== The problem ==
See description of the problem in the initial patch of this patch set.
== The solution ==
The patch provides much more reliable in-kernel solution for the 2nd
part of the problem: making outgoing connecttion from desired IP.
It adds new attach types `BPF_CGROUP_INET4_CONNECT` and
`BPF_CGROUP_INET6_CONNECT` for program type
`BPF_PROG_TYPE_CGROUP_SOCK_ADDR` that can be used to override both
source and destination of a connection at connect(2) time.
Local end of connection can be bound to desired IP using newly
introduced BPF-helper `bpf_bind()`. It allows to bind to only IP though,
and doesn't support binding to port, i.e. leverages
`IP_BIND_ADDRESS_NO_PORT` socket option. There are two reasons for this:
* looking for a free port is expensive and can affect performance
significantly;
* there is no use-case for port.
As for remote end (`struct sockaddr *` passed by user), both parts of it
can be overridden, remote IP and remote port. It's useful if an
application inside cgroup wants to connect to another application inside
same cgroup or to itself, but knows nothing about IP assigned to the
cgroup.
Support is added for IPv4 and IPv6, for TCP and UDP.
IPv4 and IPv6 have separate attach types for same reason as sys_bind
hooks, i.e. to prevent reading from / writing to e.g. user_ip6 fields
when user passes sockaddr_in since it'd be out-of-bound.
== Implementation notes ==
The patch introduces new field in `struct proto`: `pre_connect` that is
a pointer to a function with same signature as `connect` but is called
before it. The reason is in some cases BPF hooks should be called way
before control is passed to `sk->sk_prot->connect`. Specifically
`inet_dgram_connect` autobinds socket before calling
`sk->sk_prot->connect` and there is no way to call `bpf_bind()` from
hooks from e.g. `ip4_datagram_connect` or `ip6_datagram_connect` since
it'd cause double-bind. On the other hand `proto.pre_connect` provides a
flexible way to add BPF hooks for connect only for necessary `proto` and
call them at desired time before `connect`. Since `bpf_bind()` is
allowed to bind only to IP and autobind in `inet_dgram_connect` binds
only port there is no chance of double-bind.
bpf_bind() sets `force_bind_address_no_port` to bind to only IP despite
of value of `bind_address_no_port` socket field.
bpf_bind() sets `with_lock` to `false` when calling to __inet_bind()
and __inet6_bind() since all call-sites, where bpf_bind() is called,
already hold socket lock.
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Refactor `bind()` code to make it ready to be called from BPF helper
function `bpf_bind()` (will be added soon). Implementation of
`inet_bind()` and `inet6_bind()` is separated into `__inet_bind()` and
`__inet6_bind()` correspondingly. These function can be used from both
`sk_prot->bind` and `bpf_bind()` contexts.
New functions have two additional arguments.
`force_bind_address_no_port` forces binding to IP only w/o checking
`inet_sock.bind_address_no_port` field. It'll allow to bind local end of
a connection to desired IP in `bpf_bind()` w/o changing
`bind_address_no_port` field of a socket. It's useful since `bpf_bind()`
can return an error and we'd need to restore original value of
`bind_address_no_port` in that case if we changed this before calling to
the helper.
`with_lock` specifies whether to lock socket when working with `struct
sk` or not. The argument is set to `true` for `sk_prot->bind`, i.e. old
behavior is preserved. But it will be set to `false` for `bpf_bind()`
use-case. The reason is all call-sites, where `bpf_bind()` will be
called, already hold that socket lock.
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
== The problem ==
There is a use-case when all processes inside a cgroup should use one
single IP address on a host that has multiple IP configured. Those
processes should use the IP for both ingress and egress, for TCP and UDP
traffic. So TCP/UDP servers should be bound to that IP to accept
incoming connections on it, and TCP/UDP clients should make outgoing
connections from that IP. It should not require changing application
code since it's often not possible.
Currently it's solved by intercepting glibc wrappers around syscalls
such as `bind(2)` and `connect(2)`. It's done by a shared library that
is preloaded for every process in a cgroup so that whenever TCP/UDP
server calls `bind(2)`, the library replaces IP in sockaddr before
passing arguments to syscall. When application calls `connect(2)` the
library transparently binds the local end of connection to that IP
(`bind(2)` with `IP_BIND_ADDRESS_NO_PORT` to avoid performance penalty).
Shared library approach is fragile though, e.g.:
* some applications clear env vars (incl. `LD_PRELOAD`);
* `/etc/ld.so.preload` doesn't help since some applications are linked
with option `-z nodefaultlib`;
* other applications don't use glibc and there is nothing to intercept.
