Summary:
Ref T13222. See PHI873. Currently, when applications prompt users to enter MFA, their session upgrades as a side effect.
In some cases (like managing your email addresses) it makes sense to upgrade your session for a little while since it's common to make multiple edits in sequence (add a new address, make it primary, remove an old address). We generally want MFA to stay out of the way and not feel annoying.
In other cases, we don't expect multiple high-security actions in a row. Notably, PHI873 looks at more "one-shot" use cases where a prompt is answering a specific workflow. We already have at least one of these in the upstream: answering an MFA prompt when signing a Legalpad document.
Introduce a "token" workflow (in contrast to the existing "session") workflow that just does a one-shot prompt without upgrading your session statefully. Then, make Legalpad use this new workflow.
Note that this workflow has a significant problem: if the form submission is invalid for some other reason, we re-prompt you on resubmit. In Legalpad, this workflow looks like:
- Forget to check the "I agree" checkbox.
- Submit the form.
- Get prompted for MFA.
- Answer MFA prompt.
- Get dumped back to the form with an error.
- When you fix the error and submit again, you have to do another MFA check.
This isn't a fatal flaw in Legalpad, but would become a problem with wider adoption. I'll work on fixing this (so the MFA token sticks to the form) in the next set of changes.
Roughly, this is headed toward "MFA sticks to the form/workflow" instead of "MFA sticks to the user/session".
Test Plan:
- Signed a legalpad document with MFA enabled.
- Was prompted for MFA.
- Session no longer upgraded (no purple "session in high security" badge).
- Submitted form with error, answered MFA, fixed error, submitted form again.
- Bad behavior: got re-prompted for MFA. In the future, MFA should stick to the form.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13222
Differential Revision: https://secure.phabricator.com/D19843
Summary:
Ref T6960. Ref T13217. Ref T13216. Depends on D19811. Use the recently-introduced "%P" conversion ("Password/Secret") to load sessions in SessionEngine.
This secret isn't critical to protect (it's the //hash// of the actual secret and not useful to attackers on its own) but it shows up on every page in DarkConsole and is an obvious case where `%P` is a more appropriate conversion.
Test Plan:
Note "*********" in the middle of the output here, instead of a session key hash:
{F6012805}
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13217, T13216, T6960
Differential Revision: https://secure.phabricator.com/D19812
Summary: Depends on D19789. Ref T13217. Continue updating things to use the new %Q-flavored conversions instead of smushing a bunch of strings together.
Test Plan: Browsed around, far fewer errors. These changes are largely mechanical in nature.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13217
Differential Revision: https://secure.phabricator.com/D19790
Summary: Depends on D19785. Ref T13217. This converts many of the most common clause construction pathways to the new %Q / %LQ / %LO / %LA / %LJ semantics.
Test Plan: Browsed around a bunch, saw fewer warnings and no obvious behavioral errors. The transformations here are generally mechanical (although I did them by hand).
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: hach-que
Maniphest Tasks: T13217
Differential Revision: https://secure.phabricator.com/D19789
Summary:
See <https://hackerone.com/reports/435648>. We currently use 80-bit TOTP keys. The RFC suggests 128 as a minimum and recommends 160.
The math suggests that doing the hashing for an 80-bit key is hard (slightly beyond the reach of a highly motivated state actor, today) but there's no reason not to use 160 bits instead to put this completely out of reach.
See some additional discussion on the HackerOne report about enormous key sizes, number of required observations, etc.
Test Plan: Added a new 160-bit TOTP factor to Google Authenticator without issue.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D19792
Summary:
Ref T13216. We occasionally receive HackerOne reports concerned that you can select your username as a password. I suspect very few users actually do this and that this is mostly a compliance/checklist sort of issue, not a real security issue.
On this install, we have about 41,000 user accounts. Of these, 100 have their username as a password (account or VCS). A substantial subset of these are either explicitly intentional ("demo", "bugmenot") or obvious test accounts ("test" in name, or name is a nonsensical string of gibberish, or looks like "tryphab" or similar) or just a bunch of numbers (?), or clearly a "researcher" doing this on purpose (e.g., name includes "pentest" or "xss" or similar).
