[TEST] Cover Gluon Blackwell two-CTA tcgen05 ops

[Jokeren]

Summary

Add CPU-result coverage for Blackwell two-CTA tcgen05_mma simulation through Gluon.

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

• launches a Gluon kernel with num_ctas=2
• verifies gl.num_ctas() inside the simulated kernel
• loads A/B tiles through TMA into shared memory
• runs blackwell.tcgen05_mma into a two-CTA tensor-memory accumulator layout
• stores the result back through TMA
• checks the output against a PyTorch CPU matmul reference

The branch also threads launch num_ctas through the Gluon CPU builder so gl.num_ctas() reflects the launch metadata during simulation and is restored after each run.

Validation

• PYTHONPATH=/mnt/keren/triton-viz pytest tests/end_to_end/test_gluon_blackwell_tcgen05_two_cta_ops.py -q
• PYTHONPATH=/mnt/keren/triton-viz pytest 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_blackwell_tcgen05_two_cta_ops.py