== The solution ==
The patch provides much more reliable in-kernel solution for the 1st
part of the problem: binding TCP/UDP servers on desired IP. It does not
depend on application environment and implementation details (whether
glibc is used or not).
It adds new eBPF program type `BPF_PROG_TYPE_CGROUP_SOCK_ADDR` and
attach types `BPF_CGROUP_INET4_BIND` and `BPF_CGROUP_INET6_BIND`
(similar to already existing `BPF_CGROUP_INET_SOCK_CREATE`).
The new program type is intended to be used with sockets (`struct sock`)
in a cgroup and provided by user `struct sockaddr`. Pointers to both of
them are parts of the context passed to programs of newly added types.
The new attach types provides hooks in `bind(2)` system call for both
IPv4 and IPv6 so that one can write a program to override IP addresses
and ports user program tries to bind to and apply such a program for
whole cgroup.
== Implementation notes ==
[1]
Separate attach types for `AF_INET` and `AF_INET6` are added
intentionally to prevent reading/writing to offsets that don't make
sense for corresponding socket family. E.g. if user passes `sockaddr_in`
it doesn't make sense to read from / write to `user_ip6[]` context
fields.
[2]
The write access to `struct bpf_sock_addr_kern` is implemented using
special field as an additional "register".
There are just two registers in `sock_addr_convert_ctx_access`: `src`
with value to write and `dst` with pointer to context that can't be
changed not to break later instructions. But the fields, allowed to
write to, are not available directly and to access them address of
corresponding pointer has to be loaded first. To get additional register
the 1st not used by `src` and `dst` one is taken, its content is saved
to `bpf_sock_addr_kern.tmp_reg`, then the register is used to load
address of pointer field, and finally the register's content is restored
from the temporary field after writing `src` value.
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Donald reported that IPv6 route leaking between VRFs is not working.
The root cause is the strict argument in the call to rt6_lookup when
validating the nexthop spec.
ip6_route_check_nh validates the gateway and device (if given) of a
route spec. It in turn could call rt6_lookup (e.g., lookup in a given
table did not succeed so it falls back to a full lookup) and if so
sets the strict argument to 1. That means if the egress device is given,
the route lookup needs to return a result with the same device. This
strict requirement does not work with VRFs (IPv4 or IPv6) because the
oif in the flow struct is overridden with the index of the VRF device
to trigger a match on the l3mdev rule and force the lookup to its table.
The right long term solution is to add an l3mdev index to the flow
struct such that the oif is not overridden. That solution will not
backport well, so this patch aims for a simpler solution to relax the
strict argument if the route spec device is an l3mdev slave. As done
in other places, use the FLOWI_FLAG_SKIP_NH_OIF to know that the
RT6_LOOKUP_F_IFACE flag needs to be removed.
Fixes: ca254490c8 ("net: Add VRF support to IPv6 stack")
Reported-by: Donald Sharp <sharpd@cumulusnetworks.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Enabling TSO can lead to abysmal performances when using seg6 in
encap mode, such as with the ixgbe driver. This patch adds a call to
iptunnel_handle_offloads() to remove the encapsulation bit if needed.
Before:
root@comp4-seg6bpf:~# iperf3 -c fc00::55
Connecting to host fc00::55, port 5201
[ 4] local fc45::4 port 36592 connected to fc00::55 port 5201
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 196 KBytes 1.60 Mbits/sec 47 6.66 KBytes
[ 4] 1.00-2.00 sec 304 KBytes 2.49 Mbits/sec 100 5.33 KBytes
[ 4] 2.00-3.00 sec 284 KBytes 2.32 Mbits/sec 92 5.33 KBytes
After:
root@comp4-seg6bpf:~# iperf3 -c fc00::55
Connecting to host fc00::55, port 5201
[ 4] local fc45::4 port 43062 connected to fc00::55 port 5201
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 1.03 GBytes 8.89 Gbits/sec 0 743 KBytes
[ 4] 1.00-2.00 sec 1.03 GBytes 8.87 Gbits/sec 0 743 KBytes
[ 4] 2.00-3.00 sec 1.03 GBytes 8.87 Gbits/sec 0 743 KBytes
Reported-by: Tom Herbert <tom@quantonium.net>
Fixes: 6c8702c60b ("ipv6: sr: add support for SRH encapsulation and injection with lwtunnels")
Signed-off-by: David Lebrun <dlebrun@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Do not allow setting ipv6 routes from userspace if disable_ipv6 has been
enabled. The issue can be triggered using the following reproducer:
- sysctl net.ipv6.conf.all.disable_ipv6=1
- ip -6 route add a🅱️c:d::/64 dev em1
- ip -6 route show
a🅱️c:d::/64 dev em1 metric 1024 pref medium
Fix it checking disable_ipv6 value in ip6_route_info_create routine
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter/IPVS updates for net-next
The following patchset contains Netfilter/IPVS updates for your net-next
tree. This batch comes with more input sanitization for xtables to
address bug reports from fuzzers, preparation works to the flowtable
infrastructure and assorted updates. In no particular order, they are:
1) Make sure userspace provides a valid standard target verdict, from
Florian Westphal.
