2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-25 13:43:55 +08:00
linux-next/net/ipv6/route.c

5512 lines
134 KiB
C
Raw Normal View History

/*
* Linux INET6 implementation
* FIB front-end.
*
* Authors:
* Pedro Roque <roque@di.fc.ul.pt>
*
* 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.
*/
/* Changes:
*
* YOSHIFUJI Hideaki @USAGI
* reworked default router selection.
* - respect outgoing interface
* - select from (probably) reachable routers (i.e.
* routers in REACHABLE, STALE, DELAY or PROBE states).
* - always select the same router if it is (probably)
* reachable. otherwise, round-robin the list.
* Ville Nuorvala
* Fixed routing subtrees.
*/
#define pr_fmt(fmt) "IPv6: " fmt
#include <linux/capability.h>
#include <linux/errno.h>
#include <linux/export.h>
#include <linux/types.h>
#include <linux/times.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/route.h>
#include <linux/netdevice.h>
#include <linux/in6.h>
#include <linux/mroute6.h>
#include <linux/init.h>
#include <linux/if_arp.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/nsproxy.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include <linux/jhash.h>
#include <net/net_namespace.h>
#include <net/snmp.h>
#include <net/ipv6.h>
#include <net/ip6_fib.h>
#include <net/ip6_route.h>
#include <net/ndisc.h>
#include <net/addrconf.h>
#include <net/tcp.h>
#include <linux/rtnetlink.h>
#include <net/dst.h>
#include <net/dst_metadata.h>
#include <net/xfrm.h>
#include <net/netevent.h>
#include <net/netlink.h>
#include <net/nexthop.h>
#include <net/lwtunnel.h>
#include <net/ip_tunnels.h>
#include <net/l3mdev.h>
#include <net/ip.h>
#include <linux/uaccess.h>
#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif
static int ip6_rt_type_to_error(u8 fib6_type);
#define CREATE_TRACE_POINTS
#include <trace/events/fib6.h>
EXPORT_TRACEPOINT_SYMBOL_GPL(fib6_table_lookup);
#undef CREATE_TRACE_POINTS
enum rt6_nud_state {
RT6_NUD_FAIL_HARD = -3,
RT6_NUD_FAIL_PROBE = -2,
RT6_NUD_FAIL_DO_RR = -1,
RT6_NUD_SUCCEED = 1
};
static struct dst_entry *ip6_dst_check(struct dst_entry *dst, u32 cookie);
static unsigned int ip6_default_advmss(const struct dst_entry *dst);
static unsigned int ip6_mtu(const struct dst_entry *dst);
static struct dst_entry *ip6_negative_advice(struct dst_entry *);
static void ip6_dst_destroy(struct dst_entry *);
static void ip6_dst_ifdown(struct dst_entry *,
struct net_device *dev, int how);
static int ip6_dst_gc(struct dst_ops *ops);
static int ip6_pkt_discard(struct sk_buff *skb);
static int ip6_pkt_discard_out(struct net *net, struct sock *sk, struct sk_buff *skb);
static int ip6_pkt_prohibit(struct sk_buff *skb);
static int ip6_pkt_prohibit_out(struct net *net, struct sock *sk, struct sk_buff *skb);
static void ip6_link_failure(struct sk_buff *skb);
static void ip6_rt_update_pmtu(struct dst_entry *dst, struct sock *sk,
struct sk_buff *skb, u32 mtu);
static void rt6_do_redirect(struct dst_entry *dst, struct sock *sk,
struct sk_buff *skb);
static int rt6_score_route(struct fib6_info *rt, int oif, int strict);
static size_t rt6_nlmsg_size(struct fib6_info *rt);
static int rt6_fill_node(struct net *net, struct sk_buff *skb,
struct fib6_info *rt, struct dst_entry *dst,
struct in6_addr *dest, struct in6_addr *src,
int iif, int type, u32 portid, u32 seq,
unsigned int flags);
static struct rt6_info *rt6_find_cached_rt(struct fib6_info *rt,
struct in6_addr *daddr,
struct in6_addr *saddr);
#ifdef CONFIG_IPV6_ROUTE_INFO
static struct fib6_info *rt6_add_route_info(struct net *net,
const struct in6_addr *prefix, int prefixlen,
const struct in6_addr *gwaddr,
struct net_device *dev,
unsigned int pref);
static struct fib6_info *rt6_get_route_info(struct net *net,
const struct in6_addr *prefix, int prefixlen,
const struct in6_addr *gwaddr,
struct net_device *dev);
#endif
struct uncached_list {
spinlock_t lock;
struct list_head head;
};
static DEFINE_PER_CPU_ALIGNED(struct uncached_list, rt6_uncached_list);
xfrm: reuse uncached_list to track xdsts In early time, when freeing a xdst, it would be inserted into dst_garbage.list first. Then if it's refcnt was still held somewhere, later it would be put into dst_busy_list in dst_gc_task(). When one dev was being unregistered, the dev of these dsts in dst_busy_list would be set with loopback_dev and put this dev. So that this dev's removal wouldn't get blocked, and avoid the kmsg warning: kernel:unregister_netdevice: waiting for veth0 to become \ free. Usage count = 2 However after Commit 52df157f17e5 ("xfrm: take refcnt of dst when creating struct xfrm_dst bundle"), the xdst will not be freed with dst gc, and this warning happens. To fix it, we need to find these xdsts that are still held by others when removing the dev, and free xdst's dev and set it with loopback_dev. But unfortunately after flow_cache for xfrm was deleted, no list tracks them anymore. So we need to save these xdsts somewhere to release the xdst's dev later. To make this easier, this patch is to reuse uncached_list to track xdsts, so that the dev refcnt can be released in the event NETDEV_UNREGISTER process of fib_netdev_notifier. Thanks to Florian, we could move forward this fix quickly. Fixes: 52df157f17e5 ("xfrm: take refcnt of dst when creating struct xfrm_dst bundle") Reported-by: Jianlin Shi <jishi@redhat.com> Reported-by: Hangbin Liu <liuhangbin@gmail.com> Tested-by: Eyal Birger <eyal.birger@gmail.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
2018-02-14 19:06:02 +08:00
void rt6_uncached_list_add(struct rt6_info *rt)
{
struct uncached_list *ul = raw_cpu_ptr(&rt6_uncached_list);
rt->rt6i_uncached_list = ul;
spin_lock_bh(&ul->lock);
list_add_tail(&rt->rt6i_uncached, &ul->head);
spin_unlock_bh(&ul->lock);
}
xfrm: reuse uncached_list to track xdsts In early time, when freeing a xdst, it would be inserted into dst_garbage.list first. Then if it's refcnt was still held somewhere, later it would be put into dst_busy_list in dst_gc_task(). When one dev was being unregistered, the dev of these dsts in dst_busy_list would be set with loopback_dev and put this dev. So that this dev's removal wouldn't get blocked, and avoid the kmsg warning: kernel:unregister_netdevice: waiting for veth0 to become \ free. Usage count = 2 However after Commit 52df157f17e5 ("xfrm: take refcnt of dst when creating struct xfrm_dst bundle"), the xdst will not be freed with dst gc, and this warning happens. To fix it, we need to find these xdsts that are still held by others when removing the dev, and free xdst's dev and set it with loopback_dev. But unfortunately after flow_cache for xfrm was deleted, no list tracks them anymore. So we need to save these xdsts somewhere to release the xdst's dev later. To make this easier, this patch is to reuse uncached_list to track xdsts, so that the dev refcnt can be released in the event NETDEV_UNREGISTER process of fib_netdev_notifier. Thanks to Florian, we could move forward this fix quickly. Fixes: 52df157f17e5 ("xfrm: take refcnt of dst when creating struct xfrm_dst bundle") Reported-by: Jianlin Shi <jishi@redhat.com> Reported-by: Hangbin Liu <liuhangbin@gmail.com> Tested-by: Eyal Birger <eyal.birger@gmail.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
2018-02-14 19:06:02 +08:00
void rt6_uncached_list_del(struct rt6_info *rt)
{
if (!list_empty(&rt->rt6i_uncached)) {
struct uncached_list *ul = rt->rt6i_uncached_list;
struct net *net = dev_net(rt->dst.dev);
spin_lock_bh(&ul->lock);
list_del(&rt->rt6i_uncached);
atomic_dec(&net->ipv6.rt6_stats->fib_rt_uncache);
spin_unlock_bh(&ul->lock);
}
}
static void rt6_uncached_list_flush_dev(struct net *net, struct net_device *dev)
{
struct net_device *loopback_dev = net->loopback_dev;
int cpu;
if (dev == loopback_dev)
return;
for_each_possible_cpu(cpu) {
struct uncached_list *ul = per_cpu_ptr(&rt6_uncached_list, cpu);
struct rt6_info *rt;
spin_lock_bh(&ul->lock);
list_for_each_entry(rt, &ul->head, rt6i_uncached) {
struct inet6_dev *rt_idev = rt->rt6i_idev;
struct net_device *rt_dev = rt->dst.dev;
if (rt_idev->dev == dev) {
rt->rt6i_idev = in6_dev_get(loopback_dev);
in6_dev_put(rt_idev);
}
if (rt_dev == dev) {
rt->dst.dev = loopback_dev;
dev_hold(rt->dst.dev);
dev_put(rt_dev);
}
}
spin_unlock_bh(&ul->lock);
}
}
static inline const void *choose_neigh_daddr(const struct in6_addr *p,
struct sk_buff *skb,
const void *daddr)
{
if (!ipv6_addr_any(p))
return (const void *) p;
else if (skb)
return &ipv6_hdr(skb)->daddr;
return daddr;
}
struct neighbour *ip6_neigh_lookup(const struct in6_addr *gw,
struct net_device *dev,
struct sk_buff *skb,
const void *daddr)
{
struct neighbour *n;
daddr = choose_neigh_daddr(gw, skb, daddr);
n = __ipv6_neigh_lookup(dev, daddr);
if (n)
return n;
n = neigh_create(&nd_tbl, daddr, dev);
return IS_ERR(n) ? NULL : n;
}
static struct neighbour *ip6_dst_neigh_lookup(const struct dst_entry *dst,
struct sk_buff *skb,
const void *daddr)
{
const struct rt6_info *rt = container_of(dst, struct rt6_info, dst);
return ip6_neigh_lookup(&rt->rt6i_gateway, dst->dev, skb, daddr);
}
static void ip6_confirm_neigh(const struct dst_entry *dst, const void *daddr)
{
struct net_device *dev = dst->dev;
struct rt6_info *rt = (struct rt6_info *)dst;
daddr = choose_neigh_daddr(&rt->rt6i_gateway, NULL, daddr);
if (!daddr)
return;
if (dev->flags & (IFF_NOARP | IFF_LOOPBACK))
return;
if (ipv6_addr_is_multicast((const struct in6_addr *)daddr))
return;
__ipv6_confirm_neigh(dev, daddr);
}
static struct dst_ops ip6_dst_ops_template = {
.family = AF_INET6,
.gc = ip6_dst_gc,
.gc_thresh = 1024,
.check = ip6_dst_check,
.default_advmss = ip6_default_advmss,
.mtu = ip6_mtu,
.cow_metrics = dst_cow_metrics_generic,
.destroy = ip6_dst_destroy,
.ifdown = ip6_dst_ifdown,
.negative_advice = ip6_negative_advice,
.link_failure = ip6_link_failure,
.update_pmtu = ip6_rt_update_pmtu,
.redirect = rt6_do_redirect,
.local_out = __ip6_local_out,
.neigh_lookup = ip6_dst_neigh_lookup,
.confirm_neigh = ip6_confirm_neigh,
};
static unsigned int ip6_blackhole_mtu(const struct dst_entry *dst)
{
unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);
return mtu ? : dst->dev->mtu;
}
static void ip6_rt_blackhole_update_pmtu(struct dst_entry *dst, struct sock *sk,
struct sk_buff *skb, u32 mtu)
{
}
static void ip6_rt_blackhole_redirect(struct dst_entry *dst, struct sock *sk,
struct sk_buff *skb)
{
}
static struct dst_ops ip6_dst_blackhole_ops = {
.family = AF_INET6,
.destroy = ip6_dst_destroy,
.check = ip6_dst_check,
.mtu = ip6_blackhole_mtu,
.default_advmss = ip6_default_advmss,
.update_pmtu = ip6_rt_blackhole_update_pmtu,
.redirect = ip6_rt_blackhole_redirect,
.cow_metrics = dst_cow_metrics_generic,
.neigh_lookup = ip6_dst_neigh_lookup,
};
net: Implement read-only protection and COW'ing of metrics. Routing metrics are now copy-on-write. Initially a route entry points it's metrics at a read-only location. If a routing table entry exists, it will point there. Else it will point at the all zero metric place-holder called 'dst_default_metrics'. The writeability state of the metrics is stored in the low bits of the metrics pointer, we have two bits left to spare if we want to store more states. For the initial implementation, COW is implemented simply via kmalloc. However future enhancements will change this to place the writable metrics somewhere else, in order to increase sharing. Very likely this "somewhere else" will be the inetpeer cache. Note also that this means that metrics updates may transiently fail if we cannot COW the metrics successfully. But even by itself, this patch should decrease memory usage and increase cache locality especially for routing workloads. In those cases the read-only metric copies stay in place and never get written to. TCP workloads where metrics get updated, and those rare cases where PMTU triggers occur, will take a very slight performance hit. But that hit will be alleviated when the long-term writable metrics move to a more sharable location. Since the metrics storage went from a u32 array of RTAX_MAX entries to what is essentially a pointer, some retooling of the dst_entry layout was necessary. Most importantly, we need to preserve the alignment of the reference count so that it doesn't share cache lines with the read-mostly state, as per Eric Dumazet's alignment assertion checks. The only non-trivial bit here is the move of the 'flags' member into the writeable cacheline. This is OK since we are always accessing the flags around the same moment when we made a modification to the reference count. Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-27 12:51:05 +08:00
static const u32 ip6_template_metrics[RTAX_MAX] = {
[RTAX_HOPLIMIT - 1] = 0,
net: Implement read-only protection and COW'ing of metrics. Routing metrics are now copy-on-write. Initially a route entry points it's metrics at a read-only location. If a routing table entry exists, it will point there. Else it will point at the all zero metric place-holder called 'dst_default_metrics'. The writeability state of the metrics is stored in the low bits of the metrics pointer, we have two bits left to spare if we want to store more states. For the initial implementation, COW is implemented simply via kmalloc. However future enhancements will change this to place the writable metrics somewhere else, in order to increase sharing. Very likely this "somewhere else" will be the inetpeer cache. Note also that this means that metrics updates may transiently fail if we cannot COW the metrics successfully. But even by itself, this patch should decrease memory usage and increase cache locality especially for routing workloads. In those cases the read-only metric copies stay in place and never get written to. TCP workloads where metrics get updated, and those rare cases where PMTU triggers occur, will take a very slight performance hit. But that hit will be alleviated when the long-term writable metrics move to a more sharable location. Since the metrics storage went from a u32 array of RTAX_MAX entries to what is essentially a pointer, some retooling of the dst_entry layout was necessary. Most importantly, we need to preserve the alignment of the reference count so that it doesn't share cache lines with the read-mostly state, as per Eric Dumazet's alignment assertion checks. The only non-trivial bit here is the move of the 'flags' member into the writeable cacheline. This is OK since we are always accessing the flags around the same moment when we made a modification to the reference count. Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-27 12:51:05 +08:00
};
static const struct fib6_info fib6_null_entry_template = {
.fib6_flags = (RTF_REJECT | RTF_NONEXTHOP),
.fib6_protocol = RTPROT_KERNEL,
.fib6_metric = ~(u32)0,
.fib6_ref = ATOMIC_INIT(1),
.fib6_type = RTN_UNREACHABLE,
.fib6_metrics = (struct dst_metrics *)&dst_default_metrics,
};
static const struct rt6_info ip6_null_entry_template = {
.dst = {
.__refcnt = ATOMIC_INIT(1),
.__use = 1,
.obsolete = DST_OBSOLETE_FORCE_CHK,
.error = -ENETUNREACH,
.input = ip6_pkt_discard,
.output = ip6_pkt_discard_out,
},
.rt6i_flags = (RTF_REJECT | RTF_NONEXTHOP),
};
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
static const struct rt6_info ip6_prohibit_entry_template = {
.dst = {
.__refcnt = ATOMIC_INIT(1),
.__use = 1,
.obsolete = DST_OBSOLETE_FORCE_CHK,
.error = -EACCES,
.input = ip6_pkt_prohibit,
.output = ip6_pkt_prohibit_out,
},
.rt6i_flags = (RTF_REJECT | RTF_NONEXTHOP),
};
static const struct rt6_info ip6_blk_hole_entry_template = {
.dst = {
.__refcnt = ATOMIC_INIT(1),
.__use = 1,
.obsolete = DST_OBSOLETE_FORCE_CHK,
.error = -EINVAL,
.input = dst_discard,
.output = dst_discard_out,
},
.rt6i_flags = (RTF_REJECT | RTF_NONEXTHOP),
};
#endif
static void rt6_info_init(struct rt6_info *rt)
{
struct dst_entry *dst = &rt->dst;
memset(dst + 1, 0, sizeof(*rt) - sizeof(*dst));
INIT_LIST_HEAD(&rt->rt6i_uncached);
}
/* allocate dst with ip6_dst_ops */
struct rt6_info *ip6_dst_alloc(struct net *net, struct net_device *dev,
int flags)
{
struct rt6_info *rt = dst_alloc(&net->ipv6.ip6_dst_ops, dev,
1, DST_OBSOLETE_FORCE_CHK, flags);
if (rt) {
rt6_info_init(rt);
atomic_inc(&net->ipv6.rt6_stats->fib_rt_alloc);
}
return rt;
}
EXPORT_SYMBOL(ip6_dst_alloc);
static void ip6_dst_destroy(struct dst_entry *dst)
{
struct rt6_info *rt = (struct rt6_info *)dst;
struct fib6_info *from;
struct inet6_dev *idev;
ip_dst_metrics_put(dst);
rt6_uncached_list_del(rt);
idev = rt->rt6i_idev;
if (idev) {
rt->rt6i_idev = NULL;
in6_dev_put(idev);
}
rcu_read_lock();
from = rcu_dereference(rt->from);
rcu_assign_pointer(rt->from, NULL);
fib6_info_release(from);
rcu_read_unlock();
}
static void ip6_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
int how)
{
struct rt6_info *rt = (struct rt6_info *)dst;
struct inet6_dev *idev = rt->rt6i_idev;
struct net_device *loopback_dev =
dev_net(dev)->loopback_dev;
if (idev && idev->dev != loopback_dev) {
struct inet6_dev *loopback_idev = in6_dev_get(loopback_dev);
if (loopback_idev) {
rt->rt6i_idev = loopback_idev;
in6_dev_put(idev);
}
}
}
static bool __rt6_check_expired(const struct rt6_info *rt)
{
if (rt->rt6i_flags & RTF_EXPIRES)
return time_after(jiffies, rt->dst.expires);
else
return false;
}
static bool rt6_check_expired(const struct rt6_info *rt)
{
struct fib6_info *from;
from = rcu_dereference(rt->from);
if (rt->rt6i_flags & RTF_EXPIRES) {
if (time_after(jiffies, rt->dst.expires))
return true;
} else if (from) {
return rt->dst.obsolete != DST_OBSOLETE_FORCE_CHK ||
fib6_check_expired(from);
}
return false;
}
struct fib6_info *fib6_multipath_select(const struct net *net,
struct fib6_info *match,
struct flowi6 *fl6, int oif,
const struct sk_buff *skb,
int strict)
{
struct fib6_info *sibling, *next_sibling;
/* We might have already computed the hash for ICMPv6 errors. In such
* case it will always be non-zero. Otherwise now is the time to do it.
*/
if (!fl6->mp_hash)
fl6->mp_hash = rt6_multipath_hash(net, fl6, skb, NULL);
if (fl6->mp_hash <= atomic_read(&match->fib6_nh.nh_upper_bound))
return match;
list_for_each_entry_safe(sibling, next_sibling, &match->fib6_siblings,
fib6_siblings) {
int nh_upper_bound;
nh_upper_bound = atomic_read(&sibling->fib6_nh.nh_upper_bound);
if (fl6->mp_hash > nh_upper_bound)
continue;
if (rt6_score_route(sibling, oif, strict) < 0)
break;
match = sibling;
break;
}
return match;
}
/*
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
* Route lookup. rcu_read_lock() should be held.
*/
static inline struct fib6_info *rt6_device_match(struct net *net,
struct fib6_info *rt,
const struct in6_addr *saddr,
int oif,
int flags)
{
struct fib6_info *sprt;
if (!oif && ipv6_addr_any(saddr) &&
!(rt->fib6_nh.nh_flags & RTNH_F_DEAD))
return rt;
for (sprt = rt; sprt; sprt = rcu_dereference(sprt->fib6_next)) {
const struct net_device *dev = sprt->fib6_nh.nh_dev;
if (sprt->fib6_nh.nh_flags & RTNH_F_DEAD)
continue;
if (oif) {
if (dev->ifindex == oif)
return sprt;
} else {
if (ipv6_chk_addr(net, saddr, dev,
flags & RT6_LOOKUP_F_IFACE))
return sprt;
}
}
if (oif && flags & RT6_LOOKUP_F_IFACE)
return net->ipv6.fib6_null_entry;
return rt->fib6_nh.nh_flags & RTNH_F_DEAD ? net->ipv6.fib6_null_entry : rt;
}
#ifdef CONFIG_IPV6_ROUTER_PREF
struct __rt6_probe_work {
struct work_struct work;
struct in6_addr target;
struct net_device *dev;
};
static void rt6_probe_deferred(struct work_struct *w)
{
struct in6_addr mcaddr;
struct __rt6_probe_work *work =
container_of(w, struct __rt6_probe_work, work);
addrconf_addr_solict_mult(&work->target, &mcaddr);
ndisc_send_ns(work->dev, &work->target, &mcaddr, NULL, 0);
dev_put(work->dev);
kfree(work);
}
static void rt6_probe(struct fib6_info *rt)
{
struct __rt6_probe_work *work = NULL;
const struct in6_addr *nh_gw;
struct neighbour *neigh;
struct net_device *dev;
struct inet6_dev *idev;
/*
* Okay, this does not seem to be appropriate
* for now, however, we need to check if it
* is really so; aka Router Reachability Probing.
*
* Router Reachability Probe MUST be rate-limited
* to no more than one per minute.
*/
if (!rt || !(rt->fib6_flags & RTF_GATEWAY))
return;
nh_gw = &rt->fib6_nh.nh_gw;
dev = rt->fib6_nh.nh_dev;
rcu_read_lock_bh();
idev = __in6_dev_get(dev);
neigh = __ipv6_neigh_lookup_noref(dev, nh_gw);
if (neigh) {
if (neigh->nud_state & NUD_VALID)
goto out;
write_lock(&neigh->lock);
if (!(neigh->nud_state & NUD_VALID) &&
time_after(jiffies,
neigh->updated + idev->cnf.rtr_probe_interval)) {
work = kmalloc(sizeof(*work), GFP_ATOMIC);
if (work)
__neigh_set_probe_once(neigh);
}
write_unlock(&neigh->lock);
} else if (time_after(jiffies, rt->last_probe +
idev->cnf.rtr_probe_interval)) {
work = kmalloc(sizeof(*work), GFP_ATOMIC);
}
if (work) {
rt->last_probe = jiffies;
INIT_WORK(&work->work, rt6_probe_deferred);
work->target = *nh_gw;
dev_hold(dev);
work->dev = dev;
schedule_work(&work->work);
}
out:
rcu_read_unlock_bh();
}
#else
static inline void rt6_probe(struct fib6_info *rt)
{
}
#endif
/*
* Default Router Selection (RFC 2461 6.3.6)
*/
static inline int rt6_check_dev(struct fib6_info *rt, int oif)
{
const struct net_device *dev = rt->fib6_nh.nh_dev;
if (!oif || dev->ifindex == oif)
return 2;
return 0;
}
static inline enum rt6_nud_state rt6_check_neigh(struct fib6_info *rt)
{
enum rt6_nud_state ret = RT6_NUD_FAIL_HARD;
struct neighbour *neigh;
if (rt->fib6_flags & RTF_NONEXTHOP ||
!(rt->fib6_flags & RTF_GATEWAY))
return RT6_NUD_SUCCEED;
rcu_read_lock_bh();
neigh = __ipv6_neigh_lookup_noref(rt->fib6_nh.nh_dev,
&rt->fib6_nh.nh_gw);
if (neigh) {
read_lock(&neigh->lock);
if (neigh->nud_state & NUD_VALID)
ret = RT6_NUD_SUCCEED;
#ifdef CONFIG_IPV6_ROUTER_PREF
else if (!(neigh->nud_state & NUD_FAILED))
ret = RT6_NUD_SUCCEED;
else
ret = RT6_NUD_FAIL_PROBE;
#endif
read_unlock(&neigh->lock);
} else {
ret = IS_ENABLED(CONFIG_IPV6_ROUTER_PREF) ?
RT6_NUD_SUCCEED : RT6_NUD_FAIL_DO_RR;
}
rcu_read_unlock_bh();
return ret;
}
static int rt6_score_route(struct fib6_info *rt, int oif, int strict)
{
int m;
m = rt6_check_dev(rt, oif);
if (!m && (strict & RT6_LOOKUP_F_IFACE))
return RT6_NUD_FAIL_HARD;
#ifdef CONFIG_IPV6_ROUTER_PREF
m |= IPV6_DECODE_PREF(IPV6_EXTRACT_PREF(rt->fib6_flags)) << 2;
#endif
if (strict & RT6_LOOKUP_F_REACHABLE) {
int n = rt6_check_neigh(rt);
if (n < 0)
return n;
}
return m;
}
/* called with rc_read_lock held */
static inline bool fib6_ignore_linkdown(const struct fib6_info *f6i)
{
const struct net_device *dev = fib6_info_nh_dev(f6i);
bool rc = false;
if (dev) {
const struct inet6_dev *idev = __in6_dev_get(dev);
rc = !!idev->cnf.ignore_routes_with_linkdown;
}
return rc;
}
static struct fib6_info *find_match(struct fib6_info *rt, int oif, int strict,
int *mpri, struct fib6_info *match,
bool *do_rr)
{
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
int m;
bool match_do_rr = false;
net: ipv6 sysctl option to ignore routes when nexthop link is down Like the ipv4 patch with a similar title, this adds a sysctl to allow the user to change routing behavior based on whether or not the interface associated with the nexthop was an up or down link. The default setting preserves the current behavior, but anyone that enables it will notice that nexthops on down interfaces will no longer be selected: net.ipv6.conf.all.ignore_routes_with_linkdown = 0 net.ipv6.conf.default.ignore_routes_with_linkdown = 0 net.ipv6.conf.lo.ignore_routes_with_linkdown = 0 ... When the above sysctls are set, not only will link status be reported to userspace, but an indication that a nexthop is dead and will not be used is also reported. 1000::/8 via 7000::2 dev p7p1 metric 1024 dead linkdown pref medium 1000::/8 via 8000::2 dev p8p1 metric 1024 pref medium 7000::/8 dev p7p1 proto kernel metric 256 dead linkdown pref medium 8000::/8 dev p8p1 proto kernel metric 256 pref medium 9000::/8 via 8000::2 dev p8p1 metric 2048 pref medium 9000::/8 via 7000::2 dev p7p1 metric 1024 dead linkdown pref medium fe80::/64 dev p7p1 proto kernel metric 256 dead linkdown pref medium fe80::/64 dev p8p1 proto kernel metric 256 pref medium This also adds devconf support and notification when sysctl values change. v2: drop use of rt6i_nhflags since it is not needed right now Signed-off-by: Andy Gospodarek <gospo@cumulusnetworks.com> Signed-off-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-08-13 22:39:01 +08:00
if (rt->fib6_nh.nh_flags & RTNH_F_DEAD)
goto out;
if (fib6_ignore_linkdown(rt) &&
rt->fib6_nh.nh_flags & RTNH_F_LINKDOWN &&
!(strict & RT6_LOOKUP_F_IGNORE_LINKSTATE))
net: ipv6 sysctl option to ignore routes when nexthop link is down Like the ipv4 patch with a similar title, this adds a sysctl to allow the user to change routing behavior based on whether or not the interface associated with the nexthop was an up or down link. The default setting preserves the current behavior, but anyone that enables it will notice that nexthops on down interfaces will no longer be selected: net.ipv6.conf.all.ignore_routes_with_linkdown = 0 net.ipv6.conf.default.ignore_routes_with_linkdown = 0 net.ipv6.conf.lo.ignore_routes_with_linkdown = 0 ... When the above sysctls are set, not only will link status be reported to userspace, but an indication that a nexthop is dead and will not be used is also reported. 1000::/8 via 7000::2 dev p7p1 metric 1024 dead linkdown pref medium 1000::/8 via 8000::2 dev p8p1 metric 1024 pref medium 7000::/8 dev p7p1 proto kernel metric 256 dead linkdown pref medium 8000::/8 dev p8p1 proto kernel metric 256 pref medium 9000::/8 via 8000::2 dev p8p1 metric 2048 pref medium 9000::/8 via 7000::2 dev p7p1 metric 1024 dead linkdown pref medium fe80::/64 dev p7p1 proto kernel metric 256 dead linkdown pref medium fe80::/64 dev p8p1 proto kernel metric 256 pref medium This also adds devconf support and notification when sysctl values change. v2: drop use of rt6i_nhflags since it is not needed right now Signed-off-by: Andy Gospodarek <gospo@cumulusnetworks.com> Signed-off-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-08-13 22:39:01 +08:00
goto out;
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
if (fib6_check_expired(rt))
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
goto out;
m = rt6_score_route(rt, oif, strict);
if (m == RT6_NUD_FAIL_DO_RR) {
match_do_rr = true;
m = 0; /* lowest valid score */
} else if (m == RT6_NUD_FAIL_HARD) {
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
goto out;
}
if (strict & RT6_LOOKUP_F_REACHABLE)
rt6_probe(rt);
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
/* note that m can be RT6_NUD_FAIL_PROBE at this point */
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
if (m > *mpri) {
*do_rr = match_do_rr;
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
*mpri = m;
match = rt;
}
out:
return match;
}
static struct fib6_info *find_rr_leaf(struct fib6_node *fn,
struct fib6_info *leaf,
struct fib6_info *rr_head,
u32 metric, int oif, int strict,
bool *do_rr)
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
{
struct fib6_info *rt, *match, *cont;
int mpri = -1;
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
match = NULL;
cont = NULL;
for (rt = rr_head; rt; rt = rcu_dereference(rt->fib6_next)) {
if (rt->fib6_metric != metric) {
cont = rt;
break;
}
match = find_match(rt, oif, strict, &mpri, match, do_rr);
}
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
for (rt = leaf; rt && rt != rr_head;
rt = rcu_dereference(rt->fib6_next)) {
if (rt->fib6_metric != metric) {
cont = rt;
break;
}
match = find_match(rt, oif, strict, &mpri, match, do_rr);
}
if (match || !cont)
return match;
for (rt = cont; rt; rt = rcu_dereference(rt->fib6_next))
match = find_match(rt, oif, strict, &mpri, match, do_rr);
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
return match;
}
static struct fib6_info *rt6_select(struct net *net, struct fib6_node *fn,
int oif, int strict)
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
{
struct fib6_info *leaf = rcu_dereference(fn->leaf);
struct fib6_info *match, *rt0;
bool do_rr = false;
int key_plen;
if (!leaf || leaf == net->ipv6.fib6_null_entry)
return net->ipv6.fib6_null_entry;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rt0 = rcu_dereference(fn->rr_ptr);
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
if (!rt0)
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rt0 = leaf;
/* Double check to make sure fn is not an intermediate node
* and fn->leaf does not points to its child's leaf
* (This might happen if all routes under fn are deleted from
* the tree and fib6_repair_tree() is called on the node.)
*/
key_plen = rt0->fib6_dst.plen;
#ifdef CONFIG_IPV6_SUBTREES
if (rt0->fib6_src.plen)
key_plen = rt0->fib6_src.plen;
#endif
if (fn->fn_bit != key_plen)
return net->ipv6.fib6_null_entry;
match = find_rr_leaf(fn, leaf, rt0, rt0->fib6_metric, oif, strict,
&do_rr);
if (do_rr) {
struct fib6_info *next = rcu_dereference(rt0->fib6_next);
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
/* no entries matched; do round-robin */
if (!next || next->fib6_metric != rt0->fib6_metric)
next = leaf;
[IPV6]: Fix routing round-robin locking. As per RFC2461, section 6.3.6, item #2, when no routers on the matching list are known to be reachable or probably reachable we do round robin on those available routes so that we make sure to probe as many of them as possible to detect when one becomes reachable faster. Each routing table has a rwlock protecting the tree and the linked list of routes at each leaf. The round robin code executes during lookup and thus with the rwlock taken as a reader. A small local spinlock tries to provide protection but this does not work at all for two reasons: 1) The round-robin list manipulation, as coded, goes like this (with read lock held): walk routes finding head and tail spin_lock(); rotate list using head and tail spin_unlock(); While one thread is rotating the list, another thread can end up with stale values of head and tail and then proceed to corrupt the list when it gets the lock. This ends up causing the OOPS in fib6_add() later onthat many people have been hitting. 2) All the other code paths that run with the rwlock held as a reader do not expect the list to change on them, they expect it to remain completely fixed while they hold the lock in that way. So, simply stated, it is impossible to implement this correctly using a manipulation of the list without violating the rwlock locking semantics. Reimplement using a per-fib6_node round-robin pointer. This way we don't need to manipulate the list at all, and since the round-robin pointer can only ever point to real existing entries we don't need to perform any locking on the changing of the round-robin pointer itself. We only need to reset the round-robin pointer to NULL when the entry it is pointing to is removed. The idea is from Thomas Graf and it is very similar to how this was implemented before the advanced router selection code when in. Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-25 11:36:25 +08:00
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
if (next != rt0) {
spin_lock_bh(&leaf->fib6_table->tb6_lock);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
/* make sure next is not being deleted from the tree */
if (next->fib6_node)
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_assign_pointer(fn->rr_ptr, next);
spin_unlock_bh(&leaf->fib6_table->tb6_lock);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
}
}
return match ? match : net->ipv6.fib6_null_entry;
}
static bool rt6_is_gw_or_nonexthop(const struct fib6_info *rt)
{
return (rt->fib6_flags & (RTF_NONEXTHOP | RTF_GATEWAY));
}
#ifdef CONFIG_IPV6_ROUTE_INFO
int rt6_route_rcv(struct net_device *dev, u8 *opt, int len,
const struct in6_addr *gwaddr)
{
struct net *net = dev_net(dev);
struct route_info *rinfo = (struct route_info *) opt;
struct in6_addr prefix_buf, *prefix;
unsigned int pref;
unsigned long lifetime;
struct fib6_info *rt;
if (len < sizeof(struct route_info)) {
return -EINVAL;
}
/* Sanity check for prefix_len and length */
if (rinfo->length > 3) {
return -EINVAL;
} else if (rinfo->prefix_len > 128) {
return -EINVAL;
} else if (rinfo->prefix_len > 64) {
if (rinfo->length < 2) {
return -EINVAL;
}
} else if (rinfo->prefix_len > 0) {
if (rinfo->length < 1) {
return -EINVAL;
}
}
pref = rinfo->route_pref;
if (pref == ICMPV6_ROUTER_PREF_INVALID)
return -EINVAL;
lifetime = addrconf_timeout_fixup(ntohl(rinfo->lifetime), HZ);
if (rinfo->length == 3)
prefix = (struct in6_addr *)rinfo->prefix;
else {
/* this function is safe */
ipv6_addr_prefix(&prefix_buf,
(struct in6_addr *)rinfo->prefix,
rinfo->prefix_len);
prefix = &prefix_buf;
}
if (rinfo->prefix_len == 0)
rt = rt6_get_dflt_router(net, gwaddr, dev);
else
rt = rt6_get_route_info(net, prefix, rinfo->prefix_len,
gwaddr, dev);
if (rt && !lifetime) {
ip6_del_rt(net, rt);
rt = NULL;
}
if (!rt && lifetime)
rt = rt6_add_route_info(net, prefix, rinfo->prefix_len, gwaddr,
dev, pref);
else if (rt)
rt->fib6_flags = RTF_ROUTEINFO |
(rt->fib6_flags & ~RTF_PREF_MASK) | RTF_PREF(pref);
if (rt) {
if (!addrconf_finite_timeout(lifetime))
fib6_clean_expires(rt);
else
fib6_set_expires(rt, jiffies + HZ * lifetime);
fib6_info_release(rt);
}
return 0;
}
#endif
/*
* Misc support functions
*/
/* called with rcu_lock held */
static struct net_device *ip6_rt_get_dev_rcu(struct fib6_info *rt)
{
struct net_device *dev = rt->fib6_nh.nh_dev;
if (rt->fib6_flags & (RTF_LOCAL | RTF_ANYCAST)) {
/* for copies of local routes, dst->dev needs to be the
* device if it is a master device, the master device if
* device is enslaved, and the loopback as the default
*/
if (netif_is_l3_slave(dev) &&
!rt6_need_strict(&rt->fib6_dst.addr))
dev = l3mdev_master_dev_rcu(dev);
else if (!netif_is_l3_master(dev))
dev = dev_net(dev)->loopback_dev;
/* last case is netif_is_l3_master(dev) is true in which
* case we want dev returned to be dev
*/
}
return dev;
}
static const int fib6_prop[RTN_MAX + 1] = {
[RTN_UNSPEC] = 0,
[RTN_UNICAST] = 0,
[RTN_LOCAL] = 0,
[RTN_BROADCAST] = 0,
[RTN_ANYCAST] = 0,
[RTN_MULTICAST] = 0,
[RTN_BLACKHOLE] = -EINVAL,
[RTN_UNREACHABLE] = -EHOSTUNREACH,
[RTN_PROHIBIT] = -EACCES,
[RTN_THROW] = -EAGAIN,
[RTN_NAT] = -EINVAL,
[RTN_XRESOLVE] = -EINVAL,
};
static int ip6_rt_type_to_error(u8 fib6_type)
{
return fib6_prop[fib6_type];
}
static unsigned short fib6_info_dst_flags(struct fib6_info *rt)
{
unsigned short flags = 0;
if (rt->dst_nocount)
flags |= DST_NOCOUNT;
if (rt->dst_nopolicy)
flags |= DST_NOPOLICY;
if (rt->dst_host)
flags |= DST_HOST;
return flags;
}
static void ip6_rt_init_dst_reject(struct rt6_info *rt, struct fib6_info *ort)
{
rt->dst.error = ip6_rt_type_to_error(ort->fib6_type);
switch (ort->fib6_type) {
case RTN_BLACKHOLE:
rt->dst.output = dst_discard_out;
rt->dst.input = dst_discard;
break;
case RTN_PROHIBIT:
rt->dst.output = ip6_pkt_prohibit_out;
rt->dst.input = ip6_pkt_prohibit;
break;
case RTN_THROW:
case RTN_UNREACHABLE:
default:
rt->dst.output = ip6_pkt_discard_out;
rt->dst.input = ip6_pkt_discard;
break;
}
}
static void ip6_rt_init_dst(struct rt6_info *rt, struct fib6_info *ort)
{
if (ort->fib6_flags & RTF_REJECT) {
ip6_rt_init_dst_reject(rt, ort);
return;
}
rt->dst.error = 0;
rt->dst.output = ip6_output;
if (ort->fib6_type == RTN_LOCAL || ort->fib6_type == RTN_ANYCAST) {
rt->dst.input = ip6_input;
} else if (ipv6_addr_type(&ort->fib6_dst.addr) & IPV6_ADDR_MULTICAST) {
rt->dst.input = ip6_mc_input;
} else {
rt->dst.input = ip6_forward;
}
if (ort->fib6_nh.nh_lwtstate) {
rt->dst.lwtstate = lwtstate_get(ort->fib6_nh.nh_lwtstate);
lwtunnel_set_redirect(&rt->dst);
}
rt->dst.lastuse = jiffies;
}
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
/* Caller must already hold reference to @from */
static void rt6_set_from(struct rt6_info *rt, struct fib6_info *from)
{
rt->rt6i_flags &= ~RTF_EXPIRES;
rcu_assign_pointer(rt->from, from);
ip_dst_init_metrics(&rt->dst, from->fib6_metrics);
}
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
/* Caller must already hold reference to @ort */
static void ip6_rt_copy_init(struct rt6_info *rt, struct fib6_info *ort)
{
struct net_device *dev = fib6_info_nh_dev(ort);
ip6_rt_init_dst(rt, ort);
rt->rt6i_dst = ort->fib6_dst;
rt->rt6i_idev = dev ? in6_dev_get(dev) : NULL;
rt->rt6i_gateway = ort->fib6_nh.nh_gw;
rt->rt6i_flags = ort->fib6_flags;
rt6_set_from(rt, ort);
#ifdef CONFIG_IPV6_SUBTREES
rt->rt6i_src = ort->fib6_src;
#endif
}
static struct fib6_node* fib6_backtrack(struct fib6_node *fn,
struct in6_addr *saddr)
{
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
struct fib6_node *pn, *sn;
while (1) {
if (fn->fn_flags & RTN_TL_ROOT)
return NULL;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
pn = rcu_dereference(fn->parent);
sn = FIB6_SUBTREE(pn);
if (sn && sn != fn)
fn = fib6_node_lookup(sn, NULL, saddr);
else
fn = pn;
if (fn->fn_flags & RTN_RTINFO)
return fn;
}
}
static bool ip6_hold_safe(struct net *net, struct rt6_info **prt,
bool null_fallback)
{
struct rt6_info *rt = *prt;
if (dst_hold_safe(&rt->dst))
return true;
if (null_fallback) {
rt = net->ipv6.ip6_null_entry;
dst_hold(&rt->dst);
} else {
rt = NULL;
}
*prt = rt;
return false;
}
/* called with rcu_lock held */
static struct rt6_info *ip6_create_rt_rcu(struct fib6_info *rt)
{
unsigned short flags = fib6_info_dst_flags(rt);
struct net_device *dev = rt->fib6_nh.nh_dev;
struct rt6_info *nrt;
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!fib6_info_hold_safe(rt))
return NULL;
nrt = ip6_dst_alloc(dev_net(dev), dev, flags);
if (nrt)
ip6_rt_copy_init(nrt, rt);
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
else
fib6_info_release(rt);
return nrt;
}
static struct rt6_info *ip6_pol_route_lookup(struct net *net,
struct fib6_table *table,
struct flowi6 *fl6,
const struct sk_buff *skb,
int flags)
{
struct fib6_info *f6i;
struct fib6_node *fn;
struct rt6_info *rt;
if (fl6->flowi6_flags & FLOWI_FLAG_SKIP_NH_OIF)
flags &= ~RT6_LOOKUP_F_IFACE;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_lock();
fn = fib6_node_lookup(&table->tb6_root, &fl6->daddr, &fl6->saddr);
restart:
f6i = rcu_dereference(fn->leaf);
if (!f6i) {
f6i = net->ipv6.fib6_null_entry;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
} else {
f6i = rt6_device_match(net, f6i, &fl6->saddr,
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
fl6->flowi6_oif, flags);
if (f6i->fib6_nsiblings && fl6->flowi6_oif == 0)
f6i = fib6_multipath_select(net, f6i, fl6,
fl6->flowi6_oif, skb,
flags);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
}
if (f6i == net->ipv6.fib6_null_entry) {
fn = fib6_backtrack(fn, &fl6->saddr);
if (fn)
goto restart;
}
trace_fib6_table_lookup(net, f6i, table, fl6);
/* Search through exception table */
rt = rt6_find_cached_rt(f6i, &fl6->daddr, &fl6->saddr);
if (rt) {
if (ip6_hold_safe(net, &rt, true))
dst_use_noref(&rt->dst, jiffies);
} else if (f6i == net->ipv6.fib6_null_entry) {
rt = net->ipv6.ip6_null_entry;
dst_hold(&rt->dst);
} else {
rt = ip6_create_rt_rcu(f6i);
if (!rt) {
rt = net->ipv6.ip6_null_entry;
dst_hold(&rt->dst);
}
}
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
return rt;
}
struct dst_entry *ip6_route_lookup(struct net *net, struct flowi6 *fl6,
const struct sk_buff *skb, int flags)
{
return fib6_rule_lookup(net, fl6, skb, flags, ip6_pol_route_lookup);
}
EXPORT_SYMBOL_GPL(ip6_route_lookup);
struct rt6_info *rt6_lookup(struct net *net, const struct in6_addr *daddr,
const struct in6_addr *saddr, int oif,
const struct sk_buff *skb, int strict)
{
struct flowi6 fl6 = {
.flowi6_oif = oif,
.daddr = *daddr,
};
struct dst_entry *dst;
int flags = strict ? RT6_LOOKUP_F_IFACE : 0;
if (saddr) {
memcpy(&fl6.saddr, saddr, sizeof(*saddr));
flags |= RT6_LOOKUP_F_HAS_SADDR;
}
dst = fib6_rule_lookup(net, &fl6, skb, flags, ip6_pol_route_lookup);
if (dst->error == 0)
return (struct rt6_info *) dst;
dst_release(dst);
return NULL;
}
EXPORT_SYMBOL(rt6_lookup);
/* ip6_ins_rt is called with FREE table->tb6_lock.
