[ovs-dev] [PATCH] netdev-dpdk: Fix mempool segfault.

Stokes, Ian ian.stokes at intel.com
Mon Feb 27 14:20:31 UTC 2017


> >
> >The dpdk_mp_get() function can return a NULL pointer which leads to a
> >segfault when a mempool cannot be created. The lack of a return value
> >check for the function netdev_dpdk_mempool_configure() when called in
> >netdev_dpdk_reconfigure() can result in a segfault also as a NULL
> >pointer for the mempool will be passed to rte_eth_rx_queue_setup().
> >
> >Fix this by adding appropriate NULL pointer and return value checks to
> >dpdk_mp_get() and netdev_dpdk_reconfigure(), also flag when a memory
> >configuration error occurs in dpdk_vhost_reconfigure_helper().
> >
> >Signed-off-by: Ian Stokes <ian.stokes at intel.com>
> >Fixes: 2ae3d542 ("netdev-dpdk: Refactor dpdk_mp_get().")
> >Fixes: 0072e931 ("netdev-dpdk: add support for jumbo frames")
> 
> Hi Ian - this patch doesn't fix 0072e931, as all of the error checking
> that you've provided here was already in that commit.
> Could you remove this line please?

Thanks for the feedback Mark,

The commit 0072e931 introduced the call to netdev_dpdk_mempool_configure() in netdev_dpdk_reconfigure(), however it's missing a check for the return value (You can see other examples of the return value being checked throughout that commit).

Without this check there is the possibility that a null pointer for the mempool will be accessed during the creation of rxqs which happens later in the netdev_dpdk_reconfigure() function.

As such I think its ok to keep the fixes tags above, thoughts?
> 
> A few minor comments below also.
> 
> Cheers,
> Mark
> 
> >CC: Daniele Di Proietto <diproiettod at vmware.com>
> >CC: Mark Kavanagh <mark.b.kavanagh at intel.com>
> >---
> > lib/netdev-dpdk.c |   18 ++++++++++++++----
> > 1 files changed, 14 insertions(+), 4 deletions(-)
> >
> >diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index
> >ee53c4c..62f4409 100644
> >--- a/lib/netdev-dpdk.c
> >+++ b/lib/netdev-dpdk.c
> >@@ -529,8 +529,9 @@ dpdk_mp_get(int socket_id, int mtu)
> >         }
> >     }
> >
> >-    dmp = dpdk_mp_create(socket_id, mtu);
> >-    ovs_list_push_back(&dpdk_mp_list, &dmp->list_node);
> >+    if ((dmp = dpdk_mp_create(socket_id, mtu))) {
> >+        ovs_list_push_back(&dpdk_mp_list, &dmp->list_node);
> >+    }
> >
> > out:
> >     ovs_mutex_unlock(&dpdk_mp_mutex);
> >@@ -3131,7 +3132,11 @@ netdev_dpdk_reconfigure(struct netdev *netdev)
> >
> >     if (dev->mtu != dev->requested_mtu
> >         || dev->socket_id != dev->requested_socket_id) {
> >-        netdev_dpdk_mempool_configure(dev);
> >+        if ((err = netdev_dpdk_mempool_configure(dev))) {
> >+            VLOG_ERR("Mempool configuration failed for device %s: %s",
> 
> This log probably isn't needed, as netdev_dpdk_mempool_configure already
> flags the same thing on error.
> In any event, missing a newline char in the error string.
Sure thing, I can remove in the V2.
> 
> >+                     dev->up.name, rte_strerror(err));
> >+            goto out;
> >+        }
> >     }
> >
> >     netdev->n_txq = dev->requested_n_txq; @@ -3160,6 +3165,7 @@
> >dpdk_vhost_reconfigure_helper(struct netdev_dpdk *dev)  {
> >     dev->up.n_txq = dev->requested_n_txq;
> >     dev->up.n_rxq = dev->requested_n_rxq;
> >+    int err;
> >
> >     /* Enable TX queue 0 by default if it wasn't disabled. */
> >     if (dev->tx_q[0].map == OVS_VHOST_QUEUE_MAP_UNKNOWN) { @@ -3170,9
> >+3176,13 @@ dpdk_vhost_reconfigure_helper(struct netdev_dpdk *dev)
> >
> >     if (dev->requested_socket_id != dev->socket_id
> >         || dev->requested_mtu != dev->mtu) {
> >-        if (!netdev_dpdk_mempool_configure(dev)) {
> >+        if (!(err = netdev_dpdk_mempool_configure(dev))) {
> >             netdev_change_seq_changed(&dev->up);
> >         }
> >+        else {
> >+            VLOG_ERR("Mempool configuration failed for device %s: %s",
> >+                     dev->up.name, rte_strerror(err));
> 
> Ditto
Will remove for the v2.
> 
> >+        }
> >     }
> >
> >     if (!dev->dpdk_mp) {
> >--
> >1.7.0.7



More information about the dev mailing list