2) Sanitize error target size, also from Florian.
3) Validate that last rule in basechain matches underflow/policy since
userspace assumes this when decoding the ruleset blob that comes
from the kernel, from Florian.
4) Consolidate hook entry checks through xt_check_table_hooks(),
patch from Florian.
5) Cap ruleset allocations at 512 mbytes, 134217728 rules and reject
very large compat offset arrays, so we have a reasonable upper limit
and fuzzers don't exercise the oom-killer. Patches from Florian.
6) Several WARN_ON checks on xtables mutex helper, from Florian.
7) xt_rateest now has a hashtable per net, from Cong Wang.
8) Consolidate counter allocation in xt_counters_alloc(), from Florian.
9) Earlier xt_table_unlock() call in {ip,ip6,arp,eb}tables, patch
from Xin Long.
10) Set FLOW_OFFLOAD_DIR_* to IP_CT_DIR_* definitions, patch from
Felix Fietkau.
11) Consolidate code through flow_offload_fill_dir(), also from Felix.
12) Inline ip6_dst_mtu_forward() just like ip_dst_mtu_maybe_forward()
to remove a dependency with flowtable and ipv6.ko, from Felix.
13) Cache mtu size in flow_offload_tuple object, this is safe for
forwarding as f87c10a8aa describes, from Felix.
14) Rename nf_flow_table.c to nf_flow_table_core.o, to simplify too
modular infrastructure, from Felix.
15) Add rt0, rt2 and rt4 IPv6 routing extension support, patch from
Ahmed Abdelsalam.
16) Remove unused parameter in nf_conncount_count(), from Yi-Hung Wei.
17) Support for counting only to nf_conncount infrastructure, patch
from Yi-Hung Wei.
18) Add strict NFT_CT_{SRC_IP,DST_IP,SRC_IP6,DST_IP6} key datatypes
to nft_ct.
19) Use boolean as return value from ipt_ah and from IPVS too, patch
from Gustavo A. R. Silva.
20) Remove useless parameters in nfnl_acct_overquota() and
nf_conntrack_broadcast_help(), from Taehee Yoo.
21) Use ipv6_addr_is_multicast() from xt_cluster, also from Taehee Yoo.
22) Statify nf_tables_obj_lookup_byhandle, patch from Fengguang Wu.
23) Fix typo in xt_limit, from Geert Uytterhoeven.
24) Do no use VLAs in Netfilter code, again from Gustavo.
25) Use ADD_COUNTER from ebtables, from Taehee Yoo.
26) Bitshift support for CONNMARK and MARK targets, from Jack Ma.
27) Use pr_*() and add pr_fmt(), from Arushi Singhal.
28) Add synproxy support to ctnetlink.
29) ICMP type and IGMP matching support for ebtables, patches from
Matthias Schiffer.
30) Support for the revision infrastructure to ebtables, from
Bernie Harris.
31) String match support for ebtables, also from Bernie.
32) Documentation for the new flowtable infrastructure.
33) Use generic comparison functions in ebt_stp, from Joe Perches.
34) Demodularize filter chains in nftables.
35) Register conntrack hooks in case nftables NAT chain is added.
36) Merge assignments with return in a couple of spots in the
Netfilter codebase, also from Arushi.
37) Document that xtables percpu counters are stored in the same
memory area, from Ben Hutchings.
38) Revert mark_source_chains() sanity checks that break existing
rulesets, from Florian Westphal.
39) Use is_zero_ether_addr() in the ipset codebase, from Joe Perches.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 0d7df906a0.
Valdis Kletnieks reported that xtables is broken in linux-next since
0d7df906a0 ("netfilter: x_tables: ensure last rule in base chain
matches underflow/policy"), as kernel rejects the (well-formed) ruleset:
[ 64.402790] ip6_tables: last base chain position 1136 doesn't match underflow 1344 (hook 1)
mark_source_chains is not the correct place for such a check, as it
terminates evaluation of a chain once it sees an unconditional verdict
(following rules are known to be unreachable). It seems preferrable to
fix libiptc instead, so remove this check again.
Fixes: 0d7df906a0 ("netfilter: x_tables: ensure last rule in base chain matches underflow/policy")
Reported-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Register conntrack hooks if the user adds NAT chains. Users get confused
with the existing behaviour since they will see no packets hitting this
chain until they add the first rule that refers to conntrack.