* It takes new route entry, the addition fails by any reason the
* route is released.
* Caller must hold dst before calling it.
*/
static int __ip6_ins_rt(struct fib6_info *rt, struct nl_info *info,
struct netlink_ext_ack *extack)
{
int err;
struct fib6_table *table;
table = rt->fib6_table;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
spin_lock_bh(&table->tb6_lock);
err = fib6_add(&table->tb6_root, rt, info, extack);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
spin_unlock_bh(&table->tb6_lock);
return err;
}
int ip6_ins_rt(struct net *net, struct fib6_info *rt)
{
struct nl_info info = { .nl_net = net, };
return __ip6_ins_rt(rt, &info, NULL);
}
static struct rt6_info *ip6_rt_cache_alloc(struct fib6_info *ort,
const struct in6_addr *daddr,
const struct in6_addr *saddr)
{
struct net_device *dev;
struct rt6_info *rt;
/*
* Clone the route.
*/
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!fib6_info_hold_safe(ort))
return NULL;
dev = ip6_rt_get_dev_rcu(ort);
rt = ip6_dst_alloc(dev_net(dev), dev, 0);
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!rt) {
fib6_info_release(ort);
return NULL;
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
}
ip6_rt_copy_init(rt, ort);
rt->rt6i_flags |= RTF_CACHE;
rt->dst.flags |= DST_HOST;
rt->rt6i_dst.addr = *daddr;
rt->rt6i_dst.plen = 128;
if (!rt6_is_gw_or_nonexthop(ort)) {
if (ort->fib6_dst.plen != 128 &&
ipv6_addr_equal(&ort->fib6_dst.addr, daddr))
rt->rt6i_flags |= RTF_ANYCAST;
#ifdef CONFIG_IPV6_SUBTREES
if (rt->rt6i_src.plen && saddr) {
rt->rt6i_src.addr = *saddr;
rt->rt6i_src.plen = 128;
}
#endif
}
return rt;
}
static struct rt6_info *ip6_rt_pcpu_alloc(struct fib6_info *rt)
{
unsigned short flags = fib6_info_dst_flags(rt);
struct net_device *dev;
struct rt6_info *pcpu_rt;
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!fib6_info_hold_safe(rt))
return NULL;
rcu_read_lock();
dev = ip6_rt_get_dev_rcu(rt);
pcpu_rt = ip6_dst_alloc(dev_net(dev), dev, flags);
rcu_read_unlock();
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!pcpu_rt) {
fib6_info_release(rt);
return NULL;
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
}
ip6_rt_copy_init(pcpu_rt, rt);
pcpu_rt->rt6i_flags |= RTF_PCPU;
return pcpu_rt;
}
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
/* It should be called with rcu_read_lock() acquired */
static struct rt6_info *rt6_get_pcpu_route(struct fib6_info *rt)
{
struct rt6_info *pcpu_rt, **p;
p = this_cpu_ptr(rt->rt6i_pcpu);
pcpu_rt = *p;
if (pcpu_rt)
ip6_hold_safe(NULL, &pcpu_rt, false);
return pcpu_rt;
}
static struct rt6_info *rt6_make_pcpu_route(struct net *net,
struct fib6_info *rt)
{
struct rt6_info *pcpu_rt, *prev, **p;
pcpu_rt = ip6_rt_pcpu_alloc(rt);
if (!pcpu_rt) {
ipv6: Fix a potential deadlock when creating pcpu rt rt6_make_pcpu_route() is called under read_lock(&table->tb6_lock). rt6_make_pcpu_route() calls ip6_rt_pcpu_alloc(rt) which then calls dst_alloc(). dst_alloc() _may_ call ip6_dst_gc() which takes the write_lock(&tabl->tb6_lock). A visualized version: read_lock(&table->tb6_lock); rt6_make_pcpu_route(); => ip6_rt_pcpu_alloc(); => dst_alloc(); => ip6_dst_gc(); => write_lock(&table->tb6_lock); /* oops */ The fix is to do a read_unlock first before calling ip6_rt_pcpu_alloc(). A reported stack: [141625.537638] INFO: rcu_sched self-detected stall on CPU { 27} (t=60000 jiffies g=4159086 c=4159085 q=2139) [141625.547469] Task dump for CPU 27: [141625.550881] mtr R running task 0 22121 22081 0x00000008 [141625.558069] 0000000000000000 ffff88103f363d98 ffffffff8106e488 000000000000001b [141625.565641] ffffffff81684900 ffff88103f363db8 ffffffff810702b0 0000000008000000 [141625.573220] ffffffff81684900 ffff88103f363de8 ffffffff8108df9f ffff88103f375a00 [141625.580803] Call Trace: [141625.583345] <IRQ> [<ffffffff8106e488>] sched_show_task+0xc1/0xc6 [141625.589650] [<ffffffff810702b0>] dump_cpu_task+0x35/0x39 [141625.595144] [<ffffffff8108df9f>] rcu_dump_cpu_stacks+0x6a/0x8c [141625.601320] [<ffffffff81090606>] rcu_check_callbacks+0x1f6/0x5d4 [141625.607669] [<ffffffff810940c8>] update_process_times+0x2a/0x4f [141625.613925] [<ffffffff8109fbee>] tick_sched_handle+0x32/0x3e [141625.619923] [<ffffffff8109fc2f>] tick_sched_timer+0x35/0x5c [141625.625830] [<ffffffff81094a1f>] __hrtimer_run_queues+0x8f/0x18d [141625.632171] [<ffffffff81094c9e>] hrtimer_interrupt+0xa0/0x166 [141625.638258] [<ffffffff8102bf2a>] local_apic_timer_interrupt+0x4e/0x52 [141625.645036] [<ffffffff8102c36f>] smp_apic_timer_interrupt+0x39/0x4a [141625.651643] [<ffffffff8140b9e8>] apic_timer_interrupt+0x68/0x70 [141625.657895] <EOI> [<ffffffff81346ee8>] ? dst_destroy+0x7c/0xb5 [141625.664188] [<ffffffff813d45b5>] ? fib6_flush_trees+0x20/0x20 [141625.670272] [<ffffffff81082b45>] ? queue_write_lock_slowpath+0x60/0x6f [141625.677140] [<ffffffff8140aa33>] _raw_write_lock_bh+0x23/0x25 [141625.683218] [<ffffffff813d4553>] __fib6_clean_all+0x40/0x82 [141625.689124] [<ffffffff813d45b5>] ? fib6_flush_trees+0x20/0x20 [141625.695207] [<ffffffff813d6058>] fib6_clean_all+0xe/0x10 [141625.700854] [<ffffffff813d60d3>] fib6_run_gc+0x79/0xc8 [141625.706329] [<ffffffff813d0510>] ip6_dst_gc+0x85/0xf9 [141625.711718] [<ffffffff81346d68>] dst_alloc+0x55/0x159 [141625.717105] [<ffffffff813d09b5>] __ip6_dst_alloc.isra.32+0x19/0x63 [141625.723620] [<ffffffff813d1830>] ip6_pol_route+0x36a/0x3e8 [141625.729441] [<ffffffff813d18d6>] ip6_pol_route_output+0x11/0x13 [141625.735700] [<ffffffff813f02c8>] fib6_rule_action+0xa7/0x1bf [141625.741698] [<ffffffff813d18c5>] ? ip6_pol_route_input+0x17/0x17 [141625.748043] [<ffffffff81357c48>] fib_rules_lookup+0xb5/0x12a [141625.754050] [<ffffffff81141628>] ? poll_select_copy_remaining+0xf9/0xf9 [141625.761002] [<ffffffff813f0535>] fib6_rule_lookup+0x37/0x5c [141625.766914] [<ffffffff813d18c5>] ? ip6_pol_route_input+0x17/0x17 [141625.773260] [<ffffffff813d008c>] ip6_route_output+0x7a/0x82 [141625.779177] [<ffffffff813c44c8>] ip6_dst_lookup_tail+0x53/0x112 [141625.785437] [<ffffffff813c45c3>] ip6_dst_lookup_flow+0x2a/0x6b [141625.791604] [<ffffffff813ddaab>] rawv6_sendmsg+0x407/0x9b6 [141625.797423] [<ffffffff813d7914>] ? do_ipv6_setsockopt.isra.8+0xd87/0xde2 [141625.804464] [<ffffffff8139d4b4>] inet_sendmsg+0x57/0x8e [141625.810028] [<ffffffff81329ba3>] sock_sendmsg+0x2e/0x3c [141625.815588] [<ffffffff8132be57>] SyS_sendto+0xfe/0x143 [141625.821063] [<ffffffff813dd551>] ? rawv6_setsockopt+0x5e/0x67 [141625.827146] [<ffffffff8132c9f8>] ? sock_common_setsockopt+0xf/0x11 [141625.833660] [<ffffffff8132c08c>] ? SyS_setsockopt+0x81/0xa2 [141625.839565] [<ffffffff8140ac17>] entry_SYSCALL_64_fastpath+0x12/0x6a Fixes: d52d3997f843 ("pv6: Create percpu rt6_info") Signed-off-by: Martin KaFai Lau <kafai@fb.com> CC: Hannes Frederic Sowa <hannes@stressinduktion.org> Reported-by: Steinar H. Gunderson <sgunderson@bigfoot.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-08-15 02:05:54 +08:00
dst_hold(&net->ipv6.ip6_null_entry->dst);
return net->ipv6.ip6_null_entry;
}
dst_hold(&pcpu_rt->dst);
p = this_cpu_ptr(rt->rt6i_pcpu);
prev = cmpxchg(p, NULL, pcpu_rt);
BUG_ON(prev);
return pcpu_rt;
}
/* exception hash table implementation
*/
static DEFINE_SPINLOCK(rt6_exception_lock);
/* Remove rt6_ex from hash table and free the memory
* Caller must hold rt6_exception_lock
*/
static void rt6_remove_exception(struct rt6_exception_bucket *bucket,
struct rt6_exception *rt6_ex)
{
struct net *net;
if (!bucket || !rt6_ex)
return;
net = dev_net(rt6_ex->rt6i->dst.dev);
hlist_del_rcu(&rt6_ex->hlist);
dst_release(&rt6_ex->rt6i->dst);
kfree_rcu(rt6_ex, rcu);
WARN_ON_ONCE(!bucket->depth);
bucket->depth--;
net->ipv6.rt6_stats->fib_rt_cache--;
}
/* Remove oldest rt6_ex in bucket and free the memory
* Caller must hold rt6_exception_lock
*/
static void rt6_exception_remove_oldest(struct rt6_exception_bucket *bucket)
{
struct rt6_exception *rt6_ex, *oldest = NULL;
if (!bucket)
return;
hlist_for_each_entry(rt6_ex, &bucket->chain, hlist) {
if (!oldest || time_before(rt6_ex->stamp, oldest->stamp))
oldest = rt6_ex;
}
rt6_remove_exception(bucket, oldest);
}
static u32 rt6_exception_hash(const struct in6_addr *dst,
const struct in6_addr *src)
{
static u32 seed __read_mostly;
u32 val;
net_get_random_once(&seed, sizeof(seed));
val = jhash(dst, sizeof(*dst), seed);
#ifdef CONFIG_IPV6_SUBTREES
if (src)
val = jhash(src, sizeof(*src), val);
#endif
return hash_32(val, FIB6_EXCEPTION_BUCKET_SIZE_SHIFT);
}
/* Helper function to find the cached rt in the hash table
* and update bucket pointer to point to the bucket for this
* (daddr, saddr) pair
* Caller must hold rt6_exception_lock
*/
static struct rt6_exception *
__rt6_find_exception_spinlock(struct rt6_exception_bucket **bucket,
const struct in6_addr *daddr,
const struct in6_addr *saddr)
{
struct rt6_exception *rt6_ex;
u32 hval;
if (!(*bucket) || !daddr)
return NULL;
hval = rt6_exception_hash(daddr, saddr);
*bucket += hval;
hlist_for_each_entry(rt6_ex, &(*bucket)->chain, hlist) {
struct rt6_info *rt6 = rt6_ex->rt6i;
bool matched = ipv6_addr_equal(daddr, &rt6->rt6i_dst.addr);
#ifdef CONFIG_IPV6_SUBTREES
if (matched && saddr)
matched = ipv6_addr_equal(saddr, &rt6->rt6i_src.addr);
#endif
if (matched)
return rt6_ex;
}
return NULL;
}
/* Helper function to find the cached rt in the hash table
* and update bucket pointer to point to the bucket for this
* (daddr, saddr) pair
* Caller must hold rcu_read_lock()
*/
static struct rt6_exception *
__rt6_find_exception_rcu(struct rt6_exception_bucket **bucket,
const struct in6_addr *daddr,
const struct in6_addr *saddr)
{
struct rt6_exception *rt6_ex;
u32 hval;
WARN_ON_ONCE(!rcu_read_lock_held());
if (!(*bucket) || !daddr)
return NULL;
hval = rt6_exception_hash(daddr, saddr);
*bucket += hval;
hlist_for_each_entry_rcu(rt6_ex, &(*bucket)->chain, hlist) {
struct rt6_info *rt6 = rt6_ex->rt6i;
bool matched = ipv6_addr_equal(daddr, &rt6->rt6i_dst.addr);
#ifdef CONFIG_IPV6_SUBTREES
if (matched && saddr)
matched = ipv6_addr_equal(saddr, &rt6->rt6i_src.addr);
#endif
if (matched)
return rt6_ex;
}
return NULL;
}
static unsigned int fib6_mtu(const struct fib6_info *rt)
{
unsigned int mtu;
if (rt->fib6_pmtu) {
mtu = rt->fib6_pmtu;
} else {
struct net_device *dev = fib6_info_nh_dev(rt);
struct inet6_dev *idev;
rcu_read_lock();
idev = __in6_dev_get(dev);
mtu = idev->cnf.mtu6;
rcu_read_unlock();
}
mtu = min_t(unsigned int, mtu, IP6_MAX_MTU);
return mtu - lwtunnel_headroom(rt->fib6_nh.nh_lwtstate, mtu);
}
static int rt6_insert_exception(struct rt6_info *nrt,
struct fib6_info *ort)
{
struct net *net = dev_net(nrt->dst.dev);
struct rt6_exception_bucket *bucket;
struct in6_addr *src_key = NULL;
struct rt6_exception *rt6_ex;
int err = 0;
spin_lock_bh(&rt6_exception_lock);
if (ort->exception_bucket_flushed) {
err = -EINVAL;
goto out;
}
bucket = rcu_dereference_protected(ort->rt6i_exception_bucket,
lockdep_is_held(&rt6_exception_lock));
if (!bucket) {
bucket = kcalloc(FIB6_EXCEPTION_BUCKET_SIZE, sizeof(*bucket),
GFP_ATOMIC);
if (!bucket) {
err = -ENOMEM;
goto out;
}
rcu_assign_pointer(ort->rt6i_exception_bucket, bucket);
}
#ifdef CONFIG_IPV6_SUBTREES
/* rt6i_src.plen != 0 indicates ort is in subtree
* and exception table is indexed by a hash of
* both rt6i_dst and rt6i_src.
* Otherwise, the exception table is indexed by
* a hash of only rt6i_dst.
*/
if (ort->fib6_src.plen)
src_key = &nrt->rt6i_src.addr;
#endif
/* rt6_mtu_change() might lower mtu on ort.
* Only insert this exception route if its mtu
* is less than ort's mtu value.
*/
if (dst_metric_raw(&nrt->dst, RTAX_MTU) >= fib6_mtu(ort)) {
err = -EINVAL;
goto out;
}
rt6_ex = __rt6_find_exception_spinlock(&bucket, &nrt->rt6i_dst.addr,
src_key);
if (rt6_ex)
rt6_remove_exception(bucket, rt6_ex);
rt6_ex = kzalloc(sizeof(*rt6_ex), GFP_ATOMIC);
if (!rt6_ex) {
err = -ENOMEM;
goto out;
}
rt6_ex->rt6i = nrt;
rt6_ex->stamp = jiffies;
hlist_add_head_rcu(&rt6_ex->hlist, &bucket->chain);
bucket->depth++;
net->ipv6.rt6_stats->fib_rt_cache++;
if (bucket->depth > FIB6_MAX_DEPTH)
rt6_exception_remove_oldest(bucket);
out:
spin_unlock_bh(&rt6_exception_lock);
/* Update fn->fn_sernum to invalidate all cached dst */
if (!err) {
spin_lock_bh(&ort->fib6_table->tb6_lock);
fib6_update_sernum(net, ort);
spin_unlock_bh(&ort->fib6_table->tb6_lock);
fib6_force_start_gc(net);
}
return err;
}
void rt6_flush_exceptions(struct fib6_info *rt)
{
struct rt6_exception_bucket *bucket;
struct rt6_exception *rt6_ex;
struct hlist_node *tmp;
int i;
spin_lock_bh(&rt6_exception_lock);
/* Prevent rt6_insert_exception() to recreate the bucket list */
rt->exception_bucket_flushed = 1;
bucket = rcu_dereference_protected(rt->rt6i_exception_bucket,
lockdep_is_held(&rt6_exception_lock));
if (!bucket)
goto out;
for (i = 0; i < FIB6_EXCEPTION_BUCKET_SIZE; i++) {
hlist_for_each_entry_safe(rt6_ex, tmp, &bucket->chain, hlist)
rt6_remove_exception(bucket, rt6_ex);
WARN_ON_ONCE(bucket->depth);
bucket++;
}
out:
spin_unlock_bh(&rt6_exception_lock);
}
/* Find cached rt in the hash table inside passed in rt
* Caller has to hold rcu_read_lock()
*/
static struct rt6_info *rt6_find_cached_rt(struct fib6_info *rt,
struct in6_addr *daddr,
struct in6_addr *saddr)
{
struct rt6_exception_bucket *bucket;
struct in6_addr *src_key = NULL;
struct rt6_exception *rt6_ex;
struct rt6_info *res = NULL;
bucket = rcu_dereference(rt->rt6i_exception_bucket);
#ifdef CONFIG_IPV6_SUBTREES
/* rt6i_src.plen != 0 indicates rt is in subtree
* and exception table is indexed by a hash of
* both rt6i_dst and rt6i_src.
* Otherwise, the exception table is indexed by
* a hash of only rt6i_dst.
*/
if (rt->fib6_src.plen)
src_key = saddr;
#endif
rt6_ex = __rt6_find_exception_rcu(&bucket, daddr, src_key);
if (rt6_ex && !rt6_check_expired(rt6_ex->rt6i))
res = rt6_ex->rt6i;
return res;
}
/* Remove the passed in cached rt from the hash table that contains it */
static int rt6_remove_exception_rt(struct rt6_info *rt)
{
struct rt6_exception_bucket *bucket;
struct in6_addr *src_key = NULL;
struct rt6_exception *rt6_ex;
struct fib6_info *from;
int err;
from = rcu_dereference(rt->from);
if (!from ||
!(rt->rt6i_flags & RTF_CACHE))
return -EINVAL;
if (!rcu_access_pointer(from->rt6i_exception_bucket))
return -ENOENT;
spin_lock_bh(&rt6_exception_lock);
bucket = rcu_dereference_protected(from->rt6i_exception_bucket,
lockdep_is_held(&rt6_exception_lock));
#ifdef CONFIG_IPV6_SUBTREES
/* rt6i_src.plen != 0 indicates 'from' is in subtree
* and exception table is indexed by a hash of
* both rt6i_dst and rt6i_src.
* Otherwise, the exception table is indexed by
* a hash of only rt6i_dst.
*/
if (from->fib6_src.plen)
src_key = &rt->rt6i_src.addr;
#endif
rt6_ex = __rt6_find_exception_spinlock(&bucket,
&rt->rt6i_dst.addr,
src_key);
if (rt6_ex) {
rt6_remove_exception(bucket, rt6_ex);
err = 0;
} else {
err = -ENOENT;
}
spin_unlock_bh(&rt6_exception_lock);
return err;
}
/* Find rt6_ex which contains the passed in rt cache and
* refresh its stamp
*/
static void rt6_update_exception_stamp_rt(struct rt6_info *rt)
{
struct rt6_exception_bucket *bucket;
struct fib6_info *from = rt->from;
struct in6_addr *src_key = NULL;
struct rt6_exception *rt6_ex;
if (!from ||
!(rt->rt6i_flags & RTF_CACHE))
return;
rcu_read_lock();
bucket = rcu_dereference(from->rt6i_exception_bucket);
#ifdef CONFIG_IPV6_SUBTREES
/* rt6i_src.plen != 0 indicates 'from' is in subtree
* and exception table is indexed by a hash of
* both rt6i_dst and rt6i_src.
* Otherwise, the exception table is indexed by
* a hash of only rt6i_dst.
*/
if (from->fib6_src.plen)
src_key = &rt->rt6i_src.addr;
#endif
rt6_ex = __rt6_find_exception_rcu(&bucket,
&rt->rt6i_dst.addr,
src_key);
if (rt6_ex)
rt6_ex->stamp = jiffies;
rcu_read_unlock();
}
ipv6: Reflect MTU changes on PMTU of exceptions for MTU-less routes Currently, administrative MTU changes on a given netdevice are not reflected on route exceptions for MTU-less routes, with a set PMTU value, for that device: # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a proto kernel src 2001:db8::a metric 256 pref medium # ping6 -c 1 -q -s10000 2001:db8::b > /dev/null # ip netns exec a ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 3000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 9000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium The first issue is that since commit fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") we don't call rt6_exceptions_update_pmtu() from rt6_mtu_change_route(), which handles administrative MTU changes, if the regular route is MTU-less. However, PMTU exceptions should be always updated, as long as RTAX_MTU is not locked. Keep the check for MTU-less main route, as introduced by that commit, but, for exceptions, call rt6_exceptions_update_pmtu() regardless of that check. Once that is fixed, one problem remains: MTU changes are not reflected if the new MTU is higher than the previous one, because rt6_exceptions_update_pmtu() doesn't allow that. We should instead allow PMTU increase if the old PMTU matches the local MTU, as that implies that the old MTU was the lowest in the path, and PMTU discovery might lead to different results. The existing check in rt6_mtu_change_route() correctly took that case into account (for regular routes only), so factor it out and re-use it also in rt6_exceptions_update_pmtu(). While at it, fix comments style and grammar, and try to be a bit more descriptive. Reported-by: Xiumei Mu <xmu@redhat.com> Fixes: fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") Fixes: f5bbe7ee79c2 ("ipv6: prepare rt6_mtu_change() for exception table") Signed-off-by: Stefano Brivio <sbrivio@redhat.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-06 18:10:19 +08:00
static bool rt6_mtu_change_route_allowed(struct inet6_dev *idev,
struct rt6_info *rt, int mtu)
{
/* If the new MTU is lower than the route PMTU, this new MTU will be the
* lowest MTU in the path: always allow updating the route PMTU to
* reflect PMTU decreases.
*
* If the new MTU is higher, and the route PMTU is equal to the local
* MTU, this means the old MTU is the lowest in the path, so allow
* updating it: if other nodes now have lower MTUs, PMTU discovery will
* handle this.
*/
if (dst_mtu(&rt->dst) >= mtu)
return true;
if (dst_mtu(&rt->dst) == idev->cnf.mtu6)
return true;
return false;
}
static void rt6_exceptions_update_pmtu(struct inet6_dev *idev,
struct fib6_info *rt, int mtu)
{
struct rt6_exception_bucket *bucket;
struct rt6_exception *rt6_ex;
int i;
bucket = rcu_dereference_protected(rt->rt6i_exception_bucket,
lockdep_is_held(&rt6_exception_lock));
ipv6: Reflect MTU changes on PMTU of exceptions for MTU-less routes Currently, administrative MTU changes on a given netdevice are not reflected on route exceptions for MTU-less routes, with a set PMTU value, for that device: # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a proto kernel src 2001:db8::a metric 256 pref medium # ping6 -c 1 -q -s10000 2001:db8::b > /dev/null # ip netns exec a ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 3000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 9000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium The first issue is that since commit fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") we don't call rt6_exceptions_update_pmtu() from rt6_mtu_change_route(), which handles administrative MTU changes, if the regular route is MTU-less. However, PMTU exceptions should be always updated, as long as RTAX_MTU is not locked. Keep the check for MTU-less main route, as introduced by that commit, but, for exceptions, call rt6_exceptions_update_pmtu() regardless of that check. Once that is fixed, one problem remains: MTU changes are not reflected if the new MTU is higher than the previous one, because rt6_exceptions_update_pmtu() doesn't allow that. We should instead allow PMTU increase if the old PMTU matches the local MTU, as that implies that the old MTU was the lowest in the path, and PMTU discovery might lead to different results. The existing check in rt6_mtu_change_route() correctly took that case into account (for regular routes only), so factor it out and re-use it also in rt6_exceptions_update_pmtu(). While at it, fix comments style and grammar, and try to be a bit more descriptive. Reported-by: Xiumei Mu <xmu@redhat.com> Fixes: fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") Fixes: f5bbe7ee79c2 ("ipv6: prepare rt6_mtu_change() for exception table") Signed-off-by: Stefano Brivio <sbrivio@redhat.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-06 18:10:19 +08:00
if (!bucket)
return;
for (i = 0; i < FIB6_EXCEPTION_BUCKET_SIZE; i++) {
hlist_for_each_entry(rt6_ex, &bucket->chain, hlist) {
struct rt6_info *entry = rt6_ex->rt6i;
/* For RTF_CACHE with rt6i_pmtu == 0 (i.e. a redirected
* route), the metrics of its rt->from have already
ipv6: Reflect MTU changes on PMTU of exceptions for MTU-less routes Currently, administrative MTU changes on a given netdevice are not reflected on route exceptions for MTU-less routes, with a set PMTU value, for that device: # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a proto kernel src 2001:db8::a metric 256 pref medium # ping6 -c 1 -q -s10000 2001:db8::b > /dev/null # ip netns exec a ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 3000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 9000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium The first issue is that since commit fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") we don't call rt6_exceptions_update_pmtu() from rt6_mtu_change_route(), which handles administrative MTU changes, if the regular route is MTU-less. However, PMTU exceptions should be always updated, as long as RTAX_MTU is not locked. Keep the check for MTU-less main route, as introduced by that commit, but, for exceptions, call rt6_exceptions_update_pmtu() regardless of that check. Once that is fixed, one problem remains: MTU changes are not reflected if the new MTU is higher than the previous one, because rt6_exceptions_update_pmtu() doesn't allow that. We should instead allow PMTU increase if the old PMTU matches the local MTU, as that implies that the old MTU was the lowest in the path, and PMTU discovery might lead to different results. The existing check in rt6_mtu_change_route() correctly took that case into account (for regular routes only), so factor it out and re-use it also in rt6_exceptions_update_pmtu(). While at it, fix comments style and grammar, and try to be a bit more descriptive. Reported-by: Xiumei Mu <xmu@redhat.com> Fixes: fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") Fixes: f5bbe7ee79c2 ("ipv6: prepare rt6_mtu_change() for exception table") Signed-off-by: Stefano Brivio <sbrivio@redhat.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-06 18:10:19 +08:00
* been updated.
*/
if (dst_metric_raw(&entry->dst, RTAX_MTU) &&
ipv6: Reflect MTU changes on PMTU of exceptions for MTU-less routes Currently, administrative MTU changes on a given netdevice are not reflected on route exceptions for MTU-less routes, with a set PMTU value, for that device: # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a proto kernel src 2001:db8::a metric 256 pref medium # ping6 -c 1 -q -s10000 2001:db8::b > /dev/null # ip netns exec a ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 3000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 9000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium The first issue is that since commit fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") we don't call rt6_exceptions_update_pmtu() from rt6_mtu_change_route(), which handles administrative MTU changes, if the regular route is MTU-less. However, PMTU exceptions should be always updated, as long as RTAX_MTU is not locked. Keep the check for MTU-less main route, as introduced by that commit, but, for exceptions, call rt6_exceptions_update_pmtu() regardless of that check. Once that is fixed, one problem remains: MTU changes are not reflected if the new MTU is higher than the previous one, because rt6_exceptions_update_pmtu() doesn't allow that. We should instead allow PMTU increase if the old PMTU matches the local MTU, as that implies that the old MTU was the lowest in the path, and PMTU discovery might lead to different results. The existing check in rt6_mtu_change_route() correctly took that case into account (for regular routes only), so factor it out and re-use it also in rt6_exceptions_update_pmtu(). While at it, fix comments style and grammar, and try to be a bit more descriptive. Reported-by: Xiumei Mu <xmu@redhat.com> Fixes: fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") Fixes: f5bbe7ee79c2 ("ipv6: prepare rt6_mtu_change() for exception table") Signed-off-by: Stefano Brivio <sbrivio@redhat.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-06 18:10:19 +08:00
rt6_mtu_change_route_allowed(idev, entry, mtu))
dst_metric_set(&entry->dst, RTAX_MTU, mtu);
}
ipv6: Reflect MTU changes on PMTU of exceptions for MTU-less routes Currently, administrative MTU changes on a given netdevice are not reflected on route exceptions for MTU-less routes, with a set PMTU value, for that device: # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a proto kernel src 2001:db8::a metric 256 pref medium # ping6 -c 1 -q -s10000 2001:db8::b > /dev/null # ip netns exec a ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 3000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 9000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium The first issue is that since commit fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") we don't call rt6_exceptions_update_pmtu() from rt6_mtu_change_route(), which handles administrative MTU changes, if the regular route is MTU-less. However, PMTU exceptions should be always updated, as long as RTAX_MTU is not locked. Keep the check for MTU-less main route, as introduced by that commit, but, for exceptions, call rt6_exceptions_update_pmtu() regardless of that check. Once that is fixed, one problem remains: MTU changes are not reflected if the new MTU is higher than the previous one, because rt6_exceptions_update_pmtu() doesn't allow that. We should instead allow PMTU increase if the old PMTU matches the local MTU, as that implies that the old MTU was the lowest in the path, and PMTU discovery might lead to different results. The existing check in rt6_mtu_change_route() correctly took that case into account (for regular routes only), so factor it out and re-use it also in rt6_exceptions_update_pmtu(). While at it, fix comments style and grammar, and try to be a bit more descriptive. Reported-by: Xiumei Mu <xmu@redhat.com> Fixes: fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") Fixes: f5bbe7ee79c2 ("ipv6: prepare rt6_mtu_change() for exception table") Signed-off-by: Stefano Brivio <sbrivio@redhat.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-06 18:10:19 +08:00
bucket++;
}
}
#define RTF_CACHE_GATEWAY (RTF_GATEWAY | RTF_CACHE)
static void rt6_exceptions_clean_tohost(struct fib6_info *rt,
struct in6_addr *gateway)
{
struct rt6_exception_bucket *bucket;
struct rt6_exception *rt6_ex;
struct hlist_node *tmp;
int i;
if (!rcu_access_pointer(rt->rt6i_exception_bucket))
return;
spin_lock_bh(&rt6_exception_lock);
bucket = rcu_dereference_protected(rt->rt6i_exception_bucket,
lockdep_is_held(&rt6_exception_lock));
if (bucket) {
for (i = 0; i < FIB6_EXCEPTION_BUCKET_SIZE; i++) {
hlist_for_each_entry_safe(rt6_ex, tmp,
&bucket->chain, hlist) {
struct rt6_info *entry = rt6_ex->rt6i;
if ((entry->rt6i_flags & RTF_CACHE_GATEWAY) ==
RTF_CACHE_GATEWAY &&
ipv6_addr_equal(gateway,
&entry->rt6i_gateway)) {
rt6_remove_exception(bucket, rt6_ex);
}
}
bucket++;
}
}
spin_unlock_bh(&rt6_exception_lock);
}
static void rt6_age_examine_exception(struct rt6_exception_bucket *bucket,
struct rt6_exception *rt6_ex,
struct fib6_gc_args *gc_args,
unsigned long now)
{
struct rt6_info *rt = rt6_ex->rt6i;
/* we are pruning and obsoleting aged-out and non gateway exceptions
* even if others have still references to them, so that on next
* dst_check() such references can be dropped.
