netdev - [PATCH] net: sched: check tc_skip

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-next>] [day] [month] [year] [list]

Message-Id: <20211028135644.2258-1-xiangxia.m.yue@gmail.com>
Date:   Thu, 28 Oct 2021 21:56:44 +0800
From:   xiangxia.m.yue@...il.com
To:     netdev@...r.kernel.org
Cc:     Tonghao Zhang <xiangxia.m.yue@...il.com>,
        Willem de Bruijn <willemb@...gle.com>,
        Cong Wang <xiyou.wangcong@...il.com>,
        Jakub Kicinski <kuba@...nel.org>
Subject: [PATCH] net: sched: check tc_skip_classify as far as possible

From: Tonghao Zhang <xiangxia.m.yue@...il.com>

We look up and then check tc_skip_classify flag in net
sched layer, even though skb don't want to be classified.
That case may consume a lot of cpu cycles.

Install the rules as below:
$ for id in $(seq 1 100); do
$ 	tc filter add ... egress prio $id ... action mirred egress redirect dev ifb0
$ done

netperf:
$ taskset -c 1 netperf -t TCP_RR -H ip -- -r 32,32
$ taskset -c 1 netperf -t TCP_STREAM -H ip -- -m 32

Without this patch:
10662.33 tps
108.95 Mbit/s

With this patch:
12434.48 tps
145.89 Mbit/s

For TCP_RR, there are 16.6% improvement, TCP_STREAM 33.9%.

Cc: Willem de Bruijn <willemb@...gle.com>
Cc: Cong Wang <xiyou.wangcong@...il.com>
Cc: Jakub Kicinski <kuba@...nel.org>
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@...il.com>
---
 net/core/dev.c      | 3 ++-
 net/sched/act_api.c | 3 ---
 2 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/net/core/dev.c b/net/core/dev.c
index eb61a8821b3a..856ac1fb75b4 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4155,7 +4155,8 @@ static int __dev_queue_xmit(struct sk_buff *skb, struct net_device *sb_dev)
 #ifdef CONFIG_NET_CLS_ACT
 	skb->tc_at_ingress = 0;
 # ifdef CONFIG_NET_EGRESS
-	if (static_branch_unlikely(&egress_needed_key)) {
+	if (static_branch_unlikely(&egress_needed_key) &&
+	    !skb_skip_tc_classify(skb)) {
 		skb = sch_handle_egress(skb, &rc, dev);
 		if (!skb)
 			goto out;
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index 7dd3a2dc5fa4..bd66f27178be 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -722,9 +722,6 @@ int tcf_action_exec(struct sk_buff *skb, struct tc_action **actions,
 	int i;
 	int ret = TC_ACT_OK;
 
-	if (skb_skip_tc_classify(skb))
-		return TC_ACT_OK;
-
 restart_act_graph:
 	for (i = 0; i < nr_actions; i++) {
 		const struct tc_action *a = actions[i];
-- 
2.27.0