This patch adds new ->init() and ->free() indirections to chain types
that can be used by NAT chains to invoke the conntrack dependency.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
One module per supported filter chain family type takes too much memory
for very little code - too much modularization - place all chain filter
definitions in one single file.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Use WARN_ON() instead since it should not happen that neither family
goes over NFPROTO_NUMPROTO nor there is already a chain of this type
already registered.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Use nft_ prefix. By when I added chain types, I forgot to use the
nftables prefix. Rename enum nft_chain_type to enum nft_chain_types too,
otherwise there is an overlap.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
IPv4 was changed in commit 52a773d645 ("net: Export ip fragment
sysctl to unprivileged users")
The only sysctl that is not per-netns is not used :
ip6frag_secret_interval
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Nikolay Borisov <kernel@kyup.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move call to call_fib6_entry_notifiers for new IPv6 routes to right
before the insertion into the FIB. At this point notifier handlers can
decide the fate of the new route with a clean path to delete the
potential new entry if the notifier returns non-0.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Steffen Klassert says:
====================
pull request (net-next): ipsec-next 2018-03-29
1) Remove a redundant pointer initialization esp_input_set_header().
From Colin Ian King.
2) Mark the xfrm kmem_caches as __ro_after_init.
From Alexey Dobriyan.
3) Do the checksum for an ipsec offlad packet in software
if the device does not advertise NETIF_F_HW_ESP_TX_CSUM.
From Shannon Nelson.
4) Use booleans for true and false instead of integers
in xfrm_policy_cache_flush().
From Gustavo A. R. Silva
Please pull or let me know if there are problems.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Steffen Klassert says:
====================
pull request (net): ipsec 2018-03-29
1) Fix a rcu_read_lock/rcu_read_unlock imbalance
in the error path of xfrm_local_error().
From Taehee Yoo.
2) Some VTI MTU fixes. From Stefano Brivio.
3) Fix a too early overwritten skb control buffer
on xfrm transport mode.
Please note that this pull request has a merge conflict
in net/ipv4/ip_tunnel.c.
The conflict is between
commit f6cc9c054e ("ip_tunnel: Emit events for post-register MTU changes")
from the net tree and
commit 24fc79798b ("ip_tunnel: Clamp MTU to bounds on new link")
from the ipsec tree.
It can be solved as it is currently done in linux-next.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Synchronous pernet_operations are not allowed anymore.
All are asynchronous. So, drop the structure member.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove local ADBG macro and use netdev_dbg/pr_debug
Miscellanea:
o Remove unnecessary debug message after allocation failure as there
already is a dump_stack() on the failure paths
o Leave the allocation failure message on snmp6_alloc_dev as there
is one code path that does not do a dump_stack()
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since ipmr and ip6mr are using the same mr_mfc struct at their core, we
can now refactor the ipmr_cache_{hold,put} logic and apply refcounting
to both ipmr and ip6mr.
Signed-off-by: Yuval Mintz <yuvalm@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the ability to discern whether a given FIB rule notification relates
to the default rule inserted when registering ip6mr or a different one.
Would later be used by drivers wishing to offload ipv6 multicast routes
but unable to offload rules other than the default one.
Signed-off-by: Yuval Mintz <yuvalm@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In similar fashion to ipmr, support fib notifications for ip6mr mfc and
vif related events. This would later allow drivers to react to said
notifications and offload the IPv6 mroutes.
Signed-off-by: Yuval Mintz <yuvalm@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Prefer the direct use of octal for permissions.
Done with checkpatch -f --types=SYMBOLIC_PERMS --fix-inplace
and some typing.
Miscellanea:
o Whitespace neatening around these conversions.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, the SMC experimental TCP option in a SYN packet is lost on
the server side when SYN Cookies are active. However, the corresponding
SYNACK sent back to the client contains the SMC option. This causes an
inconsistent view of the SMC capabilities on the client and server.
This patch disables the SMC option in the SYNACK when SYN Cookies are
active to avoid this issue.
Fixes: 60e2a77807 ("tcp: TCP experimental option for SMC")
Signed-off-by: Hans Wippel <hwippel@linux.vnet.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for your net tree,
they are:
1) Don't pick fixed hash implementation for NFT_SET_EVAL sets, otherwise
userspace hits EOPNOTSUPP with valid rules using the meter statement,
from Florian Westphal.
2) If you send a batch that flushes the existing ruleset (that contains
a NAT chain) and the new ruleset definition comes with a new NAT
chain, don't bogusly hit EBUSY. Also from Florian.
3) Missing netlink policy attribute validation, from Florian.
