GitLab

This just ensures that an unsafe coercion is as localised as possible.
For example, instead of
    UnsafeCo (Int -> t1) (Int -> t2)
use 
    Int -> UnsafeCo t1 t2

The main purpose of this patch is to add a bunch of new rules
to the coercion optimiser.  They are documented in the (revised)
Appendix of the System FC paper.  

Some code has moved about:

- OptCoercion is now a separate module, mainly because it
  now uses tcMatchTy, which is defined in Unify, so OptCoercion
  must live higehr up in the hierarchy

- Functions that manipulate Kinds has moved from 
  Type.lhs to Coercion.lhs.  Reason: the function typeKind
  now needs to call coercionKind.  And in any case, a Kind is
  a flavour of Type, so it builds on top of Type; indeed Coercions
  and Kinds are both flavours of Type.

  This change required fiddling with a number of imports, hence
  the one-line changes to otherwise-unrelated modules

- The representation of CoTyCons in TyCon has changed.   Instead of
  an extensional representation (a kind checker) there is now an
  intensional representation (namely TyCon.CoTyConDesc).  This was
  needed for one of the new coercion optimisations.

* Fix a bug that meant that 
     (right (inst (forall tv.co) ty)) 
  wasn't getting optimised.  This showed up in the
  compiled code for ByteCodeItbls

* Add a substitution to optCoercion, so that it simultaneously
  substitutes and optimises.  Both call sites wanted this, and
  optCoercion itself can use it, so it seems a win all round.

In CSE we were getting lots of apprarently-unequal expressions with
the same hash code.  In fact they were perfectly equal -- but we were
using a cheap-and-cheerful equality tests for CoreExpr that said False
for any lambda expression!

This patch adds a proper equality test for Core, with alpha-renaming.
It's easy to do, and will avoid silly cases of CSE failing to fire.

We should get less of this:
  WARNING: file compiler/simplCore/CSE.lhs line 326
  extendCSEnv: long list, length 18
from a compiler built with -DDEBUG

* Core Lint now does full checking of kinds and coercion terms
  which picks up kind errors in coercions that were previously
  simply not checked for

* Coercion.lhs now provides optCoercion which optimises coercion
  terms.  It implements all of Dimitrios's rules

* The constructors for coercion terms now make no attempt to be
  "smart"; instead we rely solely on the coercion optimiser

* CoercionTyCons in TyCon.lhs always had a "custom" kinding rule
  (the coKindFun field of CoercionTyCon) but its type was not 
  clever enough to do both 
     (a) *figure out the result kind*, assuming the whole thing
         is well-kinded in the first place
     (b) *check* the kinds of everything, failing gracefully if
         they aren't right. 
  We need (b) for the new CoreLint stuff. The field now has type
        CoTyConKindChecker
  which does the job nicely.

In preparation for implementing the PushC rule for coercion-swizzling
in the Simplifier, I had to inmplement the three new decomposition
operators for coercions, which I've called csel1, csel2, and cselR.

     co :: ((s1~t1) => r1) ~ ((s2~t2) => r2)
     ---------------------------------------
              csel1 co :: s1~s2

and similarly csel2, cselR.

On the way I fixed the coercionKind function for types of form
          (s1~t2) => r2
which currently are expressed as a forall type.  

And I refactored quite a bit to help myself understand what is
going on.

* Remove trace from optCoercion

* Use simplCoercion for type arguments in the Simplifier
  (because they might be coercions)

Coercion terms can get big (see Trac #2859 for example), so this
patch puts the infrastructure in place to optimise them:

  * Adds Coercion.optCoercion :: Coercion -> Coercion

  * Calls optCoercion in Simplify.lhs

The optimiser doesn't work right at the moment, so it is 
commented out, but Tom is going to work on it.

- During fianlisation we use to occasionally swivel variable-variable equalities
- Now, normalisation ensures that they are always oriented as appropriate for
  instantation.
- Also fixed #1899 properly; the previous fix fixed a symptom, not the cause.

Modules that need it import it themselves instead.

Previously, cheapEqExpr would always return False if it encountered a cast.
This was bad for two reasons. Firstly, CSE (which uses cheapEqExpr to compare
expressions) never eliminated expressions which contained casts and secondly,
it was inconsistent with exprIsBig. This patch fixes this.

