-
-
Notifications
You must be signed in to change notification settings - Fork 2.1k
Conversation
c83f58c
to
c1a2f5f
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lgtm I think.
reminder that we should test this stuff carefully from a range of clients after deploying it, as it's fragile and the sort of thing that we can accidentally break and not notice until our usage stats drop off a few days later.
# If a password hash was previously stored we will not attempt to | ||
# re-hash and store it. | ||
# | ||
# Note that if the password changes throughout the authentication | ||
# flow this might break, but the data is meant to be consistent | ||
# throughout the flow. | ||
password_hash = await self.auth_handler.get_session_data( | ||
session_id, "password_hash", None | ||
) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I struggled for a while to understand what was going on here, and tbh the comment didn't do much to help.
I think this comment - especially the bit about what happens if the password changes mid-flow - should move down to inside the except InteractiveAuthIncompleteError
block at line 533ish, and we can just say "extract the previously-hashed password from the session" here.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I struggled with where to put this, moving it further up sounds like it would be an improvement.
Is there a particular piece of knowledge that should be documented in comments?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I struggled with where to put this, moving it further up sounds like it would be an improvement.
I'm mostly suggesting the code gets left where it is, and the comment moves down.
Is there a particular piece of knowledge that should be documented in comments?
I don't think so, tbh. I think this code speaks for itself, but I only realised that after I stopped trying to understand the comment, which doesn't really describe what is happening here :)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I re-arranged and re-worded the comments in e05118b, I'm hoping that helps?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yup, lgtm.
I think I've previously written tests for some of the assumptions in various client applications, but definitely agree this needs careful testing! |
I did some testing with this and Element iOS and the registration behavior seems to be the same as with I was using the following configuration:
I then went through the following steps:
This works and the registration completes OK. The behavior that isn't great is that if the link is visited on my desktop (so both the iOS app and web app are "running" at the same time) then sometimes only one of the apps completes registration successfully and the other just spins. Other times both complete registration and you immediately get a prompt to verify the other session. I think the two apps essentially race since we delete the threepid validation information after it is validated. This doesn't seem to match the rest of the UI Auth information which we purposefully leave around as "valid" for a period of time. Note that this behavior is not due to this PR. |
And...I was testing the wrong branch. 🤦 Some of the behavior on this branch does seem a little different. On develop I seem to always end up at the case where both sessions end up valid (e.g. the race isn't triggered). I'm going to look a bit more into this. |
After some investigation it seems that Element iOS is requesting to be sent back to Element Web (via a next link link https://app.element.io/#/register?client_secret=98BC6BD7-CBA4-43AE-83A1-D968C292F5BC-44237-0000FDA97FBAB43A&hs_url=http%3A%2F%2Flocalhost%3A8081&session_id=gTGEIrTURJBbxcATejjmHsJL&is_url=https%3A%2F%2Fvector.im&sid=xDzfSQZUKctJQsRZ on the requestToken API for validating email addresses). This seems wrong, but is not a Synapse issue. |
Synapse 1.19.0rc1 (2020-08-13) ============================== Removal warning --------------- As outlined in the [previous release](https://github.com/matrix-org/synapse/releases/tag/v1.18.0), we are no longer publishing Docker images with the `-py3` tag suffix. On top of that, we have also removed the `latest-py3` tag. Please see [the announcement in the upgrade notes for 1.18.0](https://github.com/matrix-org/synapse/blob/develop/UPGRADE.rst#upgrading-to-v1180). Features -------- - Add option to allow server admins to join rooms which fail complexity checks. Contributed by @lugino-emeritus. ([\#7902](#7902)) - Add an option to purge room or not with delete room admin endpoint (`POST /_synapse/admin/v1/rooms/<room_id>/delete`). Contributed