@@ -121,7 +121,8 @@ int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
}
EXPORT_SYMBOL(rdma_copy_addr);
-int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr,
+int rdma_translate_ip(const struct sockaddr *addr,
+ struct rdma_dev_addr *dev_addr,
u16 *vlan_id)
{
struct net_device *dev;
@@ -139,7 +140,7 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr,
switch (addr->sa_family) {
case AF_INET:
dev = ip_dev_find(dev_addr->net,
- ((struct sockaddr_in *) addr)->sin_addr.s_addr);
+ ((const struct sockaddr_in *)addr)->sin_addr.s_addr);
if (!dev)
return ret;
@@ -154,7 +155,7 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr,
rcu_read_lock();
for_each_netdev_rcu(dev_addr->net, dev) {
if (ipv6_chk_addr(dev_addr->net,
- &((struct sockaddr_in6 *) addr)->sin6_addr,
+ &((const struct sockaddr_in6 *)addr)->sin6_addr,
dev, 1)) {
ret = rdma_copy_addr(dev_addr, dev, NULL);
if (vlan_id)
@@ -198,7 +199,8 @@ static void queue_req(struct addr_req *req)
mutex_unlock(&lock);
}
-static int dst_fetch_ha(struct dst_entry *dst, struct rdma_dev_addr *dev_addr, void *daddr)
+static int dst_fetch_ha(struct dst_entry *dst, struct rdma_dev_addr *dev_addr,
+ const void *daddr)
{
struct neighbour *n;
int ret;
@@ -222,8 +224,9 @@ static int dst_fetch_ha(struct dst_entry *dst, struct rdma_dev_addr *dev_addr, v
}
static int addr4_resolve(struct sockaddr_in *src_in,
- struct sockaddr_in *dst_in,
- struct rdma_dev_addr *addr)
+ const struct sockaddr_in *dst_in,
+ struct rdma_dev_addr *addr,
+ struct rtable **prt)
{
__be32 src_ip = src_in->sin_addr.s_addr;
__be32 dst_ip = dst_in->sin_addr.s_addr;
@@ -243,36 +246,23 @@ static int addr4_resolve(struct sockaddr_in *src_in,
src_in->sin_family = AF_INET;
src_in->sin_addr.s_addr = fl4.saddr;
- if (rt->dst.dev->flags & IFF_LOOPBACK) {
- ret = rdma_translate_ip((struct sockaddr *)dst_in, addr, NULL);
- if (!ret)
- memcpy(addr->dst_dev_addr, addr->src_dev_addr, MAX_ADDR_LEN);
- goto put;
- }
-
- /* If the device does ARP internally, return 'done' */
- if (rt->dst.dev->flags & IFF_NOARP) {
- ret = rdma_copy_addr(addr, rt->dst.dev, NULL);
- goto put;
- }
-
/* If there's a gateway, we're definitely in RoCE v2 (as RoCE v1 isn't
* routable) and we could set the network type accordingly.
*/
if (rt->rt_uses_gateway)
addr->network = RDMA_NETWORK_IPV4;
- ret = dst_fetch_ha(&rt->dst, addr, &fl4.daddr);
-put:
- ip_rt_put(rt);
+ *prt = rt;
+ return 0;
out:
return ret;
}
#if IS_ENABLED(CONFIG_IPV6)
static int addr6_resolve(struct sockaddr_in6 *src_in,
- struct sockaddr_in6 *dst_in,
- struct rdma_dev_addr *addr)
+ const struct sockaddr_in6 *dst_in,
+ struct rdma_dev_addr *addr,
+ struct dst_entry **pdst)
{
struct flowi6 fl6;
struct dst_entry *dst;
@@ -299,49 +289,109 @@ static int addr6_resolve(struct sockaddr_in6 *src_in,
src_in->sin6_addr = fl6.saddr;
}
- if (dst->dev->flags & IFF_LOOPBACK) {
- ret = rdma_translate_ip((struct sockaddr *)dst_in, addr, NULL);
- if (!ret)
- memcpy(addr->dst_dev_addr, addr->src_dev_addr, MAX_ADDR_LEN);
- goto put;
- }
-
- /* If the device does ARP internally, return 'done' */
- if (dst->dev->flags & IFF_NOARP) {
- ret = rdma_copy_addr(addr, dst->dev, NULL);
- goto put;
- }
-
/* If there's a gateway, we're definitely in RoCE v2 (as RoCE v1 isn't
* routable) and we could set the network type accordingly.
