From e0cc7de4462db91c27345a2e842d4dfdb36ba3a4 Mon Sep 17 00:00:00 2001 From: Wenkai Du <43822138+wenkaidu@users.noreply.github.com> Date: Fri, 13 Oct 2023 15:24:59 -0700 Subject: [PATCH] NPKit: add xcc_id field (#918) [ROCm/rccl commit: 39812ce757ece7c1cf8042fd1c38265eb15c27b9] --- projects/rccl/src/collectives/device/msccl_kernel_impl.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/projects/rccl/src/collectives/device/msccl_kernel_impl.h b/projects/rccl/src/collectives/device/msccl_kernel_impl.h index 6f50d09ac4..f7787fbc0c 100644 --- a/projects/rccl/src/collectives/device/msccl_kernel_impl.h +++ b/projects/rccl/src/collectives/device/msccl_kernel_impl.h @@ -264,7 +264,7 @@ __device__ __forceinline__ void mscclRunInterpreter( #if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__) asm volatile ("s_getreg_b32 %0, hwreg(HW_REG_XCC_ID)" : "=s" (xcc_id)); #endif - NpKit::CollectGpuEventLDS(NPKIT_EVENT_MSCCL_RUN_ENTRY, mscclShmem.work.sizePerMscclChunk*mscclShmem.work.nChunksPerLoop, 0, NPKIT_GET_GPU_TIMESTAMP()); + NpKit::CollectGpuEventLDS(NPKIT_EVENT_MSCCL_RUN_ENTRY, mscclShmem.work.sizePerMscclChunk*mscclShmem.work.nChunksPerLoop, xcc_id, NPKIT_GET_GPU_TIMESTAMP()); } #endif