This repository has been archived by the owner on Feb 1, 2023. It is now read-only.
fix multiple data races #76
Merged
Conversation
This file contains 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
ghost
assigned 
Calling `wg.Add` after `wg.Wait` has returned is invalid. This change swaps the wait group for a plain rwmutex. (caught with the race detector)
Stebalien
force-pushed
the
fix/pubsub-wait-race
branch
from
1972fcd to
13f4ed3
Compare
Stebalien
force-pushed
the
fix/pubsub-wait-race
branch
from
059b761 to
0981b54
Compare
Stebalien
force-pushed
the
fix/pubsub-wait-race
branch
from
0981b54 to
e6a2a40
Compare
notifications/notifications.go
Outdated
| return | ||
| default: | ||
| } | ||
|
|
||
| ps.wrapped.Pub(block, block.Cid().KeyString()) | ||
| } | ||
|
|
||
| // Not safe to call more than once. |
There was a problem hiding this comment.
Pretty sure this comment no longer relevant. I mean not that you should, but looks like it'll just return early if you do
notifications/notifications.go
Outdated
| @@ -107,9 +100,10 @@ func (ps *impl) Subscribe(ctx context.Context, keys ...cid.Cid) <-chan blocks.Bl | |||
| if !ok { | |||
| return | |||
| } | |||
| // We could end up blocking here if the client | |||
There was a problem hiding this comment.
I see that the waitgroup is the source of the race, and I get removing it, but I'm not 100% on why removing the cancel channel is needed, and it does seem like a nice way to insure these go routines get cleaned up, even if the subscribe client doesn't cancel context.
There was a problem hiding this comment.
I mostly just removed it because it wasn't absolutely necessary. We introduced it in the first place to avoid a lock, IIRC.
However, you're right. There's no reason not to have it.
Ensures that we don't leave goroutines behind, even if the client forgets to unsubscribe.
|
@hannahhoward fixed (and I've fixed another test race). |
Jorropo
pushed a commit
to Jorropo/go-libipfs
that referenced
this pull request
Jan 26, 2023
fix multiple data races This commit was moved from ipfs/go-bitswap@722239f
# for free
to subscribe to this conversation on GitHub.
Already have an account?
#.
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.
wg.Addafterwg.Waithas returned is invalid. This change swaps the wait group for a plain rwmutex.(caught with the race detector)