The IPv4 RFCs are specific (though obtuse) that an echo response
packet needs to contain all the options from the echo request,
much as if it been routed back to the sender, though apparently
with a new TTL. They suggest copying the incoming packet header
to achieve this so that is what this patch does.
PiperOrigin-RevId: 335559176
As per relevant IP RFCS (see code comments), broadcast (for IPv4) and
multicast addresses are not allowed. Currently checks for these are
done at the transport layer, but since it is explicitly forbidden at
the IP layers, check for them there.
This change also removes the UDP.InvalidSourceAddress stat since there
is no longer a need for it.
Test: ip_test.TestSourceAddressValidation
PiperOrigin-RevId: 334490971
Currently expired IP fragments are discarded only if another fragment for the
same IP datagram is received after timeout or the total size of the fragment
queue exceeded a predefined value.
Test: fragmentation.TestReassemblingTimeout
Fixes#3960
PiperOrigin-RevId: 334423710
* Remove Capabilities and NICID methods from NetworkEndpoint.
* Remove linkEP and stack parameters from NetworkProtocol.NewEndpoint.
The LinkEndpoint can be fetched from the NetworkInterface. The stack
is passed to the NetworkProtocol when it is created so the
NetworkEndpoint can get it from its protocol.
* Remove stack parameter from TransportProtocol.NewEndpoint.
Like the NetworkProtocol/Endpoint, the stack is passed to the
TransportProtocol when it is created.
PiperOrigin-RevId: 334332721
* Add network address to network endpoints.
Hold network-specific state in the NetworkEndpoint instead of the stack.
This results in the stack no longer needing to "know" about the network
endpoints and special case certain work for various endpoints
(e.g. IPv6 DAD).
* Provide NetworkEndpoints with an NetworkInterface interface.
Instead of just passing the NIC ID of a NIC, pass an interface so the
network endpoint may query other information about the NIC such as
whether or not it is a loopback device.
* Move NDP code and state to the IPv6 package.
NDP is IPv6 specific so there is no need for it to live in the stack.
* Control forwarding through NetworkProtocols instead of Stack
Forwarding should be controlled on a per-network protocol basis so
forwarding configurations are now controlled through network protocols.
* Remove stack.referencedNetworkEndpoint.
Now that addresses are exposed via AddressEndpoint and only one
NetworkEndpoint is created per interface, there is no need for a
referenced NetworkEndpoint.
* Assume network teardown methods are infallible.
Fixes#3871, #3916
PiperOrigin-RevId: 334319433
Network or transport protocols may want to reach the stack. Support this
by letting the stack create the protocol instances so it can pass a
reference to itself at protocol creation time.
Note, protocols do not yet use the stack in this CL but later CLs will
make use of the stack from protocols.
PiperOrigin-RevId: 334260210
SO_LINGER is a socket level option and should be stored on all endpoints even
though it is used to linger only for TCP endpoints.
PiperOrigin-RevId: 332369252
The routing table (in its current) form should not be used to make
decisions about whether a remote address is a broadcast address or
not (for IPv4).
Note, a destination subnet does not always map to a network.
E.g. RouterA may have a route to 192.168.0.0/22 through RouterB,
but RouterB may be configured with 4x /24 subnets on 4 different
interfaces.
See https://github.com/google/gvisor/issues/3938.
PiperOrigin-RevId: 331819868
The existing implementation for TransportProtocol.{Set}Option take
arguments of an empty interface type which all types (implicitly)
implement; any type may be passed to the functions.
This change introduces marker interfaces for transport protocol options
that may be set or queried which transport protocol option types
implement to ensure that invalid types are caught at compile time.
Different interfaces are used to allow the compiler to enforce read-only
or set-only socket options.
RELNOTES: n/a
PiperOrigin-RevId: 330559811
Accept on gVisor will return an error if a socket in the accept queue was closed
before Accept() was called. Linux will return the new fd even if the returned
socket is already closed by the peer say due to a RST being sent by the peer.
