net/tap: fix mbuf and mem leak during queue release
authorYunjian Wang <wangyunjian@huawei.com>
Thu, 16 Apr 2020 03:04:25 +0000 (11:04 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Tue, 21 Apr 2020 11:57:08 +0000 (13:57 +0200)
For the tap PMD, we should release mbufs and iovecs from the Rx queue
when closing device. In order to remove duplicated code,
rte_pmd_tap_remove() calls tap_dev_close().

Fixes: 0781f5762cfe ("net/tap: support segmented mbufs")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
drivers/net/tap/rte_eth_tap.c

index 6d54c25..9c88b92 100644 (file)
@@ -1020,15 +1020,25 @@ tap_dev_close(struct rte_eth_dev *dev)
        int i;
        struct pmd_internals *internals = dev->data->dev_private;
        struct pmd_process_private *process_private = dev->process_private;
+       struct rx_queue *rxq;
 
        tap_link_set_down(dev);
-       tap_flow_flush(dev, NULL);
-       tap_flow_implicit_flush(internals, NULL);
+       if (internals->nlsk_fd != -1) {
+               tap_flow_flush(dev, NULL);
+               tap_flow_implicit_flush(internals, NULL);
+               tap_nl_final(internals->nlsk_fd);
+               internals->nlsk_fd = -1;
+       }
 
        for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
                if (process_private->rxq_fds[i] != -1) {
+                       rxq = &internals->rxq[i];
                        close(process_private->rxq_fds[i]);
                        process_private->rxq_fds[i] = -1;
+                       rte_pktmbuf_free(rxq->pool);
+                       rte_free(rxq->iovecs);
+                       rxq->pool = NULL;
+                       rxq->iovecs = NULL;
                }
                if (process_private->txq_fds[i] != -1) {
                        close(process_private->txq_fds[i]);
@@ -2398,8 +2408,6 @@ rte_pmd_tap_remove(struct rte_vdev_device *dev)
 {
        struct rte_eth_dev *eth_dev = NULL;
        struct pmd_internals *internals;
-       struct pmd_process_private *process_private;
-       int i;
 
        /* find the ethdev entry */
        eth_dev = rte_eth_dev_allocated(rte_vdev_device_name(dev));
@@ -2412,28 +2420,12 @@ rte_pmd_tap_remove(struct rte_vdev_device *dev)
        if (rte_eal_process_type() != RTE_PROC_PRIMARY)
                return rte_eth_dev_release_port(eth_dev);
 
-       internals = eth_dev->data->dev_private;
-       process_private = eth_dev->process_private;
+       tap_dev_close(eth_dev);
 
+       internals = eth_dev->data->dev_private;
        TAP_LOG(DEBUG, "Closing %s Ethernet device on numa %u",
                tuntap_types[internals->type], rte_socket_id());
 
-       if (internals->nlsk_fd) {
-               tap_flow_flush(eth_dev, NULL);
-               tap_flow_implicit_flush(internals, NULL);
-               tap_nl_final(internals->nlsk_fd);
-       }
-       for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
-               if (process_private->rxq_fds[i] != -1) {
-                       close(process_private->rxq_fds[i]);
-                       process_private->rxq_fds[i] = -1;
-               }
-               if (process_private->txq_fds[i] != -1) {
-                       close(process_private->txq_fds[i]);
-                       process_private->txq_fds[i] = -1;
-               }
-       }
-
        close(internals->ioctl_sock);
        rte_free(eth_dev->process_private);
        if (tap_devices_count == 1)