* EXPIRES exceptions - e.g. pmtu-generated ones are pruned when
* expired, independently from their aging, as per RFC 8201 section 4
*/
if (!(rt->rt6i_flags & RTF_EXPIRES)) {
if (time_after_eq(now, rt->dst.lastuse + gc_args->timeout)) {
RT6_TRACE("aging clone %p\n", rt);
rt6_remove_exception(bucket, rt6_ex);
return;
}
} else if (time_after(jiffies, rt->dst.expires)) {
RT6_TRACE("purging expired route %p\n", rt);
rt6_remove_exception(bucket, rt6_ex);
return;
}
if (rt->rt6i_flags & RTF_GATEWAY) {
struct neighbour *neigh;
__u8 neigh_flags = 0;
ipv6: fix possible deadlock in rt6_age_examine_exception() syzbot reported a LOCKDEP splat [1] in rt6_age_examine_exception() rt6_age_examine_exception() is called while rt6_exception_lock is held. This lock is the lower one in the lock hierarchy, thus we can not call dst_neigh_lookup() function, as it can fallback to neigh_create() We should instead do a pure RCU lookup. As a bonus we avoid a pair of atomic operations on neigh refcount. [1] WARNING: possible circular locking dependency detected 4.16.0-rc4+ #277 Not tainted syz-executor7/4015 is trying to acquire lock: (&ndev->lock){++--}, at: [<00000000416dce19>] __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 but task is already holding lock: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&tbl->lock){++-.}: __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __neigh_create+0x87e/0x1d90 net/core/neighbour.c:528 neigh_create include/net/neighbour.h:315 [inline] ip6_neigh_lookup+0x9a7/0xba0 net/ipv6/route.c:228 dst_neigh_lookup include/net/dst.h:405 [inline] rt6_age_examine_exception net/ipv6/route.c:1609 [inline] rt6_age_exceptions+0x381/0x660 net/ipv6/route.c:1645 fib6_age+0xfb/0x140 net/ipv6/ip6_fib.c:2033 fib6_clean_node+0x389/0x580 net/ipv6/ip6_fib.c:1919 fib6_walk_continue+0x46c/0x8a0 net/ipv6/ip6_fib.c:1845 fib6_walk+0x91/0xf0 net/ipv6/ip6_fib.c:1893 fib6_clean_tree+0x1e6/0x340 net/ipv6/ip6_fib.c:1970 __fib6_clean_all+0x1f4/0x3a0 net/ipv6/ip6_fib.c:1986 fib6_clean_all net/ipv6/ip6_fib.c:1997 [inline] fib6_run_gc+0x16b/0x3c0 net/ipv6/ip6_fib.c:2053 ndisc_netdev_event+0x3c2/0x4a0 net/ipv6/ndisc.c:1781 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #2 (rt6_exception_lock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] rt6_flush_exceptions+0x21/0x210 net/ipv6/route.c:1367 fib6_del_route net/ipv6/ip6_fib.c:1677 [inline] fib6_del+0x624/0x12c0 net/ipv6/ip6_fib.c:1761 __ip6_del_rt+0xc7/0x120 net/ipv6/route.c:2980 ip6_del_rt+0x132/0x1a0 net/ipv6/route.c:2993 __ipv6_dev_ac_dec+0x3b1/0x600 net/ipv6/anycast.c:332 ipv6_dev_ac_dec net/ipv6/anycast.c:345 [inline] ipv6_sock_ac_close+0x2b4/0x3e0 net/ipv6/anycast.c:200 inet6_release+0x48/0x70 net/ipv6/af_inet6.c:433 sock_release+0x8d/0x1e0 net/socket.c:594 sock_close+0x16/0x20 net/socket.c:1149 __fput+0x327/0x7e0 fs/file_table.c:209 ____fput+0x15/0x20 fs/file_table.c:243 task_work_run+0x199/0x270 kernel/task_work.c:113 exit_task_work include/linux/task_work.h:22 [inline] do_exit+0x9bb/0x1ad0 kernel/exit.c:865 do_group_exit+0x149/0x400 kernel/exit.c:968 get_signal+0x73a/0x16d0 kernel/signal.c:2469 do_signal+0x90/0x1e90 arch/x86/kernel/signal.c:809 exit_to_usermode_loop+0x258/0x2f0 arch/x86/entry/common.c:162 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline] syscall_return_slowpath arch/x86/entry/common.c:265 [inline] do_syscall_64+0x6ec/0x940 arch/x86/entry/common.c:292 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #1 (&(&tb->tb6_lock)->rlock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] __ip6_ins_rt+0x56/0x90 net/ipv6/route.c:1007 ip6_route_add+0x141/0x190 net/ipv6/route.c:2955 addrconf_prefix_route+0x44f/0x620 net/ipv6/addrconf.c:2359 fixup_permanent_addr net/ipv6/addrconf.c:3368 [inline] addrconf_permanent_addr net/ipv6/addrconf.c:3391 [inline] addrconf_notify+0x1ad2/0x2310 net/ipv6/addrconf.c:3460 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x15d/0x430 net/core/dev.c:6958 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 do_setlink+0xa22/0x3bb0 net/core/rtnetlink.c:2357 rtnl_newlink+0xf37/0x1a50 net/core/rtnetlink.c:2965 rtnetlink_rcv_msg+0x57f/0xb10 net/core/rtnetlink.c:4641 netlink_rcv_skb+0x14b/0x380 net/netlink/af_netlink.c:2444 rtnetlink_rcv+0x1c/0x20 net/core/rtnetlink.c:4659 netlink_unicast_kernel net/netlink/af_netlink.c:1308 [inline] netlink_unicast+0x4c4/0x6b0 net/netlink/af_netlink.c:1334 netlink_sendmsg+0xa4a/0xe60 net/netlink/af_netlink.c:1897 sock_sendmsg_nosec net/socket.c:629 [inline] sock_sendmsg+0xca/0x110 net/socket.c:639 ___sys_sendmsg+0x767/0x8b0 net/socket.c:2047 __sys_sendmsg+0xe5/0x210 net/socket.c:2081 SYSC_sendmsg net/socket.c:2092 [inline] SyS_sendmsg+0x2d/0x50 net/socket.c:2088 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #0 (&ndev->lock){++--}: lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 other info that might help us debug this: Chain exists of: &ndev->lock --> rt6_exception_lock --> &tbl->lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&tbl->lock); lock(rt6_exception_lock); lock(&tbl->lock); lock(&ndev->lock); *** DEADLOCK *** 2 locks held by syz-executor7/4015: #0: (rtnl_mutex){+.+.}, at: [<00000000a2f16daa>] rtnl_lock+0x17/0x20 net/core/rtnetlink.c:74 #1: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 stack backtrace: CPU: 0 PID: 4015 Comm: syz-executor7 Not tainted 4.16.0-rc4+ #277 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x24d lib/dump_stack.c:53 print_circular_bug.isra.38+0x2cd/0x2dc kernel/locking/lockdep.c:1223 check_prev_add kernel/locking/lockdep.c:1863 [inline] check_prevs_add kernel/locking/lockdep.c:1976 [inline] validate_chain kernel/locking/lockdep.c:2417 [inline] __lock_acquire+0x30a8/0x3e00 kernel/locking/lockdep.c:3431 lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 Fixes: c757faa8bfa2 ("ipv6: prepare fib6_age() for exception table") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Cc: Martin KaFai Lau <kafai@fb.com> Acked-by: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 22:56:58 +08:00
neigh = __ipv6_neigh_lookup_noref(rt->dst.dev, &rt->rt6i_gateway);
if (neigh)
neigh_flags = neigh->flags;
ipv6: fix possible deadlock in rt6_age_examine_exception() syzbot reported a LOCKDEP splat [1] in rt6_age_examine_exception() rt6_age_examine_exception() is called while rt6_exception_lock is held. This lock is the lower one in the lock hierarchy, thus we can not call dst_neigh_lookup() function, as it can fallback to neigh_create() We should instead do a pure RCU lookup. As a bonus we avoid a pair of atomic operations on neigh refcount. [1] WARNING: possible circular locking dependency detected 4.16.0-rc4+ #277 Not tainted syz-executor7/4015 is trying to acquire lock: (&ndev->lock){++--}, at: [<00000000416dce19>] __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 but task is already holding lock: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&tbl->lock){++-.}: __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __neigh_create+0x87e/0x1d90 net/core/neighbour.c:528 neigh_create include/net/neighbour.h:315 [inline] ip6_neigh_lookup+0x9a7/0xba0 net/ipv6/route.c:228 dst_neigh_lookup include/net/dst.h:405 [inline] rt6_age_examine_exception net/ipv6/route.c:1609 [inline] rt6_age_exceptions+0x381/0x660 net/ipv6/route.c:1645 fib6_age+0xfb/0x140 net/ipv6/ip6_fib.c:2033 fib6_clean_node+0x389/0x580 net/ipv6/ip6_fib.c:1919 fib6_walk_continue+0x46c/0x8a0 net/ipv6/ip6_fib.c:1845 fib6_walk+0x91/0xf0 net/ipv6/ip6_fib.c:1893 fib6_clean_tree+0x1e6/0x340 net/ipv6/ip6_fib.c:1970 __fib6_clean_all+0x1f4/0x3a0 net/ipv6/ip6_fib.c:1986 fib6_clean_all net/ipv6/ip6_fib.c:1997 [inline] fib6_run_gc+0x16b/0x3c0 net/ipv6/ip6_fib.c:2053 ndisc_netdev_event+0x3c2/0x4a0 net/ipv6/ndisc.c:1781 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #2 (rt6_exception_lock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] rt6_flush_exceptions+0x21/0x210 net/ipv6/route.c:1367 fib6_del_route net/ipv6/ip6_fib.c:1677 [inline] fib6_del+0x624/0x12c0 net/ipv6/ip6_fib.c:1761 __ip6_del_rt+0xc7/0x120 net/ipv6/route.c:2980 ip6_del_rt+0x132/0x1a0 net/ipv6/route.c:2993 __ipv6_dev_ac_dec+0x3b1/0x600 net/ipv6/anycast.c:332 ipv6_dev_ac_dec net/ipv6/anycast.c:345 [inline] ipv6_sock_ac_close+0x2b4/0x3e0 net/ipv6/anycast.c:200 inet6_release+0x48/0x70 net/ipv6/af_inet6.c:433 sock_release+0x8d/0x1e0 net/socket.c:594 sock_close+0x16/0x20 net/socket.c:1149 __fput+0x327/0x7e0 fs/file_table.c:209 ____fput+0x15/0x20 fs/file_table.c:243 task_work_run+0x199/0x270 kernel/task_work.c:113 exit_task_work include/linux/task_work.h:22 [inline] do_exit+0x9bb/0x1ad0 kernel/exit.c:865 do_group_exit+0x149/0x400 kernel/exit.c:968 get_signal+0x73a/0x16d0 kernel/signal.c:2469 do_signal+0x90/0x1e90 arch/x86/kernel/signal.c:809 exit_to_usermode_loop+0x258/0x2f0 arch/x86/entry/common.c:162 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline] syscall_return_slowpath arch/x86/entry/common.c:265 [inline] do_syscall_64+0x6ec/0x940 arch/x86/entry/common.c:292 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #1 (&(&tb->tb6_lock)->rlock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] __ip6_ins_rt+0x56/0x90 net/ipv6/route.c:1007 ip6_route_add+0x141/0x190 net/ipv6/route.c:2955 addrconf_prefix_route+0x44f/0x620 net/ipv6/addrconf.c:2359 fixup_permanent_addr net/ipv6/addrconf.c:3368 [inline] addrconf_permanent_addr net/ipv6/addrconf.c:3391 [inline] addrconf_notify+0x1ad2/0x2310 net/ipv6/addrconf.c:3460 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x15d/0x430 net/core/dev.c:6958 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 do_setlink+0xa22/0x3bb0 net/core/rtnetlink.c:2357 rtnl_newlink+0xf37/0x1a50 net/core/rtnetlink.c:2965 rtnetlink_rcv_msg+0x57f/0xb10 net/core/rtnetlink.c:4641 netlink_rcv_skb+0x14b/0x380 net/netlink/af_netlink.c:2444 rtnetlink_rcv+0x1c/0x20 net/core/rtnetlink.c:4659 netlink_unicast_kernel net/netlink/af_netlink.c:1308 [inline] netlink_unicast+0x4c4/0x6b0 net/netlink/af_netlink.c:1334 netlink_sendmsg+0xa4a/0xe60 net/netlink/af_netlink.c:1897 sock_sendmsg_nosec net/socket.c:629 [inline] sock_sendmsg+0xca/0x110 net/socket.c:639 ___sys_sendmsg+0x767/0x8b0 net/socket.c:2047 __sys_sendmsg+0xe5/0x210 net/socket.c:2081 SYSC_sendmsg net/socket.c:2092 [inline] SyS_sendmsg+0x2d/0x50 net/socket.c:2088 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #0 (&ndev->lock){++--}: lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 other info that might help us debug this: Chain exists of: &ndev->lock --> rt6_exception_lock --> &tbl->lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&tbl->lock); lock(rt6_exception_lock); lock(&tbl->lock); lock(&ndev->lock); *** DEADLOCK *** 2 locks held by syz-executor7/4015: #0: (rtnl_mutex){+.+.}, at: [<00000000a2f16daa>] rtnl_lock+0x17/0x20 net/core/rtnetlink.c:74 #1: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 stack backtrace: CPU: 0 PID: 4015 Comm: syz-executor7 Not tainted 4.16.0-rc4+ #277 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x24d lib/dump_stack.c:53 print_circular_bug.isra.38+0x2cd/0x2dc kernel/locking/lockdep.c:1223 check_prev_add kernel/locking/lockdep.c:1863 [inline] check_prevs_add kernel/locking/lockdep.c:1976 [inline] validate_chain kernel/locking/lockdep.c:2417 [inline] __lock_acquire+0x30a8/0x3e00 kernel/locking/lockdep.c:3431 lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 Fixes: c757faa8bfa2 ("ipv6: prepare fib6_age() for exception table") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Cc: Martin KaFai Lau <kafai@fb.com> Acked-by: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 22:56:58 +08:00
if (!(neigh_flags & NTF_ROUTER)) {
RT6_TRACE("purging route %p via non-router but gateway\n",
rt);
rt6_remove_exception(bucket, rt6_ex);
return;
}
}
gc_args->more++;
}
void rt6_age_exceptions(struct fib6_info *rt,
struct fib6_gc_args *gc_args,
unsigned long now)
{
struct rt6_exception_bucket *bucket;
struct rt6_exception *rt6_ex;
struct hlist_node *tmp;
int i;
if (!rcu_access_pointer(rt->rt6i_exception_bucket))
return;
ipv6: fix possible deadlock in rt6_age_examine_exception() syzbot reported a LOCKDEP splat [1] in rt6_age_examine_exception() rt6_age_examine_exception() is called while rt6_exception_lock is held. This lock is the lower one in the lock hierarchy, thus we can not call dst_neigh_lookup() function, as it can fallback to neigh_create() We should instead do a pure RCU lookup. As a bonus we avoid a pair of atomic operations on neigh refcount. [1] WARNING: possible circular locking dependency detected 4.16.0-rc4+ #277 Not tainted syz-executor7/4015 is trying to acquire lock: (&ndev->lock){++--}, at: [<00000000416dce19>] __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 but task is already holding lock: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&tbl->lock){++-.}: __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __neigh_create+0x87e/0x1d90 net/core/neighbour.c:528 neigh_create include/net/neighbour.h:315 [inline] ip6_neigh_lookup+0x9a7/0xba0 net/ipv6/route.c:228 dst_neigh_lookup include/net/dst.h:405 [inline] rt6_age_examine_exception net/ipv6/route.c:1609 [inline] rt6_age_exceptions+0x381/0x660 net/ipv6/route.c:1645 fib6_age+0xfb/0x140 net/ipv6/ip6_fib.c:2033 fib6_clean_node+0x389/0x580 net/ipv6/ip6_fib.c:1919 fib6_walk_continue+0x46c/0x8a0 net/ipv6/ip6_fib.c:1845 fib6_walk+0x91/0xf0 net/ipv6/ip6_fib.c:1893 fib6_clean_tree+0x1e6/0x340 net/ipv6/ip6_fib.c:1970 __fib6_clean_all+0x1f4/0x3a0 net/ipv6/ip6_fib.c:1986 fib6_clean_all net/ipv6/ip6_fib.c:1997 [inline] fib6_run_gc+0x16b/0x3c0 net/ipv6/ip6_fib.c:2053 ndisc_netdev_event+0x3c2/0x4a0 net/ipv6/ndisc.c:1781 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #2 (rt6_exception_lock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] rt6_flush_exceptions+0x21/0x210 net/ipv6/route.c:1367 fib6_del_route net/ipv6/ip6_fib.c:1677 [inline] fib6_del+0x624/0x12c0 net/ipv6/ip6_fib.c:1761 __ip6_del_rt+0xc7/0x120 net/ipv6/route.c:2980 ip6_del_rt+0x132/0x1a0 net/ipv6/route.c:2993 __ipv6_dev_ac_dec+0x3b1/0x600 net/ipv6/anycast.c:332 ipv6_dev_ac_dec net/ipv6/anycast.c:345 [inline] ipv6_sock_ac_close+0x2b4/0x3e0 net/ipv6/anycast.c:200 inet6_release+0x48/0x70 net/ipv6/af_inet6.c:433 sock_release+0x8d/0x1e0 net/socket.c:594 sock_close+0x16/0x20 net/socket.c:1149 __fput+0x327/0x7e0 fs/file_table.c:209 ____fput+0x15/0x20 fs/file_table.c:243 task_work_run+0x199/0x270 kernel/task_work.c:113 exit_task_work include/linux/task_work.h:22 [inline] do_exit+0x9bb/0x1ad0 kernel/exit.c:865 do_group_exit+0x149/0x400 kernel/exit.c:968 get_signal+0x73a/0x16d0 kernel/signal.c:2469 do_signal+0x90/0x1e90 arch/x86/kernel/signal.c:809 exit_to_usermode_loop+0x258/0x2f0 arch/x86/entry/common.c:162 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline] syscall_return_slowpath arch/x86/entry/common.c:265 [inline] do_syscall_64+0x6ec/0x940 arch/x86/entry/common.c:292 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #1 (&(&tb->tb6_lock)->rlock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] __ip6_ins_rt+0x56/0x90 net/ipv6/route.c:1007 ip6_route_add+0x141/0x190 net/ipv6/route.c:2955 addrconf_prefix_route+0x44f/0x620 net/ipv6/addrconf.c:2359 fixup_permanent_addr net/ipv6/addrconf.c:3368 [inline] addrconf_permanent_addr net/ipv6/addrconf.c:3391 [inline] addrconf_notify+0x1ad2/0x2310 net/ipv6/addrconf.c:3460 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x15d/0x430 net/core/dev.c:6958 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 do_setlink+0xa22/0x3bb0 net/core/rtnetlink.c:2357 rtnl_newlink+0xf37/0x1a50 net/core/rtnetlink.c:2965 rtnetlink_rcv_msg+0x57f/0xb10 net/core/rtnetlink.c:4641 netlink_rcv_skb+0x14b/0x380 net/netlink/af_netlink.c:2444 rtnetlink_rcv+0x1c/0x20 net/core/rtnetlink.c:4659 netlink_unicast_kernel net/netlink/af_netlink.c:1308 [inline] netlink_unicast+0x4c4/0x6b0 net/netlink/af_netlink.c:1334 netlink_sendmsg+0xa4a/0xe60 net/netlink/af_netlink.c:1897 sock_sendmsg_nosec net/socket.c:629 [inline] sock_sendmsg+0xca/0x110 net/socket.c:639 ___sys_sendmsg+0x767/0x8b0 net/socket.c:2047 __sys_sendmsg+0xe5/0x210 net/socket.c:2081 SYSC_sendmsg net/socket.c:2092 [inline] SyS_sendmsg+0x2d/0x50 net/socket.c:2088 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #0 (&ndev->lock){++--}: lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 other info that might help us debug this: Chain exists of: &ndev->lock --> rt6_exception_lock --> &tbl->lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&tbl->lock); lock(rt6_exception_lock); lock(&tbl->lock); lock(&ndev->lock); *** DEADLOCK *** 2 locks held by syz-executor7/4015: #0: (rtnl_mutex){+.+.}, at: [<00000000a2f16daa>] rtnl_lock+0x17/0x20 net/core/rtnetlink.c:74 #1: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 stack backtrace: CPU: 0 PID: 4015 Comm: syz-executor7 Not tainted 4.16.0-rc4+ #277 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x24d lib/dump_stack.c:53 print_circular_bug.isra.38+0x2cd/0x2dc kernel/locking/lockdep.c:1223 check_prev_add kernel/locking/lockdep.c:1863 [inline] check_prevs_add kernel/locking/lockdep.c:1976 [inline] validate_chain kernel/locking/lockdep.c:2417 [inline] __lock_acquire+0x30a8/0x3e00 kernel/locking/lockdep.c:3431 lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 Fixes: c757faa8bfa2 ("ipv6: prepare fib6_age() for exception table") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Cc: Martin KaFai Lau <kafai@fb.com> Acked-by: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 22:56:58 +08:00
rcu_read_lock_bh();
spin_lock(&rt6_exception_lock);
bucket = rcu_dereference_protected(rt->rt6i_exception_bucket,
lockdep_is_held(&rt6_exception_lock));
if (bucket) {
for (i = 0; i < FIB6_EXCEPTION_BUCKET_SIZE; i++) {
hlist_for_each_entry_safe(rt6_ex, tmp,
&bucket->chain, hlist) {
rt6_age_examine_exception(bucket, rt6_ex,
gc_args, now);
}
bucket++;
}
}
ipv6: fix possible deadlock in rt6_age_examine_exception() syzbot reported a LOCKDEP splat [1] in rt6_age_examine_exception() rt6_age_examine_exception() is called while rt6_exception_lock is held. This lock is the lower one in the lock hierarchy, thus we can not call dst_neigh_lookup() function, as it can fallback to neigh_create() We should instead do a pure RCU lookup. As a bonus we avoid a pair of atomic operations on neigh refcount. [1] WARNING: possible circular locking dependency detected 4.16.0-rc4+ #277 Not tainted syz-executor7/4015 is trying to acquire lock: (&ndev->lock){++--}, at: [<00000000416dce19>] __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 but task is already holding lock: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&tbl->lock){++-.}: __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __neigh_create+0x87e/0x1d90 net/core/neighbour.c:528 neigh_create include/net/neighbour.h:315 [inline] ip6_neigh_lookup+0x9a7/0xba0 net/ipv6/route.c:228 dst_neigh_lookup include/net/dst.h:405 [inline] rt6_age_examine_exception net/ipv6/route.c:1609 [inline] rt6_age_exceptions+0x381/0x660 net/ipv6/route.c:1645 fib6_age+0xfb/0x140 net/ipv6/ip6_fib.c:2033 fib6_clean_node+0x389/0x580 net/ipv6/ip6_fib.c:1919 fib6_walk_continue+0x46c/0x8a0 net/ipv6/ip6_fib.c:1845 fib6_walk+0x91/0xf0 net/ipv6/ip6_fib.c:1893 fib6_clean_tree+0x1e6/0x340 net/ipv6/ip6_fib.c:1970 __fib6_clean_all+0x1f4/0x3a0 net/ipv6/ip6_fib.c:1986 fib6_clean_all net/ipv6/ip6_fib.c:1997 [inline] fib6_run_gc+0x16b/0x3c0 net/ipv6/ip6_fib.c:2053 ndisc_netdev_event+0x3c2/0x4a0 net/ipv6/ndisc.c:1781 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #2 (rt6_exception_lock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] rt6_flush_exceptions+0x21/0x210 net/ipv6/route.c:1367 fib6_del_route net/ipv6/ip6_fib.c:1677 [inline] fib6_del+0x624/0x12c0 net/ipv6/ip6_fib.c:1761 __ip6_del_rt+0xc7/0x120 net/ipv6/route.c:2980 ip6_del_rt+0x132/0x1a0 net/ipv6/route.c:2993 __ipv6_dev_ac_dec+0x3b1/0x600 net/ipv6/anycast.c:332 ipv6_dev_ac_dec net/ipv6/anycast.c:345 [inline] ipv6_sock_ac_close+0x2b4/0x3e0 net/ipv6/anycast.c:200 inet6_release+0x48/0x70 net/ipv6/af_inet6.c:433 sock_release+0x8d/0x1e0 net/socket.c:594 sock_close+0x16/0x20 net/socket.c:1149 __fput+0x327/0x7e0 fs/file_table.c:209 ____fput+0x15/0x20 fs/file_table.c:243 task_work_run+0x199/0x270 kernel/task_work.c:113 exit_task_work include/linux/task_work.h:22 [inline] do_exit+0x9bb/0x1ad0 kernel/exit.c:865 do_group_exit+0x149/0x400 kernel/exit.c:968 get_signal+0x73a/0x16d0 kernel/signal.c:2469 do_signal+0x90/0x1e90 arch/x86/kernel/signal.c:809 exit_to_usermode_loop+0x258/0x2f0 arch/x86/entry/common.c:162 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline] syscall_return_slowpath arch/x86/entry/common.c:265 [inline] do_syscall_64+0x6ec/0x940 arch/x86/entry/common.c:292 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #1 (&(&tb->tb6_lock)->rlock){+.-.}: __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168 spin_lock_bh include/linux/spinlock.h:315 [inline] __ip6_ins_rt+0x56/0x90 net/ipv6/route.c:1007 ip6_route_add+0x141/0x190 net/ipv6/route.c:2955 addrconf_prefix_route+0x44f/0x620 net/ipv6/addrconf.c:2359 fixup_permanent_addr net/ipv6/addrconf.c:3368 [inline] addrconf_permanent_addr net/ipv6/addrconf.c:3391 [inline] addrconf_notify+0x1ad2/0x2310 net/ipv6/addrconf.c:3460 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x15d/0x430 net/core/dev.c:6958 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 do_setlink+0xa22/0x3bb0 net/core/rtnetlink.c:2357 rtnl_newlink+0xf37/0x1a50 net/core/rtnetlink.c:2965 rtnetlink_rcv_msg+0x57f/0xb10 net/core/rtnetlink.c:4641 netlink_rcv_skb+0x14b/0x380 net/netlink/af_netlink.c:2444 rtnetlink_rcv+0x1c/0x20 net/core/rtnetlink.c:4659 netlink_unicast_kernel net/netlink/af_netlink.c:1308 [inline] netlink_unicast+0x4c4/0x6b0 net/netlink/af_netlink.c:1334 netlink_sendmsg+0xa4a/0xe60 net/netlink/af_netlink.c:1897 sock_sendmsg_nosec net/socket.c:629 [inline] sock_sendmsg+0xca/0x110 net/socket.c:639 ___sys_sendmsg+0x767/0x8b0 net/socket.c:2047 __sys_sendmsg+0xe5/0x210 net/socket.c:2081 SYSC_sendmsg net/socket.c:2092 [inline] SyS_sendmsg+0x2d/0x50 net/socket.c:2088 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 -> #0 (&ndev->lock){++--}: lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 other info that might help us debug this: Chain exists of: &ndev->lock --> rt6_exception_lock --> &tbl->lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&tbl->lock); lock(rt6_exception_lock); lock(&tbl->lock); lock(&ndev->lock); *** DEADLOCK *** 2 locks held by syz-executor7/4015: #0: (rtnl_mutex){+.+.}, at: [<00000000a2f16daa>] rtnl_lock+0x17/0x20 net/core/rtnetlink.c:74 #1: (&tbl->lock){++-.}, at: [<00000000b5cb1d65>] neigh_ifdown+0x3d/0x250 net/core/neighbour.c:292 stack backtrace: CPU: 0 PID: 4015 Comm: syz-executor7 Not tainted 4.16.0-rc4+ #277 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x24d lib/dump_stack.c:53 print_circular_bug.isra.38+0x2cd/0x2dc kernel/locking/lockdep.c:1223 check_prev_add kernel/locking/lockdep.c:1863 [inline] check_prevs_add kernel/locking/lockdep.c:1976 [inline] validate_chain kernel/locking/lockdep.c:2417 [inline] __lock_acquire+0x30a8/0x3e00 kernel/locking/lockdep.c:3431 lock_acquire+0x1d5/0x580 kernel/locking/lockdep.c:3920 __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline] _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312 __ipv6_dev_mc_dec+0x45/0x350 net/ipv6/mcast.c:928 ipv6_dev_mc_dec+0x110/0x1f0 net/ipv6/mcast.c:961 pndisc_destructor+0x21a/0x340 net/ipv6/ndisc.c:392 pneigh_ifdown net/core/neighbour.c:695 [inline] neigh_ifdown+0x149/0x250 net/core/neighbour.c:294 rt6_disable_ip+0x537/0x700 net/ipv6/route.c:3874 addrconf_ifdown+0x14b/0x14f0 net/ipv6/addrconf.c:3633 addrconf_notify+0x5f8/0x2310 net/ipv6/addrconf.c:3557 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x32/0x70 net/core/dev.c:1707 call_netdevice_notifiers net/core/dev.c:1725 [inline] __dev_notify_flags+0x262/0x430 net/core/dev.c:6960 dev_change_flags+0xf5/0x140 net/core/dev.c:6994 devinet_ioctl+0x126a/0x1ac0 net/ipv4/devinet.c:1080 inet_ioctl+0x184/0x310 net/ipv4/af_inet.c:919 packet_ioctl+0x1ff/0x310 net/packet/af_packet.c:4066 sock_do_ioctl+0xef/0x390 net/socket.c:957 sock_ioctl+0x36b/0x610 net/socket.c:1081 vfs_ioctl fs/ioctl.c:46 [inline] do_vfs_ioctl+0x1b1/0x1520 fs/ioctl.c:686 SYSC_ioctl fs/ioctl.c:701 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:692 do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x42/0xb7 Fixes: c757faa8bfa2 ("ipv6: prepare fib6_age() for exception table") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Cc: Martin KaFai Lau <kafai@fb.com> Acked-by: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-23 22:56:58 +08:00
spin_unlock(&rt6_exception_lock);
rcu_read_unlock_bh();
}
/* must be called with rcu lock held */
struct fib6_info *fib6_table_lookup(struct net *net, struct fib6_table *table,
int oif, struct flowi6 *fl6, int strict)
{
struct fib6_node *fn, *saved_fn;
struct fib6_info *f6i;
fn = fib6_node_lookup(&table->tb6_root, &fl6->daddr, &fl6->saddr);
saved_fn = fn;
if (fl6->flowi6_flags & FLOWI_FLAG_SKIP_NH_OIF)
oif = 0;
redo_rt6_select:
f6i = rt6_select(net, fn, oif, strict);
if (f6i == net->ipv6.fib6_null_entry) {
fn = fib6_backtrack(fn, &fl6->saddr);
if (fn)
goto redo_rt6_select;
else if (strict & RT6_LOOKUP_F_REACHABLE) {
/* also consider unreachable route */
strict &= ~RT6_LOOKUP_F_REACHABLE;
fn = saved_fn;
goto redo_rt6_select;
}
}
trace_fib6_table_lookup(net, f6i, table, fl6);
return f6i;
}
struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table,
int oif, struct flowi6 *fl6,
const struct sk_buff *skb, int flags)
{
struct fib6_info *f6i;
struct rt6_info *rt;
int strict = 0;
strict |= flags & RT6_LOOKUP_F_IFACE;
strict |= flags & RT6_LOOKUP_F_IGNORE_LINKSTATE;
if (net->ipv6.devconf_all->forwarding == 0)
strict |= RT6_LOOKUP_F_REACHABLE;
rcu_read_lock();
f6i = fib6_table_lookup(net, table, oif, fl6, strict);
if (f6i->fib6_nsiblings)
f6i = fib6_multipath_select(net, f6i, fl6, oif, skb, strict);
if (f6i == net->ipv6.fib6_null_entry) {
rt = net->ipv6.ip6_null_entry;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
dst_hold(&rt->dst);
return rt;
}
/*Search through exception table */
rt = rt6_find_cached_rt(f6i, &fl6->daddr, &fl6->saddr);
if (rt) {
if (ip6_hold_safe(net, &rt, true))
dst_use_noref(&rt->dst, jiffies);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
return rt;
} else if (unlikely((fl6->flowi6_flags & FLOWI_FLAG_KNOWN_NH) &&
!(f6i->fib6_flags & RTF_GATEWAY))) {
/* Create a RTF_CACHE clone which will not be
* owned by the fib6 tree. It is for the special case where
* the daddr in the skb during the neighbor look-up is different
* from the fl6->daddr used to look-up route here.
*/
struct rt6_info *uncached_rt;
uncached_rt = ip6_rt_cache_alloc(f6i, &fl6->daddr, NULL);
rcu_read_unlock();
if (uncached_rt) {
/* Uncached_rt's refcnt is taken during ip6_rt_cache_alloc()
* No need for another dst_hold()
*/
rt6_uncached_list_add(uncached_rt);
atomic_inc(&net->ipv6.rt6_stats->fib_rt_uncache);
} else {
uncached_rt = net->ipv6.ip6_null_entry;
dst_hold(&uncached_rt->dst);
}
return uncached_rt;
} else {
/* Get a percpu copy */
struct rt6_info *pcpu_rt;
local_bh_disable();
pcpu_rt = rt6_get_pcpu_route(f6i);
if (!pcpu_rt)
pcpu_rt = rt6_make_pcpu_route(net, f6i);
local_bh_enable();
rcu_read_unlock();
return pcpu_rt;
}
}
EXPORT_SYMBOL_GPL(ip6_pol_route);
static struct rt6_info *ip6_pol_route_input(struct net *net,
struct fib6_table *table,
struct flowi6 *fl6,
const struct sk_buff *skb,
int flags)
{
return ip6_pol_route(net, table, fl6->flowi6_iif, fl6, skb, flags);
}
struct dst_entry *ip6_route_input_lookup(struct net *net,
struct net_device *dev,
struct flowi6 *fl6,
const struct sk_buff *skb,
int flags)
{
if (rt6_need_strict(&fl6->daddr) && dev->type != ARPHRD_PIMREG)
flags |= RT6_LOOKUP_F_IFACE;
return fib6_rule_lookup(net, fl6, skb, flags, ip6_pol_route_input);
}
EXPORT_SYMBOL_GPL(ip6_route_input_lookup);
static void ip6_multipath_l3_keys(const struct sk_buff *skb,
struct flow_keys *keys,
struct flow_keys *flkeys)
{
const struct ipv6hdr *outer_iph = ipv6_hdr(skb);
const struct ipv6hdr *key_iph = outer_iph;
struct flow_keys *_flkeys = flkeys;
const struct ipv6hdr *inner_iph;
const struct icmp6hdr *icmph;
struct ipv6hdr _inner_iph;
ipv6: fix uninit-value in ip6_multipath_l3_keys() syzbot/KMSAN reported an uninit-value in ip6_multipath_l3_keys(), root caused to a bad assumption of ICMP header being already pulled in skb->head ip_multipath_l3_keys() does the correct thing, so it is an IPv6 only bug. BUG: KMSAN: uninit-value in ip6_multipath_l3_keys net/ipv6/route.c:1830 [inline] BUG: KMSAN: uninit-value in rt6_multipath_hash+0x5c4/0x640 net/ipv6/route.c:1858 CPU: 0 PID: 4507 Comm: syz-executor661 Not tainted 4.16.0+ #87 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x185/0x1d0 lib/dump_stack.c:53 kmsan_report+0x142/0x240 mm/kmsan/kmsan.c:1067 __msan_warning_32+0x6c/0xb0 mm/kmsan/kmsan_instr.c:683 ip6_multipath_l3_keys net/ipv6/route.c:1830 [inline] rt6_multipath_hash+0x5c4/0x640 net/ipv6/route.c:1858 ip6_route_input+0x65a/0x920 net/ipv6/route.c:1884 ip6_rcv_finish+0x413/0x6e0 net/ipv6/ip6_input.c:69 NF_HOOK include/linux/netfilter.h:288 [inline] ipv6_rcv+0x1e16/0x2340 net/ipv6/ip6_input.c:208 __netif_receive_skb_core+0x47df/0x4a90 net/core/dev.c:4562 __netif_receive_skb net/core/dev.c:4627 [inline] netif_receive_skb_internal+0x49d/0x630 net/core/dev.c:4701 netif_receive_skb+0x230/0x240 net/core/dev.c:4725 tun_rx_batched drivers/net/tun.c:1555 [inline] tun_get_user+0x740f/0x7c60 drivers/net/tun.c:1962 tun_chr_write_iter+0x1d4/0x330 drivers/net/tun.c:1990 call_write_iter include/linux/fs.h:1782 [inline] new_sync_write fs/read_write.c:469 [inline] __vfs_write+0x7fb/0x9f0 fs/read_write.c:482 vfs_write+0x463/0x8d0 fs/read_write.c:544 SYSC_write+0x172/0x360 fs/read_write.c:589 SyS_write+0x55/0x80 fs/read_write.c:581 do_syscall_64+0x309/0x430 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x3d/0xa2 Fixes: 23aebdacb05d ("ipv6: Compute multipath hash for ICMP errors from offending packet") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Jakub Sitnicki <jkbs@redhat.com> Acked-by: Jakub Sitnicki <jkbs@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-30 00:54:59 +08:00
struct icmp6hdr _icmph;
if (likely(outer_iph->nexthdr != IPPROTO_ICMPV6))
goto out;
ipv6: fix uninit-value in ip6_multipath_l3_keys() syzbot/KMSAN reported an uninit-value in ip6_multipath_l3_keys(), root caused to a bad assumption of ICMP header being already pulled in skb->head ip_multipath_l3_keys() does the correct thing, so it is an IPv6 only bug. BUG: KMSAN: uninit-value in ip6_multipath_l3_keys net/ipv6/route.c:1830 [inline] BUG: KMSAN: uninit-value in rt6_multipath_hash+0x5c4/0x640 net/ipv6/route.c:1858 CPU: 0 PID: 4507 Comm: syz-executor661 Not tainted 4.16.0+ #87 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x185/0x1d0 lib/dump_stack.c:53 kmsan_report+0x142/0x240 mm/kmsan/kmsan.c:1067 __msan_warning_32+0x6c/0xb0 mm/kmsan/kmsan_instr.c:683 ip6_multipath_l3_keys net/ipv6/route.c:1830 [inline] rt6_multipath_hash+0x5c4/0x640 net/ipv6/route.c:1858 ip6_route_input+0x65a/0x920 net/ipv6/route.c:1884 ip6_rcv_finish+0x413/0x6e0 net/ipv6/ip6_input.c:69 NF_HOOK include/linux/netfilter.h:288 [inline] ipv6_rcv+0x1e16/0x2340 net/ipv6/ip6_input.c:208 __netif_receive_skb_core+0x47df/0x4a90 net/core/dev.c:4562 __netif_receive_skb net/core/dev.c:4627 [inline] netif_receive_skb_internal+0x49d/0x630 net/core/dev.c:4701 netif_receive_skb+0x230/0x240 net/core/dev.c:4725 tun_rx_batched drivers/net/tun.c:1555 [inline] tun_get_user+0x740f/0x7c60 drivers/net/tun.c:1962 tun_chr_write_iter+0x1d4/0x330 drivers/net/tun.c:1990 call_write_iter include/linux/fs.h:1782 [inline] new_sync_write fs/read_write.c:469 [inline] __vfs_write+0x7fb/0x9f0 fs/read_write.c:482 vfs_write+0x463/0x8d0 fs/read_write.c:544 SYSC_write+0x172/0x360 fs/read_write.c:589 SyS_write+0x55/0x80 fs/read_write.c:581 do_syscall_64+0x309/0x430 arch/x86/entry/common.c:287 entry_SYSCALL_64_after_hwframe+0x3d/0xa2 Fixes: 23aebdacb05d ("ipv6: Compute multipath hash for ICMP errors from offending packet") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Jakub Sitnicki <jkbs@redhat.com> Acked-by: Jakub Sitnicki <jkbs@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-30 00:54:59 +08:00
icmph = skb_header_pointer(skb, skb_transport_offset(skb),
sizeof(_icmph), &_icmph);
if (!icmph)
goto out;
if (icmph->icmp6_type != ICMPV6_DEST_UNREACH &&
icmph->icmp6_type != ICMPV6_PKT_TOOBIG &&
icmph->icmp6_type != ICMPV6_TIME_EXCEED &&
icmph->icmp6_type != ICMPV6_PARAMPROB)
goto out;
inner_iph = skb_header_pointer(skb,
skb_transport_offset(skb) + sizeof(*icmph),
sizeof(_inner_iph), &_inner_iph);
if (!inner_iph)
goto out;
key_iph = inner_iph;
_flkeys = NULL;
out:
if (_flkeys) {
keys->addrs.v6addrs.src = _flkeys->addrs.v6addrs.src;
keys->addrs.v6addrs.dst = _flkeys->addrs.v6addrs.dst;
keys->tags.flow_label = _flkeys->tags.flow_label;
keys->basic.ip_proto = _flkeys->basic.ip_proto;
} else {
keys->addrs.v6addrs.src = key_iph->saddr;
keys->addrs.v6addrs.dst = key_iph->daddr;
keys->tags.flow_label = ip6_flowlabel(key_iph);
keys->basic.ip_proto = key_iph->nexthdr;
}
}
/* if skb is set it will be used and fl6 can be NULL */
u32 rt6_multipath_hash(const struct net *net, const struct flowi6 *fl6,
const struct sk_buff *skb, struct flow_keys *flkeys)
{
struct flow_keys hash_keys;
u32 mhash;
switch (ip6_multipath_hash_policy(net)) {
case 0:
memset(&hash_keys, 0, sizeof(hash_keys));
hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV6_ADDRS;
if (skb) {
ip6_multipath_l3_keys(skb, &hash_keys, flkeys);
} else {
hash_keys.addrs.v6addrs.src = fl6->saddr;
hash_keys.addrs.v6addrs.dst = fl6->daddr;
hash_keys.tags.flow_label = (__force u32)flowi6_get_flowlabel(fl6);
hash_keys.basic.ip_proto = fl6->flowi6_proto;
}
break;
case 1:
if (skb) {
unsigned int flag = FLOW_DISSECTOR_F_STOP_AT_ENCAP;
struct flow_keys keys;
/* short-circuit if we already have L4 hash present */
if (skb->l4_hash)
return skb_get_hash_raw(skb) >> 1;
memset(&hash_keys, 0, sizeof(hash_keys));
if (!flkeys) {
skb_flow_dissect_flow_keys(skb, &keys, flag);
flkeys = &keys;
}
hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV6_ADDRS;
hash_keys.addrs.v6addrs.src = flkeys->addrs.v6addrs.src;
hash_keys.addrs.v6addrs.dst = flkeys->addrs.v6addrs.dst;
hash_keys.ports.src = flkeys->ports.src;
hash_keys.ports.dst = flkeys->ports.dst;
hash_keys.basic.ip_proto = flkeys->basic.ip_proto;
} else {
memset(&hash_keys, 0, sizeof(hash_keys));
hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV6_ADDRS;
hash_keys.addrs.v6addrs.src = fl6->saddr;
hash_keys.addrs.v6addrs.dst = fl6->daddr;
hash_keys.ports.src = fl6->fl6_sport;
hash_keys.ports.dst = fl6->fl6_dport;
hash_keys.basic.ip_proto = fl6->flowi6_proto;
}
break;
}
mhash = flow_hash_from_keys(&hash_keys);
return mhash >> 1;
}
void ip6_route_input(struct sk_buff *skb)
{
const struct ipv6hdr *iph = ipv6_hdr(skb);
struct net *net = dev_net(skb->dev);
int flags = RT6_LOOKUP_F_HAS_SADDR;
struct ip_tunnel_info *tun_info;
struct flowi6 fl6 = {
.flowi6_iif = skb->dev->ifindex,
.daddr = iph->daddr,
.saddr = iph->saddr,
.flowlabel = ip6_flowinfo(iph),
.flowi6_mark = skb->mark,
.flowi6_proto = iph->nexthdr,
};
struct flow_keys *flkeys = NULL, _flkeys;
tun_info = skb_tunnel_info(skb);
if (tun_info && !(tun_info->mode & IP_TUNNEL_INFO_TX))
fl6.flowi6_tun_key.tun_id = tun_info->key.tun_id;
if (fib6_rules_early_flow_dissect(net, skb, &fl6, &_flkeys))
flkeys = &_flkeys;
if (unlikely(fl6.flowi6_proto == IPPROTO_ICMPV6))
fl6.mp_hash = rt6_multipath_hash(net, &fl6, skb, flkeys);
skb_dst_drop(skb);
skb_dst_set(skb,
ip6_route_input_lookup(net, skb->dev, &fl6, skb, flags));
}
static struct rt6_info *ip6_pol_route_output(struct net *net,
struct fib6_table *table,
struct flowi6 *fl6,
const struct sk_buff *skb,
int flags)
{
return ip6_pol_route(net, table, fl6->flowi6_oif, fl6, skb, flags);
}
struct dst_entry *ip6_route_output_flags(struct net *net, const struct sock *sk,
struct flowi6 *fl6, int flags)
{
bool any_src;
if (ipv6_addr_type(&fl6->daddr) &
(IPV6_ADDR_MULTICAST | IPV6_ADDR_LINKLOCAL)) {
struct dst_entry *dst;
dst = l3mdev_link_scope_lookup(net, fl6);
if (dst)
return dst;
}
fl6->flowi6_iif = LOOPBACK_IFINDEX;
any_src = ipv6_addr_any(&fl6->saddr);
if ((sk && sk->sk_bound_dev_if) || rt6_need_strict(&fl6->daddr) ||
(fl6->flowi6_oif && any_src))
flags |= RT6_LOOKUP_F_IFACE;
if (!any_src)
flags |= RT6_LOOKUP_F_HAS_SADDR;
else if (sk)
flags |= rt6_srcprefs2flags(inet6_sk(sk)->srcprefs);
return fib6_rule_lookup(net, fl6, NULL, flags, ip6_pol_route_output);
}
EXPORT_SYMBOL_GPL(ip6_route_output_flags);
struct dst_entry *ip6_blackhole_route(struct net *net, struct dst_entry *dst_orig)
{
struct rt6_info *rt, *ort = (struct rt6_info *) dst_orig;
struct net_device *loopback_dev = net->loopback_dev;
struct dst_entry *new = NULL;
rt = dst_alloc(&ip6_dst_blackhole_ops, loopback_dev, 1,
DST_OBSOLETE_DEAD, 0);
if (rt) {
rt6_info_init(rt);
atomic_inc(&net->ipv6.rt6_stats->fib_rt_alloc);
new = &rt->dst;
new->__use = 1;
new->input = dst_discard;
new->output = dst_discard_out;
dst_copy_metrics(new, &ort->dst);
rt->rt6i_idev = in6_dev_get(loopback_dev);
rt->rt6i_gateway = ort->rt6i_gateway;
rt->rt6i_flags = ort->rt6i_flags & ~RTF_PCPU;
memcpy(&rt->rt6i_dst, &ort->rt6i_dst, sizeof(struct rt6key));
#ifdef CONFIG_IPV6_SUBTREES
memcpy(&rt->rt6i_src, &ort->rt6i_src, sizeof(struct rt6key));
#endif
}
dst_release(dst_orig);
return new ? new : ERR_PTR(-ENOMEM);
}
/*
* Destination cache support functions
*/
static bool fib6_check(struct fib6_info *f6i, u32 cookie)
ipv6: Stop rt6_info from using inet_peer's metrics inet_peer is indexed by the dst address alone. However, the fib6 tree could have multiple routing entries (rt6_info) for the same dst. For example, 1. A /128 dst via multiple gateways. 2. A RTF_CACHE route cloned from a /128 route. In the above cases, all of them will share the same metrics and step on each other. This patch will steer away from inet_peer's metrics and use dst_cow_metrics_generic() for everything. Change Highlights: 1. Remove rt6_cow_metrics() which currently acquires metrics from inet_peer for DST_HOST route (i.e. /128 route). 2. Add rt6i_pmtu to take care of the pmtu update to avoid creating a full size metrics just to override the RTAX_MTU. 3. After (2), the RTF_CACHE route can also share the metrics with its dst.from route, by: dst_init_metrics(&cache_rt->dst, dst_metrics_ptr(cache_rt->dst.from), true); 4. Stop creating RTF_CACHE route by cloning another RTF_CACHE route. Instead, directly clone from rt->dst. [ Currently, cloning from another RTF_CACHE is only possible during rt6_do_redirect(). Also, the old clone is removed from the tree immediately after the new clone is added. ] In case of cloning from an older redirect RTF_CACHE, it should work as before. In case of cloning from an older pmtu RTF_CACHE, this patch will forget the pmtu and re-learn it (if there is any) from the redirected route. The _rt6i_peer and DST_METRICS_FORCE_OVERWRITE will be removed in the next cleanup patch. Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Cc: Steffen Klassert <steffen.klassert@secunet.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-04-29 04:03:06 +08:00
{
u32 rt_cookie = 0;
if (!fib6_get_cookie_safe(f6i, &rt_cookie) || rt_cookie != cookie)
return false;
if (fib6_check_expired(f6i))
return false;
return true;
ipv6: Stop rt6_info from using inet_peer's metrics inet_peer is indexed by the dst address alone. However, the fib6 tree could have multiple routing entries (rt6_info) for the same dst. For example, 1. A /128 dst via multiple gateways. 2. A RTF_CACHE route cloned from a /128 route. In the above cases, all of them will share the same metrics and step on each other. This patch will steer away from inet_peer's metrics and use dst_cow_metrics_generic() for everything. Change Highlights: 1. Remove rt6_cow_metrics() which currently acquires metrics from inet_peer for DST_HOST route (i.e. /128 route). 2. Add rt6i_pmtu to take care of the pmtu update to avoid creating a full size metrics just to override the RTAX_MTU. 3. After (2), the RTF_CACHE route can also share the metrics with its dst.from route, by: dst_init_metrics(&cache_rt->dst, dst_metrics_ptr(cache_rt->dst.from), true); 4. Stop creating RTF_CACHE route by cloning another RTF_CACHE route. Instead, directly clone from rt->dst. [ Currently, cloning from another RTF_CACHE is only possible during rt6_do_redirect(). Also, the old clone is removed from the tree immediately after the new clone is added. ] In case of cloning from an older redirect RTF_CACHE, it should work as before. In case of cloning from an older pmtu RTF_CACHE, this patch will forget the pmtu and re-learn it (if there is any) from the redirected route. The _rt6i_peer and DST_METRICS_FORCE_OVERWRITE will be removed in the next cleanup patch. Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Cc: Steffen Klassert <steffen.klassert@secunet.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-04-29 04:03:06 +08:00
}
static struct dst_entry *rt6_check(struct rt6_info *rt,
struct fib6_info *from,
u32 cookie)
{
u32 rt_cookie = 0;
if ((from && !fib6_get_cookie_safe(from, &rt_cookie)) ||
rt_cookie != cookie)
return NULL;
if (rt6_check_expired(rt))
return NULL;
return &rt->dst;
}
static struct dst_entry *rt6_dst_from_check(struct rt6_info *rt,
struct fib6_info *from,
u32 cookie)
{
if (!__rt6_check_expired(rt) &&
rt->dst.obsolete == DST_OBSOLETE_FORCE_CHK &&
fib6_check(from, cookie))
return &rt->dst;
else
return NULL;
}
static struct dst_entry *ip6_dst_check(struct dst_entry *dst, u32 cookie)
{
struct dst_entry *dst_ret;
struct fib6_info *from;
struct rt6_info *rt;
rt = container_of(dst, struct rt6_info, dst);
rcu_read_lock();
/* All IPV6 dsts are created with ->obsolete set to the value
* DST_OBSOLETE_FORCE_CHK which forces validation calls down
* into this function always.
