Review main Merged

67548[FEATURE] Introduce MFA in Core

Oli Bartsch CR +2 V +2 Nightmare +5564 −15 updated 02/19/21, 14:38

Commit message

A new API is introduced, providing multi-factor
authentication for the Core. The API is furthermore
directly used to add two MFA providers by default:

* TOTP (time-based one-time passwords)
* Recovery codes

Even if the API is designed to allow MFA in both,
backend and frontend, it is currently only implemented
into the backend. Users can therefore configure their
available MFA providers in a new backend module,
accessible via their user settings.

There are also some configuration options for
administrators to e.g. define a recommended provider
or to disallow available providers for specific users
or user groups.

Administration of the users' MFA providers is possible
for administrators in the corresponding user records.

New providers can be introduced by implementing the
MfaProviderInterface and tagging the service with the
`mfa.provider` tag.

Note that the API is currently marked as internal since
changes in upcoming patches are to be expected.

Following dependencies are introduced:

* bacon/bacon-qr-code "^2.0"
* christian-riesen/base32 "^1.5"

Possible features that could follow later-on:

* MFA frontend integration
* Webauthn core provider for FIDO2 and U2F.
* Forcing users to set up MFA on login
* Password-recovery with active MFA

Review

  • Code review CR +2 Done
    • Benni Mack +2
    • Benjamin Franzke +1
    • Christian Kuhn +1
  • Verified V +2 Done
    • Benni Mack +2
    • Benjamin Franzke +1
    • TYPO3com +1 CI
    • core-ci +1 CI

Issues 1

Files 71

Show 51 more files

Review log 97207

Show 81182 earlier messages
  1. Oli Bartsch uploaded patch set 1 01/29/21, 00:37

  2. Oli Bartsch marked it work in progress 01/29/21, 00:37

  3. TYPO3com voted on patch set 1 V -1 01/29/21, 00:52 Completed build in 14m 20s on Fri, 29 Jan, 12:51 AM Test Summary: 815 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9092

  4. Oli Bartsch uploaded patch set 2 01/29/21, 13:22

  5. TYPO3com voted on patch set 2 V -1 01/29/21, 13:31 Completed build in 08m 55s on Fri, 29 Jan, 01:31 PM Test Summary: 694 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9100

  6. Oli Bartsch uploaded patch set 3 01/29/21, 14:00

  7. TYPO3com voted on patch set 3 V -1 01/29/21, 14:10 Completed build in 09m 53s on Fri, 29 Jan, 02:10 PM Test Summary: 673 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9101

  8. Oli Bartsch updated the commit message in patch set 4 01/29/21, 16:29

  9. Oli Bartsch updated the commit message in patch set 5 01/29/21, 16:40

  10. TYPO3com voted on patch set 4 V -1 01/29/21, 16:41 Completed build in 11m 39s on Fri, 29 Jan, 04:41 PM Test Summary: 2188 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9103

  11. TYPO3com voted on patch set 5 V -1 01/29/21, 16:52 Completed build in 10m 53s on Fri, 29 Jan, 04:52 PM Test Summary: 380 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9104

  12. Torben Hansen uploaded patch set 6 01/31/21, 14:04

  13. Torben Hansen commented on patch set 6 01/31/21, 14:07

    It must be ensured, that the verify action can not be brute forced. This can be archived by adding a max. amount of verify form submission to the AbstractOtpProvider, so custom MFA providers can define a custom amount of max. verify form submissions. If max. form submissions is reached, the user must be logged out and redirected to the login form. When the MFA auth form is shown, the current MFA provider is given as part of the URL parameter "&identifier=". A user can change this to a non activated provider by changing the parameter (e.g. &identifier=hotp). It must be ensured in the MfaController, that the current provider is active for the user. For the HOTP and TOTP implementation, it should be ensured, that the issuer used for the QR code contains the [SYS][sitename], so it is easily possible to create accounts in authenticator apps without the need to change to name manually. It would be nice, if the OTP field for setup and edit action would have auto focus (autofocus="autofocus" does not seem to work in iframes for all browsers)
  14. TYPO3com voted on patch set 6 V -1 01/31/21, 14:15 Completed build in 09m 42s on Sun, 31 Jan, 02:15 PM Test Summary: 1318 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9116

  15. Torben Hansen commented on patch set 6 01/31/21, 18:05

    When the MFA verification fails, we should also consider to: * Add a Logentry to sys_log stating that MFA login failed for the user account * Add a PSR-14 event to provide custom code (e.g. additional logging, brute force detection, ...)
  16. Oli Bartsch uploaded patch set 7 02/01/21, 10:07

  17. TYPO3com voted on patch set 7 V -1 02/01/21, 10:18 Completed build in 09m 33s on Mon, 1 Feb, 10:18 AM Test Summary: 521 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9131

  18. Oli Bartsch uploaded patch set 8 02/01/21, 14:56

  19. TYPO3com voted on patch set 8 V -1 02/01/21, 14:59 Completed build in 39s on Mon, 1 Feb, 02:59 PM Test Summary: No tests found Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9153

