ciflow/periodic-rocm-mi300/192524: [c10d] Scope the symmetric-memory lifecycle review fixes to ROCm
- PyTorch 近 90 天出现 1077 次
- PyTorch 近 90 天第 1064 次发版
- 上一次:同一天稍早 · trunk/20665f569c2701914aa26fc95db15a453c10fa2a: Revert "Add CPython test `test_dataclasses` (#195855)"
发生了什么
A review pass over the previous commit found residual defects in the symmetric-memory registry lifecycle. All of them share one root cause: ncclComm_t addresses are recycled. A process group's predecessor is destroyed before its successor is created, so the allocator routinely hands the successor the predecessor's address, and any bookkeeping that identifies a registration by pointer alone cannot tell the two apart.…
摘要按规则整理自下方来源原文