Commit Graph

84 Commits (9939374c48aff28ea9bee63a749869312d0954ef)

Author SHA1 Message Date
Jordan Whited f0230ce0b5
go.mod,net/tstun,wgengine/netstack: implement gVisor TCP GRO for Linux (#12921)
This commit implements TCP GRO for packets being written to gVisor on
Linux. Windows support will follow later. The wireguard-go dependency is
updated in order to make use of newly exported IP checksum functions.
gVisor is updated in order to make use of newly exported
stack.PacketBuffer GRO logic.

TCP throughput towards gVisor, i.e. TUN write direction, is dramatically
improved as a result of this commit. Benchmarks show substantial
improvement, sometimes as high as 2x. High bandwidth-delay product
paths remain receive window limited, bottlenecked by gVisor's default
TCP receive socket buffer size. This will be addressed in a  follow-on
commit.

The iperf3 results below demonstrate the effect of this commit between
two Linux computers with i5-12400 CPUs. There is roughly ~13us of round
trip latency between them.

The first result is from commit 57856fc without TCP GRO.

Starting Test: protocol: TCP, 1 streams, 131072 byte blocks
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  4.77 GBytes  4.10 Gbits/sec   20 sender
[  5]   0.00-10.00  sec  4.77 GBytes  4.10 Gbits/sec      receiver

The second result is from this commit with TCP GRO.

Starting Test: protocol: TCP, 1 streams, 131072 byte blocks
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  10.6 GBytes  9.14 Gbits/sec   20 sender
[  5]   0.00-10.00  sec  10.6 GBytes  9.14 Gbits/sec      receiver

Updates #6816

Signed-off-by: Jordan Whited <jordan@tailscale.com>
4 months ago
Jordan Whited 7bc2ddaedc
go.mod,net/tstun,wgengine/netstack: implement gVisor TCP GSO for Linux (#12869)
This commit implements TCP GSO for packets being read from gVisor on
Linux. Windows support will follow later. The wireguard-go dependency is
updated in order to make use of newly exported GSO logic from its tun
package.

A new gVisor stack.LinkEndpoint implementation has been established
(linkEndpoint) that is loosely modeled after its predecessor
(channel.Endpoint). This new implementation supports GSO of monster TCP
segments up to 64K in size, whereas channel.Endpoint only supports up to
32K. linkEndpoint will also be required for GRO, which will be
implemented in a follow-on commit.

TCP throughput from gVisor, i.e. TUN read direction, is dramatically
improved as a result of this commit. Benchmarks show substantial
improvement through a wide range of RTT and loss conditions, sometimes
as high as 5x.

The iperf3 results below demonstrate the effect of this commit between
two Linux computers with i5-12400 CPUs. There is roughly ~13us of round
trip latency between them.

The first result is from commit 57856fc without TCP GSO.

Starting Test: protocol: TCP, 1 streams, 131072 byte blocks
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  2.51 GBytes  2.15 Gbits/sec  154 sender
[  5]   0.00-10.00  sec  2.49 GBytes  2.14 Gbits/sec      receiver

The second result is from this commit with TCP GSO.

Starting Test: protocol: TCP, 1 streams, 131072 byte blocks
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  12.6 GBytes  10.8 Gbits/sec    6 sender
[  5]   0.00-10.00  sec  12.6 GBytes  10.8 Gbits/sec      receiver

Updates #6816

Signed-off-by: Jordan Whited <jordan@tailscale.com>
4 months ago
Jordan Whited f0b9d3f477
net/tstun: fix docstring for Wrapper.SetWGConfig (#12796)
Updates #cleanup

Signed-off-by: Jordan Whited <jordan@tailscale.com>
5 months ago
Brad Fitzpatrick c6af5bbfe8 all: add test for package comments, fix, add comments as needed
Updates #cleanup

Change-Id: Ic4304e909d2131a95a38b26911f49e7b1729aaef
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 months ago
Maisem Ali 2238ca8a05 go.mod: bump bart
Updates #bart

Signed-off-by: Maisem Ali <maisem@tailscale.com>
5 months ago
Maisem Ali 1f51bb6891 net/tstun: do SNAT after filterPacketOutboundToWireGuard
In a configuration where the local node (ip1) has a different IP (ip2)
that it uses to communicate with a peer (ip3) we would do UDP flow
tracking on the `ip2->ip3` tuple. When we receive the response from
the peer `ip3->ip2` we would dnat it back to `ip3->ip1` which would
then not match the flow track state and the packet would get dropped.

To fix this, we should do flow tracking on the `ip1->ip3` tuple instead
of `ip2->ip3` which requires doing SNAT after the running filterPacketOutboundToWireGuard.

