Hi @AbhigyaShridhar! Can you edit the title of this PR to follow the format of other PRs? Please also run pre-commit and update your commit with those changes.

Hi @tim-schilling! Thank you for pointing me to the right process. I've updated my PR.

Update: Added tests for the case where allow_relation() returns None, verifying that Django's default documented behaviour is still valid.

@Credok12 I noticed you've left similar brief reviews on a number of other repos recently. Thanks for taking the time to look at these! Unfortunately, since you're new to the project, a review that says "ok" doesn't provide us with enough context to be actionable. We're not sure what you specifically reviewed, what you consider to be okay, or where your expertise lies. We absolutely welcome you as a reviewer, but to make the most of your contributions, we need more detail about what you've reviewed and any concerns or observations. If you'd like to know what we value in a code review, please review our docs.

Uncovered lines in changed files

-------------
django/contrib/contenttypes/fields.py (100%)
django/db/models/fields/related_descriptors.py (100%)
-------------
Total:   4 lines
Missing: 0 lines
Coverage: 100%
-------------

Note: Missing lines are warnings only. Some database-specific lines may not be measured. More information

Update: The previous missing coverage concerned me since the change is entirely composed of adding and removing conditions. On tracing back, I found it was actually unreachable code. RelatedManager.add() calls check_and_update_obj(), which calls setattr() on the FK field. The FK descriptor's __set__ already calls router.allow_relation(), so a denied relation is caught there before reaching the if statement I had added. For RelatedManager.add(), all that was needed was to remove the redundant _state.db equality check that raised the misleading exception. The GenericRelatedObjectManager.add() change remains as-is since generic relations don't go through the FK descriptor. The test cases remain unchanged.

@tim-schilling @jacobtylerwalls Is there anything else needed on this PR?

Thanks for updates 👍

There are ~100 PRs waiting for a review, so we can't promise any specific timelines. We ask for your patience.