linux/net/ipv4/devinet.c

1688 lines
40 KiB
C
Raw Normal View History

/*
* NET3 IP device support routines.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version
* 2 of the License, or (at your option) any later version.
*
* Derived from the IP parts of dev.c 1.0.19
* Authors: Ross Biro
* Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
* Mark Evans, <evansmp@uhura.aston.ac.uk>
*
* Additional Authors:
* Alan Cox, <gw4pts@gw4pts.ampr.org>
* Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
*
* Changes:
* Alexey Kuznetsov: pa_* fields are replaced with ifaddr
* lists.
* Cyrus Durgin: updated for kmod
* Matthias Andree: in devinet_ioctl, compare label and
* address (4.4BSD alias style support),
* fall back to comparing just the label
* if no match found.
*/
#include <asm/uaccess.h>
#include <asm/system.h>
#include <linux/bitops.h>
#include <linux/capability.h>
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/mm.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/in.h>
#include <linux/errno.h>
#include <linux/interrupt.h>
#include <linux/if_addr.h>
#include <linux/if_ether.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <linux/init.h>
#include <linux/notifier.h>
#include <linux/inetdevice.h>
#include <linux/igmp.h>
#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif
#include <linux/kmod.h>
#include <net/arp.h>
#include <net/ip.h>
#include <net/route.h>
#include <net/ip_fib.h>
#include <net/rtnetlink.h>
#include <net/net_namespace.h>
static struct ipv4_devconf ipv4_devconf = {
.data = {
[NET_IPV4_CONF_ACCEPT_REDIRECTS - 1] = 1,
[NET_IPV4_CONF_SEND_REDIRECTS - 1] = 1,
[NET_IPV4_CONF_SECURE_REDIRECTS - 1] = 1,
[NET_IPV4_CONF_SHARED_MEDIA - 1] = 1,
},
};
static struct ipv4_devconf ipv4_devconf_dflt = {
.data = {
[NET_IPV4_CONF_ACCEPT_REDIRECTS - 1] = 1,
[NET_IPV4_CONF_SEND_REDIRECTS - 1] = 1,
[NET_IPV4_CONF_SECURE_REDIRECTS - 1] = 1,
[NET_IPV4_CONF_SHARED_MEDIA - 1] = 1,
[NET_IPV4_CONF_ACCEPT_SOURCE_ROUTE - 1] = 1,
},
};
#define IPV4_DEVCONF_DFLT(net, attr) \
IPV4_DEVCONF((*net->ipv4.devconf_dflt), attr)
static const struct nla_policy ifa_ipv4_policy[IFA_MAX+1] = {
[IFA_LOCAL] = { .type = NLA_U32 },
[IFA_ADDRESS] = { .type = NLA_U32 },
[IFA_BROADCAST] = { .type = NLA_U32 },
[IFA_LABEL] = { .type = NLA_STRING, .len = IFNAMSIZ - 1 },
};
static void rtmsg_ifa(int event, struct in_ifaddr *, struct nlmsghdr *, u32);
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
static BLOCKING_NOTIFIER_HEAD(inetaddr_chain);
static void inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
int destroy);
#ifdef CONFIG_SYSCTL
static void devinet_sysctl_register(struct in_device *idev);
static void devinet_sysctl_unregister(struct in_device *idev);
#else
static inline void devinet_sysctl_register(struct in_device *idev)
{
}
static inline void devinet_sysctl_unregister(struct in_device *idev)
{
}
#endif
/* Locks all the inet devices. */
static struct in_ifaddr *inet_alloc_ifa(void)
{
return kzalloc(sizeof(struct in_ifaddr), GFP_KERNEL);
}
static void inet_rcu_free_ifa(struct rcu_head *head)
{
struct in_ifaddr *ifa = container_of(head, struct in_ifaddr, rcu_head);
if (ifa->ifa_dev)
in_dev_put(ifa->ifa_dev);
kfree(ifa);
}
static inline void inet_free_ifa(struct in_ifaddr *ifa)
{
call_rcu(&ifa->rcu_head, inet_rcu_free_ifa);
}
void in_dev_finish_destroy(struct in_device *idev)
{
struct net_device *dev = idev->dev;
WARN_ON(idev->ifa_list);
WARN_ON(idev->mc_list);
#ifdef NET_REFCNT_DEBUG
printk(KERN_DEBUG "in_dev_finish_destroy: %p=%s\n",
idev, dev ? dev->name : "NIL");
#endif
dev_put(dev);
if (!idev->dead)
printk("Freeing alive in_device %p\n", idev);
else {
kfree(idev);
}
}
static struct in_device *inetdev_init(struct net_device *dev)
{
struct in_device *in_dev;
ASSERT_RTNL();
in_dev = kzalloc(sizeof(*in_dev), GFP_KERNEL);
if (!in_dev)
goto out;
memcpy(&in_dev->cnf, dev_net(dev)->ipv4.devconf_dflt,
sizeof(in_dev->cnf));
in_dev->cnf.sysctl = NULL;
in_dev->dev = dev;
if ((in_dev->arp_parms = neigh_parms_alloc(dev, &arp_tbl)) == NULL)
goto out_kfree;
if (IPV4_DEVCONF(in_dev->cnf, FORWARDING))
dev_disable_lro(dev);
/* Reference in_dev->dev */
dev_hold(dev);
/* Account for reference dev->ip_ptr (below) */
in_dev_hold(in_dev);
devinet_sysctl_register(in_dev);
ip_mc_init_dev(in_dev);
if (dev->flags & IFF_UP)
ip_mc_up(in_dev);
/* we can receive as soon as ip_ptr is set -- do this last */
rcu_assign_pointer(dev->ip_ptr, in_dev);
out:
return in_dev;
out_kfree:
kfree(in_dev);
in_dev = NULL;
goto out;
}
static void in_dev_rcu_put(struct rcu_head *head)
{
struct in_device *idev = container_of(head, struct in_device, rcu_head);
in_dev_put(idev);
}
static void inetdev_destroy(struct in_device *in_dev)
{
struct in_ifaddr *ifa;
struct net_device *dev;
ASSERT_RTNL();
dev = in_dev->dev;
in_dev->dead = 1;
ip_mc_destroy_dev(in_dev);
while ((ifa = in_dev->ifa_list) != NULL) {
inet_del_ifa(in_dev, &in_dev->ifa_list, 0);
inet_free_ifa(ifa);
}
dev->ip_ptr = NULL;
devinet_sysctl_unregister(in_dev);
neigh_parms_release(&arp_tbl, in_dev->arp_parms);
arp_ifdown(dev);
call_rcu(&in_dev->rcu_head, in_dev_rcu_put);
}
int inet_addr_onlink(struct in_device *in_dev, __be32 a, __be32 b)
{
rcu_read_lock();
for_primary_ifa(in_dev) {
if (inet_ifa_match(a, ifa)) {
if (!b || inet_ifa_match(b, ifa)) {
rcu_read_unlock();
return 1;
}
}
} endfor_ifa(in_dev);
rcu_read_unlock();
return 0;
}
static void __inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
int destroy, struct nlmsghdr *nlh, u32 pid)
{
struct in_ifaddr *promote = NULL;
struct in_ifaddr *ifa, *ifa1 = *ifap;
struct in_ifaddr *last_prim = in_dev->ifa_list;
struct in_ifaddr *prev_prom = NULL;
int do_promote = IN_DEV_PROMOTE_SECONDARIES(in_dev);
ASSERT_RTNL();
/* 1. Deleting primary ifaddr forces deletion all secondaries
* unless alias promotion is set
**/
if (!(ifa1->ifa_flags & IFA_F_SECONDARY)) {
struct in_ifaddr **ifap1 = &ifa1->ifa_next;
while ((ifa = *ifap1) != NULL) {
if (!(ifa->ifa_flags & IFA_F_SECONDARY) &&
ifa1->ifa_scope <= ifa->ifa_scope)
last_prim = ifa;
if (!(ifa->ifa_flags & IFA_F_SECONDARY) ||
ifa1->ifa_mask != ifa->ifa_mask ||
!inet_ifa_match(ifa1->ifa_address, ifa)) {
ifap1 = &ifa->ifa_next;
prev_prom = ifa;
continue;
}
if (!do_promote) {
*ifap1 = ifa->ifa_next;
rtmsg_ifa(RTM_DELADDR, ifa, nlh, pid);
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
blocking_notifier_call_chain(&inetaddr_chain,
NETDEV_DOWN, ifa);
inet_free_ifa(ifa);
} else {
promote = ifa;
break;
}
}
}
/* 2. Unlink it */
*ifap = ifa1->ifa_next;
/* 3. Announce address deletion */
/* Send message first, then call notifier.
