On Fri, Sep 18, 2020 at 05:54:47PM -0700, Harshad Shirwadkar wrote: > fs/jbd2/commit.c | 42 ++ > fs/jbd2/journal.c | 119 +++- Why are these changes here instead of the previous commit (jbd2: add fast commit machinery)? > diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c > index ba35ecb18616..dadd9994e74b 100644 > --- a/fs/jbd2/commit.c > +++ b/fs/jbd2/commit.c > @@ -202,6 +202,47 @@ static int journal_submit_inode_data_buffers(struct address_space *mapping, > return ret; > } > > +/* Send all the data buffers related to an inode */ > +int jbd2_submit_inode_data(journal_t *journal, struct jbd2_inode *jinode) > +{ > + struct address_space *mapping; > + loff_t dirty_start; > + loff_t dirty_end; > + int ret; > + > + if (!jinode) > + return 0; > + > + dirty_start = jinode->i_dirty_start; > + dirty_end = jinode->i_dirty_end; > + > + if (!(jinode->i_flags & JI_WRITE_DATA)) > + return 0; > + > + dirty_start = jinode->i_dirty_start; > + dirty_end = jinode->i_dirty_end; Why is dirty_start and dirty_end initialized twice? Also, this is going to conflcit with Mauricio's data=journal patches, which you'll notice when you rebase these patches on the current dev branch. (The dev branch temporarily had your v9 patches merged in, so we could get the test bots to comment on your changes, but I've since pulled the fc patches back out.) - Ted