| Age | Commit message (Collapse) | Author |
|---|
|
* legoscia/tls_ipv6_dist:
Add inet6_tls_dist, for TLS distribution over IPv6
Conflicts:
lib/ssl/src/ssl_tls_dist_proxy.erl
OTP-13391
|
|
* legoscia/fix-ssl-example/PR-976/OTP-13363:
ssl: Modernize utility function
Fix ssl example
|
|
* legoscia/crl-check-without-dp:
Be suspicious of certificates without CRL DPs
Refactor ssl_crl_SUITE: extract crl_verify_error/6
OTP-13378
|
|
* legoscia/critical-extension-verify-none:
ssl: with verify_none, accept critical extensions
OTP-13377
|
|
Use application:ensure_all_started/2 instead of hard coding dependencies
|
|
Older SSL/TLS versions have cipher suites
that look like {key_exchange(), cipher(), MAC::hash()}
and the hash function used by the PRF (Pseudo Random function)
is implicit and always the same for that protocol version.
In TLS 1.2 a cipher suite is
{key_exchange(), cipher(), MAC::hash(), PRF::hash()}.
Internally a cipher suite is always a four tuple but
for backwards compatibility older cipher suites
will be presented as a three tuples, however new cipher suites
should be presented as four tuples.
|
|
Make sure that options only relevant for one role (client|server) is set
to undefined when the other role is invoked. As there are many options to
ssl, and many are optional, we choose to filter out all undefined options
to avoid overwhelming the user with not relevant information.
This way there is no need for any special handling of the role specific options
which is also nice.
|
|
Previously, if certificate revocation checking was turned on, and a
certificate didn't contain a CRL Distribution Points extension, and
there was no relevant CRL in the cache, then ssl_handshake:crl_check
would accept the certificate even if the crl_check option was set to
reject certificates for which the revocation status could not be
determined. With this change, such certificates will only be accepted
if the crl_check option was set to best_effort.
The process for CRL validation is described in section 6.3 of RFC
5280. The text doesn't mention any special treatment to be given to
certificates without distribution points: it just says "For each
distribution point..." (section 6.3.3), which would leave the
revocation status undetermined, unless there were "any available CRLs
not specified in a distribution point but issued by the certificate
issuer". Thus the result of this algorithm should be UNDETERMINED in
this case, not UNREVOKED, and the crl_check option should govern how
the implementation reacts to this result.
|
|
When establishing a TLS connection with {verify, verify_none}, if the
server has a certificate with a critical extension, for example a
"Netscape Cert Type" extension, certificate verification would fail,
which is surprising given that the name of the option suggests that no
verification would be performed.
With this change, certificate extensions marked as critical are
ignored when using verify_none.
|
|
When searching for a certificate's issuer in the `CertDB`, verify the signature
against the original DER certificate from the handshake instead of a re-encoding
of the parsed certificate. This avoids false negatives due to differences
between DER encoding implementations of OTP and other platforms.
|
|
|
|
* ia/ssl/validator:
ssl: Only start a new session validator if the old one has finished its work
|
|
* legoscia/tls-dist-listen-ip:
TLS distribution: bind erts socket to localhost
OTP-13300
|
|
Generalise much of inet_tls_dist, so that inet6_tls_dist can reuse it.
|
|
* legoscia/tls-dist-connect-options:
ssl_dist_SUITE: don't use deprecated functions
TLS distribution: support inet_dist_connect_options
OTP-13285
|
|
* legoscia/tls_dist_wait_for_code_server:
TLS distribution: wait for code server
OTP-13268
|
|
If the session table is big the validator may not have finshed before
the validation interval is up, in this case we should not start a new
validator adding to the cpu load.
|
|
|
|
|
|
Allow adding extra options for outgoing TLS distribution connnections,
as supported for plain TCP connections.
|
|
* legoscia/tls_dist_error_reporting:
Report bad options for outgoing TLS distribution
Save error reasons for TLS distribution connections
Report bad options for TLS distribution connections
OTP-13219
|
|
There is no reason for the socket on the erts side of the proxy to
accept connections from other hosts, so let's bind it to the loopback
interface.
Also change {ip, {127,0,0,1}} to {ip, loopback} for the erts side of
the socket for outgoing connections, to avoid hardcoding IPv4.
|
|
|
|
If ssl:connect/3 returns an error related to options, let's log that
so we have a chance to see it and fix it.
|
|
When establishing an outbound connection for TLS distribution, let's
hold on to the failure reasons and use them as exit reasons. These
exit reasons are normally invisible, but they can be seen in the logs
after calling net_kernel:verbose(1).
While there are trace messages in the code already, those require
recompiling the module with a special flag, which is more cumbersome
than changing the net_kernel verbosity level at run time.
|
|
* ia/ssl-prepare-release:
ssl: Correct spec
ssl: Prepare for release
|
|
|
|
|
|
* rlipscombe/rl-ssl-options:
Ensure single 'raw' option is handled correctly
Pass 'raw' options through
OTP-13166
|
|
* ia/ssl/sslv3-completeness:
ssl: SSLv3 completeness
|
|
We are considering removing default support for DES cipher suites.