At first sight, FIB update triggered by notifier
will refer to already deleted ifaddr, that could confuse
netlink listeners. It is not true: look, gated sees
that route deleted and if it still thinks that ifaddr
is valid, it will try to restore deleted routes... Grr.
So that, this order is correct.
*/
rtmsg_ifa(RTM_DELADDR, ifa1, nlh, pid);
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
blocking_notifier_call_chain(&inetaddr_chain, NETDEV_DOWN, ifa1);
if (promote) {
if (prev_prom) {
prev_prom->ifa_next = promote->ifa_next;
promote->ifa_next = last_prim->ifa_next;
last_prim->ifa_next = promote;
}
promote->ifa_flags &= ~IFA_F_SECONDARY;
rtmsg_ifa(RTM_NEWADDR, promote, nlh, pid);
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
blocking_notifier_call_chain(&inetaddr_chain,
NETDEV_UP, promote);
for (ifa = promote->ifa_next; ifa; ifa = ifa->ifa_next) {
if (ifa1->ifa_mask != ifa->ifa_mask ||
!inet_ifa_match(ifa1->ifa_address, ifa))
continue;
fib_add_ifaddr(ifa);
}
}
if (destroy)
inet_free_ifa(ifa1);
}
static void inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
int destroy)
{
__inet_del_ifa(in_dev, ifap, destroy, NULL, 0);
}
static int __inet_insert_ifa(struct in_ifaddr *ifa, struct nlmsghdr *nlh,
u32 pid)
{
struct in_device *in_dev = ifa->ifa_dev;
struct in_ifaddr *ifa1, **ifap, **last_primary;
ASSERT_RTNL();
if (!ifa->ifa_local) {
inet_free_ifa(ifa);
return 0;
}
ifa->ifa_flags &= ~IFA_F_SECONDARY;
last_primary = &in_dev->ifa_list;
for (ifap = &in_dev->ifa_list; (ifa1 = *ifap) != NULL;
ifap = &ifa1->ifa_next) {
if (!(ifa1->ifa_flags & IFA_F_SECONDARY) &&
ifa->ifa_scope <= ifa1->ifa_scope)
last_primary = &ifa1->ifa_next;
if (ifa1->ifa_mask == ifa->ifa_mask &&
inet_ifa_match(ifa1->ifa_address, ifa)) {
if (ifa1->ifa_local == ifa->ifa_local) {
inet_free_ifa(ifa);
return -EEXIST;
}
if (ifa1->ifa_scope != ifa->ifa_scope) {
inet_free_ifa(ifa);
return -EINVAL;
}
ifa->ifa_flags |= IFA_F_SECONDARY;
}
}
if (!(ifa->ifa_flags & IFA_F_SECONDARY)) {
net_srandom(ifa->ifa_local);
ifap = last_primary;
}
ifa->ifa_next = *ifap;
*ifap = ifa;
/* Send message first, then call notifier.
