/**
* del_gid - Delete GID table entry
*
* @ib_dev: IB device whose GID entry to be deleted
* @port: Port number of the IB device
* @table: GID table of the IB device for a port
* @ix: GID entry index to delete
*
*/
static void del_gid(struct ib_device *ib_dev, u8 port,
struct ib_gid_table *table, int ix)
{
struct ib_gid_table_entry *entry;
lockdep_assert_held(&table->lock);
dev_dbg(&ib_dev->dev, "%s port=%d index=%d gid %pI6\n", __func__, port,
ix, table->data_vec[ix]->attr.gid.raw);
write_lock_irq(&table->rwlock);
entry = table->data_vec[ix];
entry->state = GID_TABLE_ENTRY_PENDING_DEL;
/*
* For non RoCE protocol, GID entry slot is ready to use.
*/
if (!rdma_protocol_roce(ib_dev, port))
table->data_vec[ix] = NULL;
write_unlock_irq(&table->rwlock);
if (rdma_cap_roce_gid_table(ib_dev, port))
ib_dev->ops.del_gid(&entry->attr, &entry->context);
put_gid_entry_locked(entry);
}
static int gid_table_reserve_default(struct ib_device *ib_dev, u8 port,
struct ib_gid_table *table)
{
if (rdma_protocol_roce(ib_dev, port)) {
struct ib_gid_table_entry *entry = &table->data_vec[0];
entry->props |= GID_TABLE_ENTRY_DEFAULT;
}
return 0;
}
/**
* ib_find_gid_by_filter - Returns the GID table index where a specified
* GID value occurs
* @device: The device to query.
* @gid: The GID value to search for.
* @port_num: The port number of the device where the GID value could be
* searched.
* @filter: The filter function is executed on any matching GID in the table.
* If the filter function returns true, the corresponding index is returned,
* otherwise, we continue searching the GID table. It's guaranteed that
* while filter is executed, ndev field is valid and the structure won't
* change. filter is executed in an atomic context. filter must not be NULL.
* @index: The index into the cached GID table where the GID was found. This
* parameter may be NULL.
*
* ib_cache_gid_find_by_filter() searches for the specified GID value
* of which the filter function returns true in the port's GID table.
* This function is only supported on RoCE ports.
*
*/
static int ib_cache_gid_find_by_filter(struct ib_device *ib_dev,
const union ib_gid *gid,
u8 port,
bool (*filter)(const union ib_gid *,
const struct ib_gid_attr *,
void *),
void *context,
u16 *index)
{
struct ib_gid_table **ports_table = ib_dev->cache.gid_cache;
struct ib_gid_table *table;
unsigned int i;
bool found = false;
if (!ports_table)
return -EOPNOTSUPP;
if (port < rdma_start_port(ib_dev) ||
port > rdma_end_port(ib_dev) ||
!rdma_protocol_roce(ib_dev, port))
return -EPROTONOSUPPORT;
table = ports_table[port - rdma_start_port(ib_dev)];
for (i = 0; i < table->sz; i++) {
struct ib_gid_attr attr;
unsigned long flags;
read_lock_irqsave(&table->data_vec[i].lock, flags);
if (table->data_vec[i].props & GID_TABLE_ENTRY_INVALID)
goto next;
if (memcmp(gid, &table->data_vec[i].gid, sizeof(*gid)))
goto next;
memcpy(&attr, &table->data_vec[i].attr, sizeof(attr));
if (filter(gid, &attr, context))
found = true;
next:
read_unlock_irqrestore(&table->data_vec[i].lock, flags);
if (found)
break;
}
if (!found)
return -ENOENT;
if (index)
*index = i;
return 0;
}
/**
* add_modify_gid - Add or modify GID table entry
*
* @table: GID table in which GID to be added or modified
* @attr: Attributes of the GID
*
* Returns 0 on success or appropriate error code. It accepts zero
* GID addition for non RoCE ports for HCA's who report them as valid
* GID. However such zero GIDs are not added to the cache.
*/
static int add_modify_gid(struct ib_gid_table *table,
const struct ib_gid_attr *attr)
{
struct ib_gid_table_entry *entry;
int ret = 0;
/*
* Invalidate any old entry in the table to make it safe to write to
* this index.
*/
if (is_gid_entry_valid(table->data_vec[attr->index]))
del_gid(attr->device, attr->port_num, table, attr->index);
/*
* Some HCA's report multiple GID entries with only one valid GID, and
* leave other unused entries as the zero GID. Convert zero GIDs to
* empty table entries instead of storing them.
