| Age | Commit message (Collapse) | Author |
|---|
|
Drain the notification channel after first accept as in case the first accept
never blocked then the notification for the first accept will still be in the
channel causing the second accept to fail as it will try to wait on the channel
and return immediately due to the older notification even though there is no
connection yet in the accept queue.
PiperOrigin-RevId: 338710062
|
|
The SO_ACCEPTCONN option is used only on getsockopt(). When this option is
specified, getsockopt() indicates whether socket listening is enabled for
the socket. A value of zero indicates that socket listening is disabled;
non-zero that it is enabled.
PiperOrigin-RevId: 338703206
|
|
Previously, the NIC local address used when completing link resolution
was held in the neighbor entry. A neighbor is not identified by any
NIC local address so remove it.
PiperOrigin-RevId: 338699695
|
|
Earlier the count was dropped only after calling e.deliverAccepted. This lead to
an issue where there were no connections in SYN-RCVD state for the listening
endpoint but e.synRcvdCount would not be zero because it was being reduced only
when handleSynSegment returned after deliverAccepted returned.
This issue is seen when the Nth SYN for a listen backlog of size N which would
cause the listen backlog to be full gets dropped occasionally. This happens when
the new SYN comes at when the previous completed endpoint has been delivered to
the accept queue but the synRcvdCount hasn't yet been decremented because the
goroutine running handleSynSegment has not yet completed.
PiperOrigin-RevId: 338690646
|
|
Our current reference leak checker uses finalizers to verify whether an object
has reached zero references before it is garbage collected. There are multiple
problems with this mechanism, so a rewrite is in order.
With finalizers, there is no way to guarantee that a finalizer will run before
the program exits. When an unreachable object with a finalizer is garbage
collected, its finalizer will be added to a queue and run asynchronously. The
best we can do is run garbage collection upon sandbox exit to make sure that
all finalizers are enqueued.
Furthermore, if there is a chain of finalized objects, e.g. A points to B
points to C, garbage collection needs to run multiple times before all of the
finalizers are enqueued. The first GC run will register the finalizer for A but
not free it. It takes another GC run to free A, at which point B's finalizer
can be registered. As a result, we need to run GC as many times as the length
of the longest such chain to have a somewhat reliable leak checker.
Finally, a cyclical chain of structs pointing to one another will never be
garbage collected if a finalizer is set. This is a well-known issue with Go
finalizers (https://github.com/golang/go/issues/7358). Using leak checking on
filesystem objects that produce cycles will not work and even result in memory
leaks.
The new leak checker stores reference counted objects in a global map when
leak check is enabled and removes them once they are destroyed. At sandbox
exit, any remaining objects in the map are considered as leaked. This provides
a deterministic way of detecting leaks without relying on the complexities of
finalizers and garbage collection.
This approach has several benefits over the former, including:
- Always detects leaks of objects that should be destroyed very close to
sandbox exit. The old checker very rarely detected these leaks, because it
relied on garbage collection to be run in a short window of time.
- Panics if we forgot to enable leak check on a ref-counted object (we will try
to remove it from the map when it is destroyed, but it will never have been
added).
- Can store extra logging information in the map values without adding to the
size of the ref count struct itself. With the size of just an int64, the ref
count object remains compact, meaning frequent operations like IncRef/DecRef
are more cache-efficient.
- Can aggregate leak results in a single report after the sandbox exits.
Instead of having warnings littered in the log, which were
non-deterministically triggered by garbage collection, we can print all
warning messages at once. Note that this could also be a limitation--the
sandbox must exit properly for leaks to be detected.
Some basic benchmarking indicates that this change does not significantly
affect performance when leak checking is enabled, which is understandable
since registering/unregistering is only done once for each filesystem object.
Updates #1486.
PiperOrigin-RevId: 338685972
|
|
Previously a link endpoint was passed to
stack.LinkAddressResolver.LinkAddressRequest. With this change,
implementations that want a route for the link address request may
find one through the stack. Other implementations that want to send
a packet without a route may continue to do so using the network
interface directly.
Test: - arp_test.TestLinkAddressRequest
- ipv6.TestLinkAddressRequest
PiperOrigin-RevId: 338577474
|
|
Also enforce the minimum MTU for IPv4 and IPv6, and discard packets if the
minimum is not met.
