This allows multiple SOCKS connections to share the available proxies,
and in particular prevents a SOCKS connection from being starved of a
proxy when the maximum proxy capacity is less then the number of the
number of SOCKS connections.
This is option 4 from https://bugs.torproject.org/33519.
The client opts into turbotunnel mode by sending a magic token at the
beginning of each WebSocket connection (before sending even the
ClientID). The token is just a random byte string I generated. The
server peeks at the token and, if it matches, uses turbotunnel mode.
Otherwise, it unreads the token and continues in the old
one-session-per-WebSocket mode.
Formerly we waiting until *both* directions finished. What this meant in
practice is that when the remote connection ended, copyLoop would become
useless but would continue blocking its caller until something else
finally closed the socks connection.
We are no longer checking for nil BrokerChannels in Catch because this
case is caught from the return values of NewBrokerChannel. This change
caused a no longer necessary unit test to hang.
The call was
copyLoop(socks, snowflake)
but the function signature was
func copyLoop(WebRTC, SOCKS io.ReadWriter) {
The mistake was mostly harmless, because both arguments were treated the
same, except that error logs would have reported the wrong direction.
Instead of returning nil from NewBrokerChannel and having
WebRTCDialer.Catch check for nil, let NewBrokerChannel return an error
and bail out before calling WebRTCDialer.Catch.
Suggested by cohosh.
https://bugs.torproject.org/33040#comment:3
These functions are called for their side effect of sending a PT error
message on stdout; they also return a representation of the error
message as an error object for the caller to use if it wishes. These
functions *always* return a non-nil error object; it is not something to
be logged, any more than the return value of errors.New is.
The mistaken logging was added in
https://bugs.torproject.org/31794b26c7a7a733ec9dd19faed3d42e1ec
There were a few tests that needed refreshing since the introduction of
the pion library. Also added a few tests for the ICE server parsing
function in the client.
last was initialised twice (creating a shadow), the second time inside
a case statement. The second initialisation is removed, keeping the use
of last aligned to the isame style as its use other parts of the case
statement.
- Error strings are no longer capitalized nor end with punctuation
- Alias import
- Remove extraneous initilisation code (No need to provide zero value
for variables, because the compiler does that anyway)
We need to set up the pion/webrtc logger to write output to the
snowflake log, otherwise the warnings we are getting from the pion
library are being lost.
Note: this requires go version 1.13 and later in order to use the
`log.Writer()` function.
This commit fixes a small error introduced in a previous commit. Servers
given by command line options weren't being added to the configuration
because we were checking for `iceServers` to be nil instead of not nil.
Modified the snowflake client to use pion/webrtc as the webrtc library.
This involved a few small changes to match function signatures as well
as several larger ones:
- OnNegotiationNeeded is no longer supported, so CreateOffer and
SetLocalDescription have been moved to a go routine called after the
other peer connection callbacks are set
- We need our own deserialize/serialize functions
- We need to use a SettingEngine in order to access the
OnICEGatheringStateChange callback
MaxBytesReader is only documented for server side reads, so we're using
a local limitedRead function instead that uses an io.LimitedReader.
Declared limits in a commented constant
