netdev - [PATCH net-next 5/9] ipv6: Only Replay routes of interest to new listeners

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20191223132820.888247-6-idosch@idosch.org>
Date:   Mon, 23 Dec 2019 15:28:16 +0200
From:   Ido Schimmel <idosch@...sch.org>
To:     netdev@...r.kernel.org
Cc:     davem@...emloft.net, dsahern@...il.com, roopa@...ulusnetworks.com,
        jakub.kicinski@...ronome.com, jiri@...lanox.com,
        Ido Schimmel <idosch@...lanox.com>
Subject: [PATCH net-next 5/9] ipv6: Only Replay routes of interest to new listeners

From: Ido Schimmel <idosch@...lanox.com>

When a new listener is registered to the FIB notification chain it
receives a dump of all the available routes in the system. Instead, make
sure to only replay the IPv6 routes that are actually used in the data
path and are of any interest to the new listener.

This is done by iterating over all the routing tables in the given
namespace, but from each traversed node only the first route ('leaf') is
notified. Multipath routes are notified in a single notification instead
of one for each nexthop.

Add fib6_rt_dump_tmp() to do that. Later on in the patch set it will be
renamed to fib6_rt_dump() instead of the existing one.

Signed-off-by: Ido Schimmel <idosch@...lanox.com>
Reviewed-by: Jiri Pirko <jiri@...lanox.com>
---
 net/ipv6/ip6_fib.c | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index 7cf9554888b0..51cf848e38f0 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -370,6 +370,21 @@ static int call_fib6_entry_notifier(struct notifier_block *nb,
 	return call_fib6_notifier(nb, event_type, &info.info);
 }
 
+static int call_fib6_multipath_entry_notifier(struct notifier_block *nb,
+					      enum fib_event_type event_type,
+					      struct fib6_info *rt,
+					      unsigned int nsiblings,
+					      struct netlink_ext_ack *extack)
+{
+	struct fib6_entry_notifier_info info = {
+		.info.extack = extack,
+		.rt = rt,
+		.nsiblings = nsiblings,
+	};
+
+	return call_fib6_notifier(nb, event_type, &info.info);
+}
+
 int call_fib6_entry_notifiers(struct net *net,
 			      enum fib_event_type event_type,
 			      struct fib6_info *rt,
@@ -414,16 +429,41 @@ static int fib6_rt_dump(struct fib6_info *rt, struct fib6_dump_arg *arg)
 					rt, arg->extack);
 }
 
+static int fib6_rt_dump_tmp(struct fib6_info *rt, struct fib6_dump_arg *arg)
+{
+	enum fib_event_type fib_event = FIB_EVENT_ENTRY_REPLACE_TMP;
+	int err;
+
+	if (!rt || rt == arg->net->ipv6.fib6_null_entry)
+		return 0;
+
+	if (rt->fib6_nsiblings)
+		err = call_fib6_multipath_entry_notifier(arg->nb, fib_event,
+							 rt,
+							 rt->fib6_nsiblings,
+							 arg->extack);
+	else
+		err = call_fib6_entry_notifier(arg->nb, fib_event, rt,
+					       arg->extack);
+
+	return err;
+}
+
 static int fib6_node_dump(struct fib6_walker *w)
 {
 	struct fib6_info *rt;
 	int err = 0;
 
+	err = fib6_rt_dump_tmp(w->leaf, w->args);
+	if (err)
+		goto out;
+
 	for_each_fib6_walker_rt(w) {
 		err = fib6_rt_dump(rt, w->args);
 		if (err)
 			break;
 	}
+out:
 	w->leaf = NULL;
 	return err;
 }
-- 
2.24.1