writeback: introduce .tagged_writepages for the WB_SYNC_NONE sync stage

sync(2) is performed in two stages: the WB_SYNC_NONE sync and the
WB_SYNC_ALL sync. Identify the first stage with .tagged_writepages and
do livelock prevention for it, too.

Jan's commit f446daaea9 ("mm: implement writeback livelock avoidance
using page tagging") is a partial fix in that it only fixed the
WB_SYNC_ALL phase livelock.

Although ext4 is tested to no longer livelock with commit f446daaea9,
it may due to some "redirty_tail() after pages_skipped" effect which
is by no means a guarantee for _all_ the file systems.

Note that writeback_inodes_sb() is called by not only sync(), they are
treated the same because the other callers also need livelock prevention.

Impact:  It changes the order in which pages/inodes are synced to disk.
Now in the WB_SYNC_NONE stage, it won't proceed to write the next inode
until finished with the current inode.

Acked-by: Jan Kara <jack@suse.cz>
CC: Dave Chinner <david@fromorbit.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>

1 files changed, 9 insertions, 8 deletions

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 0f015a0468de..5ed2ce9a28d0 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -36,6 +36,7 @@ struct wb_writeback_work {
         long nr_pages;
         struct super_block *sb;
         enum writeback_sync_modes sync_mode;
+        unsigned int tagged_writepages:1;
         unsigned int for_kupdate:1;
         unsigned int range_cyclic:1;
         unsigned int for_background:1;
@@ -650,6 +651,7 @@ static long wb_writeback(struct bdi_writeback *wb,
 {
         struct writeback_control wbc = {
                 .sync_mode              = work->sync_mode,
+                .tagged_writepages      = work->tagged_writepages,
                 .older_than_this        = NULL,
                 .for_kupdate            = work->for_kupdate,
                 .for_background         = work->for_background,
@@ -657,7 +659,7 @@ static long wb_writeback(struct bdi_writeback *wb,
         };
         unsigned long oldest_jif;
         long wrote = 0;
-        long write_chunk;
+        long write_chunk = MAX_WRITEBACK_PAGES;
         struct inode *inode;
 
         if (wbc.for_kupdate) {
@@ -683,9 +685,7 @@ static long wb_writeback(struct bdi_writeback *wb,
          *                   (quickly) tag currently dirty pages
          *                   (maybe slowly) sync all tagged pages
          */
-        if (wbc.sync_mode == WB_SYNC_NONE)
-                write_chunk = MAX_WRITEBACK_PAGES;
-        else
+        if (wbc.sync_mode == WB_SYNC_ALL || wbc.tagged_writepages)
                 write_chunk = LONG_MAX;
 
         wbc.wb_start = jiffies; /* livelock avoidance */
@@ -1188,10 +1188,11 @@ void writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr)
 {
         DECLARE_COMPLETION_ONSTACK(done);
         struct wb_writeback_work work = {
-                .sb             = sb,
-                .sync_mode      = WB_SYNC_NONE,
-                .done           = &done,
-                .nr_pages       = nr,
+                .sb                     = sb,
+                .sync_mode              = WB_SYNC_NONE,
+                .tagged_writepages      = 1,
+                .done                   = &done,
+                .nr_pages               = nr,
         };
 
         WARN_ON(!rwsem_is_locked(&sb->s_umount));