So I'm not sure real users are actually very inclined to do this, and we can't really ever stop them from picking awful passwords anyway. But we //can// stop researchers from reporting that this is an issue.
Don't allow users to select passwords which contain words in a blocklist: their username, real name, email addresses, or the install's domain name. These words also aren't allowed to contain the password (that is, neither your password nor your username may be a substring of the other one). We also do a little normalization to try to split apart email addresses, domains, and real names, so I can't have "evan1234" as my password.
Test Plan:
- Added unit tests and made them pass.
- Tried to set my password to a bunch of variations of my username / email / domain name / real name / etc, got rejected.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13216
Differential Revision: https://secure.phabricator.com/D19776
Summary:
Ref T13202. See <https://discourse.phabricator-community.org/t/2fa-input-box-isnt-hinted-as-a-password-so-browsers-suggest-auto-fills/1959>.
If browsers are autofilling this, I think browser behavior here is bad, but behavior is probably better on the balance if we hint this as `autocomplete="off"` and this is a minor concesssion.
Test Plan:
- I couldn't immediately get any browser to try to autofill this field (perhaps I've disabled autofill, or just not enabled it aggressively?), but this change didn't break anything.
- After the change, answered a TOTP prompt normally.
- After the change, inspected page content and saw `autocomplete="off"` on the `<input />` node.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13202
Differential Revision: https://secure.phabricator.com/D19722
Summary:
Ref T13197. See PHI873. Record when a user has MFA'd and add a little icon to the transaction, similar to the exiting "Silent" icon.
For now, this just makes this stuff more auditable. Future changes may add ways to require MFA for certain specific transactions, outside of the ones that already always require MFA (like revealing credentials).
Test Plan: {F5877960}
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13197
Differential Revision: https://secure.phabricator.com/D19665
Summary:
Depends on D19585. Ref T13164. This is a precursor for D19586, which causes Editors to start doing more explicit CAN_EDIT checks.
Passwords have an Editor, but don't actually define a CAN_EDIT capability. Define one (you can edit a password if you can edit the object the password is associated with).
(Today, this object is always a User -- this table just unified VCS passwords and Account passwords so they can be handled more consistently.)
Test Plan:
- With D19586, ran unit tests and got a pass.
- Edited my own password.
- Tried to edit another user's password and wasn't permitted to.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13164
Differential Revision: https://secure.phabricator.com/D19592
Summary:
Fixes T12397. Ref T13164. See PHI801.
Several installs have hit various use cases where the path on disk where Phabricator lives changes at runtime. Currently, `bin/ssh-auth` caches a flat file which includes the path to `bin/ssh-exec`, so this may fall out of date if `phabricator/` moves.
These use cases have varying strengths of legitimacy, but "we're migrating to a new set of hosts and the pool is half old machines and half new machines" seems reasonably compelling and not a problem entirely of one's own making.
Test Plan:
- Compared output on `master` to output after change, found them byte-for-byte identical.
- Moved `phabricator/` to `phabricator2/`, ran `bin/ssh-auth`, got updated output.
- Added a new SSH key, saw it appear in the output.
- Grepped for `AUTHFILE_CACHEKEY` (no hits).
- Dropped the cache, verified that the file regenerates cleanly.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13164, T12397
Differential Revision: https://secure.phabricator.com/D19568
Summary:
Ref T13065. `mailKey`s are a private secret for each object. In some mail configurations, they help us ensure that inbound mail is authentic: when we send you mail, the "Reply-To" is "T123+456+abcdef".
- The `T123` is the object you're actually replying to.
- The `456` is your user ID.
- The `abcdef` is a hash of your user account with the `mailKey`.
Knowing this hash effectively proves that Phabricator has sent you mail about the object before, i.e. that you legitimately control the account you're sending from. Without this, anyone could send mail to any object "From" someone else, and have comments post under their username.
To generate this hash, we need a stable secret per object. (We can't use properties like the PHID because the secret has to be legitimately secret.)
Today, we store these in `mailKey` properties on the actual objects, and manually generate them. This results in tons and tons and tons of copies of this same ~10 lines of code.
