[TEST] Cover Gluon shared memory descriptor ops

[Jokeren]

Summary

Add CPU-result coverage for Gluon shared-memory descriptor slicing.

This adds a focused end-to-end test that:

• loads a 2D tile into shared memory
• slices the shared-memory descriptor into left/right column halves
• loads both descriptor slices
• checks their sum against a PyTorch CPU reference

The branch also patches shared_memory_descriptor.slice through the Gluon CPU simulation scope. Current Gluon does not mark this method as a builtin, so the generic builtin patcher does not inject the builder-backed semantic automatically.

Validation

• PYTHONPATH=/mnt/keren/triton-viz pytest tests/end_to_end/test_gluon_shared_memory_descriptor_ops.py -q
• PYTHONPATH=/mnt/keren/triton-viz pytest tests/end_to_end/test_gluon_shared_memory_descriptor_ops.py tests/end_to_end/test_gluon_blackwell_tcgen05_multicast_ops.py tests/end_to_end/test_gluon_blackwell_tcgen05_two_cta_ops.py tests/end_to_end/test_gluon_blackwell_tensor_memory_reduction_ops.py tests/end_to_end/test_gluon_blackwell_tcgen05_copy_ops.py tests/end_to_end/test_gluon_blackwell_tcgen05_scaled_ops.py tests/end_to_end/test_gluon_blackwell_tcgen05_ops.py tests/end_to_end/test_gluon_blackwell_tensor_memory_ops.py tests/end_to_end/test_gluon_wgmma_ops.py tests/end_to_end/test_gluon_tma_im2col_ops.py tests/end_to_end/test_gluon_blackwell_tma_ops.py tests/end_to_end/test_gluon_tma_ops.py tests/end_to_end/test_gluon_async_copy_ops.py tests/end_to_end/test_gluon_core_ops.py tests/end_to_end/test_gluon.py tests/end_to_end/test_core.py tests/unit/test_adapters.py -q
• pre-commit run --files triton_viz/core/simulation/gluon.py tests/end_to_end/test_gluon_shared_memory_descriptor_ops.py