Notifier will trigger FIB update, so that
listeners of netlink will know about new ifaddr */
rtmsg_ifa(RTM_NEWADDR, ifa, nlh, pid);
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
blocking_notifier_call_chain(&inetaddr_chain, NETDEV_UP, ifa);
return 0;
}
static int inet_insert_ifa(struct in_ifaddr *ifa)
{
return __inet_insert_ifa(ifa, NULL, 0);
}
static int inet_set_ifa(struct net_device *dev, struct in_ifaddr *ifa)
{
struct in_device *in_dev = __in_dev_get_rtnl(dev);
ASSERT_RTNL();
if (!in_dev) {
inet_free_ifa(ifa);
return -ENOBUFS;
}
ipv4_devconf_setall(in_dev);
if (ifa->ifa_dev != in_dev) {
WARN_ON(ifa->ifa_dev);
in_dev_hold(in_dev);
ifa->ifa_dev = in_dev;
}
if (ipv4_is_loopback(ifa->ifa_local))
ifa->ifa_scope = RT_SCOPE_HOST;
return inet_insert_ifa(ifa);
}
struct in_device *inetdev_by_index(struct net *net, int ifindex)
{
struct net_device *dev;
struct in_device *in_dev = NULL;
read_lock(&dev_base_lock);
dev = __dev_get_by_index(net, ifindex);
if (dev)
in_dev = in_dev_get(dev);
read_unlock(&dev_base_lock);
return in_dev;
}
/* Called only from RTNL semaphored context. No locks. */
struct in_ifaddr *inet_ifa_byprefix(struct in_device *in_dev, __be32 prefix,
__be32 mask)
{
ASSERT_RTNL();
for_primary_ifa(in_dev) {
if (ifa->ifa_mask == mask && inet_ifa_match(prefix, ifa))
return ifa;
} endfor_ifa(in_dev);
return NULL;
}
static int inet_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
{
struct net *net = sock_net(skb->sk);
struct nlattr *tb[IFA_MAX+1];
struct in_device *in_dev;
struct ifaddrmsg *ifm;
struct in_ifaddr *ifa, **ifap;
int err = -EINVAL;
ASSERT_RTNL();
err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv4_policy);
if (err < 0)
goto errout;
ifm = nlmsg_data(nlh);
in_dev = inetdev_by_index(net, ifm->ifa_index);
if (in_dev == NULL) {
err = -ENODEV;
goto errout;
}
__in_dev_put(in_dev);
for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
ifap = &ifa->ifa_next) {
if (tb[IFA_LOCAL] &&
ifa->ifa_local != nla_get_be32(tb[IFA_LOCAL]))
continue;
if (tb[IFA_LABEL] && nla_strcmp(tb[IFA_LABEL], ifa->ifa_label))
continue;
if (tb[IFA_ADDRESS] &&
(ifm->ifa_prefixlen != ifa->ifa_prefixlen ||
!inet_ifa_match(nla_get_be32(tb[IFA_ADDRESS]), ifa)))
continue;
__inet_del_ifa(in_dev, ifap, 1, nlh, NETLINK_CB(skb).pid);
return 0;
}
err = -EADDRNOTAVAIL;
errout:
return err;
}
static struct in_ifaddr *rtm_to_ifaddr(struct net *net, struct nlmsghdr *nlh)
{
struct nlattr *tb[IFA_MAX+1];
struct in_ifaddr *ifa;
struct ifaddrmsg *ifm;
struct net_device *dev;
struct in_device *in_dev;
int err;
err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv4_policy);
if (err < 0)
goto errout;
ifm = nlmsg_data(nlh);
err = -EINVAL;
if (ifm->ifa_prefixlen > 32 || tb[IFA_LOCAL] == NULL)
goto errout;
dev = __dev_get_by_index(net, ifm->ifa_index);
err = -ENODEV;
if (dev == NULL)
goto errout;
in_dev = __in_dev_get_rtnl(dev);
err = -ENOBUFS;
if (in_dev == NULL)
goto errout;
ifa = inet_alloc_ifa();
if (ifa == NULL)
/*
* A potential indev allocation can be left alive, it stays
* assigned to its device and is destroy with it.
*/
goto errout;
ipv4_devconf_setall(in_dev);
in_dev_hold(in_dev);
if (tb[IFA_ADDRESS] == NULL)
tb[IFA_ADDRESS] = tb[IFA_LOCAL];
ifa->ifa_prefixlen = ifm->ifa_prefixlen;
ifa->ifa_mask = inet_make_mask(ifm->ifa_prefixlen);
ifa->ifa_flags = ifm->ifa_flags;
ifa->ifa_scope = ifm->ifa_scope;
ifa->ifa_dev = in_dev;
ifa->ifa_local = nla_get_be32(tb[IFA_LOCAL]);
ifa->ifa_address = nla_get_be32(tb[IFA_ADDRESS]);
if (tb[IFA_BROADCAST])
ifa->ifa_broadcast = nla_get_be32(tb[IFA_BROADCAST]);
if (tb[IFA_LABEL])
nla_strlcpy(ifa->ifa_label, tb[IFA_LABEL], IFNAMSIZ);
else
memcpy(ifa->ifa_label, dev->name, IFNAMSIZ);
return ifa;
errout:
return ERR_PTR(err);
}
static int inet_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
{
struct net *net = sock_net(skb->sk);
struct in_ifaddr *ifa;
ASSERT_RTNL();
ifa = rtm_to_ifaddr(net, nlh);
if (IS_ERR(ifa))
return PTR_ERR(ifa);
return __inet_insert_ifa(ifa, nlh, NETLINK_CB(skb).pid);
}
/*
* Determine a default network mask, based on the IP address.
*/
static __inline__ int inet_abc_len(__be32 addr)
{
int rc = -1; /* Something else, probably a multicast. */
if (ipv4_is_zeronet(addr))
rc = 0;
else {
__u32 haddr = ntohl(addr);
if (IN_CLASSA(haddr))
rc = 8;
else if (IN_CLASSB(haddr))
rc = 16;
else if (IN_CLASSC(haddr))
rc = 24;
}
return rc;
}
int devinet_ioctl(struct net *net, unsigned int cmd, void __user *arg)
{
struct ifreq ifr;
struct sockaddr_in sin_orig;
struct sockaddr_in *sin = (struct sockaddr_in *)&ifr.ifr_addr;
struct in_device *in_dev;
struct in_ifaddr **ifap = NULL;
struct in_ifaddr *ifa = NULL;
struct net_device *dev;
char *colon;
int ret = -EFAULT;
int tryaddrmatch = 0;
/*
* Fetch the caller's info block into kernel space
*/
if (copy_from_user(&ifr, arg, sizeof(struct ifreq)))
goto out;
ifr.ifr_name[IFNAMSIZ - 1] = 0;
/* save original address for comparison */
memcpy(&sin_orig, sin, sizeof(*sin));
colon = strchr(ifr.ifr_name, ':');
if (colon)
*colon = 0;
dev_load(net, ifr.ifr_name);
switch (cmd) {
case SIOCGIFADDR: /* Get interface address */
case SIOCGIFBRDADDR: /* Get the broadcast address */
case SIOCGIFDSTADDR: /* Get the destination address */
case SIOCGIFNETMASK: /* Get the netmask for the interface */
/* Note that these ioctls will not sleep,
so that we do not impose a lock.
One day we will be forced to put shlock here (I mean SMP)
*/
tryaddrmatch = (sin_orig.sin_family == AF_INET);
memset(sin, 0, sizeof(*sin));
sin->sin_family = AF_INET;
break;
case SIOCSIFFLAGS:
ret = -EACCES;
if (!capable(CAP_NET_ADMIN))
goto out;
break;
case SIOCSIFADDR: /* Set interface address (and family) */
case SIOCSIFBRDADDR: /* Set the broadcast address */
case SIOCSIFDSTADDR: /* Set the destination address */
case SIOCSIFNETMASK: /* Set the netmask for the interface */
ret = -EACCES;
if (!capable(CAP_NET_ADMIN))
goto out;
ret = -EINVAL;
if (sin->sin_family != AF_INET)
goto out;
break;
default:
ret = -EINVAL;
goto out;
}
rtnl_lock();
ret = -ENODEV;
if ((dev = __dev_get_by_name(net, ifr.ifr_name)) == NULL)
goto done;
if (colon)
*colon = ':';
if ((in_dev = __in_dev_get_rtnl(dev)) != NULL) {
if (tryaddrmatch) {
/* Matthias Andree */
/* compare label and address (4.4BSD style) */
/* note: we only do this for a limited set of ioctls
and only if the original address family was AF_INET.
