mirrors
/
tailscale
mirror of https://github.com/tailscale/tailscale.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
2,682 Commits 732 Branches 162 Tags 317 MiB
Commit Graph

2682 Commits

Author SHA1 Message Date
Josh Bleecher Snyder ae2f24ec4e remove some TODOs, make ip address extraction equally awful 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder ba49da429a simplify change point 
at the cost of some very, very naughty unsafe
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 83742afabf convert manual tests into automated tests 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder b668e5d185 tewak 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 5410042dcd start cleaning up code 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 75efd794a3 ipv6 support, UNTESTED, super hacky 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 36a4741bc5 comments 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 208e6eb0db fix io_uring tun device reads 
Well that's an interesting one!
Apparently if you issue multiple concurrent preadv calls on
a TUN device using io_uring, the TUN device falls over.

Possibly corrupting memory along the way.
Which might be why the kernel hung on shutdown...
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder fec66b4100 make it easy to toggle uring on/off 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder eff65381f2 switch file.go to test reads instead of writes 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder a8df9fa7cc simpler init 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 78fd2b7880 stick closer to upstream tun reading code 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 001dec84de make cgo pointer rules happy 
well that was a mess
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder a8a7208dbd use io_uring for sendmsg 
and clean up some dead code and unify some things
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder f254f779b5 disable polling for now 
we'll maybe bring it back with a token bucket or something.
and/or do multi-sqe submission.
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 3d91c5b369 they work with extra junk 
thanks, dave
no thanks, kernel devs

write no work
use writev
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 2a2ed7cd17 tun writes...not working yet (but why not??) 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 55c1ce00be use polling instead of syscall 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 4013c0edbb smush together two return values from C 
avoids a per-packet alloc

i will atone for my sins later

if only C let you return multiple values.
or Go let you pass in a pointer w/o it being on the heap.
 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 783d2d4327 keep 8 requests in the queue at all times 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 71f35bda1a remove all allocs 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder b83ac004f1 remove all allocations 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 59512181b5 mve sockaddr_in to Go 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder 1ee40d1670 move iovec to go 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder bbccf68a76 sheesh 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder d7a7e2d17d move msghdr to Go 2021-07-08 12:56:49 -07:00
Josh Bleecher Snyder f26c0fcbd5 use io_uring 2021-07-08 12:56:49 -07:00
Christine Dodrill 1e83b97498
tstest/integration/vms: outgoing SSH test (#2349) 
This does a few things:

1. Rewrites the tests so that we get a log of what individual tests
   failed at the end of a test run.
2. Adds a test that runs an HTTP server via the tester tailscale node and
   then has the VMs connect to that over Tailscale.
3. Dials the VM over Tailscale and ensures it answers SSH requests.
4. Other minor framework refactoring.

Signed-off-by: Christine Dodrill <xe@tailscale.com>
 2021-07-08 11:38:01 -04:00
Christine Dodrill 97279a0fe0
tstest/integration/vms: add Oracle Linux image (#2328) 
Oracle Linux[1] is a CentOS fork. It is not very special. I am adding it
to the integration jungle because I am adding it to pkgs and the website
directions.

[1]: https://www.oracle.com/linux/

Signed-off-by: Christine Dodrill <xe@tailscale.com>
 2021-07-08 10:26:20 -04:00
Brad Fitzpatrick a9fc583211 cmd/tailscale/cli: document the web subcommand a bit more 
Fixes #2326

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 21:16:33 -07:00
Josh Bleecher Snyder 0ad92b89a6 net/tstun: fix data races 
To remove some multi-case selects, we intentionally allowed
sends on closed channels (cc23049cd2).

However, we also introduced concurrent sends and closes,
which is a data race.

This commit fixes the data race. The mutexes here are uncontended,
and thus very cheap.

Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
 2021-07-07 16:15:29 -07:00
Brad Fitzpatrick 7d417586a8 tstest/integration: help bust cmd/go's test caching 
It was caching too aggressively, as it didn't see our deps due to our
running "go install tailscaled" as a child process.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 13:14:21 -07:00
Brad Fitzpatrick 3dcd18b6c8 tailcfg: note RegionID 900-999 reservation 2021-07-07 12:23:41 -07:00
Brad Fitzpatrick ddb8726c98 util/deephash: don't reflect.Copy if element type is a defined uint8 
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 11:58:04 -07:00
Brad Fitzpatrick df176c82f5 util/deephash: skip alloc test under race detector 
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 11:40:28 -07:00
Brad Fitzpatrick 6dc38ff25c util/deephash: optimize hashing of byte arrays, reduce allocs in Hash 
name              old time/op    new time/op    delta
Hash-6               173µs ± 4%     101µs ± 3%   -41.69%  (p=0.000 n=10+9)
HashMapAcyclic-6     101µs ± 5%     105µs ± 3%    +3.52%  (p=0.001 n=9+10)
TailcfgNode-6       29.4µs ± 2%    16.4µs ± 3%   -44.25%  (p=0.000 n=8+10)

name              old alloc/op   new alloc/op   delta
Hash-6              3.60kB ± 0%    1.13kB ± 0%   -68.70%  (p=0.000 n=10+10)
HashMapAcyclic-6    2.53kB ± 0%    2.53kB ± 0%      ~     (p=0.137 n=10+8)
TailcfgNode-6         528B ± 0%        0B       -100.00%  (p=0.000 n=10+10)

name              old allocs/op  new allocs/op  delta
Hash-6                84.0 ± 0%      40.0 ± 0%   -52.38%  (p=0.000 n=10+10)
HashMapAcyclic-6       202 ± 0%       202 ± 0%      ~     (all equal)
TailcfgNode-6         11.0 ± 0%       0.0       -100.00%  (p=0.000 n=10+10)

