Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To:     linux-kernel@vger.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        stable@vger.kernel.org, Eric Dumazet <edumazet@google.com>,
        "Eric W. Biederman" <ebiederm@xmission.com>,
        "David S. Miller" <davem@davemloft.net>
Subject: [PATCH 5.16 0986/1039] netns: add schedule point in ops_exit_list()
Date:   Mon, 24 Jan 2022 19:46:15 +0100
Message-Id: <20220124184158.418801561@linuxfoundation.org>
In-Reply-To: <20220124184125.121143506@linuxfoundation.org>
References: <20220124184125.121143506@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Eric Dumazet <edumazet@google.com>

commit 2836615aa22de55b8fca5e32fe1b27a67cda625e upstream.

When under stress, cleanup_net() can have to dismantle
netns in big numbers. ops_exit_list() currently calls
many helpers [1] that have no schedule point, and we can
end up with soft lockups, particularly on hosts
with many cpus.

Even for moderate amount of netns processed by cleanup_net()
this patch avoids latency spikes.

[1] Some of these helpers like fib_sync_up() and fib_sync_down_dev()
are very slow because net/ipv4/fib_semantics.c uses host-wide hash tables,
and ifindex is used as the only input of two hash functions.
    ifindexes tend to be the same for all netns (lo.ifindex==1 per instance)
    This will be fixed in a separate patch.

Fixes: 72ad937abd0a ("net: Add support for batching network namespace cleanups")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 net/core/net_namespace.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

--- a/net/core/net_namespace.c
+++ b/net/core/net_namespace.c
@@ -164,8 +164,10 @@ static void ops_exit_list(const struct p
 {
 	struct net *net;
 	if (ops->exit) {
-		list_for_each_entry(net, net_exit_list, exit_list)
+		list_for_each_entry(net, net_exit_list, exit_list) {
 			ops->exit(net);
+			cond_resched();
+		}
 	}
 	if (ops->exit_batch)
 		ops->exit_batch(net_exit_list);