*/
if (rt->rt6i_flags & RTF_GATEWAY)
addr->network = RDMA_NETWORK_IPV6;
- ret = dst_fetch_ha(dst, addr, &fl6.daddr);
+ *pdst = dst;
+ return 0;
put:
dst_release(dst);
return ret;
}
#else
static int addr6_resolve(struct sockaddr_in6 *src_in,
- struct sockaddr_in6 *dst_in,
- struct rdma_dev_addr *addr)
+ const struct sockaddr_in6 *dst_in,
+ struct rdma_dev_addr *addr,
+ struct dst_entry **pdst)
{
return -EADDRNOTAVAIL;
}
#endif
+static int addr_resolve_neigh(struct dst_entry *dst,
+ const struct sockaddr *dst_in,
+ struct rdma_dev_addr *addr)
+{
+ if (dst->dev->flags & IFF_LOOPBACK) {
+ int ret;
+
+ ret = rdma_translate_ip(dst_in, addr, NULL);
+ if (!ret)
+ memcpy(addr->dst_dev_addr, addr->src_dev_addr,
+ MAX_ADDR_LEN);
+
+ return ret;
+ }
+
+ /* If the device does ARP internally */
+ if (!(dst->dev->flags & IFF_NOARP)) {
+ const struct sockaddr_in *dst_in4 =
+ (const struct sockaddr_in *)dst_in;
+ const struct sockaddr_in6 *dst_in6 =
+ (const struct sockaddr_in6 *)dst_in;
+
+ return dst_fetch_ha(dst, addr,
+ dst_in->sa_family == AF_INET ?
+ (const void *)&dst_in4->sin_addr.s_addr :
+ (const void *)&dst_in6->sin6_addr);
+ }
+
+ return rdma_copy_addr(addr, dst->dev, NULL);
+}
+
static int addr_resolve(struct sockaddr *src_in,
- struct sockaddr *dst_in,
- struct rdma_dev_addr *addr)
+ const struct sockaddr *dst_in,
+ struct rdma_dev_addr *addr,
+ bool resolve_neigh)
{
+ struct net_device *ndev;
+ struct dst_entry *dst;
+ int ret;
+
if (src_in->sa_family == AF_INET) {
- return addr4_resolve((struct sockaddr_in *) src_in,
- (struct sockaddr_in *) dst_in, addr);
- } else
- return addr6_resolve((struct sockaddr_in6 *) src_in,
- (struct sockaddr_in6 *) dst_in, addr);
+ struct rtable *rt = NULL;
+ const struct sockaddr_in *dst_in4 =
+ (const struct sockaddr_in *)dst_in;
+
+ ret = addr4_resolve((struct sockaddr_in *)src_in,
+ dst_in4, addr, &rt);
+ if (ret)
+ return ret;
+
+ if (resolve_neigh)
+ ret = addr_resolve_neigh(&rt->dst, dst_in, addr);
+
+ ndev = rt->dst.dev;
+ dev_hold(ndev);
+
+ ip_rt_put(rt);
+ } else {
+ const struct sockaddr_in6 *dst_in6 =
+ (const struct sockaddr_in6 *)dst_in;
+
+ ret = addr6_resolve((struct sockaddr_in6 *)src_in,
+ dst_in6, addr,
+ &dst);
+ if (ret)
+ return ret;
+
+ if (resolve_neigh)
+ ret = addr_resolve_neigh(dst, dst_in, addr);
+
+ ndev = dst->dev;
+ dev_hold(ndev);
+
+ dst_release(dst);
+ }
+
+ addr->bound_dev_if = ndev->ifindex;
+ addr->net = dev_net(ndev);
+ dev_put(ndev);
+
+ return ret;
}
static void process_req(struct work_struct *work)
@@ -357,7 +407,8 @@ static void process_req(struct work_struct *work)
if (req->status == -ENODATA) {