This is checked above. */
for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
ifap = &ifa->ifa_next) {
if (!strcmp(ifr.ifr_name, ifa->ifa_label) &&
sin_orig.sin_addr.s_addr ==
ifa->ifa_address) {
break; /* found */
}
}
}
/* we didn't get a match, maybe the application is
4.3BSD-style and passed in junk so we fall back to
comparing just the label */
if (!ifa) {
for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
ifap = &ifa->ifa_next)
if (!strcmp(ifr.ifr_name, ifa->ifa_label))
break;
}
}
ret = -EADDRNOTAVAIL;
if (!ifa && cmd != SIOCSIFADDR && cmd != SIOCSIFFLAGS)
goto done;
switch (cmd) {
case SIOCGIFADDR: /* Get interface address */
sin->sin_addr.s_addr = ifa->ifa_local;
goto rarok;
case SIOCGIFBRDADDR: /* Get the broadcast address */
sin->sin_addr.s_addr = ifa->ifa_broadcast;
goto rarok;
case SIOCGIFDSTADDR: /* Get the destination address */
sin->sin_addr.s_addr = ifa->ifa_address;
goto rarok;
case SIOCGIFNETMASK: /* Get the netmask for the interface */
sin->sin_addr.s_addr = ifa->ifa_mask;
goto rarok;
case SIOCSIFFLAGS:
if (colon) {
ret = -EADDRNOTAVAIL;
if (!ifa)
break;
ret = 0;
if (!(ifr.ifr_flags & IFF_UP))
inet_del_ifa(in_dev, ifap, 1);
break;
}
ret = dev_change_flags(dev, ifr.ifr_flags);
break;
case SIOCSIFADDR: /* Set interface address (and family) */
ret = -EINVAL;
if (inet_abc_len(sin->sin_addr.s_addr) < 0)
break;
if (!ifa) {
ret = -ENOBUFS;
if ((ifa = inet_alloc_ifa()) == NULL)
break;
if (colon)
memcpy(ifa->ifa_label, ifr.ifr_name, IFNAMSIZ);
else
memcpy(ifa->ifa_label, dev->name, IFNAMSIZ);
} else {
ret = 0;
if (ifa->ifa_local == sin->sin_addr.s_addr)
break;
inet_del_ifa(in_dev, ifap, 0);
ifa->ifa_broadcast = 0;
ifa->ifa_scope = 0;
}
ifa->ifa_address = ifa->ifa_local = sin->sin_addr.s_addr;
if (!(dev->flags & IFF_POINTOPOINT)) {
ifa->ifa_prefixlen = inet_abc_len(ifa->ifa_address);
ifa->ifa_mask = inet_make_mask(ifa->ifa_prefixlen);
if ((dev->flags & IFF_BROADCAST) &&
ifa->ifa_prefixlen < 31)
ifa->ifa_broadcast = ifa->ifa_address |
~ifa->ifa_mask;
} else {
ifa->ifa_prefixlen = 32;
ifa->ifa_mask = inet_make_mask(32);
}
ret = inet_set_ifa(dev, ifa);
break;
case SIOCSIFBRDADDR: /* Set the broadcast address */
ret = 0;
if (ifa->ifa_broadcast != sin->sin_addr.s_addr) {
inet_del_ifa(in_dev, ifap, 0);
ifa->ifa_broadcast = sin->sin_addr.s_addr;
inet_insert_ifa(ifa);
}
break;
case SIOCSIFDSTADDR: /* Set the destination address */
ret = 0;
if (ifa->ifa_address == sin->sin_addr.s_addr)
break;
ret = -EINVAL;
if (inet_abc_len(sin->sin_addr.s_addr) < 0)
break;
ret = 0;
inet_del_ifa(in_dev, ifap, 0);
ifa->ifa_address = sin->sin_addr.s_addr;
inet_insert_ifa(ifa);
break;
case SIOCSIFNETMASK: /* Set the netmask for the interface */
/*
* The mask we set must be legal.
*/
ret = -EINVAL;
if (bad_mask(sin->sin_addr.s_addr, 0))
break;
ret = 0;
if (ifa->ifa_mask != sin->sin_addr.s_addr) {
__be32 old_mask = ifa->ifa_mask;
inet_del_ifa(in_dev, ifap, 0);
ifa->ifa_mask = sin->sin_addr.s_addr;
ifa->ifa_prefixlen = inet_mask_len(ifa->ifa_mask);
/* See if current broadcast address matches
* with current netmask, then recalculate
* the broadcast address. Otherwise it's a
* funny address, so don't touch it since
* the user seems to know what (s)he's doing...
*/
if ((dev->flags & IFF_BROADCAST) &&
(ifa->ifa_prefixlen < 31) &&
(ifa->ifa_broadcast ==
(ifa->ifa_local|~old_mask))) {
ifa->ifa_broadcast = (ifa->ifa_local |
~sin->sin_addr.s_addr);
}
inet_insert_ifa(ifa);
}
break;
}
done:
rtnl_unlock();
out:
return ret;
rarok:
rtnl_unlock();
ret = copy_to_user(arg, &ifr, sizeof(struct ifreq)) ? -EFAULT : 0;
goto out;
}
static int inet_gifconf(struct net_device *dev, char __user *buf, int len)
{
struct in_device *in_dev = __in_dev_get_rtnl(dev);
struct in_ifaddr *ifa;
struct ifreq ifr;
int done = 0;
if (!in_dev || (ifa = in_dev->ifa_list) == NULL)
goto out;
for (; ifa; ifa = ifa->ifa_next) {
if (!buf) {
done += sizeof(ifr);
continue;
}
if (len < (int) sizeof(ifr))
break;
memset(&ifr, 0, sizeof(struct ifreq));
if (ifa->ifa_label)
strcpy(ifr.ifr_name, ifa->ifa_label);
else
strcpy(ifr.ifr_name, dev->name);
(*(struct sockaddr_in *)&ifr.ifr_addr).sin_family = AF_INET;
(*(struct sockaddr_in *)&ifr.ifr_addr).sin_addr.s_addr =
ifa->ifa_local;
if (copy_to_user(buf, &ifr, sizeof(struct ifreq))) {
done = -EFAULT;
break;
}
buf += sizeof(struct ifreq);
len -= sizeof(struct ifreq);
done += sizeof(struct ifreq);
}
out:
return done;
}
__be32 inet_select_addr(const struct net_device *dev, __be32 dst, int scope)
{
__be32 addr = 0;
struct in_device *in_dev;
struct net *net = dev_net(dev);
rcu_read_lock();
in_dev = __in_dev_get_rcu(dev);
if (!in_dev)
goto no_in_dev;
for_primary_ifa(in_dev) {
if (ifa->ifa_scope > scope)
continue;
if (!dst || inet_ifa_match(dst, ifa)) {
addr = ifa->ifa_local;
break;
}
if (!addr)
addr = ifa->ifa_local;
} endfor_ifa(in_dev);
no_in_dev:
rcu_read_unlock();
if (addr)
goto out;
/* Not loopback addresses on loopback should be preferred
in this case. It is importnat that lo is the first interface
in dev_base list.
