refactor: replace magic layout ints with TensorLayout enum

- Add TensorLayout enum (C++ + Python) to replace magic layout ints
- Add C10_CUDA_CHECK post-launch error checking to all kernel entries
- Add CUDAGuard + freqs_cis shape validation to rotary_emb.cu
- Cache SM count to eliminate per-call cudaDeviceGetAttribute
- Add DISPATCH_CAUSAL_MASK macro to deduplicate dispatcher if/else
- Convert mask type hints from X|None to Optional[X]

csrc/kernels/attn_decode.cu

@@ -16,11 +16,12 @@ torch::Tensor attn_decode(
     TORCH_CHECK(p.head_dim % 32 == 0, "head_dim must be multiple of 32");
 
     auto O = torch::empty_strided(q.sizes(), q.strides(), q.options());
-    auto O_view = (layout == 1) ? O.transpose(1, 2) : O;
+    auto O_view = (layout == BLHD) ? O.transpose(1, 2) : O;
     p.o = (bf16*)O_view.data_ptr();
 
     alloc_split_partials(p);
     DISPATCH_HEAD_DIM(p.head_dim, dispatch_decode, p);
+    C10_CUDA_CHECK(cudaGetLastError());
     return O;
 }
 
@@ -32,6 +33,6 @@ PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) {
         py::arg("mask") = py::none(),
         py::arg("causal_offset") = -1,
         py::arg("scale") = 0.0,
-        py::arg("layout") = 0,
+        py::arg("layout") = (int64_t)BHLD,
         "GQA decode (tensor-core head-packing on sm_80+, scalar fallback)");
 }