writeback: bdi_writeback iteration must not skip dying ones

bdi_for_each_wb() is used in several places to wake up or issue
writeback work items to all wb's (bdi_writeback's) on a given bdi.
The iteration is performed by walking bdi->cgwb_tree; however, the
tree only indexes wb's which are currently active.

For example, when a memcg gets associated with a different blkcg, the
old wb is removed from the tree so that the new one can be indexed.
The old wb starts dying from then on but will linger till all its
inodes are drained.  As these dying wb's may still host dirty inodes,
writeback operations which affect all wb's must include them.
bdi_for_each_wb() skipping dying wb's led to sync(2) missing and
failing to sync the inodes belonging to those wb's.

This patch adds a RCU protected @bdi->wb_list which lists all wb's
beloinging to that bdi.  wb's are added on creation and removed on
release rather than on the start of destruction.  bdi_for_each_wb()
usages are replaced with list_for_each[_continue]_rcu() iterations
over @bdi->wb_list and bdi_for_each_wb() and its helpers are removed.

v2: Updated as per Jan.  last_wb ref leak in bdi_split_work_to_wbs()
    fixed and unnecessary list head severing in cgwb_bdi_destroy()
    removed.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-and-tested-by: Artem Bityutskiy <dedekind1@gmail.com>
Fixes: ebe41ab0c79d ("writeback: implement bdi_for_each_wb()")
Link: http://lkml.kernel.org/g/1443012552.19983.209.camel@gmail.com
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: Jens Axboe <axboe@fb.com>

2 files changed, 14 insertions, 3 deletions

diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index 2df8ddcb0ca0..e92d77937fd3 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -480,6 +480,10 @@ static void cgwb_release_workfn(struct work_struct *work)
 						release_work);
 	struct backing_dev_info *bdi = wb->bdi;
 
+	spin_lock_irq(&cgwb_lock);
+	list_del_rcu(&wb->bdi_node);
+	spin_unlock_irq(&cgwb_lock);
+
 	wb_shutdown(wb);
 
 	css_put(wb->memcg_css);
@@ -575,6 +579,7 @@ static int cgwb_create(struct backing_dev_info *bdi,
 		ret = radix_tree_insert(&bdi->cgwb_tree, memcg_css->id, wb);
 		if (!ret) {
 			atomic_inc(&bdi->usage_cnt);
+			list_add_tail_rcu(&wb->bdi_node, &bdi->wb_list);
 			list_add(&wb->memcg_node, memcg_cgwb_list);
 			list_add(&wb->blkcg_node, blkcg_cgwb_list);
 			css_get(memcg_css);
@@ -764,15 +769,22 @@ static void cgwb_bdi_destroy(struct backing_dev_info *bdi) { }
 
 int bdi_init(struct backing_dev_info *bdi)
 {
+	int ret;
+
 	bdi->dev = NULL;
 
 	bdi->min_ratio = 0;
 	bdi->max_ratio = 100;
 	bdi->max_prop_frac = FPROP_FRAC_BASE;
 	INIT_LIST_HEAD(&bdi->bdi_list);
+	INIT_LIST_HEAD(&bdi->wb_list);
 	init_waitqueue_head(&bdi->wb_waitq);
 
-	return cgwb_bdi_init(bdi);
+	ret = cgwb_bdi_init(bdi);
+
+	list_add_tail_rcu(&bdi->wb.bdi_node, &bdi->wb_list);
+
+	return ret;
 }
 EXPORT_SYMBOL(bdi_init);
 
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index 902e5f215e57..0f1a94e9f351 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -1956,7 +1956,6 @@ void laptop_mode_timer_fn(unsigned long data)
 	int nr_pages = global_page_state(NR_FILE_DIRTY) +
 		global_page_state(NR_UNSTABLE_NFS);
 	struct bdi_writeback *wb;
-	struct wb_iter iter;
 
 	/*
 	 * We want to write everything out, not just down to the dirty
@@ -1966,7 +1965,7 @@ void laptop_mode_timer_fn(unsigned long data)
 		return;
 
 	rcu_read_lock();
-	bdi_for_each_wb(wb, &q->backing_dev_info, &iter, 0)
+	list_for_each_entry_rcu(wb, &q->backing_dev_info.wb_list, bdi_node)
 		if (wb_has_dirty_io(wb))
 			wb_start_writeback(wb, nr_pages, true,
 					   WB_REASON_LAPTOP_TIMER);