Adding TLS 1.3 constants

Adding TLS 1.3 constants

BUG=

Committed: https://crrev.com/d3fd3ee31713d3992dc4bcfd385b7ccc7f9c1afc
Cr-Commit-Position: refs/heads/master@{#406007}

[svaldez, 2016-06-20 20:28:49]

svaldez@chromium.org changed reviewers:
+ davidben@chromium.org

[svaldez, 2016-06-20 20:28:49]

Adding the TLS 1.3 constants (without updating maxVersion) in order to simplify
the steps necessary to do interop testing, and to keep the constants separate
from the CL to add a flag for TLS 1.3.

[davidben, 2016-06-20 20:41:59]

On 2016/06/20 20:28:49, svaldez wrote:
> Adding the TLS 1.3 constants (without updating maxVersion) in order to
simplify
> the steps necessary to do interop testing, and to keep the constants separate
> from the CL to add a flag for TLS 1.3.

Do we necessarily know these will be the only changes? I could imagine some of
the cipher suite stuff would demand that we do other things weirdly. Are you
expecting we'll give instructions for other people to build a different
Chromium? I don't think anyone will do that.

We should prioritize getting all the code in in a working form (which means
getting the Go half caught up for testing), so it's just an about:flags toggle
like things normally work. Otherwise, landing code that doesn't run hasn't
worked well thus far.

[davidben, 2016-07-18 09:46:40]

https://codereview.chromium.org/2083743002/diff/20001/extensions/browser/api/...
File extensions/browser/api/socket/tls_socket.cc (right):

https://codereview.chromium.org/2083743002/diff/20001/extensions/browser/api/...
extensions/browser/api/socket/tls_socket.cc:36: version =
net::SSL_PROTOCOL_VERSION_TLS1_3;
We probably should not expose this until it's working. Maybe leave a TODO or
something.

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
File extensions/common/api/socket.idl (right):

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
extensions/common/api/socket.idl:117: // or <code>tls1.3</code>.
Ditto re not wanting to expose this

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
File extensions/common/api/sockets_tcp.idl (right):

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
extensions/common/api/sockets_tcp.idl:86: // or <code>tls1.3</code>.
Ditto.

https://codereview.chromium.org/2083743002/diff/20001/net/socket/client_socke...
File net/socket/client_socket_pool_manager.cc (right):

https://codereview.chromium.org/2083743002/diff/20001/net/socket/client_socke...
net/socket/client_socket_pool_manager.cc:120: if
(ssl_config_for_origin.version_max != kDefaultSSLVersionMax) {
Mismerge? (This was for the fallback, which is gone now. Hopefully the new one,
if it exists, will be below the socket pools.)

https://codereview.chromium.org/2083743002/diff/20001/net/socket/ssl_client_s...
File net/socket/ssl_client_socket_impl.cc (right):

https://codereview.chromium.org/2083743002/diff/20001/net/socket/ssl_client_s...
net/socket/ssl_client_socket_impl.cc:400: NULL,
[We'll want to wait for the roll which will include this change.]

https://codereview.chromium.org/2083743002/diff/20001/net/test/spawned_test_s...
File net/test/spawned_test_server/base_test_server.h (right):

https://codereview.chromium.org/2083743002/diff/20001/net/test/spawned_test_s...
net/test/spawned_test_server/base_test_server.h:128: TLS_INTOLERANT_TLS1_2 = 4, 
// Intolerant of TLS 1.3 or higher.
Typo? (I guess leave this alone for now until we need to add fallback and tests.
Especially since the rest isn't hooked up yet)

[svaldez, 2016-07-18 10:04:01]

https://codereview.chromium.org/2083743002/diff/20001/extensions/browser/api/...
File extensions/browser/api/socket/tls_socket.cc (right):

https://codereview.chromium.org/2083743002/diff/20001/extensions/browser/api/...
extensions/browser/api/socket/tls_socket.cc:36: version =
net::SSL_PROTOCOL_VERSION_TLS1_3;
On 2016/07/18 09:46:40, davidben (OOO until 7-25) wrote:
> We probably should not expose this until it's working. Maybe leave a TODO or
> something.

Done.

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
File extensions/common/api/socket.idl (right):

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
extensions/common/api/socket.idl:117: // or <code>tls1.3</code>.
On 2016/07/18 09:46:40, davidben (OOO until 7-25) wrote:
> Ditto re not wanting to expose this

Done.

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
File extensions/common/api/sockets_tcp.idl (right):

https://codereview.chromium.org/2083743002/diff/20001/extensions/common/api/s...
extensions/common/api/sockets_tcp.idl:86: // or <code>tls1.3</code>.
On 2016/07/18 09:46:40, davidben (OOO until 7-25) wrote:
> Ditto.

Done.

https://codereview.chromium.org/2083743002/diff/20001/net/socket/client_socke...
File net/socket/client_socket_pool_manager.cc (right):

https://codereview.chromium.org/2083743002/diff/20001/net/socket/client_socke...
net/socket/client_socket_pool_manager.cc:120: if
(ssl_config_for_origin.version_max != kDefaultSSLVersionMax) {
On 2016/07/18 09:46:40, davidben (OOO until 7-25) wrote:
> Mismerge? (This was for the fallback, which is gone now. Hopefully the new
one,
> if it exists, will be below the socket pools.)

Done.

https://codereview.chromium.org/2083743002/diff/20001/net/socket/ssl_client_s...
File net/socket/ssl_client_socket_impl.cc (right):

https://codereview.chromium.org/2083743002/diff/20001/net/socket/ssl_client_s...
net/socket/ssl_client_socket_impl.cc:400: NULL,
On 2016/07/18 09:46:40, davidben (OOO until 7-25) wrote:
> [We'll want to wait for the roll which will include this change.]

Done.

[davidben, 2016-07-18 10:08:57]

lgtm, though probably wait for the roll to go in first. Though I guess we'll
just be doing Fake TLS 1.3? *shrug*

[davidben, 2016-07-18 14:40:41]

lgtm

[svaldez, 2016-07-18 14:41:29]

svaldez@chromium.org changed reviewers:
+ asvitkine@chromium.org

[svaldez, 2016-07-18 14:41:30]

Hi,

asvitkine: Could you look at the histograms change and provide owners approval.

-Steven

[Alexei Svitkine (slow), 2016-07-18 14:47:09]

lgtm

[svaldez, 2016-07-18 14:53:35]

The CQ bit was checked by svaldez@chromium.org

[commit-bot: I haz the power, 2016-07-18 14:53:56]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-07-18 16:06:25]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2016-07-18 16:06:40]

CQ bit was unchecked.

[commit-bot: I haz the power, 2016-07-18 16:09:32]

Description was changed from

==========
Adding TLS 1.3 constants

BUG=
==========

to

==========
Adding TLS 1.3 constants

BUG=

Committed: https://crrev.com/d3fd3ee31713d3992dc4bcfd385b7ccc7f9c1afc
Cr-Commit-Position: refs/heads/master@{#406007}
==========

[commit-bot: I haz the power, 2016-07-18 16:09:33]

Patchset 4 (id:??) landed as
https://crrev.com/d3fd3ee31713d3992dc4bcfd385b7ccc7f9c1afc
Cr-Commit-Position: refs/heads/master@{#406007}