Re: [PATCH 3/3] f2fs: fix roll-forward missing scenarios

[Huang Ying <ying.huang@intel.com>]

On Sat, 2014-09-20 at 20:36 -0700, Jaegeuk Kim wrote:
> On Sun, Sep 21, 2014 at 07:22:32AM +0800, Huang Ying wrote:
> > On Sat, 2014-09-20 at 09:23 -0700, Jaegeuk Kim wrote:
> > > On Thu, Sep 18, 2014 at 09:04:11PM +0800, huang ying wrote:
> > > > On Thu, Sep 18, 2014 at 1:51 PM, Jaegeuk Kim <jaegeuk@kernel.org> wrote:
> > > > 
> > > > > We can summarize the roll forward recovery scenarios as follows.
> > > > >
> > > > > [Term] F: fsync_mark, D: dentry_mark
> > > > >
> > > > > 1. inode(x) | CP | inode(x) | dnode(F)
> > > > > -> Update the latest inode(x).
> > > > >
> > > > > 2. inode(x) | CP | inode(F) | dnode(F)
> > > > > -> No problem.
> > > > >
> > > > > 3. inode(x) | CP | dnode(F) | inode(x)
> > > > > -> Recover to the latest dnode(F), and drop the last inode(x)
> > > > >
> > > > > 4. inode(x) | CP | dnode(F) | inode(F)
> > > > > -> No problem.
> > > > >
> > > > > 5. CP | inode(x) | dnode(F)
> > > > > -> The inode(DF) was missing. Should drop this dnode(F).
> > > > >
> > > > > 6. CP | inode(DF) | dnode(F)
> > > > > -> No problem.
> > > > >
> > > > > 7. CP | dnode(F) | inode(DF)
> > > > > -> If f2fs_iget fails, then goto next to find inode(DF).
> > > > >
> > > > > 8. CP | dnode(F) | inode(x)
> > > > > -> If f2fs_iget fails, then goto next to find inode(DF).
> > > > >    But it will fail due to no inode(DF).
> > > > >
> > > > > So, this patch adds some missing points such as #1, #5, #7, and #8.
> > > > >
> > > > > Signed-off-by: Huang Ying <ying.huang@intel.com>
> > > > > Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
> > > > > ---
> > > > >  fs/f2fs/recovery.c | 71
> > > > > +++++++++++++++++++++++++++++++++++++++++++++---------
> > > > >  1 file changed, 60 insertions(+), 11 deletions(-)
> > > > >
> > > > > diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c
> > > > > index 36d4f73..a4eb978 100644
> > > > > --- a/fs/f2fs/recovery.c
> > > > > +++ b/fs/f2fs/recovery.c
> > > > > @@ -14,6 +14,37 @@
> > > > >  #include "node.h"
> > > > >  #include "segment.h"
> > > > >
> > > > > +/*
> > > > > + * Roll forward recovery scenarios.
> > > > > + *
> > > > > + * [Term] F: fsync_mark, D: dentry_mark
> > > > > + *
> > > > > + * 1. inode(x) | CP | inode(x) | dnode(F)
> > > > > + * -> Update the latest inode(x).
> > > > > + *
> > > > > + * 2. inode(x) | CP | inode(F) | dnode(F)
> > > > > + * -> No problem.
> > > > > + *
> > > > > + * 3. inode(x) | CP | dnode(F) | inode(x)
> > > > > + * -> Recover to the latest dnode(F), and drop the last inode(x)
> > > > > + *
> > > > > + * 4. inode(x) | CP | dnode(F) | inode(F)
> > > > > + * -> No problem.
> > > > > + *
> > > > > + * 5. CP | inode(x) | dnode(F)
> > > > > + * -> The inode(DF) was missing. Should drop this dnode(F).
> > > > > + *
> > > > > + * 6. CP | inode(DF) | dnode(F)
> > > > > + * -> No problem.
> > > > > + *
> > > > > + * 7. CP | dnode(F) | inode(DF)
> > > > > + * -> If f2fs_iget fails, then goto next to find inode(DF).
> > > > > + *
> > > > > + * 8. CP | dnode(F) | inode(x)
> > > > > + * -> If f2fs_iget fails, then goto next to find inode(DF).
> > > > > + *    But it will fail due to no inode(DF).
> > > > > + */
> > > > > +
> > > > >  static struct kmem_cache *fsync_entry_slab;
> > > > >
> > > > >  bool space_for_roll_forward(struct f2fs_sb_info *sbi)
> > > > > @@ -110,27 +141,32 @@ out:
> > > > >         return err;
> > > > >  }
> > > > >
> > > > > -static int recover_inode(struct inode *inode, struct page *node_page)
> > > > > +static void __recover_inode(struct inode *inode, struct page *page)
> > > > >  {
> > > > > -       struct f2fs_inode *raw_inode = F2FS_INODE(node_page);
> > > > > +       struct f2fs_inode *raw = F2FS_INODE(page);
> > > > > +
> > > > > +       inode->i_mode = le16_to_cpu(raw->i_mode);
> > > > > +       i_size_write(inode, le64_to_cpu(raw->i_size));
> > > > > +       inode->i_atime.tv_sec = le64_to_cpu(raw->i_mtime);
> > > > > +       inode->i_ctime.tv_sec = le64_to_cpu(raw->i_ctime);
> > > > > +       inode->i_mtime.tv_sec = le64_to_cpu(raw->i_mtime);
> > > > > +       inode->i_atime.tv_nsec = le32_to_cpu(raw->i_mtime_nsec);
> > > > > +       inode->i_ctime.tv_nsec = le32_to_cpu(raw->i_ctime_nsec);
> > > > > +       inode->i_mtime.tv_nsec = le32_to_cpu(raw->i_mtime_nsec);
> > > > > +}
> > > > >
> > > > > +static int recover_inode(struct inode *inode, struct page *node_page)
> > > > > +{
