fs/jffs2/TODO

   1 $Id: TODO,v 1.10 2002/09/09 16:31:21 dwmw2 Exp $
   2
   3  - disable compression in commit_write()?
   4  - fine-tune the allocation / GC thresholds
   5  - chattr support - turning on/off and tuning compression per-inode
   6  - checkpointing (do we need this? scan is quite fast)
   7  - make the scan code populate real inodes so read_inode just after
   8         mount doesn't have to read the flash twice for large files.
   9         Make this a per-inode option, changable with chattr, so you can
  10         decide which inodes should be in-core immediately after mount.
  11  - test, test, test
  12
  13  - NAND flash support:
  14         - flush_wbuf using GC to fill it, don't just pad.
  15         - Deal with write errors. Data don't get lost - we just have to write
  16           the affected node(s) out again somewhere else.
  17         - make fsync flush only if actually required
  18         - make sys_sync() work.
  19         - reboot notifier
  20         - timed flush of old wbuf
  21         - fix magical second arg of jffs2_flush_wbuf(). Split into two or more functions instead.
  22
  23
  24  - Optimisations:
  25    - Stop GC from decompressing and immediately recompressing nodes which could
  26      just be copied intact. (We now keep track of REF_PRISTINE flag. Easy now.)
  27    - Furthermore, in the case where it could be copied intact we don't even need
  28      to call iget() for it -- if we use (raw_node_raw->flash_offset & 2) as a flag
  29      to show a node can be copied intact and it's _not_ in icache, we could just do
  30      it, fix up the next_in_ino list and move on. We would need a way to find out
  31      _whether_ it's in icache though -- if it's in icache we also need to do the
  32      fragment lists, etc. P'raps a flag or pointer in the jffs2_inode_cache could
  33      help. (We have half of this now.)
  34    - Stop keeping name in-core with struct jffs2_full_dirent. If we keep the hash in
  35      the full dirent, we only need to go to the flash in lookup() when we think we've
  36      got a match, and in readdir().
  37    - Doubly-linked next_in_ino list to allow us to free obsoleted raw_node_refs immediately?
  38    - Remove totlen from jffs2_raw_node_ref? Need to have totlen passed into
  39         jffs2_mark_node_obsolete(). Can all callers work it out?
  40    - Remove size from jffs2_raw_node_frag.