Skip to content

Commit

Permalink
writeback: sync_inodes_sb() must write out I_DIRTY_TIME inodes and al…
Browse files Browse the repository at this point in the history
…ways call wait_sb_inodes()

e797291 ("writeback: don't issue wb_writeback_work if clean")
updated writeback path to avoid kicking writeback work items if there
are no inodes to be written out; unfortunately, the avoidance logic
was too aggressive and broke sync_inodes_sb().

* sync_inodes_sb() must write out I_DIRTY_TIME inodes but I_DIRTY_TIME
  inodes dont't contribute to bdi/wb_has_dirty_io() tests and were
  being skipped over.

* inodes are taken off wb->b_dirty/io/more_io lists after writeback
  starts on them.  sync_inodes_sb() skipping wait_sb_inodes() when
  bdi_has_dirty_io() breaks it by making it return while writebacks
  are in-flight.

This patch fixes the breakages by

* Removing bdi_has_dirty_io() shortcut from bdi_split_work_to_wbs().
  The callers are already testing the condition.

* Removing bdi_has_dirty_io() shortcut from sync_inodes_sb() so that
  it always calls into bdi_split_work_to_wbs() and wait_sb_inodes().

* Making bdi_split_work_to_wbs() consider the b_dirty_time list for
  WB_SYNC_ALL writebacks.

Kudos to Eryu, Dave and Jan for tracking down the issue.

Signed-off-by: Tejun Heo <tj@kernel.org>
Fixes: e797291 ("writeback: don't issue wb_writeback_work if clean")
Link: http://lkml.kernel.org/g/20150812101204.GE17933@dhcp-13-216.nay.redhat.com
Reported-and-bisected-by: Eryu Guan <eguan@redhat.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jan Kara <jack@suse.com>
Cc: Ted Ts'o <tytso@google.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
  • Loading branch information
htejun authored and axboe committed Aug 25, 2015
1 parent c13dcf9 commit 006a097
Showing 1 changed file with 13 additions and 9 deletions.
22 changes: 13 additions & 9 deletions fs/fs-writeback.c
Expand Up @@ -844,14 +844,15 @@ static void bdi_split_work_to_wbs(struct backing_dev_info *bdi,
struct wb_iter iter;

might_sleep();

if (!bdi_has_dirty_io(bdi))
return;
restart:
rcu_read_lock();
bdi_for_each_wb(wb, bdi, &iter, next_blkcg_id) {
if (!wb_has_dirty_io(wb) ||
(skip_if_busy && writeback_in_progress(wb)))
/* SYNC_ALL writes out I_DIRTY_TIME too */
if (!wb_has_dirty_io(wb) &&
(base_work->sync_mode == WB_SYNC_NONE ||
list_empty(&wb->b_dirty_time)))
continue;
if (skip_if_busy && writeback_in_progress(wb))
continue;

base_work->nr_pages = wb_split_bdi_pages(wb, nr_pages);
Expand Down Expand Up @@ -899,8 +900,7 @@ static void bdi_split_work_to_wbs(struct backing_dev_info *bdi,
{
might_sleep();

if (bdi_has_dirty_io(bdi) &&
(!skip_if_busy || !writeback_in_progress(&bdi->wb))) {
if (!skip_if_busy || !writeback_in_progress(&bdi->wb)) {
base_work->auto_free = 0;
base_work->single_wait = 0;
base_work->single_done = 0;
Expand Down Expand Up @@ -2275,8 +2275,12 @@ void sync_inodes_sb(struct super_block *sb)
};
struct backing_dev_info *bdi = sb->s_bdi;

/* Nothing to do? */
if (!bdi_has_dirty_io(bdi) || bdi == &noop_backing_dev_info)
/*
* Can't skip on !bdi_has_dirty() because we should wait for !dirty
* inodes under writeback and I_DIRTY_TIME inodes ignored by
* bdi_has_dirty() need to be written out too.
*/
if (bdi == &noop_backing_dev_info)
return;
WARN_ON(!rwsem_is_locked(&sb->s_umount));

Expand Down

0 comments on commit 006a097

Please sign in to comment.