Commit graph

742 commits

Author SHA1 Message Date
Neil Alexander
dacee648f7
Federation for v3/v4 rooms (#954)
* Update gomatrixserverlib

* Default to room version 4

* Update gomatrixserverlib

* Limit prev_events and auth_events

* Fix auth_events, prev_events

* Fix linter issues

* Update gomatrixserverlib

* Fix getState

* Update sytest-whitelist

* Squashed commit of the following:

commit 067b875063
Author: Neil Alexander <neilalexander@users.noreply.github.com>
Date:   Fri Apr 3 14:29:06 2020 +0100

    Invites v2 endpoint (#952)

    * Start converting v1 invite endpoint to v2

    * Update gomatrixserverlib

    * Early federationsender code for sending invites

    * Sending invites sorta happens now

    * Populate invite request with stripped state

    * Remodel a bit, don't reflect received invites

    * Handle invite_room_state

    * Handle room versions a bit better

    * Update gomatrixserverlib

    * Tweak order in destinationQueue.next

    * Revert check in processMessage

    * Tweak federation sender destination queue code a bit

    * Add comments

commit 955244c092
Author: Ben B <benne@klimlive.de>
Date:   Fri Apr 3 12:40:50 2020 +0200

    use custom http client instead of the http DefaultClient (#823)

    This commit replaces the default client from the http lib with a custom one.
    The previously used default client doesn't come with a timeout. This could cause
    unwanted locks.
    That solution chosen here creates a http client in the base component dendrite
    with a constant timeout of 30 seconds. If it should be necessary to overwrite
    this, we could include the timeout in the dendrite configuration.
    Here it would be a good idea to extend the type "Address" by a timeout and
    create an http client for each service.

    Closes #820

    Signed-off-by: Benedikt Bongartz <benne@klimlive.de>

    Co-authored-by: Kegsay <kegan@matrix.org>

* Update sytest-whitelist, sytest-blacklist

* Update go.mod/go.sum

* Add some error wrapping for debug

* Add a NOTSPEC to common/events.go

* Perform state resolution at send_join

* Set default room version to v2 again

* Tweak GetCapabilities

* Add comments to ResolveConflictsAdhoc

* Update sytest-blacklist

* go mod tidy

* Update sytest-whitelist, sytest-blacklist

* Update versions

* Updates from review comments

* Update sytest-blacklist, sytest-whitelist

* Check room versions compatible at make_join, add some comments, update gomatrixserverlib, other tweaks

* Set default room version back to v2

* Update gomatrixserverlib, sytest-whitelist
2020-04-09 15:46:06 +01:00
Neil Alexander
067b875063
Invites v2 endpoint (#952)
* Start converting v1 invite endpoint to v2

* Update gomatrixserverlib

* Early federationsender code for sending invites

* Sending invites sorta happens now

* Populate invite request with stripped state

* Remodel a bit, don't reflect received invites

* Handle invite_room_state

* Handle room versions a bit better

* Update gomatrixserverlib

* Tweak order in destinationQueue.next

* Revert check in processMessage

* Tweak federation sender destination queue code a bit

* Add comments
2020-04-03 14:29:06 +01:00
Ben B
955244c092
use custom http client instead of the http DefaultClient (#823)
This commit replaces the default client from the http lib with a custom one.
The previously used default client doesn't come with a timeout. This could cause
unwanted locks.
That solution chosen here creates a http client in the base component dendrite
with a constant timeout of 30 seconds. If it should be necessary to overwrite
this, we could include the timeout in the dendrite configuration.
Here it would be a good idea to extend the type "Address" by a timeout and
create an http client for each service.

Closes #820

Signed-off-by: Benedikt Bongartz <benne@klimlive.de>

Co-authored-by: Kegsay <kegan@matrix.org>
2020-04-03 11:40:50 +01:00
Alex Chen
2c8950221e
Fix: show-expected-fail-tests.sh ignores # in whitelist and strips quotes in test names in output (#943)
Signed-off-by: Alex Chen <minecnly@gmail.com>
2020-03-31 15:53:09 +01:00
Kegsay
8fbe9f4078
Implement typing over federation (#949)
Also fix a pet peeve of mine: not putting units on things!!!

