_heads.
@@ -185,6 +187,27 @@ static int journal_wait_on_commit_record(struct buffer_head *bh)
 }
 
 /*
+ * write the filemap data using writepage() address_space_operations.
+ * We don't do block allocation here even for delalloc. We don't
+ * use writepages() because with dealyed allocation we may be doing
+ * block allocation in writepages().
+ */
+static int journal_submit_inode_data_buffers(struct address_space *mapping)
+{
+        int ret;
+        struct writeback_control wbc = {
+                .sync_mode =  WB_SYNC_ALL,
+                .nr_to_write = mapping->nrpages * 2,
+                .range_start = 0,
+                .range_end = i_size_read(mapping->host),
+                .for_writepages = 1,
+        };
+
+        ret = generic_writepages(mapping, &wbc);
+        return ret;
+}
+
+/*
  * Submit all the data buffers of inode associated with the transaction to
  * disk.
  *
@@ -192,7 +215,7 @@ static int journal_wait_on_commit_record(struct buffer_head *bh)
  * our inode list. We use JI_COMMIT_RUNNING flag to protect inode we currently
  * operate on from being released while we write out pages.
  */
-static int journal_submit_inode_data_buffers(journal_t *journal,
+static int journal_submit_data_buffers(journal_t *journal,
                 transaction_t *commit_transaction)
 {
         struct jbd2_inode *jinode;
@@ -204,8 +227,13 @@ static int journal_submit_inode_data_buffers(journal_t *journal,
                 mapping = jinode->i_vfs_inode->i_mapping;
                 jinode->i_flags |= JI_COMMIT_RUNNING;
                 spin_unlock(&journal->j_list_lock);
-                err = filemap_fdatawrite_range(mapping, 0,
-                                        i_size_read(jinode->i_vfs_inode));
+                /*
+                 * submit the inode data buffers. We use writepage
+                 * instead of writepages. Because writepages can do
+                 * block allocation  with delalloc. We need to write
+                 * only allocated blocks here.
+                 */
+                err = journal_submit_inode_data_buffers(mapping);
                 if (!ret)
                         ret = err;
                 spin_lock(&journal->j_list_lock);
@@ -228,7 +256,7 @@ static int journal_finish_inode_data_buffers(journal_t *journal,
         struct jbd2_inode *jinode, *next_i;
         int err, ret = 0;
 
-        /* For locking, see the comment in journal_submit_inode_data_buffers() */
+        /* For locking, see the comment in journal_submit_data_buffers() */
         spin_lock(&journal->j_list_lock);
         list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
                 jinode->i_flags |= JI_COMMIT_RUNNING;
@@ -431,7 +459,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
          * Now start flushing things to disk, in the order they appear
          * on the transaction lists.  Data blocks go first.
          */
-        err = journal_submit_inode_data_buffers(journal, commit_transaction);
+        err = journal_submit_data_buffers(journal, commit_transaction);
         if (err)
                 jbd2_journal_abort(journal, err);