| name | concurrency-and-threading-principles |
| description | Concurrency/parallelism patterns: race prevention, deadlock avoidance, message passing, async/channels/mutexes/worker pools. |
| user-invocable | false |
Concurrency and Threading Principles
1. Race Conditions
Multiple threads + shared data + write + no sync = race.
Prevention: locks/mutexes, immutability, message passing, thread-local storage.
Detection: Go=-race, Rust=Miri, C/C++=TSan, Java=JCStress/FindBugs.
2. Deadlocks
Two+ threads waiting for each other. Four conditions (ALL required): mutual exclusion, hold-and-wait, no preemption, circular wait.
Prevention (break any one): lock ordering (same order always), try_lock with timeout, avoid nested locks, lock-free data structures.
3. Prefer Immutability
Immutable = thread-safe. No sync needed. Share freely. If must change → message passing.
4. Message Passing > Shared Memory
"Share memory by communicating" (Go proverb). Channels/queues instead of shared memory. Reduces lock need. Easier to reason about and test.
5. Graceful Degradation
Timeouts, retries, circuit breakers. Don't crash app on one thread failure. Supervisors for fault tolerance. Backpressure for producer-consumer.
Models by Use Case
- I/O-bound: async/await, event loops, coroutines, green threads
- CPU-bound: OS threads, thread pools, parallel processing
- Actor: Erlang/Akka (message passing, isolated state)
- CSP: Go channels, Rust channels
Testing
- Controlled concurrency (deterministic execution)
- Timeout + resource exhaustion scenarios
- Pool full, queue full scenarios
Related
- Resources @.gemini/skills/resources-and-memory-management/SKILL.md
- Error Handling GEMINI.md § Error Handling Principles
- Testing Strategy GEMINI.md § Testing Strategy
