diff mbox series

[net-next,v2] net-loopback: allow lo dev initial state to be controlled

Message ID 20201117231412.2054663-1-jianyang.kernel@gmail.com (mailing list archive)
State Deferred
Delegated to: Netdev Maintainers
Headers show
Series [net-next,v2] net-loopback: allow lo dev initial state to be controlled | expand

Checks

Context Check Description
netdev/cover_letter success Link
netdev/fixes_present success Link
netdev/patch_count success Link
netdev/tree_selection success Clearly marked for net-next
netdev/subject_prefix success Link
netdev/source_inline success Was 0 now: 0
netdev/verify_signedoff success Link
netdev/module_param success Was 0 now: 0
netdev/build_32bit success Errors and warnings before: 7729 this patch: 7729
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/verify_fixes success Link
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 65 lines checked
netdev/build_allmodconfig_warn success Errors and warnings before: 8097 this patch: 8097
netdev/header_inline success Link
netdev/stable success Stable not CCed

Commit Message

Jian Yang Nov. 17, 2020, 11:14 p.m. UTC 
From: Mahesh Bandewar <maheshb@google.com>

Traditionally loopback devices comes up with initial state as DOWN for
any new network-namespace. This would mean that anyone needing this
device (which is mostly true except sandboxes where networking in not
needed at all), would have to bring this UP by issuing something like
'ip link set lo up' which can be avoided if the initial state can be set
as UP. Also ICMP error propagation needs loopback to be UP.

The default value for this sysctl is set to ZERO which will preserve the
backward compatible behavior for the root-netns while changing the
sysctl will only alter the behavior of the newer network namespaces.

Signed-off-by: Mahesh Bandewar <maheshb@google.com>
Signed-off-by: Jian Yang <jianyang@google.com>
Reported-by: kernel test robot <lkp@intel.com>
---
v2:
  * Updated sysctl name from `netdev_loopback_state` to `loopback_init_state`
  * Fixed the linking error when CONFIG_SYSCTL is not defined

 Documentation/admin-guide/sysctl/net.rst | 11 +++++++++++
 drivers/net/loopback.c                   |  9 +++++++++
 include/linux/netdevice.h                |  1 +
 net/core/sysctl_net_core.c               | 14 ++++++++++++++
 4 files changed, 35 insertions(+)
diff mbox series

Patch

diff --git a/Documentation/admin-guide/sysctl/net.rst b/Documentation/admin-guide/sysctl/net.rst
index f2ab8a5b6a4b..76698371d86e 100644
--- a/Documentation/admin-guide/sysctl/net.rst
+++ b/Documentation/admin-guide/sysctl/net.rst
@@ -268,6 +268,17 @@  Maximum number of microseconds in one NAPI polling cycle. Polling
 will exit when either netdev_budget_usecs have elapsed during the
 poll cycle or the number of packets processed reaches netdev_budget.
 
+loopback_init_state
+---------------------
+
+Controls the loopback device initial state for any new network namespaces. By
+default, we keep the initial state as DOWN.
+
+If set to 1, the loopback device will be brought UP during namespace creation.
+This will only apply to all new network namespaces.
+
+Default : 0  (for compatibility reasons)
+
 netdev_max_backlog
 ------------------
 
diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index a1c77cc00416..bf6f20095ebe 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -219,6 +219,15 @@  static __net_init int loopback_net_init(struct net *net)
 
 	BUG_ON(dev->ifindex != LOOPBACK_IFINDEX);
 	net->loopback_dev = dev;
+
+#ifdef CONFIG_SYSCTL
+	if (sysctl_loopback_init_state) {
+		/* Bring loopback device UP */
+		rtnl_lock();
+		dev_open(dev, NULL);
+		rtnl_unlock();
+	}
+#endif
 	return 0;
 
 out_free_netdev:
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 7ce648a564f7..172a6f9eb517 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -625,6 +625,7 @@  struct netdev_queue {
 
 extern int sysctl_fb_tunnels_only_for_init_net;
 extern int sysctl_devconf_inherit_init_net;
+extern int sysctl_loopback_init_state;
 
 /*
  * sysctl_fb_tunnels_only_for_init_net == 0 : For all netns
diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c
index d86d8d11cfe4..224f4c3c6c87 100644
--- a/net/core/sysctl_net_core.c
+++ b/net/core/sysctl_net_core.c
@@ -35,6 +35,11 @@  static int net_msg_warn;	/* Unused, but still a sysctl */
 int sysctl_fb_tunnels_only_for_init_net __read_mostly = 0;
 EXPORT_SYMBOL(sysctl_fb_tunnels_only_for_init_net);
 
+/* 0 - default (backward compatible) state: DOWN by default
+ * 1 - UP by default (for all new network namespaces)
+ */
+int sysctl_loopback_init_state __read_mostly;
+
 /* 0 - Keep current behavior:
  *     IPv4: inherit all current settings from init_net
  *     IPv6: reset all settings to default
@@ -507,6 +512,15 @@  static struct ctl_table net_core_table[] = {
 		.proc_handler	= set_default_qdisc
 	},
 #endif
+	{
+		.procname	= "loopback_init_state",
+		.data		= &sysctl_loopback_init_state,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= SYSCTL_ZERO,
+		.extra2		= SYSCTL_ONE
+	},
 #endif /* CONFIG_NET */
 	{
 		.procname	= "netdev_budget",