PiperOrigin-RevId: 338404225
|
|
PiperOrigin-RevId: 338168977
|
|
PiperOrigin-RevId: 338156438
|
|
//pkg/tcpip/stack:stack_x_test_nogo
//pkg/tcpip/transport/raw:raw_nogo
PiperOrigin-RevId: 338153265
|
|
Before this change, if a link header was included in an incoming packet
that is forwarded, the packet that gets sent out will take the original
packet and add a link header to it while keeping the old link header.
This would make the sent packet look like:
OUTGOING LINK HDR | INCOMING LINK HDR | NETWORK HDR | ...
Obviously this is incorrect as we should drop the incoming link header
and only include the outgoing link header. This change fixes this bug.
Test: integration_test.TestForwarding
PiperOrigin-RevId: 337571447
|
|
The IPv4 header checksum has not been checked, at least in recent times,
so add code to do so. Fix all the tests that fail because they never
needed to set the checksum.
Fixes #4484
PiperOrigin-RevId: 337556243
|
|
Currently, fragmentation can only occur during WritePacket(). This enables
it for WritePackets() and WriteIncludedHeaderPacket() as well.
IPv4 unit tests were refactored to be consistent with the IPv6 unit tests.
This removes the extraHeaderReserveLength field and the related
"prependable bytes" unit tests (for both IPv4 and IPv6) because it was only
testing a panic condition when the value was too low.
Fixes #3796
PiperOrigin-RevId: 337550061
|
|
Allow writing an IPv6 packet where the IPv6 header is a provided by
the user.
* Introduce an error to let callers know a header is malformed.
We previously useed tcpip.ErrInvalidOptionValue but that did not seem
appropriate for generic malformed header errors.
* Populate network header in WriteHeaderIncludedPacket
IPv4's implementation of WriteHeaderIncludedPacket did not previously
populate the packet buffer's network header. This change fixes that.
Fixes #4527
Test: ip_test.TestWriteHeaderIncludedPacket
PiperOrigin-RevId: 337534548
|
|
RFC 4861 section 4.4 comments the Target link-layer address option is sometimes
optional in a Neighbor Advertisement packet:
"When responding to a unicast Neighbor Solicitation this option SHOULD be
included."
Tests:
pkg/tcpip/stack:stack_test
- TestEntryStaleToReachableWhenSolicitedConfirmationWithoutAddress
- TestEntryDelayToReachableWhenSolicitedConfirmationWithoutAddress
- TestEntryProbeToReachableWhenSolicitedConfirmationWithoutAddress
pkg/tcpip/network/ipv6:ipv6_test
- TestCallsToNeighborCache
PiperOrigin-RevId: 337396493
|
|
Test helpers should be used for test setup/teardown, not actual
testing. Use cmp.Diff instead of bytes.Equal to improve readability.
PiperOrigin-RevId: 337323242
|
|
This change also brings back the stack.Route.ResolveWith method so that
we can immediately resolve a route when sending an NA in response to a
a NS with a source link layer address option.
Test: ipv6_test.TestNeighorSolicitationResponse
PiperOrigin-RevId: 337185461
|
|
PiperOrigin-RevId: 336974095
|
|
Remove the duplicate NA size variable while I'm here.
See https://tools.ietf.org/html/rfc4861#section-4.4 for the packet format.
PiperOrigin-RevId: 336943206
|
|
Use the correct constant (Solicit, not Advert) while I'm here.
PiperOrigin-RevId: 336924605
|
|
|
|
The fix in commit 028e045da93b7c1c26417e80e4b4e388b86a713d was incorrect as
it can cause the right edge of the window to shrink when we announce
a zero window due to receive buffer being full as its done before the check
for seeing if the window is being shrunk because of the selected window.
Further the window was calculated purely on available space but in cases where
we are getting full sized segments it makes more sense to use the actual bytes
being held. This CL changes to use the lower of the total available space vs
the available space in the maximal window we could advertise minus the actual
payload bytes being held.
This change also cleans up the code so that the window selection logic is
not duplicated between getSendParams() and windowCrossedACKThresholdLocked.
PiperOrigin-RevId: 336404827
|
|
RACK detects packet reordering by checking if the sender received ACK for
the packet which has the sequence number less than the already acknowledged
packets.
PiperOrigin-RevId: 336397526
|
|
PiperOrigin-RevId: 336339194
|
|
PiperOrigin-RevId: 336304024
|
|
When a completed entry exists for a neighbor, there is no need to block
while reachability is (re)confirmed. The stack should continue to use
the neighbor's link address while NUD is performed.
Test: stack_test.TestNeighborCacheReplace
PiperOrigin-RevId: 336199043
|
|
When a response needs to be sent to an incoming packet, the stack should
consult its neighbour table to determine the remote address's link
address.
