NFS: Don't clear desc->pg_moreio in nfs_do_recoalesce()
authorTrond Myklebust <trond.myklebust@primarydata.com>
Fri, 24 Jul 2015 16:31:33 +0000 (12:31 -0400)
committerTrond Myklebust <trond.myklebust@primarydata.com>
Mon, 27 Jul 2015 14:33:12 +0000 (10:33 -0400)
Recoalescing does not affect whether or not we've already sent off
I/O, and doing so means that we end up sending a bunch of synchronous
for cases where we actually need to be using unstable writes.

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

index 8ea5920fb777576486bee23c6cd029d41f491aa0..4984bbe55ff1eed1623df2196bc0de0de41a4304 100644 (file)
@@ -1100,8 +1100,6 @@ static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc)
                mirror->pg_base = 0;
                mirror->pg_recoalesce = 0;
 
-               desc->pg_moreio = 0;
-
                while (!list_empty(&head)) {
                        struct nfs_page *req;