xprtrdma: Fix occasional transport deadlock

Under high I/O workloads, I've noticed that an RPC/RDMA transport
occasionally deadlocks (IOPS goes to zero, and doesn't recover).
Diagnosis shows that the sendctx queue is empty, but when sendctxs
are returned to the queue, the xprt_write_space wake-up never
occurs. The wake-up logic in rpcrdma_sendctx_put_locked is racy.

I noticed that both EMPTY_SCQ and XPRT_WRITE_SPACE are implemented
via an atomic bit. Just one of those is sufficient. Removing
EMPTY_SCQ in favor of the generic bit mechanism makes the deadlock
un-reproducible.

Without EMPTY_SCQ, rpcrdma_buffer::rb_flags is no longer used and
is therefore removed.

Unfortunately this patch does not apply cleanly to stable. If
needed, someone will have to port it and test it.

Fixes: 2fad659209d5 ("xprtrdma: Wait on empty sendctx queue")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>

1 files changed, 12 insertions, 14 deletions

diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c
index 97bfb804b6c6..59b214ba8813 100644
--- a/net/sunrpc/xprtrdma/rpc_rdma.c
+++ b/net/sunrpc/xprtrdma/rpc_rdma.c
@@ -699,22 +699,28 @@ rpcrdma_prepare_send_sges(struct rpcrdma_xprt *r_xprt,
 			  struct rpcrdma_req *req, u32 hdrlen,
 			  struct xdr_buf *xdr, enum rpcrdma_chunktype rtype)
 {
+	int ret;
+
+	ret = -EAGAIN;
 	req->rl_sendctx = rpcrdma_sendctx_get_locked(r_xprt);
 	if (!req->rl_sendctx)
-		return -EAGAIN;
+		goto err;
 	req->rl_sendctx->sc_wr.num_sge = 0;
 	req->rl_sendctx->sc_unmap_count = 0;
 	req->rl_sendctx->sc_req = req;
 	__clear_bit(RPCRDMA_REQ_F_TX_RESOURCES, &req->rl_flags);
 
+	ret = -EIO;
 	if (!rpcrdma_prepare_hdr_sge(r_xprt, req, hdrlen))
-		return -EIO;
-
+		goto err;
 	if (rtype != rpcrdma_areadch)
 		if (!rpcrdma_prepare_msg_sges(r_xprt, req, xdr, rtype))
-			return -EIO;
-
+			goto err;
 	return 0;
+
+err:
+	trace_xprtrdma_prepsend_failed(&req->rl_slot, ret);
+	return ret;
 }
 
 /**
@@ -877,15 +883,7 @@ rpcrdma_marshal_req(struct rpcrdma_xprt *r_xprt, struct rpc_rqst *rqst)
 
 out_err:
 	trace_xprtrdma_marshal_failed(rqst, ret);
-	switch (ret) {
-	case -EAGAIN:
-		xprt_wait_for_buffer_space(rqst->rq_xprt);
-		break;
-	case -ENOBUFS:
-		break;
-	default:
-		r_xprt->rx_stats.failed_marshal_count++;
-	}
+	r_xprt->rx_stats.failed_marshal_count++;
 	return ret;
 }