This seems to be intentional in linux more details on the github issue.
Fixes#3780
PiperOrigin-RevId: 329828404
The existing implementation for {G,S}etSockOpt take arguments of an
empty interface type which all types (implicitly) implement; any
type may be passed to the functions.
This change introduces marker interfaces for socket options that may be
set or queried which socket option types implement to ensure that invalid
types are caught at compile time. Different interfaces are used to allow
the compiler to enforce read-only or set-only socket options.
Fixes#3714.
RELNOTES: n/a
PiperOrigin-RevId: 328832161
In an upcoming CL, socket option types are made to implement a marker
interface with pointer receivers. Since this results in calling methods
of an interface with a pointer, we incur an allocation when attempting
to get an Endpoint's last error with the current implementation.
When calling the method of an interface, the compiler is unable to
determine what the interface implementation does with the pointer
(since calling a method on an interface uses virtual dispatch at runtime
so the compiler does not know what the interface method will do) so it
allocates on the heap to be safe incase an implementation continues to
hold the pointer after the functioon returns (the reference escapes the
scope of the object).
In the example below, the compiler does not know what b.foo does with
the reference to a it allocates a on the heap as the reference to a may
escape the scope of a.
```
var a int
var b someInterface
b.foo(&a)
```
This change removes the opportunity for that allocation.
RELNOTES: n/a
PiperOrigin-RevId: 328796559
This change adds an option to replace the current implementation of ARP through
linkAddrCache, with an implementation of NUD through neighborCache. Switching
to using NUD for both ARP and NDP is beneficial for the reasons described by
RFC 4861 Section 3.1:
"[Using NUD] significantly improves the robustness of packet delivery in the
presence of failing routers, partially failing or partitioned links, or nodes
that change their link-layer addresses. For instance, mobile nodes can move
off-link without losing any connectivity due to stale ARP caches."
"Unlike ARP, Neighbor Unreachability Detection detects half-link failures and
avoids sending traffic to neighbors with which two-way connectivity is
absent."
Along with these changes exposes the API for querying and operating the
neighbor cache. Operations include:
- Create a static entry
- List all entries
- Delete all entries
- Remove an entry by address
This also exposes the API to change the NUD protocol constants on a per-NIC
basis to allow Neighbor Discovery to operate over links with widely varying
performance characteristics. See [RFC 4861 Section 10][1] for the list of
constants.
Finally, an API for subscribing to NUD state changes is exposed through
NUDDispatcher. See [RFC 4861 Appendix C][3] for the list of edges.
Tests:
pkg/tcpip/network/arp:arp_test
+ TestDirectRequest
pkg/tcpip/network/ipv6:ipv6_test
+ TestLinkResolution
+ TestNDPValidation
+ TestNeighorAdvertisementWithTargetLinkLayerOption
+ TestNeighorSolicitationResponse
+ TestNeighorSolicitationWithSourceLinkLayerOption
+ TestRouterAdvertValidation
pkg/tcpip/stack:stack_test
+ TestCacheWaker
+ TestForwardingWithFakeResolver
+ TestForwardingWithFakeResolverManyPackets
+ TestForwardingWithFakeResolverManyResolutions
+ TestForwardingWithFakeResolverPartialTimeout
+ TestForwardingWithFakeResolverTwoPackets
+ TestIPv6SourceAddressSelectionScopeAndSameAddress
[1]: https://tools.ietf.org/html/rfc4861#section-10
[2]: https://tools.ietf.org/html/rfc4861#appendix-CFixes#1889Fixes#1894Fixes#1895Fixes#1947Fixes#1948Fixes#1949Fixes#1950
PiperOrigin-RevId: 328365034
The NetworkEndpoint does not need to be created for each address.
Most of the work the NetworkEndpoint does is address agnostic.
PiperOrigin-RevId: 326759605
Formerly, when a packet is constructed or parsed, all headers are set by the
client code. This almost always involved prepending to pk.Header buffer or
trimming pk.Data portion. This is known to prone to bugs, due to the complexity
and number of the invariants assumed across netstack to maintain.