4) Detach conntrack template from skbuff if IP_NODEFRAG is set on,
from Paolo Abeni.
5) Cache device names in flowtable object, otherwise we may end up
walking over devices going aways given no rtnl_lock is held.
6) Fix incorrect net_device ingress with ingress hooks.
7) Fix crash when trying to read more data than available in UDP
packets from the nf_socket infrastructure, from Subash.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
skb_header_pointer will copy data into a buffer if data is non linear,
otherwise it will return a pointer in the linear section of the data.
nf_sk_lookup_slow_v{4,6} always copies data of size udphdr but later
accesses memory within the size of tcphdr (th->doff) in case of TCP
packets. This causes a crash when running with KASAN with the following
call stack -
BUG: KASAN: stack-out-of-bounds in xt_socket_lookup_slow_v4+0x524/0x718
net/netfilter/xt_socket.c:178
Read of size 2 at addr ffffffe3d417a87c by task syz-executor/28971
CPU: 2 PID: 28971 Comm: syz-executor Tainted: G B W O 4.9.65+ #1
Call trace:
[<ffffff9467e8d390>] dump_backtrace+0x0/0x428 arch/arm64/kernel/traps.c:76
[<ffffff9467e8d7e0>] show_stack+0x28/0x38 arch/arm64/kernel/traps.c:226
[<ffffff946842d9b8>] __dump_stack lib/dump_stack.c:15 [inline]
[<ffffff946842d9b8>] dump_stack+0xd4/0x124 lib/dump_stack.c:51
[<ffffff946811d4b0>] print_address_description+0x68/0x258 mm/kasan/report.c:248
[<ffffff946811d8c8>] kasan_report_error mm/kasan/report.c:347 [inline]
[<ffffff946811d8c8>] kasan_report.part.2+0x228/0x2f0 mm/kasan/report.c:371
[<ffffff946811df44>] kasan_report+0x5c/0x70 mm/kasan/report.c:372
[<ffffff946811bebc>] check_memory_region_inline mm/kasan/kasan.c:308 [inline]
[<ffffff946811bebc>] __asan_load2+0x84/0x98 mm/kasan/kasan.c:739
[<ffffff94694d6f04>] __tcp_hdrlen include/linux/tcp.h:35 [inline]
[<ffffff94694d6f04>] xt_socket_lookup_slow_v4+0x524/0x718 net/netfilter/xt_socket.c:178
Fix this by copying data into appropriate size headers based on protocol.
Fixes: a583636a83 ("inet: refactor inet[6]_lookup functions to take skb")
Signed-off-by: Tejaswi Tanikella <tejaswit@codeaurora.org>
Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Fun set of conflict resolutions here...
For the mac80211 stuff, these were fortunately just parallel
adds. Trivially resolved.
In drivers/net/phy/phy.c we had a bug fix in 'net' that moved the
function phy_disable_interrupts() earlier in the file, whilst in
'net-next' the phy_error() call from this function was removed.
In net/ipv4/xfrm4_policy.c, David Ahern's changes to remove the
'rt_table_id' member of rtable collided with a bug fix in 'net' that
added a new struct member "rt_mtu_locked" which needs to be copied
over here.
The mlxsw driver conflict consisted of net-next separating
the span code and definitions into separate files, whilst
a 'net' bug fix made some changes to that moved code.
The mlx5 infiniband conflict resolution was quite non-trivial,
the RDMA tree's merge commit was used as a guide here, and
here are their notes:
====================
Due to bug fixes found by the syzkaller bot and taken into the for-rc
branch after development for the 4.17 merge window had already started
being taken into the for-next branch, there were fairly non-trivial
merge issues that would need to be resolved between the for-rc branch
and the for-next branch. This merge resolves those conflicts and
provides a unified base upon which ongoing development for 4.17 can
be based.
Conflicts:
drivers/infiniband/hw/mlx5/main.c - Commit 42cea83f95
(IB/mlx5: Fix cleanup order on unload) added to for-rc and
commit b5ca15ad7e (IB/mlx5: Add proper representors support)
add as part of the devel cycle both needed to modify the
init/de-init functions used by mlx5. To support the new
representors, the new functions added by the cleanup patch
needed to be made non-static, and the init/de-init list
added by the representors patch needed to be modified to
match the init/de-init list changes made by the cleanup
patch.
Updates:
drivers/infiniband/hw/mlx5/mlx5_ib.h - Update function
prototypes added by representors patch to reflect new function
names as changed by cleanup patch
drivers/infiniband/hw/mlx5/ib_rep.c - Update init/de-init
stage list to match new order from cleanup patch
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The current logic of flags | TUNNEL_SEQ is always non-zero and hence
sequence numbers are always incremented no matter the setting of the
TUNNEL_SEQ bit. Fix this by using & instead of |.
