linux-dev/net/openvswitch, branch linus/master

net: openvswitch: fix parsing of nw_proto for IPv6 fragments

2022-06-23T09:44:01Z

When a packet enters the OVS datapath and does not match any existing flows installed in the kernel flow cache, the packet will be sent to userspace to be parsed, and a new flow will be created. The kernel and OVS rely on each other to parse packet fields in the same way so that packets will be handled properly. As per the design document linked below, OVS expects all later IPv6 fragments to have nw_proto=44 in the flow key, so they can be correctly matched on OpenFlow rules. OpenFlow controllers create pipelines based on this design. This behavior was changed by the commit in the Fixes tag so that nw_proto equals the next_header field of the last extension header. However, there is no counterpart for this change in OVS userspace, meaning that this field is parsed differently between OVS and the kernel. This is a problem because OVS creates actions based on what is parsed in userspace, but the kernel-provided flow key is used as a match criteria, as described in Documentation/networking/openvswitch.rst. This leads to issues such as packets incorrectly matching on a flow and thus the wrong list of actions being applied to the packet. Such changes in packet parsing cannot be implemented without breaking the userspace. The offending commit is partially reverted to restore the expected behavior. The change technically made sense and there is a good reason that it was implemented, but it does not comply with the original design of OVS. If in the future someone wants to implement such a change, then it must be user-configurable and disabled by default to preserve backwards compatibility with existing OVS versions. Cc: stable@vger.kernel.org Fixes: fa642f08839b ("openvswitch: Derive IP protocol number for IPv6 later frags") Link: https://docs.openvswitch.org/en/latest/topics/design/#fragments Signed-off-by: Rosemarie O'Riorden Acked-by: Eelco Chaudron Link: https://lore.kernel.org/r/20220621204845.9721-1-roriorden@redhat.com Signed-off-by: Paolo Abeni

net: openvswitch: fix misuse of the cached connection on tuple changes

2022-06-09T03:49:52Z

If packet headers changed, the cached nfct is no longer relevant for the packet and attempt to re-use it leads to the incorrect packet classification. This issue is causing broken connectivity in OpenStack deployments with OVS/OVN due to hairpin traffic being unexpectedly dropped. The setup has datapath flows with several conntrack actions and tuple changes between them: actions:ct(commit,zone=8,mark=0/0x1,nat(src)), set(eth(src=00:00:00:00:00:01,dst=00:00:00:00:00:06)), set(ipv4(src=172.18.2.10,dst=192.168.100.6,ttl=62)), ct(zone=8),recirc(0x4) After the first ct() action the packet headers are almost fully re-written. The next ct() tries to re-use the existing nfct entry and marks the packet as invalid, so it gets dropped later in the pipeline. Clearing the cached conntrack entry whenever packet tuple is changed to avoid the issue. The flow key should not be cleared though, because we should still be able to match on the ct_state if the recirculation happens after the tuple change but before the next ct() action. Cc: stable@vger.kernel.org Fixes: 7f8a436eaa2c ("openvswitch: Add conntrack action") Reported-by: Frode Nordahl Link: https://mail.openvswitch.org/pipermail/ovs-discuss/2022-May/051829.html Link: https://bugs.launchpad.net/ubuntu/+source/ovn/+bug/1967856 Signed-off-by: Ilya Maximets Link: https://lore.kernel.org/r/20220606221140.488984-1-i.maximets@ovn.org Signed-off-by: Jakub Kicinski

openvswitch: fix OOB access in reserve_sfa_size()

2022-04-15T10:50:02Z

Given a sufficiently large number of actions, while copying and reserving memory for a new action of a new flow, if next_offset is greater than MAX_ACTIONS_BUFSIZE, the function reserve_sfa_size() does not return -EMSGSIZE as expected, but it allocates MAX_ACTIONS_BUFSIZE bytes increasing actions_len by req_size. This can then lead to an OOB write access, especially when further actions need to be copied. Fix it by rearranging the flow action size check. KASAN splat below: ================================================================== BUG: KASAN: slab-out-of-bounds in reserve_sfa_size+0x1ba/0x380 [openvswitch] Write of size 65360 at addr ffff888147e4001c by task handler15/836 CPU: 1 PID: 836 Comm: handler15 Not tainted 5.18.0-rc1+ #27 ... Call Trace: dump_stack_lvl+0x45/0x5a print_report.cold+0x5e/0x5db ? __lock_text_start+0x8/0x8 ? reserve_sfa_size+0x1ba/0x380 [openvswitch] kasan_report+0xb5/0x130 ? reserve_sfa_size+0x1ba/0x380 [openvswitch] kasan_check_range+0xf5/0x1d0 memcpy+0x39/0x60 reserve_sfa_size+0x1ba/0x380 [openvswitch] __add_action+0x24/0x120 [openvswitch] ovs_nla_add_action+0xe/0x20 [openvswitch] ovs_ct_copy_action+0x29d/0x1130 [openvswitch] ? __kernel_text_address+0xe/0x30 ? unwind_get_return_address+0x56/0xa0 ? create_prof_cpu_mask+0x20/0x20 ? ovs_ct_verify+0xf0/0xf0 [openvswitch] ? prep_compound_page+0x198/0x2a0 ? __kasan_check_byte+0x10/0x40 ? kasan_unpoison+0x40/0x70 ? ksize+0x44/0x60 ? reserve_sfa_size+0x75/0x380 [openvswitch] __ovs_nla_copy_actions+0xc26/0x2070 [openvswitch] ? __zone_watermark_ok+0x420/0x420 ? validate_set.constprop.0+0xc90/0xc90 [openvswitch] ? __alloc_pages+0x1a9/0x3e0 ? __alloc_pages_slowpath.constprop.0+0x1da0/0x1da0 ? unwind_next_frame+0x991/0x1e40 ? __mod_node_page_state+0x99/0x120 ? __mod_lruvec_page_state+0x2e3/0x470 ? __kasan_kmalloc_large+0x90/0xe0 ovs_nla_copy_actions+0x1b4/0x2c0 [openvswitch] ovs_flow_cmd_new+0x3cd/0xb10 [openvswitch] ... Cc: stable@vger.kernel.org Fixes: f28cd2af22a0 ("openvswitch: fix flow actions reallocation") Signed-off-by: Paolo Valerio Acked-by: Eelco Chaudron Signed-off-by: David S. Miller