Manually tested on p2p and works well, with some fudge factor delay.
2020-03-30 16:40:28 +01:00
Kegsay
11a8059bba
Rename the typing server to EDU server (#948)
* Blunt move and sed rename

* Rename common/ refs to typing

* Rename internal stuff in eduserver

* Rename docs and scripts

* Rename constants/filenames, goimports everything to re-order imports
2020-03-30 15:02:20 +01:00
Kegan Dougal
f72b759426 Update to latest go-sqite-js 2020-03-30 12:47:19 +01:00
Neil Alexander
664f31ec98 Ensure state res results are unique 2020-03-30 09:51:45 +01:00
Neil Alexander
05e1ae8745
Further room version wiring (#936)
* Room version 2 by default, other wiring updates, update gomatrixserverlib

* Fix nil pointer exception

* Fix some more nil pointer exceptions hopefully

* Update gomatrixserverlib

* Send all room versions when joining, not just stable ones

* Remove room version cquery

* Get room version when getting events from the roomserver database

* Reset default back to room version 2

* Don't generate event IDs unless needed

* Revert "Remove room version cquery"

This reverts commit a170d58733.

* Query room version in federation API, client API as needed

* Improvements to make_join send_join dance

* Make room server producers use headered events

* Lint tweaks

* Update gomatrixserverlib

* Versioned SendJoin

* Query room version in syncapi backfill

* Handle transaction marshalling/unmarshalling within Dendrite

* Sorta fix federation (kinda)

* whoops commit federation API too

* Use NewEventFromTrustedJSON when getting events from the database

* Update gomatrixserverlib

* Strip headers on federationapi endpoints

* Fix bug in clientapi profile room version query

* Update gomatrixserverlib

* Return more useful error if room version query doesn't find the room

* Update gomatrixserverlib

* Update gomatrixserverlib

* Maybe fix federation

* Fix formatting directive

* Update sytest whitelist and blacklist

* Temporarily disable room versions 3 and 4 until gmsl is fixed

* Fix count of EDUs in logging

* Update gomatrixserverlib

* Update gomatrixserverlib

* Update gomatrixserverlib

* Rely on EventBuilder in gmsl to generate the event IDs for us

* Some review comments fixed

* Move function out of common and into gmsl

* Comment in federationsender destinationqueue

* Update gomatrixserverlib
2020-03-27 16:28:22 +00:00
Suvid Sahay
314da91f1d
Fixed Didn't call Close on UserStreamListener log messages when syncing (#940)
Signed-off-by: Suvid Sahay <sahay.suvid61@gmail.com>

Removing the extra imports
2020-03-26 09:57:18 +00:00
Neil Alexander
0b732d6f45
Use HeaderedEvents in appservice component (#939)
* App service HeaderedEvents

* Fix database queries

* Fix lint error
2020-03-24 15:46:17 +00:00
Kegan Dougal
951b5d5e68 Missing whitelist 2020-03-24 13:39:50 +00:00
Kegsay
6bac7e5efd
Implement backfill over federation (#938)
* Implement history visibility checks for /backfill

Required for p2p to show history correctly.

* Add sytest

* Logging

* Fix two backfill bugs which prevented backfill from working correctly

- When receiving backfill requests, do not send the event that was in the original request.
- When storing backfill results, correctly update the backwards extremity for the room.

* hack: make backfill work multiple times

* add sqlite impl and remove logging

* Linting
2020-03-24 12:20:10 +00:00
Alex Chen
5a1a1ded1b
Fix dendrite config file location in docker guide (#934) 2020-03-23 11:16:17 +00:00
Prateek Sachan
c5b7a17272
Added special case for m.room.member event in appservice component (#874)
* Added special case for m.room.member event in appservice component

* fix review changes

Co-authored-by: Kegsay <kegan@matrix.org>
2020-03-20 21:40:24 +00:00
Neil Alexander
f2030286de
Room server changes for room versions (#930)
* Rearrange state package a bit, add some code to look up the right state resolution algorithm

* Remove shared

* Add GetRoomVersionForRoomNID

* Try to use room version to get correct state resolution algorithm

* Fix room joins over federation

* nolint resolveConflictsV2 because all attempts to break it up so far just result in it being awfully less obvious how it works

* Rename Prepare to NewStateResolution

* Update comments

* Re-add missing tests
2020-03-19 18:33:04 +00:00
Neil Alexander
944d454cb0 Fix SQLite problems after sync API changes 2020-03-19 13:15:35 +00:00
Neil Alexander
ad5849d222
HeaderedEvents in sync API (#922)
* Use HeaderedEvent in syncapi

* Update notifier test

* Fix persisting headered event

* Clean up unused API function

* Fix overshadowed err from linter

* Write headered JSON to invites table too

* Rename event_json to headered_event_json in syncapi database schemae

* Fix invites_table queries

* Update QueryRoomVersionCapabilitiesResponse comment

* Fix syncapi SQLite
2020-03-19 12:07:01 +00:00
Kegsay
bfbf96eec9
p2p: Implement published rooms (#923)
* Create and glue ExternalPublicRoomsProvider into the public rooms component

This is how we will link p2p stuff to dendrite proper.

