When a mechine has hundreds of memory cgroups, and some cgroups
generate more or less dirty pages, but a cgroup of them has lots
of memory pressure and always tries to reclaim dirty page, then it
will trigger all cgroups to writeback, which is less efficient:
1.if the used memory in a memory cgroup reaches its limit,
it is useless to writeback other cgroups.
2.other cgroups can wait more time to merge write request
so replace the full flush with flushing writeback of memory cgroup
whose tasks tries to reclaim memory and trigger writeback, if
nothing is writeback, then start a full flush
Signed-off-by: Li RongQing <lirongqing@xxxxxxxxx>
---
fs/fs-writeback.c | 22 ++++++++++++++++++++++
1 file changed, 22 insertions(+)
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index abafc9bd4622..a27efc4711a9 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -2001,6 +2001,28 @@ void wakeup_flusher_threads(enum wb_reason reason)
if (blk_needs_flush_plug(current))
blk_schedule_flush_plug(current);
+#ifdef CONFIG_CGROUP_WRITEBACK
+ if (reason == WB_REASON_VMSCAN) {
+ int flush = 0;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) {
+ struct bdi_writeback *wb = wb_find_current(bdi);
+
+ if (wb && wb != &bdi->wb) {
+ if (wb_has_dirty_io(wb)) {
+ wb_start_writeback(wb, reason);
+ flush++;
+ }
+ }
+ }
+ rcu_read_unlock();
+
+ if (flush)
+ return;
+ }
+#endif
+
rcu_read_lock();
list_for_each_entry_rcu(bdi, &bdi_list, bdi_list)
__wakeup_flusher_threads_bdi(bdi, reason);
--
2.16.2
