test: fix 3 stale tests from the prior pull (registry list + memory_store msg)

test_bench_registry EXPECTED_NAMES: add the 3 milestone benches the pull registered (milestone-1h-ccl, milestone-1h-gemm, milestone-gqa-headline) — 11 benches, alphabetical. test_memory_store::test_shape_mismatch_raises: the pull made same-size reshapes byte-conserving (allowed) and an over-large read raise 'Out-of-bounds read' (not 'Shape mismatch') — match the new message. Full suite now 820 passed / 1 skipped / 0 failed.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

tests/test_bench_registry.py

@@ -12,6 +12,9 @@ EXPECTED_NAMES = [
     "gpt3-qkv",
     "ipcq-allreduce",
     "matmul-composite",
+    "milestone-1h-ccl",
+    "milestone-1h-gemm",
+    "milestone-gqa-headline",
     "qkv-gemm",
     "qkv-gemm-multi-pe",
     "va-offset-verify",

tests/test_memory_store.py

@@ -58,10 +58,12 @@ def test_reshape():
 
 
 def test_shape_mismatch_raises():
+    # A read larger than the stored region is rejected. (Same-size reshapes
+    # are byte-conserving and allowed; (5,5)=100B exceeds the 48B region.)
     store = MemoryStore()
     data = np.arange(12, dtype=np.float32)
     store.write("tcm", 0x0, data)
-    with pytest.raises(ValueError, match="Shape mismatch"):
+    with pytest.raises(ValueError, match="Out-of-bounds read"):
         store.read("tcm", 0x0, shape=(5, 5))