Comment 26 for bug 1737363

Wayne-

I gave your comment a bit more thought. I think it is probably better to have an explicit call to update the pin cache, even at the expense of an extra lock/unlock cycle. I've adjusted the patch to place the pin update call in SCH_COMPONENT::ResolveAll. This is explicitly called when we need to ensure the component libid is valid, so it is the logical place to update the pin cache as well.

Attached is the corrected patch.