Instead, just store them in the Mail application and generate them on demand. This change also anticipates possibly adding flags like "must encrypt" and "original subject", which are other "durable metadata about mail transmission" properties we may have use cases for eventually.
Test Plan:
- See next change for additional testing and context.
- Sent mail about Herald rules (next change); saw mail keys generate cleanly.
- Destroyed a Herald rule with a mail key, saw the mail properties get nuked.
- Grepped for `getMailKey()` and converted all callsites I could which aren't the copy/pasted boilerplate present in 50 places.
- Used `bin/mail receive-test --to T123` to test normal mail receipt of older-style objects and make sure that wasn't broken.
Reviewers: amckinley
Reviewed By: amckinley
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Maniphest Tasks: T13065
Differential Revision: https://secure.phabricator.com/D19399
Summary:
Ref T4340. Some "Register/Login" and "Link External Account" buttons are forms which submit to third-party sites. Whitelist these targets when pages render an OAuth form.
Safari, at least, also prevents a redirect to a third-party domain after a form submission to the local domain, so when we first redirect locally (as with Twitter and other OAuth1 providers) we need to authorize an additional URI.
Test Plan: Clicked all my registration buttons locally without hitting CSP issues.
Maniphest Tasks: T4340
Differential Revision: https://secure.phabricator.com/D19159
Summary:
Depends on D19155. Ref T13094. Ref T4340.
We can't currently implement a strict `form-action 'self'` content security policy because some file downloads rely on a `<form />` which sometimes POSTs to the CDN domain.
Broadly, stop generating these forms. We just redirect instead, and show an interstitial confirm dialog if no CDN domain is configured. This makes the UX for installs with no CDN domain a little worse and the UX for everyone else better.
Then, implement the stricter Content-Security-Policy.
This also removes extra confirm dialogs for downloading Harbormaster build logs and data exports.
Test Plan:
- Went through the plain data export, data export with bulk jobs, ssh key generation, calendar ICS download, Diffusion data, Paste data, Harbormaster log data, and normal file data download workflows with a CDN domain.
- Went through all those workflows again without a CDN domain.
- Grepped for affected symbols (`getCDNURI()`, `getDownloadURI()`).
- Added an evil form to a page, tried to submit it, was rejected.
- Went through the ReCaptcha and Stripe flows again to see if they're submitting any forms.
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Maniphest Tasks: T13094, T4340
Differential Revision: https://secure.phabricator.com/D19156
Summary: See D19117. Instead of automatically figuring this out inside `phutil_tag()`, explicitly add rel="noreferrer" at the application level to all external links.
Test Plan:
- Grepped for `_blank`, `isValidRemoteURIForLink`, checked all callsites for user-controlled data.
- Created a link menu item, verified noreferrer in markup.
- Created a link custom field, verified no referrer in markup.
- Verified noreferrer for `{nav href=...}`.
Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam
Differential Revision: https://secure.phabricator.com/D19118
Summary: Ref T13054. Fixes T12714. Applies read locks to all transactions instead of only a very select subset (chat messages in Conpherence).
Test Plan: See <T13054#235650> for discussion and testing.
Maniphest Tasks: T13054, T12714
Differential Revision: https://secure.phabricator.com/D19059
Summary:
Depends on D19009. Ref T13053. For "Must Encrypt" mail, we must currently strip the "Thread-Topic" header because it sometimes contains sensitive information about the object.
I don't actually know if this header is useful or anyting uses it. My understanding is that it's an Outlook/Exchange thing, but we also implement "Thread-Index" which I think is what Outlook/Exchange actually look at. This header may have done something before we implemented "Thread-Index", or maybe never done anything. Or maybe older versions of Excel/Outlook did something with it and newer versions don't, or do less. So it's possible that an even better fix here would be to simply remove this, but I wasn't able to convince myself of that after Googling for 10 minutes and I don't think it's worth hours of installing Exchange/Outlook to figure out. Instead, I'm just trying to simplify our handling of this header for now, and maybe some day we'll learn more about Exchange/Outlook and can remove it.