Updates tailscale/corp#19971, tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
7 months ago
Maisem Ali af97e7a793 tailcfg,all: add/plumb Node.IsJailed
This adds a new bool that can be sent down from control
to do jailing on the client side. Previously this would
only be done from control by modifying the packet filter
we sent down to clients. This would result in a lot of
additional work/CPU on control, we could instead just
do this on the client. This has always been a TODO which
we keep putting off, might as well do it now.

Updates tailscale/corp#19623

Signed-off-by: Maisem Ali <maisem@tailscale.com>
7 months ago
Maisem Ali e67069550b ipn/ipnlocal,net/tstun,wgengine: create and plumb jailed packet filter
This plumbs a packet filter for jailed nodes through to the
tstun.Wrapper; the filter for a jailed node is equivalent to a "shields
up" filter. Currently a no-op as there is no way for control to
tell the client whether a peer is jailed.

Updates tailscale/corp#19623

Co-authored-by: Andrew Dunham <andrew@du.nham.ca>
Signed-off-by: Maisem Ali <maisem@tailscale.com>
Change-Id: I5ccc5f00e197fde15dd567485b2a99d8254391ad
7 months ago
Maisem Ali 5ef178fdca net/tstun: refactor peerConfig to allow storing more details
This refactors the peerConfig struct to allow storing more
details about a peer and not just the masq addresses. To be
used in a follow up change.

As a side effect, this also makes the DNAT logic on the inbound
packet stricter. Previously it would only match against the packets
dst IP, not it also takes the src IP into consideration. The beahvior
is at parity with the SNAT case.

Updates tailscale/corp#19623

Co-authored-by: Andrew Dunham <andrew@du.nham.ca>
Signed-off-by: Maisem Ali <maisem@tailscale.com>
Change-Id: I5f40802bebbf0f055436eb8824e4511d0052772d
7 months ago
Andrew Dunham be663c84c1 net/tstun: rename natConfig to peerConfig
So that we can use this for additional, non-NAT configuration without it
being confusing.

Updates #cleanup

Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I1658d59c9824217917a94ee76d2d08f0a682986f
7 months ago
Andrew Dunham 10497acc95 net/tstun: refactor natConfig to not be per-family
This was a holdover from the older, pre-BART days and is no longer
necessary.

Updates #cleanup

Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I71b892bab1898077767b9ff51cef33d59c08faf8
7 months ago
Percy Wegmann 8b8b315258 net/tstun: use gaissmai/bart instead of tempfork/device
This implementation uses less memory than tempfork/device,
which helps avoid OOM conditions in the iOS VPN extension when
switching to a Tailnet with ExitNode routing enabled.

Updates tailscale/corp#18514

Signed-off-by: Percy Wegmann <percy@tailscale.com>
8 months ago
Andrew Dunham 62cf83eb92 go.mod: bump gvisor
The `stack.PacketBufferPtr` type no longer exists; replace it with
`*stack.PacketBuffer` instead.

Updates #8043

Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ib56ceff09166a042aa3d9b80f50b2aa2d34b3683
9 months ago
Andrew Dunham 3dd8ae2f26 net/tstun: fix spelling of "WireGuard"
Updates #cleanup

Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ida7e30f4689bc18f5f7502f53a0adb5ac3c7981a
9 months ago
Andrew Lytvynov 2716250ee8
all: cleanup unused code, part 2 (#10670)
And enable U1000 check in staticcheck.

Updates #cleanup

Signed-off-by: Andrew Lytvynov <awly@tailscale.com>
11 months ago
Maisem Ali 5297bd2cff cmd/tailscaled,net/tstun: fix data race on start-up in TUN mode
Fixes #7894

Change-Id: Ice3f8019405714dd69d02bc07694f3872bb598b8

Co-authored-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Signed-off-by: Maisem Ali <maisem@tailscale.com>
1 year ago
Maisem Ali 9d96e05267 net/packet: split off checksum munging into different pkg
The current structure meant that we were embedding netstack in
the tailscale CLI and in the GUIs. This removes that by isolating
the checksum munging to a different pkg which is only called from
`net/tstun`.

Fixes #9756

Signed-off-by: Maisem Ali <maisem@tailscale.com>
1 year ago
Maisem Ali 78a083e144 types/ipproto: drop IPProto from IPProtoVersion
Based on https://github.com/golang/go/wiki/CodeReviewComments#package-names.

Updates #cleanup

Signed-off-by: Maisem Ali <maisem@tailscale.com>
1 year ago
Tom DNetto da1b917575 net/tstun: finish wiring IPv6 NAT support
Updates https://github.com/tailscale/corp/issues/11202
Updates ENG-991
Signed-off-by: Tom DNetto <tom@tailscale.com>
1 year ago
Maisem Ali 3056a98bbd net/tstun: add better logging of natV4Config
It might as well have been spewing out gibberish. This adds
a nicer output format for us to be able to read and identify
whats going on.