  20. Oli Bartsch uploaded patch set 9 02/01/21, 15:25

  21. TYPO3com voted on patch set 9 V -1 02/01/21, 15:37 Completed build in 11m 32s on Mon, 1 Feb, 03:37 PM Test Summary: 712 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9155

  22. Oli Bartsch commented on patch set 9 02/01/21, 15:46

    Reply

    Hi Torben. I implemented a "lock/unlock" functionality. For OTP, the provider now gets locked after 3 invalid attempts. A locked provider can then only be unlocked by the user in the backend (if he authenticated with an alternative active provider) or by an admin (using SwitchUser). I implemented the isActive check now already in the initializeAction to prevent switching to a non activated provider. I'll have a look at the other suggestions (including your second comment) as soon as this POC becomes the actual patch.
  23. Oli Bartsch uploaded patch set 10 02/01/21, 15:49

  24. TYPO3com voted on patch set 10 V -1 02/01/21, 15:59 Completed build in 10m 03s on Mon, 1 Feb, 03:59 PM Test Summary: 615 of 32842 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9157

  25. Oli Bartsch uploaded patch set 11 02/04/21, 22:00

  26. TYPO3com voted on patch set 11 V -1 02/04/21, 22:07 Completed build in 07m 43s on Thu, 4 Feb, 10:07 PM Test Summary: 1441 of 32826 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9269

  27. core-ci voted on patch set 11 V -1 02/04/21, 22:16 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/386

  28. Oliver Hader commented on patch set 11 02/08/21, 10:08

    MfaProviderInterface.php:75

    In case Providers would be allowed to "do anything", it might be a `ResponseInterface` - merging body and potential headers (if any).

    MfaProviderInterface.php:102

    Next to `unlock`, does it need a `lock` or `revoke` command?

    MfaProviderInterface.php:113

    Not really required yet, but could be some `getMeta()` or `getManifest()` returning a `MfaProviderManifestInterface`

    MfaProviderRepository.php:38

    Remove default value for context to make it explicit. In case the repository might be acting on more that just users, the should be a `Actor` DTO combining userId/groupId/actorId and context.

    MfaProviderRepository.php:55

    This is "domain logic", a decision the controller/application should take, but not the repository as peer to a low-level database.

    MfaProviderRepository.php:118

    (see above for `updateProviderProperties`) - it probably might be `public function hasProviderEntry(): bool`

    MfaProviderRepository.php:136

    (see above for `updateProviderProperties`) - `public create...`
  29. Oli Bartsch uploaded patch set 12 02/08/21, 10:31

  30. TYPO3com voted on patch set 12 V -1 02/08/21, 10:39 Completed build in 08m 19s on Mon, 8 Feb, 10:39 AM Test Summary: 401 of 32826 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9346

  31. core-ci voted on patch set 12 V -1 02/08/21, 10:48 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/475

  32. Oli Bartsch uploaded patch set 13 02/08/21, 12:15

  33. Oli Bartsch commented on patch set 13 02/08/21, 12:18

    MfaProviderInterface.php:75 Reply Unresolved

    Currently not quite sure how to achieve this since the providers content is placed inside a form which has a predefined action and the provider identifier as hidden field. Should providers deal with this by themself and call the correct action in the TYPO3 controller?

    MfaProviderInterface.php:102 Reply Unresolved

    Also thought about this while implementing this feature. But since these methods are called by TYPO3 for specific events and TYPO3 can not know when to initiate a lock event, this must be handled by each provider itself. For example, the OTP providers allow 3 wrong attempts - this is a very easy check. However other providers could initiate a lock based on an ip address, or a completely external factor. TYPO3 just provides a way to unlock a provider and the possibility to allow providers to lock themself but can not initiate this call. While writing this, I think there could may be some call, in case TYPO3 detects some unusual behaviour. This however must be independent of the provider used. So TYPO3 could force a provider to handle such finding then. But for now, I can't imagine what such unusual event could be, since it must be very general.

    MfaProviderInterface.php:113 Reply Unresolved

    Great idea, since it separates the functional methods from the meta ones. I added such interface but did not include the identifier because this is not just an information used in some views, but required internally e.g. for registration of the provider.

    MfaProviderRepository.php:38 Reply

    The context is now implicit since the information are now stored in the corresponding user table.

    MfaProviderRepository.php:55 Reply

    Done

    MfaProviderRepository.php:118 Reply

    Done

    MfaProviderRepository.php:136 Reply

    Done
  34. TYPO3com voted on patch set 13 V -1 02/08/21, 12:26 Completed build in 10m 06s on Mon, 8 Feb, 12:26 PM Test Summary: 1010 of 32826 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9347

  35. core-ci voted on patch set 13 V -1 02/08/21, 12:26 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/476

  36. Benni Mack commented on patch set 13 02/08/21, 16:45

    AbstractUserAuthentication.php:1237

    should go into a separate factory IMHO

    TotpProviderManifest.php:27

    we could move the manifest information into the registration process (Services.yaml), or attach it directly to the provider - what do you think @bnf?
  37. Oliver Hader commented on patch set 13 02/09/21, 16:46

    AbstractOtp.php:144

    wrap in GeneralUtility::hmac() to have the secret prefix change every time...