* Use gmsl structs rather than our own

* Implement federated public rooms

- Make thirdparty endpoint r0 so riot-web loads the public room list

* Typo

* Missing callsites
2020-03-19 11:04:08 +00:00
Prateek Sachan
dc06c69887
Implement /joined_rooms (#911)
* Implemented /joined_rooms

* Removed account endpoint added by mistake

* trigger ci
2020-03-19 10:25:36 +00:00
Abhinav Krishna C K
ec38783192
maintenance: Fix matrix-org#896 use %w format verb to wrap errors (#916)
* maintenance: Fix matrix-org#896 use %w format verb to wrap errors

* In Go version 1.13 a new formatting verb introduced for fmt.Errorf
  %w https://blog.golang.org/go1.13-errors

* update %s to %w to wrap errors.

* Update all instances of error type to use %w

Signed-off-by: Abhinav Krishna C K <me@abhy.me>

Co-authored-by: Kegsay <kegan@matrix.org>
2020-03-18 12:48:51 +00:00
Prateek Sachan
c019ad7086
Log errors from rows.Close (#920)
* Log errors from rows.Close

* fixed imports

* Added contextual messages

* fixed review changes
2020-03-18 10:17:18 +00:00
Neil Alexander
c2bd0b97b3
Get room versions from database (#918)
* Retrieve room version where known in roomserver

* Get room versions in alias code

* Increase gocyclothreshold to 13, since we hit that number a lot

* Remove gocyclo nolint from StoreEvent

* Update interface to get room version from room ID instead of NID

* Remove new API

* Fixed this query for SQLite but not for Postgres
2020-03-17 18:00:10 +00:00
Kegsay
1467cc10d8
bugfix: Fix a bug which caused failures to join rooms over federation (#917)
* bugfix: Fix a bug which caused failures to join rooms over federation

The cause of this was the semantics of `/send_join`'s `auth_chain` response.

Previously, we would only send back the auth chain *for the join event* and
not the entire room state. However, we would then try to check that the
room state is valid, and then be missing auth events.

Now, we send back the entire auth chain for all room state in `/send_join`.
The spec needs to be clarified that this is what the chain should be.

* refactor: split out grabbing state to reduce cyclo complexity
2020-03-17 17:18:48 +00:00
Kegsay
4af8323df3
bugfix: Fix a bug which caused prev_content not to be sent to clients (#919)
I don't know how this ever passed QA... also fix a missing rows.Close()
2020-03-17 16:45:40 +00:00
Neil Alexander
a66c701b29
Add room version to room create request, persist in storage (#915)
* Add room version into createRoomReq

* Extract room version from m.room.create event when persisting

* Reduce cyclomatic complexity

* Update whitelist, gomatrixserverlib, tweaks to roomserver

* Update sytest-whitelist again
2020-03-17 15:12:01 +00:00
Neil Alexander
aebf347a79
Implement gomatrixserverlib.HeaderedEvent in roomserver Kafka output (#914)
* Use Event.Headered

* Use HeaderedEvent in roomserver kafka output

* Fix syncserver-integration-tests

* Update producers to roomserver inputs

* Update gomatrixserverlib

* Update gomatrixserverlib

* Update gomatrixserverlib

* Update gomatrixserverlib

* Update gomatrixserverlib

* Update gomatrixserverlib
2020-03-17 11:01:25 +00:00
Kegsay
9f74a8798e
bugfix: Fix #908 by setting the correct state after the event (#913)
* bugfix: Fix #908 by setting the correct state after the event

Previously, this would only happen if the state already existed
previously!

* Structured logging
2020-03-16 17:51:58 +00:00
Neil Alexander
acb505b717
Implement gomatrixserverlib.HeaderedEvent in roomserver query API (#912)
* Implement gomatrixserverlib.HeaderedEvent, which should allow us to store room version headers along with the event across API boundaries and consumers/producers, and intercept unmarshalling to get the event structure right

* Add federationsender to previous
2020-03-16 17:29:52 +00:00
Neil Alexander
452f393dd7
Update room version descriptors, add error handling (#906)
* Update room version descriptors, add error handling

* Fix database queries

* Drop Get from version package

* Fix database wrapping, add comments for version descriptions

* Don't set default room_version value in SQL
2020-03-16 16:05:29 +00:00
Kegsay
dfd8b93d93
Federation: Add event_auth endpoint (#905)
* Federation: Add event_auth endpoint

