pathfinder: classify cudla as a CTK library

Move cudla into the CTK descriptor block so its packaging classification matches
how it is shipped in toolkit installs and the optional nvidia-cudla wheel.
This keeps the catalog organization consistent with the current understanding of
cudla as a CUDA Toolkit component rather than a third-party add-on.

Made-with: Cursor

Author: rwgk

cuda_pathfinder/cuda/pathfinder/_dynamic_libs/descriptor_catalog.py

@@ -290,6 +290,12 @@ class DescriptorSpec:
         anchor_rel_dirs_windows=("extras/CUPTI/lib64", "bin"),
         ctk_root_canary_anchor_libnames=("cudart",),
     ),
+    DescriptorSpec(
+        name="cudla",
+        packaged_with="ctk",
+        linux_sonames=("libcudla.so.1",),
+        site_packages_linux=("nvidia/cu13/lib",),
+    ),
     # -----------------------------------------------------------------------
     # Third-party / separately packaged libraries
     # -----------------------------------------------------------------------
@@ -324,12 +330,6 @@ class DescriptorSpec:
         site_packages_windows=("nvidia/cu13/bin", "nvidia/cu12/bin"),
         dependencies=("nvrtc",),
     ),
-    DescriptorSpec(
-        name="cudla",
-        packaged_with="other",
-        linux_sonames=("libcudla.so.1",),
-        site_packages_linux=("nvidia/cu13/lib",),
-    ),
     DescriptorSpec(
         name="cudss",
         packaged_with="other",