fix(test): require concurrent_managed_access for advise tests that hit real hardware

These three tests call cuMemAdvise on real CUDA devices and verify
memory range attributes. On devices without concurrent_managed_access
(e.g. Windows/WDDM), set_read_mostly silently no-ops and
set_preferred_location fails with CUDA_ERROR_INVALID_DEVICE. Use the
stricter _skip_if_managed_location_ops_unsupported guard, matching the
pattern already used by test_managed_memory_functions_accept_raw_pointer_ranges.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: rparolin

cuda_core/tests/test_memory.py

@@ -1207,7 +1207,7 @@ def test_managed_memory_prefetch_supports_managed_pool_allocations(init_cuda):
 
 def test_managed_memory_advise_supports_external_managed_allocations(init_cuda):
     device = Device()
-    _skip_if_managed_allocation_unsupported(device)
+    _skip_if_managed_location_ops_unsupported(device)
     device.set_current()
 
     buffer = DummyUnifiedMemoryResource(device).allocate(_MANAGED_TEST_ALLOCATION_SIZE)
@@ -1390,7 +1390,7 @@ def test_managed_memory_operation_validation(init_cuda):
 def test_managed_memory_advise_location_validation(init_cuda):
     """Verify doc-specified location constraints for each advice kind."""
     device = Device()
-    _skip_if_managed_allocation_unsupported(device)
+    _skip_if_managed_location_ops_unsupported(device)
     device.set_current()
 
     buffer = DummyUnifiedMemoryResource(device).allocate(_MANAGED_TEST_ALLOCATION_SIZE)
@@ -1422,7 +1422,7 @@ def test_managed_memory_advise_location_validation(init_cuda):
 def test_managed_memory_advise_accepts_enum_value(init_cuda):
     """advise() accepts CUmem_advise enum values directly, not just string aliases."""
     device = Device()
-    _skip_if_managed_allocation_unsupported(device)
+    _skip_if_managed_location_ops_unsupported(device)
     device.set_current()
 
     buffer = DummyUnifiedMemoryResource(device).allocate(_MANAGED_TEST_ALLOCATION_SIZE)