Conversation
previously, the second next could return nil, which would be the reason for panic on line 275: memTx := next.Value.(*mempoolTx) Closes #5408
Codecov Report
@@ Coverage Diff @@
## master #5412 +/- ##
==========================================
+ Coverage 61.24% 61.28% +0.04%
==========================================
Files 259 259
Lines 23437 23438 +1
==========================================
+ Hits 14354 14364 +10
+ Misses 7625 7619 -6
+ Partials 1458 1455 -3
|
erikgrinaker
left a comment
There was a problem hiding this comment.
Makes sense. Have we verified that there aren't any other races here?
This comment has been minimized.
This comment has been minimized.
| // regression test for https://github.com/tendermint/tendermint/issues/5408 | ||
| func TestReactorConcurrency(t *testing.T) { | ||
| config := cfg.TestConfig() | ||
| const N = 2 |
There was a problem hiding this comment.
What does this actually do? Seems like we're only testing two reactors regardless of this setting, in which case I think we can just drop this constant.
There was a problem hiding this comment.
we use it when creating reactors makeAndConnectReactors. the number can be increased to 4 for example. 2 is only required when we want to assert the order of transactions on the receiving side
There was a problem hiding this comment.
Sure, but what purpose would it serve to start more reactors than the two we're testing?
There was a problem hiding this comment.
I can't think of any
previously, the second next could return nil, which would be the reason for panic on line 275: memTx := next.Value.(*mempoolTx) Closes #5408
* mempool: fix nil pointer dereference (tendermint#5412) previously, the second next could return nil, which would be the reason for panic on line 275: memTx := next.Value.(*mempoolTx) Closes tendermint#5408 * docs: specify TM version in go tutorials (tendermint#5427) Closes tendermint#5425 * privval: allow passing options to NewSignerDialerEndpoint (tendermint#5434) Required for tendermint#5291 to set timeouts for remote signers. Co-authored-by: Anton Kaliaev <anton.kalyaev@gmail.com> Co-authored-by: Erik Grinaker <erik@interchain.berlin>
previously, the second next could return nil, which would be the reason
for panic on line 275:
memTx := next.Value.(*mempoolTx)
Closes #5408