diff --git a/drivers/net/wireless/iwlwifi/dvm/sta.c b/drivers/net/wireless/iwlwifi/dvm/sta.c
index b29b798f7550..fe36a38f3505 100644
--- a/drivers/net/wireless/iwlwifi/dvm/sta.c
+++ b/drivers/net/wireless/iwlwifi/dvm/sta.c
@@ -150,7 +150,7 @@ int iwl_send_add_sta(struct iwl_priv *priv,
 		       sta_id, sta->sta.addr, flags & CMD_ASYNC ?  "a" : "");
 
 	if (!(flags & CMD_ASYNC)) {
-		cmd.flags |= CMD_WANT_SKB;
+		cmd.flags |= CMD_WANT_SKB | CMD_WANT_HCMD;
 		might_sleep();
 	}
 
diff --git a/drivers/net/wireless/iwlwifi/iwl-trans.h b/drivers/net/wireless/iwlwifi/iwl-trans.h
index 8ac72a6ef227..ff1154232885 100644
--- a/drivers/net/wireless/iwlwifi/iwl-trans.h
+++ b/drivers/net/wireless/iwlwifi/iwl-trans.h
@@ -184,14 +184,20 @@ struct iwl_rx_packet {
  * @CMD_SYNC: The caller will be stalled until the fw responds to the command
  * @CMD_ASYNC: Return right away and don't want for the response
  * @CMD_WANT_SKB: valid only with CMD_SYNC. The caller needs the buffer of the
- *	response.
+ *	response. The caller needs to call iwl_free_resp when done.
+ * @CMD_WANT_HCMD: The caller needs to get the HCMD that was sent in the
+ *	response handler. Chunks flagged by %IWL_HCMD_DFL_NOCOPY won't be
+ *	copied. The pointer passed to the response handler is in the transport
+ *	ownership and don't need to be freed by the op_mode. This also means
+ *	that the pointer is invalidated after the op_mode's handler returns.
  * @CMD_ON_DEMAND: This command is sent by the test mode pipe.
  */
 enum CMD_MODE {
 	CMD_SYNC = 0,
 	CMD_ASYNC = BIT(0),
 	CMD_WANT_SKB = BIT(1),
-	CMD_ON_DEMAND = BIT(2),
+	CMD_WANT_HCMD = BIT(2),
+	CMD_ON_DEMAND = BIT(3),
 };
 
 #define DEF_CMD_PAYLOAD_SIZE 320
diff --git a/drivers/net/wireless/iwlwifi/pcie/internal.h b/drivers/net/wireless/iwlwifi/pcie/internal.h
index d9694c58208c..3ef8d5adc991 100644
--- a/drivers/net/wireless/iwlwifi/pcie/internal.h
+++ b/drivers/net/wireless/iwlwifi/pcie/internal.h
@@ -184,6 +184,7 @@ struct iwl_queue {
 
 struct iwl_pcie_tx_queue_entry {
 	struct iwl_device_cmd *cmd;
+	struct iwl_device_cmd *copy_cmd;
 	struct sk_buff *skb;
 	struct iwl_cmd_meta meta;
 };
diff --git a/drivers/net/wireless/iwlwifi/pcie/rx.c b/drivers/net/wireless/iwlwifi/pcie/rx.c
index 39a6ca1f009c..d80604a2bb1a 100644
--- a/drivers/net/wireless/iwlwifi/pcie/rx.c
+++ b/drivers/net/wireless/iwlwifi/pcie/rx.c
@@ -421,13 +421,23 @@ static void iwl_rx_handle_rxbuf(struct iwl_trans *trans,
 		index = SEQ_TO_INDEX(sequence);
 		cmd_index = get_cmd_index(&txq->q, index);
 
-		if (reclaim)
-			cmd = txq->entries[cmd_index].cmd;
-		else
+		if (reclaim) {
+			struct iwl_pcie_tx_queue_entry *ent;
+			ent = &txq->entries[cmd_index];
+			cmd = ent->copy_cmd;
+			WARN_ON_ONCE(!cmd && ent->meta.flags & CMD_WANT_HCMD);
+		} else {
 			cmd = NULL;
+		}
 
 		err = iwl_op_mode_rx(trans->op_mode, &rxcb, cmd);
 