*/
if (rdma_is_zero_gid(&attr->gid))
return 0;
entry = alloc_gid_entry(attr);
if (!entry)
return -ENOMEM;
if (rdma_protocol_roce(attr->device, attr->port_num)) {
ret = add_roce_gid(entry);
if (ret)
goto done;
}
store_gid_entry(table, entry);
return 0;
done:
put_gid_entry(entry);
return ret;
}
static void ib_cache_update(struct ib_device *device,
u8 port,
bool enforce_security)
{
struct ib_port_attr *tprops = NULL;
struct ib_pkey_cache *pkey_cache = NULL, *old_pkey_cache;
int i;
int ret;
if (!rdma_is_port_valid(device, port))
return;
tprops = kmalloc(sizeof *tprops, GFP_KERNEL);
if (!tprops)
return;
ret = ib_query_port(device, port, tprops);
if (ret) {
dev_warn(&device->dev, "ib_query_port failed (%d)\n", ret);
goto err;
}
if (!rdma_protocol_roce(device, port)) {
ret = config_non_roce_gid_cache(device, port,
tprops->gid_tbl_len);
if (ret)
goto err;
}
pkey_cache = kmalloc(struct_size(pkey_cache, table,
tprops->pkey_tbl_len),
GFP_KERNEL);
if (!pkey_cache)
goto err;
pkey_cache->table_len = tprops->pkey_tbl_len;
for (i = 0; i < pkey_cache->table_len; ++i) {
ret = ib_query_pkey(device, port, i, pkey_cache->table + i);
if (ret) {
dev_warn(&device->dev,
"ib_query_pkey failed (%d) for index %d\n",
ret, i);
goto err;
}
}
write_lock_irq(&device->cache.lock);
old_pkey_cache = device->cache.ports[port -
rdma_start_port(device)].pkey;
device->cache.ports[port - rdma_start_port(device)].pkey = pkey_cache;
device->cache.ports[port - rdma_start_port(device)].lmc = tprops->lmc;
device->cache.ports[port - rdma_start_port(device)].port_state =
tprops->state;
device->cache.ports[port - rdma_start_port(device)].subnet_prefix =
tprops->subnet_prefix;
write_unlock_irq(&device->cache.lock);
if (enforce_security)
ib_security_cache_change(device,
port,
tprops->subnet_prefix);
kfree(old_pkey_cache);
kfree(tprops);
return;
err:
kfree(pkey_cache);
kfree(tprops);
}
int ib_init_ah_from_wc(struct ib_device *device, u8 port_num,
const struct ib_wc *wc, const struct ib_grh *grh,
struct ib_ah_attr *ah_attr)
{
u32 flow_class;
u16 gid_index;
int ret;
enum rdma_network_type net_type = RDMA_NETWORK_IB;
enum ib_gid_type gid_type = IB_GID_TYPE_IB;
int hoplimit = 0xff;
union ib_gid dgid;
union ib_gid sgid;
memset(ah_attr, 0, sizeof *ah_attr);
if (rdma_cap_eth_ah(device, port_num)) {
if (wc->wc_flags & IB_WC_WITH_NETWORK_HDR_TYPE)
net_type = wc->network_hdr_type;
else
net_type = ib_get_net_type_by_grh(device, port_num, grh);
gid_type = ib_network_to_gid_type(net_type);
}
ret = get_gids_from_rdma_hdr((union rdma_network_hdr *)grh, net_type,
&sgid, &dgid);
if (ret)
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 (!device->get_netdev)
return -EOPNOTSUPP;
idev = device->get_netdev(device, port_num);
if (!idev)
return -ENODEV;
ret = rdma_addr_find_l2_eth_by_grh(&dgid, &sgid,
ah_attr->dmac,
wc->wc_flags & IB_WC_WITH_VLAN ?
NULL : &vlan_id,
&if_index, &hoplimit);
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;
}
ah_attr->dlid = wc->slid;
ah_attr->sl = wc->sl;
ah_attr->src_path_bits = wc->dlid_path_bits;
ah_attr->port_num = port_num;
if (wc->wc_flags & IB_WC_GRH) {
ah_attr->ah_flags = IB_AH_GRH;
ah_attr->grh.dgid = sgid;
if (!rdma_cap_eth_ah(device, port_num)) {
if (dgid.global.interface_id != cpu_to_be64(IB_SA_WELL_KNOWN_GUID)) {
ret = ib_find_cached_gid_by_port(device, &dgid,
IB_GID_TYPE_IB,
port_num, NULL,
&gid_index);
if (ret)
return ret;
} else {
gid_index = 0;
}
}
ah_attr->grh.sgid_index = (u8) gid_index;
flow_class = be32_to_cpu(grh->version_tclass_flow);
ah_attr->grh.flow_label = flow_class & 0xFFFFF;
ah_attr->grh.hop_limit = hoplimit;
ah_attr->grh.traffic_class = (flow_class >> 20) & 0xFF;
}