*/
ipv6: ip6_dst_check needs to check for expired dst_entries On receiving a packet too big icmp error we check if our current cached dst_entry in the socket is still valid. This validation check did not care about the expiration of the (cached) route. The error path I traced down: The socket receives a packet too big mtu notification. It still has a valid dst_entry and thus issues the ip6_rt_pmtu_update on this dst_entry, setting RTF_EXPIRE and updates the dst.expiration value (which could fail because of not up-to-date expiration values, see previous patch). In some seldom cases we race with a) the ip6_fib gc or b) another routing lookup which would result in a recreation of the cached rt6_info from its parent non-cached rt6_info. While copying the rt6_info we reinitialize the metrics store by copying it over from the parent thus invalidating the just installed pmtu update (both dsts use the same key to the inetpeer storage). The dst_entry with the just invalidated metrics data would just get its RTF_EXPIRES flag cleared and would continue to stay valid for the socket. We should have not issued the pmtu update on the already expired dst_entry in the first placed. By checking the expiration on the dst entry and doing a relookup in case it is out of date we close the race because we would install a new rt6_info into the fib before we issue the pmtu update, thus closing this race. Not reliably updating the dst.expire value was fixed by the patch "ipv6: reset dst.expires value when clearing expire flag". Reported-by: Steinar H. Gunderson <sgunderson@bigfoot.com> Reported-by: Valentijn Sessink <valentyn@blub.net> Cc: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Reviewed-by: Eric Dumazet <edumazet@google.com> Tested-by: Valentijn Sessink <valentyn@blub.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-24 13:48:24 +08:00
from = rcu_dereference(rt->from);
ipv6: Stop rt6_info from using inet_peer's metrics inet_peer is indexed by the dst address alone. However, the fib6 tree could have multiple routing entries (rt6_info) for the same dst. For example, 1. A /128 dst via multiple gateways. 2. A RTF_CACHE route cloned from a /128 route. In the above cases, all of them will share the same metrics and step on each other. This patch will steer away from inet_peer's metrics and use dst_cow_metrics_generic() for everything. Change Highlights: 1. Remove rt6_cow_metrics() which currently acquires metrics from inet_peer for DST_HOST route (i.e. /128 route). 2. Add rt6i_pmtu to take care of the pmtu update to avoid creating a full size metrics just to override the RTAX_MTU. 3. After (2), the RTF_CACHE route can also share the metrics with its dst.from route, by: dst_init_metrics(&cache_rt->dst, dst_metrics_ptr(cache_rt->dst.from), true); 4. Stop creating RTF_CACHE route by cloning another RTF_CACHE route. Instead, directly clone from rt->dst. [ Currently, cloning from another RTF_CACHE is only possible during rt6_do_redirect(). Also, the old clone is removed from the tree immediately after the new clone is added. ] In case of cloning from an older redirect RTF_CACHE, it should work as before. In case of cloning from an older pmtu RTF_CACHE, this patch will forget the pmtu and re-learn it (if there is any) from the redirected route. The _rt6i_peer and DST_METRICS_FORCE_OVERWRITE will be removed in the next cleanup patch. Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Cc: Steffen Klassert <steffen.klassert@secunet.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-04-29 04:03:06 +08:00
if (from && (rt->rt6i_flags & RTF_PCPU ||
unlikely(!list_empty(&rt->rt6i_uncached))))
dst_ret = rt6_dst_from_check(rt, from, cookie);
else
dst_ret = rt6_check(rt, from, cookie);
rcu_read_unlock();
return dst_ret;
}
static struct dst_entry *ip6_negative_advice(struct dst_entry *dst)
{
struct rt6_info *rt = (struct rt6_info *) dst;
if (rt) {
if (rt->rt6i_flags & RTF_CACHE) {
rcu_read_lock();
if (rt6_check_expired(rt)) {
rt6_remove_exception_rt(rt);
dst = NULL;
}
rcu_read_unlock();
} else {
dst_release(dst);
dst = NULL;
}
}
return dst;
}
static void ip6_link_failure(struct sk_buff *skb)
{
struct rt6_info *rt;
icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_ADDR_UNREACH, 0);
rt = (struct rt6_info *) skb_dst(skb);
if (rt) {
rcu_read_lock();
ipv6: in case of link failure remove route directly instead of letting it expire We could end up expiring a route which is part of an ecmp route set. Doing so would invalidate the rt->rt6i_nsiblings calculations and could provoke the following panic: [ 80.144667] ------------[ cut here ]------------ [ 80.145172] kernel BUG at net/ipv6/ip6_fib.c:733! [ 80.145172] invalid opcode: 0000 [#1] SMP [ 80.145172] Modules linked in: 8021q nf_conntrack_netbios_ns nf_conntrack_broadcast ipt_MASQUERADE ip6table_mangle ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_filter ebtables ip6table_filter ip6_tables +snd_hda_intel snd_hda_codec snd_hwdep snd_seq snd_seq_device snd_pcm snd_page_alloc snd_timer virtio_balloon snd soundcore i2c_piix4 i2c_core virtio_net virtio_blk [ 80.145172] CPU: 1 PID: 786 Comm: ping6 Not tainted 3.10.0+ #118 [ 80.145172] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 80.145172] task: ffff880117fa0000 ti: ffff880118770000 task.ti: ffff880118770000 [ 80.145172] RIP: 0010:[<ffffffff815f3b5d>] [<ffffffff815f3b5d>] fib6_add+0x75d/0x830 [ 80.145172] RSP: 0018:ffff880118771798 EFLAGS: 00010202 [ 80.145172] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffff88011350e480 [ 80.145172] RDX: ffff88011350e238 RSI: 0000000000000004 RDI: ffff88011350f738 [ 80.145172] RBP: ffff880118771848 R08: ffff880117903280 R09: 0000000000000001 [ 80.145172] R10: 0000000000000000 R11: 0000000000000000 R12: ffff88011350f680 [ 80.145172] R13: ffff880117903280 R14: ffff880118771890 R15: ffff88011350ef90 [ 80.145172] FS: 00007f02b5127740(0000) GS:ffff88011fd00000(0000) knlGS:0000000000000000 [ 80.145172] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 80.145172] CR2: 00007f981322a000 CR3: 00000001181b1000 CR4: 00000000000006e0 [ 80.145172] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 80.145172] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 80.145172] Stack: [ 80.145172] 0000000000000001 ffff880100000000 ffff880100000000 ffff880117903280 [ 80.145172] 0000000000000000 ffff880119a4cf00 0000000000000400 00000000000007fa [ 80.145172] 0000000000000000 0000000000000000 0000000000000000 ffff88011350f680 [ 80.145172] Call Trace: [ 80.145172] [<ffffffff815eeceb>] ? rt6_bind_peer+0x4b/0x90 [ 80.145172] [<ffffffff815ed985>] __ip6_ins_rt+0x45/0x70 [ 80.145172] [<ffffffff815eee35>] ip6_ins_rt+0x35/0x40 [ 80.145172] [<ffffffff815ef1e4>] ip6_pol_route.isra.44+0x3a4/0x4b0 [ 80.145172] [<ffffffff815ef34a>] ip6_pol_route_output+0x2a/0x30 [ 80.145172] [<ffffffff81616077>] fib6_rule_action+0xd7/0x210 [ 80.145172] [<ffffffff815ef320>] ? ip6_pol_route_input+0x30/0x30 [ 80.145172] [<ffffffff81553026>] fib_rules_lookup+0xc6/0x140 [ 80.145172] [<ffffffff81616374>] fib6_rule_lookup+0x44/0x80 [ 80.145172] [<ffffffff815ef320>] ? ip6_pol_route_input+0x30/0x30 [ 80.145172] [<ffffffff815edea3>] ip6_route_output+0x73/0xb0 [ 80.145172] [<ffffffff815dfdf3>] ip6_dst_lookup_tail+0x2c3/0x2e0 [ 80.145172] [<ffffffff813007b1>] ? list_del+0x11/0x40 [ 80.145172] [<ffffffff81082a4c>] ? remove_wait_queue+0x3c/0x50 [ 80.145172] [<ffffffff815dfe4d>] ip6_dst_lookup_flow+0x3d/0xa0 [ 80.145172] [<ffffffff815fda77>] rawv6_sendmsg+0x267/0xc20 [ 80.145172] [<ffffffff815a8a83>] inet_sendmsg+0x63/0xb0 [ 80.145172] [<ffffffff8128eb93>] ? selinux_socket_sendmsg+0x23/0x30 [ 80.145172] [<ffffffff815218d6>] sock_sendmsg+0xa6/0xd0 [ 80.145172] [<ffffffff81524a68>] SYSC_sendto+0x128/0x180 [ 80.145172] [<ffffffff8109825c>] ? update_curr+0xec/0x170 [ 80.145172] [<ffffffff81041d09>] ? kvm_clock_get_cycles+0x9/0x10 [ 80.145172] [<ffffffff810afd1e>] ? __getnstimeofday+0x3e/0xd0 [ 80.145172] [<ffffffff8152509e>] SyS_sendto+0xe/0x10 [ 80.145172] [<ffffffff8164efd9>] system_call_fastpath+0x16/0x1b [ 80.145172] Code: fe ff ff 41 f6 45 2a 06 0f 85 ca fe ff ff 49 8b 7e 08 4c 89 ee e8 94 ef ff ff e9 b9 fe ff ff 48 8b 82 28 05 00 00 e9 01 ff ff ff <0f> 0b 49 8b 54 24 30 0d 00 00 40 00 89 83 14 01 00 00 48 89 53 [ 80.145172] RIP [<ffffffff815f3b5d>] fib6_add+0x75d/0x830 [ 80.145172] RSP <ffff880118771798> [ 80.387413] ---[ end trace 02f20b7a8b81ed95 ]--- [ 80.390154] Kernel panic - not syncing: Fatal exception in interrupt Cc: Nicolas Dichtel <nicolas.dichtel@6wind.com> Cc: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-11 05:00:57 +08:00
if (rt->rt6i_flags & RTF_CACHE) {
rt6_remove_exception_rt(rt);
} else {
struct fib6_info *from;
struct fib6_node *fn;
from = rcu_dereference(rt->from);
if (from) {
fn = rcu_dereference(from->fib6_node);
if (fn && (rt->rt6i_flags & RTF_DEFAULT))
fn->fn_sernum = -1;
}
ipv6: in case of link failure remove route directly instead of letting it expire We could end up expiring a route which is part of an ecmp route set. Doing so would invalidate the rt->rt6i_nsiblings calculations and could provoke the following panic: [ 80.144667] ------------[ cut here ]------------ [ 80.145172] kernel BUG at net/ipv6/ip6_fib.c:733! [ 80.145172] invalid opcode: 0000 [#1] SMP [ 80.145172] Modules linked in: 8021q nf_conntrack_netbios_ns nf_conntrack_broadcast ipt_MASQUERADE ip6table_mangle ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_filter ebtables ip6table_filter ip6_tables +snd_hda_intel snd_hda_codec snd_hwdep snd_seq snd_seq_device snd_pcm snd_page_alloc snd_timer virtio_balloon snd soundcore i2c_piix4 i2c_core virtio_net virtio_blk [ 80.145172] CPU: 1 PID: 786 Comm: ping6 Not tainted 3.10.0+ #118 [ 80.145172] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 80.145172] task: ffff880117fa0000 ti: ffff880118770000 task.ti: ffff880118770000 [ 80.145172] RIP: 0010:[<ffffffff815f3b5d>] [<ffffffff815f3b5d>] fib6_add+0x75d/0x830 [ 80.145172] RSP: 0018:ffff880118771798 EFLAGS: 00010202 [ 80.145172] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffff88011350e480 [ 80.145172] RDX: ffff88011350e238 RSI: 0000000000000004 RDI: ffff88011350f738 [ 80.145172] RBP: ffff880118771848 R08: ffff880117903280 R09: 0000000000000001 [ 80.145172] R10: 0000000000000000 R11: 0000000000000000 R12: ffff88011350f680 [ 80.145172] R13: ffff880117903280 R14: ffff880118771890 R15: ffff88011350ef90 [ 80.145172] FS: 00007f02b5127740(0000) GS:ffff88011fd00000(0000) knlGS:0000000000000000 [ 80.145172] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 80.145172] CR2: 00007f981322a000 CR3: 00000001181b1000 CR4: 00000000000006e0 [ 80.145172] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 80.145172] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 80.145172] Stack: [ 80.145172] 0000000000000001 ffff880100000000 ffff880100000000 ffff880117903280 [ 80.145172] 0000000000000000 ffff880119a4cf00 0000000000000400 00000000000007fa [ 80.145172] 0000000000000000 0000000000000000 0000000000000000 ffff88011350f680 [ 80.145172] Call Trace: [ 80.145172] [<ffffffff815eeceb>] ? rt6_bind_peer+0x4b/0x90 [ 80.145172] [<ffffffff815ed985>] __ip6_ins_rt+0x45/0x70 [ 80.145172] [<ffffffff815eee35>] ip6_ins_rt+0x35/0x40 [ 80.145172] [<ffffffff815ef1e4>] ip6_pol_route.isra.44+0x3a4/0x4b0 [ 80.145172] [<ffffffff815ef34a>] ip6_pol_route_output+0x2a/0x30 [ 80.145172] [<ffffffff81616077>] fib6_rule_action+0xd7/0x210 [ 80.145172] [<ffffffff815ef320>] ? ip6_pol_route_input+0x30/0x30 [ 80.145172] [<ffffffff81553026>] fib_rules_lookup+0xc6/0x140 [ 80.145172] [<ffffffff81616374>] fib6_rule_lookup+0x44/0x80 [ 80.145172] [<ffffffff815ef320>] ? ip6_pol_route_input+0x30/0x30 [ 80.145172] [<ffffffff815edea3>] ip6_route_output+0x73/0xb0 [ 80.145172] [<ffffffff815dfdf3>] ip6_dst_lookup_tail+0x2c3/0x2e0 [ 80.145172] [<ffffffff813007b1>] ? list_del+0x11/0x40 [ 80.145172] [<ffffffff81082a4c>] ? remove_wait_queue+0x3c/0x50 [ 80.145172] [<ffffffff815dfe4d>] ip6_dst_lookup_flow+0x3d/0xa0 [ 80.145172] [<ffffffff815fda77>] rawv6_sendmsg+0x267/0xc20 [ 80.145172] [<ffffffff815a8a83>] inet_sendmsg+0x63/0xb0 [ 80.145172] [<ffffffff8128eb93>] ? selinux_socket_sendmsg+0x23/0x30 [ 80.145172] [<ffffffff815218d6>] sock_sendmsg+0xa6/0xd0 [ 80.145172] [<ffffffff81524a68>] SYSC_sendto+0x128/0x180 [ 80.145172] [<ffffffff8109825c>] ? update_curr+0xec/0x170 [ 80.145172] [<ffffffff81041d09>] ? kvm_clock_get_cycles+0x9/0x10 [ 80.145172] [<ffffffff810afd1e>] ? __getnstimeofday+0x3e/0xd0 [ 80.145172] [<ffffffff8152509e>] SyS_sendto+0xe/0x10 [ 80.145172] [<ffffffff8164efd9>] system_call_fastpath+0x16/0x1b [ 80.145172] Code: fe ff ff 41 f6 45 2a 06 0f 85 ca fe ff ff 49 8b 7e 08 4c 89 ee e8 94 ef ff ff e9 b9 fe ff ff 48 8b 82 28 05 00 00 e9 01 ff ff ff <0f> 0b 49 8b 54 24 30 0d 00 00 40 00 89 83 14 01 00 00 48 89 53 [ 80.145172] RIP [<ffffffff815f3b5d>] fib6_add+0x75d/0x830 [ 80.145172] RSP <ffff880118771798> [ 80.387413] ---[ end trace 02f20b7a8b81ed95 ]--- [ 80.390154] Kernel panic - not syncing: Fatal exception in interrupt Cc: Nicolas Dichtel <nicolas.dichtel@6wind.com> Cc: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-11 05:00:57 +08:00
}
rcu_read_unlock();
}
}
static void rt6_update_expires(struct rt6_info *rt0, int timeout)
{
if (!(rt0->rt6i_flags & RTF_EXPIRES)) {
struct fib6_info *from;
rcu_read_lock();
from = rcu_dereference(rt0->from);
if (from)
rt0->dst.expires = from->expires;
rcu_read_unlock();
}
dst_set_expires(&rt0->dst, timeout);
rt0->rt6i_flags |= RTF_EXPIRES;
}
static void rt6_do_update_pmtu(struct rt6_info *rt, u32 mtu)
{
struct net *net = dev_net(rt->dst.dev);
dst_metric_set(&rt->dst, RTAX_MTU, mtu);
rt->rt6i_flags |= RTF_MODIFIED;
rt6_update_expires(rt, net->ipv6.sysctl.ip6_rt_mtu_expires);
}
ipv6: Avoid creating RTF_CACHE from a rt that is not managed by fib6 tree The original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1272571 The setup has a IPv4 GRE tunnel running in a IPSec. The bug happens when ndisc starts sending router solicitation at the gre interface. The simplified oops stack is like: __lock_acquire+0x1b2/0x1c30 lock_acquire+0xb9/0x140 _raw_write_lock_bh+0x3f/0x50 __ip6_ins_rt+0x2e/0x60 ip6_ins_rt+0x49/0x50 ~~~~~~~~ __ip6_rt_update_pmtu.part.54+0x145/0x250 ip6_rt_update_pmtu+0x2e/0x40 ~~~~~~~~ ip_tunnel_xmit+0x1f1/0xf40 __gre_xmit+0x7a/0x90 ipgre_xmit+0x15a/0x220 dev_hard_start_xmit+0x2bd/0x480 __dev_queue_xmit+0x696/0x730 dev_queue_xmit+0x10/0x20 neigh_direct_output+0x11/0x20 ip6_finish_output2+0x21f/0x770 ip6_finish_output+0xa7/0x1d0 ip6_output+0x56/0x190 ~~~~~~~~ ndisc_send_skb+0x1d9/0x400 ndisc_send_rs+0x88/0xc0 ~~~~~~~~ The rt passed to ip6_rt_update_pmtu() is created by icmp6_dst_alloc() and it is not managed by the fib6 tree, so its rt6i_table == NULL. When __ip6_rt_update_pmtu() creates a RTF_CACHE clone, the newly created clone also has rt6i_table == NULL and it causes the ip6_ins_rt() oops. During pmtu update, we only want to create a RTF_CACHE clone from a rt which is currently managed (or owned) by the fib6 tree. It means either rt->rt6i_node != NULL or rt is a RTF_PCPU clone. It is worth to note that rt6i_table may not be NULL even it is not (yet) managed by the fib6 tree (e.g. addrconf_dst_alloc()). Hence, rt6i_node is a better check instead of rt6i_table. Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reported-by: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-11-12 03:51:06 +08:00
static bool rt6_cache_allowed_for_pmtu(const struct rt6_info *rt)
{
bool from_set;
rcu_read_lock();
from_set = !!rcu_dereference(rt->from);
rcu_read_unlock();
ipv6: Avoid creating RTF_CACHE from a rt that is not managed by fib6 tree The original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1272571 The setup has a IPv4 GRE tunnel running in a IPSec. The bug happens when ndisc starts sending router solicitation at the gre interface. The simplified oops stack is like: __lock_acquire+0x1b2/0x1c30 lock_acquire+0xb9/0x140 _raw_write_lock_bh+0x3f/0x50 __ip6_ins_rt+0x2e/0x60 ip6_ins_rt+0x49/0x50 ~~~~~~~~ __ip6_rt_update_pmtu.part.54+0x145/0x250 ip6_rt_update_pmtu+0x2e/0x40 ~~~~~~~~ ip_tunnel_xmit+0x1f1/0xf40 __gre_xmit+0x7a/0x90 ipgre_xmit+0x15a/0x220 dev_hard_start_xmit+0x2bd/0x480 __dev_queue_xmit+0x696/0x730 dev_queue_xmit+0x10/0x20 neigh_direct_output+0x11/0x20 ip6_finish_output2+0x21f/0x770 ip6_finish_output+0xa7/0x1d0 ip6_output+0x56/0x190 ~~~~~~~~ ndisc_send_skb+0x1d9/0x400 ndisc_send_rs+0x88/0xc0 ~~~~~~~~ The rt passed to ip6_rt_update_pmtu() is created by icmp6_dst_alloc() and it is not managed by the fib6 tree, so its rt6i_table == NULL. When __ip6_rt_update_pmtu() creates a RTF_CACHE clone, the newly created clone also has rt6i_table == NULL and it causes the ip6_ins_rt() oops. During pmtu update, we only want to create a RTF_CACHE clone from a rt which is currently managed (or owned) by the fib6 tree. It means either rt->rt6i_node != NULL or rt is a RTF_PCPU clone. It is worth to note that rt6i_table may not be NULL even it is not (yet) managed by the fib6 tree (e.g. addrconf_dst_alloc()). Hence, rt6i_node is a better check instead of rt6i_table. Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reported-by: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-11-12 03:51:06 +08:00
return !(rt->rt6i_flags & RTF_CACHE) &&
(rt->rt6i_flags & RTF_PCPU || from_set);
ipv6: Avoid creating RTF_CACHE from a rt that is not managed by fib6 tree The original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1272571 The setup has a IPv4 GRE tunnel running in a IPSec. The bug happens when ndisc starts sending router solicitation at the gre interface. The simplified oops stack is like: __lock_acquire+0x1b2/0x1c30 lock_acquire+0xb9/0x140 _raw_write_lock_bh+0x3f/0x50 __ip6_ins_rt+0x2e/0x60 ip6_ins_rt+0x49/0x50 ~~~~~~~~ __ip6_rt_update_pmtu.part.54+0x145/0x250 ip6_rt_update_pmtu+0x2e/0x40 ~~~~~~~~ ip_tunnel_xmit+0x1f1/0xf40 __gre_xmit+0x7a/0x90 ipgre_xmit+0x15a/0x220 dev_hard_start_xmit+0x2bd/0x480 __dev_queue_xmit+0x696/0x730 dev_queue_xmit+0x10/0x20 neigh_direct_output+0x11/0x20 ip6_finish_output2+0x21f/0x770 ip6_finish_output+0xa7/0x1d0 ip6_output+0x56/0x190 ~~~~~~~~ ndisc_send_skb+0x1d9/0x400 ndisc_send_rs+0x88/0xc0 ~~~~~~~~ The rt passed to ip6_rt_update_pmtu() is created by icmp6_dst_alloc() and it is not managed by the fib6 tree, so its rt6i_table == NULL. When __ip6_rt_update_pmtu() creates a RTF_CACHE clone, the newly created clone also has rt6i_table == NULL and it causes the ip6_ins_rt() oops. During pmtu update, we only want to create a RTF_CACHE clone from a rt which is currently managed (or owned) by the fib6 tree. It means either rt->rt6i_node != NULL or rt is a RTF_PCPU clone. It is worth to note that rt6i_table may not be NULL even it is not (yet) managed by the fib6 tree (e.g. addrconf_dst_alloc()). Hence, rt6i_node is a better check instead of rt6i_table. Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reported-by: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-11-12 03:51:06 +08:00
}
static void __ip6_rt_update_pmtu(struct dst_entry *dst, const struct sock *sk,
const struct ipv6hdr *iph, u32 mtu)
{
const struct in6_addr *daddr, *saddr;
struct rt6_info *rt6 = (struct rt6_info *)dst;
if (dst_metric_locked(dst, RTAX_MTU))
return;
if (iph) {
daddr = &iph->daddr;
saddr = &iph->saddr;
} else if (sk) {
daddr = &sk->sk_v6_daddr;
saddr = &inet6_sk(sk)->saddr;
} else {
daddr = NULL;
saddr = NULL;
}
dst_confirm_neigh(dst, daddr);
mtu = max_t(u32, mtu, IPV6_MIN_MTU);
if (mtu >= dst_mtu(dst))
return;
ipv6: Avoid creating RTF_CACHE from a rt that is not managed by fib6 tree The original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1272571 The setup has a IPv4 GRE tunnel running in a IPSec. The bug happens when ndisc starts sending router solicitation at the gre interface. The simplified oops stack is like: __lock_acquire+0x1b2/0x1c30 lock_acquire+0xb9/0x140 _raw_write_lock_bh+0x3f/0x50 __ip6_ins_rt+0x2e/0x60 ip6_ins_rt+0x49/0x50 ~~~~~~~~ __ip6_rt_update_pmtu.part.54+0x145/0x250 ip6_rt_update_pmtu+0x2e/0x40 ~~~~~~~~ ip_tunnel_xmit+0x1f1/0xf40 __gre_xmit+0x7a/0x90 ipgre_xmit+0x15a/0x220 dev_hard_start_xmit+0x2bd/0x480 __dev_queue_xmit+0x696/0x730 dev_queue_xmit+0x10/0x20 neigh_direct_output+0x11/0x20 ip6_finish_output2+0x21f/0x770 ip6_finish_output+0xa7/0x1d0 ip6_output+0x56/0x190 ~~~~~~~~ ndisc_send_skb+0x1d9/0x400 ndisc_send_rs+0x88/0xc0 ~~~~~~~~ The rt passed to ip6_rt_update_pmtu() is created by icmp6_dst_alloc() and it is not managed by the fib6 tree, so its rt6i_table == NULL. When __ip6_rt_update_pmtu() creates a RTF_CACHE clone, the newly created clone also has rt6i_table == NULL and it causes the ip6_ins_rt() oops. During pmtu update, we only want to create a RTF_CACHE clone from a rt which is currently managed (or owned) by the fib6 tree. It means either rt->rt6i_node != NULL or rt is a RTF_PCPU clone. It is worth to note that rt6i_table may not be NULL even it is not (yet) managed by the fib6 tree (e.g. addrconf_dst_alloc()). Hence, rt6i_node is a better check instead of rt6i_table. Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reported-by: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Chris Siebenmann <cks-rhbugzilla@cs.toronto.edu> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-11-12 03:51:06 +08:00
if (!rt6_cache_allowed_for_pmtu(rt6)) {
rt6_do_update_pmtu(rt6, mtu);
/* update rt6_ex->stamp for cache */
if (rt6->rt6i_flags & RTF_CACHE)
rt6_update_exception_stamp_rt(rt6);
} else if (daddr) {
struct fib6_info *from;
struct rt6_info *nrt6;
rcu_read_lock();
from = rcu_dereference(rt6->from);
nrt6 = ip6_rt_cache_alloc(from, daddr, saddr);
if (nrt6) {
rt6_do_update_pmtu(nrt6, mtu);
if (rt6_insert_exception(nrt6, from))
dst_release_immediate(&nrt6->dst);
}
rcu_read_unlock();
}
}
static void ip6_rt_update_pmtu(struct dst_entry *dst, struct sock *sk,
struct sk_buff *skb, u32 mtu)
{
__ip6_rt_update_pmtu(dst, sk, skb ? ipv6_hdr(skb) : NULL, mtu);
}
void ip6_update_pmtu(struct sk_buff *skb, struct net *net, __be32 mtu,
int oif, u32 mark, kuid_t uid)
{
const struct ipv6hdr *iph = (struct ipv6hdr *) skb->data;
struct dst_entry *dst;
struct flowi6 fl6 = {
.flowi6_oif = oif,
.flowi6_mark = mark ? mark : IP6_REPLY_MARK(net, skb->mark),
.daddr = iph->daddr,
.saddr = iph->saddr,
.flowlabel = ip6_flowinfo(iph),
.flowi6_uid = uid,
};
dst = ip6_route_output(net, NULL, &fl6);
if (!dst->error)
__ip6_rt_update_pmtu(dst, NULL, iph, ntohl(mtu));
dst_release(dst);
}
EXPORT_SYMBOL_GPL(ip6_update_pmtu);
void ip6_sk_update_pmtu(struct sk_buff *skb, struct sock *sk, __be32 mtu)
{
int oif = sk->sk_bound_dev_if;
ipv6: datagram: Update dst cache of a connected datagram sk during pmtu update There is a case in connected UDP socket such that getsockopt(IPV6_MTU) will return a stale MTU value. The reproducible sequence could be the following: 1. Create a connected UDP socket 2. Send some datagrams out 3. Receive a ICMPV6_PKT_TOOBIG 4. No new outgoing datagrams to trigger the sk_dst_check() logic to update the sk->sk_dst_cache. 5. getsockopt(IPV6_MTU) returns the mtu from the invalid sk->sk_dst_cache instead of the newly created RTF_CACHE clone. This patch updates the sk->sk_dst_cache for a connected datagram sk during pmtu-update code path. Note that the sk->sk_v6_daddr is used to do the route lookup instead of skb->data (i.e. iph). It is because a UDP socket can become connected after sending out some datagrams in un-connected state. or It can be connected multiple times to different destinations. Hence, iph may not be related to where sk is currently connected to. It is done under '!sock_owned_by_user(sk)' condition because the user may make another ip6_datagram_connect() (i.e changing the sk->sk_v6_daddr) while dst lookup is happening in the pmtu-update code path. For the sock_owned_by_user(sk) == true case, the next patch will introduce a release_cb() which will update the sk->sk_dst_cache. Test: Server (Connected UDP Socket): ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Route Details: [root@arch-fb-vm1 ~]# ip -6 r show | egrep '2fac' 2fac::/64 dev eth0 proto kernel metric 256 pref medium 2fac:face::/64 via 2fac::face dev eth0 metric 1024 pref medium A simple python code to create a connected UDP socket: import socket import errno HOST = '2fac::1' PORT = 8080 s = socket.socket(socket.AF_INET6, socket.SOCK_DGRAM) s.bind((HOST, PORT)) s.connect(('2fac:face::face', 53)) print("connected") while True: try: data = s.recv(1024) except socket.error as se: if se.errno == errno.EMSGSIZE: pmtu = s.getsockopt(41, 24) print("PMTU:%d" % pmtu) break s.close() Python program output after getting a ICMPV6_PKT_TOOBIG: [root@arch-fb-vm1 ~]# python2 ~/devshare/kernel/tasks/fib6/udp-connect-53-8080.py connected PMTU:1300 Cache routes after recieving TOOBIG: [root@arch-fb-vm1 ~]# ip -6 r show table cache 2fac:face::face via 2fac::face dev eth0 metric 0 cache expires 463sec mtu 1300 pref medium Client (Send the ICMPV6_PKT_TOOBIG): ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ scapy is used to generate the TOOBIG message. Here is the scapy script I have used: >>> p=Ether(src='da:75:4d:36:ac:32', dst='52:54:00:12:34:66', type=0x86dd)/IPv6(src='2fac::face', dst='2fac::1')/ICMPv6PacketTooBig(mtu=1300)/IPv6(src='2fac:: 1',dst='2fac:face::face', nh='UDP')/UDP(sport=8080,dport=53) >>> sendp(p, iface='qemubr0') Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu exception") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reported-by: Wei Wang <weiwan@google.com> Cc: Cong Wang <xiyou.wangcong@gmail.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-12 06:29:36 +08:00
struct dst_entry *dst;
if (!oif && skb->dev)
oif = l3mdev_master_ifindex(skb->dev);
ip6_update_pmtu(skb, sock_net(sk), mtu, oif, sk->sk_mark, sk->sk_uid);
ipv6: datagram: Update dst cache of a connected datagram sk during pmtu update There is a case in connected UDP socket such that getsockopt(IPV6_MTU) will return a stale MTU value. The reproducible sequence could be the following: 1. Create a connected UDP socket 2. Send some datagrams out 3. Receive a ICMPV6_PKT_TOOBIG 4. No new outgoing datagrams to trigger the sk_dst_check() logic to update the sk->sk_dst_cache. 5. getsockopt(IPV6_MTU) returns the mtu from the invalid sk->sk_dst_cache instead of the newly created RTF_CACHE clone. This patch updates the sk->sk_dst_cache for a connected datagram sk during pmtu-update code path. Note that the sk->sk_v6_daddr is used to do the route lookup instead of skb->data (i.e. iph). It is because a UDP socket can become connected after sending out some datagrams in un-connected state. or It can be connected multiple times to different destinations. Hence, iph may not be related to where sk is currently connected to. It is done under '!sock_owned_by_user(sk)' condition because the user may make another ip6_datagram_connect() (i.e changing the sk->sk_v6_daddr) while dst lookup is happening in the pmtu-update code path. For the sock_owned_by_user(sk) == true case, the next patch will introduce a release_cb() which will update the sk->sk_dst_cache. Test: Server (Connected UDP Socket): ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Route Details: [root@arch-fb-vm1 ~]# ip -6 r show | egrep '2fac' 2fac::/64 dev eth0 proto kernel metric 256 pref medium 2fac:face::/64 via 2fac::face dev eth0 metric 1024 pref medium A simple python code to create a connected UDP socket: import socket import errno HOST = '2fac::1' PORT = 8080 s = socket.socket(socket.AF_INET6, socket.SOCK_DGRAM) s.bind((HOST, PORT)) s.connect(('2fac:face::face', 53)) print("connected") while True: try: data = s.recv(1024) except socket.error as se: if se.errno == errno.EMSGSIZE: pmtu = s.getsockopt(41, 24) print("PMTU:%d" % pmtu) break s.close() Python program output after getting a ICMPV6_PKT_TOOBIG: [root@arch-fb-vm1 ~]# python2 ~/devshare/kernel/tasks/fib6/udp-connect-53-8080.py connected PMTU:1300 Cache routes after recieving TOOBIG: [root@arch-fb-vm1 ~]# ip -6 r show table cache 2fac:face::face via 2fac::face dev eth0 metric 0 cache expires 463sec mtu 1300 pref medium Client (Send the ICMPV6_PKT_TOOBIG): ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ scapy is used to generate the TOOBIG message. Here is the scapy script I have used: >>> p=Ether(src='da:75:4d:36:ac:32', dst='52:54:00:12:34:66', type=0x86dd)/IPv6(src='2fac::face', dst='2fac::1')/ICMPv6PacketTooBig(mtu=1300)/IPv6(src='2fac:: 1',dst='2fac:face::face', nh='UDP')/UDP(sport=8080,dport=53) >>> sendp(p, iface='qemubr0') Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu exception") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reported-by: Wei Wang <weiwan@google.com> Cc: Cong Wang <xiyou.wangcong@gmail.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-12 06:29:36 +08:00
dst = __sk_dst_get(sk);
if (!dst || !dst->obsolete ||
dst->ops->check(dst, inet6_sk(sk)->dst_cookie))
return;
bh_lock_sock(sk);
if (!sock_owned_by_user(sk) && !ipv6_addr_v4mapped(&sk->sk_v6_daddr))
ip6_datagram_dst_update(sk, false);
bh_unlock_sock(sk);
}
EXPORT_SYMBOL_GPL(ip6_sk_update_pmtu);
void ip6_sk_dst_store_flow(struct sock *sk, struct dst_entry *dst,
const struct flowi6 *fl6)
{
#ifdef CONFIG_IPV6_SUBTREES
struct ipv6_pinfo *np = inet6_sk(sk);
#endif
ip6_dst_store(sk, dst,
ipv6_addr_equal(&fl6->daddr, &sk->sk_v6_daddr) ?
&sk->sk_v6_daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
ipv6_addr_equal(&fl6->saddr, &np->saddr) ?
&np->saddr :
#endif
NULL);
}
/* Handle redirects */
struct ip6rd_flowi {
struct flowi6 fl6;
struct in6_addr gateway;
};
static struct rt6_info *__ip6_route_redirect(struct net *net,
struct fib6_table *table,
struct flowi6 *fl6,
const struct sk_buff *skb,
int flags)
{
struct ip6rd_flowi *rdfl = (struct ip6rd_flowi *)fl6;
struct rt6_info *ret = NULL, *rt_cache;
struct fib6_info *rt;
struct fib6_node *fn;
/* Get the "current" route for this destination and
* check if the redirect has come from appropriate router.
*
* RFC 4861 specifies that redirects should only be
* accepted if they come from the nexthop to the target.
* Due to the way the routes are chosen, this notion
* is a bit fuzzy and one might need to check all possible
* routes.
*/
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_lock();
fn = fib6_node_lookup(&table->tb6_root, &fl6->daddr, &fl6->saddr);
restart:
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
for_each_fib6_node_rt_rcu(fn) {
if (rt->fib6_nh.nh_flags & RTNH_F_DEAD)
continue;
if (fib6_check_expired(rt))
continue;
if (rt->fib6_flags & RTF_REJECT)
break;
if (!(rt->fib6_flags & RTF_GATEWAY))
continue;
if (fl6->flowi6_oif != rt->fib6_nh.nh_dev->ifindex)
continue;
/* rt_cache's gateway might be different from its 'parent'
* in the case of an ip redirect.
* So we keep searching in the exception table if the gateway
* is different.
*/
if (!ipv6_addr_equal(&rdfl->gateway, &rt->fib6_nh.nh_gw)) {
rt_cache = rt6_find_cached_rt(rt,
&fl6->daddr,
&fl6->saddr);
if (rt_cache &&
ipv6_addr_equal(&rdfl->gateway,
&rt_cache->rt6i_gateway)) {
ret = rt_cache;
break;
}
continue;
}
break;
}
if (!rt)
rt = net->ipv6.fib6_null_entry;
else if (rt->fib6_flags & RTF_REJECT) {
ret = net->ipv6.ip6_null_entry;
goto out;
}
if (rt == net->ipv6.fib6_null_entry) {
fn = fib6_backtrack(fn, &fl6->saddr);
if (fn)
goto restart;
}
out:
if (ret)
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
ip6_hold_safe(net, &ret, true);
else
ret = ip6_create_rt_rcu(rt);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
trace_fib6_table_lookup(net, rt, table, fl6);
return ret;
};
static struct dst_entry *ip6_route_redirect(struct net *net,
const struct flowi6 *fl6,
const struct sk_buff *skb,
const struct in6_addr *gateway)
{
int flags = RT6_LOOKUP_F_HAS_SADDR;
struct ip6rd_flowi rdfl;
rdfl.fl6 = *fl6;
rdfl.gateway = *gateway;
return fib6_rule_lookup(net, &rdfl.fl6, skb,
flags, __ip6_route_redirect);
}
void ip6_redirect(struct sk_buff *skb, struct net *net, int oif, u32 mark,
kuid_t uid)
{
const struct ipv6hdr *iph = (struct ipv6hdr *) skb->data;
struct dst_entry *dst;
struct flowi6 fl6 = {
.flowi6_iif = LOOPBACK_IFINDEX,
.flowi6_oif = oif,
.flowi6_mark = mark,
.daddr = iph->daddr,
.saddr = iph->saddr,
.flowlabel = ip6_flowinfo(iph),
.flowi6_uid = uid,
};
dst = ip6_route_redirect(net, &fl6, skb, &ipv6_hdr(skb)->saddr);
rt6_do_redirect(dst, NULL, skb);
dst_release(dst);
}
EXPORT_SYMBOL_GPL(ip6_redirect);
void ip6_redirect_no_header(struct sk_buff *skb, struct net *net, int oif)
{
const struct ipv6hdr *iph = ipv6_hdr(skb);
const struct rd_msg *msg = (struct rd_msg *)icmp6_hdr(skb);
struct dst_entry *dst;
struct flowi6 fl6 = {
.flowi6_iif = LOOPBACK_IFINDEX,
.flowi6_oif = oif,
.daddr = msg->dest,
.saddr = iph->daddr,
.flowi6_uid = sock_net_uid(net, NULL),
};
dst = ip6_route_redirect(net, &fl6, skb, &iph->saddr);
rt6_do_redirect(dst, NULL, skb);
dst_release(dst);
}
void ip6_sk_redirect(struct sk_buff *skb, struct sock *sk)
{
ip6_redirect(skb, sock_net(sk), sk->sk_bound_dev_if, sk->sk_mark,
sk->sk_uid);
}
EXPORT_SYMBOL_GPL(ip6_sk_redirect);
static unsigned int ip6_default_advmss(const struct dst_entry *dst)
{
struct net_device *dev = dst->dev;
unsigned int mtu = dst_mtu(dst);
struct net *net = dev_net(dev);
mtu -= sizeof(struct ipv6hdr) + sizeof(struct tcphdr);
if (mtu < net->ipv6.sysctl.ip6_rt_min_advmss)
mtu = net->ipv6.sysctl.ip6_rt_min_advmss;
/*
* Maximal non-jumbo IPv6 payload is IPV6_MAXPLEN and
* corresponding MSS is IPV6_MAXPLEN - tcp_header_size.
