aboutsummaryrefslogtreecommitdiff
path: root/fs/jffs2/TODO
diff options
context:
space:
mode:
Diffstat (limited to 'fs/jffs2/TODO')
-rw-r--r--fs/jffs2/TODO38
1 files changed, 19 insertions, 19 deletions
diff --git a/fs/jffs2/TODO b/fs/jffs2/TODO
index 2bff82fd221..d0e23b26fa5 100644
--- a/fs/jffs2/TODO
+++ b/fs/jffs2/TODO
@@ -1,5 +1,11 @@
-$Id: TODO,v 1.10 2002/09/09 16:31:21 dwmw2 Exp $
+$Id: TODO,v 1.18 2005/09/22 11:24:56 dedekind Exp $
+ - support asynchronous operation -- add a per-fs 'reserved_space' count,
+ let each outstanding write reserve the _maximum_ amount of physical
+ space it could take. Let GC flush the outstanding writes because the
+ reservations will necessarily be pessimistic. With this we could even
+ do shared writable mmap, if we can have a fs hook for do_wp_page() to
+ make the reservation.
- disable compression in commit_write()?
- fine-tune the allocation / GC thresholds
- chattr support - turning on/off and tuning compression per-inode
@@ -11,26 +17,15 @@ $Id: TODO,v 1.10 2002/09/09 16:31:21 dwmw2 Exp $
- test, test, test
- NAND flash support:
- - flush_wbuf using GC to fill it, don't just pad.
- - Deal with write errors. Data don't get lost - we just have to write
- the affected node(s) out again somewhere else.
- - make fsync flush only if actually required
- - make sys_sync() work.
- - reboot notifier
- - timed flush of old wbuf
- - fix magical second arg of jffs2_flush_wbuf(). Split into two or more functions instead.
-
+ - almost done :)
+ - use bad block check instead of the hardwired byte check
- Optimisations:
- - Stop GC from decompressing and immediately recompressing nodes which could
- just be copied intact. (We now keep track of REF_PRISTINE flag. Easy now.)
- - Furthermore, in the case where it could be copied intact we don't even need
- to call iget() for it -- if we use (raw_node_raw->flash_offset & 2) as a flag
- to show a node can be copied intact and it's _not_ in icache, we could just do
- it, fix up the next_in_ino list and move on. We would need a way to find out
- _whether_ it's in icache though -- if it's in icache we also need to do the
- fragment lists, etc. P'raps a flag or pointer in the jffs2_inode_cache could
- help. (We have half of this now.)
+ - Split writes so they go to two separate blocks rather than just c->nextblock.
+ By writing _new_ nodes to one block, and garbage-collected REF_PRISTINE
+ nodes to a different one, we can separate clean nodes from those which
+ are likely to become dirty, and end up with blocks which are each far
+ closer to 100% or 0% clean, hence speeding up later GC progress dramatically.
- Stop keeping name in-core with struct jffs2_full_dirent. If we keep the hash in
the full dirent, we only need to go to the flash in lookup() when we think we've
got a match, and in readdir().
@@ -38,3 +33,8 @@ $Id: TODO,v 1.10 2002/09/09 16:31:21 dwmw2 Exp $
- Remove totlen from jffs2_raw_node_ref? Need to have totlen passed into
jffs2_mark_node_obsolete(). Can all callers work it out?
- Remove size from jffs2_raw_node_frag.
+
+dedekind:
+1. __jffs2_flush_wbuf() has a strange 'pad' parameter. Eliminate.
+2. get_sb()->build_fs()->scan() path... Why get_sb() removes scan()'s crap in
+ case of failure? scan() does not clean everything. Fix.