kanshan/nccl-tests
1
0
Fork 0
mirror of https://github.com/NVIDIA/nccl-tests.git synced 2026-04-23 16:08:20 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Fix Clang compilation errors with VLA initialization

Browse Source 
Code failed to compile with Clang on following:

```
common.cu(1347): error: variable "initGpuMem" may not be initialized
    int64_t initGpuMem[nThreads] = {0};
            ^
common.cu(1348): error: variable "bufferMemory" may not be initialized
    int64_t bufferMemory[nThreads] = {0};
            ^
2 errors detected in the compilation of "common.cu".
```
This commit is contained in:
Marcin Malagowski 2026-02-09 15:10:55 +02:00
parent 9938d5a657
commit d7451ee3f0
1 changed files with 4 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/common.cu
View File

@ -1344,8 +1344,10 @@ testResult_t run() {
#if NCCL_VERSION_CODE >= NCCL_VERSION(2,28,0)
ncclDevComm devComms[nThreads*nGpus];
#endif
int64_t initGpuMem[nThreads] = {0};
int64_t bufferMemory[nThreads] = {0};
int64_t initGpuMem[nThreads];
int64_t bufferMemory[nThreads];
memset(initGpuMem, 0, sizeof(initGpuMem));
memset(bufferMemory, 0, sizeof(bufferMemory));
if (!parallel_init) {
// Capture the memory used by the GPUs before initializing the NCCL communicators
int64_t* initFreeGpuMem = (int64_t*)calloc(nGpus*3, sizeof(int64_t));