| Age | Commit message (Collapse) | Author |
|---|
|
Also makes the behavior of raw sockets WRT fragmentation clearer, and makes the
ICMPv4 header-length check explicit.
Fixes #3160.
PiperOrigin-RevId: 380033450
|
|
Fixes #3159.
PiperOrigin-RevId: 379814096
|
|
This change prepares for a later change which actually handles the
Prf field in RAs to discover default routers with preference values,
as per RFC 4191.
Updates #6172.
Test: header_test.TestNDPRouterAdvert
PiperOrigin-RevId: 379421710
|
|
This change prepares for a later change which supports the NDP
Route Information option to discover more-specific routes, as
per RFC 4191.
Updates #6172.
PiperOrigin-RevId: 379361330
|
|
There are many references to unimplemented iptables features that link to #170,
but that bug is about Istio support specifically. Istio is supported, so the
references should change.
Some TODOs are addressed, some removed because they are not features requested
by users, and some are left as implementation notes.
Fixes #170.
PiperOrigin-RevId: 379328488
|
|
Before this change, the NDPDispatcher was allowed to "cancel" the
discovery of default routers/prefixes and auto-generate addresses.
No use case exists for this today so we drop this for now. If a
use case comes up in the future, we should instead invalidate the
discovered configuration through the stack instead of during
discovery.
PiperOrigin-RevId: 379327009
|
|
PiperOrigin-RevId: 378506076
|
|
Running multiple instances of netstack in the same network namespace can
cause collisions when enabling packet fanout for fdbased endpoints. The
only bulletproof fix is to run in different network namespaces, but by
using `getpid()` instead of 0 as the fanout ID starting point we can
avoid collisions in the common case, particularly when
testing/experimenting.
Addresses #6124
|
|
...as address add/removal updates multicast group memberships and NDP
state.
This partially reverts the change made to the IPv6 endpoint in
https://github.com/google/gvisor/commit/ebebb3059f7c5dbe42af85715f1c51c.
PiperOrigin-RevId: 378061726
|
|
As per https://linux.die.net/man/8/iptables,
```
Parameters
-i, --in-interface [!] name
Name of an interface via which a packet was received (only for
packets entering the INPUT, FORWARD and PREROUTING chains).
```
Before this change, iptables would use the NIC that a packet was
delivered to after forwarding a packet locally (when forwarding is
enabled) instead of the NIC the packet arrived at.
Updates #170, #3549.
Test: iptables_test.TestInputHookWithLocalForwarding
PiperOrigin-RevId: 377714971
|
|
If the ACK completing the handshake has FIN or data, requeue the segment
for further processing by the newly established endpoint. Otherwise,
the segments would have to be retransmitted by the peer to be processed
by the established endpoint. Doing this, keeps the behavior in parity
with Linux.
This also addresses a test flake with TCPNonBlockingConnectClose where
the ACK (completing the handshake) and multiple retransmitted FINACKs
from the peer could be dropped by the listener, when using syncookies
and the accept queue is full. The handshake could eventually get
completed with a retransmitted FINACK, without actual processing of
FIN. This can cause the poll with POLLRDHUP on the accepted socket to
sometimes time out before the next FINACK retransmission.
PiperOrigin-RevId: 377651695
|
|
Address a race with non-blocking connect and socket close, causing the
FIN (because of socket close) to not be sent out, even after completing
the handshake.
The race occurs with this sequence:
(1) endpoint Connect starts handshake, sending out SYN
(2) handshake complete() releases endpoint lock, waiting on sleeper.Fetch()
(3) endpoint Close acquires endpoint lock, does not enqueue FIN (as the
endpoint is not yet connected) and asserts notifyClose
(4) SYNACK from peer gets enqueued asserting newSegmentWaker
(5) handshake complete() re-aqcuires lock, first processes newSegmentWaker
event, transitions to ESTABLISHED and proceeds to protocolMainLoop()
(6) protocolMainLoop() exits while processing notifyClose
When the execution follows the above sequence, no FIN is sent to the peer.
