Add grafted torch2.9+cu130 build variant

Add grafted torch2.9+cu130 build variant

full build matrix: torch2.9-2.12 x cu126-132, x86_64+aarch64 (fused-dense-matrix-1)

fix __init__: drop comprehension-scoped _name from del

fused_dense_lib kernels build: torch2.9 cu12.8 cp312 sm80/90/100/120 (odyssey-v2.8.3-fused-1)

initial commit