    AbstractOtp.php:147

    What's the limit for the QR code - 10 characters?
  38. Oli Bartsch uploaded patch set 14 02/09/21, 23:14

  39. Oli Bartsch rebased onto patch set 15 02/09/21, 23:14

  40. Oli Bartsch commented on patch set 15 02/09/21, 23:21

    AbstractUserAuthentication.php:1237 Reply

    Ack

    AbstractOtp.php:144 Reply

    Done

    AbstractOtp.php:147 Reply Unresolved

    I'll have a look into this again. The main problem here is that if we pass more than 10 characters to the base32 method, it applies multiple `=` (pads) on the right side. Such secret is not accepted by e.g. GoogleAuthenticator.

    TotpProviderManifest.php:27 Reply

    Yes, though about this already but then we can not longer benefit from auto configuration based on the interface.
  41. TYPO3com voted on patch set 14 V -1 02/09/21, 23:24 Completed build in 09m 54s on Tue, 9 Feb, 11:24 PM Test Summary: 677 of 32838 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9422

  42. TYPO3com voted on patch set 15 V -1 02/09/21, 23:25 Completed build in 10m 38s on Tue, 9 Feb, 11:25 PM Test Summary: 573 of 32838 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9423

  43. core-ci voted on patch set 14 V -1 02/09/21, 23:25 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/560

  44. core-ci voted on patch set 15 V -1 02/09/21, 23:28 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/561

  45. Oli Bartsch uploaded patch set 16 02/09/21, 23:44

  46. TYPO3com voted on patch set 16 V -1 02/09/21, 23:52 Completed build in 08m 37s on Tue, 9 Feb, 11:52 PM Test Summary: 466 of 32838 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9425

  47. core-ci voted on patch set 16 V -1 02/09/21, 23:55 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/563

  48. Oli Bartsch uploaded patch set 17 02/10/21, 13:01

  49. TYPO3com voted on patch set 17 V -1 02/10/21, 13:10 Completed build in 08m 37s on Wed, 10 Feb, 01:10 PM Test Summary: 408 of 32841 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9439

  50. core-ci voted on patch set 17 V -1 02/10/21, 13:12 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/581

  51. Oli Bartsch uploaded patch set 18 02/10/21, 23:17

  52. TYPO3com voted on patch set 18 V -1 02/10/21, 23:25 Completed build in 08m 17s on Wed, 10 Feb, 11:25 PM Test Summary: 694 of 32841 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9455

  53. core-ci voted on patch set 18 V -1 02/10/21, 23:28 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/619

  54. Benjamin Franzke commented on patch set 18 02/11/21, 07:46

    TotpProviderManifest.php:27 Reply

    I somehow feel that the manifest should rather be a separate service, as the manifest may be desired to be used without the provider to be instantiated at all, no? I somehow think about the console command story, where not every command should be instantiated, just to be able to list the available commands. On top of that, or as another Idea, I could think of using the MfaProviderManifest both as description, but also als "lazy-loader" for the provider. Given following yaml: TYPO3\CMS\Core\Authentication\Mfa\Provider\Otp\TotpProvider: tags: - name: mfa.provider title: 'Time-based one-time password' description: 'Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua.' icon: 'content-coffe' defaultProviderAllowed: true Then the MfaProviderManifest would be automatically be created in the compiler pass, configured with the given property, *and* the (totp)privder passed a service locator (that's a lazy-factory) to resolve the TotpProvider on demand. Then MfaProviderRegistry would then have an array of MfaProviderManifests, which allow to resolve the real MfaProvider (if needed). And once we have php8 annotations, we can move these tags into the TotpProvider and then use full-interface based autoconfiguration. MfaProviderManifests is then a better structed/types array, than our usual registers have ;) If you are interested, and this sounds too mazy, I can provide this as a diff/patch in the next days.
  55. Torben Hansen commented on patch set 18 02/11/21, 09:54

    TotpProviderManifest.php:27 Reply

    +1 for auto creating the MfaProviderManifest. We must just keep in mind, that both the title and the description of MFA providers must be translatable. So if MfaProviderManifest would be created automatically, getTitle() and getDescription() should handle possible localization scenarios for both strings (e.g. "EXT:my_extension/Resources/Private/Language/locallang.xlf:mfa_provider.title").
  56. Oliver Hader uploaded patch set 19 02/11/21, 12:39

  57. TYPO3com voted on patch set 19 V -1 02/11/21, 12:40 Completed build in 18s on Thu, 11 Feb, 12:40 PM Test Summary: No tests found Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9461

  58. core-ci voted on patch set 19 V -1 02/11/21, 12:41 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/638

