NFSv4.1: Fix a race in nfs4_proc_layoutget

If the task calling layoutget is signalled, then it is possible for the
calls to nfs4_sequence_free_slot() and nfs4_layoutget_prepare() to race,
in which case we leak a slot.
The fix is to move the call to nfs4_sequence_free_slot() into the
nfs4_layoutget_release() so that it gets called at task teardown time.

Fixes: 2e80dbe7ac51 ("NFSv4.1: Close callback races for OPEN, LAYOUTGET...")
Cc: stable@vger.kernel.org # v4.8+
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 4e43fd1270c410125d9b91c779b4e9e8749839fc..dbfa18900e25a38a0998a2d429644a860c559ac2 100644 (file)
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -8417,6 +8417,7 @@ static void nfs4_layoutget_release(void *calldata)
        size_t max_pages = max_response_pages(server);
 
        dprintk("--> %s\n", __func__);
+       nfs4_sequence_free_slot(&lgp->res.seq_res);
        nfs4_free_pages(lgp->args.layout.pages, max_pages);
        pnfs_put_layout_hdr(NFS_I(inode)->layout);
        put_nfs_open_context(lgp->args.ctx);
@@ -8491,7 +8492,6 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout, gfp_t gfp_flags)
        /* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */
        if (status == 0 && lgp->res.layoutp->len)
                lseg = pnfs_layout_process(lgp);
-       nfs4_sequence_free_slot(&lgp->res.seq_res);
        rpc_put_task(task);
        dprintk("<-- %s status=%d\n", __func__, status);
        if (status)