Zircon mem bug fix

[mbstrange2]

Fixed an issue in the Lake port.py code where a port would wait for the next valid token before committing a word to memory. This is unnecessary because the address stream can be observed independently from the data stream.

[steveri]

Hm I see that the gold test is failing....would you like me to fix that for you? The gold model has to be updated to reflect your new changes, and unfortunately that's not as straightforward as it ought to be. But since it's my code, I think it would probably be relatively easier for me to fix it...

[yuchen-mei]

Hm I see that the gold test is failing....would you like me to fix that for you? The gold model has to be updated to reflect your new changes, and unfortunately that's not as straightforward as it ought to be. But since it's my code, I think it would probably be relatively easier for me to fix it...

Hi Steve, we have discussed and decided to let Max push his fixes to this AHA branch first. Once the Zircon chip is brought up, we will relax the Zircon RTL gold check and merge Max's changes in.

[steveri]

Hm I see that the gold test is failing....would you like me to fix that for you? The gold model has to be updated to reflect your new changes, and unfortunately that's not as straightforward as it ought to be. But since it's my code, I think it would probably be relatively easier for me to fix it...

Hi Steve, we have discussed and decided to let Max push his fixes to this AHA branch first. Once the Zircon chip is brought up, we will relax the Zircon RTL gold check and merge Max's changes in.

Fair enough, that makes sense. So you are freezing the RTL until you are ready to start the next spin on the chip...