Merge remote-tracking branch 'nccl/master' into develop

Bu işleme şunda yer alıyor:
Wenkai Du committed 2021-11-11 14:22:12 -08:00
işleme 3a919c1f49
51 files changed
+1774 -1252

No files matched your search

+7 -1
Dosyayı Görüntüle
@@ -52,10 +52,16 @@ ncclResult_t ArgsCheck(struct ncclInfo* info) {
}
if (info->coll == ncclFuncAllGather || info->coll == ncclFuncReduceScatter) info->nBytes *= info->comm->nRanks; // count is per rank
if (info->op < 0 || info->op >= ncclNumOps) {
if (info->op < 0 || ncclMaxRedOp < info->op) {
WARN("%s : invalid reduction operation %d", info->opName, info->op);
return ncclInvalidArgument;
}
int opIx = int(ncclUserRedOpMangle(info->comm, info->op)) - int(ncclNumOps);
if (ncclNumOps <= info->op &&
(info->comm->userRedOpCapacity <= opIx || info->comm->userRedOps[opIx].freeNext != -1)) {
WARN("%s : reduction operation %d unknown to this communicator", info->opName, info->op);
return ncclInvalidArgument;
}
if (info->comm->checkPointers) {
if (info->coll == ncclFuncSendRecv) {
+1 -1
Dosyayı Görüntüle
@@ -60,7 +60,7 @@ ncclResult_t wrap_ibv_symbols(void) {
if (!ibvhandle) {
ibvhandle=dlopen("libibverbs.so.1", RTLD_NOW);
if (!ibvhandle) {
WARN("Failed to open libibverbs.so[.1]");
INFO(NCCL_INIT, "Failed to open libibverbs.so[.1]");
goto teardown;
}
}