*/
read_lock(&dev_base_lock);
rcu_read_lock();
for_each_netdev(net, dev) {
if ((in_dev = __in_dev_get_rcu(dev)) == NULL)
continue;
for_primary_ifa(in_dev) {
if (ifa->ifa_scope != RT_SCOPE_LINK &&
ifa->ifa_scope <= scope) {
addr = ifa->ifa_local;
goto out_unlock_both;
}
} endfor_ifa(in_dev);
}
out_unlock_both:
read_unlock(&dev_base_lock);
rcu_read_unlock();
out:
return addr;
}
static __be32 confirm_addr_indev(struct in_device *in_dev, __be32 dst,
__be32 local, int scope)
{
int same = 0;
__be32 addr = 0;
for_ifa(in_dev) {
if (!addr &&
(local == ifa->ifa_local || !local) &&
ifa->ifa_scope <= scope) {
addr = ifa->ifa_local;
if (same)
break;
}
if (!same) {
same = (!local || inet_ifa_match(local, ifa)) &&
(!dst || inet_ifa_match(dst, ifa));
if (same && addr) {
if (local || !dst)
break;
/* Is the selected addr into dst subnet? */
if (inet_ifa_match(addr, ifa))
break;
/* No, then can we use new local src? */
if (ifa->ifa_scope <= scope) {
addr = ifa->ifa_local;
break;
}
/* search for large dst subnet for addr */
same = 0;
}
}
} endfor_ifa(in_dev);
return same? addr : 0;
}
/*
* Confirm that local IP address exists using wildcards:
* - in_dev: only on this interface, 0=any interface
* - dst: only in the same subnet as dst, 0=any dst
* - local: address, 0=autoselect the local address
* - scope: maximum allowed scope value for the local address
*/
__be32 inet_confirm_addr(struct in_device *in_dev,
__be32 dst, __be32 local, int scope)
{
__be32 addr = 0;
struct net_device *dev;
struct net *net;
if (scope != RT_SCOPE_LINK)
return confirm_addr_indev(in_dev, dst, local, scope);
net = dev_net(in_dev->dev);
read_lock(&dev_base_lock);
rcu_read_lock();
for_each_netdev(net, dev) {
if ((in_dev = __in_dev_get_rcu(dev))) {
addr = confirm_addr_indev(in_dev, dst, local, scope);
if (addr)
break;
}
}
rcu_read_unlock();
read_unlock(&dev_base_lock);
return addr;
}
/*
* Device notifier
*/
int register_inetaddr_notifier(struct notifier_block *nb)
{
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
return blocking_notifier_chain_register(&inetaddr_chain, nb);
}
int unregister_inetaddr_notifier(struct notifier_block *nb)
{
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 09:16:30 +00:00
return blocking_notifier_chain_unregister(&inetaddr_chain, nb);
}
/* Rename ifa_labels for a device name change. Make some effort to preserve existing
* alias numbering and to create unique labels if possible.
*/
static void inetdev_changename(struct net_device *dev, struct in_device *in_dev)
{
struct in_ifaddr *ifa;
int named = 0;
for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) {
char old[IFNAMSIZ], *dot;
memcpy(old, ifa->ifa_label, IFNAMSIZ);
memcpy(ifa->ifa_label, dev->name, IFNAMSIZ);
if (named++ == 0)
goto skip;
dot = strchr(old, ':');
if (dot == NULL) {
sprintf(old, ":%d", named);
dot = old;
}
if (strlen(dot) + strlen(dev->name) < IFNAMSIZ) {
strcat(ifa->ifa_label, dot);
} else {
strcpy(ifa->ifa_label + (IFNAMSIZ - strlen(dot) - 1), dot);
}
skip:
rtmsg_ifa(RTM_NEWADDR, ifa, NULL, 0);
}
}
static inline bool inetdev_valid_mtu(unsigned mtu)
{
return mtu >= 68;
}
/* Called only under RTNL semaphore */
static int inetdev_event(struct notifier_block *this, unsigned long event,
void *ptr)
{
struct net_device *dev = ptr;
struct in_device *in_dev = __in_dev_get_rtnl(dev);
ASSERT_RTNL();
if (!in_dev) {
if (event == NETDEV_REGISTER) {
in_dev = inetdev_init(dev);
if (!in_dev)
return notifier_from_errno(-ENOMEM);
if (dev->flags & IFF_LOOPBACK) {
IN_DEV_CONF_SET(in_dev, NOXFRM, 1);
IN_DEV_CONF_SET(in_dev, NOPOLICY, 1);
}
} else if (event == NETDEV_CHANGEMTU) {
/* Re-enabling IP */
if (inetdev_valid_mtu(dev->mtu))
in_dev = inetdev_init(dev);
}
goto out;
}
switch (event) {
case NETDEV_REGISTER:
printk(KERN_DEBUG "inetdev_event: bug\n");
dev->ip_ptr = NULL;
break;
case NETDEV_UP:
if (!inetdev_valid_mtu(dev->mtu))
break;
if (dev->flags & IFF_LOOPBACK) {
struct in_ifaddr *ifa;
if ((ifa = inet_alloc_ifa()) != NULL) {
ifa->ifa_local =
ifa->ifa_address = htonl(INADDR_LOOPBACK);
ifa->ifa_prefixlen = 8;
ifa->ifa_mask = inet_make_mask(8);
in_dev_hold(in_dev);
ifa->ifa_dev = in_dev;
ifa->ifa_scope = RT_SCOPE_HOST;
memcpy(ifa->ifa_label, dev->name, IFNAMSIZ);
inet_insert_ifa(ifa);
}
}
ip_mc_up(in_dev);
/* fall through */
case NETDEV_CHANGEADDR:
/* Send gratuitous ARP to notify of link change */
if (IN_DEV_ARP_NOTIFY(in_dev)) {
struct in_ifaddr *ifa = in_dev->ifa_list;
if (ifa)
arp_send(ARPOP_REQUEST, ETH_P_ARP,
ifa->ifa_address, dev,
ifa->ifa_address, NULL,
dev->dev_addr, NULL);
}
break;
case NETDEV_DOWN:
ip_mc_down(in_dev);
break;
case NETDEV_BONDING_OLDTYPE:
ip_mc_unmap(in_dev);
break;
case NETDEV_BONDING_NEWTYPE:
ip_mc_remap(in_dev);
break;
case NETDEV_CHANGEMTU:
if (inetdev_valid_mtu(dev->mtu))
break;
/* disable IP when MTU is not enough */
case NETDEV_UNREGISTER:
inetdev_destroy(in_dev);
break;
case NETDEV_CHANGENAME:
/* Do not notify about label change, this event is
* not interesting to applications using netlink.
