aboutsummaryrefslogtreecommitdiff
path: root/tun/tun_linux.go
AgeCommit message (Collapse)Author
2023-12-11tun: implement UDP GSO/GRO for LinuxJordan Whited
Implement UDP GSO and GRO for the Linux tun.Device, which is made possible by virtio extensions in the kernel's TUN driver starting in v6.2. secnetperf, a QUIC benchmark utility from microsoft/msquic@8e1eb1a, is used to demonstrate the effect of this commit between two Linux computers with i5-12400 CPUs. There is roughly ~13us of round trip latency between them. secnetperf was invoked with the following command line options: -stats:1 -exec:maxtput -test:tput -download:10000 -timed:1 -encrypt:0 The first result is from commit 2e0774f without UDP GSO/GRO on the TUN. [conn][0x55739a144980] STATS: EcnCapable=0 RTT=3973 us SendTotalPackets=55859 SendSuspectedLostPackets=61 SendSpuriousLostPackets=59 SendCongestionCount=27 SendEcnCongestionCount=0 RecvTotalPackets=2779122 RecvReorderedPackets=0 RecvDroppedPackets=0 RecvDuplicatePackets=0 RecvDecryptionFailures=0 Result: 3654977571 bytes @ 2922821 kbps (10003.972 ms). The second result is with UDP GSO/GRO on the TUN. [conn][0x56493dfd09a0] STATS: EcnCapable=0 RTT=1216 us SendTotalPackets=165033 SendSuspectedLostPackets=64 SendSpuriousLostPackets=61 SendCongestionCount=53 SendEcnCongestionCount=0 RecvTotalPackets=11845268 RecvReorderedPackets=25267 RecvDroppedPackets=0 RecvDuplicatePackets=0 RecvDecryptionFailures=0 Result: 15574671184 bytes @ 12458214 kbps (10001.222 ms). Signed-off-by: Jordan Whited <jordan@tailscale.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-03-17tun: replace ErrorBatch() with errors.Join()Jordan Whited
Reviewed-by: Maisem Ali <maisem@tailscale.com> Signed-off-by: Jordan Whited <jordan@tailscale.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-03-13global: buff -> bufJason A. Donenfeld
This always struck me as kind of weird and non-standard. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-03-10conn: inch BatchSize toward being non-dynamicJason A. Donenfeld
There's not really a use at the moment for making this configurable, and once bind_windows.go behaves like bind_std.go, we'll be able to use constants everywhere. So begin that simplification now. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-03-10conn, device, tun: implement vectorized I/O on LinuxJordan Whited
Implement TCP offloading via TSO and GRO for the Linux tun.Device, which is made possible by virtio extensions in the kernel's TUN driver. Delete conn.LinuxSocketEndpoint in favor of a collapsed conn.StdNetBind. conn.StdNetBind makes use of recvmmsg() and sendmmsg() on Linux. All platforms now fall under conn.StdNetBind, except for Windows, which remains in conn.WinRingBind, which still needs to be adjusted to handle multiple packets. Also refactor sticky sockets support to eventually be applicable on platforms other than just Linux. However Linux remains the sole platform that fully implements it for now. Co-authored-by: James Tucker <james@tailscale.com> Signed-off-by: James Tucker <james@tailscale.com> Signed-off-by: Jordan Whited <jordan@tailscale.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-03-10conn, device, tun: implement vectorized I/O plumbingJordan Whited
Accept packet vectors for reading and writing in the tun.Device and conn.Bind interfaces, so that the internal plumbing between these interfaces now passes a vector of packets. Vectors move untouched between these interfaces, i.e. if 128 packets are received from conn.Bind.Read(), 128 packets are passed to tun.Device.Write(). There is no internal buffering. Currently, existing implementations are only adjusted to have vectors of length one. Subsequent patches will improve that. Also, as a related fixup, use the unix and windows packages rather than the syscall package when possible. Co-authored-by: James Tucker <james@tailscale.com> Signed-off-by: James Tucker <james@tailscale.com> Signed-off-by: Jordan Whited <jordan@tailscale.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-02-09tun: guard Device.Events() against chan writesJordan Whited
Signed-off-by: Jordan Whited <jordan@tailscale.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2023-02-07global: bump copyright yearJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2022-09-20global: bump copyright yearJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2022-07-04conn, device, tun: set CLOEXEC on fdsBrad Fitzpatrick
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2022-06-01tun: use ByteSliceToString from golang.org/x/sys/unixTobias Klauser
Use unix.ByteSliceToString in (*NativeTun).nameSlice to convert the TUNGETIFF ioctl result []byte to a string. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-12-09global: apply gofumptJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-09-23tun: avoid leaking sock fd in CreateTUN error casesTobias Klauser
At these points, the socket file descriptor is not yet wrapped in an *os.File, so it needs to be closed explicitly on error. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-05-20tun: linux: account for interface removal from outsideJason A. Donenfeld