However this cipher suite is currently allowed in TLS and missing from
SSL.
|
|
If upper limit is reached invalidate the current cache entries, e.i the session
lifetime is the max time a session will be keept, but it may be invalidated
earlier if the max limit for the table is reached. This will keep the ssl
manager process well behaved, not exhusting memeory. Invalidating the entries
will incrementally empty the cache to make room for fresh sessions entries.
|
|
|
|
* legoscia/tls_dist_options:
Test interface listen option for TLS distribution
Test socket listen options for TLS distribution
Test port options for TLS distribution
TLS Dist: Use inet_dist_ options
Conflicts:
lib/ssl/src/ssl_tls_dist_proxy.erl
lib/ssl/test/ssl_dist_SUITE.erl
OTP-12838
|
|
* legoscia/ssl_connection_terminate_crash:
Avoid crash for SSL connections with nonexistent keyfile
OTP-13144
|
|
* legoscia/tls_dist_nodelay:
Add test for dist_nodelay option
Honour dist_nodelay socket option in tls_dist proxy
OTP-13143
|
|
* legoscia/ssl-dist-error-handling:
In ssl_tls_dist_proxy, pass along EPMD registration errors
OTP-13142
|
|
* ppikula/fix-24h-macro:
fix incorrect number of seconds in 24h macro The previous commit - 7b93f5d8a224a0a076a420294c95a666a763ee60 fixed the macro only in one place.
OTP-13141
|
|
Add test that checks that the option inet_dist_listen_options is used
when starting a node with TLS distribution.
This test was adapted from inet_dist_options_options in
erl_distribution_SUITE.
|
|
The inet_dist_ options, such as min/max port numbers aren't used
with TLS distribution. This commits uses those settings in the
same way as they're used in inet_tcp_dist.erl
|
|
* legoscia/tls-dist-shutdown:
Adjust shutdown strategies for distribution over TLS
OTP-13134
|
|
willing to support
Refactor highest_protocol_version so that code is symmetrical with lowest_protocol_version. For clarity and possible future use cases of highest_protocol_version/2
|
|
As described in the comments in the patch, doing a TLS handshake
requires the crypto module to be loaded. The crypto module needs the
code server to find its NIF library. However, there is a time window
between opening the listening ports for distribution and starting the
code server, and if we get an incoming connection in that time window,
the node would believe that it's alive, but it wouldn't actually
accept any more connections.
|
|
If ssl:ssl_accept/2 returns an error related to options, it's most
likely something we want to log. In particular, if the specified
certificate file doesn't exist, this is where the error ends up, so we
shouldn't just throw the error away.
|
|
Starting an SSL connection with a nonexistent keyfile will obviously
return an error:
> ssl:connect("www.google.com", 443, [{keyfile, "nonexistent"}]).
{error,{options,{keyfile,"nonexistent",{error,enoent}}}}
But it also generates an error report with the following backtrace:
** Reason for termination =
** {badarg,[{ets,select_delete,
[undefined,[{{{undefined,'_','_'},'_'},[],[true]}]],
[]},
{ets,match_delete,2,[{file,"ets.erl"},{line,700}]},
{ssl_pkix_db,remove_certs,2,[{file,"ssl_pkix_db.erl"},{line,243}]},
{ssl_connection,terminate,3,
[{file,"ssl_connection.erl"},{line,941}]},
{tls_connection,terminate,3,
[{file,"tls_connection.erl"},{line,335}]},
{gen_fsm,terminate,7,[{file,"gen_fsm.erl"},{line,610}]},
{gen_fsm,handle_msg,7,[{file,"gen_fsm.erl"},{line,532}]},
{proc_lib,init_p_do_apply,3,[{file,"proc_lib.erl"},{line,240}]}]}
This happens because the ssl_connection process receives its cert_db
while handling the {start, Timeout} message, but if the handshake
fails, the cert_db will never be inserted into the state data, and the
terminate function will use 'undefined' as an ETS table name.
Avoid this by checking for 'undefined' in the handle_trusted_certs_db
function.
|
|
The duplicate_name error returned from erl_epmd:register_node elicits a
particularly precise error message from net_kernel, so let's pass it
along to our caller.
Not doing this for the other things that could go wrong here, since for
those having the line number will likely aid debugging.
|
|
The previous commit - 7b93f5d8a224a0a076a420294c95a666a763ee60 fixed the macro
only in one place.
|
|
Change ssl_dist_sup to be considered as a supervisor with infinite
shutdown time.
Change the ssl_connection_dist instance of tls_connection_sup to have
infinite shutdown time.
This avoids spurious error messages when shutting down a node that
uses distribution over TLS.
|
|
In Erlang R16B03-1, I've been passing raw options to ssl:listen as
follows, and it's been working fine:
% The constants are defined elsewhere.
LOpts = [{raw, ?IPPROTO_TCP, ?TCP_MAXSEG, <<MSS:32/native>>} | ...],
{ok, LSocket} = ssl:listen(0, LOpts)
In Erlang 17.3, this fails with
{option_not_a_key_value_tuple,{raw,6,2,<<64,2,0,0>>}}
I originally reported this in
http://erlang.org/pipermail/erlang-questions/2014-October/081226.html
I need to pass this particular raw option to ssl:listen, because it
needs to be applied when the socket is first opened -- between inet:open
and prim_inet:listen -- it cannot be applied later by setopts. This
means that it needs to be done by inet_tcp:listen/2 -- well, actually by
inet:open/8, but...
Otherwise it's racey -- a client could connect between prim_inet:listen
and the setopts call. The MSS option is advertised in the SYN,ACK
packet, and can't be changed later.
|
|
Soft upgrade test did not work as expected due to that the upgrade
frame work keeps the control of the test case process to itself,
so we need a proxy process to receive messages from ssl test framework.
|