> > > > >         if (!IS_INODE(node_page))
> > > > >                 return 0;
> > > > >
> > > > > -       inode->i_mode = le16_to_cpu(raw_inode->i_mode);
> > > > > -       i_size_write(inode, le64_to_cpu(raw_inode->i_size));
> > > > > -       inode->i_atime.tv_sec = le64_to_cpu(raw_inode->i_mtime);
> > > > > -       inode->i_ctime.tv_sec = le64_to_cpu(raw_inode->i_ctime);
> > > > > -       inode->i_mtime.tv_sec = le64_to_cpu(raw_inode->i_mtime);
> > > > > -       inode->i_atime.tv_nsec = le32_to_cpu(raw_inode->i_mtime_nsec);
> > > > > -       inode->i_ctime.tv_nsec = le32_to_cpu(raw_inode->i_ctime_nsec);
> > > > > -       inode->i_mtime.tv_nsec = le32_to_cpu(raw_inode->i_mtime_nsec);
> > > > > +       __recover_inode(inode, node_page);
> > > > >
> > > > >         if (is_dent_dnode(node_page))
> > > > >                 return recover_dentry(node_page, inode);
> > > > >
> > > > >         f2fs_msg(inode->i_sb, KERN_NOTICE, "recover_inode: ino = %x, name
> > > > > = %s",
> > > > > -                       ino_of_node(node_page), raw_inode->i_name);
> > > > > +                       ino_of_node(node_page),
> > > > > F2FS_INODE(node_page)->i_name);
> > > > >         return 0;
> > > > >  }
> > > > >
> > > > > @@ -183,10 +219,16 @@ static int find_fsync_dnodes(struct f2fs_sb_info
> > > > > *sbi, struct list_head *head)
> > > > >                                 break;
> > > > >                         }
> > > > >
> > > > > +                       /*
> > > > > +                        * CP | dnode(F) | inode(DF)
> > > > > +                        * For this case, we should not give up now.
> > > > > +                        */
> > > > >                         entry->inode = f2fs_iget(sbi->sb,
> > > > > ino_of_node(page));
> > > > >                         if (IS_ERR(entry->inode)) {
> > > > >                                 err = PTR_ERR(entry->inode);
> > > > >                                 kmem_cache_free(fsync_entry_slab, entry);
> > > > > +                               if (err == -ENOENT)
> > > > > +                                       goto next;
> > > > >                                 break;
> > > > >                         }
> > > > >                         list_add_tail(&entry->list, head);
> > > > > @@ -423,6 +465,13 @@ static int recover_data(struct f2fs_sb_info *sbi,
> > > > >                 entry = get_fsync_inode(head, ino_of_node(page));
> > > > >                 if (!entry)
> > > > >                         goto next;
> > > > > +               /*
> > > > > +                * inode(x) | CP | inode(x) | dnode(F)
> > > > > +                * In this case, we can lose the latest inode(x).
> > > > > +                * So, call __recover_inode for the inode update.
> > > > > +                */
> > > > >
> > > > 
> > > > With 2/3, because both IS_CHECKPOINTED and HAS_FSYNCED_INODE flag are unset
> > > > for inode, we will append a inode(F).
> > > 
> > > No, inode(F) is not appended.
> > > Please check the fsync rule #1.
> > 
> > >From implementation of need_inode_block_update, we do not append an
> > inode(F) or inode(DF), only if:
> > 
> >   get_nat_flag(e, HAS_LAST_FSYNC) &&
> >           (get_nat_flag(e, IS_CHECKPOINTED) ||
> >            get_nat_flag(e, HAS_FSYNCED_INODE)))
> > 
> > e is nat entry for the inode.
> > 
> > For inode(x) | CP | inode(x) | dnode(F)
> > 
> > We have:
> > 
> > HAS_LAST_FSYNC:    true
> > IS_CHECKPOINTED:   false
> 
> IS_CHECKPOINTED: true
> 
> I changed the rule for CHECKPOINTED too.

Sorry for the noise.  I did not notice that.

Best Regards,
Huang, Ying

> > HAS_FSYNCED_INODE: false
> > 
> > So we will append a inode(DF) here.
> > 
> > > > So we do not need to call
> > > > __recover_inode here?
> > 
> > Thought again, __recover_inode here may be helpful here for the
> > following situation:
> > 
> > inode(x) | CP | inode(x) | dnode(F) | inode(DF)
> >                                     |
> >                                     v
> >                                Sudden power off
> > 
> > That is, sudden power off before writing inode(DF).
> > 
> > So I think we should keep the code, but maybe change the comments?
> > 
> > Best Regards,
> > Huang, Ying
> > 
> > > > 
> > > > 
> > > > > +               if (IS_INODE(page))
> > > > > +                       __recover_inode(entry->inode, page);
> > > > >
> > > > >                 err = do_recover_data(sbi, entry->inode, page, blkaddr);
> > > > >                 if (err) {
> > > > > --
> > > > >