Andrew,
I'd like to propose a cleaner way of using the s_dirty, s_io, s_more_io
queues for the writeback of dirty inodes. The basic idea is to clearly
define the function of the queues, especially to decouple s_diry from
s_io/s_more_io. The details are in the changelog of patch 2.
The patches are some cleanups on top of Andrew's s_dirty time-ordering
patches and Ken's s_more_io patch:
[PATCH 1/4] writeback: check time-ordering of s_io and s_more_io
[PATCH 2/4] writeback: 3-queue based writeback schedule
[PATCH 3/4] writeback: function renames and cleanups
[PATCH 4/4] writeback: fix ntfs with sb_has_dirty_inodes()
fs/fs-writeback.c | 196 +++++++++++++++++++++++--------------------
fs/ntfs/super.c | 4
include/linux/fs.h | 1
3 files changed, 108 insertions(+), 93 deletions(-)
Note that the patches need rework for inserting into the right place of
your queue of -mm patches. I'll take care of it in the next take.
Thank you,
Fengguang
--