src_in = (struct sockaddr *) &req->src_addr;
dst_in = (struct sockaddr *) &req->dst_addr;
- req->status = addr_resolve(src_in, dst_in, req->addr);
+ req->status = addr_resolve(src_in, dst_in, req->addr,
+ true);
if (req->status && time_after_eq(jiffies, req->timeout))
req->status = -ETIMEDOUT;
else if (req->status == -ENODATA)
@@ -417,7 +468,7 @@ int rdma_resolve_ip(struct rdma_addr_client *client,
req->client = client;
atomic_inc(&client->refcount);
- req->status = addr_resolve(src_in, dst_in, addr);
+ req->status = addr_resolve(src_in, dst_in, addr, true);
switch (req->status) {
case 0:
req->timeout = jiffies;
@@ -439,6 +490,25 @@ err:
}
EXPORT_SYMBOL(rdma_resolve_ip);
+int rdma_resolve_ip_route(struct sockaddr *src_addr,
+ const struct sockaddr *dst_addr,
+ struct rdma_dev_addr *addr)
+{
+ struct sockaddr_storage ssrc_addr;
+ struct sockaddr *src_in = (struct sockaddr *)&ssrc_addr;
+
+ if (src_addr->sa_family != dst_addr->sa_family)
+ return -EINVAL;
+
+ if (src_addr)
+ memcpy(src_in, src_addr, rdma_addr_size(src_addr));
+ else
+ src_in->sa_family = dst_addr->sa_family;
+
+ return addr_resolve(src_in, dst_addr, addr, false);
+}
+EXPORT_SYMBOL(rdma_resolve_ip_route);
+
void rdma_addr_cancel(struct rdma_dev_addr *addr)
{
struct addr_req *req, *temp_req;
@@ -471,7 +541,7 @@ static void resolve_cb(int status, struct sockaddr *src_addr,
}
int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgid,
- u8 *dmac, u16 *vlan_id, int if_index)
+ u8 *dmac, u16 *vlan_id, int *if_index)
{
int ret = 0;
struct rdma_dev_addr dev_addr;
@@ -489,7 +559,8 @@ int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgi
rdma_gid2ip(&dgid_addr._sockaddr, dgid);
memset(&dev_addr, 0, sizeof(dev_addr));
- dev_addr.bound_dev_if = if_index;
+ if (if_index)
+ dev_addr.bound_dev_if = *if_index;
dev_addr.net = &init_net;
ctx.addr = &dev_addr;
@@ -505,6 +576,8 @@ int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgi
dev = dev_get_by_index(&init_net, dev_addr.bound_dev_if);
if (!dev)
return -ENODEV;
+ if (if_index)
+ *if_index = dev_addr.bound_dev_if;
if (vlan_id)
*vlan_id = rdma_vlan_dev_vlan_id(dev);
dev_put(dev);
@@ -1646,8 +1646,11 @@ static int cm_req_handler(struct cm_work *work)
cm_id_priv->av.ah_attr.grh.sgid_index,
&gid, &gid_attr);
if (!ret) {
- if (gid_attr.ndev)
+ if (gid_attr.ndev) {
+ work->path[0].ifindex = gid_attr.ndev->ifindex;
+ work->path[0].net = dev_net(gid_attr.ndev);
dev_put(gid_attr.ndev);
+ }
work->path[0].gid_type = gid_attr.gid_type;
ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av);
}
@@ -1656,8 +1659,11 @@ static int cm_req_handler(struct cm_work *work)
work->port->port_num, 0,
