This patch adds support for ECMP hash policy choice via a new sysctl
called fib_multipath_hash_policy and also adds support for L4 hashes.
The current values for fib_multipath_hash_policy are:
 0 - layer 3 (default)
 1 - layer 4
If there's an skb hash already set and it matches the chosen policy then it
will be used instead of being calculated (currently only for L4).
In L3 mode we always calculate the hash due to the ICMP error special
case, the flow dissector's field consistentification should handle the
address order thus we can remove the address reversals.
If the skb is provided we always use it for the hash calculation,
otherwise we fallback to fl4, that is if skb is NULL fl4 has to be set.
Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Whenever a FIB rule is added or removed, a notification is sent in the
FIB notification chain. However, listeners don't have a way to tell
which rule was added or removed.

This is problematic as we would like to give listeners the ability to
decide which action to execute based on the notified rule. Specifically,
offloading drivers should be able to determine if they support the
reflection of the notified FIB rule and flush their LPM tables in case
they don't.

Do that by adding a notifier info to these notifications and embed the
common FIB rule struct in it.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, when non-default (custom) FIB rules are used, devices capable
of layer 3 offloading flush their tables and let the kernel do the
forwarding instead.

When these devices' drivers are loaded they register to the FIB
notification chain, which lets them know about the existence of any
custom FIB rules. This is done by sending a RULE_ADD notification based
on the value of 'net->ipv4.fib_has_custom_rules'.

This approach is problematic when VRF offload is taken into account, as
upon the creation of the first VRF netdev, a l3mdev rule is programmed
to direct skbs to the VRF's table.

Instead of merely reading the above value and sending a single RULE_ADD
notification, we should iterate over all the FIB rules and send a
detailed notification for each, thereby allowing offloading drivers to
sanitize the rules they don't support and potentially flush their
tables.

While l3mdev rules are uniquely marked, the default rules are not.
Therefore, when they are being notified they might invoke offloading
drivers to unnecessarily flush their tables.

Solve this by adding an helper to check if a FIB rule is a default rule.
Namely, its selector should match all packets and its action should
point to the local, main or default tables.

As noted by David Ahern, uniquely marking the default rules is
insufficient. When using VRFs, it's common to avoid false hits by moving
the rule for the local table to just before the main table:

Default configuration:
$ ip rule show
0:      from all lookup local
32766:  from all lookup main
32767:  from all lookup default

Common configuration with VRFs:
$ ip rule show
1000:   from all lookup [l3mdev-table]
32765:  from all lookup local
32766:  from all lookup main
32767:  from all lookup default
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We always pass the same event type to fib_notify() and
fib_rules_notify(), so we can safely drop this argument.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Most of the code concerned with the FIB notification chain currently
resides in fib_trie.c, but this isn't really appropriate, as the FIB
notification chain is also used for FIB rules.

Therefore, it makes sense to move the common FIB notification code to a
separate file and have it export the relevant functions, which can be
invoked by its different users (e.g., fib_trie.c, fib_rules.c).
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The FIB notification chain currently uses the NLM_F_{REPLACE,APPEND}
flags to signal routes being replaced or appended.

Instead of using netlink flags for in-kernel notifications we can simply
introduce two new events in the FIB notification chain. This has the
added advantage of making the API cleaner, thereby making it clear that
these events should be supported by listeners of the notification chain.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
CC: Patrick McHardy <kaber@trash.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a multipath route is hit the kernel doesn't consider nexthops that
are DEAD or LINKDOWN when IN_DEV_IGNORE_ROUTES_WITH_LINKDOWN is set.
Devices that offload multipath routes need to be made aware of nexthop
status changes. Otherwise, the device will keep forwarding packets to
non-functional nexthops.

Add the FIB_EVENT_NH_{ADD,DEL} events to the fib notification chain,
which notify capable devices when they should add or delete a nexthop
from their tables.