  59. Benjamin Franzke commented on patch set 19 02/11/21, 20:50

    TotpProviderManifest.php:27 Reply

    My proposal for merging manifest and provider for the backend code here: https://review.typo3.org/c/Packages/TYPO3.CMS/+/67872 The extensions simply implement MfaProviderInterface, but the annotated MfaProviderManifestInterface (which now extends/adapts MfaProviderInterface) is then seen by all core code that uses the providers. So, that's a bit easier to implement, but still easy to be used by core. @Oli: Would you be ok with that change, or does that block you or change too much of what you intended to do? @Torben: Good point regarding translations. With above example the trasnslations could be transparently handled by the intermeidate ProviderManifest class. So one would simply write: title: EXT:my_extension/Resources/Private/Language/locallang.xlf:mfa_provider.title in the Service.yaml file

    RecoveryCodesProvider.php:139

    Hm.. I think it would be better to hide the backup codes card from the user, as long it isn't usable. I wanteded to start over there, as my thinking is: Before I start to add a custom provider, I need backup codes first, for safety. And on top of that: I think we need to discuss how we "better" force users to create backup codes. I usually like those implementations that automatically create backend codes as soon as I register one MFA factor. I fear that uses are lazy (at least I am) and will not add backup codes, if not forced to.

    RecoveryCodesProvider.php:339

    typo here, should be activeProvidersExist

    RecoveryCodesProvider.php:341

    Should/Could be injected.
  60. Oli Bartsch uploaded patch set 20 02/11/21, 21:53

  61. TYPO3com voted on patch set 20 V -1 02/11/21, 21:54 Completed build in 22s on Thu, 11 Feb, 09:54 PM Test Summary: No tests found Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9489

  62. Oli Bartsch uploaded patch set 21 02/11/21, 21:58

  63. TYPO3com voted on patch set 21 V -1 02/11/21, 22:07 Completed build in 09m 26s on Thu, 11 Feb, 10:07 PM Test Summary: 574 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9492

  64. core-ci voted on patch set 20 V -1 02/11/21, 22:08 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/673

  65. core-ci voted on patch set 21 V -1 02/11/21, 22:17 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/676

  66. Oli Bartsch uploaded patch set 22 02/11/21, 23:20

  67. TYPO3com voted on patch set 22 V -1 02/11/21, 23:26 Completed build in 06m 25s on Thu, 11 Feb, 11:26 PM Test Summary: 372 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9494

  68. core-ci voted on patch set 22 V -1 02/11/21, 23:32 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/678

  69. Oli Bartsch uploaded patch set 23 02/11/21, 23:51

  70. TYPO3com voted on patch set 23 V -1 02/11/21, 23:51 Completed build in 12s on Thu, 11 Feb, 11:51 PM Test Summary: No tests found Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9495

  71. core-ci voted on patch set 23 V -1 02/11/21, 23:52 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/679

  72. Oli Bartsch uploaded patch set 24 02/11/21, 23:52

  73. Oli Bartsch commented on patch set 22 02/11/21, 23:53

    TotpProviderManifest.php:27 Reply Unresolved

    Thanks a lot!! I merged your proposal and made some adjustments since I've had already pushed a new patchset in the meantime. Could you please do me a favor and adjust the necessary parts in the commit message? Regarding the renaming of the ProviderPropertyManager, I'm fully with you, especially because I'm still not really happy with some details. For example, I think about performing a database update just once in a request. During runtime, it should be enough to just update the internal state. We could then use something like __destruct() to finally perform the necessary database update. What do you think about this, and the class in general? @Torben Absolutely, will add this in one of the next patchsets.

    RecoveryCodesProvider.php:139 Reply Unresolved

    I totally agree with you. We then have to extend the interface once again with another setting, because the providers are currently not able to decide whether they are shown in the overview. I thought about that, too. My second idea, after integrating the recovery codes as separate provider, was to adjust the API to enable every provide to feature their own recovery codes. The drawback is, a user then has to store multiple recovery code sets, which are only valid for the corresponding provider. This will may be a bad idea in terms of UX. Additionally this may also lead to a lot of duplicate code. Maybe you have another idea how to improve this?

    RecoveryCodesProvider.php:339 Reply

    Done

    RecoveryCodesProvider.php:341 Reply

    Done
  74. TYPO3com voted on patch set 24 V -1 02/11/21, 23:59 Completed build in 06m 36s on Thu, 11 Feb, 11:59 PM Test Summary: 1555 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9496

  75. core-ci voted on patch set 24 V -1 02/12/21, 00:04 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/680

  76. Oli Bartsch uploaded patch set 25 02/12/21, 10:05

  77. TYPO3com voted on patch set 25 V -1 02/12/21, 10:13 Completed build in 07m 40s on Fri, 12 Feb, 10:13 AM Test Summary: 845 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9499

  78. core-ci voted on patch set 25 V -1 02/12/21, 10:16 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/685

  79. Oli Bartsch uploaded patch set 26 02/12/21, 11:59

  80. TYPO3com voted on patch set 26 V -1 02/12/21, 12:07 Completed build in 07m 23s on Fri, 12 Feb, 12:07 PM Test Summary: 812 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9501

  81. core-ci voted on patch set 26 V -1 02/12/21, 12:09 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/687