&work->path[0].sgid,
&gid_attr);
- if (!err && gid_attr.ndev)
+ if (!err && gid_attr.ndev) {
+ work->path[0].ifindex = gid_attr.ndev->ifindex;
+ work->path[0].net = dev_net(gid_attr.ndev);
dev_put(gid_attr.ndev);
+ }
work->path[0].gid_type = gid_attr.gid_type;
ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_GID,
&work->path[0].sgid, sizeof work->path[0].sgid,
@@ -454,8 +454,20 @@ static inline int cma_validate_port(struct ib_device *device, u8 port,
if ((dev_type != ARPHRD_INFINIBAND) && rdma_protocol_ib(device, port))
return ret;
- if (dev_type == ARPHRD_ETHER)
+ if (dev_type == ARPHRD_ETHER) {
ndev = dev_get_by_index(&init_net, bound_if_index);
+ if (ndev && ndev->flags & IFF_LOOPBACK) {
+ pr_info("detected loopback device\n");
+ dev_put(ndev);
+
+ if (!device->get_netdev)
+ return -EOPNOTSUPP;
+
+ ndev = device->get_netdev(device, port);
+ if (!ndev)
+ return -ENODEV;
+ }
+ }
ret = ib_find_cached_gid_by_port(device, gid, IB_GID_TYPE_IB, port,
ndev, NULL);
@@ -2325,8 +2337,22 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv)
if (addr->dev_addr.bound_dev_if) {
ndev = dev_get_by_index(&init_net, addr->dev_addr.bound_dev_if);
+ if (!ndev)
+ return -ENODEV;
+
+ if (ndev->flags & IFF_LOOPBACK) {
+ dev_put(ndev);
+ if (!id_priv->id.device->get_netdev)
+ return -EOPNOTSUPP;
+
+ ndev = id_priv->id.device->get_netdev(id_priv->id.device,
+ id_priv->id.port_num);
+ if (!ndev)
+ return -ENODEV;
+ }
+
route->path_rec->net = &init_net;
- route->path_rec->ifindex = addr->dev_addr.bound_dev_if;
+ route->path_rec->ifindex = ndev->ifindex;
route->path_rec->gid_type = id_priv->gid_type;
}
if (!ndev) {
@@ -49,7 +49,9 @@
#include <net/netlink.h>
#include <uapi/rdma/ib_user_sa.h>
#include <rdma/ib_marshall.h>
+#include <rdma/ib_addr.h>
#include "sa.h"
+#include "core_priv.h"
MODULE_AUTHOR("Roland Dreier");
MODULE_DESCRIPTION("InfiniBand subnet administration query support");
@@ -994,7 +996,8 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num,
{
int ret;
u16 gid_index;
- int force_grh;
+ int use_roce;
+ struct net_device *ndev = NULL;
memset(ah_attr, 0, sizeof *ah_attr);
ah_attr->dlid = be16_to_cpu(rec->dlid);
@@ -1004,16 +1007,71 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num,
ah_attr->port_num = port_num;
ah_attr->static_rate = rec->rate;
- force_grh = rdma_cap_eth_ah(device, port_num);
+ use_roce = rdma_cap_eth_ah(device, port_num);
+
+ if (use_roce) {
+ struct net_device *idev;
+ struct net_device *resolved_dev;
+ struct rdma_dev_addr dev_addr = {.bound_dev_if = rec->ifindex,
+ .net = rec->net ? rec->net :
+ &init_net};
+ union {
+ struct sockaddr _sockaddr;
