diff --git a/tests/kernels/moe/test_deepep_moe.py b/tests/kernels/moe/test_deepep_moe.py index ce9e4f343d..8aa1be8423 100644 --- a/tests/kernels/moe/test_deepep_moe.py +++ b/tests/kernels/moe/test_deepep_moe.py @@ -488,10 +488,17 @@ def test_deep_ep_moe( world_dp_size: tuple[int, int], per_act_token_quant: bool, workspace_init, + monkeypatch, ): low_latency_mode = False use_fp8_dispatch = False + if current_platform.is_rocm(): + # The cooperative kernel launches on this path segfault inside the ROCm + # HSA runtime at process exit while rocprofiler-sdk is attached, and + # torch attaches it implicitly. Workers are spawned, so they inherit it. + monkeypatch.setenv("ROCPROFILER_REGISTER_ENABLED", "0") + set_random_seed(7) world_size, dp_size = world_dp_size config = TestConfig(dtype=dtype, topk=topk, m=m, k=k, n=n, num_experts=num_experts)