* IPV6_MAXPLEN is also valid and means: "any MSS,
* rely only on pmtu discovery"
*/
if (mtu > IPV6_MAXPLEN - sizeof(struct tcphdr))
mtu = IPV6_MAXPLEN;
return mtu;
}
static unsigned int ip6_mtu(const struct dst_entry *dst)
{
struct inet6_dev *idev;
unsigned int mtu;
ipv6: Stop rt6_info from using inet_peer's metrics inet_peer is indexed by the dst address alone. However, the fib6 tree could have multiple routing entries (rt6_info) for the same dst. For example, 1. A /128 dst via multiple gateways. 2. A RTF_CACHE route cloned from a /128 route. In the above cases, all of them will share the same metrics and step on each other. This patch will steer away from inet_peer's metrics and use dst_cow_metrics_generic() for everything. Change Highlights: 1. Remove rt6_cow_metrics() which currently acquires metrics from inet_peer for DST_HOST route (i.e. /128 route). 2. Add rt6i_pmtu to take care of the pmtu update to avoid creating a full size metrics just to override the RTAX_MTU. 3. After (2), the RTF_CACHE route can also share the metrics with its dst.from route, by: dst_init_metrics(&cache_rt->dst, dst_metrics_ptr(cache_rt->dst.from), true); 4. Stop creating RTF_CACHE route by cloning another RTF_CACHE route. Instead, directly clone from rt->dst. [ Currently, cloning from another RTF_CACHE is only possible during rt6_do_redirect(). Also, the old clone is removed from the tree immediately after the new clone is added. ] In case of cloning from an older redirect RTF_CACHE, it should work as before. In case of cloning from an older pmtu RTF_CACHE, this patch will forget the pmtu and re-learn it (if there is any) from the redirected route. The _rt6i_peer and DST_METRICS_FORCE_OVERWRITE will be removed in the next cleanup patch. Signed-off-by: Martin KaFai Lau <kafai@fb.com> Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Cc: Steffen Klassert <steffen.klassert@secunet.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-04-29 04:03:06 +08:00
mtu = dst_metric_raw(dst, RTAX_MTU);
if (mtu)
goto out;
mtu = IPV6_MIN_MTU;
rcu_read_lock();
idev = __in6_dev_get(dst->dev);
if (idev)
mtu = idev->cnf.mtu6;
rcu_read_unlock();
out:
mtu = min_t(unsigned int, mtu, IP6_MAX_MTU);
return mtu - lwtunnel_headroom(dst->lwtstate, mtu);
}
/* MTU selection:
* 1. mtu on route is locked - use it
* 2. mtu from nexthop exception
* 3. mtu from egress device
*
* based on ip6_dst_mtu_forward and exception logic of
* rt6_find_cached_rt; called with rcu_read_lock
*/
u32 ip6_mtu_from_fib6(struct fib6_info *f6i, struct in6_addr *daddr,
struct in6_addr *saddr)
{
struct rt6_exception_bucket *bucket;
struct rt6_exception *rt6_ex;
struct in6_addr *src_key;
struct inet6_dev *idev;
u32 mtu = 0;
if (unlikely(fib6_metric_locked(f6i, RTAX_MTU))) {
mtu = f6i->fib6_pmtu;
if (mtu)
goto out;
}
src_key = NULL;
#ifdef CONFIG_IPV6_SUBTREES
if (f6i->fib6_src.plen)
src_key = saddr;
#endif
bucket = rcu_dereference(f6i->rt6i_exception_bucket);
rt6_ex = __rt6_find_exception_rcu(&bucket, daddr, src_key);
if (rt6_ex && !rt6_check_expired(rt6_ex->rt6i))
mtu = dst_metric_raw(&rt6_ex->rt6i->dst, RTAX_MTU);
if (likely(!mtu)) {
struct net_device *dev = fib6_info_nh_dev(f6i);
mtu = IPV6_MIN_MTU;
idev = __in6_dev_get(dev);
if (idev && idev->cnf.mtu6 > mtu)
mtu = idev->cnf.mtu6;
}
mtu = min_t(unsigned int, mtu, IP6_MAX_MTU);
out:
return mtu - lwtunnel_headroom(fib6_info_nh_lwt(f6i), mtu);
}
struct dst_entry *icmp6_dst_alloc(struct net_device *dev,
struct flowi6 *fl6)
{
struct dst_entry *dst;
struct rt6_info *rt;
struct inet6_dev *idev = in6_dev_get(dev);
struct net *net = dev_net(dev);
if (unlikely(!idev))
return ERR_PTR(-ENODEV);
rt = ip6_dst_alloc(net, dev, 0);
if (unlikely(!rt)) {
in6_dev_put(idev);
dst = ERR_PTR(-ENOMEM);
goto out;
}
rt->dst.flags |= DST_HOST;
rt->dst.input = ip6_input;
rt->dst.output = ip6_output;
rt->rt6i_gateway = fl6->daddr;
rt->rt6i_dst.addr = fl6->daddr;
rt->rt6i_dst.plen = 128;
rt->rt6i_idev = idev;
dst_metric_set(&rt->dst, RTAX_HOPLIMIT, 0);
/* Add this dst into uncached_list so that rt6_disable_ip() can
* do proper release of the net_device
*/
rt6_uncached_list_add(rt);
atomic_inc(&net->ipv6.rt6_stats->fib_rt_uncache);
dst = xfrm_lookup(net, &rt->dst, flowi6_to_flowi(fl6), NULL, 0);
out:
return dst;
}
static int ip6_dst_gc(struct dst_ops *ops)
{
struct net *net = container_of(ops, struct net, ipv6.ip6_dst_ops);
int rt_min_interval = net->ipv6.sysctl.ip6_rt_gc_min_interval;
int rt_max_size = net->ipv6.sysctl.ip6_rt_max_size;
int rt_elasticity = net->ipv6.sysctl.ip6_rt_gc_elasticity;
int rt_gc_timeout = net->ipv6.sysctl.ip6_rt_gc_timeout;
unsigned long rt_last_gc = net->ipv6.ip6_rt_last_gc;
int entries;
entries = dst_entries_get_fast(ops);
if (time_after(rt_last_gc + rt_min_interval, jiffies) &&
entries <= rt_max_size)
goto out;
net->ipv6.ip6_rt_gc_expire++;
fib6_run_gc(net->ipv6.ip6_rt_gc_expire, net, true);
entries = dst_entries_get_slow(ops);
if (entries < ops->gc_thresh)
net->ipv6.ip6_rt_gc_expire = rt_gc_timeout>>1;
out:
net->ipv6.ip6_rt_gc_expire -= net->ipv6.ip6_rt_gc_expire>>rt_elasticity;
return entries > rt_max_size;
}
net: ipv6: Use passed in table for nexthop lookups Similar to 3bfd847203c6 ("net: Use passed in table for nexthop lookups") for IPv4, if the route spec contains a table id use that to lookup the next hop first and fall back to a full lookup if it fails (per the fix 4c9bcd117918b ("net: Fix nexthop lookups")). Example: root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 RTNETLINK answers: No route to host Route add fails even though 2100:1::64 is a reachable next hop: root@kenny:~# ping6 -I red 2100:1::64 ping6: Warning: source address might be selected on device other than red. PING 2100:1::64(2100:1::64) from 2100:1::1 red: 56 data bytes 64 bytes from 2100:1::64: icmp_seq=1 ttl=64 time=1.33 ms With this patch: root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium 2100:3::/64 via 2100:1::64 dev eth1 metric 1024 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-25 12:26:04 +08:00
static struct rt6_info *ip6_nh_lookup_table(struct net *net,
struct fib6_config *cfg,
const struct in6_addr *gw_addr,
u32 tbid, int flags)
net: ipv6: Use passed in table for nexthop lookups Similar to 3bfd847203c6 ("net: Use passed in table for nexthop lookups") for IPv4, if the route spec contains a table id use that to lookup the next hop first and fall back to a full lookup if it fails (per the fix 4c9bcd117918b ("net: Fix nexthop lookups")). Example: root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 RTNETLINK answers: No route to host Route add fails even though 2100:1::64 is a reachable next hop: root@kenny:~# ping6 -I red 2100:1::64 ping6: Warning: source address might be selected on device other than red. PING 2100:1::64(2100:1::64) from 2100:1::1 red: 56 data bytes 64 bytes from 2100:1::64: icmp_seq=1 ttl=64 time=1.33 ms With this patch: root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium 2100:3::/64 via 2100:1::64 dev eth1 metric 1024 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-25 12:26:04 +08:00
{
struct flowi6 fl6 = {
.flowi6_oif = cfg->fc_ifindex,
.daddr = *gw_addr,
.saddr = cfg->fc_prefsrc,
};
struct fib6_table *table;
struct rt6_info *rt;
table = fib6_get_table(net, tbid);
net: ipv6: Use passed in table for nexthop lookups Similar to 3bfd847203c6 ("net: Use passed in table for nexthop lookups") for IPv4, if the route spec contains a table id use that to lookup the next hop first and fall back to a full lookup if it fails (per the fix 4c9bcd117918b ("net: Fix nexthop lookups")). Example: root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 RTNETLINK answers: No route to host Route add fails even though 2100:1::64 is a reachable next hop: root@kenny:~# ping6 -I red 2100:1::64 ping6: Warning: source address might be selected on device other than red. PING 2100:1::64(2100:1::64) from 2100:1::1 red: 56 data bytes 64 bytes from 2100:1::64: icmp_seq=1 ttl=64 time=1.33 ms With this patch: root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium 2100:3::/64 via 2100:1::64 dev eth1 metric 1024 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-25 12:26:04 +08:00
if (!table)
return NULL;
if (!ipv6_addr_any(&cfg->fc_prefsrc))
flags |= RT6_LOOKUP_F_HAS_SADDR;
flags |= RT6_LOOKUP_F_IGNORE_LINKSTATE;
rt = ip6_pol_route(net, table, cfg->fc_ifindex, &fl6, NULL, flags);
net: ipv6: Use passed in table for nexthop lookups Similar to 3bfd847203c6 ("net: Use passed in table for nexthop lookups") for IPv4, if the route spec contains a table id use that to lookup the next hop first and fall back to a full lookup if it fails (per the fix 4c9bcd117918b ("net: Fix nexthop lookups")). Example: root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 RTNETLINK answers: No route to host Route add fails even though 2100:1::64 is a reachable next hop: root@kenny:~# ping6 -I red 2100:1::64 ping6: Warning: source address might be selected on device other than red. PING 2100:1::64(2100:1::64) from 2100:1::1 red: 56 data bytes 64 bytes from 2100:1::64: icmp_seq=1 ttl=64 time=1.33 ms With this patch: root@kenny:~# ip -6 ro add table red 2100:3::/64 via 2100:1::64 root@kenny:~# ip -6 ro ls table red local 2100:1::1 dev lo proto none metric 0 pref medium 2100:1::/120 dev eth1 proto kernel metric 256 pref medium local 2100:2::1 dev lo proto none metric 0 pref medium 2100:2::/120 dev eth2 proto kernel metric 256 pref medium 2100:3::/64 via 2100:1::64 dev eth1 metric 1024 pref medium local fe80::e0:f9ff:fe09:3cac dev lo proto none metric 0 pref medium local fe80::e0:f9ff:fe1c:b974 dev lo proto none metric 0 pref medium fe80::/64 dev eth1 proto kernel metric 256 pref medium fe80::/64 dev eth2 proto kernel metric 256 pref medium ff00::/8 dev red metric 256 pref medium ff00::/8 dev eth1 metric 256 pref medium ff00::/8 dev eth2 metric 256 pref medium unreachable default dev lo metric 240 error -113 pref medium Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-25 12:26:04 +08:00
/* if table lookup failed, fall back to full lookup */
if (rt == net->ipv6.ip6_null_entry) {
ip6_rt_put(rt);
rt = NULL;
}
return rt;
}
static int ip6_route_check_nh_onlink(struct net *net,
struct fib6_config *cfg,
const struct net_device *dev,
struct netlink_ext_ack *extack)
{
u32 tbid = l3mdev_fib_table(dev) ? : RT_TABLE_MAIN;
const struct in6_addr *gw_addr = &cfg->fc_gateway;
u32 flags = RTF_LOCAL | RTF_ANYCAST | RTF_REJECT;
struct rt6_info *grt;
int err;
err = 0;
grt = ip6_nh_lookup_table(net, cfg, gw_addr, tbid, 0);
if (grt) {
if (!grt->dst.error &&
/* ignore match if it is the default route */
grt->from && !ipv6_addr_any(&grt->from->fib6_dst.addr) &&
(grt->rt6i_flags & flags || dev != grt->dst.dev)) {
NL_SET_ERR_MSG(extack,
"Nexthop has invalid gateway or device mismatch");
err = -EINVAL;
}
ip6_rt_put(grt);
}
return err;
}
static int ip6_route_check_nh(struct net *net,
struct fib6_config *cfg,
struct net_device **_dev,
struct inet6_dev **idev)
{
const struct in6_addr *gw_addr = &cfg->fc_gateway;
struct net_device *dev = _dev ? *_dev : NULL;
struct rt6_info *grt = NULL;
int err = -EHOSTUNREACH;
if (cfg->fc_table) {
int flags = RT6_LOOKUP_F_IFACE;
grt = ip6_nh_lookup_table(net, cfg, gw_addr,
cfg->fc_table, flags);
if (grt) {
if (grt->rt6i_flags & RTF_GATEWAY ||
(dev && dev != grt->dst.dev)) {
ip6_rt_put(grt);
grt = NULL;
}
}
}
if (!grt)
grt = rt6_lookup(net, gw_addr, NULL, cfg->fc_ifindex, NULL, 1);
if (!grt)
goto out;
if (dev) {
if (dev != grt->dst.dev) {
ip6_rt_put(grt);
goto out;
}
} else {
*_dev = dev = grt->dst.dev;
*idev = grt->rt6i_idev;
dev_hold(dev);
in6_dev_hold(grt->rt6i_idev);
}
if (!(grt->rt6i_flags & RTF_GATEWAY))
err = 0;
ip6_rt_put(grt);
out:
return err;
}
static int ip6_validate_gw(struct net *net, struct fib6_config *cfg,
struct net_device **_dev, struct inet6_dev **idev,
struct netlink_ext_ack *extack)
{
const struct in6_addr *gw_addr = &cfg->fc_gateway;
int gwa_type = ipv6_addr_type(gw_addr);
net/ipv6: Change address check to always take a device argument ipv6_chk_addr_and_flags determines if an address is a local address and optionally if it is an address on a specific device. For example, it is called by ip6_route_info_create to determine if a given gateway address is a local address. The address check currently does not consider L3 domains and as a result does not allow a route to be added in one VRF if the nexthop points to an address in a second VRF. e.g., $ ip route add 2001:db8:1::/64 vrf r2 via 2001:db8:102::23 Error: Invalid gateway address. where 2001:db8:102::23 is an address on an interface in vrf r1. ipv6_chk_addr_and_flags needs to allow callers to always pass in a device with a separate argument to not limit the address to the specific device. The device is used used to determine the L3 domain of interest. To that end add an argument to skip the device check and update callers to always pass a device where possible and use the new argument to mean any address in the domain. Update a handful of users of ipv6_chk_addr with a NULL dev argument. This patch handles the change to these callers without adding the domain check. ip6_validate_gw needs to handle 2 cases - one where the device is given as part of the nexthop spec and the other where the device is resolved. There is at least 1 VRF case where deferring the check to only after the route lookup has resolved the device fails with an unintuitive error "RTNETLINK answers: No route to host" as opposed to the preferred "Error: Gateway can not be a local address." The 'no route to host' error is because of the fallback to a full lookup. The check is done twice to avoid this error. Signed-off-by: David Ahern <dsahern@gmail.com> Reviewed-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-13 23:29:37 +08:00
bool skip_dev = gwa_type & IPV6_ADDR_LINKLOCAL ? false : true;
const struct net_device *dev = *_dev;
net/ipv6: Change address check to always take a device argument ipv6_chk_addr_and_flags determines if an address is a local address and optionally if it is an address on a specific device. For example, it is called by ip6_route_info_create to determine if a given gateway address is a local address. The address check currently does not consider L3 domains and as a result does not allow a route to be added in one VRF if the nexthop points to an address in a second VRF. e.g., $ ip route add 2001:db8:1::/64 vrf r2 via 2001:db8:102::23 Error: Invalid gateway address. where 2001:db8:102::23 is an address on an interface in vrf r1. ipv6_chk_addr_and_flags needs to allow callers to always pass in a device with a separate argument to not limit the address to the specific device. The device is used used to determine the L3 domain of interest. To that end add an argument to skip the device check and update callers to always pass a device where possible and use the new argument to mean any address in the domain. Update a handful of users of ipv6_chk_addr with a NULL dev argument. This patch handles the change to these callers without adding the domain check. ip6_validate_gw needs to handle 2 cases - one where the device is given as part of the nexthop spec and the other where the device is resolved. There is at least 1 VRF case where deferring the check to only after the route lookup has resolved the device fails with an unintuitive error "RTNETLINK answers: No route to host" as opposed to the preferred "Error: Gateway can not be a local address." The 'no route to host' error is because of the fallback to a full lookup. The check is done twice to avoid this error. Signed-off-by: David Ahern <dsahern@gmail.com> Reviewed-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-13 23:29:37 +08:00
bool need_addr_check = !dev;
int err = -EINVAL;
/* if gw_addr is local we will fail to detect this in case
* address is still TENTATIVE (DAD in progress). rt6_lookup()
* will return already-added prefix route via interface that
* prefix route was assigned to, which might be non-loopback.
*/
net/ipv6: Change address check to always take a device argument ipv6_chk_addr_and_flags determines if an address is a local address and optionally if it is an address on a specific device. For example, it is called by ip6_route_info_create to determine if a given gateway address is a local address. The address check currently does not consider L3 domains and as a result does not allow a route to be added in one VRF if the nexthop points to an address in a second VRF. e.g., $ ip route add 2001:db8:1::/64 vrf r2 via 2001:db8:102::23 Error: Invalid gateway address. where 2001:db8:102::23 is an address on an interface in vrf r1. ipv6_chk_addr_and_flags needs to allow callers to always pass in a device with a separate argument to not limit the address to the specific device. The device is used used to determine the L3 domain of interest. To that end add an argument to skip the device check and update callers to always pass a device where possible and use the new argument to mean any address in the domain. Update a handful of users of ipv6_chk_addr with a NULL dev argument. This patch handles the change to these callers without adding the domain check. ip6_validate_gw needs to handle 2 cases - one where the device is given as part of the nexthop spec and the other where the device is resolved. There is at least 1 VRF case where deferring the check to only after the route lookup has resolved the device fails with an unintuitive error "RTNETLINK answers: No route to host" as opposed to the preferred "Error: Gateway can not be a local address." The 'no route to host' error is because of the fallback to a full lookup. The check is done twice to avoid this error. Signed-off-by: David Ahern <dsahern@gmail.com> Reviewed-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-13 23:29:37 +08:00
if (dev &&
ipv6_chk_addr_and_flags(net, gw_addr, dev, skip_dev, 0, 0)) {
NL_SET_ERR_MSG(extack, "Gateway can not be a local address");
goto out;
}
if (gwa_type != (IPV6_ADDR_LINKLOCAL | IPV6_ADDR_UNICAST)) {
/* IPv6 strictly inhibits using not link-local
* addresses as nexthop address.
* Otherwise, router will not able to send redirects.
* It is very good, but in some (rare!) circumstances
* (SIT, PtP, NBMA NOARP links) it is handy to allow
* some exceptions. --ANK
* We allow IPv4-mapped nexthops to support RFC4798-type
* addressing
*/
if (!(gwa_type & (IPV6_ADDR_UNICAST | IPV6_ADDR_MAPPED))) {
NL_SET_ERR_MSG(extack, "Invalid gateway address");
goto out;
}
if (cfg->fc_flags & RTNH_F_ONLINK)
err = ip6_route_check_nh_onlink(net, cfg, dev, extack);
else
err = ip6_route_check_nh(net, cfg, _dev, idev);
if (err)
goto out;
}
/* reload in case device was changed */
dev = *_dev;
err = -EINVAL;
if (!dev) {
NL_SET_ERR_MSG(extack, "Egress device not specified");
goto out;
} else if (dev->flags & IFF_LOOPBACK) {
NL_SET_ERR_MSG(extack,
"Egress device can not be loopback device for this route");
goto out;
}
net/ipv6: Change address check to always take a device argument ipv6_chk_addr_and_flags determines if an address is a local address and optionally if it is an address on a specific device. For example, it is called by ip6_route_info_create to determine if a given gateway address is a local address. The address check currently does not consider L3 domains and as a result does not allow a route to be added in one VRF if the nexthop points to an address in a second VRF. e.g., $ ip route add 2001:db8:1::/64 vrf r2 via 2001:db8:102::23 Error: Invalid gateway address. where 2001:db8:102::23 is an address on an interface in vrf r1. ipv6_chk_addr_and_flags needs to allow callers to always pass in a device with a separate argument to not limit the address to the specific device. The device is used used to determine the L3 domain of interest. To that end add an argument to skip the device check and update callers to always pass a device where possible and use the new argument to mean any address in the domain. Update a handful of users of ipv6_chk_addr with a NULL dev argument. This patch handles the change to these callers without adding the domain check. ip6_validate_gw needs to handle 2 cases - one where the device is given as part of the nexthop spec and the other where the device is resolved. There is at least 1 VRF case where deferring the check to only after the route lookup has resolved the device fails with an unintuitive error "RTNETLINK answers: No route to host" as opposed to the preferred "Error: Gateway can not be a local address." The 'no route to host' error is because of the fallback to a full lookup. The check is done twice to avoid this error. Signed-off-by: David Ahern <dsahern@gmail.com> Reviewed-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-13 23:29:37 +08:00
/* if we did not check gw_addr above, do so now that the
* egress device has been resolved.
*/
if (need_addr_check &&
ipv6_chk_addr_and_flags(net, gw_addr, dev, skip_dev, 0, 0)) {
NL_SET_ERR_MSG(extack, "Gateway can not be a local address");
goto out;
}
err = 0;
out:
return err;
}
static struct fib6_info *ip6_route_info_create(struct fib6_config *cfg,
gfp_t gfp_flags,
struct netlink_ext_ack *extack)
{
struct net *net = cfg->fc_nlinfo.nl_net;
struct fib6_info *rt = NULL;
struct net_device *dev = NULL;
struct inet6_dev *idev = NULL;
struct fib6_table *table;
int addr_type;
int err = -EINVAL;
net: ipv6: RTF_PCPU should not be settable from userspace Andrey reported a fault in the IPv6 route code: kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Modules linked in: CPU: 1 PID: 4035 Comm: a.out Not tainted 4.11.0-rc7+ #250 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 task: ffff880069809600 task.stack: ffff880062dc8000 RIP: 0010:ip6_rt_cache_alloc+0xa6/0x560 net/ipv6/route.c:975 RSP: 0018:ffff880062dced30 EFLAGS: 00010206 RAX: dffffc0000000000 RBX: ffff8800670561c0 RCX: 0000000000000006 RDX: 0000000000000003 RSI: ffff880062dcfb28 RDI: 0000000000000018 RBP: ffff880062dced68 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: ffff880062dcfb28 R14: dffffc0000000000 R15: 0000000000000000 FS: 00007feebe37e7c0(0000) GS:ffff88006cb00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000205a0fe4 CR3: 000000006b5c9000 CR4: 00000000000006e0 Call Trace: ip6_pol_route+0x1512/0x1f20 net/ipv6/route.c:1128 ip6_pol_route_output+0x4c/0x60 net/ipv6/route.c:1212 ... Andrey's syzkaller program passes rtmsg.rtmsg_flags with the RTF_PCPU bit set. Flags passed to the kernel are blindly copied to the allocated rt6_info by ip6_route_info_create making a newly inserted route appear as though it is a per-cpu route. ip6_rt_cache_alloc sees the flag set and expects rt->dst.from to be set - which it is not since it is not really a per-cpu copy. The subsequent call to __ip6_dst_alloc then generates the fault. Fix by checking for the flag and failing with EINVAL. Fixes: d52d3997f843f ("ipv6: Create percpu rt6_info") Reported-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Tested-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-20 05:19:43 +08:00
/* RTF_PCPU is an internal flag; can not be set by userspace */
if (cfg->fc_flags & RTF_PCPU) {
NL_SET_ERR_MSG(extack, "Userspace can not set RTF_PCPU");
net: ipv6: RTF_PCPU should not be settable from userspace Andrey reported a fault in the IPv6 route code: kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Modules linked in: CPU: 1 PID: 4035 Comm: a.out Not tainted 4.11.0-rc7+ #250 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 task: ffff880069809600 task.stack: ffff880062dc8000 RIP: 0010:ip6_rt_cache_alloc+0xa6/0x560 net/ipv6/route.c:975 RSP: 0018:ffff880062dced30 EFLAGS: 00010206 RAX: dffffc0000000000 RBX: ffff8800670561c0 RCX: 0000000000000006 RDX: 0000000000000003 RSI: ffff880062dcfb28 RDI: 0000000000000018 RBP: ffff880062dced68 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: ffff880062dcfb28 R14: dffffc0000000000 R15: 0000000000000000 FS: 00007feebe37e7c0(0000) GS:ffff88006cb00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000205a0fe4 CR3: 000000006b5c9000 CR4: 00000000000006e0 Call Trace: ip6_pol_route+0x1512/0x1f20 net/ipv6/route.c:1128 ip6_pol_route_output+0x4c/0x60 net/ipv6/route.c:1212 ... Andrey's syzkaller program passes rtmsg.rtmsg_flags with the RTF_PCPU bit set. Flags passed to the kernel are blindly copied to the allocated rt6_info by ip6_route_info_create making a newly inserted route appear as though it is a per-cpu route. ip6_rt_cache_alloc sees the flag set and expects rt->dst.from to be set - which it is not since it is not really a per-cpu copy. The subsequent call to __ip6_dst_alloc then generates the fault. Fix by checking for the flag and failing with EINVAL. Fixes: d52d3997f843f ("ipv6: Create percpu rt6_info") Reported-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Tested-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-20 05:19:43 +08:00
goto out;
}
net: ipv6: RTF_PCPU should not be settable from userspace Andrey reported a fault in the IPv6 route code: kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Modules linked in: CPU: 1 PID: 4035 Comm: a.out Not tainted 4.11.0-rc7+ #250 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 task: ffff880069809600 task.stack: ffff880062dc8000 RIP: 0010:ip6_rt_cache_alloc+0xa6/0x560 net/ipv6/route.c:975 RSP: 0018:ffff880062dced30 EFLAGS: 00010206 RAX: dffffc0000000000 RBX: ffff8800670561c0 RCX: 0000000000000006 RDX: 0000000000000003 RSI: ffff880062dcfb28 RDI: 0000000000000018 RBP: ffff880062dced68 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: ffff880062dcfb28 R14: dffffc0000000000 R15: 0000000000000000 FS: 00007feebe37e7c0(0000) GS:ffff88006cb00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000205a0fe4 CR3: 000000006b5c9000 CR4: 00000000000006e0 Call Trace: ip6_pol_route+0x1512/0x1f20 net/ipv6/route.c:1128 ip6_pol_route_output+0x4c/0x60 net/ipv6/route.c:1212 ... Andrey's syzkaller program passes rtmsg.rtmsg_flags with the RTF_PCPU bit set. Flags passed to the kernel are blindly copied to the allocated rt6_info by ip6_route_info_create making a newly inserted route appear as though it is a per-cpu route. ip6_rt_cache_alloc sees the flag set and expects rt->dst.from to be set - which it is not since it is not really a per-cpu copy. The subsequent call to __ip6_dst_alloc then generates the fault. Fix by checking for the flag and failing with EINVAL. Fixes: d52d3997f843f ("ipv6: Create percpu rt6_info") Reported-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Tested-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-20 05:19:43 +08:00
ipv6: prevent user from adding cached routes Cached routes should only be created by the system when receiving pmtu discovery or ip redirect msg. Users should not be allowed to create cached routes. Furthermore, after the patch series to move cached routes into exception table, user added cached routes will trigger the following warning in fib6_add(): WARNING: CPU: 0 PID: 2985 at net/ipv6/ip6_fib.c:1137 fib6_add+0x20d9/0x2c10 net/ipv6/ip6_fib.c:1137 Kernel panic - not syncing: panic_on_warn set ... CPU: 0 PID: 2985 Comm: syzkaller320388 Not tainted 4.14.0-rc3+ #74 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:16 [inline] dump_stack+0x194/0x257 lib/dump_stack.c:52 panic+0x1e4/0x417 kernel/panic.c:181 __warn+0x1c4/0x1d9 kernel/panic.c:542 report_bug+0x211/0x2d0 lib/bug.c:183 fixup_bug+0x40/0x90 arch/x86/kernel/traps.c:178 do_trap_no_signal arch/x86/kernel/traps.c:212 [inline] do_trap+0x260/0x390 arch/x86/kernel/traps.c:261 do_error_trap+0x120/0x390 arch/x86/kernel/traps.c:298 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:311 invalid_op+0x18/0x20 arch/x86/entry/entry_64.S:905 RIP: 0010:fib6_add+0x20d9/0x2c10 net/ipv6/ip6_fib.c:1137 RSP: 0018:ffff8801cf09f6a0 EFLAGS: 00010297 RAX: ffff8801ce45e340 RBX: 1ffff10039e13eec RCX: ffff8801d749c814 RDX: 0000000000000000 RSI: ffff8801d749c700 RDI: ffff8801d749c780 RBP: ffff8801cf09fa08 R08: 0000000000000000 R09: ffff8801cf09f360 R10: ffff8801cf09f2d8 R11: 1ffff10039c8befb R12: 0000000000000001 R13: dffffc0000000000 R14: ffff8801d749c700 R15: ffffffff860655c0 __ip6_ins_rt+0x6c/0x90 net/ipv6/route.c:1011 ip6_route_add+0x148/0x1a0 net/ipv6/route.c:2782 ipv6_route_ioctl+0x4d5/0x690 net/ipv6/route.c:3291 inet6_ioctl+0xef/0x1e0 net/ipv6/af_inet6.c:521 sock_do_ioctl+0x65/0xb0 net/socket.c:961 sock_ioctl+0x2c2/0x440 net/socket.c:1058 vfs_ioctl fs/ioctl.c:45 [inline] do_vfs_ioctl+0x1b1/0x1530 fs/ioctl.c:685 SYSC_ioctl fs/ioctl.c:700 [inline] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:691 entry_SYSCALL_64_fastpath+0x1f/0xbe So we fix this by failing the attemp to add cached routes from userspace with returning EINVAL error. Fixes: 2b760fcf5cfb ("ipv6: hook up exception table to store dst cache") Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-28 08:30:12 +08:00
/* RTF_CACHE is an internal flag; can not be set by userspace */
if (cfg->fc_flags & RTF_CACHE) {
NL_SET_ERR_MSG(extack, "Userspace can not set RTF_CACHE");
goto out;
}
if (cfg->fc_type > RTN_MAX) {
NL_SET_ERR_MSG(extack, "Invalid route type");
goto out;
}
if (cfg->fc_dst_len > 128) {
NL_SET_ERR_MSG(extack, "Invalid prefix length");
goto out;
}
if (cfg->fc_src_len > 128) {
NL_SET_ERR_MSG(extack, "Invalid source address length");
goto out;
}
#ifndef CONFIG_IPV6_SUBTREES
if (cfg->fc_src_len) {
NL_SET_ERR_MSG(extack,
"Specifying source address requires IPV6_SUBTREES to be enabled");
goto out;
}
#endif
if (cfg->fc_ifindex) {
err = -ENODEV;
dev = dev_get_by_index(net, cfg->fc_ifindex);
if (!dev)
goto out;
idev = in6_dev_get(dev);
if (!idev)
goto out;
}
if (cfg->fc_metric == 0)
cfg->fc_metric = IP6_RT_PRIO_USER;
if (cfg->fc_flags & RTNH_F_ONLINK) {
if (!dev) {
NL_SET_ERR_MSG(extack,
"Nexthop device required for onlink");
err = -ENODEV;
goto out;
}
if (!(dev->flags & IFF_UP)) {
NL_SET_ERR_MSG(extack, "Nexthop device is not up");
err = -ENETDOWN;
goto out;
}
}
err = -ENOBUFS;
if (cfg->fc_nlinfo.nlh &&
!(cfg->fc_nlinfo.nlh->nlmsg_flags & NLM_F_CREATE)) {
table = fib6_get_table(net, cfg->fc_table);
if (!table) {
pr_warn("NLM_F_CREATE should be specified when creating new route\n");
table = fib6_new_table(net, cfg->fc_table);
}
} else {
table = fib6_new_table(net, cfg->fc_table);
}
if (!table)
goto out;
err = -ENOMEM;
rt = fib6_info_alloc(gfp_flags);
if (!rt)
goto out;
rt->fib6_metrics = ip_fib_metrics_init(net, cfg->fc_mx, cfg->fc_mx_len,
extack);
if (IS_ERR(rt->fib6_metrics)) {
err = PTR_ERR(rt->fib6_metrics);
/* Do not leave garbage there. */
rt->fib6_metrics = (struct dst_metrics *)&dst_default_metrics;
goto out;
}
if (cfg->fc_flags & RTF_ADDRCONF)
rt->dst_nocount = true;
if (cfg->fc_flags & RTF_EXPIRES)
fib6_set_expires(rt, jiffies +
clock_t_to_jiffies(cfg->fc_expires));
else
fib6_clean_expires(rt);
if (cfg->fc_protocol == RTPROT_UNSPEC)
cfg->fc_protocol = RTPROT_BOOT;
rt->fib6_protocol = cfg->fc_protocol;
addr_type = ipv6_addr_type(&cfg->fc_dst);
if (cfg->fc_encap) {
struct lwtunnel_state *lwtstate;
err = lwtunnel_build_state(cfg->fc_encap_type,
cfg->fc_encap, AF_INET6, cfg,
&lwtstate, extack);
if (err)
goto out;
rt->fib6_nh.nh_lwtstate = lwtstate_get(lwtstate);
}
ipv6_addr_prefix(&rt->fib6_dst.addr, &cfg->fc_dst, cfg->fc_dst_len);
rt->fib6_dst.plen = cfg->fc_dst_len;
if (rt->fib6_dst.plen == 128)
rt->dst_host = true;
ipv6: do not overwrite inetpeer metrics prematurely If an IPv6 host route with metrics exists, an attempt to add a new route for the same target with different metrics fails but rewrites the metrics anyway: 12sp0:~ # ip route add fec0::1 dev eth0 rto_min 1000 12sp0:~ # ip -6 route show fe80::/64 dev eth0 proto kernel metric 256 fec0::1 dev eth0 metric 1024 rto_min lock 1s 12sp0:~ # ip route add fec0::1 dev eth0 rto_min 1500 RTNETLINK answers: File exists 12sp0:~ # ip -6 route show fe80::/64 dev eth0 proto kernel metric 256 fec0::1 dev eth0 metric 1024 rto_min lock 1.5s This is caused by all IPv6 host routes using the metrics in their inetpeer (or the shared default). This also holds for the new route created in ip6_route_add() which shares the metrics with the already existing route and thus ip6_route_add() rewrites the metrics even if the new route ends up not being used at all. Another problem is that old metrics in inetpeer can reappear unexpectedly for a new route, e.g. 12sp0:~ # ip route add fec0::1 dev eth0 rto_min 1000 12sp0:~ # ip route del fec0::1 12sp0:~ # ip route add fec0::1 dev eth0 12sp0:~ # ip route change fec0::1 dev eth0 hoplimit 10 12sp0:~ # ip -6 route show fe80::/64 dev eth0 proto kernel metric 256 fec0::1 dev eth0 metric 1024 hoplimit 10 rto_min lock 1s Resolve the first problem by moving the setting of metrics down into fib6_add_rt2node() to the point we are sure we are inserting the new route into the tree. Second problem is addressed by introducing new flag DST_METRICS_FORCE_OVERWRITE which is set for a new host route in ip6_route_add() and makes ipv6_cow_metrics() always overwrite the metrics in inetpeer (even if they are not "new"); it is reset after that. v5: use a flag in _metrics member rather than one in flags v4: fix a typo making a condition always true (thanks to Hannes Frederic Sowa) v3: rewritten based on David Miller's idea to move setting the metrics (and allocation in non-host case) down to the point we already know the route is to be inserted. Also rebased to net-next as it is quite late in the cycle. Signed-off-by: Michal Kubecek <mkubecek@suse.cz> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-27 20:04:08 +08:00
#ifdef CONFIG_IPV6_SUBTREES
ipv6_addr_prefix(&rt->fib6_src.addr, &cfg->fc_src, cfg->fc_src_len);
rt->fib6_src.plen = cfg->fc_src_len;
#endif
rt->fib6_metric = cfg->fc_metric;
rt->fib6_nh.nh_weight = 1;
rt->fib6_type = cfg->fc_type;
/* We cannot add true routes via loopback here,
they would result in kernel looping; promote them to reject routes
*/
if ((cfg->fc_flags & RTF_REJECT) ||
(dev && (dev->flags & IFF_LOOPBACK) &&
!(addr_type & IPV6_ADDR_LOOPBACK) &&
!(cfg->fc_flags & RTF_LOCAL))) {
/* hold loopback dev/idev if we haven't done so. */
if (dev != net->loopback_dev) {
if (dev) {
dev_put(dev);
in6_dev_put(idev);
}
dev = net->loopback_dev;
dev_hold(dev);
idev = in6_dev_get(dev);
if (!idev) {
err = -ENODEV;
goto out;
}
}
rt->fib6_flags = RTF_REJECT|RTF_NONEXTHOP;
goto install_route;
}
if (cfg->fc_flags & RTF_GATEWAY) {
err = ip6_validate_gw(net, cfg, &dev, &idev, extack);
if (err)
goto out;
rt->fib6_nh.nh_gw = cfg->fc_gateway;
}
err = -ENODEV;
if (!dev)
goto out;
if (idev->cnf.disable_ipv6) {
NL_SET_ERR_MSG(extack, "IPv6 is disabled on nexthop device");
err = -EACCES;
goto out;
}
if (!(dev->flags & IFF_UP)) {
NL_SET_ERR_MSG(extack, "Nexthop device is not up");
err = -ENETDOWN;
goto out;
}
if (!ipv6_addr_any(&cfg->fc_prefsrc)) {
if (!ipv6_chk_addr(net, &cfg->fc_prefsrc, dev, 0)) {
NL_SET_ERR_MSG(extack, "Invalid source address");
err = -EINVAL;
goto out;
}
rt->fib6_prefsrc.addr = cfg->fc_prefsrc;
rt->fib6_prefsrc.plen = 128;
} else
rt->fib6_prefsrc.plen = 0;
rt->fib6_flags = cfg->fc_flags;
install_route:
if (!(rt->fib6_flags & (RTF_LOCAL | RTF_ANYCAST)) &&
!netif_carrier_ok(dev))
rt->fib6_nh.nh_flags |= RTNH_F_LINKDOWN;
rt->fib6_nh.nh_flags |= (cfg->fc_flags & RTNH_F_ONLINK);
rt->fib6_nh.nh_dev = dev;
rt->fib6_table = table;
if (idev)
in6_dev_put(idev);
return rt;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
out:
if (dev)
dev_put(dev);
if (idev)
in6_dev_put(idev);
fib6_info_release(rt);
return ERR_PTR(err);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
}
int ip6_route_add(struct fib6_config *cfg, gfp_t gfp_flags,
struct netlink_ext_ack *extack)
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
{
struct fib6_info *rt;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
int err;
rt = ip6_route_info_create(cfg, gfp_flags, extack);
if (IS_ERR(rt))
return PTR_ERR(rt);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
err = __ip6_ins_rt(rt, &cfg->fc_nlinfo, extack);
fib6_info_release(rt);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
return err;
}
static int __ip6_del_rt(struct fib6_info *rt, struct nl_info *info)
{
struct net *net = info->nl_net;
struct fib6_table *table;
int err;
if (rt == net->ipv6.fib6_null_entry) {
err = -ENOENT;
goto out;
}
table = rt->fib6_table;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
spin_lock_bh(&table->tb6_lock);
err = fib6_del(rt, info);
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
spin_unlock_bh(&table->tb6_lock);
out:
fib6_info_release(rt);
return err;
}
int ip6_del_rt(struct net *net, struct fib6_info *rt)
{
struct nl_info info = { .nl_net = net };
return __ip6_del_rt(rt, &info);
}
static int __ip6_del_rt_siblings(struct fib6_info *rt, struct fib6_config *cfg)
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
{
struct nl_info *info = &cfg->fc_nlinfo;
struct net *net = info->nl_net;
struct sk_buff *skb = NULL;
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
struct fib6_table *table;
int err = -ENOENT;
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
if (rt == net->ipv6.fib6_null_entry)
goto out_put;
table = rt->fib6_table;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
spin_lock_bh(&table->tb6_lock);
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
if (rt->fib6_nsiblings && cfg->fc_delete_all_nh) {
struct fib6_info *sibling, *next_sibling;
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
/* prefer to send a single notification with all hops */
skb = nlmsg_new(rt6_nlmsg_size(rt), gfp_any());
if (skb) {
u32 seq = info->nlh ? info->nlh->nlmsg_seq : 0;
if (rt6_fill_node(net, skb, rt, NULL,
NULL, NULL, 0, RTM_DELROUTE,
info->portid, seq, 0) < 0) {
kfree_skb(skb);
skb = NULL;
} else
info->skip_notify = 1;
}
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
list_for_each_entry_safe(sibling, next_sibling,
&rt->fib6_siblings,
fib6_siblings) {
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
err = fib6_del(sibling, info);
if (err)
goto out_unlock;
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
}
}
err = fib6_del(rt, info);
out_unlock:
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
spin_unlock_bh(&table->tb6_lock);
out_put:
fib6_info_release(rt);
if (skb) {
rtnl_notify(skb, net, info->portid, RTNLGRP_IPV6_ROUTE,
info->nlh, gfp_any());
}
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
return err;
}
static int ip6_del_cached_rt(struct rt6_info *rt, struct fib6_config *cfg)
{
int rc = -ESRCH;
if (cfg->fc_ifindex && rt->dst.dev->ifindex != cfg->fc_ifindex)
goto out;
if (cfg->fc_flags & RTF_GATEWAY &&
!ipv6_addr_equal(&cfg->fc_gateway, &rt->rt6i_gateway))
goto out;
rc = rt6_remove_exception_rt(rt);
out:
return rc;
}
static int ip6_route_del(struct fib6_config *cfg,
struct netlink_ext_ack *extack)
{
struct rt6_info *rt_cache;
struct fib6_table *table;
struct fib6_info *rt;
struct fib6_node *fn;
int err = -ESRCH;
table = fib6_get_table(cfg->fc_nlinfo.nl_net, cfg->fc_table);
if (!table) {
NL_SET_ERR_MSG(extack, "FIB table does not exist");
return err;
}
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_lock();
fn = fib6_locate(&table->tb6_root,
&cfg->fc_dst, cfg->fc_dst_len,
&cfg->fc_src, cfg->fc_src_len,
!(cfg->fc_flags & RTF_CACHE));
if (fn) {
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
for_each_fib6_node_rt_rcu(fn) {
if (cfg->fc_flags & RTF_CACHE) {
int rc;
rt_cache = rt6_find_cached_rt(rt, &cfg->fc_dst,
&cfg->fc_src);
if (rt_cache) {
rc = ip6_del_cached_rt(rt_cache, cfg);
net/ipv6: fix lock imbalance in ip6_route_del() WARNING: lock held when returning to user space! 4.17.0-rc3+ #37 Not tainted syz-executor1/27662 is leaving the kernel with locks still held! 1 lock held by syz-executor1/27662: #0: 00000000f661aee7 (rcu_read_lock){....}, at: ip6_route_del+0xea/0x13f0 net/ipv6/route.c:3206 BUG: scheduling while atomic: syz-executor1/27662/0x00000002 INFO: lockdep is turned off. Modules linked in: Kernel panic - not syncing: scheduling while atomic CPU: 1 PID: 27662 Comm: syz-executor1 Not tainted 4.17.0-rc3+ #37 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1b9/0x294 lib/dump_stack.c:113 panic+0x22f/0x4de kernel/panic.c:184 __schedule_bug.cold.85+0xdf/0xdf kernel/sched/core.c:3290 schedule_debug kernel/sched/core.c:3307 [inline] __schedule+0x139e/0x1e30 kernel/sched/core.c:3412 schedule+0xef/0x430 kernel/sched/core.c:3549 exit_to_usermode_loop+0x220/0x310 arch/x86/entry/common.c:152 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline] syscall_return_slowpath arch/x86/entry/common.c:265 [inline] do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x455979 RSP: 002b:00007fbf4051dc68 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 RAX: 0000000000000000 RBX: 00007fbf4051e6d4 RCX: 0000000000455979 RDX: 00000000200001c0 RSI: 000000000000890c RDI: 0000000000000013 RBP: 000000000072bea0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff R13: 00000000000003c8 R14: 00000000006f9b60 R15: 0000000000000000 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 23fb93a4d3f1 ("net/ipv6: Cleanup exception and cache route handling") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: David Ahern <dsahern@gmail.com> Reported-by: syzbot <syzkaller@googlegroups.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-10 01:05:46 +08:00
if (rc != -ESRCH) {
rcu_read_unlock();
return rc;
net/ipv6: fix lock imbalance in ip6_route_del() WARNING: lock held when returning to user space! 4.17.0-rc3+ #37 Not tainted syz-executor1/27662 is leaving the kernel with locks still held! 1 lock held by syz-executor1/27662: #0: 00000000f661aee7 (rcu_read_lock){....}, at: ip6_route_del+0xea/0x13f0 net/ipv6/route.c:3206 BUG: scheduling while atomic: syz-executor1/27662/0x00000002 INFO: lockdep is turned off. Modules linked in: Kernel panic - not syncing: scheduling while atomic CPU: 1 PID: 27662 Comm: syz-executor1 Not tainted 4.17.0-rc3+ #37 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1b9/0x294 lib/dump_stack.c:113 panic+0x22f/0x4de kernel/panic.c:184 __schedule_bug.cold.85+0xdf/0xdf kernel/sched/core.c:3290 schedule_debug kernel/sched/core.c:3307 [inline] __schedule+0x139e/0x1e30 kernel/sched/core.c:3412 schedule+0xef/0x430 kernel/sched/core.c:3549 exit_to_usermode_loop+0x220/0x310 arch/x86/entry/common.c:152 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline] syscall_return_slowpath arch/x86/entry/common.c:265 [inline] do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x455979 RSP: 002b:00007fbf4051dc68 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 RAX: 0000000000000000 RBX: 00007fbf4051e6d4 RCX: 0000000000455979 RDX: 00000000200001c0 RSI: 000000000000890c RDI: 0000000000000013 RBP: 000000000072bea0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff R13: 00000000000003c8 R14: 00000000006f9b60 R15: 0000000000000000 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 23fb93a4d3f1 ("net/ipv6: Cleanup exception and cache route handling") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: David Ahern <dsahern@gmail.com> Reported-by: syzbot <syzkaller@googlegroups.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-10 01:05:46 +08:00
}
}
continue;
}
if (cfg->fc_ifindex &&
(!rt->fib6_nh.nh_dev ||
rt->fib6_nh.nh_dev->ifindex != cfg->fc_ifindex))
continue;
if (cfg->fc_flags & RTF_GATEWAY &&
!ipv6_addr_equal(&cfg->fc_gateway, &rt->fib6_nh.nh_gw))
continue;
if (cfg->fc_metric && cfg->fc_metric != rt->fib6_metric)
continue;
if (cfg->fc_protocol && cfg->fc_protocol != rt->fib6_protocol)
continue;
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!fib6_info_hold_safe(rt))
continue;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
/* if gateway was specified only delete the one hop */
if (cfg->fc_flags & RTF_GATEWAY)
return __ip6_del_rt(rt, &cfg->fc_nlinfo);
return __ip6_del_rt_siblings(rt, cfg);
}
}
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
return err;
}
static void rt6_do_redirect(struct dst_entry *dst, struct sock *sk, struct sk_buff *skb)
{
struct netevent_redirect netevent;
struct rt6_info *rt, *nrt = NULL;
struct ndisc_options ndopts;
struct inet6_dev *in6_dev;
struct neighbour *neigh;
struct fib6_info *from;
struct rd_msg *msg;
int optlen, on_link;
u8 *lladdr;
optlen = skb_tail_pointer(skb) - skb_transport_header(skb);
optlen -= sizeof(*msg);
if (optlen < 0) {
net_dbg_ratelimited("rt6_do_redirect: packet too short\n");
return;
}
msg = (struct rd_msg *)icmp6_hdr(skb);
if (ipv6_addr_is_multicast(&msg->dest)) {
net_dbg_ratelimited("rt6_do_redirect: destination address is multicast\n");
return;
}
on_link = 0;
if (ipv6_addr_equal(&msg->dest, &msg->target)) {
on_link = 1;
} else if (ipv6_addr_type(&msg->target) !=
(IPV6_ADDR_UNICAST|IPV6_ADDR_LINKLOCAL)) {
net_dbg_ratelimited("rt6_do_redirect: target address is not link-local unicast\n");
return;
}
in6_dev = __in6_dev_get(skb->dev);
if (!in6_dev)
return;
if (in6_dev->cnf.forwarding || !in6_dev->cnf.accept_redirects)
return;
/* RFC2461 8.1:
* The IP source address of the Redirect MUST be the same as the current
* first-hop router for the specified ICMP Destination Address.