Older GHCs can't parse OPTIONS_GHC.
This also changes the URL referenced for the -w options from
WorkingConventions#Warnings to CodingStyle#Warnings for the compiler
modules.

This patch introduces type checking for type families of which associated
type synonyms are a special case. E.g.

        type family Sum n m

        type instance Sum Zero n = n
        type instance Sum (Succ n) m = Succ (Sum n m)

where

        data Zero       -- empty type
        data Succ n     -- empty type

In addition we support equational constraints of the form:

        ty1 ~ ty2

(where ty1 and ty2 are arbitrary tau types) in any context where
type class constraints are already allowed, e.g.

        data Equals a b where
                Equals :: a ~ b => Equals a b

The above two syntactical extensions are disabled by default. Enable
with the -XTypeFamilies flag.

For further documentation about the patch, see:

        * the master plan
          http://hackage.haskell.org/trac/ghc/wiki/TypeFunctions

        * the user-level documentation
          http://haskell.org/haskellwiki/GHC/Indexed_types

The patch is mostly backwards compatible, except for:

        * Some error messages have been changed slightly.

        * Type checking of GADTs now requires a bit more type declarations:
          not only should the type of a GADT case scrutinee be given, but also
          that of any identifiers used in the branches and the return type.

Please report any unexpected behavior and incomprehensible error message 
for existing code.

Contributors (code and/or ideas):
        Tom Schrijvers
        Manuel Chakravarty
        Simon Peyton-Jones
        Martin Sulzmann 
with special thanks to Roman Leshchinskiy

I've forgotten the precise details already, but this patch
significantly refactors the way newtypes are handled, fixes
the foreign-export problem Trac #736 (which concerned newtypes),
and gets rid of a bogus unsafeCoerce in the foreign export
desugaring.

- Type synonym instances are turned into representation synonym tycons
- They are entered into the pool of family instances (FamInst environments)
  in the same way as data/newtype instances
- Still missing is writing the parent tycon information into ifaces and 
  various well-formedness checks.

At the moment GHC really does very little simplification of coercions.
This patch improves matters, but it's still not great, especially when
you have chains linked together with 'trans'. 

I'm also concerned that I have not yet implemented the 'leftc' and 'rightc'
coercions we added to the paper. 

But at least things are better than they were.  In particular
	g `trans` sym g
now cancels to give the identity.

A type variable has a flag saying whether it is a *type* variable or
a *coercion* variable.  This patch adds assertions to check the flag.

And it adds fixes to places which were Wrong (and hence fired the
assertion)! 

Also removed isCoVar from Coercion, since it's done by Var.isCoVar.

This large commit combines several interrelated changes:

  - IfaceSyn now contains actual Names rather than the special
    IfaceExtName type.  The binary interface file contains
    a symbol table of Names, where each entry is a (package,
    ModuleName, OccName) triple.  Names in the IfaceSyn point
    to entries in the symbol table.

    This reduces the size of interface files, which should
    hopefully improve performance (not measured yet).

    The toIfaceXXX functions now do not need to pass around
    a function from Name -> IfaceExtName, which makes that
    code simpler.

  - Names now do not point directly to their parents, and the
    nameParent operation has gone away.  It turned out to be hard to
    keep this information consistent in practice, and the parent info
    was only valid in some Names.  Instead we made the following
    changes:

    * ImportAvails contains a new field 
          imp_parent :: NameEnv AvailInfo
      which gives the family info for any Name in scope, and
      is used by the renamer when renaming export lists, amongst
      other things.  This info is thrown away after renaming.

    * The mi_ver_fn field of ModIface now maps to
      (OccName,Version) instead of just Version, where the
      OccName is the parent name.  This mapping is used when
      constructing the usage info for dependent modules.
      There may be entries in mi_ver_fn for things that are not in
      scope, whereas imp_parent only deals with in-scope things.

    * The md_exports field of ModDetails now contains
      [AvailInfo] rather than NameSet.  This gives us
      family info for the exported names of a module.

Also:

   - ifaceDeclSubBinders moved to IfaceSyn (seems like the
     right place for it).

   - heavily refactored renaming of import/export lists.

   - Unfortunately external core is now broken, as it relied on
     IfaceSyn.  It requires some attention.

Before the coKindFun could be applied to too many arguments;
now it expects exactly the right number of arguments.  That
makes it easier to write the coKindFuns, and localises the work.