Cc: Roopa Prabhu <roopa@cumulusnetworks.com>
Cc: David Ahern <dsa@cumulusnetworks.com>
Cc: Andy Gospodarek <andy@greyhouse.net>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Reviewed-by: NAndy Gospodarek <gospo@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

fib_select_default has a single caller within the same file.
Make it static.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit b90eb754 ("fib: introduce FIB notification infrastructure")
introduced a new notification chain to notify listeners (f.e., switchdev
drivers) about addition and deletion of routes.

However, upon registration to the chain the FIB tables can already be
populated, which means potential listeners will have an incomplete view
of the tables.

Solve that by dumping the FIB tables and replaying the events to the
passed notification block. The dump itself is done using RCU in order
not to starve consumers that need RTNL to make progress.

The integrity of the dump is ensured by reading the FIB change sequence
counter before and after the dump under RTNL. This allows us to avoid
the problematic situation in which the dumping process sends a ENTRY_ADD
notification following ENTRY_DEL generated by another process holding
RTNL.

Callers of the registration function may pass a callback that is
executed in case the dump was inconsistent with current FIB tables.

The number of retries until a consistent dump is achieved is set to a
fixed number to prevent callers from looping for long periods of time.
In case current limit proves to be problematic in the future, it can be
easily converted to be configurable using a sysctl.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As explained in the previous commit, modules are going to need to take a
reference on fib info and then drop it using fib_info_put().

Add the fib_info_hold() helper to make the code more readable and also
symmetric with fib_info_put().
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Suggested-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The patch that removed the FIB offload infrastructure was a bit too
aggressive and also removed code needed to clean up us splitting the table
if additional rules were added.  Specifically the function
fib_trie_flush_external was called at the end of a new rule being added to
flush the foreign trie entries from the main trie.

I updated the code so that we only call fib_trie_flush_external on the main
table so that we flush the entries for local from main.  This way we don't
call it for every rule change which is what was happening previously.

Fixes: 347e3b28 ("switchdev: remove FIB offload infrastructure")
Reported-by: NEric Dumazet <edumazet@google.com>
Cc: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since this is now taken care of by FIB notifier, remove the code, with
all unused dependencies.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

These helpers are to be used in case someone offloads the FIB entry. The
result is that if the entry is offloaded to at least one device, the
offload flag is set.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Reviewed-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This allows to pass information about added/deleted FIB entries/rules to
whoever is interested. This is done in a very similar way as devinet
notifies address additions/removals.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When deleting an IP address from an interface, there is a clean-up of
routes which refer to this local address. However, there was no check to
see that the VRF matched. This meant that deletion wasn't confined to
the VRF it should have been.

To solve this, a new field has been added to fib_info to hold a table
id. When removing fib entries corresponding to a local ip address, this
table id is also used in the comparison.

The table id is populated when the fib_info is created. This was already
done in some places, but not in ip_rt_ioctl(). This has now been fixed.

Fixes: 021dd3b8 ("net: Add routes to the table associated with the device")
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Tested-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NMark Tomlinson <mark.tomlinson@alliedtelesis.co.nz>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since the gc of ipv4 route was removed, the route cached would has
no chance to be removed, and even it has been timeout, it still could
be used, cause no code to check it's expires.

Fix this issue by checking  and removing route cache when we get route.
Signed-off-by: NXin Long <lucien.xin@gmail.com>
Acked-by: NHannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

fib_multipath_hash() computes a hash using __be32 values, force
cast these to u32 to pacify sparse.
Signed-off-by: NLance Richardson <lrichard@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When fib_netdev_event calls fib_disable_ip on NETDEV_DOWN event
we should not delete the local routes if the local address
is still present. The confusion comes from the fact that both
fib_netdev_event and fib_inetaddr_event use the NETDEV_DOWN
constant. Fix it by returning back the variable 'force'.