In a number of cases we already use the object monogram or PHID as a "Thread-Topic" without users ever complaining, so I think that if this header is useful it probably isn't shown to users, or isn't shown very often (e.g., only in a specific "conversation" sub-view?). Just use the object PHID (which should be unique and stable) as a thread-topic, everywhere, automatically.
Then allow this header through for "Must Encrypt" mail.
Test Plan: Processed some local mail, saw object PHIDs for "Thread-Topic" headers.
Reviewers: amckinley
Maniphest Tasks: T13053
Differential Revision: https://secure.phabricator.com/D19012
Summary: Depends on D18928. Ref T13043. Add some automated test coverage for SSH revocation rules.
Test Plan: Ran tests, got a clean bill of health.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18929
Summary:
Ref T13043. In an earlier change I updated this langauge from "Deactivate" to "Revoke", but the behavior doesn't quite match.
This table has a unique key on `<isActive, keyBody>`, which enforces the rule that "a key can only be active for one unique user".
However, we set `isActive` to `null` when we revoke a key, and multiple rows are allowed to have the value `<null, "asdf">` (since a `null` column in a unique key basically means "don't enforce this unique key").
This is intentional, to support this workflow:
- You add key X to bot A.
- Whoops, wrong account.
- You revoke key X from bot A.
- You add key X to bot B.
This isn't necessarily a great workflow -- ideally, you'd throw key X away and go generate a new key after you realize you made a mistake -- but it's the sort of practical workflow that users are likely to expect and want to see work ("I don't want to generate a new key, it's already being used by 5 other services and cycling it is a ton of work and this is just a test install for my dog anyway."), and there's no technical reason we can't support it.
To prevent users from adding keys on the revocation list back to their account, just check explicitly.
(This is probably better in general anyway, because "cert-authority" support from PHI269 may mean that two keys are "equivalent" even if their text differs, and we may not be able to rely on a database test anyway.)
Test Plan:
- Added the key `ssh-rsa asdf` to my account.
- Revoked it.
- Tried to add it again.
- Before patch: worked.
- After patch: error, "this key has been revoked".
- Added it to a different account (the "I put it on the wrong bot" workflow).
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18928
Summary:
Depends on D18908. Ref T13043. Allow users to get information about what revokers do with a new `--list` flag.
You can use `--list --type <key>` to get information about a specfic revoker.
Test Plan: Ran `bin/auth revoke --list`, saw a list of revokers with useful information.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18910
Summary:
Depends on D18907. Ref T13043. Ref T12509. We have some weird old password digest behavior that isn't terribly concerning, but also isn't great.
Specifically, old passwords were digested in weird ways before being hashed. Notably, account passwords were digested with usernames, so your password stops working if your username is chagned. Not the end of the world, but silly.
Mark all existing hashes as "v1", and automatically upgrade then when they're used or changed. Some day, far in the future, we could stop supporting these legacy digests and delete the code and passwords and just issue upgrade advice ("Passwords which haven't been used in more than two years no longer work."). But at least get things on a path toward sane, modern behavior.
Test Plan: Ran migration. Spot-checked that everthing in the database got marked as "v1". Used an existing password to login successfully. Verified that it was upgraded to a `null` (modern) digest. Logged in with it again.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043, T12509
Differential Revision: https://secure.phabricator.com/D18908
Summary:
Depends on D18906. Ref T13043. When SSH keys are edited, we normally include a warning that if you don't recognize the activity you might have problems in the mail body.
Currently, this warning is also shown for revocations with `bin/auth revoke --type ssh`. However, these revocations are safe (revocations are generally not dangerous anyway) and almost certainly legitimate and administrative, so don't warn users about them.
Test Plan:
- Created and revoked a key.
- Creation mail still had warning; revocation mail no longer did.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18907
Summary:
Depends on D18904. Ref T13043. If an attacker compromises a victim's session and bypasses their MFA, they can try to guess the user's current account password by making repeated requests to change it: if they guess the right "Old Password", they get a different error than if they don't.