This causes the listener side to have a half-open connection sitting in
the accept queue.
Fix this by ensuring that the protocolMainLoop() performs clean shutdown
when the endpoint state is still ESTABLISHED.
This would not be a bug, if during handshake complete(), sleeper.Fetch()
prioritized notificationWaker over newSegmentWaker. In that case, the
handshake would not have completed in (5) above.
Fixes #6067
PiperOrigin-RevId: 376994395
|
|
The current implementation has a bug where TCP listener does not ignore
RSTs from the peer. While handling RST+ACK from the peer, this bug can
complete handshakes that use syncookies. This results in half-open
connection delivered to the accept queue.
Fixes #6076
PiperOrigin-RevId: 376868749
|
|
- Typos
- Unused arguments
- Useless conversions
PiperOrigin-RevId: 376362730
|
|
Adds support for the SO_BINDTODEVICE socket option in ICMP sockets with an
accompanying packetimpact test to exercise use of this socket option.
Adds a unit test to exercise the NIC selection logic introduced by this change.
The remaining unit tests for ICMP sockets need to be added in a subsequent CL.
See https://gvisor.dev/issues/5623 for the list of remaining unit tests.
Adds a "timeout" field to PacketimpactTestInfo, necessary due to the long
runtime of the newly added packetimpact test.
Fixes #5678
Fixes #4896
Updates #5623
Updates #5681
Updates #5763
Updates #5956
Updates #5966
Updates #5967
PiperOrigin-RevId: 376271581
|
|
Testing only TestBindToDeviceDistribution decreased from 24s to 11s, and with
TSAN from 186s to 21s.
Note: using `t.Parallel()` actually slows the test down.
PiperOrigin-RevId: 376251420
|
|
...except TCP tests and NDP tests that mutate globals. These will be
undertaken later.
Updates #5940.
PiperOrigin-RevId: 376145608
|
|
- Don't shadow package name
- Don't defer in a loop
- Remove unnecessary type conversion
PiperOrigin-RevId: 376137822
|
|
...except in tests.
Note this replaces some uses of a cryptographic RNG with a plain RNG.
PiperOrigin-RevId: 376070666
|
|
PiperOrigin-RevId: 376022495
|
|
PiperOrigin-RevId: 376001032
|
|
PiperOrigin-RevId: 375977977
|
|
...and use manual clocks in forwarding and link resolution tests.
Fixes #5141.
Fixes #6012.
PiperOrigin-RevId: 375939167
|
|
Updates #5939.
Updates #6012.
RELNOTES: n/a
PiperOrigin-RevId: 375931554
|
|
Introduce tcpip.MonotonicTime; replace int64 in tcpip.Clock method
returns with time.Time and MonotonicTime to improve type safety and
ensure that monotonic clock readings are never compared to wall clock
readings.
PiperOrigin-RevId: 375775907
|
|
Remove redundant interface.
PiperOrigin-RevId: 375756254
|
|
PiperOrigin-RevId: 375728461
|
|
PiperOrigin-RevId: 375705200
|
|
Use it everywhere.
PiperOrigin-RevId: 375539262
|
|
PiperOrigin-RevId: 375507298
|
|
PiperOrigin-RevId: 375490676
|
|
Signed-off-by: Koichi Shiraishi <zchee.io@gmail.com>
|
|
- Unused constants
- Unused functions
- Unused arguments
- Unkeyed literals
- Unnecessary conversions
PiperOrigin-RevId: 375253464
|
|
This change also includes miscellaneous improvements:
* UnknownProtocolRcvdPackets has been separated into two stats, to
specify at which layer the unknown protocol was found (L3 or L4)
* MalformedRcvdPacket is not aggregated across every endpoint anymore.
Doing it this way did not add useful information, and it was also error-prone
(example: ipv6 forgot to increment this aggregated stat, it only
incremented its own ipv6.MalformedPacketsReceived). It is now only incremented
the NIC.
* Removed TestStatsString test which was outdated and had no real
utility.