Steps to reproduce:
modprobe dummy
ifconfig dummy0 192.168.168.1 up
ifconfig dummy0 down
ip route list table local | grep dummy | grep host
local 192.168.168.1 dev dummy0  proto kernel  scope host  src 192.168.168.1

Fixes: 8a3d0316 ("net: track link-status of ipv4 nexthops")
Signed-off-by: NJulian Anastasov <ja@ssi.bg>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

VRF device needs the same path selection following lookup to set source
address. Rather than duplicating code, move existing code into a
function that is exported to modules.

Code move only; no functional change.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Replaces the per-packet multipath with a hash-based multipath using
source and destination address.
Signed-off-by: NPeter Nørlund <pch@ordbogen.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Man page of ip-route(8) says following about route types:

  unreachable - these destinations are unreachable.  Packets are dis‐
  carded and the ICMP message host unreachable is generated.  The local
  senders get an EHOSTUNREACH error.

  blackhole - these destinations are unreachable.  Packets are dis‐
  carded silently.  The local senders get an EINVAL error.

  prohibit - these destinations are unreachable.  Packets are discarded
  and the ICMP message communication administratively prohibited is
  generated.  The local senders get an EACCES error.

In the inet6 address family, this was correct, except the local senders
got ENETUNREACH error instead of EHOSTUNREACH in case of unreachable route.
In the inet address family, all three route types generated ICMP message
net unreachable, and the local senders got ENETUNREACH error.

In both address families all three route types now behave consistently
with documentation.
Signed-off-by: NNikola Forró <nforro@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

fib_select_default considers alternative routes only when
res->fi is for the first alias in res->fa_head. In the
common case this can happen only when the initial lookup
matches the first alias with highest TOS value. This
prevents the alternative routes to require specific TOS.

This patch solves the problem as follows:

- routes that require specific TOS should be returned by
fib_select_default only when TOS matches, as already done
in fib_table_lookup. This rule implies that depending on the
TOS we can have many different lists of alternative gateways
and we have to keep the last used gateway (fa_default) in first
alias for the TOS instead of using single tb_default value.

- as the aliases are ordered by many keys (TOS desc,
fib_priority asc), we restrict the possible results to
routes with matching TOS and lowest metric (fib_priority)
and routes that match any TOS, again with lowest metric.

For example, packet with TOS 8 can not use gw3 (not lowest
metric), gw4 (different TOS) and gw6 (not lowest metric),
all other gateways can be used:

tos 8 via gw1 metric 2 <--- res->fa_head and res->fi
tos 8 via gw2 metric 2
tos 8 via gw3 metric 3
tos 4 via gw4
tos 0 via gw5
tos 0 via gw6 metric 1
Reported-by: NHagen Paul Pfeifer <hagen@jauu.net>
Signed-off-by: NJulian Anastasov <ja@ssi.bg>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds support in ipv4 fib functions to parse user
provided encap attributes and attach encap state data to fib_nh
and rtable.
Signed-off-by: NRoopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This feature is only enabled with the new per-interface or ipv4 global
sysctls called 'ignore_routes_with_linkdown'.

net.ipv4.conf.all.ignore_routes_with_linkdown = 0
net.ipv4.conf.default.ignore_routes_with_linkdown = 0
net.ipv4.conf.lo.ignore_routes_with_linkdown = 0
...

When the above sysctls are set, will report to userspace that a route is
dead and will no longer resolve to this nexthop when performing a fib
lookup.  This will signal to userspace that the route will not be
selected.  The signalling of a RTNH_F_DEAD is only passed to userspace
if the sysctl is enabled and link is down.  This was done as without it
the netlink listeners would have no idea whether or not a nexthop would
be selected.   The kernel only sets RTNH_F_DEAD internally if the
interface has IFF_UP cleared.