  82. Oli Bartsch uploaded patch set 27 02/12/21, 12:55

  83. TYPO3com voted on patch set 27 V -1 02/12/21, 13:04 Completed build in 08m 28s on Fri, 12 Feb, 01:04 PM Test Summary: 1006 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9502

  84. core-ci voted on patch set 27 V -1 02/12/21, 13:06 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/688

  85. Oli Bartsch uploaded patch set 28 02/12/21, 14:07

  86. TYPO3com voted on patch set 28 V -1 02/12/21, 14:07 Completed build in 12s on Fri, 12 Feb, 02:07 PM Test Summary: No tests found Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9503

  87. core-ci voted on patch set 28 V -1 02/12/21, 14:08 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/689

  88. Oli Bartsch uploaded patch set 29 02/12/21, 14:46

  89. TYPO3com voted on patch set 29 V -1 02/12/21, 14:55 Completed build in 08m 27s on Fri, 12 Feb, 02:55 PM Test Summary: 2162 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9504

  90. core-ci voted on patch set 29 V -1 02/12/21, 14:57 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/690

  91. Oli Bartsch uploaded patch set 30 02/12/21, 17:58

  92. TYPO3com voted on patch set 30 V -1 02/12/21, 18:07 Completed build in 08m 17s on Fri, 12 Feb, 06:07 PM Test Summary: 659 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9507

  93. core-ci voted on patch set 30 V -1 02/12/21, 18:10 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/693

  94. Oli Bartsch uploaded patch set 31 02/14/21, 23:45

  95. TYPO3com voted on patch set 31 V -1 02/14/21, 23:53 Completed build in 07m 44s on Sun, 14 Feb, 11:53 PM Test Summary: 590 of 32852 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9527

  96. core-ci voted on patch set 31 V -1 02/14/21, 23:56 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/716

  97. Oli Bartsch uploaded patch set 32 02/16/21, 01:41

  98. TYPO3com voted on patch set 32 V -1 02/16/21, 01:50 Completed build in 08m 26s on Tue, 16 Feb, 01:50 AM Test Summary: 672 of 32862 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9560

  99. core-ci voted on patch set 32 V -1 02/16/21, 01:52 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/756

  100. Oli Bartsch uploaded patch set 33 02/16/21, 18:10

  101. TYPO3com voted on patch set 33 V -1 02/16/21, 18:18 Completed build in 07m 24s on Tue, 16 Feb, 06:18 PM Test Summary: 393 of 32862 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9577

  102. core-ci voted on patch set 33 V -1 02/16/21, 18:20 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/782

  103. Oli Bartsch uploaded patch set 34 02/16/21, 21:59

  104. TYPO3com voted on patch set 34 V -1 02/16/21, 22:06 Completed build in 07m 01s on Tue, 16 Feb, 10:06 PM Test Summary: 1073 of 32862 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9580

  105. core-ci voted on patch set 34 V -1 02/16/21, 22:11 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/786

  106. Oli Bartsch uploaded patch set 35 02/17/21, 01:07

  107. TYPO3com voted on patch set 35 V -1 02/17/21, 01:15 Completed build in 08m 13s on Wed, 17 Feb, 01:15 AM Test Summary: 536 of 32862 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9581

  108. core-ci voted on patch set 35 V -1 02/17/21, 01:17 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/787

  109. Oli Bartsch uploaded patch set 36 02/17/21, 13:10

  110. TYPO3com voted on patch set 36 V -1 02/17/21, 13:19 Completed build in 08m 40s on Wed, 17 Feb, 01:19 PM Test Summary: 416 of 32862 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9589

  111. core-ci voted on patch set 36 V -1 02/17/21, 13:23 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/798

  112. Benni Mack uploaded patch set 37 02/17/21, 14:14

  113. TYPO3com voted on patch set 37 V -1 02/17/21, 14:22 Completed build in 08m 00s on Wed, 17 Feb, 02:22 PM Test Summary: 611 of 32862 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9590

  114. core-ci voted on patch set 37 V -1 02/17/21, 14:27 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/799

  115. Benni Mack uploaded patch set 38 02/17/21, 15:01

  116. TYPO3com voted on patch set 38 V +1 02/17/21, 15:08 Completed build in 06m 36s on Wed, 17 Feb, 03:07 PM Test Summary: 32621 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9593

  117. core-ci voted on patch set 38 V +1 02/17/21, 15:12 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/802

  118. Benni Mack uploaded patch set 39 02/17/21, 15:33

  119. Benni Mack updated the commit message in patch set 40 02/17/21, 15:38

  120. TYPO3com voted on patch set 39 V +1 02/17/21, 15:42 Completed build in 08m 38s on Wed, 17 Feb, 03:42 PM Test Summary: 32621 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9597

  121. TYPO3com voted on patch set 40 V +1 02/17/21, 15:44 Completed build in 05m 57s on Wed, 17 Feb, 03:44 PM Test Summary: 32621 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9598

  122. core-ci voted on patch set 39 V +1 02/17/21, 15:44 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/806

  123. core-ci voted on patch set 40 V +1 02/17/21, 15:48 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/807

