Kernel mixin transformation desugars mixin applications into normal
classes. Mixed-in type is pulled into interfaces list.
However, dart:mirrors needs to know the original mixed-in type of
a mixin application.
This change solves this problem by propagating a 'isTransformedMixinApplication'
attribute of a class through kernel AST, kernel binary and VM objects
into dart:mirrors implementation.
Fixes: https://github.com/dart-lang/sdk/issues/33240
Change-Id: I98ca69294e1ad445402a5ca91d90c30447aabcb2
Reviewed-on: https://dart-review.googlesource.com/56721
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
Reviewed-by: Kevin Millikin <kmillikin@google.com>
This splits create_archive.py into two steps. One that creates the .tar
and one that creates the .cc for the .tar. This way on Fuchsia, the
second step will take the prebuilt instead of building it.
Change-Id: I7f407f9e70db3135884eeeb61aa5aec3091f1fcc
Reviewed-on: https://dart-review.googlesource.com/56291
Reviewed-by: Ryan Macnak <rmacnak@google.com>
This reverts commit 3e50ea32b5.
Reason for revert: Severe performance regressions on many aot-v2 benchmarks.
Original change's description:
> [vm] Support definition of entry-points via @pragma('vm.extern') annotations.
>
> The `@pragma` annotations are evaluated by the constants transformation and
> visible to TFA and the precompiler, which match on the "options" field of the
> annotation to determine whether to mark the class/procedure as a root.
>
> This required enabling the transformation of annotation constants by default.
>
> # Test Plan
>
> The "vmservice_io.main" entry-point is removed from `main.cc` and annotated with
> `@pragma`. All precompiler tests will crash if "vmservice_io.main" is not
> available at runtime.
>
> Debug/release precompiler bots are visible in "cl-linux" button.
>
> Change-Id: I03c5d6ba7918672ed9905fcaee8dabe675a93a5d
> Reviewed-on: https://dart-review.googlesource.com/56660
> Commit-Queue: Samir Jindel <sjindel@google.com>
> Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
TBR=vegorov@google.com,alexmarkov@google.com,sjindel@google.com
Change-Id: I779c17d003659129a4b3fcf284423104948f60e2
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/56820
Reviewed-by: Samir Jindel <sjindel@google.com>
Commit-Queue: Samir Jindel <sjindel@google.com>
The `@pragma` annotations are evaluated by the constants transformation and
visible to TFA and the precompiler, which match on the "options" field of the
annotation to determine whether to mark the class/procedure as a root.
This required enabling the transformation of annotation constants by default.
# Test Plan
The "vmservice_io.main" entry-point is removed from `main.cc` and annotated with
`@pragma`. All precompiler tests will crash if "vmservice_io.main" is not
available at runtime.
Debug/release precompiler bots are visible in "cl-linux" button.
Change-Id: I03c5d6ba7918672ed9905fcaee8dabe675a93a5d
Reviewed-on: https://dart-review.googlesource.com/56660
Commit-Queue: Samir Jindel <sjindel@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Since the downcast is explicit (using "as") the type will be checked
even in unchecked mode, so technically there is a slight risk of
breakage. But since the callback is so near the creation of the
ButtonElement, I'm not concerned. Also, I'm not concerned about the
performance penalty of the explicit downcast because the cost is only
paid once per user click.
Change-Id: I64fbb442985f6ba6f745f665b36a3321498025e4
Reviewed-on: https://dart-review.googlesource.com/56704
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
For brevity we explicitly downcast the parameter type (using "as")
rather than implicitly downcasting it (by assigning to a fresh
variable). Since this is test code, I'm not concerned about the
performance penalty of the explicit downcast.
Change-Id: I2bf1556ec11db124978c6631df8fcdd5adbc298d
Reviewed-on: https://dart-review.googlesource.com/56701
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
In some cases we need to specify a type because the inferred type
would be too specific. For example, in the code:
final content = [new SampleBufferControlElement(...)];
...
content.addAll([new BRElement(), ...]);
Strong mode would infer a type of `List<SampleBufferControlElement>`
for `content`, making it impossible to add other kinds of HTML
elements to the list later. So we have to specify explicitly that the
list element type is meant to be `HTMLElement`.
In other cases we need to specify a type because the inferred type is
too general. For example, in the code:
Future<List<String>> complete(...) {
var result = [];
...add values of type String to `result`...
return new Future.value(result);
}
Strong mode would infer a type of `List<dynamic>` for `result`, making
it impossible to enclose in a `Future<List<String>>`. So we have to
specify explicitly that the list element type is meant to be `String`.
