Flurry of late patches for 7.10.2

[Alan & Kim Zimmerman]

You may have noticed that I have submitted a lot of patches for the 7.10.2
release at a very
 late stage.

I apologise for this, but it has been triggered by an exciting test
conducted by Matt Pickering
 using the new ghc-exactprint[1] based on the API Annotations in 7.10.1

He ran it in round-trip mode over all the source files in hackage, with the
following results

Total: 62,741 files.
Parse fails: 6,611
CPP: 4,487 (not currently processed, coming soon)

So the total succesfully processed was 51,643 files.

Of these,  3,042 failed to reproduce the source exactly.

The patches coming in are to bring the 3,042 count down even further.

Regards
  Alan

[1]  https://github.com/alanz/ghc-exactprint
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-devs/attachments/20150408/e5613696/attachment.html>