  124. Benni Mack uploaded patch set 41 02/17/21, 16:05

  125. Benni Mack commented on patch set 41 02/17/21, 16:07

    This change is ready for review.
    Side nodes / ToDo: * Remove custom again CSS (not needed) * Remove HOTP and Abstract OTP (move everything into TOTP) * Add unit tests for providers and registries * Add acceptance tests for setting up MFA
  126. TYPO3com voted on patch set 41 V +1 02/17/21, 16:14 Completed build in 08m 59s on Wed, 17 Feb, 04:14 PM Test Summary: 32621 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9599

  127. core-ci voted on patch set 41 V +1 02/17/21, 16:14 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/808

  128. Oli Bartsch uploaded patch set 42 02/17/21, 17:41

  129. TYPO3com voted on patch set 42 V +1 02/17/21, 17:49 Completed build in 08m 13s on Wed, 17 Feb, 05:49 PM Test Summary: 32621 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9601

  130. core-ci voted on patch set 42 V +1 02/17/21, 17:53 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/810

  131. Oli Bartsch uploaded patch set 43 02/18/21, 00:10

  132. Oli Bartsch commented on patch set 43 02/18/21, 00:12

    FYI: I removed HOTP in the latest patchset. It's now available as a separate extension and can be used e.g. for testing. See https://github.com/o-ba/hotp.
  133. TYPO3com voted on patch set 43 V -1 02/18/21, 00:18 Completed build in 07m 58s on Thu, 18 Feb, 12:18 AM Test Summary: 32623 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9606

  134. core-ci voted on patch set 43 V -1 02/18/21, 00:20 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/821

  135. Oli Bartsch uploaded patch set 44 02/18/21, 00:34

  136. TYPO3com voted on patch set 44 V +1 02/18/21, 00:42 Completed build in 08m 22s on Thu, 18 Feb, 12:42 AM Test Summary: 32623 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9607

  137. core-ci voted on patch set 44 V +1 02/18/21, 00:44 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/822

  138. Benni Mack commented on patch set 45 02/18/21, 08:51

    Published edit on patch set 44.
  139. TYPO3com voted on patch set 45 V +1 02/18/21, 08:59 Completed build in 08m 13s on Thu, 18 Feb, 08:59 AM Test Summary: 32623 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9610

  140. core-ci voted on patch set 45 V +1 02/18/21, 09:02 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/828

  141. Benni Mack uploaded patch set 46 02/18/21, 09:17

  142. TYPO3com voted on patch set 46 V +1 02/18/21, 09:25 Completed build in 08m 29s on Thu, 18 Feb, 09:25 AM Test Summary: 32623 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9613

  143. core-ci voted on patch set 46 V +1 02/18/21, 09:28 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/831

  144. Oli Bartsch uploaded patch set 47 02/18/21, 10:36

  145. TYPO3com voted on patch set 47 V +1 02/18/21, 10:45 Completed build in 08m 50s on Thu, 18 Feb, 10:45 AM Test Summary: 32623 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9614

  146. core-ci voted on patch set 47 V +1 02/18/21, 10:47 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/832

  147. Christian Kuhn commented on patch set 47 02/18/21, 11:25

    Feature-93526-MultiFactorAuthentication.rst:18

    something you own

    Feature-93526-MultiFactorAuthentication.rst:77

    it's not explained yet what a 'locked' provider actually is. maybe drop that sentence or rephrase?

    Feature-93526-MultiFactorAuthentication.rst:85

    'There it's for example possible' denglish

    Feature-93526-MultiFactorAuthentication.rst:91

    unlocks -> unlock

    Feature-93526-MultiFactorAuthentication.rst:92

    requires -> require

    Feature-93526-MultiFactorAuthentication.rst:121

    dot at end of sentence.

    Feature-93526-MultiFactorAuthentication.rst:141

    ['BE'] (missing tick)

    Feature-93526-MultiFactorAuthentication.rst:155

    tab vs. 4 spaces

    Feature-93526-MultiFactorAuthentication.rst:157

    rest for 'note': .. note:: note text next section

    Feature-93526-MultiFactorAuthentication.rst:187

    tab vs. spaces

    Feature-93526-MultiFactorAuthentication.rst:192

    by default

    Feature-93526-MultiFactorAuthentication.rst:200

    tab vs. spaces

    Feature-93526-MultiFactorAuthentication.rst:229

    can be useful

    Feature-93526-MultiFactorAuthentication.rst:246

    MfaProviderProperyManager typo MfaProviderProper-t-yManager ?!
  148. Oli Bartsch uploaded patch set 48 02/18/21, 12:10

  149. Oli Bartsch commented on patch set 48 02/18/21, 12:12

    Feature-93526-MultiFactorAuthentication.rst:18 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:77 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:85 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:91 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:92 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:121 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:141 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:155 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:157 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:187 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:192 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:200 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:229 Reply

    Done

    Feature-93526-MultiFactorAuthentication.rst:246 Reply

    Done
  150. TYPO3com voted on patch set 48 V +1 02/18/21, 12:18 Completed build in 08m 13s on Thu, 18 Feb, 12:18 PM Test Summary: 32623 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9615