When an entry does not exist in the stack's neighbor table, the stack
should queue the packet while link resolution completes. See comments.
PiperOrigin-RevId: 336185457
|
|
When the neighbor table already has link address for a neighbor but is
trying to confirm reachability, it may send unicast probes to the
neighbor.
PiperOrigin-RevId: 336166711
|
|
It was originally set to 30s for IPv6 (same as IPv4) but this is not
what RFC 8200 prescibes. Linux also defaults to 60s [1].
[1] https://github.com/torvalds/linux/blob/47ec5303d73ea344e84f46660fff693c57641386/include/net/ipv6.h#L456
PiperOrigin-RevId: 336034636
|
|
...per RFC 4861 s7.1.1.
PiperOrigin-RevId: 335742851
|
|
Most of the IPv4 fragmentation code was moved in the fragmentation
package and it is reused by IPv6 fragmentation.
Test:
- pkg/tcpip/network/ipv4:ipv4_test
- pkg/tcpip/network/ipv6:ipv6_test
- pkg/tcpip/network/fragmentation:fragmentation_test
Fixes #4389
PiperOrigin-RevId: 335714280
|
|
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
|
|
We can get the network endpoint directly from the NIC.
This is a preparatory CL for when a Route needs to hold a dedicated NIC
as its output interface. This is because when forwarding is enabled,
packets may be sent from a NIC different from the NIC a route's local
address is associated with.
PiperOrigin-RevId: 335484500
|
|
We are currently tracking the minimum RTT for RACK as smoothed RTT. As per RFC
minimum RTT can be a global minimum of all RTTs or filtered value of recent
RTT measurements. In this cl minimum RTT is updated to global minimum of all
RTTs for the connection.
PiperOrigin-RevId: 335061518
|
|
Adds support for the IPv6-compatible redirect target. Redirection is a limited
form of DNAT, where the destination is always the localhost.
Updates #3549.
PiperOrigin-RevId: 334698344
|
|
Before this change, OutgoingPacketErrors was incremented in the
stack.Route methods. This was going to be a problem once
IPv4/IPv6 WritePackets support fragmentation because Route.WritePackets
might now know how many packets are left after an error occurs.
Test:
- pkg/tcpip/network/ipv4:ipv4_test
- pkg/tcpip/network/ipv6:ipv6_test
PiperOrigin-RevId: 334687983
|
|
Add code in IPv6 to send ICMP packets while processing extension headers.
Add some accounting in processing IPV6 Extension headers which
allows us to report meaningful information back in ICMP parameter
problem packets.
IPv4 also needs to send a message when an unsupported protocol
is requested.
Add some tests to generate both ipv4 and ipv6 packets with
various errors and check the responses.
Add some new checkers and cleanup some inconsistencies in
the messages in that file.
Add new error types for the ICMPv4/6 generators.
Fix a bug in the ICMPv4 generator that stopped it from generating
"Unknown protocol" messages.
Updates #2211
PiperOrigin-RevId: 334661716
|
|
PiperOrigin-RevId: 334652998
|
|
PiperOrigin-RevId: 334535896
|
|
PiperOrigin-RevId: 334531794
|
|
Test: stack_test.TestGetMainNICAddressWhenNICDisabled
PiperOrigin-RevId: 334513286
|
|
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
|
|
Like matchers, targets should use a module-like register/lookup system. This
replaces the brittle switch statements we had before.
The only behavior change is supporing IPT_GET_REVISION_TARGET. This makes it
much easier to add IPv6 redirect in the next change.
Updates #3549.
PiperOrigin-RevId: 334469418
|
|
Linux doesn't generate a link-local address for the loopback interface.
Test: integration_test.TestInitialLoopbackAddresses
PiperOrigin-RevId: 334453182
|
|
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
|
|
In TestReceiveBufferAutoTuning we now send a keep-alive packet to measure the
current window rather than a 1 byte segment as the returned window value in the
latter case is reduced due to the 1 byte segment now being held in the receive
buffer and can cause the test to flake if the segment overheads were to change.
In getSendParams in rcv.go we were advertising a non-zero window even if
available window space was zero after we received the previous segment. In such
a case newWnd and curWnd will be the same and we end up advertising a tiny but
non-zero window and this can cause the next segment to be dropped.
PiperOrigin-RevId: 334314070
|
|
When the socket is set with SO_LINGER and close()'d in the initial state, it
should not linger and return immediately.
PiperOrigin-RevId: 334263149
|
|
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
|