In the new PacketHeader API, client will call Push()/Consume() method to
construct/parse an outgoing/incoming packet. All invariants, such as slicing
and trimming, are maintained by the API itself.
NewPacketBuffer() is introduced to create new PacketBuffer. Zero value is no
longer valid.
PacketBuffer now assumes the packet is a concatenation of following portions:
* LinkHeader
* NetworkHeader
* TransportHeader
* Data
Any of them could be empty, or zero-length.
PiperOrigin-RevId: 326507688
Packets MUST NOT use a non-unicast source address for ICMP
Echo Replies.
Test: integration_test.TestPingMulticastBroadcast
PiperOrigin-RevId: 325634380
When sending packets to a known network's broadcast address, use the
broadcast MAC address.
Test:
- stack_test.TestOutgoingSubnetBroadcast
- udp_test.TestOutgoingSubnetBroadcast
PiperOrigin-RevId: 324062407
Changes the API of tcpip.Clock to also provide a method for scheduling and
rescheduling work after a specified duration. This change also implements the
AfterFunc method for existing implementations of tcpip.Clock.
This is the groundwork required to mock time within tests. All references to
CancellableTimer has been replaced with the tcpip.Job interface, allowing for
custom implementations of scheduling work.
This is a BREAKING CHANGE for clients that implement their own tcpip.Clock or
use tcpip.CancellableTimer. Migration plan:
1. Add AfterFunc(d, f) to tcpip.Clock
2. Replace references of tcpip.CancellableTimer with tcpip.Job
3. Replace calls to tcpip.CancellableTimer#StopLocked with tcpip.Job#Cancel
4. Replace calls to tcpip.CancellableTimer#Reset with tcpip.Job#Schedule
5. Replace calls to tcpip.NewCancellableTimer with tcpip.NewJob.
PiperOrigin-RevId: 322906897
RFC-1122 (and others) specify that UDP should not receive
datagrams that have a source address that is a multicast address.
Packets should never be received FROM a multicast address.
See also, RFC 768: 'User Datagram Protocol'
J. Postel, ISI, 28 August 1980
A UDP datagram received with an invalid IP source address
(e.g., a broadcast or multicast address) must be discarded
by UDP or by the IP layer (see rfc 1122 Section 3.2.1.3).
This CL does not address TCP or broadcast which is more complicated.
Also adds a test for both ipv6 and ipv4 UDP.
Fixes#3154
PiperOrigin-RevId: 320547674
RFC 6864 imposes various restrictions on the uniqueness of the IPv4
Identification field for non-atomic datagrams, defined as an IP datagram that
either can be fragmented (DF=0) or is already a fragment (MF=1 or positive
fragment offset). In order to be compliant, the ID field is assigned for all
non-atomic datagrams.
Add a TCP unit test that induces retransmissions and checks that the IPv4
ID field is unique every time. Add basic handling of the IP_MTU_DISCOVER
socket option so that the option can be used to disable PMTU discovery,
effectively setting DF=0. Attempting to set the sockopt to anything other
than disabled will fail because PMTU discovery is currently not implemented,
and the default behavior matches that of disabled.
PiperOrigin-RevId: 320081842
SO_NO_CHECK is used to skip the UDP checksum generation on a TX socket
(UDP checksum is optional on IPv4).
Test:
- TestNoChecksum
- SoNoCheckOffByDefault (UdpSocketTest)
- SoNoCheck (UdpSocketTest)
Fixes#3055
PiperOrigin-RevId: 318575215
Linux controls socket send/receive buffers using a few sysctl variables
- net.core.rmem_default
- net.core.rmem_max
- net.core.wmem_max
- net.core.wmem_default
- net.ipv4.tcp_rmem
- net.ipv4.tcp_wmem
The first 4 control the default socket buffer sizes for all sockets
raw/packet/tcp/udp and also the maximum permitted socket buffer that can be
specified in setsockopt(SOL_SOCKET, SO_(RCV|SND)BUF,...).