*/
if (!ndisc_parse_options(skb->dev, msg->opt, optlen, &ndopts)) {
net_dbg_ratelimited("rt6_redirect: invalid ND options\n");
return;
}
lladdr = NULL;
if (ndopts.nd_opts_tgt_lladdr) {
lladdr = ndisc_opt_addr_data(ndopts.nd_opts_tgt_lladdr,
skb->dev);
if (!lladdr) {
net_dbg_ratelimited("rt6_redirect: invalid link-layer address length\n");
return;
}
}
rt = (struct rt6_info *) dst;
if (rt->rt6i_flags & RTF_REJECT) {
net_dbg_ratelimited("rt6_redirect: source isn't a valid nexthop for redirect target\n");
return;
}
/* Redirect received -> path was valid.
* Look, redirects are sent only in response to data packets,
* so that this nexthop apparently is reachable. --ANK
*/
dst_confirm_neigh(&rt->dst, &ipv6_hdr(skb)->saddr);
neigh = __neigh_lookup(&nd_tbl, &msg->target, skb->dev, 1);
if (!neigh)
return;
/*
* We have finally decided to accept it.
*/
ndisc_update(skb->dev, neigh, lladdr, NUD_STALE,
NEIGH_UPDATE_F_WEAK_OVERRIDE|
NEIGH_UPDATE_F_OVERRIDE|
(on_link ? 0 : (NEIGH_UPDATE_F_OVERRIDE_ISROUTER|
NEIGH_UPDATE_F_ISROUTER)),
NDISC_REDIRECT, &ndopts);
rcu_read_lock();
from = rcu_dereference(rt->from);
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
/* This fib6_info_hold() is safe here because we hold reference to rt
* and rt already holds reference to fib6_info.
*/
fib6_info_hold(from);
rcu_read_unlock();
nrt = ip6_rt_cache_alloc(from, &msg->dest, NULL);
if (!nrt)
goto out;
nrt->rt6i_flags = RTF_GATEWAY|RTF_UP|RTF_DYNAMIC|RTF_CACHE;
if (on_link)
nrt->rt6i_flags &= ~RTF_GATEWAY;
nrt->rt6i_gateway = *(struct in6_addr *)neigh->primary_key;
/* No need to remove rt from the exception table if rt is
* a cached route because rt6_insert_exception() will
* takes care of it
*/
if (rt6_insert_exception(nrt, from)) {
dst_release_immediate(&nrt->dst);
goto out;
}
netevent.old = &rt->dst;
netevent.new = &nrt->dst;
netevent.daddr = &msg->dest;
netevent.neigh = neigh;
call_netevent_notifiers(NETEVENT_REDIRECT, &netevent);
out:
fib6_info_release(from);
neigh_release(neigh);
}
#ifdef CONFIG_IPV6_ROUTE_INFO
static struct fib6_info *rt6_get_route_info(struct net *net,
const struct in6_addr *prefix, int prefixlen,
const struct in6_addr *gwaddr,
struct net_device *dev)
{
u32 tb_id = l3mdev_fib_table(dev) ? : RT6_TABLE_INFO;
int ifindex = dev->ifindex;
struct fib6_node *fn;
struct fib6_info *rt = NULL;
struct fib6_table *table;
table = fib6_get_table(net, tb_id);
if (!table)
return NULL;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_lock();
fn = fib6_locate(&table->tb6_root, prefix, prefixlen, NULL, 0, true);
if (!fn)
goto out;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
for_each_fib6_node_rt_rcu(fn) {
if (rt->fib6_nh.nh_dev->ifindex != ifindex)
continue;
if ((rt->fib6_flags & (RTF_ROUTEINFO|RTF_GATEWAY)) != (RTF_ROUTEINFO|RTF_GATEWAY))
continue;
if (!ipv6_addr_equal(&rt->fib6_nh.nh_gw, gwaddr))
continue;
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (!fib6_info_hold_safe(rt))
continue;
break;
}
out:
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
return rt;
}
static struct fib6_info *rt6_add_route_info(struct net *net,
const struct in6_addr *prefix, int prefixlen,
const struct in6_addr *gwaddr,
struct net_device *dev,
unsigned int pref)
{
struct fib6_config cfg = {
.fc_metric = IP6_RT_PRIO_USER,
.fc_ifindex = dev->ifindex,
.fc_dst_len = prefixlen,
.fc_flags = RTF_GATEWAY | RTF_ADDRCONF | RTF_ROUTEINFO |
RTF_UP | RTF_PREF(pref),
.fc_protocol = RTPROT_RA,
.fc_type = RTN_UNICAST,
.fc_nlinfo.portid = 0,
.fc_nlinfo.nlh = NULL,
.fc_nlinfo.nl_net = net,
};
cfg.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_INFO,
cfg.fc_dst = *prefix;
cfg.fc_gateway = *gwaddr;
/* We should treat it as a default route if prefix length is 0. */
if (!prefixlen)
cfg.fc_flags |= RTF_DEFAULT;
ip6_route_add(&cfg, GFP_ATOMIC, NULL);
return rt6_get_route_info(net, prefix, prefixlen, gwaddr, dev);
}
#endif
struct fib6_info *rt6_get_dflt_router(struct net *net,
const struct in6_addr *addr,
struct net_device *dev)
{
u32 tb_id = l3mdev_fib_table(dev) ? : RT6_TABLE_DFLT;
struct fib6_info *rt;
struct fib6_table *table;
table = fib6_get_table(net, tb_id);
if (!table)
return NULL;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_lock();
for_each_fib6_node_rt_rcu(&table->tb6_root) {
if (dev == rt->fib6_nh.nh_dev &&
((rt->fib6_flags & (RTF_ADDRCONF | RTF_DEFAULT)) == (RTF_ADDRCONF | RTF_DEFAULT)) &&
ipv6_addr_equal(&rt->fib6_nh.nh_gw, addr))
break;
}
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
if (rt && !fib6_info_hold_safe(rt))
rt = NULL;
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
return rt;
}
struct fib6_info *rt6_add_dflt_router(struct net *net,
const struct in6_addr *gwaddr,
struct net_device *dev,
unsigned int pref)
{
struct fib6_config cfg = {
.fc_table = l3mdev_fib_table(dev) ? : RT6_TABLE_DFLT,
.fc_metric = IP6_RT_PRIO_USER,
.fc_ifindex = dev->ifindex,
.fc_flags = RTF_GATEWAY | RTF_ADDRCONF | RTF_DEFAULT |
RTF_UP | RTF_EXPIRES | RTF_PREF(pref),
.fc_protocol = RTPROT_RA,
.fc_type = RTN_UNICAST,
.fc_nlinfo.portid = 0,
.fc_nlinfo.nlh = NULL,
.fc_nlinfo.nl_net = net,
};
cfg.fc_gateway = *gwaddr;
if (!ip6_route_add(&cfg, GFP_ATOMIC, NULL)) {
struct fib6_table *table;
table = fib6_get_table(dev_net(dev), cfg.fc_table);
if (table)
table->flags |= RT6_TABLE_HAS_DFLT_ROUTER;
}
return rt6_get_dflt_router(net, gwaddr, dev);
}
static void __rt6_purge_dflt_routers(struct net *net,
struct fib6_table *table)
{
struct fib6_info *rt;
restart:
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_lock();
for_each_fib6_node_rt_rcu(&table->tb6_root) {
struct net_device *dev = fib6_info_nh_dev(rt);
struct inet6_dev *idev = dev ? __in6_dev_get(dev) : NULL;
if (rt->fib6_flags & (RTF_DEFAULT | RTF_ADDRCONF) &&
ipv6: use fib6_info_hold_safe() when necessary In the code path where only rcu read lock is held, e.g. in the route lookup code path, it is not safe to directly call fib6_info_hold() because the fib6_info may already have been deleted but still exists in the rcu grace period. Holding reference to it could cause double free and crash the kernel. This patch adds a new function fib6_info_hold_safe() and replace fib6_info_hold() in all necessary places. Syzbot reported 3 crash traces because of this. One of them is: 8021q: adding VLAN 0 to HW filter on device team0 IPv6: ADDRCONF(NETDEV_CHANGE): team0: link becomes ready dst_release: dst:(____ptrval____) refcnt:-1 dst_release: dst:(____ptrval____) refcnt:-2 WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 dst_hold include/net/dst.h:239 [inline] WARNING: CPU: 1 PID: 4845 at include/net/dst.h:239 ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 dst_release: dst:(____ptrval____) refcnt:-1 Kernel panic - not syncing: panic_on_warn set ... CPU: 1 PID: 4845 Comm: syz-executor493 Not tainted 4.18.0-rc3+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x2b4 lib/dump_stack.c:113 panic+0x238/0x4e7 kernel/panic.c:184 dst_release: dst:(____ptrval____) refcnt:-2 dst_release: dst:(____ptrval____) refcnt:-3 __warn.cold.8+0x163/0x1ba kernel/panic.c:536 dst_release: dst:(____ptrval____) refcnt:-4 report_bug+0x252/0x2d0 lib/bug.c:186 fixup_bug arch/x86/kernel/traps.c:178 [inline] do_error_trap+0x1fc/0x4d0 arch/x86/kernel/traps.c:296 dst_release: dst:(____ptrval____) refcnt:-5 do_invalid_op+0x1b/0x20 arch/x86/kernel/traps.c:316 invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:992 RIP: 0010:dst_hold include/net/dst.h:239 [inline] RIP: 0010:ip6_setup_cork+0xd66/0x1830 net/ipv6/ip6_output.c:1204 Code: c1 ed 03 89 9d 18 ff ff ff 48 b8 00 00 00 00 00 fc ff df 41 c6 44 05 00 f8 e9 2d 01 00 00 4c 8b a5 c8 fe ff ff e8 1a f6 e6 fa <0f> 0b e9 6a fc ff ff e8 0e f6 e6 fa 48 8b 85 d0 fe ff ff 48 8d 78 RSP: 0018:ffff8801a8fcf178 EFLAGS: 00010293 RAX: ffff8801a8eba5c0 RBX: 0000000000000000 RCX: ffffffff869511e6 RDX: 0000000000000000 RSI: ffffffff869515b6 RDI: 0000000000000005 RBP: ffff8801a8fcf2c8 R08: ffff8801a8eba5c0 R09: ffffed0035ac8338 R10: ffffed0035ac8338 R11: ffff8801ad6419c3 R12: ffff8801a8fcf720 R13: ffff8801a8fcf6a0 R14: ffff8801ad6419c0 R15: ffff8801ad641980 ip6_make_skb+0x2c8/0x600 net/ipv6/ip6_output.c:1768 udpv6_sendmsg+0x2c90/0x35f0 net/ipv6/udp.c:1376 inet_sendmsg+0x1a1/0x690 net/ipv4/af_inet.c:798 sock_sendmsg_nosec net/socket.c:641 [inline] sock_sendmsg+0xd5/0x120 net/socket.c:651 ___sys_sendmsg+0x51d/0x930 net/socket.c:2125 __sys_sendmmsg+0x240/0x6f0 net/socket.c:2220 __do_sys_sendmmsg net/socket.c:2249 [inline] __se_sys_sendmmsg net/socket.c:2246 [inline] __x64_sys_sendmmsg+0x9d/0x100 net/socket.c:2246 do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x49/0xbe RIP: 0033:0x446ba9 Code: e8 cc bb 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 eb 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007fb39a469da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000006dcc54 RCX: 0000000000446ba9 RDX: 00000000000000b8 RSI: 0000000020001b00 RDI: 0000000000000003 RBP: 00000000006dcc50 R08: 00007fb39a46a700 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 45c828efc7a64843 R13: e6eeb815b9d8a477 R14: 5068caf6f713c6fc R15: 0000000000000001 Dumping ftrace buffer: (ftrace buffer empty) Kernel Offset: disabled Rebooting in 86400 seconds.. Fixes: 93531c674315 ("net/ipv6: separate handling of FIB entries from dst based routes") Reported-by: syzbot+902e2a1bcd4f7808cef5@syzkaller.appspotmail.com Reported-by: syzbot+8ae62d67f647abeeceb9@syzkaller.appspotmail.com Reported-by: syzbot+3f08feb14086930677d0@syzkaller.appspotmail.com Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-22 11:56:32 +08:00
(!idev || idev->cnf.accept_ra != 2) &&
fib6_info_hold_safe(rt)) {
rcu_read_unlock();
ip6_del_rt(net, rt);
goto restart;
}
}
ipv6: replace rwlock with rcu and spinlock in fib6_table With all the preparation work before, we are now ready to replace rwlock with rcu and spinlock in fib6_table. That means now all fib6_node in fib6_table are protected by rcu. And when freeing fib6_node, call_rcu() is used to wait for the rcu grace period before releasing the memory. When accessing fib6_node, corresponding rcu APIs need to be used. And all previous sessions protected by the write lock will now be protected by the spin lock per table. All previous sessions protected by read lock will now be protected by rcu_read_lock(). A couple of things to note here: 1. As part of the work of replacing rwlock with rcu, the linked list of fn->leaf now has to be rcu protected as well. So both fn->leaf and rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are used when manipulating them. 2. For fn->rr_ptr, first of all, it also needs to be rcu protected now and is tagged with __rcu and rcu APIs are used in corresponding places. Secondly, fn->rr_ptr is changed in rt6_select() which is a reader thread. This makes the issue a bit complicated. We think a valid solution for it is to let rt6_select() grab the tb6_lock if it decides to change it. As it is not in the normal operation and only happens when there is no valid neighbor cache for the route, we think the performance impact should be low. 3. fib6_walk_continue() has to be called with tb6_lock held even in the route dumping related functions, e.g. inet6_dump_fib(), fib6_tables_dump() and ipv6_route_seq_ops. It is because fib6_walk_continue() makes modifications to the walker structure, and so are fib6_repair_tree() and fib6_del_route(). In order to do proper syncing between them, we need to let fib6_walk_continue() hold the lock. We may be able to do further improvement on the way we do the tree walk to get rid of the need for holding the spin lock. But not for now. 4. When fib6_del_route() removes a route from the tree, we no longer mark rt->dst.rt6_next to NULL to make simultaneous reader be able to further traverse the list with rcu. However, rt->dst.rt6_next is only valid within this same rcu period. No one should access it later. 5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be performed before we publish this route (either by linking it to fn->leaf or insert it in the list pointed by fn->leaf) just to be safe because as soon as we publish the route, some read thread will be able to access it. Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-07 03:06:10 +08:00
rcu_read_unlock();
table->flags &= ~RT6_TABLE_HAS_DFLT_ROUTER;
}
void rt6_purge_dflt_routers(struct net *net)
{
struct fib6_table *table;
struct hlist_head *head;
unsigned int h;
rcu_read_lock();
for (h = 0; h < FIB6_TABLE_HASHSZ; h++) {
head = &net->ipv6.fib_table_hash[h];
hlist_for_each_entry_rcu(table, head, tb6_hlist) {
if (table->flags & RT6_TABLE_HAS_DFLT_ROUTER)
__rt6_purge_dflt_routers(net, table);
}
}
rcu_read_unlock();
}
static void rtmsg_to_fib6_config(struct net *net,
struct in6_rtmsg *rtmsg,
struct fib6_config *cfg)
{
*cfg = (struct fib6_config){
.fc_table = l3mdev_fib_table_by_index(net, rtmsg->rtmsg_ifindex) ?
: RT6_TABLE_MAIN,
.fc_ifindex = rtmsg->rtmsg_ifindex,
.fc_metric = rtmsg->rtmsg_metric,
.fc_expires = rtmsg->rtmsg_info,
.fc_dst_len = rtmsg->rtmsg_dst_len,
.fc_src_len = rtmsg->rtmsg_src_len,
.fc_flags = rtmsg->rtmsg_flags,
.fc_type = rtmsg->rtmsg_type,
.fc_nlinfo.nl_net = net,
.fc_dst = rtmsg->rtmsg_dst,
.fc_src = rtmsg->rtmsg_src,
.fc_gateway = rtmsg->rtmsg_gateway,
};
}
int ipv6_route_ioctl(struct net *net, unsigned int cmd, void __user *arg)
{
struct fib6_config cfg;
struct in6_rtmsg rtmsg;
int err;
switch (cmd) {
case SIOCADDRT: /* Add a route */
case SIOCDELRT: /* Delete a route */
net: Allow userns root to control ipv6 Allow an unpriviled user who has created a user namespace, and then created a network namespace to effectively use the new network namespace, by reducing capable(CAP_NET_ADMIN) and capable(CAP_NET_RAW) calls to be ns_capable(net->user_ns, CAP_NET_ADMIN), or capable(net->user_ns, CAP_NET_RAW) calls. Settings that merely control a single network device are allowed. Either the network device is a logical network device where restrictions make no difference or the network device is hardware NIC that has been explicity moved from the initial network namespace. In general policy and network stack state changes are allowed while resource control is left unchanged. Allow the SIOCSIFADDR ioctl to add ipv6 addresses. Allow the SIOCDIFADDR ioctl to delete ipv6 addresses. Allow the SIOCADDRT ioctl to add ipv6 routes. Allow the SIOCDELRT ioctl to delete ipv6 routes. Allow creation of ipv6 raw sockets. Allow setting the IPV6_JOIN_ANYCAST socket option. Allow setting the IPV6_FL_A_RENEW parameter of the IPV6_FLOWLABEL_MGR socket option. Allow setting the IPV6_TRANSPARENT socket option. Allow setting the IPV6_HOPOPTS socket option. Allow setting the IPV6_RTHDRDSTOPTS socket option. Allow setting the IPV6_DSTOPTS socket option. Allow setting the IPV6_IPSEC_POLICY socket option. Allow setting the IPV6_XFRM_POLICY socket option. Allow sending packets with the IPV6_2292HOPOPTS control message. Allow sending packets with the IPV6_2292DSTOPTS control message. Allow sending packets with the IPV6_RTHDRDSTOPTS control message. Allow setting the multicast routing socket options on non multicast routing sockets. Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL, and SIOCDELTUNNEL ioctls for setting up, changing and deleting tunnels over ipv6. Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL, SIOCDELTUNNEL ioctls for setting up, changing and deleting ipv6 over ipv4 tunnels. Allow the SIOCADDPRL, SIOCDELPRL, SIOCCHGPRL ioctls for adding, deleting, and changing the potential router list for ISATAP tunnels. Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-11-16 11:03:06 +08:00
if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
return -EPERM;
err = copy_from_user(&rtmsg, arg,
sizeof(struct in6_rtmsg));
if (err)
return -EFAULT;
rtmsg_to_fib6_config(net, &rtmsg, &cfg);
rtnl_lock();
switch (cmd) {
case SIOCADDRT:
err = ip6_route_add(&cfg, GFP_KERNEL, NULL);
break;
case SIOCDELRT:
err = ip6_route_del(&cfg, NULL);
break;
default:
err = -EINVAL;
}
rtnl_unlock();
return err;
}
return -EINVAL;
}
/*
* Drop the packet on the floor
*/
static int ip6_pkt_drop(struct sk_buff *skb, u8 code, int ipstats_mib_noroutes)
{
int type;
struct dst_entry *dst = skb_dst(skb);
switch (ipstats_mib_noroutes) {
case IPSTATS_MIB_INNOROUTES:
type = ipv6_addr_type(&ipv6_hdr(skb)->daddr);
if (type == IPV6_ADDR_ANY) {
IP6_INC_STATS(dev_net(dst->dev),
__in6_dev_get_safely(skb->dev),
IPSTATS_MIB_INADDRERRORS);
break;
}
/* FALLTHROUGH */
case IPSTATS_MIB_OUTNOROUTES:
IP6_INC_STATS(dev_net(dst->dev), ip6_dst_idev(dst),
ipstats_mib_noroutes);
break;
}
icmpv6_send(skb, ICMPV6_DEST_UNREACH, code, 0);
kfree_skb(skb);
return 0;
}
static int ip6_pkt_discard(struct sk_buff *skb)
{
return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_INNOROUTES);
}
static int ip6_pkt_discard_out(struct net *net, struct sock *sk, struct sk_buff *skb)
{
skb->dev = skb_dst(skb)->dev;
return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_OUTNOROUTES);
}
static int ip6_pkt_prohibit(struct sk_buff *skb)
{
return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_INNOROUTES);
}
static int ip6_pkt_prohibit_out(struct net *net, struct sock *sk, struct sk_buff *skb)
{
skb->dev = skb_dst(skb)->dev;
return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_OUTNOROUTES);
}
/*
* Allocate a dst for local (unicast / anycast) address.
*/
struct fib6_info *addrconf_f6i_alloc(struct net *net,
struct inet6_dev *idev,
const struct in6_addr *addr,
bool anycast, gfp_t gfp_flags)
{
u32 tb_id;
struct net_device *dev = idev->dev;
struct fib6_info *f6i;
f6i = fib6_info_alloc(gfp_flags);
if (!f6i)
return ERR_PTR(-ENOMEM);
f6i->fib6_metrics = ip_fib_metrics_init(net, NULL, 0, NULL);
f6i->dst_nocount = true;
f6i->dst_host = true;
f6i->fib6_protocol = RTPROT_KERNEL;
f6i->fib6_flags = RTF_UP | RTF_NONEXTHOP;
if (anycast) {
f6i->fib6_type = RTN_ANYCAST;
f6i->fib6_flags |= RTF_ANYCAST;
} else {
f6i->fib6_type = RTN_LOCAL;
f6i->fib6_flags |= RTF_LOCAL;
}
f6i->fib6_nh.nh_gw = *addr;
dev_hold(dev);
f6i->fib6_nh.nh_dev = dev;
f6i->fib6_dst.addr = *addr;
f6i->fib6_dst.plen = 128;
tb_id = l3mdev_fib_table(idev->dev) ? : RT6_TABLE_LOCAL;
f6i->fib6_table = fib6_get_table(net, tb_id);
return f6i;
}
/* remove deleted ip from prefsrc entries */
struct arg_dev_net_ip {
struct net_device *dev;
struct net *net;
struct in6_addr *addr;
};
static int fib6_remove_prefsrc(struct fib6_info *rt, void *arg)
{
struct net_device *dev = ((struct arg_dev_net_ip *)arg)->dev;
struct net *net = ((struct arg_dev_net_ip *)arg)->net;
struct in6_addr *addr = ((struct arg_dev_net_ip *)arg)->addr;
if (((void *)rt->fib6_nh.nh_dev == dev || !dev) &&
rt != net->ipv6.fib6_null_entry &&
ipv6_addr_equal(addr, &rt->fib6_prefsrc.addr)) {
spin_lock_bh(&rt6_exception_lock);
/* remove prefsrc entry */
rt->fib6_prefsrc.plen = 0;
spin_unlock_bh(&rt6_exception_lock);
}
return 0;
}
void rt6_remove_prefsrc(struct inet6_ifaddr *ifp)
{
struct net *net = dev_net(ifp->idev->dev);
struct arg_dev_net_ip adni = {
.dev = ifp->idev->dev,
.net = net,
.addr = &ifp->addr,
};
fib6_clean_all(net, fib6_remove_prefsrc, &adni);
}
#define RTF_RA_ROUTER (RTF_ADDRCONF | RTF_DEFAULT | RTF_GATEWAY)
/* Remove routers and update dst entries when gateway turn into host. */
static int fib6_clean_tohost(struct fib6_info *rt, void *arg)
{
struct in6_addr *gateway = (struct in6_addr *)arg;
if (((rt->fib6_flags & RTF_RA_ROUTER) == RTF_RA_ROUTER) &&
ipv6_addr_equal(gateway, &rt->fib6_nh.nh_gw)) {
return -1;
}
/* Further clean up cached routes in exception table.
* This is needed because cached route may have a different
* gateway than its 'parent' in the case of an ip redirect.
*/
rt6_exceptions_clean_tohost(rt, gateway);
return 0;
}
void rt6_clean_tohost(struct net *net, struct in6_addr *gateway)
{
fib6_clean_all(net, fib6_clean_tohost, gateway);
}
struct arg_netdev_event {
const struct net_device *dev;
union {
unsigned int nh_flags;
unsigned long event;
};
};
static struct fib6_info *rt6_multipath_first_sibling(const struct fib6_info *rt)
{
struct fib6_info *iter;
struct fib6_node *fn;
fn = rcu_dereference_protected(rt->fib6_node,
lockdep_is_held(&rt->fib6_table->tb6_lock));
iter = rcu_dereference_protected(fn->leaf,
lockdep_is_held(&rt->fib6_table->tb6_lock));
while (iter) {
if (iter->fib6_metric == rt->fib6_metric &&
rt6_qualify_for_ecmp(iter))
return iter;
iter = rcu_dereference_protected(iter->fib6_next,
lockdep_is_held(&rt->fib6_table->tb6_lock));
}
return NULL;
}
static bool rt6_is_dead(const struct fib6_info *rt)
{
if (rt->fib6_nh.nh_flags & RTNH_F_DEAD ||
(rt->fib6_nh.nh_flags & RTNH_F_LINKDOWN &&
fib6_ignore_linkdown(rt)))
return true;
return false;
}
static int rt6_multipath_total_weight(const struct fib6_info *rt)
{
struct fib6_info *iter;
int total = 0;
if (!rt6_is_dead(rt))
total += rt->fib6_nh.nh_weight;
list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings) {
if (!rt6_is_dead(iter))
total += iter->fib6_nh.nh_weight;
}
return total;
}
static void rt6_upper_bound_set(struct fib6_info *rt, int *weight, int total)
{
int upper_bound = -1;
if (!rt6_is_dead(rt)) {
*weight += rt->fib6_nh.nh_weight;
upper_bound = DIV_ROUND_CLOSEST_ULL((u64) (*weight) << 31,
total) - 1;
}
atomic_set(&rt->fib6_nh.nh_upper_bound, upper_bound);
}
static void rt6_multipath_upper_bound_set(struct fib6_info *rt, int total)
{
struct fib6_info *iter;
int weight = 0;
rt6_upper_bound_set(rt, &weight, total);
list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings)
rt6_upper_bound_set(iter, &weight, total);
}
void rt6_multipath_rebalance(struct fib6_info *rt)
{
struct fib6_info *first;
int total;
/* In case the entire multipath route was marked for flushing,
* then there is no need to rebalance upon the removal of every
* sibling route.
*/
if (!rt->fib6_nsiblings || rt->should_flush)
return;
/* During lookup routes are evaluated in order, so we need to
* make sure upper bounds are assigned from the first sibling
* onwards.
*/
first = rt6_multipath_first_sibling(rt);
if (WARN_ON_ONCE(!first))
return;
total = rt6_multipath_total_weight(first);
rt6_multipath_upper_bound_set(first, total);
}
static int fib6_ifup(struct fib6_info *rt, void *p_arg)
{
const struct arg_netdev_event *arg = p_arg;
struct net *net = dev_net(arg->dev);
if (rt != net->ipv6.fib6_null_entry && rt->fib6_nh.nh_dev == arg->dev) {
rt->fib6_nh.nh_flags &= ~arg->nh_flags;
fib6_update_sernum_upto_root(net, rt);
rt6_multipath_rebalance(rt);
}
return 0;
}
void rt6_sync_up(struct net_device *dev, unsigned int nh_flags)
{
struct arg_netdev_event arg = {
.dev = dev,
{
.nh_flags = nh_flags,
},
};
if (nh_flags & RTNH_F_DEAD && netif_carrier_ok(dev))
arg.nh_flags |= RTNH_F_LINKDOWN;
fib6_clean_all(dev_net(dev), fib6_ifup, &arg);
}
static bool rt6_multipath_uses_dev(const struct fib6_info *rt,
const struct net_device *dev)
{
struct fib6_info *iter;
if (rt->fib6_nh.nh_dev == dev)
return true;
list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings)
if (iter->fib6_nh.nh_dev == dev)
return true;
return false;
}
static void rt6_multipath_flush(struct fib6_info *rt)
{
struct fib6_info *iter;
rt->should_flush = 1;
list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings)
iter->should_flush = 1;
}
static unsigned int rt6_multipath_dead_count(const struct fib6_info *rt,
const struct net_device *down_dev)
{
struct fib6_info *iter;
unsigned int dead = 0;
if (rt->fib6_nh.nh_dev == down_dev ||
rt->fib6_nh.nh_flags & RTNH_F_DEAD)
dead++;
list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings)
if (iter->fib6_nh.nh_dev == down_dev ||
iter->fib6_nh.nh_flags & RTNH_F_DEAD)
dead++;
return dead;
}
static void rt6_multipath_nh_flags_set(struct fib6_info *rt,
const struct net_device *dev,
unsigned int nh_flags)
{
struct fib6_info *iter;
if (rt->fib6_nh.nh_dev == dev)
rt->fib6_nh.nh_flags |= nh_flags;
list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings)
if (iter->fib6_nh.nh_dev == dev)
iter->fib6_nh.nh_flags |= nh_flags;
}
net: ipv6: Keep nexthop of multipath route on admin down IPv6 deletes route entries associated with multipath routes on an admin down where IPv4 does not. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 metric 64 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.100.2.254 dev eth2 weight 1 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.4 10.100.2.0/24 dev eth2 proto kernel scope link src 10.100.2.4 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2:: dev red proto none metric 0 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:11::/120 via 2001:db8:1::16 dev eth1 metric 1024 pref medium 2001:db8:11::/120 via 2001:db8:2::17 dev eth2 metric 1024 pref medium ... Set link down: $ ip li set eth1 down IPv4 retains the multihop route but flags eth1 route as dead: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.16 dev eth1 weight 1 dead linkdown nexthop via 10.100.2.16 dev eth2 weight 1 10.100.2.0/24 dev eth2 proto kernel scope link src 10.100.2.4 and IPv6 deletes the route as part of flushing all routes for the device: $ ip -6 ro ls vrf red 2001:db8:2:: dev red proto none metric 0 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:11::/120 via 2001:db8:2::17 dev eth2 metric 1024 pref medium ... Worse, on admin up of the device the multipath route has to be deleted to get this leg of the route re-added. This patch keeps routes that are part of a multipath route if ignore_routes_with_linkdown is set with the dead and linkdown flags enabling consistency between IPv4 and IPv6: $ ip -6 ro ls vrf red 2001:db8:2:: dev red proto none metric 0 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:11::/120 via 2001:db8:1::16 dev eth1 metric 1024 dead linkdown pref medium 2001:db8:11::/120 via 2001:db8:2::17 dev eth2 metric 1024 pref medium ... Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-18 23:40:36 +08:00
/* called with write lock held for table with rt */
static int fib6_ifdown(struct fib6_info *rt, void *p_arg)
{
const struct arg_netdev_event *arg = p_arg;
const struct net_device *dev = arg->dev;
struct net *net = dev_net(dev);
if (rt == net->ipv6.fib6_null_entry)
return 0;
switch (arg->event) {
case NETDEV_UNREGISTER:
return rt->fib6_nh.nh_dev == dev ? -1 : 0;
case NETDEV_DOWN:
if (rt->should_flush)
return -1;
if (!rt->fib6_nsiblings)
return rt->fib6_nh.nh_dev == dev ? -1 : 0;
if (rt6_multipath_uses_dev(rt, dev)) {
unsigned int count;
count = rt6_multipath_dead_count(rt, dev);
if (rt->fib6_nsiblings + 1 == count) {
rt6_multipath_flush(rt);
return -1;
}
rt6_multipath_nh_flags_set(rt, dev, RTNH_F_DEAD |
RTNH_F_LINKDOWN);
fib6_update_sernum(net, rt);
rt6_multipath_rebalance(rt);
}
return -2;
case NETDEV_CHANGE:
if (rt->fib6_nh.nh_dev != dev ||
rt->fib6_flags & (RTF_LOCAL | RTF_ANYCAST))
break;
rt->fib6_nh.nh_flags |= RTNH_F_LINKDOWN;
rt6_multipath_rebalance(rt);
break;
}
return 0;
}
void rt6_sync_down_dev(struct net_device *dev, unsigned long event)
{
struct arg_netdev_event arg = {
.dev = dev,
{
.event = event,
},
};
struct net *net = dev_net(dev);
if (net->ipv6.sysctl.skip_notify_on_dev_down)
fib6_clean_all_skip_notify(net, fib6_ifdown, &arg);
else
fib6_clean_all(net, fib6_ifdown, &arg);
}
void rt6_disable_ip(struct net_device *dev, unsigned long event)
{
rt6_sync_down_dev(dev, event);
rt6_uncached_list_flush_dev(dev_net(dev), dev);
neigh_ifdown(&nd_tbl, dev);
}
struct rt6_mtu_change_arg {
struct net_device *dev;
unsigned int mtu;
};
static int rt6_mtu_change_route(struct fib6_info *rt, void *p_arg)
{
struct rt6_mtu_change_arg *arg = (struct rt6_mtu_change_arg *) p_arg;
struct inet6_dev *idev;
/* In IPv6 pmtu discovery is not optional,
so that RTAX_MTU lock cannot disable it.
We still use this lock to block changes
caused by addrconf/ndisc.
*/
idev = __in6_dev_get(arg->dev);
if (!idev)
return 0;
/* For administrative MTU increase, there is no way to discover
IPv6 PMTU increase, so PMTU increase should be updated here.
