Fasm annotation support rebased

[acomodi]

This PR is a rebased version of #32.

Below there is a list of outstanding discussions from #32.

Current issues

• Attribute on wires: AFAIK this still needs to be implemented. One thing that is not entirely clear though is the use case here. Why would we need to add attributes on wires?
• FASM prefix annotation embedded in the module name: IMO the two ways of assigning the FASM prefix (through the FASM attribute or through the model name) are not different in a restriction point of view. It is not clear to me why embedding the FASM prefix within the cell name would be better than just add the FASM_PREFIX attribute on top of the cell.
• FASM prefix in generate statement: Unclear what is the issue here. Generate statements are useful when, for instance, there are multiple FFs with the same structure, but that requires different FASM prefixes.
  An example of this would be the FF for the xc7. Using the generate statement would allow us to have a result such as the one in the example.
• Missing FASM feature mode test: I believe that we have this test already in this PR?

[acomodi]

@mithro FYI