I don't think this is really a very serious concern (the attacker already got a session and MFA, if configured, somehow; many installs don't use passwords anyway) but we get occasional reports about it from HackerOne. Technically, it's better policy to rate limit it, and this should reduce the reports we receive.
Test Plan: Tried to change password over and over again, eventually got rated limited. Used `bin/auth unlimit` to clear the limit, changed password normally without issues.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18906
Summary:
Ref T13043. This moves user account passwords to the new shared infrastructure.
There's a lot of code changes here, but essentially all of it is the same as the VCS password logic in D18898.
Test Plan:
- Ran migration.
- Spot checked table for general sanity.
- Logged in with an existing password.
- Hit all error conditions on "change password", "set password", "register new account" flows.
- Verified that changing password logs out other sessions.
- Verified that revoked passwords of a different type can't be selected.
- Changed passwords a bunch.
- Verified that salt regenerates properly after password change.
- Tried to login with the wrong password, which didn't work.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18903
Summary:
Ref T13043. We have ~4 copies of this logic (registration, lost password recovery, set password, set VCS password).
Currently it varies a bit from case to case, but since it's all going to be basically identical once account passwords swap to the new infrastructure, bring it into the Engine so it can live in one place.
This also fixes VCS passwords not being affected by `account.minimum-password-length`.
Test Plan: Hit all errors in "VCS Password" panel. Successfully changed password.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18902
Summary:
Ref T13043. This cleans some things up to prepare for moving account passwords to shared infrastructure.
Currently, the (very old, fairly unusual) `bin/accountadmin` tool can set account passwords. This is a bit weird, generally not great, and makes upgrading to shared infrastructure more difficult. Just get rid of this to simplify things. Many installs don't have passwords and this is pointless and unhelpful in those cases.
Instead, let `bin/auth recover` recover any account, not just administrator accounts. This was a guardrail against administrative abuse, but it has always seemed especially flimsy (since anyone who can run the tool can easily comment out the checks) and I use this tool in cluster support with some frequency, occasionally just commenting out the checks. This is generally a better solution than actually setting a password on accounts anyway. Just get rid of the check and give users enough rope to shoot themselves in the foot with if they truly desire.
Test Plan:
- Ran `bin/accountadmin`, didn't get prompted to swap passwords anymore.
- Ran `bin/auth recover` to recover a non-admin account.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18901
Summary:
Ref T13043. In D18898 I moved VCS passwords to the new shared infrastructure.
Before account passwords can move, we need to make two changes:
- For legacy reasons, VCS passwords and Account passwords have different "digest" algorithms. Both are more complicated than they should be, but we can't easily fix it without breaking existing passwords. Add a `PasswordHashInterface` so that objects which can have passwords hashes can implement custom digest logic for each password type.
- Account passwords have a dedicated external salt (`PhabricatorUser->passwordSalt`). This is a generally reasonable thing to support (since not all hashers are self-salting) and we need to keep it around so existing passwords still work. Add salt support to `AuthPassword` and make it generate/regenerate when passwords are updated.
Then add a nice story about password digestion.
Test Plan: Ran migrations. Used an existing VCS password; changed VCS password. Tried to use a revoked password. Unit tests still pass. Grepped for callers to legacy `PhabricatorHash::digestPassword()`, found none.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18900
Summary:
Ref T13043. When we verify a password and a better hasher is available, we automatically upgrade the stored hash to the stronger hasher.
Add test coverage for this workflow and fix a few bugs and issues, mostly related to shuffling the old hasher name into the transaction.
This doesn't touch anything user-visible yet.
Test Plan: Ran unit tests.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18897
Summary:
Ref T13043. This provides a new piece of shared infrastructure that VCS passwords and account passwords can use to validate passwords that users enter.
This isn't reachable by anything yet.
The test coverage of the "upgrade" flow (where we rehash a password to use a stronger hasher) isn't great in this diff, I'll expand that in the next change and then start migrating things.
Test Plan: Added a bunch of unit tests.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18896
Summary: Ref T13043. I'd like to replace the manual credential revocation in the Phacility export workflow with shared code in `bin/auth revoke`, but we need it to run non-interactively. Add a `--force` flag purely to make our lives easier.
