DEVEX-780: fix race condition in php-common-bump workflow

[pdodgen-revparts]

Description

Fix race condition in php-common-bump.yaml that causes workflow failures when main branch moves during execution.

Jira Issue: https://revolutionparts.atlassian.net/browse/DEVEX-780

Background

The upgrade-common workflow in downstream repos (like batch) was consistently failing with errors like:

! [remote rejected] main -> main (cannot lock ref 'refs/heads/main': is at <sha1> but expected <sha2>)

Root cause: The workflow does a simple git push origin main which fails if another commit lands on main between checkout and push (typically ~1 minute window). This happens frequently when:

• Multiple PRs merge to main in quick succession
• Each merge triggers an upgrade-common workflow
• The workflows race to push their common bump commits

Fix: Added fetch + rebase before pushing, with retry logic (up to 5 attempts) to handle concurrent pushes.

Testing Information

• Verified the shell script logic handles:
  • No changes case (exits early without push)
  • Successful push on first attempt
  • Retry on push failure after rebase
  • Rebase conflict handling (abort and retry)
• The fix follows the same retry pattern already used in the workflow for composer install and composer require

[cursor]

PR Summary

Low Risk
CI-only workflow scripting change with limited blast radius; main risk is unexpected rebase/push behavior causing workflow failures or repeated retries.

Overview
Hardens the php-common-bump.yaml workflow against concurrent updates to main.

The commit step now exits early when there are no changes (still emitting the current SHA), and replaces the direct git push with a fetch+rebase loop that retries up to 5 times and aborts/recovers from failed rebases before retrying.

^{Written by Cursor Bugbot for commit 076d0ad. This will update automatically on new commits. Configure here.}

[aspencer]

The amount of logic in this github workflow made me raise my eyebrow. Makes me wonder if there's some sort of way to test these things.

But... I guess it's fine?