Check result type of redirecting factory in checked mode.

Check result type of redirecting factory in checked mode.
Committed: https://code.google.com/p/dart/source/detail?r=14862

[srdjan, 2012-11-13 18:33:35]

LGTM with one comment

https://codereview.chromium.org/11369219/diff/1/runtime/vm/flow_graph_builder.cc
File runtime/vm/flow_graph_builder.cc (right):

https://codereview.chromium.org/11369219/diff/1/runtime/vm/flow_graph_builder...
runtime/vm/flow_graph_builder.cc:623: }
Why are you not using BuildAssignableValue as below? Maybe just move the
VisitAssignableNode below to EffectGraphVisitor

[regis, 2012-11-13 18:38:53]

Thanks!

https://codereview.chromium.org/11369219/diff/1/runtime/vm/flow_graph_builder.cc
File runtime/vm/flow_graph_builder.cc (right):

https://codereview.chromium.org/11369219/diff/1/runtime/vm/flow_graph_builder...
runtime/vm/flow_graph_builder.cc:623: }
On 2012/11/13 18:33:35, srdjan wrote:
> Why are you not using BuildAssignableValue as below? Maybe just move the
> VisitAssignableNode below to EffectGraphVisitor

I tried, but it does not get simpler. BuildAssertAssignable would need to return
a definition instead of a value, however, this does not work when the test is
skipped, because returning the definition of the input value breaks invariants.
I'm open to suggestions and I can revisit this cl.

[hausner, 2012-11-13 19:06:37]

LGTM

https://codereview.chromium.org/11369219/diff/1/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

https://codereview.chromium.org/11369219/diff/1/runtime/vm/parser.cc#newcode9446
runtime/vm/parser.cc:9446: const String& dst_name =
String::ZoneHandle(Symbols::New("factory result"));
Would it make sense to add this string to the symbol table?

[regis, 2012-11-13 19:29:26]

https://codereview.chromium.org/11369219/diff/1/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

https://codereview.chromium.org/11369219/diff/1/runtime/vm/parser.cc#newcode9446
runtime/vm/parser.cc:9446: const String& dst_name =
String::ZoneHandle(Symbols::New("factory result"));
On 2012/11/13 19:06:37, hausner wrote:
> Would it make sense to add this string to the symbol table?   

I'll do it in a following cl. Thanks.