@@ -585,6 +585,8 @@ enum rtnetlink_groups {
#define RTNLGRP_PHONET_IFADDR RTNLGRP_PHONET_IFADDR
RTNLGRP_PHONET_ROUTE,
#define RTNLGRP_PHONET_ROUTE RTNLGRP_PHONET_ROUTE
+ RTNLGRP_DCB,
+#define RTNLGRP_DCB RTNLGRP_DCB
__RTNLGRP_MAX
};
#define RTNLGRP_MAX (__RTNLGRP_MAX - 1)
@@ -31,6 +31,8 @@ struct dcb_app_type {
u8 dcb_setapp(struct net_device *, struct dcb_app *);
u8 dcb_getapp(struct net_device *, struct dcb_app *);
+int dcbnl_notify(struct net_device *dev, int event, int cmd, u32 seq, u32 pid);
+
/*
* Ops struct for the netlink callbacks. Used by DCB-enabled drivers through
* the netdevice struct.
@@ -1166,64 +1166,6 @@ err:
return ret;
}
-/* Handle IEEE 802.1Qaz SET commands. If any requested operation can not
- * be completed the entire msg is aborted and error value is returned.
- * No attempt is made to reconcile the case where only part of the
- * cmd can be completed.
- */
-static int dcbnl_ieee_set(struct net_device *netdev, struct nlattr **tb,
- u32 pid, u32 seq, u16 flags)
-{
- const struct dcbnl_rtnl_ops *ops = netdev->dcbnl_ops;
- struct nlattr *ieee[DCB_ATTR_IEEE_MAX + 1];
- int err = -EOPNOTSUPP;
-
- if (!ops)
- goto err;
-
- err = nla_parse_nested(ieee, DCB_ATTR_IEEE_MAX,
- tb[DCB_ATTR_IEEE], dcbnl_ieee_policy);
- if (err)
- goto err;
-
- if (ieee[DCB_ATTR_IEEE_ETS] && ops->ieee_setets) {
- struct ieee_ets *ets = nla_data(ieee[DCB_ATTR_IEEE_ETS]);
- err = ops->ieee_setets(netdev, ets);
- if (err)
- goto err;
- }
-
- if (ieee[DCB_ATTR_IEEE_PFC] && ops->ieee_setpfc) {
- struct ieee_pfc *pfc = nla_data(ieee[DCB_ATTR_IEEE_PFC]);
- err = ops->ieee_setpfc(netdev, pfc);
- if (err)
- goto err;
- }
-
- if (ieee[DCB_ATTR_IEEE_APP_TABLE]) {
- struct nlattr *attr;
- int rem;
-
- nla_for_each_nested(attr, ieee[DCB_ATTR_IEEE_APP_TABLE], rem) {
- struct dcb_app *app_data;
- if (nla_type(attr) != DCB_ATTR_IEEE_APP)
- continue;
- app_data = nla_data(attr);
- if (ops->ieee_setapp)
- err = ops->ieee_setapp(netdev, app_data);
- else
- err = dcb_setapp(netdev, app_data);
- if (err)
- goto err;
- }
- }
-
-err:
- dcbnl_reply(err, RTM_SETDCB, DCB_CMD_IEEE_SET, DCB_ATTR_IEEE,
- pid, seq, flags);
- return err;
-}
-
static int dcbnl_build_peer_app(struct net_device *netdev, struct sk_buff* skb,
int app_nested_type, int app_info_type,
int app_entry_type)
@@ -1279,30 +1221,13 @@ nla_put_failure:
