Details
Details
- Reviewers
Fabien - Group Reviewers
Restricted Project - Commits
- rABC763d666f73c2: [net processing] Annotate m_recently_announced_{invs|proofs} as guarded by…
With Debug
ninja all check-all
Diff Detail
Diff Detail
- Repository
- rABC Bitcoin ABC
- Lint
Lint Passed - Unit
No Test Coverage - Build Status
Buildable 26685 Build 52938: Build Diff lint-circular-dependencies · build-without-wallet · build-diff · build-debug · build-clang-tidy · build-clang Build 52937: arc lint + arc unit
Event Timeline
| src/net_processing.cpp | ||
|---|---|---|
| 514 ↗ | (On Diff #44613) | We could probably use the m_recently_announced_invs for proofs rather than duplicating |
| 3235 ↗ | (On Diff #44613) | Do you still need cs_main ? |
| 4662 ↗ | (On Diff #44613) | Do you still need cs_main ? |
| 7871 ↗ | (On Diff #44613) | Are you holding NetEventsInterface::g_msgproc_mutex here ? |
| 7980 ↗ | (On Diff #44613) | Same question |
Comment Actions
remove unneeded cs_main lock
| src/net_processing.cpp | ||
|---|---|---|
| 514 ↗ | (On Diff #44613) | There are some differences, I think the TxRelay is not necessarily initialized if a node does not relay transactions whereas the ProofRelay is not necessarily initialized if avalanche is not enabled. |
| 7871 ↗ | (On Diff #44613) | Yes, see line 7510. bool PeerManagerImpl::SendMessages(const Config &config, CNode *pto) {
AssertLockHeld(g_msgproc_mutex);
... |
| 7980 ↗ | (On Diff #44613) | yes, we are still in PeerManagerImpl::SendMessages |