Fix KeyError in verbose mode when common workload has unparseable JSON

[Copilot]

common is derived from file presence via find_common_workloads(), but baseline_data/test_data from load_comparison_data() silently drop files that fail to parse. Iterating over common in verbose mode could therefore hit a KeyError for any workload whose JSON exists on disk but is invalid.

Changes

• hwqueue_pipeline.py: Replace for wl_name in common with for wl_name in baseline_data.keys() & test_data.keys() in the verbose summary loop, restricting iteration to workloads successfully loaded in both datasets.

# Before — KeyError if any file in `common` failed to parse
for wl_name in common:
    b = baseline_data[wl_name]
    t = test_data[wl_name]

# After — safe: only workloads present in both loaded dicts
for wl_name in baseline_data.keys() & test_data.keys():
    b = baseline_data[wl_name]
    t = test_data[wl_name]

---

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.