*/
inetdev_changename(dev, in_dev);
devinet_sysctl_unregister(in_dev);
devinet_sysctl_register(in_dev);
break;
}
out:
return NOTIFY_DONE;
}
static struct notifier_block ip_netdev_notifier = {
.notifier_call = inetdev_event,
};
static inline size_t inet_nlmsg_size(void)
{
return NLMSG_ALIGN(sizeof(struct ifaddrmsg))
+ nla_total_size(4) /* IFA_ADDRESS */
+ nla_total_size(4) /* IFA_LOCAL */
+ nla_total_size(4) /* IFA_BROADCAST */
+ nla_total_size(IFNAMSIZ); /* IFA_LABEL */
}
static int inet_fill_ifaddr(struct sk_buff *skb, struct in_ifaddr *ifa,
u32 pid, u32 seq, int event, unsigned int flags)
{
struct ifaddrmsg *ifm;
struct nlmsghdr *nlh;
nlh = nlmsg_put(skb, pid, seq, event, sizeof(*ifm), flags);
if (nlh == NULL)
return -EMSGSIZE;
ifm = nlmsg_data(nlh);
ifm->ifa_family = AF_INET;
ifm->ifa_prefixlen = ifa->ifa_prefixlen;
ifm->ifa_flags = ifa->ifa_flags|IFA_F_PERMANENT;
ifm->ifa_scope = ifa->ifa_scope;
ifm->ifa_index = ifa->ifa_dev->dev->ifindex;
if (ifa->ifa_address)
NLA_PUT_BE32(skb, IFA_ADDRESS, ifa->ifa_address);
if (ifa->ifa_local)
NLA_PUT_BE32(skb, IFA_LOCAL, ifa->ifa_local);
if (ifa->ifa_broadcast)
NLA_PUT_BE32(skb, IFA_BROADCAST, ifa->ifa_broadcast);
if (ifa->ifa_label[0])
NLA_PUT_STRING(skb, IFA_LABEL, ifa->ifa_label);
return nlmsg_end(skb, nlh);
nla_put_failure:
nlmsg_cancel(skb, nlh);
return -EMSGSIZE;
}
static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
{
struct net *net = sock_net(skb->sk);
int idx, ip_idx;
struct net_device *dev;
struct in_device *in_dev;
struct in_ifaddr *ifa;
int s_ip_idx, s_idx = cb->args[0];
s_ip_idx = ip_idx = cb->args[1];
idx = 0;
for_each_netdev(net, dev) {
if (idx < s_idx)
goto cont;
if (idx > s_idx)
s_ip_idx = 0;
if ((in_dev = __in_dev_get_rtnl(dev)) == NULL)
goto cont;
for (ifa = in_dev->ifa_list, ip_idx = 0; ifa;
ifa = ifa->ifa_next, ip_idx++) {
if (ip_idx < s_ip_idx)
continue;
if (inet_fill_ifaddr(skb, ifa, NETLINK_CB(cb->skb).pid,
cb->nlh->nlmsg_seq,
RTM_NEWADDR, NLM_F_MULTI) <= 0)
goto done;
}
cont:
idx++;
}
done:
cb->args[0] = idx;
cb->args[1] = ip_idx;
return skb->len;
}
static void rtmsg_ifa(int event, struct in_ifaddr *ifa, struct nlmsghdr *nlh,
u32 pid)
{
struct sk_buff *skb;
u32 seq = nlh ? nlh->nlmsg_seq : 0;
int err = -ENOBUFS;
struct net *net;
net = dev_net(ifa->ifa_dev->dev);
skb = nlmsg_new(inet_nlmsg_size(), GFP_KERNEL);
if (skb == NULL)
goto errout;
err = inet_fill_ifaddr(skb, ifa, pid, seq, event, 0);
if (err < 0) {
/* -EMSGSIZE implies BUG in inet_nlmsg_size() */
WARN_ON(err == -EMSGSIZE);
kfree_skb(skb);
goto errout;
}
2009-02-25 07:18:28 +00:00
rtnl_notify(skb, net, pid, RTNLGRP_IPV4_IFADDR, nlh, GFP_KERNEL);
return;
errout:
if (err < 0)
rtnl_set_sk_err(net, RTNLGRP_IPV4_IFADDR, err);
}
#ifdef CONFIG_SYSCTL
static void devinet_copy_dflt_conf(struct net *net, int i)
{
struct net_device *dev;
read_lock(&dev_base_lock);
for_each_netdev(net, dev) {
struct in_device *in_dev;
rcu_read_lock();
in_dev = __in_dev_get_rcu(dev);
if (in_dev && !test_bit(i, in_dev->cnf.state))
in_dev->cnf.data[i] = net->ipv4.devconf_dflt->data[i];
rcu_read_unlock();
}
read_unlock(&dev_base_lock);
}
static void inet_forward_change(struct net *net)
{
struct net_device *dev;
int on = IPV4_DEVCONF_ALL(net, FORWARDING);
IPV4_DEVCONF_ALL(net, ACCEPT_REDIRECTS) = !on;
IPV4_DEVCONF_DFLT(net, FORWARDING) = on;
read_lock(&dev_base_lock);
for_each_netdev(net, dev) {
struct in_device *in_dev;
if (on)
dev_disable_lro(dev);
rcu_read_lock();
in_dev = __in_dev_get_rcu(dev);
if (in_dev)
IN_DEV_CONF_SET(in_dev, FORWARDING, on);
rcu_read_unlock();
}
read_unlock(&dev_base_lock);
}
static int devinet_conf_proc(ctl_table *ctl, int write,
void __user *buffer,
size_t *lenp, loff_t *ppos)
{
int ret = proc_dointvec(ctl, write, buffer, lenp, ppos);
if (write) {
struct ipv4_devconf *cnf = ctl->extra1;
struct net *net = ctl->extra2;
int i = (int *)ctl->data - cnf->data;
set_bit(i, cnf->state);
if (cnf == net->ipv4.devconf_dflt)
devinet_copy_dflt_conf(net, i);
}
return ret;
}
static int devinet_conf_sysctl(ctl_table *table,
void __user *oldval, size_t __user *oldlenp,
void __user *newval, size_t newlen)
{
struct ipv4_devconf *cnf;
struct net *net;
int *valp = table->data;
int new;
int i;
if (!newval || !newlen)
return 0;
if (newlen != sizeof(int))
return -EINVAL;
if (get_user(new, (int __user *)newval))
return -EFAULT;
if (new == *valp)
return 0;
if (oldval && oldlenp) {
size_t len;
if (get_user(len, oldlenp))
return -EFAULT;
if (len) {
if (len > table->maxlen)
len = table->maxlen;
if (copy_to_user(oldval, valp, len))
return -EFAULT;
if (put_user(len, oldlenp))