Sample output
```
natV4Config{nativeAddr: 100.83.114.95, listenAddrs: [10.32.80.33], dstMasqAddrs: [10.32.80.33: 407 peers]}
```

Fixes tailscale/corp#14650

Signed-off-by: Maisem Ali <maisem@tailscale.com>
1 year ago
Tom DNetto c08cf2a9c6 all: declare & plumb IPv6 masquerade address for peer
This PR plumbs through awareness of an IPv6 SNAT/masquerade address from the wire protocol
through to the low-level (tstun / wgengine). This PR is the first in two PRs for implementing
IPv6 NAT support to/from peers.

A subsequent PR will implement the data-plane changes to implement IPv6 NAT - this is just plumbing.

Signed-off-by: Tom DNetto <tom@tailscale.com>
Updates ENG-991
1 year ago
Brad Fitzpatrick dc7aa98b76 all: use set.Set consistently instead of map[T]struct{}
I didn't clean up the more idiomatic map[T]bool with true values, at
least yet.  I just converted the relatively awkward struct{}-valued
maps.

Updates #cleanup

Change-Id: I758abebd2bb1f64bc7a9d0f25c32298f4679c14f
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
1 year ago
Brad Fitzpatrick e8551d6b40 all: use Go 1.21 slices, maps instead of x/exp/{slices,maps}
Updates #8419

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
1 year ago
Maisem Ali 3ae7140690 net/tstun: handle exit nodes in NAT configs
In the case where the exit node requires SNAT, we would SNAT all traffic not just the
traffic meant to go through the exit node. This was a result of the default route being
added to the routing table which would match basically everything.

In this case, we need to account for all peers in the routing table not just the ones
that require NAT.

Fix and add a test.

Updates tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Andrew Dunham 04a3118d45 net/tstun: add tests for captureHook
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I630f852d9f16c951c721b34f2bc4128e68fe9475
2 years ago
Maisem Ali 64bbf1738e tailcfg: make SelfNodeV4MasqAddrForThisPeer a pointer
This makes `omitempty` actually work, and saves bytes in each map response.

Updates tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Tom DNetto 6a627e5a33 net, wgengine/capture: encode NAT addresses in pcap stream
Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Maisem Ali 985535aebc net/tstun,wgengine/*: add support for NAT to routes
This adds support to make exit nodes and subnet routers work
when in scenarios where NAT is required.

It also updates the NATConfig to be generated from a `wgcfg.Config` as
that handles merging prefs with the netmap, so it has the required information
about whether an exit node is already configured and whether routes are accepted.

Updates tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Maisem Ali d2fd101eb4 net/tstun: only log natConfig on changes
Updates tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Maisem Ali 0bf8c8e710 net/tstun: use p.Buffer() in more places
Updates tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Maisem Ali bb31fd7d1c net/tstun: add inital support for NAT v4
This adds support in tstun to utitilize the SelfNodeV4MasqAddrForThisPeer and
perform the necessary modifications to the packet as it passes through tstun.

Currently this only handles ICMP, UDP and TCP traffic.
Subnet routers and Exit Nodes are also unsupported.

Updates tailscale/corp#8020

Co-authored-by: Melanie Warrick <warrick@tailscale.com>
Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Maisem Ali 535fad16f8 net/tstun: rename filterIn/filterOut methods to be more descriptive
Updates tailscale/corp#8020

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Tom DNetto 2ca6dd1f1d wgengine: start logging DISCO frames to pcap stream
Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Tom DNetto 99b9d7a621 all: implement pcap streaming for datapath debugging
Updates: tailscale/corp#8470

Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Will Norris 71029cea2d all: update copyright and license headers
This updates all source files to use a new standard header for copyright
and license declaration.  Notably, copyright no longer includes a date,
and we now use the standard SPDX-License-Identifier header.

This commit was done almost entirely mechanically with perl, and then
some minimal manual fixes.

Updates #6865

Signed-off-by: Will Norris <will@tailscale.com>
2 years ago
andig 14e8afe444 go.mod, etc: bump gvisor
Fixes #6554

Change-Id: Ia04ae37a47b67fa57091c9bfe1d45a1842589aa8
Signed-off-by: andig <cpuidle@gmx.de>
2 years ago
Jordan Whited 55b24009f7
net/tstun: don't return early from a partial tun.Read() (#6745)
Fixes #6730

Signed-off-by: Jordan Whited <jordan@tailscale.com>
2 years ago
Jordan Whited ea5ee6f87c
all: update golang.zx2c4.com/wireguard to github.com/tailscale/wireguard-go (#6692)
This is temporary while we work to upstream performance work in
https://github.com/WireGuard/wireguard-go/pull/64. A replace directive
is less ideal as it breaks dependent code without duplication of the
directive.