Detected by CoverityScan, CID#1466039 ("Operands don't affect result")
Fixes: 77a5196a80 ("gre: add sequence number for collect md mode.")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For multipath routes the ONLINK flag can be specified per nexthop in
rtnh_flags or globally in rtm_flags. Update ip6_route_multipath_add
to consider the ONLINK setting coming from rtnh_flags. Each loop over
nexthops the config for the sibling route is initialized to the global
config and then per nexthop settings overlayed. The flag is 'or'ed into
fib6_config to handle the ONLINK flag coming from either rtm_flags or
rtnh_flags.
Fixes: fc1e64e109 ("net/ipv6: Add support for onlink flag")
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These pernet_operations register and unregister sysctl.
Also, there is inet_frags_exit_net() called in exit method,
which has to be safe after a560002437 "net: Fix hlist
corruptions in inet_evict_bucket()".
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes: 2f987a76a9 ("net: ipv6: keep sk status consistent after datagram connect failure")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch exposes synproxy information per-conntrack. Moreover, send
sequence adjustment events once server sends us the SYN,ACK packet, so
we can synchronize the sequence adjustment too for packets going as
reply from the server, as part of the synproxy logic.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
We shouldn't allow a tunnel to have IP_MAX_MTU as MTU, because
another IPv6 header is going on top of our packets. Without this
patch, we might end up building packets bigger than IP_MAX_MTU.
Fixes: b96f9afee4 ("ipv4/6: use core net MTU range checking")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
In vti6_link_config(), if MTU is already given on link creation
or change, validate and use it instead of recomputing it. To do
that, we need to propagate the knowledge that MTU was set by
userspace all the way down to vti6_link_config().
To keep this simple, vti6_dev_init() sets the new 'keep_mtu'
argument of vti6_link_config() to true: on initialization, we
don't have convenient access to netlink attributes there, but we
will anyway check whether dev->mtu is set in vti6_link_config().
If it's non-zero, it was set to the value of the IFLA_MTU
attribute during creation. Otherwise, determine a reasonable
value.
Fixes: ed1efb2aef ("ipv6: Add support for IPsec virtual tunnel interfaces")
Fixes: 53c81e95df ("ip6_vti: adjust vti mtu according to mtu of lower device")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
If a lower device is found, we don't need to subtract
LL_MAX_HEADER to calculate our MTU: just use its MTU, the link
layer headers are already taken into account by it.
If the lower device is not found, start from ETH_DATA_LEN
instead, and only in this case subtract a worst-case
LL_MAX_HEADER.
We then need to subtract our additional IPv6 header from the
calculation.
While at it, note that vti6 doesn't have a hardware header, so
it doesn't need to set dev->hard_header_len. And as
vti6_link_config() now always sets the MTU, there's no need to
set a default value in vti6_dev_setup().
This makes the behaviour consistent with IPv4 vti, after
commit a32452366b ("vti4: Don't count header length twice."),
which was accidentally reverted by merge commit f895f0cfbb
("Merge branch 'master' of
git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec").
While commit 53c81e95df ("ip6_vti: adjust vti mtu according to
mtu of lower device") improved on the original situation, this
was still not ideal. As reported in that commit message itself,
if we start from an underlying veth MTU of 9000, we end up with
an MTU of 8832, that is, 9000 - LL_MAX_HEADER - sizeof(ipv6hdr).
This should simply be 8880, or 9000 - sizeof(ipv6hdr) instead:
we found the lower device (veth) and we know we don't have any
additional link layer header, so there's no need to subtract an
hypothetical worst-case number.
Fixes: 53c81e95df ("ip6_vti: adjust vti mtu according to mtu of lower device")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
This patch moves the udp_rmem_min, udp_wmem_min
to namespace and init the udp_l3mdev_accept explicitly.
The udp_rmem_min/udp_wmem_min affect udp rx/tx queue,
with this patch namespaces can set them differently.
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Lookup the L3 master device for the passed in device. Only consider
addresses on netdev's with the same master device. If the device is
not enslaved or is NULL, then the l3mdev is NULL which means only
devices not enslaved (ie, in the default domain) are considered.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
Move gateway validation code from ip6_route_info_create into
ip6_validate_gw. Code move plus adjustments to handle the potential
reset of dev and idev and to make checkpatch happy.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Steffen Klassert says:
====================
pull request (net): ipsec 2018-03-13
1) Refuse to insert 32 bit userspace socket policies on 64
bit systems like we do it for standard policies. We don't
have a compat layer, so inserting socket policies from
32 bit userspace will lead to a broken configuration.
