2D convolution performance improvements

[JuanPedroGHM]

• Combine padding and get_halo for to a have a single communication step
• Create a padded_halod_signal buffer at the start to avoid concatenations
• Possibly delete original signal and return it from the padded_halod_signal at the end of convolution

[github-actions]

This issue is stale because it has been open for 60 days with no activity.