Added multi-profile support for attestation on chromeos.

Added multi-profile support for attestation on chromeos.

All certified keys and certificates will be associated with the correct
profile when multiple profiles are used.

BUG=chromium:205206
TEST=unit, manual

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=229891

[Darren Krahn, 2013-10-11 22:15:09]

+satorux: chromeos/dbus, chromeos/cryptohome
+bartfab: OWNERS, attestation, extension api
+pastarmovj: OWNERS
+mpcomplete: approve OWNERS for extension api

[Matt Perry, 2013-10-11 22:26:59]

lgtm

[bartfab (slow), 2013-10-13 16:31:35]

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
File chrome/browser/chromeos/attestation/platform_verification_flow.cc (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
chrome/browser/chromeos/attestation/platform_verification_flow.cc:241:
chromeos::User* user = user_manager_->GetUserByProfile(
#include "chrome/browser/chromeos/login/user.h"

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
chrome/browser/chromeos/attestation/platform_verification_flow.cc:242:
Profile::FromBrowserContext(web_contents->GetBrowserContext()));
#include "content/public/browser/browser_context.h"
#include "chrome/browser/profiles/profile.h"

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
File chrome/browser/chromeos/attestation/platform_verification_flow.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
chrome/browser/chromeos/attestation/platform_verification_flow.h:182: void
OnCertificateReady(const std::string& user_email,
We have many different names for the same thing: user_email == user_name ==
user_id == account_id. To tame this a bit, we decided to only use the term
user_id in all new code.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
File
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h
(right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h:95:
const std::string& user_email,
As in PlatformVerificationFlow, please call this user_id instead of user_email.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h:109:
const std::string& user_email,
As above, s/user_email/user_id/.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h:117:
const std::string& user_email,
As above, s/user_email/user_id/.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
File chromeos/attestation/attestation_flow.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
chromeos/attestation/attestation_flow.h:162: //   user_email - The active user's
canonical email.
As elsewhere, please use user_id.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
File chromeos/attestation/attestation_flow_unittest.cc (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
chromeos/attestation/attestation_flow_unittest.cc:118: "fake_email",
Could you use something with an @ symbol in it? This particular code may not
care but we had so many cases where an e-mail ended up getting canonicalized
somewhere, causing a crash because of the missing @. It is best not to create
bad precedents. Something like user@example.com would be safe.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/cryptohome...
File chromeos/cryptohome/async_method_caller.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/cryptohome...
chromeos/cryptohome/async_method_caller.h:138: const std::string& user_email,
As in the other files, please use user_id.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/dbus/crypt...
File chromeos/dbus/cryptohome_client.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/dbus/crypt...
chromeos/dbus/cryptohome_client.h:294: const std::string& user_email,
As in the other files, please use user_id.

[pastarmovj, 2013-10-14 08:32:35]

LGTM for the OWNERS files.

[Darren Krahn, 2013-10-15 16:49:05]

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
File chrome/browser/chromeos/attestation/platform_verification_flow.cc (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
chrome/browser/chromeos/attestation/platform_verification_flow.cc:241:
chromeos::User* user = user_manager_->GetUserByProfile(
On 2013/10/13 16:31:35, bartfab wrote:
> #include "chrome/browser/chromeos/login/user.h"

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
chrome/browser/chromeos/attestation/platform_verification_flow.cc:242:
Profile::FromBrowserContext(web_contents->GetBrowserContext()));
On 2013/10/13 16:31:35, bartfab wrote:
> #include "content/public/browser/browser_context.h"
> #include "chrome/browser/profiles/profile.h"

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
File chrome/browser/chromeos/attestation/platform_verification_flow.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/chro...
chrome/browser/chromeos/attestation/platform_verification_flow.h:182: void
OnCertificateReady(const std::string& user_email,
On 2013/10/13 16:31:35, bartfab wrote:
> We have many different names for the same thing: user_email == user_name ==
> user_id == account_id. To tame this a bit, we decided to only use the term
> user_id in all new code.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
File
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h
(right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h:95:
const std::string& user_email,
On 2013/10/13 16:31:35, bartfab wrote:
> As in PlatformVerificationFlow, please call this user_id instead of
user_email.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h:109:
const std::string& user_email,
On 2013/10/13 16:31:35, bartfab wrote:
> As above, s/user_email/user_id/.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chrome/browser/exte...
chrome/browser/extensions/api/enterprise_platform_keys_private/enterprise_platform_keys_private_api.h:117:
const std::string& user_email,
On 2013/10/13 16:31:35, bartfab wrote:
> As above, s/user_email/user_id/.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
File chromeos/attestation/attestation_flow.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
chromeos/attestation/attestation_flow.h:162: //   user_email - The active user's
canonical email.
On 2013/10/13 16:31:35, bartfab wrote:
> As elsewhere, please use user_id.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
File chromeos/attestation/attestation_flow_unittest.cc (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/attestatio...
chromeos/attestation/attestation_flow_unittest.cc:118: "fake_email",
On 2013/10/13 16:31:35, bartfab wrote:
> Could you use something with an @ symbol in it? This particular code may not
> care but we had so many cases where an e-mail ended up getting canonicalized
> somewhere, causing a crash because of the missing @. It is best not to create
> bad precedents. Something like mailto:user@example.com would be safe.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/cryptohome...
File chromeos/cryptohome/async_method_caller.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/cryptohome...
chromeos/cryptohome/async_method_caller.h:138: const std::string& user_email,
On 2013/10/13 16:31:35, bartfab wrote:
> As in the other files, please use user_id.

Done.

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/dbus/crypt...
File chromeos/dbus/cryptohome_client.h (right):

https://chromiumcodereview.appspot.com/27044004/diff/3001/chromeos/dbus/crypt...
chromeos/dbus/cryptohome_client.h:294: const std::string& user_email,
On 2013/10/13 16:31:35, bartfab wrote:
> As in the other files, please use user_id.

Done.

[Darren Krahn, 2013-10-15 16:49:57]

+davidyu: CC for extension API

[satorux1, 2013-10-16 01:52:09]

chromeos/dbus chromeos/cryptohome lgtm

[bartfab (slow), 2013-10-16 12:53:46]

Thanks for changing user_email to user_id. I am still uneasy about the comments
referring to a "canonical e-mail address". Yes, this is what a user_email has
traditionally always been. Yes, this is what it will always be for GAIA
accounts. But we have other account types now where the user_id is anything but
an e-mail address. In order to make legacy code happy, we ensure that the
user_id contains an @ symbol. But that is about it. It is just a string with an
@ in it, not an e-mail address. Once again, old comments are welcome to stay as
they are. But I think we should not be referring to canonical e-mail addresses
in new comments.

[Darren Krahn, 2013-10-17 22:59:17]

On 2013/10/16 12:53:46, bartfab wrote:
> Thanks for changing user_email to user_id. I am still uneasy about the
comments
> referring to a "canonical e-mail address". Yes, this is what a user_email has
> traditionally always been. Yes, this is what it will always be for GAIA
> accounts. But we have other account types now where the user_id is anything
but
> an e-mail address. In order to make legacy code happy, we ensure that the
> user_id contains an @ symbol. But that is about it. It is just a string with
an
> @ in it, not an e-mail address. Once again, old comments are welcome to stay
as
> they are. But I think we should not be referring to canonical e-mail addresses
> in new comments.

Comments clarified.

[bartfab (slow), 2013-10-18 13:02:12]

lgtm

[commit-bot: I haz the power, 2013-10-21 17:01:42]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dkrahn@chromium.org/27044004/162001

[commit-bot: I haz the power, 2013-10-21 19:57:27]

Change committed as 229891