2) Make the policy hold queue work without the flowcache.
Dummy bundles are not chached anymore, so we need to
generate a new one on each lookup as long as the SAs
are not yet in place.
3) Fix the validation of the esn replay attribute. The
The sanity check in verify_replay() is bypassed if
the XFRM_STATE_ESN flag is not set. Fix this by doing
the sanity check uncoditionally.
From Florian Westphal.
4) After most of the dst_entry garbage collection code
is removed, we may leak xfrm_dst entries as they are
neither cached nor tracked somewhere. Fix this by
reusing the 'uncached_list' to track xfrm_dst entries
too. From Xin Long.
5) Fix a rcu_read_lock/rcu_read_unlock imbalance in
xfrm_get_tos() From Xin Long.
6) Fix an infinite loop in xfrm_get_dst_nexthop. On
transport mode we fetch the child dst_entry after
we continue, so this pointer is never updated.
Fix this by fetching it before we continue.
7) Fix ESN sequence number gap after IPsec GSO packets.
We accidentally increment the sequence number counter
on the xfrm_state by one packet too much in the ESN
case. Fix this by setting the sequence number to the
correct value.
8) Reset the ethernet protocol after decapsulation only if a
mac header was set. Otherwise it breaks configurations
with TUN devices. From Yossi Kuperman.
9) Fix __this_cpu_read() usage in preemptible code. Use
this_cpu_read() instead in ipcomp_alloc_tfms().
From Greg Hackmann.
Please pull or let me know if there are problems.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
On unsuccesful ip6_datagram_connect(), if the failure is caused by
ip6_datagram_dst_update(), the sk peer information are cleared, but
the sk->sk_state is preserved.
If the socket was already in an established status, the overall sk
status is inconsistent and fouls later checks in datagram code.
Fix this saving the old peer information and restoring them in
case of failure. This also aligns ipv6 datagram connect() behavior
with ipv4.
v1 -> v2:
- added missing Fixes tag
Fixes: 85cb73ff9b ("net: ipv6: reset daddr and dport in sk if connect() fails")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The patch adds skb_cow_header() to ensure enough headroom
at ip6erspan_tunnel_xmit before pushing the erspan header
to the skb.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When users fill in incorrect erspan version number through
the struct erspan_metadata uapi, current code skips pushing
the erspan header but continue pushing the gre header, which
is incorrect. The patch fixes it by returning error.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The patch adds the erspan v2 proto in ip6gre_tunnel_lookup
so the erspan v2 tunnel can be found correctly.
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fallback tunnels (like tunl0, gre0, gretap0, erspan0, sit0,
ip6tnl0, ip6gre0) are automatically created when the corresponding
module is loaded.
These tunnels are also automatically created when a new network
namespace is created, at a great cost.
In many cases, netns are used for isolation purposes, and these
extra network devices are a waste of resources. We are using
thousands of netns per host, and hit the netns creation/delete
bottleneck a lot. (Many thanks to Kirill for recent work on this)
Add a new sysctl so that we can opt-out from this automatic creation.
Note that these tunnels are still created for the initial namespace,
to be the least intrusive for typical setups.
Tested:
lpk43:~# cat add_del_unshare.sh
for i in `seq 1 40`
do
(for j in `seq 1 100` ; do unshare -n /bin/true >/dev/null ; done) &
done
wait
lpk43:~# echo 0 >/proc/sys/net/core/fb_tunnels_only_for_init_net
lpk43:~# time ./add_del_unshare.sh
real 0m37.521s
user 0m0.886s
sys 7m7.084s
lpk43:~# echo 1 >/proc/sys/net/core/fb_tunnels_only_for_init_net
lpk43:~# time ./add_del_unshare.sh
real 0m4.761s
user 0m0.851s
sys 1m8.343s
lpk43:~#
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix the following slab-out-of-bounds kasan report in
ndisc_fill_redirect_hdr_option when the incoming ipv6 packet is not
linear and the accessed data are not in the linear data region of orig_skb.
[ 1503.122508] ==================================================================
[ 1503.122832] BUG: KASAN: slab-out-of-bounds in ndisc_send_redirect+0x94e/0x990
[ 1503.123036] Read of size 1184 at addr ffff8800298ab6b0 by task netperf/1932
[ 1503.123220] CPU: 0 PID: 1932 Comm: netperf Not tainted 4.16.0-rc2+ #124
[ 1503.123347] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.10.2-2.fc27 04/01/2014
[ 1503.123527] Call Trace:
[ 1503.123579] <IRQ>
[ 1503.123638] print_address_description+0x6e/0x280
[ 1503.123849] kasan_report+0x233/0x350
[ 1503.123946] memcpy+0x1f/0x50
[ 1503.124037] ndisc_send_redirect+0x94e/0x990
[ 1503.125150] ip6_forward+0x1242/0x13b0
[...]