net: openvswitch: fix leak of nested actions

2022-04-06T12:36:50Z

While parsing user-provided actions, openvswitch module may dynamically allocate memory and store pointers in the internal copy of the actions. So this memory has to be freed while destroying the actions. Currently there are only two such actions: ct() and set(). However, there are many actions that can hold nested lists of actions and ovs_nla_free_flow_actions() just jumps over them leaking the memory. For example, removal of the flow with the following actions will lead to a leak of the memory allocated by nf_ct_tmpl_alloc(): actions:clone(ct(commit),0) Non-freed set() action may also leak the 'dst' structure for the tunnel info including device references. Under certain conditions with a high rate of flow rotation that may cause significant memory leak problem (2MB per second in reporter's case). The problem is also hard to mitigate, because the user doesn't have direct control over the datapath flows generated by OVS. Fix that by iterating over all the nested actions and freeing everything that needs to be freed recursively. New build time assertion should protect us from this problem if new actions will be added in the future. Unfortunately, openvswitch module doesn't use NLA_F_NESTED, so all attributes has to be explicitly checked. sample() and clone() actions are mixing extra attributes into the user-provided action list. That prevents some code generalization too. Fixes: 34ae932a4036 ("openvswitch: Make tunnel set action attach a metadata dst") Link: https://mail.openvswitch.org/pipermail/ovs-dev/2022-March/392922.html Reported-by: Stéphane Graber Signed-off-by: Ilya Maximets Acked-by: Aaron Conole Signed-off-by: David S. Miller

net: openvswitch: don't send internal clone attribute to the userspace.

2022-04-06T00:57:54Z

'OVS_CLONE_ATTR_EXEC' is an internal attribute that is used for performance optimization inside the kernel. It's added by the kernel while parsing user-provided actions and should not be sent during the flow dump as it's not part of the uAPI. The issue doesn't cause any significant problems to the ovs-vswitchd process, because reported actions are not really used in the application lifecycle and only supposed to be shown to a human via ovs-dpctl flow dump. However, the action list is still incorrect and causes the following error if the user wants to look at the datapath flows: # ovs-dpctl add-dp system@ovs-system # ovs-dpctl add-flow "" "clone(ct(commit),0)" # ovs-dpctl dump-flows , packets:0, bytes:0, used:never, actions:clone(bad length 4, expected -1 for: action0(01 00 00 00), ct(commit),0) With the fix: # ovs-dpctl dump-flows , packets:0, bytes:0, used:never, actions:clone(ct(commit),0) Additionally fixed an incorrect attribute name in the comment. Fixes: b233504033db ("openvswitch: kernel datapath clone action") Signed-off-by: Ilya Maximets Acked-by: Aaron Conole Link: https://lore.kernel.org/r/20220404104150.2865736-1-i.maximets@ovn.org Signed-off-by: Jakub Kicinski

openvswitch: Add recirc_id to recirc warning

2022-03-31T15:52:48Z

When hitting the recirculation limit, the kernel would currently log something like this: [ 58.586597] openvswitch: ovs-system: deferred action limit reached, drop recirc action Which isn't all that useful to debug as we only have the interface name to go on but can't track it down to a specific flow. With this change, we now instead get: [ 58.586597] openvswitch: ovs-system: deferred action limit reached, drop recirc action (recirc_id=0x9e) Which can now be correlated with the flow entries from OVS. Suggested-by: Frode Nordahl Signed-off-by: Stéphane Graber Tested-by: Stephane Graber Acked-by: Eelco Chaudron Link: https://lore.kernel.org/r/20220330194244.3476544-1-stgraber@ubuntu.com Signed-off-by: Jakub Kicinski

openvswitch: Fixed nd target mask field in the flow dump.