  151. core-ci voted on patch set 48 V +1 02/18/21, 12:21 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/833

  152. Benni Mack uploaded patch set 49 02/18/21, 12:26

  153. Benni Mack voted on patch set 49 V +1 02/18/21, 12:31

  154. TYPO3com voted on patch set 49 V +1 02/18/21, 12:35 Completed build in 08m 23s on Thu, 18 Feb, 12:35 PM Test Summary: 32632 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9616

  155. core-ci voted on patch set 49 V +1 02/18/21, 12:36 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/834

  156. Benjamin Franzke commented on patch set 49 V +1 CR -1 02/18/21, 12:59

    Works good so far, also EXT:mfa_webauthn and EXT:sf_yubikey (still) work. on it, reviewing…

    Setup.html:31

    What is the usecase for explicitly naming "recovery codes"?
  157. Christian Kuhn commented on patch set 49 CR -1 02/18/21, 13:25

    on it, too.
  158. Benjamin Franzke commented on patch set 49 CR -2 02/18/21, 13:31

    Hm, we got a problem. The security relies on the "fact" that the user is unable to retrieve a valid token for the backend routes, without fullfilling the MFA. The user session already counts as "fully authenticated", as as soon as the password is verified, but before MFA is done. Now that means, the user only needs a way to retrieve the token to bypass MFA: The problem: There is an easy way: The admin panel. And I'm unsure how to properly fix this, this is not just the mistake of adminpanel, as we'll get the same issue for frontend MFA where no tokens are used at all.
    changes in upcoming patches are to be expected.
    As Webauthn includes U2F support, I'd prefer: Webauthn core provider for FIDO2 and U2F.
  159. Christian Kuhn commented on patch set 49 02/18/21, 14:01

    MfaStatusViewHelper.php:39

    minor detail: I'd prefer userUid here - 'uid' over 'id' is the general wording we use in core, and it's also the table primary key field name.

    RecoveryCodes.php:36

    this may need an adaption: always working with 'defaultHashInstance' will fail when a user created some codes and the default hash instance is later changed.

    RecoveryCodes.php:46

    this method could have some unit test coverage.
  160. Oli Bartsch uploaded patch set 50 02/18/21, 15:13

  161. TYPO3com voted on patch set 50 V -1 02/18/21, 15:22 Completed build in 08m 31s on Thu, 18 Feb, 03:22 PM Test Summary: 2 of 32889 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9618

  162. core-ci voted on patch set 50 V -1 02/18/21, 15:23 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/836

  163. Benni Mack uploaded patch set 51 02/18/21, 21:36

  164. TYPO3com voted on patch set 51 V -1 02/18/21, 21:36 Completed build in 13s on Thu, 18 Feb, 09:36 PM Test Summary: No tests found Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9620

  165. core-ci voted on patch set 51 V -1 02/18/21, 21:37 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/838

  166. Benjamin Franzke uploaded patch set 52 02/18/21, 21:55

  167. Benjamin Franzke commented on patch set 52 02/18/21, 21:55

    Reply

    Done

    Reply

    Done

    MfaProviderPropertyManager.php:214

    Looks like we could only have `createFromManifest`, and then we'd maybe just name it `::create()`. but that is (very) low priority right now, and could be adapted later on.

    Setup.html:15

    Implies backend usage. Later on this should be an ajax action, but for now window.location.refresh() would be sufficent(?)
  168. Benni Mack uploaded patch set 53 02/18/21, 21:59

  169. TYPO3com voted on patch set 52 V -1 02/18/21, 22:03 Completed build in 08m 04s on Thu, 18 Feb, 10:03 PM Test Summary: 32648 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9621

  170. TYPO3com voted on patch set 53 V -1 02/18/21, 22:06 Completed build in 07m 22s on Thu, 18 Feb, 10:06 PM Test Summary: 2 of 32889 failed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9622

  171. core-ci voted on patch set 52 V -1 02/18/21, 22:07 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/839

  172. core-ci voted on patch set 53 V -1 02/18/21, 22:12 Core CI is not happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/840

  173. Benjamin Franzke uploaded patch set 54 02/18/21, 22:27

  174. TYPO3com voted on patch set 54 V +1 02/18/21, 22:36 Completed build in 08m 54s on Thu, 18 Feb, 10:36 PM Test Summary: 32648 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9623

  175. core-ci voted on patch set 54 V +1 02/18/21, 22:38 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/841

  176. Benjamin Franzke uploaded patch set 55 02/18/21, 23:12

  177. Benjamin Franzke commented on patch set 55 V +1 CR 0 02/18/21, 23:18

    Reply

    Thank you Benni, the issue of token-bypass and is resolved since patchset 51. Therefore removing my -2.
    will review tomorrow.