Signed-off-by: Sudhanshu Jaiswal <sudsjaiswal@gmail.com>

* Update metric names

Signed-off-by: Sudhanshu Jaiswal <sudsjaiswal@gmail.com>

* Rebase

* Add comment from @APwhitehat

Co-authored-by: Sudhanshu Jaiswal <sudsjaiswal@gmail.com>
2020-03-14 20:08:54 +00:00
Kegan Dougal
7353de7dbb Make P2P use JSServer 2020-03-13 10:15:01 +00:00
Kegsay
8bc5084d8d
p2p: Use JSServer for comms rather than GoJsConn (#888)
* p2p: Use JSServer for comms rather than GoJsConn

This has several benefits:
 - it fixes a bug whereby you could not transmit >4k bytes to/from JS/Go land.
 - it more clearly exposes the interface point between Go and JS: a single
   global function call.
 - it presents a nicer API shape than the previous `net.Conn`.
 - it doesn't needlessly 'stream' data which is already sitting in-memory.

This is currently only active for local CS API traffic, another PR will
add Federation P2P support.

* Typo
2020-03-11 12:18:37 +00:00
Kegsay
d71b72816d
p2p: use separate topics! (#886)
Without it, setting a display name fails and /sync wedges forever.
2020-03-10 18:10:10 +00:00
Kegsay
09dfd4839a
p2p: Use a single ed25519 key for event signing/verifying and p2p peer IDs (#885)
* Pass in the private key seed to NewP2PLocalNode to use the same key for p2p

* Extract the ed25519 key from the server name and use it for event checks

This allows HSes which no longer exist to still have events verified for
them. This means that new members can join rooms that has missing servers.

* Remove no-op ipfs uri which we don't use anymore
2020-03-10 14:13:15 +00:00
Neil Alexander
176f722d53 Update .gitignore 2020-03-10 11:42:40 +00:00
Kegsay
cdc1157785
Improve logging when sending events (#883)
We have some failing sytests on sqlite but it's very difficult to debug
due to lack of useful logging. This adds a log line for when a new event
is sent (incl. logging the event ID) as well as adding a user_id field
for all contextual logs so we know who initiated certain actions.
2020-03-09 14:37:51 +00:00
Kegsay
c31cb02271
bugfix: Fix a race condition when creating guest accounts (#882)
* bugfix: Fix a race condition when creating guest accounts

It was possible to both select the same next numeric ID and then both
attempt to INSERT this into the table. This would cause a UNIQUE violation
which then presented itself as an error in sqlite because it does not
implement `common.IsUniqueConstraintViolationErr`.

The fix here is NOT to implement `common.IsUniqueConstraintViolationErr`
otherwise the 2 users would get the SAME guest account. Instead, all of
these operations should be done inside a transaction. This is what this
PR does.

* Update postgres

* Typo

* Actually use the txn when creating accounts

* bugfix for database is locked on guest reg
2020-03-06 18:00:07 +00:00
Neil Alexander
6a1111c3d4
Try to recursively find auth events (to a point) if they are missing (#881)
* Try to recursively find auth events (to a point) if they are missing

* Remove recursion limit for now and other review fixes

* Simplify error handling for recursion

* Pass room version 1 only to MakeJoin until room version support comes later
2020-03-06 16:58:10 +00:00
Kegan Dougal
87283e9de7 bugfix: fix sytest 155 by actually returning depth+1 and not 0 2020-03-06 14:31:12 +00:00
Kegsay
a97b8eafd4
Add peer-to-peer support into Dendrite via libp2p and fetch (#880)
* Use a fork of pq which supports userCurrent on wasm

* Use sqlite3_js driver when running in JS

* Add cmd/dendritejs to pull in sqlite3_js driver for wasm only

* Update to latest go-sqlite-js version

* Replace prometheus with a stub. sigh

* Hard-code a config and don't use opentracing

* Latest go-sqlite3-js version

* Generate a key for now

* Listen for fetch traffic rather than HTTP

* Latest hacks for js

* libp2p support

* More libp2p

* Fork gjson to allow us to enforce auth checks as before

Previously, all events would come down redacted because the hash
checks would fail. They would fail because sjson.DeleteBytes didn't
remove keys not used for hashing. This didn't work because of a build
tag which included a file which no-oped the index returned.

See https://github.com/tidwall/gjson/issues/157

When it's resolved, let's go back to mainline.

* Use gjson@1.6.0 as it fixes https://github.com/tidwall/gjson/issues/157

* Use latest gomatrixserverlib for sig checks

* Fix a bug which could cause exclude_from_sync to not be set

Caused when sending events over federation.

* Use query variadic to make lookups actually work!

* Latest gomatrixserverlib

* Add notes on getting p2p up and running

Partly so I don't forget myself!