+ struct sockaddr_in _sockaddr_in;
+ struct sockaddr_in6 _sockaddr_in6;
+ } sgid_addr, dgid_addr;
+
+ if (!device->get_netdev)
+ return -EOPNOTSUPP;
+
+ rdma_gid2ip(&sgid_addr._sockaddr, &rec->sgid);
+ rdma_gid2ip(&dgid_addr._sockaddr, &rec->dgid);
+
+ /* validate the route */
+ ret = rdma_resolve_ip_route(&sgid_addr._sockaddr,
+ &dgid_addr._sockaddr, &dev_addr);
+ if (ret)
+ return ret;
- if (rec->hop_limit > 1 || force_grh) {
- struct net_device *ndev = ib_get_ndev_from_path(rec);
+ if ((dev_addr.network == RDMA_NETWORK_IPV4 ||
+ dev_addr.network == RDMA_NETWORK_IPV6) &&
+ rec->gid_type != IB_GID_TYPE_ROCE_UDP_ENCAP)
+ return -EINVAL;
+
+ idev = device->get_netdev(device, port_num);
+ if (!idev)
+ return -ENODEV;
+
+ resolved_dev = dev_get_by_index(dev_addr.net,
+ dev_addr.bound_dev_if);
+ if (resolved_dev->flags & IFF_LOOPBACK) {
+ dev_put(resolved_dev);
+ resolved_dev = idev;
+ dev_hold(resolved_dev);
+ }
+ ndev = ib_get_ndev_from_path(rec);
+ rcu_read_lock();
+ if ((ndev && ndev != resolved_dev) ||
+ (resolved_dev != idev &&
+ !rdma_is_upper_dev_rcu(idev, resolved_dev)))
+ ret = -EHOSTUNREACH;
+ rcu_read_unlock();
+ dev_put(idev);
+ dev_put(resolved_dev);
+ if (ret) {
+ if (ndev)
+ dev_put(ndev);
+ return ret;
+ }
+ }
+ if (rec->hop_limit > 1 || use_roce) {
ah_attr->ah_flags = IB_AH_GRH;
ah_attr->grh.dgid = rec->dgid;
- ret = ib_find_cached_gid(device, &rec->sgid, rec->gid_type, ndev,
- &port_num, &gid_index);
+ ret = ib_find_cached_gid_by_port(device, &rec->sgid,
+ rec->gid_type, port_num, ndev,
+ &gid_index);
if (ret) {
if (ndev)
dev_put(ndev);
@@ -1027,9 +1085,10 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num,
if (ndev)
dev_put(ndev);
}
- if (force_grh) {
+
+ if (use_roce)
memcpy(ah_attr->dmac, rec->dmac, ETH_ALEN);
- }
+
return 0;
}
EXPORT_SYMBOL(ib_init_ah_from_path);
@@ -457,30 +457,52 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num,
return ret;
if (rdma_protocol_roce(device, port_num)) {
+ int if_index = 0;
u16 vlan_id = wc->wc_flags & IB_WC_WITH_VLAN ?
wc->vlan_id : 0xffff;
+ struct net_device *idev;
+ struct net_device *resolved_dev;
if (!(wc->wc_flags & IB_WC_GRH))
return -EPROTOTYPE;
- if (!(wc->wc_flags & IB_WC_WITH_SMAC) ||
- !(wc->wc_flags & IB_WC_WITH_VLAN)) {
- ret = rdma_addr_find_dmac_by_grh(&dgid, &sgid,
- ah_attr->dmac,
- wc->wc_flags & IB_WC_WITH_VLAN ?
- NULL : &vlan_id,
- 0);
- if (ret)
- return ret;
+ if (!device->get_netdev)
+ return -EOPNOTSUPP;
+
+ idev = device->get_netdev(device, port_num);
+ if (!idev)
+ return -ENODEV;
+
+ ret = rdma_addr_find_dmac_by_grh(&dgid, &sgid,
+ ah_attr->dmac,
+ wc->wc_flags & IB_WC_WITH_VLAN ?