Since RFC 1981 doesn't include administrative MTU increase
update PMTU increase is a MUST. (i.e. jumbo frame)
*/
if (rt->fib6_nh.nh_dev == arg->dev &&
!fib6_metric_locked(rt, RTAX_MTU)) {
u32 mtu = rt->fib6_pmtu;
if (mtu >= arg->mtu ||
(mtu < arg->mtu && mtu == idev->cnf.mtu6))
fib6_metric_set(rt, RTAX_MTU, arg->mtu);
spin_lock_bh(&rt6_exception_lock);
ipv6: Reflect MTU changes on PMTU of exceptions for MTU-less routes Currently, administrative MTU changes on a given netdevice are not reflected on route exceptions for MTU-less routes, with a set PMTU value, for that device: # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a proto kernel src 2001:db8::a metric 256 pref medium # ping6 -c 1 -q -s10000 2001:db8::b > /dev/null # ip netns exec a ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 3000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium # ip link set dev vti_a mtu 9000 # ip -6 route get 2001:db8::b 2001:db8::b from :: dev vti_a src 2001:db8::a metric 0 cache expires 571sec mtu 4926 pref medium The first issue is that since commit fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") we don't call rt6_exceptions_update_pmtu() from rt6_mtu_change_route(), which handles administrative MTU changes, if the regular route is MTU-less. However, PMTU exceptions should be always updated, as long as RTAX_MTU is not locked. Keep the check for MTU-less main route, as introduced by that commit, but, for exceptions, call rt6_exceptions_update_pmtu() regardless of that check. Once that is fixed, one problem remains: MTU changes are not reflected if the new MTU is higher than the previous one, because rt6_exceptions_update_pmtu() doesn't allow that. We should instead allow PMTU increase if the old PMTU matches the local MTU, as that implies that the old MTU was the lowest in the path, and PMTU discovery might lead to different results. The existing check in rt6_mtu_change_route() correctly took that case into account (for regular routes only), so factor it out and re-use it also in rt6_exceptions_update_pmtu(). While at it, fix comments style and grammar, and try to be a bit more descriptive. Reported-by: Xiumei Mu <xmu@redhat.com> Fixes: fb56be83e43d ("net-ipv6: on device mtu change do not add mtu to mtu-less routes") Fixes: f5bbe7ee79c2 ("ipv6: prepare rt6_mtu_change() for exception table") Signed-off-by: Stefano Brivio <sbrivio@redhat.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-06 18:10:19 +08:00
rt6_exceptions_update_pmtu(idev, rt, arg->mtu);
spin_unlock_bh(&rt6_exception_lock);
}
return 0;
}
void rt6_mtu_change(struct net_device *dev, unsigned int mtu)
{
struct rt6_mtu_change_arg arg = {
.dev = dev,
.mtu = mtu,
};
fib6_clean_all(dev_net(dev), rt6_mtu_change_route, &arg);
}
static const struct nla_policy rtm_ipv6_policy[RTA_MAX+1] = {
[RTA_GATEWAY] = { .len = sizeof(struct in6_addr) },
[RTA_PREFSRC] = { .len = sizeof(struct in6_addr) },
[RTA_OIF] = { .type = NLA_U32 },
[RTA_IIF] = { .type = NLA_U32 },
[RTA_PRIORITY] = { .type = NLA_U32 },
[RTA_METRICS] = { .type = NLA_NESTED },
[RTA_MULTIPATH] = { .len = sizeof(struct rtnexthop) },
[RTA_PREF] = { .type = NLA_U8 },
[RTA_ENCAP_TYPE] = { .type = NLA_U16 },
[RTA_ENCAP] = { .type = NLA_NESTED },
[RTA_EXPIRES] = { .type = NLA_U32 },
[RTA_UID] = { .type = NLA_U32 },
[RTA_MARK] = { .type = NLA_U32 },
[RTA_TABLE] = { .type = NLA_U32 },
[RTA_IP_PROTO] = { .type = NLA_U8 },
[RTA_SPORT] = { .type = NLA_U16 },
[RTA_DPORT] = { .type = NLA_U16 },
};
static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
struct fib6_config *cfg,
struct netlink_ext_ack *extack)
{
struct rtmsg *rtm;
struct nlattr *tb[RTA_MAX+1];
unsigned int pref;
int err;
err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv6_policy,
extack);
if (err < 0)
goto errout;
err = -EINVAL;
rtm = nlmsg_data(nlh);
*cfg = (struct fib6_config){
.fc_table = rtm->rtm_table,
.fc_dst_len = rtm->rtm_dst_len,
.fc_src_len = rtm->rtm_src_len,
.fc_flags = RTF_UP,
.fc_protocol = rtm->rtm_protocol,
.fc_type = rtm->rtm_type,
.fc_nlinfo.portid = NETLINK_CB(skb).portid,
.fc_nlinfo.nlh = nlh,
.fc_nlinfo.nl_net = sock_net(skb->sk),
};
if (rtm->rtm_type == RTN_UNREACHABLE ||
rtm->rtm_type == RTN_BLACKHOLE ||
rtm->rtm_type == RTN_PROHIBIT ||
rtm->rtm_type == RTN_THROW)
cfg->fc_flags |= RTF_REJECT;
if (rtm->rtm_type == RTN_LOCAL)
cfg->fc_flags |= RTF_LOCAL;
if (rtm->rtm_flags & RTM_F_CLONED)
cfg->fc_flags |= RTF_CACHE;
cfg->fc_flags |= (rtm->rtm_flags & RTNH_F_ONLINK);
if (tb[RTA_GATEWAY]) {
cfg->fc_gateway = nla_get_in6_addr(tb[RTA_GATEWAY]);
cfg->fc_flags |= RTF_GATEWAY;
}
if (tb[RTA_DST]) {
int plen = (rtm->rtm_dst_len + 7) >> 3;
if (nla_len(tb[RTA_DST]) < plen)
goto errout;
nla_memcpy(&cfg->fc_dst, tb[RTA_DST], plen);
}
if (tb[RTA_SRC]) {
int plen = (rtm->rtm_src_len + 7) >> 3;
if (nla_len(tb[RTA_SRC]) < plen)
goto errout;
nla_memcpy(&cfg->fc_src, tb[RTA_SRC], plen);
}
if (tb[RTA_PREFSRC])
cfg->fc_prefsrc = nla_get_in6_addr(tb[RTA_PREFSRC]);
if (tb[RTA_OIF])
cfg->fc_ifindex = nla_get_u32(tb[RTA_OIF]);
if (tb[RTA_PRIORITY])
cfg->fc_metric = nla_get_u32(tb[RTA_PRIORITY]);
if (tb[RTA_METRICS]) {
cfg->fc_mx = nla_data(tb[RTA_METRICS]);
cfg->fc_mx_len = nla_len(tb[RTA_METRICS]);
}
if (tb[RTA_TABLE])
cfg->fc_table = nla_get_u32(tb[RTA_TABLE]);
if (tb[RTA_MULTIPATH]) {
cfg->fc_mp = nla_data(tb[RTA_MULTIPATH]);
cfg->fc_mp_len = nla_len(tb[RTA_MULTIPATH]);
lwtunnel: fix autoload of lwt modules Trying to add an mpls encap route when the MPLS modules are not loaded hangs. For example: CONFIG_MPLS=y CONFIG_NET_MPLS_GSO=m CONFIG_MPLS_ROUTING=m CONFIG_MPLS_IPTUNNEL=m $ ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 The ip command hangs: root 880 826 0 21:25 pts/0 00:00:00 ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 $ cat /proc/880/stack [<ffffffff81065a9b>] call_usermodehelper_exec+0xd6/0x134 [<ffffffff81065efc>] __request_module+0x27b/0x30a [<ffffffff814542f6>] lwtunnel_build_state+0xe4/0x178 [<ffffffff814aa1e4>] fib_create_info+0x47f/0xdd4 [<ffffffff814ae451>] fib_table_insert+0x90/0x41f [<ffffffff814a8010>] inet_rtm_newroute+0x4b/0x52 ... modprobe is trying to load rtnl-lwt-MPLS: root 881 5 0 21:25 ? 00:00:00 /sbin/modprobe -q -- rtnl-lwt-MPLS and it hangs after loading mpls_router: $ cat /proc/881/stack [<ffffffff81441537>] rtnl_lock+0x12/0x14 [<ffffffff8142ca2a>] register_netdevice_notifier+0x16/0x179 [<ffffffffa0033025>] mpls_init+0x25/0x1000 [mpls_router] [<ffffffff81000471>] do_one_initcall+0x8e/0x13f [<ffffffff81119961>] do_init_module+0x5a/0x1e5 [<ffffffff810bd070>] load_module+0x13bd/0x17d6 ... The problem is that lwtunnel_build_state is called with rtnl lock held preventing mpls_init from registering. Given the potential references held by the time lwtunnel_build_state it can not drop the rtnl lock to the load module. So, extract the module loading code from lwtunnel_build_state into a new function to validate the encap type. The new function is called while converting the user request into a fib_config which is well before any table, device or fib entries are examined. Fixes: 745041e2aaf1 ("lwtunnel: autoload of lwt modules") Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-18 06:57:36 +08:00
err = lwtunnel_valid_encap_type_attr(cfg->fc_mp,
cfg->fc_mp_len, extack);
lwtunnel: fix autoload of lwt modules Trying to add an mpls encap route when the MPLS modules are not loaded hangs. For example: CONFIG_MPLS=y CONFIG_NET_MPLS_GSO=m CONFIG_MPLS_ROUTING=m CONFIG_MPLS_IPTUNNEL=m $ ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 The ip command hangs: root 880 826 0 21:25 pts/0 00:00:00 ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 $ cat /proc/880/stack [<ffffffff81065a9b>] call_usermodehelper_exec+0xd6/0x134 [<ffffffff81065efc>] __request_module+0x27b/0x30a [<ffffffff814542f6>] lwtunnel_build_state+0xe4/0x178 [<ffffffff814aa1e4>] fib_create_info+0x47f/0xdd4 [<ffffffff814ae451>] fib_table_insert+0x90/0x41f [<ffffffff814a8010>] inet_rtm_newroute+0x4b/0x52 ... modprobe is trying to load rtnl-lwt-MPLS: root 881 5 0 21:25 ? 00:00:00 /sbin/modprobe -q -- rtnl-lwt-MPLS and it hangs after loading mpls_router: $ cat /proc/881/stack [<ffffffff81441537>] rtnl_lock+0x12/0x14 [<ffffffff8142ca2a>] register_netdevice_notifier+0x16/0x179 [<ffffffffa0033025>] mpls_init+0x25/0x1000 [mpls_router] [<ffffffff81000471>] do_one_initcall+0x8e/0x13f [<ffffffff81119961>] do_init_module+0x5a/0x1e5 [<ffffffff810bd070>] load_module+0x13bd/0x17d6 ... The problem is that lwtunnel_build_state is called with rtnl lock held preventing mpls_init from registering. Given the potential references held by the time lwtunnel_build_state it can not drop the rtnl lock to the load module. So, extract the module loading code from lwtunnel_build_state into a new function to validate the encap type. The new function is called while converting the user request into a fib_config which is well before any table, device or fib entries are examined. Fixes: 745041e2aaf1 ("lwtunnel: autoload of lwt modules") Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-18 06:57:36 +08:00
if (err < 0)
goto errout;
}
if (tb[RTA_PREF]) {
pref = nla_get_u8(tb[RTA_PREF]);
if (pref != ICMPV6_ROUTER_PREF_LOW &&
pref != ICMPV6_ROUTER_PREF_HIGH)
pref = ICMPV6_ROUTER_PREF_MEDIUM;
cfg->fc_flags |= RTF_PREF(pref);
}
if (tb[RTA_ENCAP])
cfg->fc_encap = tb[RTA_ENCAP];
lwtunnel: fix autoload of lwt modules Trying to add an mpls encap route when the MPLS modules are not loaded hangs. For example: CONFIG_MPLS=y CONFIG_NET_MPLS_GSO=m CONFIG_MPLS_ROUTING=m CONFIG_MPLS_IPTUNNEL=m $ ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 The ip command hangs: root 880 826 0 21:25 pts/0 00:00:00 ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 $ cat /proc/880/stack [<ffffffff81065a9b>] call_usermodehelper_exec+0xd6/0x134 [<ffffffff81065efc>] __request_module+0x27b/0x30a [<ffffffff814542f6>] lwtunnel_build_state+0xe4/0x178 [<ffffffff814aa1e4>] fib_create_info+0x47f/0xdd4 [<ffffffff814ae451>] fib_table_insert+0x90/0x41f [<ffffffff814a8010>] inet_rtm_newroute+0x4b/0x52 ... modprobe is trying to load rtnl-lwt-MPLS: root 881 5 0 21:25 ? 00:00:00 /sbin/modprobe -q -- rtnl-lwt-MPLS and it hangs after loading mpls_router: $ cat /proc/881/stack [<ffffffff81441537>] rtnl_lock+0x12/0x14 [<ffffffff8142ca2a>] register_netdevice_notifier+0x16/0x179 [<ffffffffa0033025>] mpls_init+0x25/0x1000 [mpls_router] [<ffffffff81000471>] do_one_initcall+0x8e/0x13f [<ffffffff81119961>] do_init_module+0x5a/0x1e5 [<ffffffff810bd070>] load_module+0x13bd/0x17d6 ... The problem is that lwtunnel_build_state is called with rtnl lock held preventing mpls_init from registering. Given the potential references held by the time lwtunnel_build_state it can not drop the rtnl lock to the load module. So, extract the module loading code from lwtunnel_build_state into a new function to validate the encap type. The new function is called while converting the user request into a fib_config which is well before any table, device or fib entries are examined. Fixes: 745041e2aaf1 ("lwtunnel: autoload of lwt modules") Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-18 06:57:36 +08:00
if (tb[RTA_ENCAP_TYPE]) {
cfg->fc_encap_type = nla_get_u16(tb[RTA_ENCAP_TYPE]);
err = lwtunnel_valid_encap_type(cfg->fc_encap_type, extack);
lwtunnel: fix autoload of lwt modules Trying to add an mpls encap route when the MPLS modules are not loaded hangs. For example: CONFIG_MPLS=y CONFIG_NET_MPLS_GSO=m CONFIG_MPLS_ROUTING=m CONFIG_MPLS_IPTUNNEL=m $ ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 The ip command hangs: root 880 826 0 21:25 pts/0 00:00:00 ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 $ cat /proc/880/stack [<ffffffff81065a9b>] call_usermodehelper_exec+0xd6/0x134 [<ffffffff81065efc>] __request_module+0x27b/0x30a [<ffffffff814542f6>] lwtunnel_build_state+0xe4/0x178 [<ffffffff814aa1e4>] fib_create_info+0x47f/0xdd4 [<ffffffff814ae451>] fib_table_insert+0x90/0x41f [<ffffffff814a8010>] inet_rtm_newroute+0x4b/0x52 ... modprobe is trying to load rtnl-lwt-MPLS: root 881 5 0 21:25 ? 00:00:00 /sbin/modprobe -q -- rtnl-lwt-MPLS and it hangs after loading mpls_router: $ cat /proc/881/stack [<ffffffff81441537>] rtnl_lock+0x12/0x14 [<ffffffff8142ca2a>] register_netdevice_notifier+0x16/0x179 [<ffffffffa0033025>] mpls_init+0x25/0x1000 [mpls_router] [<ffffffff81000471>] do_one_initcall+0x8e/0x13f [<ffffffff81119961>] do_init_module+0x5a/0x1e5 [<ffffffff810bd070>] load_module+0x13bd/0x17d6 ... The problem is that lwtunnel_build_state is called with rtnl lock held preventing mpls_init from registering. Given the potential references held by the time lwtunnel_build_state it can not drop the rtnl lock to the load module. So, extract the module loading code from lwtunnel_build_state into a new function to validate the encap type. The new function is called while converting the user request into a fib_config which is well before any table, device or fib entries are examined. Fixes: 745041e2aaf1 ("lwtunnel: autoload of lwt modules") Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-18 06:57:36 +08:00
if (err < 0)
goto errout;
}
if (tb[RTA_EXPIRES]) {
unsigned long timeout = addrconf_timeout_fixup(nla_get_u32(tb[RTA_EXPIRES]), HZ);
if (addrconf_finite_timeout(timeout)) {
cfg->fc_expires = jiffies_to_clock_t(timeout * HZ);
cfg->fc_flags |= RTF_EXPIRES;
}
}
err = 0;
errout:
return err;
}
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
struct rt6_nh {
struct fib6_info *fib6_info;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
struct fib6_config r_cfg;
struct list_head next;
};
static int ip6_route_info_append(struct net *net,
struct list_head *rt6_nh_list,
struct fib6_info *rt,
struct fib6_config *r_cfg)
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
{
struct rt6_nh *nh;
int err = -EEXIST;
list_for_each_entry(nh, rt6_nh_list, next) {
/* check if fib6_info already exists */
if (rt6_duplicate_nexthop(nh->fib6_info, rt))
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
return err;
}
nh = kzalloc(sizeof(*nh), GFP_KERNEL);
if (!nh)
return -ENOMEM;
nh->fib6_info = rt;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
memcpy(&nh->r_cfg, r_cfg, sizeof(*r_cfg));
list_add_tail(&nh->next, rt6_nh_list);
return 0;
}
static void ip6_route_mpath_notify(struct fib6_info *rt,
struct fib6_info *rt_last,
struct nl_info *info,
__u16 nlflags)
{
/* if this is an APPEND route, then rt points to the first route
* inserted and rt_last points to last route inserted. Userspace
* wants a consistent dump of the route which starts at the first
* nexthop. Since sibling routes are always added at the end of
* the list, find the first sibling of the last route appended
*/
if ((nlflags & NLM_F_APPEND) && rt_last && rt_last->fib6_nsiblings) {
rt = list_first_entry(&rt_last->fib6_siblings,
struct fib6_info,
fib6_siblings);
}
if (rt)
inet6_rt_notify(RTM_NEWROUTE, rt, info, nlflags);
}
static int ip6_route_multipath_add(struct fib6_config *cfg,
struct netlink_ext_ack *extack)
{
struct fib6_info *rt_notif = NULL, *rt_last = NULL;
struct nl_info *info = &cfg->fc_nlinfo;
struct fib6_config r_cfg;
struct rtnexthop *rtnh;
struct fib6_info *rt;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
struct rt6_nh *err_nh;
struct rt6_nh *nh, *nh_safe;
__u16 nlflags;
int remaining;
int attrlen;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
int err = 1;
int nhn = 0;
int replace = (cfg->fc_nlinfo.nlh &&
(cfg->fc_nlinfo.nlh->nlmsg_flags & NLM_F_REPLACE));
LIST_HEAD(rt6_nh_list);
nlflags = replace ? NLM_F_REPLACE : NLM_F_CREATE;
if (info->nlh && info->nlh->nlmsg_flags & NLM_F_APPEND)
nlflags |= NLM_F_APPEND;
remaining = cfg->fc_mp_len;
rtnh = (struct rtnexthop *)cfg->fc_mp;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
/* Parse a Multipath Entry and build a list (rt6_nh_list) of
* fib6_info structs per nexthop
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
*/
while (rtnh_ok(rtnh, remaining)) {
memcpy(&r_cfg, cfg, sizeof(*cfg));
if (rtnh->rtnh_ifindex)
r_cfg.fc_ifindex = rtnh->rtnh_ifindex;
attrlen = rtnh_attrlen(rtnh);
if (attrlen > 0) {
struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
nla = nla_find(attrs, attrlen, RTA_GATEWAY);
if (nla) {
r_cfg.fc_gateway = nla_get_in6_addr(nla);
r_cfg.fc_flags |= RTF_GATEWAY;
}
r_cfg.fc_encap = nla_find(attrs, attrlen, RTA_ENCAP);
nla = nla_find(attrs, attrlen, RTA_ENCAP_TYPE);
if (nla)
r_cfg.fc_encap_type = nla_get_u16(nla);
}
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
r_cfg.fc_flags |= (rtnh->rtnh_flags & RTNH_F_ONLINK);
rt = ip6_route_info_create(&r_cfg, GFP_KERNEL, extack);
if (IS_ERR(rt)) {
err = PTR_ERR(rt);
rt = NULL;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
goto cleanup;
}
net/ipv6: Do not allow device only routes via the multipath API Eric reported that reverting the patch that fixed and simplified IPv6 multipath routes means reverting back to invalid userspace notifications. eg., $ ip -6 route add 2001:db8:1::/64 nexthop dev eth0 nexthop dev eth1 only generates a single notification: 2001:db8:1::/64 dev eth0 metric 1024 pref medium While working on a fix for this problem I found another case that is just broken completely - a multipath route with a gateway followed by device followed by gateway: $ ip -6 ro add 2001:db8:103::/64 nexthop via 2001:db8:1::64 nexthop dev dummy2 nexthop via 2001:db8:3::64 In this case the device only route is dropped completely - no notification to userpsace but no addition to the FIB either: $ ip -6 ro ls 2001:db8:1::/64 dev dummy1 proto kernel metric 256 pref medium 2001:db8:2::/64 dev dummy2 proto kernel metric 256 pref medium 2001:db8:3::/64 dev dummy3 proto kernel metric 256 pref medium 2001:db8:103::/64 metric 1024 nexthop via 2001:db8:1::64 dev dummy1 weight 1 nexthop via 2001:db8:3::64 dev dummy3 weight 1 pref medium fe80::/64 dev dummy1 proto kernel metric 256 pref medium fe80::/64 dev dummy2 proto kernel metric 256 pref medium fe80::/64 dev dummy3 proto kernel metric 256 pref medium Really, IPv6 multipath is just FUBAR'ed beyond repair when it comes to device only routes, so do not allow it all. This change will break any scripts relying on the mpath api for insert, but I don't see any other way to handle the permutations. Besides, since the routes are added to the FIB as standalone (non-multipath) routes the kernel is not doing what the user requested, so it might as well tell the user that. Reported-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David Ahern <dsahern@gmail.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-16 00:35:19 +08:00
if (!rt6_qualify_for_ecmp(rt)) {
err = -EINVAL;
NL_SET_ERR_MSG(extack,
"Device only routes can not be added for IPv6 using the multipath API.");
fib6_info_release(rt);
goto cleanup;
}
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
rt->fib6_nh.nh_weight = rtnh->rtnh_hops + 1;
err = ip6_route_info_append(info->nl_net, &rt6_nh_list,
rt, &r_cfg);
if (err) {
fib6_info_release(rt);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
goto cleanup;
}
rtnh = rtnh_next(rtnh, &remaining);
}
/* for add and replace send one notification with all nexthops.
* Skip the notification in fib6_add_rt2node and send one with
* the full route when done
*/
info->skip_notify = 1;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
err_nh = NULL;
list_for_each_entry(nh, &rt6_nh_list, next) {
err = __ip6_ins_rt(nh->fib6_info, info, extack);
fib6_info_release(nh->fib6_info);
net/ipv6: prevent use after free in ip6_route_mpath_notify syzbot reported a use-after-free: BUG: KASAN: use-after-free in ip6_route_mpath_notify+0xe9/0x100 net/ipv6/route.c:4180 Read of size 4 at addr ffff8801bf789cf0 by task syz-executor756/4555 CPU: 1 PID: 4555 Comm: syz-executor756 Not tainted 4.17.0-rc7+ #78 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1b9/0x294 lib/dump_stack.c:113 print_address_description+0x6c/0x20b mm/kasan/report.c:256 kasan_report_error mm/kasan/report.c:354 [inline] kasan_report.cold.7+0x242/0x2fe mm/kasan/report.c:412 __asan_report_load4_noabort+0x14/0x20 mm/kasan/report.c:432 ip6_route_mpath_notify+0xe9/0x100 net/ipv6/route.c:4180 ip6_route_multipath_add+0x615/0x1910 net/ipv6/route.c:4303 inet6_rtm_newroute+0xe3/0x160 net/ipv6/route.c:4391 ... Allocated by task 4555: save_stack+0x43/0xd0 mm/kasan/kasan.c:448 set_track mm/kasan/kasan.c:460 [inline] kasan_kmalloc+0xc4/0xe0 mm/kasan/kasan.c:553 kasan_slab_alloc+0x12/0x20 mm/kasan/kasan.c:490 kmem_cache_alloc+0x12e/0x760 mm/slab.c:3554 dst_alloc+0xbb/0x1d0 net/core/dst.c:104 __ip6_dst_alloc+0x35/0xa0 net/ipv6/route.c:361 ip6_dst_alloc+0x29/0xb0 net/ipv6/route.c:376 ip6_route_info_create+0x4d4/0x3a30 net/ipv6/route.c:2834 ip6_route_multipath_add+0xc7e/0x1910 net/ipv6/route.c:4240 inet6_rtm_newroute+0xe3/0x160 net/ipv6/route.c:4391 ... Freed by task 4555: save_stack+0x43/0xd0 mm/kasan/kasan.c:448 set_track mm/kasan/kasan.c:460 [inline] __kasan_slab_free+0x11a/0x170 mm/kasan/kasan.c:521 kasan_slab_free+0xe/0x10 mm/kasan/kasan.c:528 __cache_free mm/slab.c:3498 [inline] kmem_cache_free+0x86/0x2d0 mm/slab.c:3756 dst_destroy+0x267/0x3c0 net/core/dst.c:140 dst_release_immediate+0x71/0x9e net/core/dst.c:205 fib6_add+0xa40/0x1650 net/ipv6/ip6_fib.c:1305 __ip6_ins_rt+0x6c/0x90 net/ipv6/route.c:1011 ip6_route_multipath_add+0x513/0x1910 net/ipv6/route.c:4267 inet6_rtm_newroute+0xe3/0x160 net/ipv6/route.c:4391 ... The problem is that rt_last can point to a deleted route if the insert fails. One reproducer is to insert a route and then add a multipath route that has a duplicate nexthop.e.g,: $ ip -6 ro add vrf red 2001:db8:101::/64 nexthop via 2001:db8:1::2 $ ip -6 ro append vrf red 2001:db8:101::/64 nexthop via 2001:db8:1::4 nexthop via 2001:db8:1::2 Fix by not setting rt_last until the it is verified the insert succeeded. Fixes: 3b1137fe7482 ("net: ipv6: Change notifications for multipath add to RTA_MULTIPATH") Cc: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: David Ahern <dsahern@gmail.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-06-05 04:41:42 +08:00
if (!err) {
/* save reference to last route successfully inserted */
rt_last = nh->fib6_info;
/* save reference to first route for notification */
if (!rt_notif)
rt_notif = nh->fib6_info;
}
/* nh->fib6_info is used or freed at this point, reset to NULL*/
nh->fib6_info = NULL;
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
if (err) {
if (replace && nhn)
NL_SET_ERR_MSG_MOD(extack,
"multipath route replace failed (check consistency of installed routes)");
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
err_nh = nh;
goto add_errout;
}
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
/* Because each route is added like a single route we remove
* these flags after the first nexthop: if there is a collision,
* we have already failed to add the first nexthop:
* fib6_add_rt2node() has rejected it; when replacing, old
* nexthops have been replaced by first new, the rest should
* be added to it.
*/
cfg->fc_nlinfo.nlh->nlmsg_flags &= ~(NLM_F_EXCL |
NLM_F_REPLACE);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
nhn++;
}
/* success ... tell user about new route */
ip6_route_mpath_notify(rt_notif, rt_last, info, nlflags);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
goto cleanup;
add_errout:
/* send notification for routes that were added so that
* the delete notifications sent by ip6_route_del are
* coherent
*/
if (rt_notif)
ip6_route_mpath_notify(rt_notif, rt_last, info, nlflags);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
/* Delete routes that were already added */
list_for_each_entry(nh, &rt6_nh_list, next) {
if (err_nh == nh)
break;
ip6_route_del(&nh->r_cfg, extack);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
}
cleanup:
list_for_each_entry_safe(nh, nh_safe, &rt6_nh_list, next) {
if (nh->fib6_info)
fib6_info_release(nh->fib6_info);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
list_del(&nh->next);
kfree(nh);
}
return err;
}
static int ip6_route_multipath_del(struct fib6_config *cfg,
struct netlink_ext_ack *extack)
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
{
struct fib6_config r_cfg;
struct rtnexthop *rtnh;
int remaining;
int attrlen;
int err = 1, last_err = 0;
remaining = cfg->fc_mp_len;
rtnh = (struct rtnexthop *)cfg->fc_mp;
/* Parse a Multipath Entry */
while (rtnh_ok(rtnh, remaining)) {
memcpy(&r_cfg, cfg, sizeof(*cfg));
if (rtnh->rtnh_ifindex)
r_cfg.fc_ifindex = rtnh->rtnh_ifindex;
attrlen = rtnh_attrlen(rtnh);
if (attrlen > 0) {
struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
nla = nla_find(attrs, attrlen, RTA_GATEWAY);
if (nla) {
nla_memcpy(&r_cfg.fc_gateway, nla, 16);
r_cfg.fc_flags |= RTF_GATEWAY;
}
}
err = ip6_route_del(&r_cfg, extack);
ipv6: fix multipath route replace error recovery Problem: The ecmp route replace support for ipv6 in the kernel, deletes the existing ecmp route too early, ie when it installs the first nexthop. If there is an error in installing the subsequent nexthops, its too late to recover the already deleted existing route leaving the fib in an inconsistent state. This patch reduces the possibility of this by doing the following: a) Changes the existing multipath route add code to a two stage process: build rt6_infos + insert them ip6_route_add rt6_info creation code is moved into ip6_route_info_create. b) This ensures that most errors are caught during building rt6_infos and we fail early c) Separates multipath add and del code. Because add needs the special two stage mode in a) and delete essentially does not care. d) In any event if the code fails during inserting a route again, a warning is printed (This should be unlikely) Before the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show /* previously added ecmp route 3000:1000:1000:1000::2 dissappears from * kernel */ After the patch: $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 /* Try replacing the route with a duplicate nexthop */ $ip -6 route change 3000:1000:1000:1000::2/128 nexthop via fe80::202:ff:fe00:b dev swp49s0 nexthop via fe80::202:ff:fe00:d dev swp49s1 nexthop via fe80::202:ff:fe00:d dev swp49s1 RTNETLINK answers: File exists $ip -6 route show 3000:1000:1000:1000::2 via fe80::202:ff:fe00:b dev swp49s0 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:d dev swp49s1 metric 1024 3000:1000:1000:1000::2 via fe80::202:ff:fe00:f dev swp49s2 metric 1024 Fixes: 27596472473a ("ipv6: fix ECMP route replacement") Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-09-09 01:53:04 +08:00
if (err)
last_err = err;
rtnh = rtnh_next(rtnh, &remaining);
}
return last_err;
}
static int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh,
struct netlink_ext_ack *extack)
{
struct fib6_config cfg;
int err;
err = rtm_to_fib6_config(skb, nlh, &cfg, extack);
if (err < 0)
return err;
if (cfg.fc_mp)
return ip6_route_multipath_del(&cfg, extack);
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
else {
cfg.fc_delete_all_nh = 1;
return ip6_route_del(&cfg, extack);
net: ipv6: Allow shorthand delete of all nexthops in multipath route IPv4 allows multipath routes to be deleted using just the prefix and length. For example: $ ip ro ls vrf red unreachable default metric 8192 1.1.1.0/24 nexthop via 10.100.1.254 dev eth1 weight 1 nexthop via 10.11.200.2 dev eth11.200 weight 1 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 $ ip ro del 1.1.1.0/24 vrf red $ ip ro ls vrf red unreachable default metric 8192 10.11.200.0/24 dev eth11.200 proto kernel scope link src 10.11.200.3 10.100.1.0/24 dev eth1 proto kernel scope link src 10.100.1.3 The same notation does not work with IPv6 because of how multipath routes are implemented for IPv6. For IPv6 only the first nexthop of a multipath route is deleted if the request contains only a prefix and length. This leads to unnecessary complexity in userspace dealing with IPv6 multipath routes. This patch allows all nexthops to be deleted without specifying each one in the delete request. Internally, this is done by walking the sibling list of the route matching the specifications given (prefix, length, metric, protocol, etc). $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium 2001:db8:200::/120 via 2001:db8:1::2 dev eth1 metric 1024 pref medium 2001:db8:200::/120 via 2001:db8:2::2 dev eth2 metric 1024 pref medium ... $ ip -6 ro del vrf red 2001:db8:200::/120 $ ip -6 ro ls vrf red 2001:db8:1::/120 dev eth1 proto kernel metric 256 pref medium 2001:db8:2::/120 dev eth2 proto kernel metric 256 pref medium ... Because IPv6 allows individual nexthops to be deleted without deleting the entire route, the ip6_route_multipath_del and non-multipath code path (ip6_route_del) have to be discriminated so that all nexthops are only deleted for the latter case. This is done by making the existing fc_type in fib6_config a u16 and then adding a new u16 field with fc_delete_all_nh as the first bit. Suggested-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-03 04:37:08 +08:00
}
}
static int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh,
struct netlink_ext_ack *extack)
{
struct fib6_config cfg;
int err;
err = rtm_to_fib6_config(skb, nlh, &cfg, extack);
if (err < 0)
return err;
if (cfg.fc_mp)
return ip6_route_multipath_add(&cfg, extack);
else
return ip6_route_add(&cfg, GFP_KERNEL, extack);
}
static size_t rt6_nlmsg_size(struct fib6_info *rt)
{
int nexthop_len = 0;
if (rt->fib6_nsiblings) {
nexthop_len = nla_total_size(0) /* RTA_MULTIPATH */
+ NLA_ALIGN(sizeof(struct rtnexthop))
+ nla_total_size(16) /* RTA_GATEWAY */
+ lwtunnel_get_encap_size(rt->fib6_nh.nh_lwtstate);
nexthop_len *= rt->fib6_nsiblings;
}
return NLMSG_ALIGN(sizeof(struct rtmsg))
+ nla_total_size(16) /* RTA_SRC */
+ nla_total_size(16) /* RTA_DST */
+ nla_total_size(16) /* RTA_GATEWAY */
+ nla_total_size(16) /* RTA_PREFSRC */
+ nla_total_size(4) /* RTA_TABLE */
+ nla_total_size(4) /* RTA_IIF */
+ nla_total_size(4) /* RTA_OIF */
+ nla_total_size(4) /* RTA_PRIORITY */
+ RTAX_MAX * nla_total_size(4) /* RTA_METRICS */
+ nla_total_size(sizeof(struct rta_cacheinfo))
+ nla_total_size(TCP_CA_NAME_MAX) /* RTAX_CC_ALGO */
+ nla_total_size(1) /* RTA_PREF */
+ lwtunnel_get_encap_size(rt->fib6_nh.nh_lwtstate)
+ nexthop_len;
}
static int rt6_nexthop_info(struct sk_buff *skb, struct fib6_info *rt,
unsigned int *flags, bool skip_oif)
{
if (rt->fib6_nh.nh_flags & RTNH_F_DEAD)
*flags |= RTNH_F_DEAD;
if (rt->fib6_nh.nh_flags & RTNH_F_LINKDOWN) {
*flags |= RTNH_F_LINKDOWN;
rcu_read_lock();
if (fib6_ignore_linkdown(rt))
*flags |= RTNH_F_DEAD;
rcu_read_unlock();
}
if (rt->fib6_flags & RTF_GATEWAY) {
if (nla_put_in6_addr(skb, RTA_GATEWAY, &rt->fib6_nh.nh_gw) < 0)
goto nla_put_failure;
}
*flags |= (rt->fib6_nh.nh_flags & RTNH_F_ONLINK);
if (rt->fib6_nh.nh_flags & RTNH_F_OFFLOAD)
*flags |= RTNH_F_OFFLOAD;
/* not needed for multipath encoding b/c it has a rtnexthop struct */
if (!skip_oif && rt->fib6_nh.nh_dev &&
nla_put_u32(skb, RTA_OIF, rt->fib6_nh.nh_dev->ifindex))
goto nla_put_failure;
if (rt->fib6_nh.nh_lwtstate &&
lwtunnel_fill_encap(skb, rt->fib6_nh.nh_lwtstate) < 0)
goto nla_put_failure;
return 0;
nla_put_failure:
return -EMSGSIZE;
}
/* add multipath next hop */
static int rt6_add_nexthop(struct sk_buff *skb, struct fib6_info *rt)
{
const struct net_device *dev = rt->fib6_nh.nh_dev;
struct rtnexthop *rtnh;
unsigned int flags = 0;
rtnh = nla_reserve_nohdr(skb, sizeof(*rtnh));
if (!rtnh)
goto nla_put_failure;
rtnh->rtnh_hops = rt->fib6_nh.nh_weight - 1;
rtnh->rtnh_ifindex = dev ? dev->ifindex : 0;
if (rt6_nexthop_info(skb, rt, &flags, true) < 0)
goto nla_put_failure;
rtnh->rtnh_flags = flags;
/* length of rtnetlink header + attributes */
rtnh->rtnh_len = nlmsg_get_pos(skb) - (void *)rtnh;
return 0;
nla_put_failure:
return -EMSGSIZE;
}
static int rt6_fill_node(struct net *net, struct sk_buff *skb,
struct fib6_info *rt, struct dst_entry *dst,
struct in6_addr *dest, struct in6_addr *src,
int iif, int type, u32 portid, u32 seq,
unsigned int flags)
{
struct rt6_info *rt6 = (struct rt6_info *)dst;
struct rt6key *rt6_dst, *rt6_src;
u32 *pmetrics, table, rt6_flags;
struct nlmsghdr *nlh;
struct rtmsg *rtm;
long expires = 0;
nlh = nlmsg_put(skb, portid, seq, type, sizeof(*rtm), flags);
if (!nlh)
return -EMSGSIZE;
if (rt6) {
rt6_dst = &rt6->rt6i_dst;
rt6_src = &rt6->rt6i_src;
rt6_flags = rt6->rt6i_flags;
} else {
rt6_dst = &rt->fib6_dst;
rt6_src = &rt->fib6_src;
rt6_flags = rt->fib6_flags;
}
rtm = nlmsg_data(nlh);
rtm->rtm_family = AF_INET6;
rtm->rtm_dst_len = rt6_dst->plen;
rtm->rtm_src_len = rt6_src->plen;
rtm->rtm_tos = 0;
if (rt->fib6_table)
table = rt->fib6_table->tb6_id;
else
table = RT6_TABLE_UNSPEC;
rtm->rtm_table = table;
if (nla_put_u32(skb, RTA_TABLE, table))
goto nla_put_failure;
rtm->rtm_type = rt->fib6_type;
rtm->rtm_flags = 0;
rtm->rtm_scope = RT_SCOPE_UNIVERSE;
rtm->rtm_protocol = rt->fib6_protocol;
if (rt6_flags & RTF_CACHE)
rtm->rtm_flags |= RTM_F_CLONED;
if (dest) {
if (nla_put_in6_addr(skb, RTA_DST, dest))
goto nla_put_failure;
rtm->rtm_dst_len = 128;
} else if (rtm->rtm_dst_len)
if (nla_put_in6_addr(skb, RTA_DST, &rt6_dst->addr))
goto nla_put_failure;
#ifdef CONFIG_IPV6_SUBTREES
if (src) {
if (nla_put_in6_addr(skb, RTA_SRC, src))
goto nla_put_failure;
rtm->rtm_src_len = 128;
} else if (rtm->rtm_src_len &&
nla_put_in6_addr(skb, RTA_SRC, &rt6_src->addr))
goto nla_put_failure;
#endif
if (iif) {
#ifdef CONFIG_IPV6_MROUTE
if (ipv6_addr_is_multicast(&rt6_dst->addr)) {
int err = ip6mr_get_route(net, skb, rtm, portid);
if (err == 0)
return 0;
if (err < 0)
goto nla_put_failure;
} else
#endif
if (nla_put_u32(skb, RTA_IIF, iif))
goto nla_put_failure;
} else if (dest) {
struct in6_addr saddr_buf;
if (ip6_route_get_saddr(net, rt, dest, 0, &saddr_buf) == 0 &&
nla_put_in6_addr(skb, RTA_PREFSRC, &saddr_buf))
goto nla_put_failure;
}
if (rt->fib6_prefsrc.plen) {
struct in6_addr saddr_buf;
saddr_buf = rt->fib6_prefsrc.addr;
if (nla_put_in6_addr(skb, RTA_PREFSRC, &saddr_buf))
goto nla_put_failure;
}
pmetrics = dst ? dst_metrics_ptr(dst) : rt->fib6_metrics->metrics;
if (rtnetlink_put_metrics(skb, pmetrics) < 0)
goto nla_put_failure;
if (nla_put_u32(skb, RTA_PRIORITY, rt->fib6_metric))
goto nla_put_failure;
/* For multipath routes, walk the siblings list and add
* each as a nexthop within RTA_MULTIPATH.