}
/* Handle IEEE 802.1Qaz GET commands. */
-static int dcbnl_ieee_get(struct net_device *netdev, struct nlattr **tb,
- u32 pid, u32 seq, u16 flags)
+static int dcbnl_ieee_fill(struct sk_buff *skb, struct net_device *netdev)
{
- struct sk_buff *skb;
- struct nlmsghdr *nlh;
- struct dcbmsg *dcb;
struct nlattr *ieee, *app;
struct dcb_app_type *itr;
const struct dcbnl_rtnl_ops *ops = netdev->dcbnl_ops;
int dcbx;
- int err;
-
- if (!ops)
- return -EOPNOTSUPP;
-
- skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
- if (!skb)
- return -ENOBUFS;
-
- nlh = NLMSG_NEW(skb, pid, seq, RTM_GETDCB, sizeof(*dcb), flags);
-
- dcb = NLMSG_DATA(nlh);
- dcb->dcb_family = AF_UNSPEC;
- dcb->cmd = DCB_CMD_IEEE_GET;
+ int err = -EMSGSIZE;
NLA_PUT_STRING(skb, DCB_ATTR_IFNAME, netdev->name);
@@ -1378,16 +1303,154 @@ static int dcbnl_ieee_get(struct net_device *netdev, struct nlattr **tb,
if (err)
goto nla_put_failure;
}
- nlmsg_end(skb, nlh);
- return rtnl_unicast(skb, &init_net, pid);
+ return 0;
+
nla_put_failure:
- nlmsg_cancel(skb, nlh);
-nlmsg_failure:
- kfree_skb(skb);
- return -1;
+ return err;
}
+int dcbnl_notify(struct net_device *dev, int event, int cmd,
+ u32 seq, u32 pid)
+{
+ struct net *net = dev_net(dev);
+ struct sk_buff *skb;
+ struct nlmsghdr *nlh;
+ struct dcbmsg *dcb;
+ const struct dcbnl_rtnl_ops *ops = dev->dcbnl_ops;
+ int err;
+
+ if (!ops)
+ return -EOPNOTSUPP;
+
+ skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+ if (!skb)
+ return -ENOBUFS;
+
+ nlh = nlmsg_put(skb, pid, 0, event, sizeof(*dcb), 0);
+ if (nlh == NULL) {
+ kfree(skb);
+ return -EMSGSIZE;
+ }
+
+ dcb = NLMSG_DATA(nlh);
+ dcb->dcb_family = AF_UNSPEC;
+ dcb->cmd = cmd;
+
+ err = dcbnl_ieee_fill(skb, dev);
+ if (err < 0) {
+ /* Report error to broadcast listeners */
+ nlmsg_cancel(skb, nlh);
+ kfree_skb(skb);
+ rtnl_set_sk_err(net, RTNLGRP_DCB, err);
+ } else {
+ /* End nlmsg and notify broadcast listeners */
+ nlmsg_end(skb, nlh);
+ rtnl_notify(skb, net, 0, RTNLGRP_DCB, NULL, GFP_KERNEL);
+ }
+
+ return err;
+}
+EXPORT_SYMBOL(dcbnl_notify);
+
+/* Handle IEEE 802.1Qaz SET commands. If any requested operation can not
+ * be completed the entire msg is aborted and error value is returned.
+ * No attempt is made to reconcile the case where only part of the
+ * cmd can be completed.
