The weight generation code also was cleaned up (and now can support
generating distributions that look like what ScrambleSuit does as
a compile time change).
Per: http://www.keithschwarz.com/darts-dice-coins/
To ease delopyment, "-genServerParams has changed".
* "-genServerParams" is now a bool, and will by default generate a
random node-id.
* "-genServerParams -genServerParamsFP=<Base16 blob>" will convert the
supplied bridge fingerprint to a node-id (the old behavior).
Either way of deriving node-id is belived to be secure.
* https://lists.torproject.org/pipermail/tor-dev/2014-May/006929.html
* https://lists.torproject.org/pipermail/tor-dev/2014-June/006936.html
The extra parameter was added because golang's flags library doesn't
support distinguishing between "set but used the default value" and
"not set, so you go the default value".
Instead of using the nonce for the secret box, just use SipHash-2-4 in
OFB mode instead. The IV is generated as part of the KDF. This
simplifies the code a decent amount and also is better on the off
chance that SipHash-2-4 does not avalanche as well as it is currently
assumed.
While here, also decouple the fact that *this implementation* of obfs4
uses a PRNG with 24 bytes of internal state for protocol polymorphism
instead of 32 bytes (that the spec requires).
THIS CHANGE BREAKS WIRE PROTCOL COMPATIBILITY.
Instead of threading the code, move the keypair generation to right
after Accept() is called. This should mask the timing differential due
to the rejection sampling with the noise from the variablity in how
long it takes for the server to get around to pulling a connection out
of the backlog, and the time taken for the client to send it's portion
of the handshake.
The downside is that anyone connecting to the obfs4 port does force us
to do a bunch of math, but the obfs4 math is relatively cheap compared
to it's precursors.
Fixes#9.
The old way was biasted towards the earlier values. Thanks to asn for
pointing this out and suggesting an alternative.
As an additional tweak, do not reuse the drbg seed when calculating the
IAT distribution, but instead run the seed through SHA256 first, for
extra tinfoil goodness.
Joining a SOCKS dialer on the list of things the Golang runtime really
should have is a HTTP CONNECT dialer. There's a full fledged HTTP
client and server there, but not this. Why? Who knows.
This fixes issue #7.
Despite the unfortunate scheme name, this really is SOCKS4, and not 4a,
as the torrc Socks4Proxy option only supports addresses and not FQDNs.
Part of issue #7.
With tor patched to support 8402, obfs4 bootstraps via a SOCKSv5 proxy
now. Other schemes will bail with a PROXY-ERROR, as the go.net/proxy
package does not support them, and I have not gotten around to writing
dialers for them yet (next on my TODO list).
Part of issue #7.
Currently obfs4proxy is hardcoded to always PROXY-ERROR, despite a
valid proxy uri being passed in the env var. Once the dialer portion
of the code is done, this will be changed.
Part of issue #7.
When enabled, inter-packet delay will be randomized between 0 and 10
ms in 100 usec intervals. As experiences from ScrambleSuit (and back
of the envelope math based on how networks work) show, this is
extremely expensive and artificially limits the throughput of the link.
When enabled, bulk transfer throughput will be limited to an average of
278 KiB/s.
* handhake_ntor_test now is considerably more comprehensive.
* The padding related constants in the spec were clarified.
This breaks wireprotocol compatibility.
This is done by maintaining a map keyed off the SipHash-2-4 digest of
the MAC_C component of the handshake. Collisions, while possible are
unlikely in the extreme and are thus treated as replays.
In concept this is fairly similar to the ScrambleSuit `replay.py` code,
with a few modifications:
* There is a upper bound on how large the replay filter can grow.
Currently this is set to 102400 entries, though it is unlikely that
this limit will be hit.
* A doubly linked list is also maintained parallel to the map, so the
filter compaction process does not need to iterate over the entire
filter.
As of `15b960d55905877a840fe605a41a8139bffb5329` goptlib supports
IsClient, IsServer, and handling the StateLocation.
Yes this means you need to use goptlib out of git.
* Fixed where the code wasn't ensuring that the MAC_[C,S] was present.
* Optimized the server side to only look at the tail of the (possibly
incomplete handshakeRequest).
Clients will now always add 87 bytes of padding to the clientRequest,
and Servers will always send the PRNG seed frame unpadded, and bundled
with the serverResponse.
Why 87 bytes? The amount of data that the server sends is 87.
This fixes#5.
Instead of including the previous secretbox in the input when
calculating the SipHash-2-4 digest used to generate the obfuscation
mask, use only the nonce. This is significantly faster, and if someone
breaks obfs4 by exploiting the low amount of input entropy between each
invocation (a counter incrementing by 1), I hope they publish the
attack on the PRF.
This breaks wire protocol compatibility.
This paves the way for having servers use the same seed for all
incoming connections, across multiple startup/shutdown cycles. As
opposed to the current situation where each Obfs4Listener will
randomly generate it's seed at creation time.
Additionally, use 256 bit seeds (128 bit SipHash-2-4 key + 16 bytes of
initial material).