Test Plan: Ran `bin/auth revoke --everywhere ...` with and without `--force`. Got prompted without, got total annihilation with.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18895
Summary:
Ref T13043. Currently:
- Passwords are stored separately in the "VCS Passwords" and "User" tables and don't share as much code as they could.
- Because User objects are all over the place in the code, password hashes are all over the place too (i.e., often somewhere in process memory). This is a very low-severity, theoretical sort of issue, but it could make leaving a stray `var_dump()` in the code somewhere a lot more dangerous than it otherwise is. Even if we never do this, third-party developers might. So it "feels nice" to imagine separating this data into a different table that we rarely load.
- Passwords can not be //revoked//. They can be //deleted//, but users can set the same password again. If you believe or suspect that a password may have been compromised, you might reasonably prefer to revoke it and force the user to select a //different// password.
This change prepares to remedy these issues by adding a new, more modern dedicated password storage table which supports storing multiple password types (account vs VCS), gives passwords real PHIDs and transactions, supports DestructionEngine, supports revocation, and supports `bin/auth revoke`.
It doesn't actually make anything use this new table yet. Future changes will migrate VCS passwords and account passwords to this table.
(This also gives third party applications a reasonable place to store password hashes in a consistent way if they have some need for it.)
Test Plan: Added some basic unit tests to cover general behavior. This is just skeleton code for now and will get more thorough testing when applications move.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18894
Summary: Ref T13043. Adds CLI support for revoking SSH keys. Also retargets UI language from "Deactivate" to "Revoke" to make it more clear that this is a one-way operation. This operation is already correctly implemented as a "Revoke" operation.
Test Plan: Used `bin/auth revoke --type ssh` to revoke keys, verified they became revoked (with proper transactions) in the UI. Revoked keys from the web UI flow.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18893
Summary: Ref T13043. Allows CLI revocation of login sessions.
Test Plan: Used `bin/auth revoke --type session` with `--from` and `--everywhere` to revoke sessions. Saw accounts get logged out in web UI.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13043
Differential Revision: https://secure.phabricator.com/D18892
Summary:
See PHI223. Ref T13024. There's a remaining registration/login order issue after the other changes in T13024: we lose track of the current URI when we go through the MFA flow, so we can lose "Set Password" at the end of the flow.
Specifically, the flow goes like this today:
- User clicks the welcome link in email.
- They get redirected to the "set password" settings panel.
- This gets pre-empted by Legalpad (although we'll potentially survive this with the URI intact).
- This also gets pre-empted by the "Set MFA" workflow. If the user completes this flow, they get redirected to a `/auth/multifactor/?id=123` sort of URI to highlight the factor they added. This causes us to lose the `/settings/panel/password/blah/blah?key=xyz` URI.
The ordering on this is also not ideal; it's preferable to start with a password, then do the other steps, so the user can return to the flow more easily if they are interrupted.
Resolve this by separating the "change your password" and "set/reset your password" flows onto two different pages. This copy/pastes a bit of code, but both flows end up simpler so it feels reasonable to me overall.
We don't require a full session for "set/reset password" (so you can do it if you don't have MFA/legalpad yet) and do it first.
This works better and is broadly simpler for users.
Test Plan:
- Required MFA + legalpad, invited a user via email, registered.
- Before: password set flow got lost when setting MFA.
- After: prompted to set password, then sign documents, then set up MFA.
- Reset password (with MFA confgiured, was required to MFA first).
- Tried to reset password without a valid reset key, wasn't successful.
- Changed password using existing flow.
- Hit various (all?) error cases (short password, common password, mismatch, missing password, etc).
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13024
Differential Revision: https://secure.phabricator.com/D18840
Summary:
See <https://discourse.phabricator-community.org/t/activation-link-in-welcome-mail-only-works-if-new-user-isnt-semi-logged-in/740/7>.
In T13024, I rewrote the main menu bar to hide potentially sensitive items (like notification and message counts and saved search filters) until users fully log in.
However, the "Log In" item got caught in this too. For clarity, rename `shouldAllowPartialSessions()` to `shouldRequireFullSession()` (since logged-out users don't have any session at all, so it would be a bit misleading to say that "Log In" "allows" a partial session). Then let "Log In" work again for logged-out users.