by @dklimpel. ([\#7964](#7964)) - Add rate limiting to users joining rooms. ([\#8008](#8008)) - Add a `/health` endpoint to every configured HTTP listener that can be used as a health check endpoint by load balancers. ([\#8048](#8048)) - Allow login to be blocked based on the values of SAML attributes. ([\#8052](#8052)) - Allow guest access to the `GET /_matrix/client/r0/rooms/{room_id}/members` endpoint, according to MSC2689. Contributed by Awesome Technologies Innovationslabor GmbH. ([\#7314](#7314)) Bugfixes -------- - Fix a bug introduced in Synapse v1.7.2 which caused inaccurate membership counts in the room directory. ([\#7977](#7977)) - Fix a long standing bug: 'Duplicate key value violates unique constraint "event_relations_id"' when message retention is configured. ([\#7978](#7978)) - Fix "no create event in auth events" when trying to reject invitation after inviter leaves. Bug introduced in Synapse v1.10.0. ([\#7980](#7980)) - Fix various comments and minor discrepencies in server notices code. ([\#7996](#7996)) - Fix a long standing bug where HTTP HEAD requests resulted in a 400 error. ([\#7999](#7999)) - Fix a long-standing bug which caused two copies of some log lines to be written when synctl was used along with a MemoryHandler logger. ([\#8011](#8011), [\#8012](#8012)) Updates to the Docker image --------------------------- - We no longer publish Docker images with the `-py3` tag suffix, as [announced in the upgrade notes](https://github.com/matrix-org/synapse/blob/develop/UPGRADE.rst#upgrading-to-v1180). ([\#8056](#8056)) Improved Documentation ---------------------- - Document how to set up a client .well-known file and fix several pieces of outdated documentation. ([\#7899](#7899)) - Improve workers docs. ([\#7990](#7990), [\#8000](#8000)) - Fix typo in `docs/workers.md`. ([\#7992](#7992)) - Add documentation for how to undo a room shutdown. ([\#7998](#7998), [\#8010](#8010)) Internal Changes ---------------- - Reduce the amount of whitespace in JSON stored and sent in responses. Contributed by David Vo. ([\#7372](#7372)) - Switch to the JSON implementation from the standard library and bump the minimum version of the canonicaljson library to 1.2.0. ([\#7936](#7936), [\#7979](#7979)) - Convert various parts of the codebase to async/await. ([\#7947](#7947), [\#7948](#7948), [\#7949](#7949), [\#7951](#7951), [\#7963](#7963), [\#7973](#7973), [\#7975](#7975), [\#7976](#7976), [\#7981](#7981), [\#7987](#7987), [\#7989](#7989), [\#8003](#8003), [\#8014](#8014), [\#8016](#8016), [\#8027](#8027), [\#8031](#8031), [\#8032](#8032), [\#8035](#8035), [\#8042](#8042), [\#8044](#8044), [\#8045](#8045), [\#8061](#8061), [\#8062](#8062), [\#8063](#8063), [\#8066](#8066), [\#8069](#8069), [\#8070](#8070)) - Move some database-related log lines from the default logger to the database/transaction loggers. ([\#7952](#7952)) - Add a script to detect source code files using non-unix line terminators. ([\#7965](#7965), [\#7970](#7970)) - Log the SAML session ID during creation. ([\#7971](#7971)) - Implement new experimental push rules for some users. ([\#7997](#7997)) - Remove redundant and unreliable signature check for v1 Identity Service lookup responses. ([\#8001](#8001)) - Improve the performance of the register endpoint. ([\#8009](#8009)) - Reduce less useful output in the newsfragment CI step. Add a link to the changelog section of the contributing guide on error. ([\#8024](#8024)) - Rename storage layer objects to be more sensible. ([\#8033](#8033)) - Change the default log config to reduce disk I/O and storage for new servers. ([\#8040](#8040)) - Add an assertion on `prev_events` in `create_new_client_event`. ([\#8041](#8041)) - Add a comment to `ServerContextFactory` about the use of `SSLv23_METHOD`. ([\#8043](#8043)) - Log `OPTIONS` requests at `DEBUG` rather than `INFO` level to reduce amount logged at `INFO`. ([\#8049](#8049)) - Reduce amount of outbound request logging at `INFO` level. ([\#8050](#8050)) - It is no longer necessary to explicitly define `filters` in the logging configuration. (Continuing to do so is redundant but harmless.) ([\#8051](#8051)) - Add and improve type hints. ([\#8058](#8058), [\#8064](#8064), [\#8060](#8060), [\#8067](#8067))
* commit '66f24449d': Improve performance of the register endpoint (#8009)
This has a few tweaks to the register (and modify password) endpoints that should improve performance a bit:
It also has a couple of other changes:
AuthHandler.check_auth
toAuthHandler.check_ui_auth
to avoid clashes withUserInteractiveAuthChecker
.This should be reviewable commit-by-commit.
Fixes #7956