diff options
| author | Sean Hefty <sean.hefty@intel.com> | 2011-12-06 21:15:18 +0000 |
|---|---|---|
| committer | Roland Dreier <roland@purestorage.com> | 2012-01-04 09:13:52 -0800 |
| commit | 46ea5061c750fc6b088b9566234287115fb70f98 (patch) | |
| tree | db5ad08595bac4e37b508f697b950246bb87aba3 /drivers/infiniband | |
| parent | 5f0a6e2d503896062f641639dacfe5055c2f593b (diff) | |
| download | op-kernel-dev-46ea5061c750fc6b088b9566234287115fb70f98.zip op-kernel-dev-46ea5061c750fc6b088b9566234287115fb70f98.tar.gz | |
RDMA/cma: Fix endianness bugs
Fix endianness bugs reported by sparse in the RDMA core stack. Note
that these are real bugs, but don't affect any existing code to the
best of my knowledge. The mlid issue would only affect kernel users
of rdma_join_multicast which have the rdma_cm attach/detach its QP.
There are no current in tree users that do this. (rdma_join_multicast
may be used called by user space applications, which does not have
this issue.) And the pkey setting is simply returned as
informational.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Diffstat (limited to 'drivers/infiniband')
| -rw-r--r-- | drivers/infiniband/core/cma.c | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index d0d4aa9..fdd8ef8 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -1110,7 +1110,7 @@ static struct rdma_id_private *cma_new_conn_id(struct rdma_cm_id *listen_id, if (cma_any_addr((struct sockaddr *) &rt->addr.src_addr)) { rt->addr.dev_addr.dev_type = ARPHRD_INFINIBAND; rdma_addr_set_sgid(&rt->addr.dev_addr, &rt->path_rec[0].sgid); - ib_addr_set_pkey(&rt->addr.dev_addr, rt->path_rec[0].pkey); + ib_addr_set_pkey(&rt->addr.dev_addr, be16_to_cpu(rt->path_rec[0].pkey)); } else { ret = rdma_translate_ip((struct sockaddr *) &rt->addr.src_addr, &rt->addr.dev_addr); @@ -2926,7 +2926,7 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast) mutex_lock(&id_priv->qp_mutex); if (!status && id_priv->id.qp) status = ib_attach_mcast(id_priv->id.qp, &multicast->rec.mgid, - multicast->rec.mlid); + be16_to_cpu(multicast->rec.mlid)); mutex_unlock(&id_priv->qp_mutex); memset(&event, 0, sizeof event); @@ -3187,7 +3187,7 @@ void rdma_leave_multicast(struct rdma_cm_id *id, struct sockaddr *addr) if (id->qp) ib_detach_mcast(id->qp, &mc->multicast.ib->rec.mgid, - mc->multicast.ib->rec.mlid); + be16_to_cpu(mc->multicast.ib->rec.mlid)); if (rdma_node_get_transport(id_priv->cma_dev->device->node_type) == RDMA_TRANSPORT_IB) { switch (rdma_port_get_link_layer(id->device, id->port_num)) { case IB_LINK_LAYER_INFINIBAND: |
