NFS: Flush reclaim writes using FLUSH_COND_STABLE
authorTrond Myklebust <trond.myklebust@primarydata.com>
Mon, 28 Dec 2015 18:34:59 +0000 (13:34 -0500)
committerTrond Myklebust <trond.myklebust@primarydata.com>
Mon, 28 Dec 2015 18:34:59 +0000 (13:34 -0500)
If there are already writes queued up for commit, then don't flush
just this page even if it is a reclaim issue.

Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
fs/nfs/write.c

index 7a4fe7d82e65e4463d290131998e8b429971bb53..1ea35f88eadbc64b5da229c6b5b10d4353169843 100644 (file)
@@ -244,7 +244,7 @@ static int wb_priority(struct writeback_control *wbc)
 {
        int ret = 0;
        if (wbc->for_reclaim)
-               return FLUSH_HIGHPRI | FLUSH_STABLE;
+               return FLUSH_HIGHPRI | FLUSH_COND_STABLE;
        if (wbc->sync_mode == WB_SYNC_ALL)
                ret = FLUSH_COND_STABLE;
        return ret;