diff --git a/tests/unit/test_fmha_v3_stage_c_full.py b/tests/unit/test_fmha_v3_stage_c_full.py index 3be5e33f..93bd5bbe 100644 --- a/tests/unit/test_fmha_v3_stage_c_full.py +++ b/tests/unit/test_fmha_v3_stage_c_full.py @@ -209,7 +209,7 @@ class FmhaV3StageC: cute.copy(tma_q, tAgQ[(None, Int32(0))], tAsQ[(None, qh.index)], tma_bar_ptr=qh.barrier) qp.tail() kvp.reset(); pk = kvp.try_acquire() - kv_coord = Int32(1) # DEBUG: hardcoded to 1 to test if TMA actually uses it + kv_coord = 0 # Plain Python int, like CUTLASS reference for kt in cutlass.range(n_kv_tiles, unroll=1): kvh = kvp.acquire_and_advance(pk) cute.copy(tma_k, tBgK[(None, kv_coord)], tBsK[(None, kvh.index)], tma_bar_ptr=kvh.barrier)