GitLab

No reason user should turn them off, just want for ghc developer
debugging uses.

   (i) Replaced a lot of clunky and fragile EvVar handling code with
       a more uniform ``flavor transformer'' API in the canonicalizer
       and the interaction solver. Now EvVars are just fields inside
       the CtFlavors.
   (ii) Significantly simplified our caching story
This patch does not validate yet and more refactoring is on the way.

A Forest is just a list, and we have a Show instance for
lists already.  We don't need a special Show instance for
Forests.  If we change our minds we can add a newtype,
but I don't think it's worth it.

Fixes Trac #5960

RdrHsSyn.extractGenericPatTyVars was a leftover from the
old generic classes.

Conflicts:
	compiler/hsSyn/Convert.lhs
	compiler/hsSyn/HsDecls.lhs

of an HsType return RdrNames rather than (Located RdrNames).

This means less clutter, and the individual locations are
a bit arbitrary if a name occurs more than once.

This extension is implied by:

  * TypeOperators: so that we can import/export things like (+)
  * TypeFamilies: because associated type synonyms use "type T"
                  to name the associated type in a subordinate list.

Conflicts:
	compiler/main/HscStats.lhs

When we see a type operator in an import or an export, we tag it
with the keyword 'type' so that it is not confused with value level
operators with the same name.

We were comparing ALIGNMENT_DOUBLE to ALIGNMENT_LONG, but really
we cared about W_ values, and sizeof(long) /= sizeof(void *) on Win64