[PATCH] Code coverage tool that uses the new coverage mapping format and clang's instrumentation based profiling data

[Alex Lorenz]

I'd like to reply to this point:

  As far as I can tell, this really only provides "demangle", and I'm not
  entirely convinced this is even a good idea. This is *very* brittle and
  pretty awkward to extend, and really difficult to make general enough to
  be valuable. How much is it really gaining us?

I think that the language operation won't be extended beyond demangling, so they really aren't needed. However, GCOV provides support for demangling, so maybe it might make sense to keep the demangling functionality in there? I kept it there now but it probably shouldn't be committed straight away, because as you pointed out it's brittle and will probably break some buildbots.

http://reviews.llvm.org/D4445