NCCL 2.4.6-1

Added detection of IBM/Power NVLink bridge device.
    Add NUMA support to PCI distance calculations.
    Added NCCL_IGNORE_CPU_AFFINITY env var.
    Fix memory leaks; GithubIssue#180
    Compiler warning fix; GithubIssue#178
    Replace non-standard variable length arrays. GithubIssue#171
    Fix Tree+Shared Memory crash. GithubPR#185
    Fix LL cleanup hang during long running DL jobs.
    Fix NCCL_RINGS environment variable handling.
    Added extra checks to catch repeat calls to ncclCommDestroy() GithubIssue#191
    Improve bootstrap socket connection reliability at scale.
    Fix hostname hashing issue. GithubIssue#187
    Code cleanup to rename all non device files from *.cu to *.cc


[ROCm/rccl commit: f40ce73e89]
This commit is contained in:
David Addison
2019-03-14 19:39:20 -07:00
parent ab2dd12f3e
commit 17c8317cb1
81 changed files with 892 additions and 692 deletions
+3 -3
View File
@@ -1,5 +1,5 @@
/*************************************************************************
* Copyright (c) 2015-2018, NVIDIA CORPORATION. All rights reserved.
* Copyright (c) 2015-2019, NVIDIA CORPORATION. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
@@ -24,7 +24,7 @@ extern int ncclDebugLevel;
extern uint64_t ncclDebugMask;
extern pthread_mutex_t ncclDebugOutputLock;
extern FILE *ncclDebugFile;
extern ncclResult_t getHostName(char* hostname, int maxlen);
extern ncclResult_t getHostName(char* hostname, int maxlen, const char delim);
extern ncclResult_t getNvmlDevice(int cudaDev, int *nvmlDev);
extern void ncclDebugLog(ncclDebugLogLevel level, unsigned long flags, const char *filefunc, int line, const char *fmt, ...);
@@ -108,7 +108,7 @@ static inline void initDebug() {
break;
case 'h': // %h = hostname
char hostname[1024];
getHostName(hostname, 1024);
getHostName(hostname, 1024, '.');
dfn += snprintf(dfn, PATH_MAX, "%s", hostname);
break;
case 'p': // %p = pid