Signed-off-by: Jordan Whited <jordan@tailscale.com>
2 years ago
Maisem Ali eb1adf629f net/tstun: reuse buffered packet from pool
We would call parsedPacketPool.Get() for all packets received in Read/Write.
This was wasteful and not necessary, fetch a single *packet.Parsed for
all packets.

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Jordan Whited 76389d8baf
net/tstun, wgengine/magicsock: enable vectorized I/O on Linux (#6663)
This commit updates the wireguard-go dependency and implements the
necessary changes to the tun.Device and conn.Bind implementations to
support passing vectors of packets in tailscaled. This significantly
improves throughput performance on Linux.

Updates #414

Signed-off-by: Jordan Whited <jordan@tailscale.com>
Signed-off-by: James Tucker <james@tailscale.com>
Co-authored-by: James Tucker <james@tailscale.com>
2 years ago
Joe Tsai 2e5d08ec4f
net/connstats: invert network logging data flow (#6272)
Previously, tstun.Wrapper and magicsock.Conn managed their
own statistics data structure and relied on an external call to
Extract to extract (and reset) the statistics.
This makes it difficult to ensure a maximum size on the statistics
as the caller has no introspection into whether the number
of unique connections is getting too large.

Invert the control flow such that a *connstats.Statistics
is registered with tstun.Wrapper and magicsock.Conn.
Methods on non-nil *connstats.Statistics are called for every packet.
This allows the implementation of connstats.Statistics (in the future)
to better control when it needs to flush to ensure
bounds on maximum sizes.

The value registered into tstun.Wrapper and magicsock.Conn could
be an interface, but that has two performance detriments:

1. Method calls on interface values are more expensive since
they must go through a virtual method dispatch.

2. The implementation would need a sync.Mutex to protect the
statistics value instead of using an atomic.Pointer.

Given that methods on constats.Statistics are called for every packet,
we want reduce the CPU cost on this hot path.

Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2 years ago
Brad Fitzpatrick f4ff26f577 types/pad32: delete package
Use Go 1.19's new 64-bit alignment ~hidden feature instead.

Fixes #5356

Change-Id: Ifcbcb115875a7da01df3bc29e9e7feadce5bc956
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2 years ago
Joe Tsai c21a3c4733
types/netlogtype: new package for network logging types (#6092)
The netlog.Message type is useful to depend on from other packages,
but doing so would transitively cause gvisor and other large packages
to be linked in.

Avoid this problem by moving all network logging types to a single package.

We also update staticcheck to take in:

	003d277bcf

Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2 years ago
Joe Tsai 1b4e4cc1e8
wgengine/netlog: new package for traffic flow logging (#5864)
The Logger type managers a logtail.Logger for extracting
statistics from a tstun.Wrapper.
So long as Shutdown is called, it ensures that logtail
and statistic gathering resources are properly cleared up.

Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2 years ago
Joe Tsai 84e8f25c21
net/tstun: rename statististics method (#5852)
Rename StatisticsEnable as SetStatisticsEnabled to be consistent
with other similarly named methods.

Rename StatisticsExtract as ExtractStatistics to follow
the convention where methods start with a verb.
It was originally named with Statistics as a prefix so that
statistics related methods would sort well in godoc,
but that property no longer holds.

Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2 years ago
Joe Tsai 24ebf161e8
net/tstun: instrument Wrapper with statistics gathering (#5847)
If Wrapper.StatisticsEnable is enabled,
then per-connection counters are maintained.
If enabled, Wrapper.StatisticsExtract must be periodically called
otherwise there is unbounded memory growth.

Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2 years ago
Josh Soref d4811f11a0 all: fix spelling mistakes
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2 years ago
Maisem Ali a9f6cd41fd all: use syncs.AtomicValue
Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Maisem Ali 9bb5a038e5 all: use atomic.Pointer
Also add some missing docs.

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2 years ago
Brad Fitzpatrick a12aad6b47 all: convert more code to use net/netip directly
perl -i -npe 's,netaddr.IPPrefixFrom,netip.PrefixFrom,' $(git grep -l -F netaddr.)
    perl -i -npe 's,netaddr.IPPortFrom,netip.AddrPortFrom,' $(git grep -l -F netaddr. )
    perl -i -npe 's,netaddr.IPPrefix,netip.Prefix,g' $(git grep -l -F netaddr. )
    perl -i -npe 's,netaddr.IPPort,netip.AddrPort,g' $(git grep -l -F netaddr. )
    perl -i -npe 's,netaddr.IP\b,netip.Addr,g' $(git grep -l -F netaddr. )
    perl -i -npe 's,netaddr.IPv6Raw\b,netip.AddrFrom16,g' $(git grep -l -F netaddr. )
    goimports -w .

Then delete some stuff from the net/netaddr shim package which is no
longer neeed.

Updates #5162

Change-Id: Ia7a86893fe21c7e3ee1ec823e8aba288d4566cd8
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2 years ago