+ */
+static int dcbnl_ieee_set(struct net_device *netdev, struct nlattr **tb,
+ u32 pid, u32 seq, u16 flags)
+{
+ const struct dcbnl_rtnl_ops *ops = netdev->dcbnl_ops;
+ struct nlattr *ieee[DCB_ATTR_IEEE_MAX + 1];
+ int err = -EOPNOTSUPP;
+
+ if (!ops)
+ return err;
+
+ err = nla_parse_nested(ieee, DCB_ATTR_IEEE_MAX,
+ tb[DCB_ATTR_IEEE], dcbnl_ieee_policy);
+ if (err)
+ return err;
+
+ if (ieee[DCB_ATTR_IEEE_ETS] && ops->ieee_setets) {
+ struct ieee_ets *ets = nla_data(ieee[DCB_ATTR_IEEE_ETS]);
+ err = ops->ieee_setets(netdev, ets);
+ if (err)
+ goto err;
+ }
+
+ if (ieee[DCB_ATTR_IEEE_PFC] && ops->ieee_setpfc) {
+ struct ieee_pfc *pfc = nla_data(ieee[DCB_ATTR_IEEE_PFC]);
+ err = ops->ieee_setpfc(netdev, pfc);
+ if (err)
+ goto err;
+ }
+
+ if (ieee[DCB_ATTR_IEEE_APP_TABLE]) {
+ struct nlattr *attr;
+ int rem;
+
+ nla_for_each_nested(attr, ieee[DCB_ATTR_IEEE_APP_TABLE], rem) {
+ struct dcb_app *app_data;
+ if (nla_type(attr) != DCB_ATTR_IEEE_APP)
+ continue;
+ app_data = nla_data(attr);
+ if (ops->ieee_setapp)
+ err = ops->ieee_setapp(netdev, app_data);
+ else
+ err = dcb_setapp(netdev, app_data);
+ if (err)
+ goto err;
+ }
+ }
+
+err:
+ dcbnl_reply(err, RTM_SETDCB, DCB_CMD_IEEE_SET, DCB_ATTR_IEEE,
+ pid, seq, flags);
+ dcbnl_notify(netdev, RTM_SETDCB, DCB_CMD_IEEE_SET, seq, 0);
+ return err;
+}
+
+static int dcbnl_ieee_get(struct net_device *netdev, struct nlattr **tb,
+ u32 pid, u32 seq, u16 flags)
+{
+ struct net *net = dev_net(netdev);
+ struct sk_buff *skb;
+ struct nlmsghdr *nlh;
+ struct dcbmsg *dcb;
+ const struct dcbnl_rtnl_ops *ops = netdev->dcbnl_ops;
+ int err;
+
+ if (!ops)
+ return -EOPNOTSUPP;
+
+ skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+ if (!skb)
+ return -ENOBUFS;
+
+ nlh = nlmsg_put(skb, pid, seq, RTM_GETDCB, sizeof(*dcb), flags);
+ if (nlh == NULL) {
+ kfree(skb);
+ return -EMSGSIZE;
+ }
+
+ dcb = NLMSG_DATA(nlh);
+ dcb->dcb_family = AF_UNSPEC;
+ dcb->cmd = DCB_CMD_IEEE_GET;
+
+ err = dcbnl_ieee_fill(skb, netdev);
+
+ if (err < 0) {
+ nlmsg_cancel(skb, nlh);
+ kfree_skb(skb);
+ } else {
+ nlmsg_end(skb, nlh);
+ err = rtnl_unicast(skb, net, pid);
+ }
+
+ return err;
+}
/* DCBX configuration */
static int dcbnl_getdcbx(struct net_device *netdev, struct nlattr **tb,
u32 pid, u32 seq, u16 flags)
Now that dcbnl is being used in many cases by more than a single agent it is beneficial to be notified when some entity either driver or user space has changed the DCB attributes. Today applications either end up polling the interface or relying on a user space database to maintain the DCB state and post events. Polling is a poor solution for obvious reasons. And relying on a user space database has its own downside. Namely it has created strange boot dependencies requiring the database be populated before any applications dependent on DCB attributes starts or the application goes into a polling loop. Populating the database requires negotiating link setting with the peer and can take anywhere from less than a second up to a few seconds depending on the switch implementation. Perhaps more importantly if another application or an embedded agent sets a DCB link attribute the database has no way of knowing other than polling the kernel. This prevents applications from responding quickly to changes in link events which at least in the FCoE case and probably any other protocols expecting a lossless link may result in IO errors. By adding a multicast group for DCB we have clean way to disseminate kernel DCB link attributes up to user space. Avoiding the need for user space to maintain a coherant database and disperse events that potentially do not reflect the current link state. Signed-off-by: John Fastabend <john.r.fastabend@intel.com> --- include/linux/rtnetlink.h | 2 include/net/dcbnl.h | 2 net/dcb/dcbnl.c | 229 +++++++++++++++++++++++++++++---------------- 3 files changed, 150 insertions(+), 83 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html