2022-03-29T07:47:41Z

IPv6 nd target mask was not getting populated in flow dump. In the function __ovs_nla_put_key the icmp code mask field was checked instead of icmp code key field to classify the flow as neighbour discovery. ufid:bdfbe3e5-60c2-43b0-a5ff-dfcac1c37328, recirc_id(0),dp_hash(0/0), skb_priority(0/0),in_port(ovs-nm1),skb_mark(0/0),ct_state(0/0), ct_zone(0/0),ct_mark(0/0),ct_label(0/0), eth(src=00:00:00:00:00:00/00:00:00:00:00:00, dst=00:00:00:00:00:00/00:00:00:00:00:00), eth_type(0x86dd), ipv6(src=::/::,dst=::/::,label=0/0,proto=58,tclass=0/0,hlimit=0/0,frag=no), icmpv6(type=135,code=0), nd(target=2001::2/::, sll=00:00:00:00:00:00/00:00:00:00:00:00, tll=00:00:00:00:00:00/00:00:00:00:00:00), packets:10, bytes:860, used:0.504s, dp:ovs, actions:ovs-nm2 Fixes: e64457191a25 (openvswitch: Restructure datapath.c and flow.c) Signed-off-by: Martin Varghese Link: https://lore.kernel.org/r/20220328054148.3057-1-martinvarghesenokia@gmail.com Signed-off-by: Paolo Abeni

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

2022-03-23T17:53:49Z

Merge in overtime fixes, no conflicts. Signed-off-by: Jakub Kicinski

openvswitch: always update flow key after nat

2022-03-22T05:29:10Z

During NAT, a tuple collision may occur. When this happens, openvswitch will make a second pass through NAT which will perform additional packet modification. This will update the skb data, but not the flow key that OVS uses. This means that future flow lookups, and packet matches will have incorrect data. This has been supported since 5d50aa83e2c8 ("openvswitch: support asymmetric conntrack"). That commit failed to properly update the sw_flow_key attributes, since it only called the ovs_ct_nat_update_key once, rather than each time ovs_ct_nat_execute was called. As these two operations are linked, the ovs_ct_nat_execute() function should always make sure that the sw_flow_key is updated after a successful call through NAT infrastructure. Fixes: 5d50aa83e2c8 ("openvswitch: support asymmetric conntrack") Cc: Dumitru Ceara Cc: Numan Siddique Signed-off-by: Aaron Conole Acked-by: Eelco Chaudron Link: https://lore.kernel.org/r/20220318124319.3056455-1-aconole@redhat.com Signed-off-by: Jakub Kicinski

net: openvswitch: fix uAPI incompatibility with existing user space

2022-03-11T04:14:52Z

Few years ago OVS user space made a strange choice in the commit [1] to define types only valid for the user space inside the copy of a kernel uAPI header. '#ifndef __KERNEL__' and another attribute was added later. This leads to the inevitable clash between user space and kernel types when the kernel uAPI is extended. The issue was unveiled with the addition of a new type for IPv6 extension header in kernel uAPI. When kernel provides the OVS_KEY_ATTR_IPV6_EXTHDRS attribute to the older user space application, application tries to parse it as OVS_KEY_ATTR_PACKET_TYPE and discards the whole netlink message as malformed. Since OVS_KEY_ATTR_IPV6_EXTHDRS is supplied along with every IPv6 packet that goes to the user space, IPv6 support is fully broken. Fixing that by bringing these user space attributes to the kernel uAPI to avoid the clash. Strictly speaking this is not the problem of the kernel uAPI, but changing it is the only way to avoid breakage of the older user space applications at this point. These 2 types are explicitly rejected now since they should not be passed to the kernel. Additionally, OVS_KEY_ATTR_TUNNEL_INFO moved out from the '#ifdef __KERNEL__' as there is no good reason to hide it from the userspace. And it's also explicitly rejected now, because it's for in-kernel use only. Comments with warnings were added to avoid the problem coming back. (1 << type) converted to (1ULL << type) to avoid integer overflow on OVS_KEY_ATTR_IPV6_EXTHDRS, since it equals 32 now. [1] beb75a40fdc2 ("userspace: Switching of L3 packets in L2 pipeline") Fixes: 28a3f0601727 ("net: openvswitch: IPv6: Add IPv6 extension header support") Link: https://lore.kernel.org/netdev/3adf00c7-fe65-3ef4-b6d7-6d8a0cad8a5f@nvidia.com Link: https://github.com/openvswitch/ovs/commit/beb75a40fdc295bfd6521b0068b4cd12f6de507c Reported-by: Roi Dayan Signed-off-by: Ilya Maximets Acked-by: Nicolas Dichtel Acked-by: Aaron Conole Link: https://lore.kernel.org/r/20220309222033.3018976-1-i.maximets@ovn.org Signed-off-by: Jakub Kicinski