return -EFAULT;
}
}
*valp = new;
cnf = table->extra1;
net = table->extra2;
i = (int *)table->data - cnf->data;
set_bit(i, cnf->state);
if (cnf == net->ipv4.devconf_dflt)
devinet_copy_dflt_conf(net, i);
return 1;
}
static int devinet_sysctl_forward(ctl_table *ctl, int write,
void __user *buffer,
size_t *lenp, loff_t *ppos)
{
int *valp = ctl->data;
int val = *valp;
int ret = proc_dointvec(ctl, write, buffer, lenp, ppos);
if (write && *valp != val) {
struct net *net = ctl->extra2;
if (valp != &IPV4_DEVCONF_DFLT(net, FORWARDING)) {
if (!rtnl_trylock())
return restart_syscall();
if (valp == &IPV4_DEVCONF_ALL(net, FORWARDING)) {
inet_forward_change(net);
} else if (*valp) {
struct ipv4_devconf *cnf = ctl->extra1;
struct in_device *idev =
container_of(cnf, struct in_device, cnf);
dev_disable_lro(idev->dev);
}
rtnl_unlock();
rt_cache_flush(net, 0);
}
}
return ret;
}
int ipv4_doint_and_flush(ctl_table *ctl, int write,
void __user *buffer,
size_t *lenp, loff_t *ppos)
{
int *valp = ctl->data;
int val = *valp;
int ret = proc_dointvec(ctl, write, buffer, lenp, ppos);
struct net *net = ctl->extra2;
if (write && *valp != val)
rt_cache_flush(net, 0);
return ret;
}
int ipv4_doint_and_flush_strategy(ctl_table *table,
void __user *oldval, size_t __user *oldlenp,
void __user *newval, size_t newlen)
{
int ret = devinet_conf_sysctl(table, oldval, oldlenp, newval, newlen);
struct net *net = table->extra2;
if (ret == 1)
rt_cache_flush(net, 0);
return ret;
}
#define DEVINET_SYSCTL_ENTRY(attr, name, mval, proc, sysctl) \
{ \
.ctl_name = NET_IPV4_CONF_ ## attr, \
.procname = name, \
.data = ipv4_devconf.data + \
NET_IPV4_CONF_ ## attr - 1, \
.maxlen = sizeof(int), \
.mode = mval, \
.proc_handler = proc, \
.strategy = sysctl, \
.extra1 = &ipv4_devconf, \
}
#define DEVINET_SYSCTL_RW_ENTRY(attr, name) \
DEVINET_SYSCTL_ENTRY(attr, name, 0644, devinet_conf_proc, \
devinet_conf_sysctl)
#define DEVINET_SYSCTL_RO_ENTRY(attr, name) \
DEVINET_SYSCTL_ENTRY(attr, name, 0444, devinet_conf_proc, \
devinet_conf_sysctl)
#define DEVINET_SYSCTL_COMPLEX_ENTRY(attr, name, proc, sysctl) \
DEVINET_SYSCTL_ENTRY(attr, name, 0644, proc, sysctl)
#define DEVINET_SYSCTL_FLUSHING_ENTRY(attr, name) \
DEVINET_SYSCTL_COMPLEX_ENTRY(attr, name, ipv4_doint_and_flush, \
ipv4_doint_and_flush_strategy)
static struct devinet_sysctl_table {
struct ctl_table_header *sysctl_header;
struct ctl_table devinet_vars[__NET_IPV4_CONF_MAX];
char *dev_name;
} devinet_sysctl = {
.devinet_vars = {
DEVINET_SYSCTL_COMPLEX_ENTRY(FORWARDING, "forwarding",
devinet_sysctl_forward,
devinet_conf_sysctl),
DEVINET_SYSCTL_RO_ENTRY(MC_FORWARDING, "mc_forwarding"),
DEVINET_SYSCTL_RW_ENTRY(ACCEPT_REDIRECTS, "accept_redirects"),
DEVINET_SYSCTL_RW_ENTRY(SECURE_REDIRECTS, "secure_redirects"),
DEVINET_SYSCTL_RW_ENTRY(SHARED_MEDIA, "shared_media"),
DEVINET_SYSCTL_RW_ENTRY(RP_FILTER, "rp_filter"),
DEVINET_SYSCTL_RW_ENTRY(SEND_REDIRECTS, "send_redirects"),
DEVINET_SYSCTL_RW_ENTRY(ACCEPT_SOURCE_ROUTE,
"accept_source_route"),
DEVINET_SYSCTL_RW_ENTRY(PROXY_ARP, "proxy_arp"),
DEVINET_SYSCTL_RW_ENTRY(MEDIUM_ID, "medium_id"),
DEVINET_SYSCTL_RW_ENTRY(BOOTP_RELAY, "bootp_relay"),
DEVINET_SYSCTL_RW_ENTRY(LOG_MARTIANS, "log_martians"),
DEVINET_SYSCTL_RW_ENTRY(TAG, "tag"),
DEVINET_SYSCTL_RW_ENTRY(ARPFILTER, "arp_filter"),
DEVINET_SYSCTL_RW_ENTRY(ARP_ANNOUNCE, "arp_announce"),
DEVINET_SYSCTL_RW_ENTRY(ARP_IGNORE, "arp_ignore"),
DEVINET_SYSCTL_RW_ENTRY(ARP_ACCEPT, "arp_accept"),
DEVINET_SYSCTL_RW_ENTRY(ARP_NOTIFY, "arp_notify"),
DEVINET_SYSCTL_FLUSHING_ENTRY(NOXFRM, "disable_xfrm"),
DEVINET_SYSCTL_FLUSHING_ENTRY(NOPOLICY, "disable_policy"),
DEVINET_SYSCTL_FLUSHING_ENTRY(FORCE_IGMP_VERSION,
"force_igmp_version"),
DEVINET_SYSCTL_FLUSHING_ENTRY(PROMOTE_SECONDARIES,
"promote_secondaries"),
},
};
static int __devinet_sysctl_register(struct net *net, char *dev_name,
int ctl_name, struct ipv4_devconf *p)
{
int i;
struct devinet_sysctl_table *t;
#define DEVINET_CTL_PATH_DEV 3
struct ctl_path devinet_ctl_path[] = {
{ .procname = "net", .ctl_name = CTL_NET, },
{ .procname = "ipv4", .ctl_name = NET_IPV4, },
{ .procname = "conf", .ctl_name = NET_IPV4_CONF, },
{ /* to be set */ },
{ },
};
t = kmemdup(&devinet_sysctl, sizeof(*t), GFP_KERNEL);
if (!t)
goto out;
for (i = 0; i < ARRAY_SIZE(t->devinet_vars) - 1; i++) {
t->devinet_vars[i].data += (char *)p - (char *)&ipv4_devconf;
t->devinet_vars[i].extra1 = p;
t->devinet_vars[i].extra2 = net;
}
/*
* Make a copy of dev_name, because '.procname' is regarded as const
* by sysctl and we wouldn't want anyone to change it under our feet
* (see SIOCSIFNAME).