+ NULL : &vlan_id,
+ &if_index);
+ if (ret) {
+ dev_put(idev);
+ return ret;
}
+ resolved_dev = dev_get_by_index(&init_net, if_index);
+ if (resolved_dev->flags & IFF_LOOPBACK) {
+ dev_put(resolved_dev);
+ resolved_dev = idev;
+ dev_hold(resolved_dev);
+ }
+ rcu_read_lock();
+ if (resolved_dev != idev && !rdma_is_upper_dev_rcu(idev,
+ resolved_dev))
+ ret = -EHOSTUNREACH;
+ rcu_read_unlock();
+ dev_put(idev);
+ dev_put(resolved_dev);
+ if (ret)
+ return ret;
+
ret = get_sgid_index_from_eth(device, port_num, vlan_id,
&dgid, gid_type, &gid_index);
if (ret)
return ret;
-
- if (wc->wc_flags & IB_WC_WITH_SMAC)
- memcpy(ah_attr->dmac, wc->smac, ETH_ALEN);
}
ah_attr->dlid = wc->slid;
@@ -1145,7 +1167,7 @@ int ib_resolve_eth_dmac(struct ib_qp *qp,
ret = rdma_addr_find_dmac_by_grh(&sgid,
&qp_attr->ah_attr.grh.dgid,
qp_attr->ah_attr.dmac,
- NULL, ifindex);
+ NULL, &ifindex);
dev_put(sgid_attr.ndev);
}
@@ -154,7 +154,7 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
(!rdma_link_local_addr((struct in6_addr *)attr->grh.dgid.raw))) {
status = rdma_addr_find_dmac_by_grh(&sgid, &attr->grh.dgid,
attr->dmac, &vlan_tag,
- sgid_attr.ndev->ifindex);
+ &sgid_attr.ndev->ifindex);
if (status) {
pr_err("%s(): Failed to resolve dmac from gid."
"status = %d\n", __func__, status);
@@ -92,8 +92,8 @@ struct rdma_dev_addr {
*
* The dev_addr->net field must be initialized.
*/
-int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr,
- u16 *vlan_id);
+int rdma_translate_ip(const struct sockaddr *addr,
+ struct rdma_dev_addr *dev_addr, u16 *vlan_id);
/**
* rdma_resolve_ip - Resolve source and destination IP addresses to
@@ -118,6 +118,10 @@ int rdma_resolve_ip(struct rdma_addr_client *client,
struct rdma_dev_addr *addr, void *context),
void *context);
+int rdma_resolve_ip_route(struct sockaddr *src_addr,
+ const struct sockaddr *dst_addr,
+ struct rdma_dev_addr *addr);
+
void rdma_addr_cancel(struct rdma_dev_addr *addr);
int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
@@ -127,7 +131,7 @@ int rdma_addr_size(struct sockaddr *addr);
int rdma_addr_find_smac_by_sgid(union ib_gid *sgid, u8 *smac, u16 *vlan_id);
int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgid,
- u8 *smac, u16 *vlan_id, int if_index);
+ u8 *smac, u16 *vlan_id, int *if_index);
static inline u16 ib_addr_get_pkey(struct rdma_dev_addr *dev_addr)
{
In order to make sure API users don't try to use SGIDs which don't conform to the routing table, validate the route before searching the RoCE GID table. Signed-off-by: Matan Barak <matanb@mellanox.com> --- drivers/infiniband/core/addr.c | 175 ++++++++++++++++++++++--------- drivers/infiniband/core/cm.c | 10 +- drivers/infiniband/core/cma.c | 30 +++++- drivers/infiniband/core/sa_query.c | 75 +++++++++++-- drivers/infiniband/core/verbs.c | 48 ++++++--- drivers/infiniband/hw/ocrdma/ocrdma_ah.c | 2 +- include/rdma/ib_addr.h | 10 +- 7 files changed, 270 insertions(+), 80 deletions(-)