net/tap: fix fd leak on creation failure
authorYunjian Wang <wangyunjian@huawei.com>
Thu, 16 Apr 2020 03:04:56 +0000 (11:04 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Tue, 21 Apr 2020 11:57:08 +0000 (13:57 +0200)
When eth_dev_tap_create() is failed, nlsk_fd and ka_fd won't be closed
thus leading fds leak. Zero is a valid fd. Ultimately leads to a valid
fd was closed by mistake.

Fixes: bf7b7f437b49 ("net/tap: create netdevice during probing")
Fixes: cb7e68da630a ("net/tap: fix cleanup on allocation failure")
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 41ea548..fea464c 100644 (file)
@@ -1823,6 +1823,8 @@ eth_dev_tap_create(struct rte_vdev_device *vdev, const char *tap_name,
        pmd->dev = dev;
        strlcpy(pmd->name, tap_name, sizeof(pmd->name));
        pmd->type = type;
+       pmd->ka_fd = -1;
+       pmd->nlsk_fd = -1;
 
        pmd->ioctl_sock = socket(AF_INET, SOCK_DGRAM, 0);
        if (pmd->ioctl_sock == -1) {
@@ -1853,7 +1855,6 @@ eth_dev_tap_create(struct rte_vdev_device *vdev, const char *tap_name,
        dev->intr_handle = &pmd->intr_handle;
 
        /* Presetup the fds to -1 as being not valid */
-       pmd->ka_fd = -1;
        for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
                process_private->rxq_fds[i] = -1;
                process_private->txq_fds[i] = -1;
@@ -1993,7 +1994,11 @@ error_remote:
        tap_flow_implicit_flush(pmd, NULL);
 
 error_exit:
-       if (pmd->ioctl_sock > 0)
+       if (pmd->nlsk_fd != -1)
+               close(pmd->nlsk_fd);
+       if (pmd->ka_fd != -1)
+               close(pmd->ka_fd);
+       if (pmd->ioctl_sock != -1)
                close(pmd->ioctl_sock);
        /* mac_addrs must not be freed alone because part of dev_private */
        dev->data->mac_addrs = NULL;