With the new sysctl set, the following behavior can be observed
(interface p8p1 is link-down):

default via 10.0.5.2 dev p9p1
10.0.5.0/24 dev p9p1  proto kernel  scope link  src 10.0.5.15
70.0.0.0/24 dev p7p1  proto kernel  scope link  src 70.0.0.1
80.0.0.0/24 dev p8p1  proto kernel  scope link  src 80.0.0.1 dead linkdown
90.0.0.0/24 via 80.0.0.2 dev p8p1  metric 1 dead linkdown
90.0.0.0/24 via 70.0.0.2 dev p7p1  metric 2
90.0.0.1 via 70.0.0.2 dev p7p1  src 70.0.0.1
    cache
local 80.0.0.1 dev lo  src 80.0.0.1
    cache <local>
80.0.0.2 via 10.0.5.2 dev p9p1  src 10.0.5.15
    cache

While the route does remain in the table (so it can be modified if
needed rather than being wiped away as it would be if IFF_UP was
cleared), the proper next-hop is chosen automatically when the link is
down.  Now interface p8p1 is linked-up:

default via 10.0.5.2 dev p9p1
10.0.5.0/24 dev p9p1  proto kernel  scope link  src 10.0.5.15
70.0.0.0/24 dev p7p1  proto kernel  scope link  src 70.0.0.1
80.0.0.0/24 dev p8p1  proto kernel  scope link  src 80.0.0.1
90.0.0.0/24 via 80.0.0.2 dev p8p1  metric 1
90.0.0.0/24 via 70.0.0.2 dev p7p1  metric 2
192.168.56.0/24 dev p2p1  proto kernel  scope link  src 192.168.56.2
90.0.0.1 via 80.0.0.2 dev p8p1  src 80.0.0.1
    cache
local 80.0.0.1 dev lo  src 80.0.0.1
    cache <local>
80.0.0.2 dev p8p1  src 80.0.0.1
    cache

and the output changes to what one would expect.

If the sysctl is not set, the following output would be expected when
p8p1 is down:

default via 10.0.5.2 dev p9p1
10.0.5.0/24 dev p9p1  proto kernel  scope link  src 10.0.5.15
70.0.0.0/24 dev p7p1  proto kernel  scope link  src 70.0.0.1
80.0.0.0/24 dev p8p1  proto kernel  scope link  src 80.0.0.1 linkdown
90.0.0.0/24 via 80.0.0.2 dev p8p1  metric 1 linkdown
90.0.0.0/24 via 70.0.0.2 dev p7p1  metric 2

Since the dead flag does not appear, there should be no expectation that
the kernel would skip using this route due to link being down.

v2: Split kernel changes into 2 patches, this actually makes a
behavioral change if the sysctl is set.  Also took suggestion from Alex
to simplify code by only checking sysctl during fib lookup and
suggestion from Scott to add a per-interface sysctl.

v3: Code clean-ups to make it more readable and efficient as well as a
reverse path check fix.

v4: Drop binary sysctl

v5: Whitespace fixups from Dave

v6: Style changes from Dave and checkpatch suggestions

v7: One more checkpatch fixup
Signed-off-by: NAndy Gospodarek <gospo@cumulusnetworks.com>
Signed-off-by: NDinesh Dutt <ddutt@cumulusnetworks.com>
Acked-by: NScott Feldman <sfeldma@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a fib flag called RTNH_F_LINKDOWN to any ipv4 nexthops that are
reachable via an interface where carrier is off.  No action is taken,
but additional flags are passed to userspace to indicate carrier status.

This also includes a cleanup to fib_disable_ip to more clearly indicate
what event made the function call to replace the more cryptic force
option previously used.

v2: Split out kernel functionality into 2 patches, this patch simply
sets and clears new nexthop flag RTNH_F_LINKDOWN.

v3: Cleanups suggested by Alex as well as a bug noticed in
fib_sync_down_dev and fib_sync_up when multipath was not enabled.

v5: Whitespace and variable declaration fixups suggested by Dave.

v6: Style fixups noticed by Dave; ran checkpatch to be sure I got them
all.
Signed-off-by: NAndy Gospodarek <gospo@cumulusnetworks.com>
Signed-off-by: NDinesh Dutt <ddutt@cumulusnetworks.com>
Acked-by: NScott Feldman <sfeldma@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is meant to collapse local and main into one by converting
tb_data from an array to a pointer.  Doing this allows us to point the
local table into the main while maintaining the same variables in the
table.