* refactor: Move p2p specific stuff to cmd/dendritejs

This is important or else the normal build of dendrite will fail
because the p2p libraries depend on syscall/js which doesn't work
on normal builds.

Also, clean up main.go to read a bit better.

* Update ho-http-js-libp2p to return errors from RoundTrip

* Add an LRU cache around the key DB

We actually need this for P2P because otherwise we can *segfault*
with things like: "runtime: unexpected return pc for runtime.handleEvent"
where the event is a `syscall/js` event, caused by spamming sql.js
caused by "Checking event signatures for 14 events of room state" which
hammers the key DB repeatedly in quick succession.

Using a cache fixes this, though the underlying cause is probably a bug
in the version of Go I'm on (1.13.7)

* breaking: Add Tracing.Enabled to toggle whether we do opentracing

Defaults to false, which is why this is a breaking change. We need
this flag because WASM builds cannot do opentracing.

* Start adding conditional builds for wasm to handle lib/pq

The general idea here is to have the wasm build have a `NewXXXDatabase`
that doesn't import any postgres package and hence we never import
`lib/pq`, which doesn't work under WASM (undefined `userCurrent`).

* Remove lib/pq for wasm for syncapi

* Add conditional building to remaining storage APIs

* Update build script to set env vars correctly for dendritejs

* sqlite bug fixes

* Docs

* Add a no-op main for dendritejs when not building under wasm

* Use the real prometheus, even for WASM

Instead, the dendrite-sw.js must mock out `process.pid` and
`fs.stat` - which must invoke the callback with an error (e.g `EINVAL`)
in order for it to work:

```
    global.process = {
        pid: 1,
    };
    global.fs.stat = function(path, cb) {
        cb({
            code: "EINVAL",
        });
    }
```

* Linting
2020-03-06 10:23:55 +00:00
Neil Alexander
0cda3c52d0 Don't fail hard if backward topology position fails, just use 1 instead 2020-03-02 18:01:24 +00:00
Neil Alexander
59a1f4b8ed
Remove httputil.LogThenError so that the line numbers are reported properly - make error reporting slightly more useful (#879) 2020-03-02 16:20:44 +00:00
Neil Alexander
72565f2eeb
Fix bug in devices endpoint (#877) 2020-02-28 15:06:16 +00:00
Neil Alexander
6460b3725d
Make sure PDUs and EDUs in transaction don't marshal to null (#876) 2020-02-28 14:54:51 +00:00
Kegan Dougal
420ee15433 Kick CI 2020-02-28 14:46:03 +00:00
Neil Alexander
0352f250b8 Fix dockerfile 2020-02-20 13:54:50 +00:00
Kegsay
5caae6f3a0
sqlite: fixes from sytest (#872)
* bugfix: fix panic on new invite events from sytest

I'm unsure why the previous code didn't work, but it's
clearer, quicker and easier to read the `LastInsertID()` way.
Previously, the code would panic as the SELECT would fail
to find the last inserted row ID.

* sqlite: Fix UNIQUE violations and close more cursors

- Add missing `defer rows.Close()`
- Do not have the state block NID as a PRIMARY KEY else it breaks for blocks
  with >1 state event in them. Instead, rejig the queries so we can still
  have monotonically increasing integers without using AUTOINCREMENT (which
  mandates PRIMARY KEY).

* sqlite: Add missing variadic function

* Use LastInsertId because empirically it works over the SELECT form (though I don't know why that is)

* sqlite: Fix invite table by using the global stream pos rather than one specific to invites

If we don't use the global, clients don't get notified about any invites
because the position is too low.

* linting: shadowing

* sqlite: do not use last rowid, we already know the stream pos!

* sqlite: Fix account data table in syncapi by commiting insert txns!

* sqlite: Fix failing federation invite

Was failing with 'database is locked' due to multiple write txns
being taken out.

* sqlite: Ensure we return exactly the number of events found in the database

Previously we would return exactly the number of *requested* events, which
meant that several zero-initialised events would bubble through the system,
failing at JSON serialisation time.

* sqlite: let's just ignore the problem for now....

* linting
2020-02-20 09:28:03 +00:00
Neil Alexander
3dabf4d4ed
More SQLite (#871)
* SQLite support for appservice

* SQLite support for mediaapi

* Copyright notices

* SQLite for public rooms API (although with some slight differences in behaviour)

* Lazy match aliases, add TODOs
2020-02-14 14:12:33 +00:00
Andrew Morgan
409fec2a48
Add whitelist/blacklist maintenance complainer to buildkite annotations (#870) 2020-02-14 13:07:14 +00:00