[ 1503.153890] Allocated by task 1932:
[ 1503.153982] kasan_kmalloc+0x9f/0xd0
[ 1503.154074] __kmalloc_track_caller+0xb5/0x160
[ 1503.154198] __kmalloc_reserve.isra.41+0x24/0x70
[ 1503.154324] __alloc_skb+0x130/0x3e0
[ 1503.154415] sctp_packet_transmit+0x21a/0x1810
[ 1503.154533] sctp_outq_flush+0xc14/0x1db0
[ 1503.154624] sctp_do_sm+0x34e/0x2740
[ 1503.154715] sctp_primitive_SEND+0x57/0x70
[ 1503.154807] sctp_sendmsg+0xaa6/0x1b10
[ 1503.154897] sock_sendmsg+0x68/0x80
[ 1503.154987] ___sys_sendmsg+0x431/0x4b0
[ 1503.155078] __sys_sendmsg+0xa4/0x130
[ 1503.155168] do_syscall_64+0x171/0x3f0
[ 1503.155259] entry_SYSCALL_64_after_hwframe+0x42/0xb7
[ 1503.155436] Freed by task 1932:
[ 1503.155527] __kasan_slab_free+0x134/0x180
[ 1503.155618] kfree+0xbc/0x180
[ 1503.155709] skb_release_data+0x27f/0x2c0
[ 1503.155800] consume_skb+0x94/0xe0
[ 1503.155889] sctp_chunk_put+0x1aa/0x1f0
[ 1503.155979] sctp_inq_pop+0x2f8/0x6e0
[ 1503.156070] sctp_assoc_bh_rcv+0x6a/0x230
[ 1503.156164] sctp_inq_push+0x117/0x150
[ 1503.156255] sctp_backlog_rcv+0xdf/0x4a0
[ 1503.156346] __release_sock+0x142/0x250
[ 1503.156436] release_sock+0x80/0x180
[ 1503.156526] sctp_sendmsg+0xbb0/0x1b10
[ 1503.156617] sock_sendmsg+0x68/0x80
[ 1503.156708] ___sys_sendmsg+0x431/0x4b0
[ 1503.156799] __sys_sendmsg+0xa4/0x130
[ 1503.156889] do_syscall_64+0x171/0x3f0
[ 1503.156980] entry_SYSCALL_64_after_hwframe+0x42/0xb7
[ 1503.157158] The buggy address belongs to the object at ffff8800298ab600
which belongs to the cache kmalloc-1024 of size 1024
[ 1503.157444] The buggy address is located 176 bytes inside of
1024-byte region [ffff8800298ab600, ffff8800298aba00)
[ 1503.157702] The buggy address belongs to the page:
[ 1503.157820] page:ffffea0000a62a00 count:1 mapcount:0 mapping:0000000000000000 index:0x0 compound_mapcount: 0
[ 1503.158053] flags: 0x4000000000008100(slab|head)
[ 1503.158171] raw: 4000000000008100 0000000000000000 0000000000000000 00000001800e000e
[ 1503.158350] raw: dead000000000100 dead000000000200 ffff880036002600 0000000000000000
[ 1503.158523] page dumped because: kasan: bad access detected
[ 1503.158698] Memory state around the buggy address:
[ 1503.158816] ffff8800298ab900: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[ 1503.158988] ffff8800298ab980: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[ 1503.159165] >ffff8800298aba00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[ 1503.159338] ^
[ 1503.159436] ffff8800298aba80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 1503.159610] ffff8800298abb00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 1503.159785] ==================================================================
[ 1503.159964] Disabling lock debugging due to kernel taint
The test scenario to trigger the issue consists of 4 devices:
- H0: data sender, connected to LAN0
- H1: data receiver, connected to LAN1
- GW0 and GW1: routers between LAN0 and LAN1. Both of them have an
ethernet connection on LAN0 and LAN1
On H{0,1} set GW0 as default gateway while on GW0 set GW1 as next hop for
data from LAN0 to LAN1.
Moreover create an ip6ip6 tunnel between H0 and H1 and send 3 concurrent
data streams (TCP/UDP/SCTP) from H0 to H1 through ip6ip6 tunnel (send
buffer size is set to 16K). While data streams are active flush the route
cache on HA multiple times.
I have not been able to identify a given commit that introduced the issue
since, using the reproducer described above, the kasan report has been
triggered from 4.14 and I have not gone back further.
Reported-by: Jianlin Shi <jishi@redhat.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
