[misc]: pin the identity-order/chunked-launch invariant; fix set_extension docstring

The identity work order (nullptr remap) cannot be offset per chunk, so the
DQ_L2_KEEP chunked launches must only run with a device-computed order.
Today that holds because the L2 transition (256K tokens) lies above the
order-kernel threshold (64K), but nothing tied the two constants together;
a static_assert now does, plus a runtime guard for external callers of the
launch API. Also drops the reference to a tests/jit_ext.py that does not
exist in the repo.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01G4KbxfFobc4FUXLZV3H4QF

fastvideo-kernel/python/fastvideo_kernel/block_sparse_attn_bwd_sm100a.py

@@ -33,8 +33,9 @@ BHSD = True
 def set_extension(module) -> None:
     """Use an already-loaded extension module exposing ``block_sparse_sm100a_bwd``.
 
-    The in-repo tests build the binding standalone (tests/jit_ext.py) and inject it here, so
-    the backend can be exercised without a fastvideo_kernel install.
+    A standalone build of ``block_sparse_bwd_sm100a.cu`` (for example through
+    ``torch.utils.cpp_extension.load`` with a ten-line pybind wrapper) can be injected here, so
+    the backend can be exercised without rebuilding the fastvideo_kernel wheel.
     """
     global _C, _BWD, _HAS_VSA_BWD_SM100A
     _C = module