Add support to meterpreter that allows for the querying and toggling of
SSL certificate verification on the fly.
In order to verify that the socket was SSL-enabled, some rejigging had
to be done of the type? method in the ssl socket class.
This patch fixes#4711.
The problem here is that the browser sometimes will shutdown some of our
exploit's connections (in my testing, all Java), and that will cause Ruby
to call a rb_sys_fail with "getpeername(2)". The error goes all the
way to Rex::IO::StreamServer's monitor_listener method, which triggers a
"break" to quit monitoring. And then this causes another chain of reactions
that eventually forces BrowserAutoPwn to quit completely (while the
JavaScript on the browser is still running)
This change emulates the auto-generated snakeoil certificate from Ubuntu 14.04. The main changes including moving to 2048-bit RSA, SHA256, a single name CN for subject/issuer, and the removal of most certificate extensions.
Note that this does not apply to reverse_tcp meterpreter clients yet, as
they do not allow certificates to be supplied. I abstracted out the SSL
certificate generation and parsing methods so that we can address this
next.
This change adds two new Rex exceptions and changes the local comm to raise the right one depending on the circumstances. The problem with the existing model is
that failed binds and failed connections both raised the same exception. This change is backwards compatible with modules that rescue Rex::AddressInUse in additi
on to Rex::ConnectionError. There were two corner cases that rescued Rex::AddressInUse specifically:
1. The 'r'-services mixin and modules caught the old exception when handling bind errors. These have been updated to use BindFailed
2. The meterpreter client had a catch for the old exception when the socket reports a bad destination (usually a network connection dropped). This has been updat
ed to use InvalidDestination as that was the intention prior to this change.
Since AddressInUse was part of ConnectionError, modules and mixins which caught both in the same rescue have been updated to just catch ConnectionError.
Instead of the eventual target where our proxy chain will connect. In
the usual case (no Proxies set), this will be the same output as before.
When proxies are given, the user will see that the first proxy
connection is actually what failed.
Otherwise, we risk getting our connections killed by particularly
aggressive DPI devices (IPS, firewalls, etc)
Squashed commit of the following:
commit 5e203851d5c9dce1fe984b106ce3031a3653e54b
Author: Tod Beardsley <tod_beardsley@rapid7.com>
Date: Wed Oct 15 10:19:04 2014 -0500
Whoops missed one
commit 477b15a08e06e74d725f1c45486b37e4b403e3c2
Author: Tod Beardsley <tod_beardsley@rapid7.com>
Date: Wed Oct 15 10:16:59 2014 -0500
Other datastore options also want TLS1 as default
commit 8d397bd9b500ff6a8462170b4c39849228494795
Author: Tod Beardsley <tod_beardsley@rapid7.com>
Date: Wed Oct 15 10:12:06 2014 -0500
TCP datastore opts default to TLS1
Old encryption is old. See also: POODLE
Note that there are some cases of host-endian left, these
are intentional because they operate on host-local memory
or services.
When in doubt, please use:
```
ri pack
```
Older versions of OpenSSL did not export the OP_NO_COMPRESSION constant,
so users running metasploit on systems with old copies of openssl
would throw a NameError since the constant did not exist.
[SeeRM #823], where Stephen was asking for SSL compression for
Meterpreter -- this isn't that, but it's at least now possible for other
Metasploit functionality.
Also disables the compression by default. TLS-level compression is almost
never used by browsers, and openssl seems to be the only one that enables
it by default.
This also kills some ruby < 1.9.3 code.
Aside from codebase-wide changes, nearly all of these tests haven't been
touched since before 2010, and there is no effort to maintain this style
of testing. We've moved on to (correctly) seperating out our tests from
our codebase.
Update Rex::Socket::SslTcp to accept verification mode string from
Rex::Socket::Parameters, which has been modified accordingly.
Add SSLVerifyMode and SSLCipher options (params and socket work
were done before, but the option was not exposed) to
Msf::Exploit::Tcp.
Testing:
```
>> sock = Rex::Socket::Tcp.create('PeerHost'=>'10.1.1.1','PeerPort'
=>443,'SSL' => true, 'SSLVerifyMode' => 'NONE')
>> sock.sslctx.verify_mode
=> 0
>> sock.close
=> nil
>> sock = Rex::Socket::Tcp.create('PeerHost'=>'10.1.1.1','PeerPort'
=>443,'SSL' => true, 'SSLVerifyMode' => 'PEER')
=> #<Socket:fd 13>
>> sock.sslctx.verify_mode
=> 1
```
Note: this should be able to resolve the recent SSL socket hackery
of exploit/linux/misc/nagios_nrpe_arguments.
we were only accepting sloppy string values and not accepting input of
the actual symbols that OpenSSL expects in the first place. Allow the
user to enter it right themselves to be compat with OpenSSL
Brent Cook
OJ
HD Moore
sinn3r
James Lee
Tod Beardsley
Joe Vennix
jvazquez-r7
Tab Assassin
David Maloney
RageLtMan
Tod Beardsley