Commit 2ec99ada authored by Ziye Yang's avatar Ziye Yang Committed by Jim Harris
Browse files

nvmf/tcp: fix the state machine issue if data is already read. 



Since we use big buffer to read the data, so the incoming
data may already be read when the req is waiting for the buffer.
So if we use the orginalstatement machine, there will be no
read event will be generated again.

The quick solution is to restore the original code, since
for req which has incapsule data, we not need to wait for the
buffer from the shared buffer pool.

Signed-off-by: default avatarZiye Yang <ziye.yang@intel.com>
Change-Id: Ib195d57cc2969235203c34664115c3322d1c9eae
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/472047


Reviewed-by: default avatarShuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: default avatarJim Harris <james.r.harris@intel.com>
Tested-by: default avatarSPDK CI Jenkins <sys_sgci@intel.com>
parent faedb249

lib/nvmf/tcp.c

+2 −2
Original line number Diff line number Diff line
@@ -2546,7 +2546,7 @@ spdk_nvmf_tcp_req_process(struct spdk_nvmf_tcp_transport *ttransport, 


			assert(tcp_req->req.xfer != SPDK_NVME_DATA_NONE);



			if (&tcp_req->req != STAILQ_FIRST(&group->pending_buf_queue)) {

			if (!tcp_req->has_incapsule_data && (&tcp_req->req != STAILQ_FIRST(&group->pending_buf_queue))) {

				SPDK_DEBUGLOG(SPDK_LOG_NVMF_TCP,

					      "Not the first element to wait for the buf for tcp_req(%p) on tqpair=%p\n",

					      tcp_req, tqpair);
@@ -2572,7 +2572,7 @@ spdk_nvmf_tcp_req_process(struct spdk_nvmf_tcp_transport *ttransport, 
				break;

			}



			STAILQ_REMOVE_HEAD(&group->pending_buf_queue, buf_link);

			STAILQ_REMOVE(&group->pending_buf_queue, &tcp_req->req, spdk_nvmf_request, buf_link);



			/* If data is transferring from host to controller, we need to do a transfer from the host. */

			if (tcp_req->req.xfer == SPDK_NVME_DATA_HOST_TO_CONTROLLER) {