(In most cases, users are prompted to log in when they take an action which requires them to be logged in -- like creating or editing an object, or adding comments -- so this item doesn't really need to exist. However, it aligns better with user expectations in many cases to have it present, and some reasonable operations like "Check if I have notifications/messages" don't have an obvious thing to click otherwise.)
Test Plan: Viewed site in an incognito window, saw "Log In" button again. Browsed normally, saw normal menu.
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D18818
Summary: Depends on D18791. Ref T13024. This clears up another initialization order issue, where an unverified address could prevent MFA enrollment.
Test Plan: Configured both verification required and MFA required, clicked "Add Factor", got a dialog for the workflow.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13024
Differential Revision: https://secure.phabricator.com/D18792
Summary:
Depends on D18790. Ref T13024. Fixes T8335. Currently, "unapproved" and "disabled" users are bundled together. This prevents users from completing some registration steps (verification, legalpad documents, MFA enrollment) before approval.
Separate approval out and move it to the end so users can do all the required enrollment stuff on their end before we roadblock them.
Test Plan: Required approval, email verification, signatures, and MFA. Registered an account. Verified email, signed documents, enrolled in MFA, and then got prompted to wait for approval.
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T13024, T8335
Differential Revision: https://secure.phabricator.com/D18791
Summary: Noticed a couple of typos in the docs, and then things got out of hand.
Test Plan:
- Stared at the words until my eyes watered and the letters began to swim on the screen.
- Consulted a dictionary.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: epriestley, yelirekim, PHID-OPKG-gm6ozazyms6q6i22gyam
Differential Revision: https://secure.phabricator.com/D18693
Summary:
See PHI78. The user was getting this message and (reasonably) interpreted it to mean "reset mail can never be sent to unverified addresses".
Reword it to be more clear, albeit an entire paragraph long. I don't really have a good solution in these cases where we'd need a whole page to explain what's happening (this, plus "we can't tell you which address you should use because an attacker could get information if we did" and "this rule defuses the risk that an opportunistic attacker may try to compromise your account after you add an email you don't own by mistake"). We could write it up separately and link to it, but I feel like that stuff tends to get out of date.
Just land somewhere in the middle.
Test Plan: {F5189105}
Reviewers: amckinley
Reviewed By: amckinley
Differential Revision: https://secure.phabricator.com/D18630
Summary: Ref T12964. This feels like a cheat, but works well. Just redirect the user back to the form they came from instead of to the key page.
Test Plan: Add a key to a user profile, add a key to an Alamanac device. Grep for PhabricatorAuthSSHKeyTableView and check all locations.
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: Korvin
Maniphest Tasks: T12964
Differential Revision: https://secure.phabricator.com/D18445
Summary: Cursory research indicates that "login" is a noun, referring to a form, and "log in" is a verb, referring to the action of logging in. I went though every instances of 'login' I could find and tried to clarify all this language. Also, we have "Phabricator" on the registration for like 4-5 times, which is a bit verbose, so I tried to simplify that language as well.
Test Plan: Tested logging in and logging out. Pages feel simpler.
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D18322
Summary:
Ref M1476. Currently, `setColor('simple')` is meaningful. Instead, `setButtonType('simple')`.
Depends on D18047.
Test Plan: Looked at UI examples, Phame, Auth. Notifications mooted by D18047.
Reviewers: chad
Reviewed By: chad
Differential Revision: https://secure.phabricator.com/D18048
Summary:
Fixes T12554. The SSH key cache contains usernames, but is not currently dirtied on username changes.
An alternative solution would be to use user PHIDs instead of usernames in the file, which would make this unnecessary, but that would make debugging a bit harder. For now, I think this small added complexity is worth the easier debugging, but we could look at this again if cache management gets harder in the future.
Test Plan:
- Added a key as `ducksey`, ran `bin/ssh-auth`, saw key immediately.
- Renamed `ducksey` to `ducker`, ran `bin/ssh-auth`, saw username change immediately.