*/
if (rt6) {
if (rt6_flags & RTF_GATEWAY &&
nla_put_in6_addr(skb, RTA_GATEWAY, &rt6->rt6i_gateway))
goto nla_put_failure;
if (dst->dev && nla_put_u32(skb, RTA_OIF, dst->dev->ifindex))
goto nla_put_failure;
} else if (rt->fib6_nsiblings) {
struct fib6_info *sibling, *next_sibling;
struct nlattr *mp;
mp = nla_nest_start(skb, RTA_MULTIPATH);
if (!mp)
goto nla_put_failure;
if (rt6_add_nexthop(skb, rt) < 0)
goto nla_put_failure;
list_for_each_entry_safe(sibling, next_sibling,
&rt->fib6_siblings, fib6_siblings) {
if (rt6_add_nexthop(skb, sibling) < 0)
goto nla_put_failure;
}
nla_nest_end(skb, mp);
} else {
if (rt6_nexthop_info(skb, rt, &rtm->rtm_flags, false) < 0)
goto nla_put_failure;
}
if (rt6_flags & RTF_EXPIRES) {
expires = dst ? dst->expires : rt->expires;
expires -= jiffies;
}
if (rtnl_put_cacheinfo(skb, dst, 0, expires, dst ? dst->error : 0) < 0)
goto nla_put_failure;
if (nla_put_u8(skb, RTA_PREF, IPV6_EXTRACT_PREF(rt6_flags)))
goto nla_put_failure;
netlink: make nlmsg_end() and genlmsg_end() void Contrary to common expectations for an "int" return, these functions return only a positive value -- if used correctly they cannot even return 0 because the message header will necessarily be in the skb. This makes the very common pattern of if (genlmsg_end(...) < 0) { ... } be a whole bunch of dead code. Many places also simply do return nlmsg_end(...); and the caller is expected to deal with it. This also commonly (at least for me) causes errors, because it is very common to write if (my_function(...)) /* error condition */ and if my_function() does "return nlmsg_end()" this is of course wrong. Additionally, there's not a single place in the kernel that actually needs the message length returned, and if anyone needs it later then it'll be very easy to just use skb->len there. Remove this, and make the functions void. This removes a bunch of dead code as described above. The patch adds lines because I did - return nlmsg_end(...); + nlmsg_end(...); + return 0; I could have preserved all the function's return values by returning skb->len, but instead I've audited all the places calling the affected functions and found that none cared. A few places actually compared the return value with <= 0 in dump functionality, but that could just be changed to < 0 with no change in behaviour, so I opted for the more efficient version. One instance of the error I've made numerous times now is also present in net/phonet/pn_netlink.c in the route_dumpit() function - it didn't check for <0 or <=0 and thus broke out of the loop every single time. I've preserved this since it will (I think) have caused the messages to userspace to be formatted differently with just a single message for every SKB returned to userspace. It's possible that this isn't needed for the tools that actually use this, but I don't even know what they are so couldn't test that changing this behaviour would be acceptable. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-01-17 05:09:00 +08:00
nlmsg_end(skb, nlh);
return 0;
nla_put_failure:
nlmsg_cancel(skb, nlh);
return -EMSGSIZE;
}
static bool fib6_info_uses_dev(const struct fib6_info *f6i,
const struct net_device *dev)
{
if (f6i->fib6_nh.nh_dev == dev)
return true;
if (f6i->fib6_nsiblings) {
struct fib6_info *sibling, *next_sibling;
list_for_each_entry_safe(sibling, next_sibling,
&f6i->fib6_siblings, fib6_siblings) {
if (sibling->fib6_nh.nh_dev == dev)
return true;
}
}
return false;
}
int rt6_dump_route(struct fib6_info *rt, void *p_arg)
{
struct rt6_rtnl_dump_arg *arg = (struct rt6_rtnl_dump_arg *) p_arg;
struct fib_dump_filter *filter = &arg->filter;
unsigned int flags = NLM_F_MULTI;
net: ipv6: ignore null_entry on route dumps lkp-robot reported a BUG: [ 10.151226] BUG: unable to handle kernel NULL pointer dereference at 00000198 [ 10.152525] IP: rt6_fill_node+0x164/0x4b8 [ 10.153307] *pdpt = 0000000012ee5001 *pde = 0000000000000000 [ 10.153309] [ 10.154492] Oops: 0000 [#1] [ 10.154987] CPU: 0 PID: 909 Comm: netifd Not tainted 4.10.0-rc4-00722-g41e8c70ee162-dirty #10 [ 10.156482] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014 [ 10.158254] task: d0deb000 task.stack: d0e0c000 [ 10.159059] EIP: rt6_fill_node+0x164/0x4b8 [ 10.159780] EFLAGS: 00010296 CPU: 0 [ 10.160404] EAX: 00000000 EBX: d10c2358 ECX: c1f7c6cc EDX: c1f6ff44 [ 10.161469] ESI: 00000000 EDI: c2059900 EBP: d0e0dc4c ESP: d0e0dbe4 [ 10.162534] DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068 [ 10.163482] CR0: 80050033 CR2: 00000198 CR3: 10d94660 CR4: 000006b0 [ 10.164535] Call Trace: [ 10.164993] ? paravirt_sched_clock+0x9/0xd [ 10.165727] ? sched_clock+0x9/0xc [ 10.166329] ? sched_clock_cpu+0x19/0xe9 [ 10.166991] ? lock_release+0x13e/0x36c [ 10.167652] rt6_dump_route+0x4c/0x56 [ 10.168276] fib6_dump_node+0x1d/0x3d [ 10.168913] fib6_walk_continue+0xab/0x167 [ 10.169611] fib6_walk+0x2a/0x40 [ 10.170182] inet6_dump_fib+0xfb/0x1e0 [ 10.170855] netlink_dump+0xcd/0x21f This happens when the loopback device is set down and a ipv6 fib route dump is requested. ip6_null_entry is the root of all ipv6 fib tables making it integrated into the table and hence passed to the ipv6 route dump code. The null_entry route uses the loopback device for dst.dev but may not have rt6i_idev set because of the order in which initializations are done -- ip6_route_net_init is run before addrconf_init has initialized the loopback device. Fixing the initialization order is a much bigger problem with no obvious solution thus far. The BUG is triggered when the loopback is set down and the netif_running check added by a1a22c1206 fails. The fill_node descends to checking rt->rt6i_idev for ignore_routes_with_linkdown and since rt6i_idev is NULL it faults. The null_entry route should not be processed in a dump request. Catch and ignore. This check is done in rt6_dump_route as it is the highest place in the callchain with knowledge of both the route and the network namespace. Fixes: a1a22c1206("net: ipv6: Keep nexthop of multipath route on admin down") Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-27 05:54:08 +08:00
struct net *net = arg->net;
if (rt == net->ipv6.fib6_null_entry)
net: ipv6: ignore null_entry on route dumps lkp-robot reported a BUG: [ 10.151226] BUG: unable to handle kernel NULL pointer dereference at 00000198 [ 10.152525] IP: rt6_fill_node+0x164/0x4b8 [ 10.153307] *pdpt = 0000000012ee5001 *pde = 0000000000000000 [ 10.153309] [ 10.154492] Oops: 0000 [#1] [ 10.154987] CPU: 0 PID: 909 Comm: netifd Not tainted 4.10.0-rc4-00722-g41e8c70ee162-dirty #10 [ 10.156482] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014 [ 10.158254] task: d0deb000 task.stack: d0e0c000 [ 10.159059] EIP: rt6_fill_node+0x164/0x4b8 [ 10.159780] EFLAGS: 00010296 CPU: 0 [ 10.160404] EAX: 00000000 EBX: d10c2358 ECX: c1f7c6cc EDX: c1f6ff44 [ 10.161469] ESI: 00000000 EDI: c2059900 EBP: d0e0dc4c ESP: d0e0dbe4 [ 10.162534] DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068 [ 10.163482] CR0: 80050033 CR2: 00000198 CR3: 10d94660 CR4: 000006b0 [ 10.164535] Call Trace: [ 10.164993] ? paravirt_sched_clock+0x9/0xd [ 10.165727] ? sched_clock+0x9/0xc [ 10.166329] ? sched_clock_cpu+0x19/0xe9 [ 10.166991] ? lock_release+0x13e/0x36c [ 10.167652] rt6_dump_route+0x4c/0x56 [ 10.168276] fib6_dump_node+0x1d/0x3d [ 10.168913] fib6_walk_continue+0xab/0x167 [ 10.169611] fib6_walk+0x2a/0x40 [ 10.170182] inet6_dump_fib+0xfb/0x1e0 [ 10.170855] netlink_dump+0xcd/0x21f This happens when the loopback device is set down and a ipv6 fib route dump is requested. ip6_null_entry is the root of all ipv6 fib tables making it integrated into the table and hence passed to the ipv6 route dump code. The null_entry route uses the loopback device for dst.dev but may not have rt6i_idev set because of the order in which initializations are done -- ip6_route_net_init is run before addrconf_init has initialized the loopback device. Fixing the initialization order is a much bigger problem with no obvious solution thus far. The BUG is triggered when the loopback is set down and the netif_running check added by a1a22c1206 fails. The fill_node descends to checking rt->rt6i_idev for ignore_routes_with_linkdown and since rt6i_idev is NULL it faults. The null_entry route should not be processed in a dump request. Catch and ignore. This check is done in rt6_dump_route as it is the highest place in the callchain with knowledge of both the route and the network namespace. Fixes: a1a22c1206("net: ipv6: Keep nexthop of multipath route on admin down") Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-27 05:54:08 +08:00
return 0;
if ((filter->flags & RTM_F_PREFIX) &&
!(rt->fib6_flags & RTF_PREFIX_RT)) {
/* success since this is not a prefix route */
return 1;
}
if (filter->filter_set) {
if ((filter->rt_type && rt->fib6_type != filter->rt_type) ||
(filter->dev && !fib6_info_uses_dev(rt, filter->dev)) ||
(filter->protocol && rt->fib6_protocol != filter->protocol)) {
return 1;
}
flags |= NLM_F_DUMP_FILTERED;
}
return rt6_fill_node(net, arg->skb, rt, NULL, NULL, NULL, 0,
RTM_NEWROUTE, NETLINK_CB(arg->cb->skb).portid,
arg->cb->nlh->nlmsg_seq, flags);
}
static int inet6_rtm_valid_getroute_req(struct sk_buff *skb,
const struct nlmsghdr *nlh,
struct nlattr **tb,
struct netlink_ext_ack *extack)
{
struct rtmsg *rtm;
int i, err;
if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*rtm))) {
NL_SET_ERR_MSG_MOD(extack,
"Invalid header for get route request");
return -EINVAL;
}
if (!netlink_strict_get_check(skb))
return nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX,
rtm_ipv6_policy, extack);
rtm = nlmsg_data(nlh);
if ((rtm->rtm_src_len && rtm->rtm_src_len != 128) ||
(rtm->rtm_dst_len && rtm->rtm_dst_len != 128) ||
rtm->rtm_table || rtm->rtm_protocol || rtm->rtm_scope ||
rtm->rtm_type) {
NL_SET_ERR_MSG_MOD(extack, "Invalid values in header for get route request");
return -EINVAL;
}
if (rtm->rtm_flags & ~RTM_F_FIB_MATCH) {
NL_SET_ERR_MSG_MOD(extack,
"Invalid flags for get route request");
return -EINVAL;
}
err = nlmsg_parse_strict(nlh, sizeof(*rtm), tb, RTA_MAX,
rtm_ipv6_policy, extack);
if (err)
return err;
if ((tb[RTA_SRC] && !rtm->rtm_src_len) ||
(tb[RTA_DST] && !rtm->rtm_dst_len)) {
NL_SET_ERR_MSG_MOD(extack, "rtm_src_len and rtm_dst_len must be 128 for IPv6");
return -EINVAL;
}
for (i = 0; i <= RTA_MAX; i++) {
if (!tb[i])
continue;
switch (i) {
case RTA_SRC:
case RTA_DST:
case RTA_IIF:
case RTA_OIF:
case RTA_MARK:
case RTA_UID:
case RTA_SPORT:
case RTA_DPORT:
case RTA_IP_PROTO:
break;
default:
NL_SET_ERR_MSG_MOD(extack, "Unsupported attribute in get route request");
return -EINVAL;
}
}
return 0;
}
static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh,
struct netlink_ext_ack *extack)
{
struct net *net = sock_net(in_skb->sk);
struct nlattr *tb[RTA_MAX+1];
int err, iif = 0, oif = 0;
struct fib6_info *from;
struct dst_entry *dst;
struct rt6_info *rt;
struct sk_buff *skb;
struct rtmsg *rtm;
struct flowi6 fl6 = {};
bool fibmatch;
err = inet6_rtm_valid_getroute_req(in_skb, nlh, tb, extack);
if (err < 0)
goto errout;
err = -EINVAL;
rtm = nlmsg_data(nlh);
fl6.flowlabel = ip6_make_flowinfo(rtm->rtm_tos, 0);
fibmatch = !!(rtm->rtm_flags & RTM_F_FIB_MATCH);
if (tb[RTA_SRC]) {
if (nla_len(tb[RTA_SRC]) < sizeof(struct in6_addr))
goto errout;
fl6.saddr = *(struct in6_addr *)nla_data(tb[RTA_SRC]);
}
if (tb[RTA_DST]) {
if (nla_len(tb[RTA_DST]) < sizeof(struct in6_addr))
goto errout;
fl6.daddr = *(struct in6_addr *)nla_data(tb[RTA_DST]);
}
if (tb[RTA_IIF])
iif = nla_get_u32(tb[RTA_IIF]);
if (tb[RTA_OIF])
oif = nla_get_u32(tb[RTA_OIF]);
if (tb[RTA_MARK])
fl6.flowi6_mark = nla_get_u32(tb[RTA_MARK]);
if (tb[RTA_UID])
fl6.flowi6_uid = make_kuid(current_user_ns(),
nla_get_u32(tb[RTA_UID]));
else
fl6.flowi6_uid = iif ? INVALID_UID : current_uid();
if (tb[RTA_SPORT])
fl6.fl6_sport = nla_get_be16(tb[RTA_SPORT]);
if (tb[RTA_DPORT])
fl6.fl6_dport = nla_get_be16(tb[RTA_DPORT]);
if (tb[RTA_IP_PROTO]) {
err = rtm_getroute_parse_ip_proto(tb[RTA_IP_PROTO],
&fl6.flowi6_proto, extack);
if (err)
goto errout;
}
if (iif) {
struct net_device *dev;
int flags = 0;
rcu_read_lock();
dev = dev_get_by_index_rcu(net, iif);
if (!dev) {
rcu_read_unlock();
err = -ENODEV;
goto errout;
}
fl6.flowi6_iif = iif;
if (!ipv6_addr_any(&fl6.saddr))
flags |= RT6_LOOKUP_F_HAS_SADDR;
dst = ip6_route_input_lookup(net, dev, &fl6, NULL, flags);
rcu_read_unlock();
} else {
fl6.flowi6_oif = oif;
ipv6: Honor specified parameters in fibmatch lookup Currently, parameters such as oif and source address are not taken into account during fibmatch lookup. Example (IPv4 for reference) before patch: $ ip -4 route show 192.0.2.0/24 dev dummy0 proto kernel scope link src 192.0.2.1 198.51.100.0/24 dev dummy1 proto kernel scope link src 198.51.100.1 $ ip -6 route show 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium 2001:db8:2::/64 dev dummy1 proto kernel metric 256 pref medium fe80::/64 dev dummy0 proto kernel metric 256 pref medium fe80::/64 dev dummy1 proto kernel metric 256 pref medium $ ip -4 route get fibmatch 192.0.2.2 oif dummy0 192.0.2.0/24 dev dummy0 proto kernel scope link src 192.0.2.1 $ ip -4 route get fibmatch 192.0.2.2 oif dummy1 RTNETLINK answers: No route to host $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy0 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy1 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium After: $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy0 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy1 RTNETLINK answers: Network is unreachable The problem stems from the fact that the necessary route lookup flags are not set based on these parameters. Instead of duplicating the same logic for fibmatch, we can simply resolve the original route from its copy and dump it instead. Fixes: 18c3a61c4264 ("net: ipv6: RTM_GETROUTE: return matched fib result when requested") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-12-20 18:28:25 +08:00
dst = ip6_route_output(net, NULL, &fl6);
}
rt = container_of(dst, struct rt6_info, dst);
if (rt->dst.error) {
err = rt->dst.error;
ip6_rt_put(rt);
goto errout;
}
if (rt == net->ipv6.ip6_null_entry) {
err = rt->dst.error;
ip6_rt_put(rt);
goto errout;
}
skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
if (!skb) {
ip6_rt_put(rt);
err = -ENOBUFS;
goto errout;
}
skb_dst_set(skb, &rt->dst);
rcu_read_lock();
from = rcu_dereference(rt->from);
if (fibmatch)
err = rt6_fill_node(net, skb, from, NULL, NULL, NULL, iif,
RTM_NEWROUTE, NETLINK_CB(in_skb).portid,
nlh->nlmsg_seq, 0);
else
err = rt6_fill_node(net, skb, from, dst, &fl6.daddr,
&fl6.saddr, iif, RTM_NEWROUTE,
NETLINK_CB(in_skb).portid, nlh->nlmsg_seq,
0);
rcu_read_unlock();
if (err < 0) {
kfree_skb(skb);
goto errout;
}
err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
errout:
return err;
}
void inet6_rt_notify(int event, struct fib6_info *rt, struct nl_info *info,
unsigned int nlm_flags)
{
struct sk_buff *skb;
struct net *net = info->nl_net;
u32 seq;
int err;
err = -ENOBUFS;
seq = info->nlh ? info->nlh->nlmsg_seq : 0;
skb = nlmsg_new(rt6_nlmsg_size(rt), gfp_any());
if (!skb)
goto errout;
err = rt6_fill_node(net, skb, rt, NULL, NULL, NULL, 0,
event, info->portid, seq, nlm_flags);
if (err < 0) {
/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
WARN_ON(err == -EMSGSIZE);
kfree_skb(skb);
goto errout;
}
rtnl_notify(skb, net, info->portid, RTNLGRP_IPV6_ROUTE,
2009-02-25 15:18:28 +08:00
info->nlh, gfp_any());
return;
errout:
if (err < 0)
rtnl_set_sk_err(net, RTNLGRP_IPV6_ROUTE, err);
}
static int ip6_route_dev_notify(struct notifier_block *this,
unsigned long event, void *ptr)
{
struct net_device *dev = netdev_notifier_info_to_dev(ptr);
struct net *net = dev_net(dev);
if (!(dev->flags & IFF_LOOPBACK))
return NOTIFY_OK;
if (event == NETDEV_REGISTER) {
net->ipv6.fib6_null_entry->fib6_nh.nh_dev = dev;
net->ipv6.ip6_null_entry->dst.dev = dev;
net->ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(dev);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
net->ipv6.ip6_prohibit_entry->dst.dev = dev;
net->ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(dev);
net->ipv6.ip6_blk_hole_entry->dst.dev = dev;
net->ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(dev);
#endif
} else if (event == NETDEV_UNREGISTER &&
dev->reg_state != NETREG_UNREGISTERED) {
/* NETDEV_UNREGISTER could be fired for multiple times by
* netdev_wait_allrefs(). Make sure we only call this once.
*/
ipv6: fix NULL dereference in ip6_route_dev_notify() Based on a syzkaller report [1], I found that a per cpu allocation failure in snmp6_alloc_dev() would then lead to NULL dereference in ip6_route_dev_notify(). It seems this is a very old bug, thus no Fixes tag in this submission. Let's add in6_dev_put_clear() helper, as we will probably use it elsewhere (once available/present in net-next) [1] kasan: CONFIG_KASAN_INLINE enabled kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 17294 Comm: syz-executor6 Not tainted 4.13.0-rc2+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 task: ffff88019f456680 task.stack: ffff8801c6e58000 RIP: 0010:__read_once_size include/linux/compiler.h:250 [inline] RIP: 0010:atomic_read arch/x86/include/asm/atomic.h:26 [inline] RIP: 0010:refcount_sub_and_test+0x7d/0x1b0 lib/refcount.c:178 RSP: 0018:ffff8801c6e5f1b0 EFLAGS: 00010202 RAX: 0000000000000037 RBX: dffffc0000000000 RCX: ffffc90005d25000 RDX: ffff8801c6e5f218 RSI: ffffffff82342bbf RDI: 0000000000000001 RBP: ffff8801c6e5f240 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff10038dcbe37 R13: 0000000000000006 R14: 0000000000000001 R15: 00000000000001b8 FS: 00007f21e0429700(0000) GS:ffff8801dc100000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000001ddbc22000 CR3: 00000001d632b000 CR4: 00000000001426e0 DR0: 0000000020000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000600 Call Trace: refcount_dec_and_test+0x1a/0x20 lib/refcount.c:211 in6_dev_put include/net/addrconf.h:335 [inline] ip6_route_dev_notify+0x1c9/0x4a0 net/ipv6/route.c:3732 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x51/0x90 net/core/dev.c:1678 call_netdevice_notifiers net/core/dev.c:1694 [inline] rollback_registered_many+0x91c/0xe80 net/core/dev.c:7107 rollback_registered+0x1be/0x3c0 net/core/dev.c:7149 register_netdevice+0xbcd/0xee0 net/core/dev.c:7587 register_netdev+0x1a/0x30 net/core/dev.c:7669 loopback_net_init+0x76/0x160 drivers/net/loopback.c:214 ops_init+0x10a/0x570 net/core/net_namespace.c:118 setup_net+0x313/0x710 net/core/net_namespace.c:294 copy_net_ns+0x27c/0x580 net/core/net_namespace.c:418 create_new_namespaces+0x425/0x880 kernel/nsproxy.c:107 unshare_nsproxy_namespaces+0xae/0x1e0 kernel/nsproxy.c:206 SYSC_unshare kernel/fork.c:2347 [inline] SyS_unshare+0x653/0xfa0 kernel/fork.c:2297 entry_SYSCALL_64_fastpath+0x1f/0xbe RIP: 0033:0x4512c9 RSP: 002b:00007f21e0428c08 EFLAGS: 00000216 ORIG_RAX: 0000000000000110 RAX: ffffffffffffffda RBX: 0000000000718150 RCX: 00000000004512c9 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000062020200 RBP: 0000000000000086 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000216 R12: 00000000004b973d R13: 00000000ffffffff R14: 000000002001d000 R15: 00000000000002dd Code: 50 2b 34 82 c7 00 f1 f1 f1 f1 c7 40 04 04 f2 f2 f2 c7 40 08 f3 f3 f3 f3 e8 a1 43 39 ff 4c 89 f8 48 8b 95 70 ff ff ff 48 c1 e8 03 <0f> b6 0c 18 4c 89 f8 83 e0 07 83 c0 03 38 c8 7c 08 84 c9 0f 85 RIP: __read_once_size include/linux/compiler.h:250 [inline] RSP: ffff8801c6e5f1b0 RIP: atomic_read arch/x86/include/asm/atomic.h:26 [inline] RSP: ffff8801c6e5f1b0 RIP: refcount_sub_and_test+0x7d/0x1b0 lib/refcount.c:178 RSP: ffff8801c6e5f1b0 ---[ end trace e441d046c6410d31 ]--- Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-15 19:09:51 +08:00
in6_dev_put_clear(&net->ipv6.ip6_null_entry->rt6i_idev);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
ipv6: fix NULL dereference in ip6_route_dev_notify() Based on a syzkaller report [1], I found that a per cpu allocation failure in snmp6_alloc_dev() would then lead to NULL dereference in ip6_route_dev_notify(). It seems this is a very old bug, thus no Fixes tag in this submission. Let's add in6_dev_put_clear() helper, as we will probably use it elsewhere (once available/present in net-next) [1] kasan: CONFIG_KASAN_INLINE enabled kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 17294 Comm: syz-executor6 Not tainted 4.13.0-rc2+ #10 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 task: ffff88019f456680 task.stack: ffff8801c6e58000 RIP: 0010:__read_once_size include/linux/compiler.h:250 [inline] RIP: 0010:atomic_read arch/x86/include/asm/atomic.h:26 [inline] RIP: 0010:refcount_sub_and_test+0x7d/0x1b0 lib/refcount.c:178 RSP: 0018:ffff8801c6e5f1b0 EFLAGS: 00010202 RAX: 0000000000000037 RBX: dffffc0000000000 RCX: ffffc90005d25000 RDX: ffff8801c6e5f218 RSI: ffffffff82342bbf RDI: 0000000000000001 RBP: ffff8801c6e5f240 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff10038dcbe37 R13: 0000000000000006 R14: 0000000000000001 R15: 00000000000001b8 FS: 00007f21e0429700(0000) GS:ffff8801dc100000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000001ddbc22000 CR3: 00000001d632b000 CR4: 00000000001426e0 DR0: 0000000020000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000600 Call Trace: refcount_dec_and_test+0x1a/0x20 lib/refcount.c:211 in6_dev_put include/net/addrconf.h:335 [inline] ip6_route_dev_notify+0x1c9/0x4a0 net/ipv6/route.c:3732 notifier_call_chain+0x136/0x2c0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 [inline] raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x51/0x90 net/core/dev.c:1678 call_netdevice_notifiers net/core/dev.c:1694 [inline] rollback_registered_many+0x91c/0xe80 net/core/dev.c:7107 rollback_registered+0x1be/0x3c0 net/core/dev.c:7149 register_netdevice+0xbcd/0xee0 net/core/dev.c:7587 register_netdev+0x1a/0x30 net/core/dev.c:7669 loopback_net_init+0x76/0x160 drivers/net/loopback.c:214 ops_init+0x10a/0x570 net/core/net_namespace.c:118 setup_net+0x313/0x710 net/core/net_namespace.c:294 copy_net_ns+0x27c/0x580 net/core/net_namespace.c:418 create_new_namespaces+0x425/0x880 kernel/nsproxy.c:107 unshare_nsproxy_namespaces+0xae/0x1e0 kernel/nsproxy.c:206 SYSC_unshare kernel/fork.c:2347 [inline] SyS_unshare+0x653/0xfa0 kernel/fork.c:2297 entry_SYSCALL_64_fastpath+0x1f/0xbe RIP: 0033:0x4512c9 RSP: 002b:00007f21e0428c08 EFLAGS: 00000216 ORIG_RAX: 0000000000000110 RAX: ffffffffffffffda RBX: 0000000000718150 RCX: 00000000004512c9 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000062020200 RBP: 0000000000000086 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000216 R12: 00000000004b973d R13: 00000000ffffffff R14: 000000002001d000 R15: 00000000000002dd Code: 50 2b 34 82 c7 00 f1 f1 f1 f1 c7 40 04 04 f2 f2 f2 c7 40 08 f3 f3 f3 f3 e8 a1 43 39 ff 4c 89 f8 48 8b 95 70 ff ff ff 48 c1 e8 03 <0f> b6 0c 18 4c 89 f8 83 e0 07 83 c0 03 38 c8 7c 08 84 c9 0f 85 RIP: __read_once_size include/linux/compiler.h:250 [inline] RSP: ffff8801c6e5f1b0 RIP: atomic_read arch/x86/include/asm/atomic.h:26 [inline] RSP: ffff8801c6e5f1b0 RIP: refcount_sub_and_test+0x7d/0x1b0 lib/refcount.c:178 RSP: ffff8801c6e5f1b0 ---[ end trace e441d046c6410d31 ]--- Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-15 19:09:51 +08:00
in6_dev_put_clear(&net->ipv6.ip6_prohibit_entry->rt6i_idev);
in6_dev_put_clear(&net->ipv6.ip6_blk_hole_entry->rt6i_idev);
#endif
}
return NOTIFY_OK;
}
/*
* /proc
*/
#ifdef CONFIG_PROC_FS
static int rt6_stats_seq_show(struct seq_file *seq, void *v)
{
struct net *net = (struct net *)seq->private;
seq_printf(seq, "%04x %04x %04x %04x %04x %04x %04x\n",
net->ipv6.rt6_stats->fib_nodes,
net->ipv6.rt6_stats->fib_route_nodes,
atomic_read(&net->ipv6.rt6_stats->fib_rt_alloc),
net->ipv6.rt6_stats->fib_rt_entries,
net->ipv6.rt6_stats->fib_rt_cache,
dst_entries_get_slow(&net->ipv6.ip6_dst_ops),
net->ipv6.rt6_stats->fib_discarded_routes);
return 0;
}
#endif /* CONFIG_PROC_FS */
#ifdef CONFIG_SYSCTL
static
int ipv6_sysctl_rtcache_flush(struct ctl_table *ctl, int write,
void __user *buffer, size_t *lenp, loff_t *ppos)
{
struct net *net;
int delay;
int ret;
if (!write)
return -EINVAL;
net = (struct net *)ctl->extra1;
delay = net->ipv6.sysctl.flush_delay;
ret = proc_dointvec(ctl, write, buffer, lenp, ppos);
if (ret)
return ret;
fib6_run_gc(delay <= 0 ? 0 : (unsigned long)delay, net, delay > 0);
return 0;
}
static int zero;
static int one = 1;
static struct ctl_table ipv6_route_table_template[] = {
{
.procname = "flush",
.data = &init_net.ipv6.sysctl.flush_delay,
.maxlen = sizeof(int),
.mode = 0200,
.proc_handler = ipv6_sysctl_rtcache_flush
},
{
.procname = "gc_thresh",
.data = &ip6_dst_ops_template.gc_thresh,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec,
},
{
.procname = "max_size",
.data = &init_net.ipv6.sysctl.ip6_rt_max_size,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec,
},
{
.procname = "gc_min_interval",
.data = &init_net.ipv6.sysctl.ip6_rt_gc_min_interval,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec_jiffies,
},
{
.procname = "gc_timeout",
.data = &init_net.ipv6.sysctl.ip6_rt_gc_timeout,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec_jiffies,
},
{
.procname = "gc_interval",
.data = &init_net.ipv6.sysctl.ip6_rt_gc_interval,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec_jiffies,
},
{
.procname = "gc_elasticity",
.data = &init_net.ipv6.sysctl.ip6_rt_gc_elasticity,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec,
},
{
.procname = "mtu_expires",
.data = &init_net.ipv6.sysctl.ip6_rt_mtu_expires,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec_jiffies,
},
{
.procname = "min_adv_mss",
.data = &init_net.ipv6.sysctl.ip6_rt_min_advmss,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec,
},
{
.procname = "gc_min_interval_ms",
.data = &init_net.ipv6.sysctl.ip6_rt_gc_min_interval,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec_ms_jiffies,
},
{
.procname = "skip_notify_on_dev_down",
.data = &init_net.ipv6.sysctl.skip_notify_on_dev_down,
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec,
.extra1 = &zero,
.extra2 = &one,
},
{ }
};
struct ctl_table * __net_init ipv6_route_sysctl_init(struct net *net)
{
struct ctl_table *table;
table = kmemdup(ipv6_route_table_template,
sizeof(ipv6_route_table_template),
GFP_KERNEL);
if (table) {
table[0].data = &net->ipv6.sysctl.flush_delay;
table[0].extra1 = net;
table[1].data = &net->ipv6.ip6_dst_ops.gc_thresh;
table[2].data = &net->ipv6.sysctl.ip6_rt_max_size;
table[3].data = &net->ipv6.sysctl.ip6_rt_gc_min_interval;
table[4].data = &net->ipv6.sysctl.ip6_rt_gc_timeout;
table[5].data = &net->ipv6.sysctl.ip6_rt_gc_interval;
table[6].data = &net->ipv6.sysctl.ip6_rt_gc_elasticity;
table[7].data = &net->ipv6.sysctl.ip6_rt_mtu_expires;
table[8].data = &net->ipv6.sysctl.ip6_rt_min_advmss;
table[9].data = &net->ipv6.sysctl.ip6_rt_gc_min_interval;
table[10].data = &net->ipv6.sysctl.skip_notify_on_dev_down;
/* Don't export sysctls to unprivileged users */
if (net->user_ns != &init_user_ns)
table[0].procname = NULL;
}
return table;
}
#endif
static int __net_init ip6_route_net_init(struct net *net)
{
int ret = -ENOMEM;
memcpy(&net->ipv6.ip6_dst_ops, &ip6_dst_ops_template,
sizeof(net->ipv6.ip6_dst_ops));
if (dst_entries_init(&net->ipv6.ip6_dst_ops) < 0)
goto out_ip6_dst_ops;
net->ipv6.fib6_null_entry = kmemdup(&fib6_null_entry_template,
sizeof(*net->ipv6.fib6_null_entry),
GFP_KERNEL);
if (!net->ipv6.fib6_null_entry)
goto out_ip6_dst_entries;
net->ipv6.ip6_null_entry = kmemdup(&ip6_null_entry_template,
sizeof(*net->ipv6.ip6_null_entry),
GFP_KERNEL);
if (!net->ipv6.ip6_null_entry)
goto out_fib6_null_entry;
net->ipv6.ip6_null_entry->dst.ops = &net->ipv6.ip6_dst_ops;
net: Implement read-only protection and COW'ing of metrics. Routing metrics are now copy-on-write. Initially a route entry points it's metrics at a read-only location. If a routing table entry exists, it will point there. Else it will point at the all zero metric place-holder called 'dst_default_metrics'. The writeability state of the metrics is stored in the low bits of the metrics pointer, we have two bits left to spare if we want to store more states. For the initial implementation, COW is implemented simply via kmalloc. However future enhancements will change this to place the writable metrics somewhere else, in order to increase sharing. Very likely this "somewhere else" will be the inetpeer cache. Note also that this means that metrics updates may transiently fail if we cannot COW the metrics successfully. But even by itself, this patch should decrease memory usage and increase cache locality especially for routing workloads. In those cases the read-only metric copies stay in place and never get written to. TCP workloads where metrics get updated, and those rare cases where PMTU triggers occur, will take a very slight performance hit. But that hit will be alleviated when the long-term writable metrics move to a more sharable location. Since the metrics storage went from a u32 array of RTAX_MAX entries to what is essentially a pointer, some retooling of the dst_entry layout was necessary. Most importantly, we need to preserve the alignment of the reference count so that it doesn't share cache lines with the read-mostly state, as per Eric Dumazet's alignment assertion checks. The only non-trivial bit here is the move of the 'flags' member into the writeable cacheline. This is OK since we are always accessing the flags around the same moment when we made a modification to the reference count. Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-27 12:51:05 +08:00
dst_init_metrics(&net->ipv6.ip6_null_entry->dst,
ip6_template_metrics, true);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
net: ipv6: avoid overhead when no custom FIB rules are installed If the user hasn't installed any custom rules, don't go through the whole FIB rules layer. This is pretty similar to f4530fa574df (ipv4: Avoid overhead when no custom FIB rules are installed). Using a micro-benchmark module [1], timing ip6_route_output() with get_cycles(), with 40,000 routes in the main routing table, before this patch: min=606 max=12911 count=627 average=1959 95th=4903 90th=3747 50th=1602 mad=821 table=254 avgdepth=21.8 maxdepth=39 value │ ┊ count 600 │▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ 199 880 │▒▒▒░░░░░░░░░░░░░░░░ 43 1160 │▒▒▒░░░░░░░░░░░░░░░░░░░░ 48 1440 │▒▒▒░░░░░░░░░░░░░░░░░░░░░░░ 43 1720 │▒▒▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░ 59 2000 │▒▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 50 2280 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 26 2560 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 31 2840 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 28 3120 │▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 17 3400 │▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 17 3680 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 8 3960 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 11 4240 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 6 4520 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 6 4800 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 9 After: min=544 max=11687 count=627 average=1776 95th=4546 90th=3585 50th=1227 mad=565 table=254 avgdepth=21.8 maxdepth=39 value │ ┊ count 540 │▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ 201 800 │▒▒▒▒▒░░░░░░░░░░░░░░░░ 63 1060 │▒▒▒▒▒░░░░░░░░░░░░░░░░░░░░░ 68 1320 │▒▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░ 39 1580 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 32 1840 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 32 2100 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 34 2360 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 33 2620 │▒▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 26 2880 │▒░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 22 3140 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 9 3400 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 8 3660 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 9 3920 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 8 4180 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 8 4440 │░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 8 At the frequency of the host during the bench (~ 3.7 GHz), this is about a 100 ns difference on the median value. A next step would be to collapse local and main tables, as in 0ddcf43d5d4a (ipv4: FIB Local/MAIN table collapse). [1]: https://github.com/vincentbernat/network-lab/blob/master/lab-routes-ipv6/kbench_mod.c Signed-off-by: Vincent Bernat <vincent@bernat.im> Reviewed-by: Jiri Pirko <jiri@mellanox.com> Acked-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-09 02:23:49 +08:00
net->ipv6.fib6_has_custom_rules = false;
net->ipv6.ip6_prohibit_entry = kmemdup(&ip6_prohibit_entry_template,
sizeof(*net->ipv6.ip6_prohibit_entry),
GFP_KERNEL);
if (!net->ipv6.ip6_prohibit_entry)
goto out_ip6_null_entry;
net->ipv6.ip6_prohibit_entry->dst.ops = &net->ipv6.ip6_dst_ops;
net: Implement read-only protection and COW'ing of metrics. Routing metrics are now copy-on-write. Initially a route entry points it's metrics at a read-only location. If a routing table entry exists, it will point there. Else it will point at the all zero metric place-holder called 'dst_default_metrics'. The writeability state of the metrics is stored in the low bits of the metrics pointer, we have two bits left to spare if we want to store more states. For the initial implementation, COW is implemented simply via kmalloc. However future enhancements will change this to place the writable metrics somewhere else, in order to increase sharing. Very likely this "somewhere else" will be the inetpeer cache. Note also that this means that metrics updates may transiently fail if we cannot COW the metrics successfully. But even by itself, this patch should decrease memory usage and increase cache locality especially for routing workloads. In those cases the read-only metric copies stay in place and never get written to. TCP workloads where metrics get updated, and those rare cases where PMTU triggers occur, will take a very slight performance hit. But that hit will be alleviated when the long-term writable metrics move to a more sharable location. Since the metrics storage went from a u32 array of RTAX_MAX entries to what is essentially a pointer, some retooling of the dst_entry layout was necessary. Most importantly, we need to preserve the alignment of the reference count so that it doesn't share cache lines with the read-mostly state, as per Eric Dumazet's alignment assertion checks. The only non-trivial bit here is the move of the 'flags' member into the writeable cacheline. This is OK since we are always accessing the flags around the same moment when we made a modification to the reference count. Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-27 12:51:05 +08:00
dst_init_metrics(&net->ipv6.ip6_prohibit_entry->dst,
ip6_template_metrics, true);
net->ipv6.ip6_blk_hole_entry = kmemdup(&ip6_blk_hole_entry_template,
sizeof(*net->ipv6.ip6_blk_hole_entry),
GFP_KERNEL);
if (!net->ipv6.ip6_blk_hole_entry)
goto out_ip6_prohibit_entry;
net->ipv6.ip6_blk_hole_entry->dst.ops = &net->ipv6.ip6_dst_ops;
net: Implement read-only protection and COW'ing of metrics. Routing metrics are now copy-on-write. Initially a route entry points it's metrics at a read-only location. If a routing table entry exists, it will point there. Else it will point at the all zero metric place-holder called 'dst_default_metrics'. The writeability state of the metrics is stored in the low bits of the metrics pointer, we have two bits left to spare if we want to store more states. For the initial implementation, COW is implemented simply via kmalloc. However future enhancements will change this to place the writable metrics somewhere else, in order to increase sharing. Very likely this "somewhere else" will be the inetpeer cache. Note also that this means that metrics updates may transiently fail if we cannot COW the metrics successfully. But even by itself, this patch should decrease memory usage and increase cache locality especially for routing workloads. In those cases the read-only metric copies stay in place and never get written to. TCP workloads where metrics get updated, and those rare cases where PMTU triggers occur, will take a very slight performance hit. But that hit will be alleviated when the long-term writable metrics move to a more sharable location. Since the metrics storage went from a u32 array of RTAX_MAX entries to what is essentially a pointer, some retooling of the dst_entry layout was necessary. Most importantly, we need to preserve the alignment of the reference count so that it doesn't share cache lines with the read-mostly state, as per Eric Dumazet's alignment assertion checks. The only non-trivial bit here is the move of the 'flags' member into the writeable cacheline. This is OK since we are always accessing the flags around the same moment when we made a modification to the reference count. Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-27 12:51:05 +08:00
dst_init_metrics(&net->ipv6.ip6_blk_hole_entry->dst,
ip6_template_metrics, true);
#endif
net->ipv6.sysctl.flush_delay = 0;
net->ipv6.sysctl.ip6_rt_max_size = 4096;
net->ipv6.sysctl.ip6_rt_gc_min_interval = HZ / 2;
net->ipv6.sysctl.ip6_rt_gc_timeout = 60*HZ;
net->ipv6.sysctl.ip6_rt_gc_interval = 30*HZ;
net->ipv6.sysctl.ip6_rt_gc_elasticity = 9;
net->ipv6.sysctl.ip6_rt_mtu_expires = 10*60*HZ;
net->ipv6.sysctl.ip6_rt_min_advmss = IPV6_MIN_MTU - 20 - 40;
net->ipv6.sysctl.skip_notify_on_dev_down = 0;
net->ipv6.ip6_rt_gc_expire = 30*HZ;
ret = 0;
out:
return ret;
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
out_ip6_prohibit_entry:
kfree(net->ipv6.ip6_prohibit_entry);
out_ip6_null_entry:
kfree(net->ipv6.ip6_null_entry);
#endif
out_fib6_null_entry:
kfree(net->ipv6.fib6_null_entry);
out_ip6_dst_entries:
dst_entries_destroy(&net->ipv6.ip6_dst_ops);
out_ip6_dst_ops:
goto out;
}
static void __net_exit ip6_route_net_exit(struct net *net)
{
kfree(net->ipv6.fib6_null_entry);
kfree(net->ipv6.ip6_null_entry);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
kfree(net->ipv6.ip6_prohibit_entry);
kfree(net->ipv6.ip6_blk_hole_entry);
#endif
dst_entries_destroy(&net->ipv6.ip6_dst_ops);
}
static int __net_init ip6_route_net_init_late(struct net *net)
{
#ifdef CONFIG_PROC_FS
proc_create_net("ipv6_route", 0, net->proc_net, &ipv6_route_seq_ops,
sizeof(struct ipv6_route_iter));
proc_create_net_single("rt6_stats", 0444, net->proc_net,
rt6_stats_seq_show, NULL);
#endif
return 0;
}
static void __net_exit ip6_route_net_exit_late(struct net *net)
{
#ifdef CONFIG_PROC_FS
remove_proc_entry("ipv6_route", net->proc_net);
remove_proc_entry("rt6_stats", net->proc_net);
#endif
}
static struct pernet_operations ip6_route_net_ops = {
.init = ip6_route_net_init,
.exit = ip6_route_net_exit,
};
static int __net_init ipv6_inetpeer_init(struct net *net)
{
struct inet_peer_base *bp = kmalloc(sizeof(*bp), GFP_KERNEL);
if (!bp)
return -ENOMEM;
inet_peer_base_init(bp);
net->ipv6.peers = bp;
return 0;
}
static void __net_exit ipv6_inetpeer_exit(struct net *net)
{
struct inet_peer_base *bp = net->ipv6.peers;
net->ipv6.peers = NULL;
inetpeer_invalidate_tree(bp);
kfree(bp);
}
static struct pernet_operations ipv6_inetpeer_ops = {
.init = ipv6_inetpeer_init,
.exit = ipv6_inetpeer_exit,
};
static struct pernet_operations ip6_route_net_late_ops = {
.init = ip6_route_net_init_late,
.exit = ip6_route_net_exit_late,
};
static struct notifier_block ip6_route_dev_notifier = {
.notifier_call = ip6_route_dev_notify,
.priority = ADDRCONF_NOTIFY_PRIORITY - 10,
};
void __init ip6_route_init_special_entries(void)
{
/* Registering of the loopback is done before this portion of code,
* the loopback reference in rt6_info will not be taken, do it
* manually for init_net */
init_net.ipv6.fib6_null_entry->fib6_nh.nh_dev = init_net.loopback_dev;
init_net.ipv6.ip6_null_entry->dst.dev = init_net.loopback_dev;
init_net.ipv6.ip6_null_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
init_net.ipv6.ip6_prohibit_entry->dst.dev = init_net.loopback_dev;
init_net.ipv6.ip6_prohibit_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
init_net.ipv6.ip6_blk_hole_entry->dst.dev = init_net.loopback_dev;
init_net.ipv6.ip6_blk_hole_entry->rt6i_idev = in6_dev_get(init_net.loopback_dev);
#endif
}
int __init ip6_route_init(void)
{
int ret;
int cpu;
ret = -ENOMEM;
ip6_dst_ops_template.kmem_cachep =
kmem_cache_create("ip6_dst_cache", sizeof(struct rt6_info), 0,
SLAB_HWCACHE_ALIGN, NULL);
if (!ip6_dst_ops_template.kmem_cachep)
goto out;
ret = dst_entries_init(&ip6_dst_blackhole_ops);
if (ret)
goto out_kmem_cache;
ret = register_pernet_subsys(&ipv6_inetpeer_ops);
if (ret)
goto out_dst_entries;
ret = register_pernet_subsys(&ip6_route_net_ops);
if (ret)
goto out_register_inetpeer;
XFRM,IPv6: initialize ip6_dst_blackhole_ops.kmem_cachep ip6_dst_blackhole_ops.kmem_cachep is not expected to be NULL (i.e. to be initialized) when dst_alloc() is called from ip6_dst_blackhole(). Otherwise, it results in the following (xfrm_larval_drop is now set to 1 by default): [ 78.697642] Unable to handle kernel paging request for data at address 0x0000004c [ 78.703449] Faulting instruction address: 0xc0097f54 [ 78.786896] Oops: Kernel access of bad area, sig: 11 [#1] [ 78.792791] PowerMac [ 78.798383] Modules linked in: btusb usbhid bluetooth b43 mac80211 cfg80211 ehci_hcd ohci_hcd sungem sungem_phy usbcore ssb [ 78.804263] NIP: c0097f54 LR: c0334a28 CTR: c002d430 [ 78.809997] REGS: eef19ad0 TRAP: 0300 Not tainted (2.6.27-rc5) [ 78.815743] MSR: 00001032 <ME,IR,DR> CR: 22242482 XER: 20000000 [ 78.821550] DAR: 0000004c, DSISR: 40000000 [ 78.827278] TASK = eef0df40[3035] 'mip6d' THREAD: eef18000 [ 78.827408] GPR00: 00001032 eef19b80 eef0df40 00000000 00008020 eef19c30 00000001 00000000 [ 78.833249] GPR08: eee5101c c05a5c10 ef9ad500 00000000 24242422 1005787c 00000000 1004f960 [ 78.839151] GPR16: 00000000 10024e90 10050040 48030018 0fe44150 00000000 00000000 eef19c30 [ 78.845046] GPR24: eef19e44 00000000 eef19bf8 efb37c14 eef19bf8 00008020 00009032 c0596064 [ 78.856671] NIP [c0097f54] kmem_cache_alloc+0x20/0x94 [ 78.862581] LR [c0334a28] dst_alloc+0x40/0xc4 [ 78.868451] Call Trace: [ 78.874252] [eef19b80] [c03c1810] ip6_dst_lookup_tail+0x1c8/0x1dc (unreliable) [ 78.880222] [eef19ba0] [c0334a28] dst_alloc+0x40/0xc4 [ 78.886164] [eef19bb0] [c03cd698] ip6_dst_blackhole+0x28/0x1cc [ 78.892090] [eef19be0] [c03d9be8] rawv6_sendmsg+0x75c/0xc88 [ 78.897999] [eef19cb0] [c038bca4] inet_sendmsg+0x4c/0x78 [ 78.903907] [eef19cd0] [c03207c8] sock_sendmsg+0xac/0xe4 [ 78.909734] [eef19db0] [c03209e4] sys_sendmsg+0x1e4/0x2a0 [ 78.915540] [eef19f00] [c03220a8] sys_socketcall+0xfc/0x210 [ 78.921406] [eef19f40] [c0014b3c] ret_from_syscall+0x0/0x38 [ 78.927295] --- Exception: c01 at 0xfe2d730 [ 78.927297] LR = 0xfe2d71c [ 78.939019] Instruction dump: [ 78.944835] 91640018 9144001c 900a0000 4bffff44 9421ffe0 7c0802a6 bf810010 7c9d2378 [ 78.950694] 90010024 7fc000a6 57c0045e 7c000124 <83e3004c> 8383005c 2f9f0000 419e0050 [ 78.956464] ---[ end trace 05fa1ed7972487a1 ]--- As commented by Benjamin Thery, the bug was introduced by f2fc6a54585a1be6669613a31fbaba2ecbadcd36, while adding network namespaces support to ipv6 routes. Signed-off-by: Arnaud Ebalard <arno@natisbad.org> Acked-by: Benjamin Thery <benjamin.thery@bull.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-01 17:37:56 +08:00
ip6_dst_blackhole_ops.kmem_cachep = ip6_dst_ops_template.kmem_cachep;
ret = fib6_init();
if (ret)
goto out_register_subsys;
ret = xfrm6_init();
if (ret)
goto out_fib6_init;
ret = fib6_rules_init();
if (ret)
goto xfrm6_init;
ret = register_pernet_subsys(&ip6_route_net_late_ops);
if (ret)
goto fib6_rules_init;
ret = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_NEWROUTE,
inet6_rtm_newroute, NULL, 0);
if (ret < 0)
goto out_register_late_subsys;
ret = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_DELROUTE,
inet6_rtm_delroute, NULL, 0);
if (ret < 0)
goto out_register_late_subsys;
ret = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETROUTE,
inet6_rtm_getroute, NULL,
RTNL_FLAG_DOIT_UNLOCKED);
if (ret < 0)
goto out_register_late_subsys;
ret = register_netdevice_notifier(&ip6_route_dev_notifier);
if (ret)
goto out_register_late_subsys;
for_each_possible_cpu(cpu) {
struct uncached_list *ul = per_cpu_ptr(&rt6_uncached_list, cpu);
INIT_LIST_HEAD(&ul->head);
spin_lock_init(&ul->lock);
}
out:
return ret;
out_register_late_subsys:
rtnl_unregister_all(PF_INET6);
unregister_pernet_subsys(&ip6_route_net_late_ops);
fib6_rules_init:
fib6_rules_cleanup();
xfrm6_init:
xfrm6_fini();
out_fib6_init:
fib6_gc_cleanup();
out_register_subsys:
unregister_pernet_subsys(&ip6_route_net_ops);
out_register_inetpeer:
unregister_pernet_subsys(&ipv6_inetpeer_ops);
out_dst_entries:
dst_entries_destroy(&ip6_dst_blackhole_ops);
out_kmem_cache:
kmem_cache_destroy(ip6_dst_ops_template.kmem_cachep);
goto out;
}
void ip6_route_cleanup(void)
{
unregister_netdevice_notifier(&ip6_route_dev_notifier);
unregister_pernet_subsys(&ip6_route_net_late_ops);
fib6_rules_cleanup();
xfrm6_fini();
fib6_gc_cleanup();
unregister_pernet_subsys(&ipv6_inetpeer_ops);
unregister_pernet_subsys(&ip6_route_net_ops);
dst_entries_destroy(&ip6_dst_blackhole_ops);
kmem_cache_destroy(ip6_dst_ops_template.kmem_cachep);
}