Introduce new chrome.easyUnlock APIs for getting the sign-in challenge and using it to sign-in the …

Introduce new chrome.easyUnlock APIs for getting the sign-in challenge and using it to sign in the user.

BUG=401636
Committed: https://crrev.com/82bf980c0c69abc54169b13d0052ec4fbbe67916
Cr-Commit-Position: refs/heads/master@{#294317}

[Tim Song, 2014-09-08 22:59:35]

tengs@chromium.org changed reviewers:
+ tbarzic@chromium.org, xiyuan@chromium.org

[Tim Song, 2014-09-08 22:59:36]

Let me know these APIs look sufficient.

[Tim Song, 2014-09-08 22:59:36]

Let me know these APIs look sufficient.

[Tim Song, 2014-09-08 23:54:23]

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/easy_unlock_private.idl (right):

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/easy_unlock_private.idl:264: static void
getSignInChallenge(DataCallback callback);
Should we pass m2 here and return the wrapped challenge SecureMessage?

[xiyuan, 2014-09-09 02:35:28]

I think this should be good enough for phase 1.

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/easy_unlock_private.idl (right):

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/easy_unlock_private.idl:264: static void
getSignInChallenge(DataCallback callback);
On 2014/09/08 23:54:23, Tim Song wrote:
> Should we pass m2 here and return the wrapped challenge SecureMessage?

Yep, makes sense.

[xiyuan, 2014-09-09 16:19:49]

One more thing we might want to consider is that a user could have multiple
phones. So it might make sense that getSignInChallenge returns a BT address (or
some other identifier) to challenge bytes dict and trySignInSecret present a BT
address with the secret.

[tbarzic, 2014-09-09 19:55:49]

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/easy_unlock_private.idl (right):

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/easy_unlock_private.idl:264: static void
getSignInChallenge(DataCallback callback);
On 2014/09/09 02:35:28, xiyuan wrote:
> On 2014/09/08 23:54:23, Tim Song wrote:
> > Should we pass m2 here and return the wrapped challenge SecureMessage?
> 
> Yep, makes sense.

I'm nit sure that we'll be able to create the message in Chrome without
knowledge about securemessage proto, so we would probably need a separate
function for signing m2. For first pass, this should be enough though.

[Tim Song, 2014-09-10 00:47:13]

> One more thing we might want to consider is that a user could have multiple
> phones. So it might make sense that getSignInChallenge returns a BT address
(or
> some other identifier) to challenge bytes dict and trySignInSecret present a
BT
> address with the secret.

Right now we're only handling one user at a time, and we may only support the
multiple user use case after migrating to native code. So Let's not worry about
that for now.

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/easy_unlock_private.idl (right):

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/easy_unlock_private.idl:264: static void
getSignInChallenge(DataCallback callback);
On 2014/09/09 19:55:49, tbarzic wrote:
> On 2014/09/09 02:35:28, xiyuan wrote:
> > On 2014/09/08 23:54:23, Tim Song wrote:
> > > Should we pass m2 here and return the wrapped challenge SecureMessage?
> > 
> > Yep, makes sense.
> 
> I'm nit sure that we'll be able to create the message in Chrome without
> knowledge about securemessage proto, so we would probably need a separate
> function for signing m2. For first pass, this should be enough though.

I'm creating the wrapper SecureMessage in the app at the moment, but this will
eventually need to be piped through the TPM.

Okay, let's leave this work for later.

[Tim Song, 2014-09-10 00:50:28]

tengs@chromium.org changed reviewers:
+ isherman@chromium.org, yoz@chromium.org

[Tim Song, 2014-09-10 00:50:28]

+yoz@ and isherman@

Please take a look.

[Yoyo Zhou, 2014-09-10 01:16:00]

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/easy_unlock_private.idl (right):

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/easy_unlock_private.idl:266: // Tries to sign-in
the current user with a secret obtained by decrypting
This says something about trying, but it doesn't say what happens if it succeeds
or fails. Is that something you want to document?

[Tim Song, 2014-09-10 03:07:32]

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
File chrome/common/extensions/api/easy_unlock_private.idl (right):

https://codereview.chromium.org/552893002/diff/20001/chrome/common/extensions...
chrome/common/extensions/api/easy_unlock_private.idl:266: // Tries to sign-in
the current user with a secret obtained by decrypting
On 2014/09/10 01:16:00, Yoyo Zhou wrote:
> This says something about trying, but it doesn't say what happens if it
succeeds
> or fails. Is that something you want to document?

Done.

[Ilya Sherman, 2014-09-10 04:12:42]

LGTM

[Yoyo Zhou, 2014-09-10 19:10:11]

LGTM

[tbarzic, 2014-09-10 19:57:34]

lgtm

[Tim Song, 2014-09-11 02:58:46]

The CQ bit was checked by tengs@chromium.org

[commit-bot: I haz the power, 2014-09-11 03:04:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patchset/552893002/40001

[commit-bot: I haz the power, 2014-09-11 03:47:15]

Committed patchset #3 (id:40001) as 05f040627387935442195bcb2c03c3682c61f10b

[commit-bot: I haz the power, 2014-09-11 04:00:14]

Patchset 3 (id:??) landed as
https://crrev.com/82bf980c0c69abc54169b13d0052ec4fbbe67916
Cr-Commit-Position: refs/heads/master@{#294317}