Updates tailscale/corp#2130

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 11:30:49 -07:00
Brad Fitzpatrick 3962744450 util/deephash: prevent infinite loop on map cycle 
Fixes #2340

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 10:57:46 -07:00
Brad Fitzpatrick aceaa70b16 util/deephash: move funcs to methods 
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-07 08:17:18 -07:00
Irshad Pananilath 9288e0d61c build_docker.sh: use build_dist.sh to inject version information 
version.sh was removed in commit 5088af68. Use `build_dist.sh shellvars`
to provide version information instead.

Signed-off-by: Irshad Pananilath <pmirshad+code@gmail.com>
 2021-07-07 06:38:04 -07:00
Christine Dodrill a8360050e7
tstest/integration/vms: make first end to end test (#2332) 
This makes sure `tailscale status` and `tailscale ping` works. It also
switches goexpect to use a batch instead of manually banging out each
line, which makes the tests so much easier to read.

Signed-off-by: Christine Dodrill <xe@tailscale.com>
 2021-07-06 12:50:19 -04:00
David Crawshaw 805d5d3cde ipnlocal: move log line inside if statement 
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
 2021-07-06 09:35:01 -07:00
Brad Fitzpatrick 14f901da6d util/deephash: fix sync.Pool usage 
Whoops.

From yesterday's 9ae3bd0939 (not yet
used by anything, fortunately)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-05 22:21:44 -07:00
Brad Fitzpatrick e0258ffd92 util/deephash: use keyed struct literal, fix vet 
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-05 21:31:30 -07:00
Brad Fitzpatrick bf9f279768 util/deephash: optimize CPU a bit by by avoiding fmt in more places 
name              old time/op    new time/op    delta
Hash-6               179µs ± 5%     173µs ± 4%   -3.12%  (p=0.004 n=10+10)
HashMapAcyclic-6     115µs ± 3%     101µs ± 5%  -11.51%  (p=0.000 n=9+9)
TailcfgNode-6       30.8µs ± 4%    29.4µs ± 2%   -4.51%  (p=0.000 n=10+8)

name              old alloc/op   new alloc/op   delta
Hash-6              3.60kB ± 0%    3.60kB ± 0%     ~     (p=0.445 n=9+10)
HashMapAcyclic-6    2.53kB ± 0%    2.53kB ± 0%     ~     (p=0.065 n=9+10)
TailcfgNode-6         528B ± 0%      528B ± 0%     ~     (all equal)

name              old allocs/op  new allocs/op  delta
Hash-6                84.0 ± 0%      84.0 ± 0%     ~     (all equal)
HashMapAcyclic-6       202 ± 0%       202 ± 0%     ~     (all equal)
TailcfgNode-6         11.0 ± 0%      11.0 ± 0%     ~     (all equal)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-05 21:28:54 -07:00
Brad Fitzpatrick 58f2ef6085 util/deephash: add a benchmark and some benchmark data 
No code changes.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-05 21:21:52 -07:00
Brad Fitzpatrick 9ae3bd0939 util/deephash: export a Hash func for use by the control plane 
name              old time/op    new time/op    delta
Hash-6              69.4µs ± 6%    68.4µs ± 4%     ~     (p=0.286 n=9+9)
HashMapAcyclic-6     115µs ± 5%     115µs ± 4%     ~     (p=1.000 n=10+10)

name              old alloc/op   new alloc/op   delta
Hash-6              2.29kB ± 0%    1.88kB ± 0%  -18.13%  (p=0.000 n=10+10)
HashMapAcyclic-6    2.53kB ± 0%    2.53kB ± 0%     ~     (all equal)

name              old allocs/op  new allocs/op  delta
Hash-6                58.0 ± 0%      54.0 ± 0%   -6.90%  (p=0.000 n=10+10)
HashMapAcyclic-6       202 ± 0%       202 ± 0%     ~     (all equal)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
 2021-07-05 11:41:44 -07:00
Brad Fitzpatrick 700badd8f8 util/deephash: move internal/deephash to util/deephash 
No code changes. Just a minor package doc addition about lack of API
stability.
 2021-07-02 21:33:02 -07:00
Josh Bleecher Snyder 7f095617f2 internal/deephash: 8 bits of output is not enough 
Running hex.Encode(b, b) is a bad idea.
The first byte of input will overwrite the first two bytes of output.
Subsequent bytes have no impact on the output.

Not related to today's IPv6 bug, but...wh::ps.

This caused us to spuriously ignore some wireguard config updates.

Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
 2021-07-02 13:48:27 -07:00
Josh Bleecher Snyder c35a832de6 net/tstun: add inner loop to poll 
This avoids re-enqueuing to t.bufferConsumed,
which makes the code a bit clearer.

Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
 2021-07-02 11:02:12 -07:00
Josh Bleecher Snyder a4cc7b6d54 net/tstun: simplify code 
Calculate whether the packet is injected directly,
rather than via an else branch.

Unify the exit paths. It is easier here than duplicating them.

Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
 2021-07-02 11:02:12 -07:00