    Viewport.ts:39

    For anyone who is wondering why I've added this. With the enhanced security in patchset 51, the request to the persistence layer caused an unauthenticated request to the user settings endpoint. This is redirected to the mfa authentication, which means a *new* mfa auth is started. Now, because mfa_webauthn needs to ask the authenticator to sign a challenge, the challenge gets too old, therefore we need to avoid this request.
  178. core-ci voted on patch set 55 V +1 02/18/21, 23:22 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/842

  179. TYPO3com voted on patch set 55 V +1 02/18/21, 23:22 Completed build in 10m 48s on Thu, 18 Feb, 11:22 PM Test Summary: 32648 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9624

  180. Oli Bartsch uploaded patch set 56 02/18/21, 23:56

  181. Oli Bartsch commented on patch set 56 02/18/21, 23:57

    MfaStatusViewHelper.php:39 Reply

    Done

    MfaProviderPropertyManager.php:214 Reply

    Then I would opt-in to only have createFromIdentifier. There is no need to pass the whole object. There are some places where we only got the identifier. Fetching the object through the registry just to pass it in the factory where only getIdentifier is called, seems a bit superflous to me.

    RecoveryCodes.php:36 Reply

    Done

    RecoveryCodes.php:46 Reply

    Done

    Setup.html:15 Reply

    Done

    Setup.html:31 Reply Unresolved

    This was just the idea to give users the possibility to assigned names for the providers so they can e.g. reference them in their storage with a unique name.
  182. TYPO3com voted on patch set 56 V +1 02/19/21, 00:05 Completed build in 08m 24s on Fri, 19 Feb, 12:05 AM Test Summary: 32684 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9625

  1. core-ci voted on patch set 56 V +1 02/19/21, 00:07 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/843

  2. Oli Bartsch uploaded patch set 57 02/19/21, 00:33

  3. Oli Bartsch commented on patch set 56 V -1 02/19/21, 00:35

    Will do an in-deep review tomorrow again. Some things need to be properly checked regarding patchset 51.

    Viewport.ts:39 Reply

    Maybe this is solved with the latest patchset?
  4. Oli Bartsch voted on patch set 57 V -1 02/19/21, 00:36

  5. TYPO3com voted on patch set 57 V +1 02/19/21, 00:42 Completed build in 08m 28s on Fri, 19 Feb, 12:42 AM Test Summary: 32684 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9626

  6. core-ci voted on patch set 57 V +1 02/19/21, 00:43 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/844

  7. Benni Mack commented on patch set 57 02/19/21, 08:09

    MfaProviderPropertyManager.php:214 Reply

    I like your proposal
  8. Benni Mack commented on patch set 57 02/19/21, 08:10

    MfaProviderPropertyManager.php:214 Reply

    Sorry, my comment was to bnf's original statement. I usually prefer handing over objects instead of identifiers because then a) we don't need to re-create or fetch objects b) don't rely on identifiers everywhere, (no strings) but on objects, which makes testing much easier.
  9. Oli Bartsch uploaded patch set 58 02/19/21, 12:03

  10. Oli Bartsch commented on patch set 58 02/19/21, 12:04

    Finished testing, so ready for review again.

    MfaProviderPropertyManager.php:214 Reply

    Done
  11. core-ci voted on patch set 58 V +1 02/19/21, 12:13 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/852

  12. TYPO3com voted on patch set 58 V +1 02/19/21, 12:14 Completed build in 10m 18s on Fri, 19 Feb, 12:14 PM Test Summary: 32684 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9631

  13. Benni Mack commented on patch set 58 V +1 CR +1 02/19/21, 12:20

    @bnf - your turn if you think it's ready to merge
  14. Benjamin Franzke uploaded patch set 59 02/19/21, 12:21

  15. Benjamin Franzke commented on patch set 59 CR -1 02/19/21, 12:22

    on it, reviewing.
  16. TYPO3com voted on patch set 59 V +1 02/19/21, 12:30 Completed build in 08m 50s on Fri, 19 Feb, 12:30 PM Test Summary: 32684 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9632

  17. core-ci voted on patch set 59 V +1 02/19/21, 12:31 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/853

  18. Oli Bartsch uploaded patch set 60 02/19/21, 13:54

  19. Benjamin Franzke voted on patch set 60 CR +1 02/19/21, 13:56

  20. TYPO3com voted on patch set 60 V +1 02/19/21, 14:03 Completed build in 08m 01s on Fri, 19 Feb, 02:03 PM Test Summary: 32684 passed Find logs and detail information at https://bamboo.typo3.com/browse/CORE-GTC-9633

  21. core-ci voted on patch set 60 V +1 02/19/21, 14:05 Core CI is happy: https://core.b13.com/typo3-core/typo3.cms/-/pipelines/854

  22. Benjamin Franzke commented on patch set 60 V +1 02/19/21, 14:14

    Reply

    @benni, yes, I think now it is ready. now it's your turn ;)
    @benni, yes I think now it is ready. Now it's your turn ;)
  23. Christian Kuhn voted on patch set 60 CR +1 02/19/21, 14:37

  24. Benni Mack voted on patch set 60 V +2 CR +2 02/19/21, 14:38

  25. Benni Mack merged patch set 61 02/19/21, 14:38