Support for UTD error codes in [Sync]TimelineEvent
#4070
Conversation
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
3 times, most recently
from
455bbd4 to
c737e86
Compare
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #4070 +/- ##
==========================================
- Coverage 84.84% 84.80% -0.04%
==========================================
Files 269 269
Lines 28800 28827 +27
==========================================
+ Hits 24434 24446 +12
- Misses 4366 4381 +15 ☔ View full report in Codecov by Sentry. |
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
8 times, most recently
from
bcd57e8 to
ab28964
Compare
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
2 times, most recently
from
03d4148 to
8fa3bf9
Compare
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
from
f702bc9 to
bad2be9
Compare
richvdh
changed the title
[Sync]TimelineEvent
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
3 times, most recently
from
06eff7f to
d00c1a2
Compare
richvdh
changed the title
[Sync]TimelineEvent[Sync]TimelineEvent
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
3 times, most recently
from
0997324 to
d838ed5
Compare
When `decrypt_sync_room_event` fails to decrypt an event, return the UTD as a `SyncTimelineEvent` instead of an Error.
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
from
d838ed5 to
63b94a3
Compare
| // key; retry in a few. | ||
| wait *= 2; | ||
| } else { | ||
| trace!("Event could not be decrypted, but waiting longer is unlikely to help: {:?}", utd_info.reason); |
There was a problem hiding this comment.
This feels useful on the nse level, should we maybe promote it to at least info?
There was a problem hiding this comment.
Well, the failure will already be logged at warning in the crypto layer, so it's not immediately obvious that logging it again is right.
I'll bump this one to debug as a compromise for now.
| matrix_sdk::deserialized_responses::TimelineEventKind::UnableToDecrypt { | ||
| utd_info, .. | ||
| } => { | ||
| trace!( |
There was a problem hiding this comment.
Again, the actual error will be logged at warning, and in this case it's just replicating the previous behaviour (note the trace!("Encryption sync failed to decrypt the event: {err}"); just below).
I can change this if you feel strongly but it's not obvious to me that we should.
There was a problem hiding this comment.
No no, that's quite okay, I didn't check the lower levels to see if we print it anywhere else.
When `decrypt_room_event` fails to decrypt an event, return the UTD as a `TimelineEvent` instead of an Error.
Make the tests behave the same way as the network code, by returning UTDs as `TimelineEventKind::UnableToDecrypt` instead of `TimelineEventKind::PlainText`.
richvdh
force-pushed
the
rav/utds_in_timeline_event
branch
from
89bd8c1 to
1cc4629
Compare
This PR takes the
TimelineEventKindthat was added in #4082, and adds a third enum variant which is specific to unable-to-decrypt events. We update the two methods which are responsible for decrypting encrypted events that have been received from the network (decrypt_sync_room_eventanddecrypt_room_event), and make them return a[Sync]TimelineEventof the relevant kind instead of a megolm error.Since pretty much everything downstream accesses UTD events via
TimelineEventKind::raw(), we don't have to change much else to keep things working.Part of #4048.