As such the tb_data was converted from an array to a pointer, and a new
array called data is added in order to still provide an object for tb_data
to point to.

In order to track the origin of the fib aliases a tb_id value was added in
a hole that existed on 64b systems.  Using this we can also reverse the
merge in the event that custom FIB rules are enabled.

With this patch I am seeing an improvement of 20ns to 30ns for routing
lookups as long as custom rules are not enabled, with custom rules enabled
we fall back to split tables and the original behavior.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Keep switchdev FIB offload model simple for now and don't allow custom ip
rules.
Signed-off-by: NScott Feldman <sfeldma@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The fib_table was wrapped in several places with an
rcu_read_lock/rcu_read_unlock however after looking over the code I found
several spots where the tables were being accessed as just standard
pointers without any protections.  This change fixes that so that all of
the proper protections are in place when accessing the table to take RCU
replacement or removal of the table into account.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There isn't any advantage to having it as a list and by making it an hlist
we make the fib_alias more compatible with the list_info in terms of the
type of list used.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change is to start cleaning up some of the rcu_read_lock/unlock
handling.  I realized while reviewing the code there are several spots that
I don't believe are being handled correctly or are masking warnings by
locally calling rcu_read_lock/unlock instead of calling them at the correct
level.

A common example is a call to fib_get_table followed by fib_table_lookup.
The rcu_read_lock/unlock ought to wrap both but there are several spots where
they were not wrapped.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In recent testing I had disabled CONFIG_IP_MULTIPLE_TABLES and as a result
when I ran "cat /proc/net/fib_trie" the main trie was displayed multiple
times.  I found that the problem line of code was in the function
fib_trie_seq_next.  Specifically the line below caused the indexes to go in
the opposite direction of our traversal:

	h = tb->tb_id & (FIB_TABLE_HASHSZ - 1);

This issue was that the RT tables are defined such that RT_TABLE_LOCAL is ID
255, while it is located at TABLE_LOCAL_INDEX of 0, and RT_TABLE_MAIN is 254
with a TABLE_MAIN_INDEX of 1.  This means that the above line will return 1
for the local table and 0 for main.  The result is that fib_trie_seq_next
will return NULL at the end of the local table, fib_trie_seq_start will
return the start of the main table, and then fib_trie_seq_next will loop on
main forever as h will always return 0.

The fix for this is to reverse the ordering of the two tables.  It has the
advantage of making it so that the tables now print in the same order
regardless of if multiple tables are enabled or not.  In order to make the
definition consistent with the multiple tables case I simply masked the to
RT_TABLE_XXX values by (FIB_TABLE_HASHSZ - 1).  This way the two table
layouts should always stay consistent.

Fixes: 93456b6d ("[IPV4]: Unify access to the routing tables")
Signed-off-by: NAlexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Lets make this hash function a bit secure, as ICMP attacks are still
in the wild.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

nh_exceptions is effectively used under rcu, but lacks proper
barriers. Between kzalloc() and setting of nh->nh_exceptions(),
we need a proper memory barrier.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Fixes: 4895c771 ("ipv4: Add FIB nexthop exceptions.")
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There are a mix of function prototypes with and without extern
in the kernel sources.  Standardize on not using extern for
function prototypes.

Function prototypes don't need to be written with extern.
extern is assumed by the compiler.  Its use is as unnecessary as
using auto to declare automatic/local variables in a block.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch removes the fib_update_nh_saddrs() declaration from
include/net/ip_fib.h, as the fib_update_nh_saddrs() method was removed in
coomit 436c3b66 ("ipv4: Invalidate nexthop cache nh_saddr more correctly").
Signed-off-by: NRami Rosen <ramirose@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit d2d68ba9 (ipv4: Cache input routes in fib_info nexthops)
assmued that "locally destined, and routed packets, never trigger
PMTU events or redirects that will be processed by us".