On Linux we can run `ip link del wg0`, in which case the fd becomes stale, and we should exit. Since this is an intentional action, don't treat it as an error. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-03-11tun: linux: do not spam events every second from hack listenerJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-02-22tun: make NativeTun.Close well behaved, not crash on double closeBrad Fitzpatrick
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-02-09rwcancel: add an explicit close callJason A. Donenfeld
This lets us collect FDs even if the GC doesn't do it for us. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-01-28global: bump copyrightJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2021-01-27tun: use %w for errors on linuxJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2020-05-02global: update header comments and modulesJason A. Donenfeld
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2020-05-02tun: return a better error message if /dev/net/tun doesn't existBrad Fitzpatrick
It was just returning "no such file or directory" (the String of the syscall.Errno returned by CreateTUN). Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-05-02tun: NetlinkListener: don't send EventDown before sending EventUpAvery Pennarun
This works around a startup race condition when competing with HackListener, which is trying to do the same job. If HackListener detects that the tundev is running while there is still an event in the netlink queue that says it isn't running, then the device receives a string of events like EventUp (HackListener) EventDown (NetlinkListener) EventUp (NetlinkListener) Unfortunately, after the first EventDown, the device stops itself, thinking incorrectly that the administrator has downed its tundev. The device is ignoring the initial EventDown anyway, so just don't emit it. Signed-off-by: Avery Pennarun <apenwarr@tailscale.com>
2020-05-02tun: fix data race on name fieldBrad Fitzpatrick
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-05-02tun: remove unused isUp methodBrad Fitzpatrick
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-03-17global: use RTMGRP_* consts from x/sys/unixTobias Klauser
Update the golang.org/x/sys/unix dependency and use the newly introduced RTMGRP_* consts instead of using the corresponding RTNLGRP_* const to create a mask. Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
2019-10-22global: fix a few typos courtesy of codespellJonathan Tooker
Signed-off-by: Jonathan Tooker <jonathan.tooker@netprotect.com>
2019-06-14tun: remove TUN prefix from types to reduce stutter elsewhereMatt Layher
Signed-off-by: Matt Layher <mdlayher@gmail.com>
2019-06-06global: fixup TODO comment spacingJason A. Donenfeld
2019-05-14global: regroup all importsJason A. Donenfeld
2019-03-21receive: implement flush semanticsJason A. Donenfeld
2019-03-07tun: linux: work out netpoll trickJason A. Donenfeld
2019-03-04tun: import mobile particularitiesJason A. Donenfeld
2019-03-01tun: allow special methods in NativeTunJason A. Donenfeld
2019-02-27tun: linux: netpoll is broken for tun's epollJason A. Donenfeld
So this mostly reverts the switch to Sysconn for Linux. Issue: https://github.com/golang/go/issues/30426
2019-02-27tun: linux: netlink sock needs cleaning up but file will be gc'dJason A. Donenfeld
2019-02-27tun: use netpoll instead of rwcancelJason A. Donenfeld
The new sysconn function of Go 1.12 makes this possible: package main import "log" import "os" import "unsafe" import "time" import "syscall" import "sync" import "golang.org/x/sys/unix" func main() { fd, err := os.OpenFile("/dev/net/tun", os.O_RDWR, 0) if err != nil { log.Fatal(err) } var ifr [unix.IFNAMSIZ + 64]byte copy(ifr[:], []byte("cheese")) *(*uint16)(unsafe.Pointer(&ifr[unix.IFNAMSIZ])) = unix.IFF_TUN var errno syscall.Errno s, _ := fd.SyscallConn() s.Control(func(fd uintptr) { _, _, errno = unix.Syscall( unix.SYS_IOCTL, fd, uintptr(unix.TUNSETIFF), uintptr(unsafe.Pointer(&ifr[0])), ) }) if errno != 0 { log.Fatal(errno) } b := [4]byte{} wait := sync.WaitGroup{} wait.Add(1) go func() { _, err := fd.Read(b[:]) log.Print("Read errored: ", err) wait.Done() }() time.Sleep(time.Second) log.Print("Closing") err = fd.Close() if err != nil { log.Print("Close errored: " , err) } wait.Wait() log.Print("Exiting") }
2019-02-27tun: use sysconn instead of .Fd with Go 1.12Jason A. Donenfeld
2019-02-18Change package pathJason A. Donenfeld
2019-02-05Update copyrightJason A. Donenfeld
2018-12-06tun: remove nonblock hack for linuxJason A. Donenfeld
This is no longer necessary and actually breaks things Reported-by: Chris Branch <cbranch@cloudflare.com>
2018-10-17tun: only call .Fd() onceJason A. Donenfeld
Doing so tends to make the tunnel blocking, so we only retrieve it once before we call SetNonblock, and then cache the result.
2018-09-16global: fix up copyright headersJason A. Donenfeld
2018-07-30Fix duplicate copyright lineJason A. Donenfeld
2018-05-24Catch EINTRJason A. Donenfeld
2018-05-23Adopt GOPATHJason A. Donenfeld
GOPATH is annoying, but the Go community pushing me to adopt it is even more annoying.
2018-05-23Move tun to subpackageJason A. Donenfeld
7apu2bq3rhoylwmucxizk54afw5jyda7pho4j5zdcqpwkufke7zdzwad.onion