PiperOrigin-RevId: 375057472
|
|
Add missing protocol state to TCPINFO struct and update packetimpact.
This re-arranges the TCP state definitions to align with Linux.
Fixes #478
PiperOrigin-RevId: 374996751
|
|
Before this change, we would silently drop packets when link resolution
failed. This change brings us into line with RFC 792 (IPv4) and RFC 4443 (IPv6),
both of which specify that gateways should return an ICMP error to the sender
when link resolution fails.
PiperOrigin-RevId: 374699789
|
|
Fuchsia was unable to build when building netstack transitively depended on
golang.org/x/unix constants not defined in Fuchsia. The packages causing this
(safemem and usermem) are no longer in the allowlist.
Tested that this failed at cl/373651666, and passes now that the dependency has
been removed.
PiperOrigin-RevId: 374570220
|
|
PiperOrigin-RevId: 374545882
|
|
PiperOrigin-RevId: 374464969
|
|
|
|
Previously, we named domain objects using numbers (e.g. "e1", "e2" etc). This
change renames objects to clarify whether they are part of the incoming or
outgoing path.
PiperOrigin-RevId: 374226859
|
|
Make sure that the initial configurations used by the DAD state is
valid.
Before this change, an invalid DAD configuration (with a zero-valued
retransmit timer) was used so the DAD state would attempt to resolve
DAD immediately.
This lead to a deadlock in TestDADResolve as when DAD resolves, the
stack notifies the NDP dispatcher which would attempt to write to an
unbuffered channel while holding a lock. The test goroutine also
attempts to obtain a stack.Route (before receiving from the channel)
which ends up attempting to take the same lock.
Test: stack_test.TestDADResolve
PiperOrigin-RevId: 373888540
|
|
...instead of per NetworkProtocol to better conform with linux
(https://www.kernel.org/doc/Documentation/networking/ip-sysctl.txt):
```
conf/interface/*
forwarding - BOOLEAN
Enable IP forwarding on this interface. This controls whether packets
received _on_ this interface can be forwarded.
```
Fixes #5932.
PiperOrigin-RevId: 373888000
|
|
headerOffset() is incorrectly taking account of previous push(), so it thinks
there is more data to consume. This change switches to use pk.reserved as
pivot point.
Reported-by: syzbot+64fef9acd509976f9ce7@syzkaller.appspotmail.com
PiperOrigin-RevId: 373846283
|
|
This change updates the forwarding path to perform the forwarding hook
with iptables so that the filter table is consulted before a packet is
forwarded
Updates #170.
Test: iptables_test.TestForwardingHook
PiperOrigin-RevId: 373702359
|
|
When recovering from a zero-receive-window situation, and asked to
send out an ACK, ensure that we apply SWS avoidance in our window
updates.
Fixes #5984
PiperOrigin-RevId: 373689578
|
|
Benchmark iperf3:
Before After
native->runsc 5.14 5.01 (Gbps)
runsc->native 4.15 4.07 (Gbps)
It did introduce overhead, mainly at the bridge between pkg/buffer and
VectorisedView, the ExtractVV method. Once endpoints start migrating away from
VV, this overhead will be gone.
Updates #2404
PiperOrigin-RevId: 373651666
|
|
...to make it clear to callers that all interfaces are updated with the
forwarding flag and that future NICs will be created with the new
forwarding state.
PiperOrigin-RevId: 373618435
|
|
Before this change, we would silently drop packets when the packet was too
big to be sent out through the NIC (and, for IPv4 packets, if DF was set).
This change brings us into line with RFC 792 (IPv4) and RFC 4443 (IPv6),
both of which specify that gateways should return an ICMP error to the sender
when the packet can't be fragmented.
PiperOrigin-RevId: 373480078
|
|
This code path is for outgoing packets, and we don't currently do memory
accounting on this path. So it wasn't breaking anything.
This change did not add a test for ref-counting issue fixed, but will switch to
the leak-checking ref-counter later when all ref-counting issues are fixed.
PiperOrigin-RevId: 373447913
|
