[m-rev.] For review by Zoltan: Fix some auto-parallelism problems.

[Paul Bone]

On Sun, Jan 23, 2011 at 04:09:52PM +1100, Zoltan Somogyi wrote:
> On 23-Jan-2011, Paul Bone <pbone at csse.unimelb.edu.au> wrote:
> > +    maybe_simplify(no, simplify_pass_pre_implicit_parallelism, Verbose, Stats,
> > +        !HLDS, [], SimplifySpecsPreImpPar, !IO),
> > +    expect(unify(contains_errors(Globals, SimplifySpecsPreImpPar), no),
> > +        this_file, "middle_pass: simplify has errors"),
> 
> Delete this expect. While simplify should not find any errors, it may find
> warnings, and those will be in SimplifySpecsPreImpPar as well.
> 
> > @@ -237,8 +244,8 @@ middle_pass(ModuleName, !HLDS, !DumpInfo
> >      % propagation and we cannot do that once the term-size profiling or deep
> >      % profiling transformations have been applied.
> >      maybe_simplify(no, simplify_pass_pre_prof_transforms, Verbose, Stats,
> > -        !HLDS, [], SimplifySpecs, !IO),
> > -    expect(unify(contains_errors(Globals, SimplifySpecs), no), this_file,
> > +        !HLDS, [], SimplifySpecsPreProf, !IO),
> > +    expect(unify(contains_errors(Globals, SimplifySpecsPreProf), no), this_file,
> 
> Same here.
> 

It looks as if the first argument to maybe_simpify controls whether warnings
will be emitted.  Is this correct?

Thanks.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 490 bytes
Desc: Digital signature
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20110123/d56f68ea/attachment.sig>