*/
t->dev_name = kstrdup(dev_name, GFP_KERNEL);
if (!t->dev_name)
goto free;
devinet_ctl_path[DEVINET_CTL_PATH_DEV].procname = t->dev_name;
devinet_ctl_path[DEVINET_CTL_PATH_DEV].ctl_name = ctl_name;
t->sysctl_header = register_net_sysctl_table(net, devinet_ctl_path,
t->devinet_vars);
if (!t->sysctl_header)
goto free_procname;
p->sysctl = t;
return 0;
free_procname:
kfree(t->dev_name);
free:
kfree(t);
out:
return -ENOBUFS;
}
static void __devinet_sysctl_unregister(struct ipv4_devconf *cnf)
{
struct devinet_sysctl_table *t = cnf->sysctl;
if (t == NULL)
return;
cnf->sysctl = NULL;
unregister_sysctl_table(t->sysctl_header);
kfree(t->dev_name);
kfree(t);
}
static void devinet_sysctl_register(struct in_device *idev)
{
neigh_sysctl_register(idev->dev, idev->arp_parms, NET_IPV4,
NET_IPV4_NEIGH, "ipv4", NULL, NULL);
__devinet_sysctl_register(dev_net(idev->dev), idev->dev->name,
idev->dev->ifindex, &idev->cnf);
}
static void devinet_sysctl_unregister(struct in_device *idev)
{
__devinet_sysctl_unregister(&idev->cnf);
neigh_sysctl_unregister(idev->arp_parms);
}
static struct ctl_table ctl_forward_entry[] = {
{
.ctl_name = NET_IPV4_FORWARD,
.procname = "ip_forward",
.data = &ipv4_devconf.data[
NET_IPV4_CONF_FORWARDING - 1],
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = devinet_sysctl_forward,
.strategy = devinet_conf_sysctl,
.extra1 = &ipv4_devconf,
.extra2 = &init_net,
},
{ },
};
static __net_initdata struct ctl_path net_ipv4_path[] = {
{ .procname = "net", .ctl_name = CTL_NET, },
{ .procname = "ipv4", .ctl_name = NET_IPV4, },
{ },
};
#endif
static __net_init int devinet_init_net(struct net *net)
{
int err;
struct ipv4_devconf *all, *dflt;
#ifdef CONFIG_SYSCTL
struct ctl_table *tbl = ctl_forward_entry;
struct ctl_table_header *forw_hdr;
#endif
err = -ENOMEM;
all = &ipv4_devconf;
dflt = &ipv4_devconf_dflt;
if (net != &init_net) {
all = kmemdup(all, sizeof(ipv4_devconf), GFP_KERNEL);
if (all == NULL)
goto err_alloc_all;
dflt = kmemdup(dflt, sizeof(ipv4_devconf_dflt), GFP_KERNEL);
if (dflt == NULL)
goto err_alloc_dflt;
#ifdef CONFIG_SYSCTL
tbl = kmemdup(tbl, sizeof(ctl_forward_entry), GFP_KERNEL);
if (tbl == NULL)
goto err_alloc_ctl;
tbl[0].data = &all->data[NET_IPV4_CONF_FORWARDING - 1];
tbl[0].extra1 = all;
tbl[0].extra2 = net;
#endif
}
#ifdef CONFIG_SYSCTL
err = __devinet_sysctl_register(net, "all",
NET_PROTO_CONF_ALL, all);
if (err < 0)
goto err_reg_all;
err = __devinet_sysctl_register(net, "default",
NET_PROTO_CONF_DEFAULT, dflt);
if (err < 0)
goto err_reg_dflt;
err = -ENOMEM;
forw_hdr = register_net_sysctl_table(net, net_ipv4_path, tbl);
if (forw_hdr == NULL)
goto err_reg_ctl;
net->ipv4.forw_hdr = forw_hdr;
#endif
net->ipv4.devconf_all = all;
net->ipv4.devconf_dflt = dflt;
return 0;
#ifdef CONFIG_SYSCTL
err_reg_ctl:
__devinet_sysctl_unregister(dflt);
err_reg_dflt:
__devinet_sysctl_unregister(all);
err_reg_all:
if (tbl != ctl_forward_entry)
kfree(tbl);
err_alloc_ctl:
#endif
if (dflt != &ipv4_devconf_dflt)
kfree(dflt);
err_alloc_dflt:
if (all != &ipv4_devconf)
kfree(all);
err_alloc_all:
return err;
}
static __net_exit void devinet_exit_net(struct net *net)
{
#ifdef CONFIG_SYSCTL
struct ctl_table *tbl;
tbl = net->ipv4.forw_hdr->ctl_table_arg;
unregister_net_sysctl_table(net->ipv4.forw_hdr);
__devinet_sysctl_unregister(net->ipv4.devconf_dflt);
__devinet_sysctl_unregister(net->ipv4.devconf_all);
kfree(tbl);
#endif
kfree(net->ipv4.devconf_dflt);
kfree(net->ipv4.devconf_all);
}
static __net_initdata struct pernet_operations devinet_ops = {
.init = devinet_init_net,
.exit = devinet_exit_net,
};
void __init devinet_init(void)
{
register_pernet_subsys(&devinet_ops);
register_gifconf(PF_INET, inet_gifconf);
register_netdevice_notifier(&ip_netdev_notifier);
rtnl_register(PF_INET, RTM_NEWADDR, inet_rtm_newaddr, NULL);
rtnl_register(PF_INET, RTM_DELADDR, inet_rtm_deladdr, NULL);
rtnl_register(PF_INET, RTM_GETADDR, NULL, inet_dump_ifaddr);
}
EXPORT_SYMBOL(in_dev_finish_destroy);
EXPORT_SYMBOL(inet_select_addr);
EXPORT_SYMBOL(inetdev_by_index);
EXPORT_SYMBOL(register_inetaddr_notifier);
EXPORT_SYMBOL(unregister_inetaddr_notifier);