+		if (reclaim) {
+			/* The original command isn't needed any more */
+			kfree(txq->entries[cmd_index].copy_cmd);
+			txq->entries[cmd_index].copy_cmd = NULL;
+		}
+
 		/*
 		 * After here, we should always check rxcb._page_stolen,
 		 * if it is true then one of the handlers took the page.
diff --git a/drivers/net/wireless/iwlwifi/pcie/trans.c b/drivers/net/wireless/iwlwifi/pcie/trans.c
index 02326287ba46..f981b7387292 100644
--- a/drivers/net/wireless/iwlwifi/pcie/trans.c
+++ b/drivers/net/wireless/iwlwifi/pcie/trans.c
@@ -492,10 +492,11 @@ static void iwl_tx_queue_free(struct iwl_trans *trans, int txq_id)
 	iwl_tx_queue_unmap(trans, txq_id);
 
 	/* De-alloc array of command/tx buffers */
-
 	if (txq_id == trans_pcie->cmd_queue)
-		for (i = 0; i < txq->q.n_window; i++)
+		for (i = 0; i < txq->q.n_window; i++) {
 			kfree(txq->entries[i].cmd);
+			kfree(txq->entries[i].copy_cmd);
+		}
 
 	/* De-alloc circular buffer of TFDs */
 	if (txq->q.n_bd) {
diff --git a/drivers/net/wireless/iwlwifi/pcie/tx.c b/drivers/net/wireless/iwlwifi/pcie/tx.c
index 6baf8deef519..392d2bc5e357 100644
--- a/drivers/net/wireless/iwlwifi/pcie/tx.c
+++ b/drivers/net/wireless/iwlwifi/pcie/tx.c
@@ -521,7 +521,7 @@ static int iwl_enqueue_hcmd(struct iwl_trans *trans, struct iwl_host_cmd *cmd)
 	u16 copy_size, cmd_size;
 	bool had_nocopy = false;
 	int i;
-	u8 *cmd_dest;
+	u32 cmd_pos;
 #ifdef CONFIG_IWLWIFI_DEVICE_TRACING
 	const void *trace_bufs[IWL_MAX_CMD_TFDS + 1] = {};
 	int trace_lens[IWL_MAX_CMD_TFDS + 1] = {};
@@ -584,15 +584,31 @@ static int iwl_enqueue_hcmd(struct iwl_trans *trans, struct iwl_host_cmd *cmd)
 					 INDEX_TO_SEQ(q->write_ptr));
 
 	/* and copy the data that needs to be copied */
-
-	cmd_dest = out_cmd->payload;
+	cmd_pos = offsetof(struct iwl_device_cmd, payload);
 	for (i = 0; i < IWL_MAX_CMD_TFDS; i++) {
 		if (!cmd->len[i])
 			continue;
 		if (cmd->dataflags[i] & IWL_HCMD_DFL_NOCOPY)
 			break;
-		memcpy(cmd_dest, cmd->data[i], cmd->len[i]);
-		cmd_dest += cmd->len[i];
+		memcpy((u8 *)out_cmd + cmd_pos, cmd->data[i], cmd->len[i]);
+		cmd_pos += cmd->len[i];
+	}
+
+	WARN_ON_ONCE(txq->entries[idx].copy_cmd);
+
+	/*
+	 * since out_cmd will be the source address of the FH, it will write
+	 * the retry count there. So when the user needs to receivce the HCMD
+	 * that corresponds to the response in the response handler, it needs
+	 * to set CMD_WANT_HCMD.
+	 */
+	if (cmd->flags & CMD_WANT_HCMD) {
+		txq->entries[idx].copy_cmd =
+			kmemdup(out_cmd, cmd_pos, GFP_ATOMIC);
+		if (unlikely(!txq->entries[idx].copy_cmd)) {
+			idx = -ENOMEM;
+			goto out;
+		}
 	}
 
 	IWL_DEBUG_HC(trans,