Asyncify test suite and reintroduce VSS-internal runtime #670
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
.. as we're about to `cargo test` anyways.
.. some of the `ReadFailed` cases didn't log why they failed. Here we fix that oversight.
|
👋 Thanks for assigning @joostjager as a reviewer! |
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
3 times, most recently
from
f817f78 to
141f186
Compare
|
🔔 1st Reminder Hey @joostjager! This PR has been waiting for your review. |
joostjager
reviewed
Oct 20, 2025
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
from
e6e2073 to
584fff6
Compare
tnull
changed the title
|
Now added changes reintroducing a secondary runtime internal to |
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
7 times, most recently
from
6d3a774 to
eedfb12
Compare
joostjager
reviewed
Oct 23, 2025
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
from
eedfb12 to
dba502d
Compare
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
2 times, most recently
from
300e9bb to
4622fb1
Compare
|
CI failure is unrelated: #679 |
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
from
4622fb1 to
e4f3ba0
Compare
|
Addressed pending feedback. @joostjager let me know if I can squash. |
Given we regularly run into issues arising from mixing sync and async contexts, we here simplify our `EventQueue` implementation by avoiding to use `Condvar::wait_while` (which parks the current thread) and rather simply us `block_on` on our `next_event_async` method.
.. as LDK Node is moving towards a more `async` core, it starts to make sense to switch our test suite over to be `async`. This change should make our CI more efficient (as not every node will spawn its independent runtime, but we just have one runtime per test created) and also makes sure we won't run into any edge cases arising from blocking test threads that are executing other async tasks.
Since it seems to make a difference to `tokio` (see https://docs.rs/tokio/latest/tokio/time/fn.timeout.html#panics) we make sure the futures are always put in an `async` closure.
In order to avoid the recently, discovered blocking-task-deadlock (in which the task holding the runtime reactor got blocked and hence stopped polling VSS write tasks), we where re-introduce an internal runtime to the `VssStore`, on which we spawn the tasks, while still using `block_on` of our regular runtime for async-sync conversions. This also finally fixes our VSS CI.
.. we previously avoided running some tests which turned out to be broken.
tnull
force-pushed
the
2025-10-asyncify-more-things
branch
from
e4f3ba0 to
8cad63b
Compare
joostjager
approved these changes
Oct 28, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
As we're looking to move to a more
async-friendly API over time (which would allow us to drop the dependency on theKVStoreSyncimplementation entirely), we here make a few prefactors and minor improvements that move in that direction.Most notably, we switch to have our tests run in async contexts, which also should make them more efficient as not every node will spawn its own runtime, but rather reuse the one runtime spawned for each test case.
Moreover, we reintroduce a secondary runtime internal to
VssStorein order to fix the blocking issue we recently hit in our CI. Additionally, we found that not all tests cfg-gated undervss_testwere being run, so we fix these test cases and have them run in CI.