linux-kernel - [RFC V3 PATCH 07/26] net/netpolicy: enable and disable NET policy

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 for Android: free password hash cracker in your pocket

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

Message-Id: <1473692159-4017-8-git-send-email-kan.liang@intel.com>
Date:   Mon, 12 Sep 2016 07:55:40 -0700
From:   kan.liang@...el.com
To:     davem@...emloft.net, linux-kernel@...r.kernel.org,
        netdev@...r.kernel.org
Cc:     jeffrey.t.kirsher@...el.com, mingo@...hat.com,
        peterz@...radead.org, kuznet@....inr.ac.ru, jmorris@...ei.org,
        yoshfuji@...ux-ipv6.org, kaber@...sh.net,
        akpm@...ux-foundation.org, keescook@...omium.org,
        viro@...iv.linux.org.uk, gorcunov@...nvz.org,
        john.stultz@...aro.org, aduyck@...antis.com, ben@...adent.org.uk,
        decot@...glers.com, fw@...len.de, alexander.duyck@...il.com,
        daniel@...earbox.net, tom@...bertland.com, rdunlap@...radead.org,
        xiyou.wangcong@...il.com, hannes@...essinduktion.org,
        stephen@...workplumber.org, alexei.starovoitov@...il.com,
        jesse.brandeburg@...el.com, andi@...stfloor.org,
        Kan Liang <kan.liang@...el.com>
Subject: [RFC V3 PATCH 07/26] net/netpolicy: enable and disable NET policy

From: Kan Liang <kan.liang@...el.com>

This patch introduces functions to enable and disable NET policy.

For enabling, it collects device and CPU information and setup CPU/queue
mapping. Also, for some drivers like i40e driver, it will get better
performance if setting IRQ affinity. This patch provides an option
irq_affinity for driver to set IRQ affinity in NET policy enabling.

For disabling, it removes the IRQ affinity (if set) and mapping
information.

np_lock should protect the enable and disable state. It will be done
later in this series.

Signed-off-by: Kan Liang <kan.liang@...el.com>
---
 include/linux/netpolicy.h |  1 +
 net/core/netpolicy.c      | 41 +++++++++++++++++++++++++++++++++++++++++
 2 files changed, 42 insertions(+)

diff --git a/include/linux/netpolicy.h b/include/linux/netpolicy.h
index a946b75c..bfab7b8 100644
--- a/include/linux/netpolicy.h
+++ b/include/linux/netpolicy.h
@@ -49,6 +49,7 @@ struct netpolicy_sys_info {
 struct netpolicy_info {
 	enum netpolicy_name	cur_policy;
 	unsigned long avail_policy[BITS_TO_LONGS(NET_POLICY_MAX)];
+	bool irq_affinity;
 	/* cpu and queue mapping information */
 	struct netpolicy_sys_info	sys_info;
 };
diff --git a/net/core/netpolicy.c b/net/core/netpolicy.c
index adcb5e3..a6e240f 100644
--- a/net/core/netpolicy.c
+++ b/net/core/netpolicy.c
@@ -162,6 +162,47 @@ static void netpolicy_set_affinity(struct net_device *dev)
 	}
 }
 
+static int netpolicy_disable(struct net_device *dev)
+{
+	if (dev->netpolicy->irq_affinity)
+		netpolicy_clear_affinity(dev);
+	netpolicy_free_sys_map(dev);
+
+	return 0;
+}
+
+static int netpolicy_enable(struct net_device *dev)
+{
+	int ret;
+	struct netpolicy_dev_info d_info;
+	u32 cpu;
+
+	if (WARN_ON(!dev->netpolicy))
+		return -EINVAL;
+
+	/* get driver information */
+	ret = netpolicy_get_dev_info(dev, &d_info);
+	if (ret)
+		return ret;
+
+	/* get cpu information */
+	cpu = netpolicy_get_cpu_information();
+
+	/* create sys map */
+	ret = netpolicy_update_sys_map(dev, &d_info, cpu);
+	if (ret) {
+		netpolicy_free_dev_info(&d_info);
+		return ret;
+	}
+
+	/* set irq affinity */
+	if (dev->netpolicy->irq_affinity)
+		netpolicy_set_affinity(dev);
+
+	netpolicy_free_dev_info(&d_info);
+	return 0;
+}
+
 const char *policy_name[NET_POLICY_MAX] = {
 	"NONE"
 };
-- 
2.5.5