2.4.2-1

Add tree algorithms for allreduce to improve performance at scale. Add ncclCommAbort() and ncclCommGetAsyncError() to properly handle network errors and be permit recover. Detect initial CPU affinity and no longer escape it. [ROCm/rccl commit: 1450d42675]
2018-12-13 15:56:12 -08:00
parent 3a7974d4c4
commit ffc55560a1
66 changed files with 3746 additions and 3251 deletions
@@ -25,6 +25,7 @@ extern uint64_t ncclDebugMask;
 extern pthread_mutex_t ncclDebugOutputLock;
 extern FILE *ncclDebugFile;
 extern ncclResult_t getHostName(char* hostname, int maxlen);
+extern ncclResult_t getNvmlDevice(int cudaDev, int *nvmlDev);

 extern void ncclDebugLog(ncclDebugLogLevel level, unsigned long flags, const char *filefunc, int line, const char *fmt, ...);