However, it seems that tunnel devices do trigger PMTU events in certain
cases. At least ip_gre, ip6_gre, sit, and ipip do use the inner flow's
skb_dst(skb)->ops->update_pmtu to propage mtu information from the
outer flows. These can cause the inner flow mtu to be decreased. If
next hop exceptions are not consulted for pmtu, IP fragmentation will
not be done properly for these routes.

It also seems that we really need to have the PMTU information always
for netfilter TCPMSS clamp-to-pmtu feature to work properly.

So for the time being, cache separate copies of input routes for
each next hop exception.
Signed-off-by: NTimo Teräs <timo.teras@iki.fi>
Reviewed-by: NJulian Anastasov <ja@ssi.bg>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

commit 13d82bf5 (ipv4: Fix flushing of cached routing informations)
added the support to flush learned pmtu information.

However, using rt_genid is quite heavy as it is bumped on route
add/change and multicast events amongst other places. These can
happen quite often, especially if using dynamic routing protocols.

While this is ok with routes (as they are just recreated locally),
the pmtu information is learned from remote systems and the icmp
notification can come with long delays. It is worthy to have separate
genid to avoid excessive pmtu resets.

Cc: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: NTimo Teräs <timo.teras@iki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

a long time ago by the commit

  commit 93456b6d
  Author: Denis V. Lunev <den@openvz.org>
  Date:   Thu Jan 10 03:23:38 2008 -0800

    [IPV4]: Unify access to the routing tables.

the defenition of FIB_HASH_TABLE size has obtained wrong dependency:
it should depend upon CONFIG_IP_MULTIPLE_TABLES (as was in the original
code) but it was depended from CONFIG_IP_ROUTE_MULTIPATH

This patch returns the situation to the original state.

The problem was spotted by Tingwei Liu.
Signed-off-by: NDenis V. Lunev <den@openvz.org>
CC: Tingwei Liu <tingw.liu@gmail.com>
CC: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

commit d2d68ba9 (ipv4: Cache input routes in fib_info nexthops.)
introduced a regression for forwarding.

This was hard to reproduce but the symptom was that packets were
delivered to local host instead of being forwarded.

David suggested to add fib_type to fib_info so that we dont
inadvertently share same fib_info for different purposes.

With help from Julian Anastasov who provided very helpful
hints, reproduced here :

<quote>
        Can it be a problem related to fib_info reuse
from different routes. For example, when local IP address
is created for subnet we have:

broadcast 192.168.0.255 dev DEV  proto kernel  scope link  src
192.168.0.1
192.168.0.0/24 dev DEV  proto kernel  scope link  src 192.168.0.1
local 192.168.0.1 dev DEV  proto kernel  scope host  src 192.168.0.1

        The "dev DEV  proto kernel  scope link  src 192.168.0.1" is
a reused fib_info structure where we put cached routes.
The result can be same fib_info for 192.168.0.255 and
192.168.0.0/24. RTN_BROADCAST is cached only for input
routes. Incoming broadcast to 192.168.0.255 can be cached
and can cause problems for traffic forwarded to 192.168.0.0/24.
So, this patch should solve the problem because it
separates the broadcast from unicast traffic.

        And the ip_route_input_slow caching will work for
local and broadcast input routes (above routes 1 and 3) just
because they differ in scope and use different fib_info.

</quote>

Many thanks to Chris Clayton for his patience and help.
Reported-by: NChris Clayton <chris2553@googlemail.com>
Bisected-by: NChris Clayton <chris2553@googlemail.com>
Reported-by: NDave Jones <davej@redhat.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Julian Anastasov <ja@ssi.bg>
Tested-by: NChris Clayton <chris2553@googlemail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>