Fix unpickle for source paths missing run_dir (#863)

On the datasource class, we require the use of paths.run_dir to
perform some operations. On older cloud-init version, the
Paths class does not have the run_dir attribute. To fix that,
we are now manually adding that attribute in the Paths
object if doesn't exist in the unpickle operation.

LP: #1899299

1 files changed, 16 insertions, 1 deletions

diff --git a/cloudinit/helpers.py b/cloudinit/helpers.py
index fc5011ec..b8f9d2c3 100644
--- a/cloudinit/helpers.py
+++ b/cloudinit/helpers.py
@@ -20,6 +20,7 @@ from cloudinit.settings import (PER_INSTANCE, PER_ALWAYS, PER_ONCE,
 
 from cloudinit import log as logging
 from cloudinit import type_utils
+from cloudinit import persistence
 from cloudinit import util
 
 LOG = logging.getLogger(__name__)
@@ -317,7 +318,9 @@ class ContentHandlers(object):
         return list(self.registered.items())
 
 
-class Paths(object):
+class Paths(persistence.CloudInitPickleMixin):
+    _ci_pkl_version = 1
+
     def __init__(self, path_cfgs, ds=None):
         self.cfgs = path_cfgs
         # Populate all the initial paths
@@ -354,6 +357,18 @@ class Paths(object):
         # Set when a datasource becomes active
         self.datasource = ds
 
+    def _unpickle(self, ci_pkl_version: int) -> None:
+        """Perform deserialization fixes for Paths."""
+        if not hasattr(self, "run_dir"):
+            # On older versions of cloud-init the Paths class do not
+            # have the run_dir attribute. This is problematic because
+            # when loading the pickle object on newer versions of cloud-init
+            # we will rely on this attribute. To fix that, we are now
+            # manually adding that attribute here.
+            self.run_dir = Paths(
+                path_cfgs=self.cfgs,
+                ds=self.datasource).run_dir
+
     # get_ipath_cur: get the current instance path for an item
     def get_ipath_cur(self, name=None):
         return self._get_path(self.instance_link, name)