Handle const constructor invocation in kernel_impact.

Handle const constructor invocation in kernel_impact.

This includes a move toward registering the type together with constructors which is need for this CL and (later) for not registering classes with factories as instantiated.

R=het@google.com

Committed: https://github.com/dart-lang/sdk/commit/9c1ad561161db490c2ae47295e302b79decc4723

[Johnni Winther, 2016-10-05 11:37:28]

Description was changed from

==========
Handle const constructor invocation in kernel_impact.
==========

to

==========
Handle const constructor invocation in kernel_impact.

This includes a move toward registering the type together with constructors
which is need for this CL and (later) for not registering classes with factories
as instantiated.
==========

[Johnni Winther, 2016-10-05 11:37:35]

johnniwinther@google.com changed reviewers:
+ het@google.com

[Johnni Winther, 2016-10-07 08:19:45]

ping

[Harry Terkelsen, 2016-10-07 15:53:21]

lgtm

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/js_bac...
File pkg/compiler/lib/src/js_backend/backend.dart (right):

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/js_bac...
pkg/compiler/lib/src/js_backend/backend.dart:3163:
backend.lookupMapAnalysis.registerInstantiatedType(staticUse.type);
just a heads up: I think we are planning to remove lookup map support since it's
unused

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/resolu...
File pkg/compiler/lib/src/resolution/members.dart (right):

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/resolu...
pkg/compiler/lib/src/resolution/members.dart:3893: ? new
StaticUse.constConstructorInvoke(
this is the output of dartfmt? looks weird

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/univer...
File pkg/compiler/lib/src/universe/use.dart (right):

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/univer...
pkg/compiler/lib/src/universe/use.dart:87: StaticUse.internal(Element element,
StaticUseKind kind, [DartType type = null])
long line

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/univer...
pkg/compiler/lib/src/universe/use.dart:223: return new
StaticUse.internal(element, StaticUseKind.CONSTRUCTOR_INVOKE, type);
long line

https://codereview.chromium.org/2392943003/diff/1/tests/compiler/dart2js/kern...
File tests/compiler/dart2js/kernel/impact_test.dart (right):

https://codereview.chromium.org/2392943003/diff/1/tests/compiler/dart2js/kern...
tests/compiler/dart2js/kernel/impact_test.dart:490: new
StaticUse.typedConstructorInvoke(effectiveTarget, null, effectiveTargetType));
long line

https://codereview.chromium.org/2392943003/diff/1/tests/compiler/dart2js/kern...
tests/compiler/dart2js/kernel/impact_test.dart:490: new
StaticUse.typedConstructorInvoke(effectiveTarget, null, effectiveTargetType));
why do we lose the distinction between const and non-const constructors in
kernel?

[Johnni Winther, 2016-10-10 09:03:17]

Description was changed from

==========
Handle const constructor invocation in kernel_impact.

This includes a move toward registering the type together with constructors
which is need for this CL and (later) for not registering classes with factories
as instantiated.
==========

to

==========
Handle const constructor invocation in kernel_impact.

This includes a move toward registering the type together with constructors
which is need for this CL and (later) for not registering classes with factories
as instantiated.

R=het@google.com

Committed:
https://github.com/dart-lang/sdk/commit/9c1ad561161db490c2ae47295e302b79decc4723
==========

[Johnni Winther, 2016-10-10 09:03:18]

Committed patchset #2 (id:20001) manually as
9c1ad561161db490c2ae47295e302b79decc4723 (presubmit successful).

[Johnni Winther, 2016-10-10 09:58:12]

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/js_bac...
File pkg/compiler/lib/src/js_backend/backend.dart (right):

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/js_bac...
pkg/compiler/lib/src/js_backend/backend.dart:3163:
backend.lookupMapAnalysis.registerInstantiatedType(staticUse.type);
On 2016/10/07 15:53:21, Harry Terkelsen wrote:
> just a heads up: I think we are planning to remove lookup map support since
it's
> unused

Acknowledged.

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/resolu...
File pkg/compiler/lib/src/resolution/members.dart (right):

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/resolu...
pkg/compiler/lib/src/resolution/members.dart:3893: ? new
StaticUse.constConstructorInvoke(
On 2016/10/07 15:53:21, Harry Terkelsen wrote:
> this is the output of dartfmt? looks weird

No. Forgot to dartfmt this CL.

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/univer...
File pkg/compiler/lib/src/universe/use.dart (right):

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/univer...
pkg/compiler/lib/src/universe/use.dart:87: StaticUse.internal(Element element,
StaticUseKind kind, [DartType type = null])
On 2016/10/07 15:53:21, Harry Terkelsen wrote:
> long line

Done.

https://codereview.chromium.org/2392943003/diff/1/pkg/compiler/lib/src/univer...
pkg/compiler/lib/src/universe/use.dart:223: return new
StaticUse.internal(element, StaticUseKind.CONSTRUCTOR_INVOKE, type);
On 2016/10/07 15:53:21, Harry Terkelsen wrote:
> long line

Done.

https://codereview.chromium.org/2392943003/diff/1/tests/compiler/dart2js/kern...
File tests/compiler/dart2js/kernel/impact_test.dart (right):

https://codereview.chromium.org/2392943003/diff/1/tests/compiler/dart2js/kern...
tests/compiler/dart2js/kernel/impact_test.dart:490: new
StaticUse.typedConstructorInvoke(effectiveTarget, null, effectiveTargetType));
On 2016/10/07 15:53:21, Harry Terkelsen wrote:
> long line

Done.

https://codereview.chromium.org/2392943003/diff/1/tests/compiler/dart2js/kern...
tests/compiler/dart2js/kernel/impact_test.dart:490: new
StaticUse.typedConstructorInvoke(effectiveTarget, null, effectiveTargetType));
On 2016/10/07 15:53:21, Harry Terkelsen wrote:
> why do we lose the distinction between const and non-const constructors in
> kernel?

Fixed.