cudnn_cudatoolkit_11: remove libcudnn_cnn_infer.so exception

With the update to CuDNN 8.1, we do not get the "maximum file size
exceeded" error anymore when patchelf'ing libcudnn_cnn_infer.so. This
change removes the exception for that library.

pkgs/development/libraries/science/math/cudnn/generic.nix

@@ -38,13 +38,7 @@ stdenv.mkDerivation {
   # See the explanation in addOpenGLRunpath.
   postFixup = ''
     for lib in $out/lib/lib*.so; do
-      # patchelf fails on libcudnn_cnn_infer due to it being too big.
-      # Most programs will still get the RPATH since they link to
-      # other things.
-      # (https://github.com/NixOS/patchelf/issues/222)
-      if [ "$(basename $lib)" != libcudnn_cnn_infer.so ]; then
-        addOpenGLRunpath $lib
-      fi
+      addOpenGLRunpath $lib
     done
   '';