- Added another key as `ducker`, ran `bin/ssh-auth`, saw key immediately.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T12554
Differential Revision: https://secure.phabricator.com/D17687
Summary: Ref T12509. This encourages code to move away from HMAC+SHA1 by making the method name more obviously undesirable.
Test Plan: `grep`, browsed around.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T12509
Differential Revision: https://secure.phabricator.com/D17632
Summary:
Ref T12509. This adds support for HMAC+SHA256 (instead of HMAC+SHA1). Although HMAC+SHA1 is not currently broken in any sense, SHA1 has a well-known collision and it's good to look at moving away from HMAC+SHA1.
The new mechanism also automatically generates and stores HMAC keys.
Currently, HMAC keys largely use a per-install constant defined in `security.hmac-key`. In theory this can be changed, but in practice essentially no install changes it.
We generally (in fact, always, I think?) don't use HMAC digests in a way where it matters that this key is well-known, but it's slightly better if this key is unique per class of use cases. Principally, if use cases have unique HMAC keys they are generally less vulnerable to precomputation attacks where an attacker might generate a large number of HMAC hashes of well-known values and use them in a nefarious way. The actual threat here is probably close to nonexistent, but we can harden against it without much extra effort.
Beyond that, this isn't something users should really have to think about or bother configuring.
Test Plan:
- Added unit tests.
- Used `bin/files integrity` to verify, strip, and recompute hashes.
- Tampered with a generated HMAC key, verified it invalidated hashes.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T12509
Differential Revision: https://secure.phabricator.com/D17630
Summary:
Ref T12464. This is a very old method which can return an existing file instead of creating a new one, if there's some existing file with the same content.
In the best case this is a bad idea. This being somewhat reasonable predates policies, temporary files, etc. Modern methods like `newFromFileData()` do this right: they share underlying data in storage, but not the actual `File` records.
Specifically, this is the case where we get into trouble:
- I upload a private file with content "X".
- You somehow generate a file with the same content by, say, viewing a raw diff in Differential.
- If the diff had the same content, you get my file, but you don't have permission to see it or whatever so everything breaks and is terrible.
Just get rid of this.
Test Plan:
- Generated an SSH key.
- Viewed a raw diff in Differential.
- (Did not test Phragment.)
Reviewers: chad
Reviewed By: chad
Subscribers: hach-que
Maniphest Tasks: T12464
Differential Revision: https://secure.phabricator.com/D17617
Summary:
Ref T11357. When creating a file, callers can currently specify a `ttl`. However, it isn't unambiguous what you're supposed to pass, and some callers get it wrong.
For example, to mean "this file expires in 60 minutes", you might pass either of these:
- `time() + phutil_units('60 minutes in seconds')`
- `phutil_units('60 minutes in seconds')`
The former means "60 minutes from now". The latter means "1 AM, January 1, 1970". In practice, because the GC normally runs only once every four hours (at least, until recently), and all the bad TTLs are cases where files are normally accessed immediately, these 1970 TTLs didn't cause any real problems.
Split `ttl` into `ttl.relative` and `ttl.absolute`, and make sure the values are sane. Then correct all callers, and simplify out the `time()` calls where possible to make switching to `PhabricatorTime` easier.
Test Plan:
- Generated an SSH keypair.
- Viewed a changeset.
- Viewed a raw diff.
- Viewed a commit's file data.
- Viewed a temporary file's details, saw expiration date and relative time.
- Ran unit tests.
- (Didn't really test Phragment.)
Reviewers: chad
Reviewed By: chad
Subscribers: hach-que
Maniphest Tasks: T11357
Differential Revision: https://secure.phabricator.com/D17616
Summary:
Ref T12313. This puts a UI on revoking credentials after a widespread compromise like Cloudbleed or a local one like copy/pasting a token into public chat.
For now, I'm only providing a revoker for conduit tokens since that's the immediate use case.
Test Plan:
- Revoked in user + type, everything + user, everywhere + type, and everything + everywhere modes.
- Verified that conduit tokens were destroyed in all cases.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T12313
Differential Revision: https://secure.phabricator.com/D17458