The last two control the TCP auto-tuning limits and override the default
specified in rmem_default/wmem_default as well as the max limits.
Netstack today only implements tcp_rmem/tcp_wmem and incorrectly uses it
to limit the maximum size in setsockopt() as well as uses it for raw/udp
sockets.
This changelist introduces the other 4 and updates the udp/raw sockets to use
the newly introduced variables. The values for min/max match the current
tcp_rmem/wmem values and the default value buffers for UDP/RAW sockets is
updated to match the linux value of 212KiB up from the really low current value
of 32 KiB.
Updates #3043Fixes#3043
PiperOrigin-RevId: 318089805
For TCP sockets, SO_REUSEADDR relaxes the rules for binding addresses.
gVisor/netstack already supported a behavior similar to SO_REUSEADDR, but did
not allow disabling it. This change brings the SO_REUSEADDR behavior closer to
the behavior implemented by Linux and adds a new SO_REUSEADDR disabled
behavior. Like Linux, SO_REUSEADDR is now disabled by default.
PiperOrigin-RevId: 317984380
On UDP sockets, SO_REUSEADDR allows multiple sockets to bind to the same
address, but only delivers packets to the most recently bound socket. This
differs from the behavior of SO_REUSEADDR on TCP sockets. SO_REUSEADDR for TCP
sockets will likely need an almost completely independent implementation.
SO_REUSEADDR has some odd interactions with the similar SO_REUSEPORT. These
interactions are tested fairly extensively and all but one particularly odd
one (that honestly seems like a bug) behave the same on gVisor and Linux.
PiperOrigin-RevId: 315844832
Netstack has traditionally parsed headers on-demand as a packet moves up the
stack. This is conceptually simple and convenient, but incompatible with
iptables, where headers can be inspected and mangled before even a routing
decision is made.
This changes header parsing to happen early in the incoming packet path, as soon
as the NIC gets the packet from a link endpoint. Even if an invalid packet is
found (e.g. a TCP header of insufficient length), the packet is passed up the
stack for proper stats bookkeeping.
PiperOrigin-RevId: 315179302
IPTables.connections contains a sync.RWMutex. Copying it will trigger copylocks
analysis. Tested by manually enabling nogo tests.
sync.RWMutex is added to IPTables for the additional race condition discovered.
PiperOrigin-RevId: 314817019
Historically we've been passing PacketBuffer by shallow copying through out
the stack. Right now, this is only correct as the caller would not use
PacketBuffer after passing into the next layer in netstack.
With new buffer management effort in gVisor/netstack, PacketBuffer will
own a Buffer (to be added). Internally, both PacketBuffer and Buffer may
have pointers and shallow copying shouldn't be used.
Updates #2404.
PiperOrigin-RevId: 314610879
These methods let users eaily break the VectorisedView abstraction, and
allowed netstack to slip into pseudo-enforcement of the "all headers are
in the first View" invariant. Removing them and replacing with PullUp(n)
breaks this reliance and will make it easier to add iptables support and
rework network buffer management.
The new View.PullUp(n) method is low cost in the common case, when when
all the headers fit in the first View.
PiperOrigin-RevId: 308163542
These methods let users eaily break the VectorisedView abstraction, and
allowed netstack to slip into pseudo-enforcement of the "all headers are
in the first View" invariant. Removing them and replacing with PullUp(n)
breaks this reliance and will make it easier to add iptables support and
rework network buffer management.
The new View.PullUp(n) method is low cost in the common case, when when
all the headers fit in the first View.
Julian Elischer
Ghanan Gowripalan
Toshi Kikuchi
Tamir Duberstein
Nayana Bidari
Bhasker Hariharan
Sam Balana
Ting-Yu Wang
gVisor bot
Tony Gong
Ian Gudger
Kevin Krakauer
Andrei Vagin
Adin Scannell