Change-Id: I1b1c884529de2775898dfff800d250cd10629811
Reviewed-on: https://dart-review.googlesource.com/56700
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
This CL makes the following changes, which should be safe since they
don't affect Dart 1.0 semantics:
- Explicitly declare the types of certain variables as `dynamic` to
avoid strong mode inferring a type that leads to errors.
- Explicitly cast certain expressions to `dynamic` to allow access to
members that are dynamically known to be present.
- Add `covariant` annotation to method parameters that are a subtype
of the corresponding parameter in the base class.
Change-Id: Ibefe4d9591f70ea954373ca4d521bc54179cca9b
Reviewed-on: https://dart-review.googlesource.com/55900
Commit-Queue: Paul Berry <paulberry@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
This CL addresses a code pattern where a method expects its parameter
to have a certain type, but that method is torn off and passed as a
callback to another method expecting its parameter type to be more
general. For example:
void f(int i) { ... }
void g(void callback(Object o)) { ... }
void h() {
g(f); // Error: () -> int is not a subtype of () -> Object
}
This is a strong mode error because the type system cannot guarantee
that the value pased to f will be an int. The solution is to broaden
the type of the callback parameter so that it matches the type
expected for the callback. In most cases, we insert an implicit
downcast (by reassigning the parameter to a local variable with the
expected type), which in Dart 2.0 semantics will result in a runtime
check (similar to what happens in Dart 1.0 checked mode).
Since the downcasts are implicit, the Dart 1.0 semantics are
unchanged, so this should be a safe change.
Change-Id: I9583ea194343b89b39305c9796cfad299a47943f
Reviewed-on: https://dart-review.googlesource.com/55907
Commit-Queue: Paul Berry <paulberry@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
This reverts commit ad4cfa0260.
Reason for revert: Assertion failures on DBC
Original change's description:
> [mirrors] Add IsolateMirror.loadUri.
>
> This allows a programmer to dynamically load code into an isolate. The closest existing API is Isolate.spawnUri, but communication with the dynamically loaded code in that case is limited to asynchronous message passing of JSON-like objects.
>
> Change-Id: Icb23e9dacfb0035622c119f11d4e0f892ba2ccd1
> Reviewed-on: https://dart-review.googlesource.com/45363
> Reviewed-by: Zach Anderson <zra@google.com>
TBR=rmacnak@google.com,zra@google.com,asiva@google.com,kmillikin@google.com
Change-Id: I80669188b9f40b3b527e8e268ade0d0d514a8753
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/56640
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
This allows a programmer to dynamically load code into an isolate. The closest existing API is Isolate.spawnUri, but communication with the dynamically loaded code in that case is limited to asynchronous message passing of JSON-like objects.
Change-Id: Icb23e9dacfb0035622c119f11d4e0f892ba2ccd1
Reviewed-on: https://dart-review.googlesource.com/45363
Reviewed-by: Zach Anderson <zra@google.com>
Summary:
When evaluating constants for direct inspection (and not for injection into the
FG), we store many intermediate values in ZoneHandles. This leaks memory, since
the only need for ZoneHandles is preserving constants which are referenced by
the FG.
Test Plan:
Existing coverage is sufficient since no functionality is changed.
Change-Id: I33951eab762a4ad626f07ebf11118247c5a9ad3e
Reviewed-on: https://dart-review.googlesource.com/56345
Commit-Queue: Samir Jindel <sjindel@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
# Summary
As part of commit 772c9bb5f, we changed the representation of partially
instantiated closures to facilitate partial instantiation of local functions.
The compile-time constant evaluator was not updated -- this revision implements
that update.
Fixes task 33211.
# Test Plan
Updated `partial_tearoff_instantiation_test.dart` to test the case which caused
incorrect behavior in task 33211.
Change-Id: I7bbd4fb83a5aea86ffc85c1d9952f1202f098d6b
Reviewed-on: https://dart-review.googlesource.com/56346
Reviewed-by: Régis Crelier <regis@google.com>
Commit-Queue: Samir Jindel <sjindel@google.com>
Given IL like this:
t3 <- Constant(#null) <--\
// ... do something ... |
StoreLocal(:temp, t4) ---/
// Do something else
StoreIndexed(t1, t2, t3)
Graph builder would use Constant(#null) as an input definition for
StoreIndexed - which leads to wrong generated code in unoptimized mode
on X64 where StoreIndexed has specialization for constant values. In
optimized code SSA renaming would actually replace t3 with a correct
reaching definition.
This CL introduces a new IL instruction MakeTemp(...) which has exactly
the same effect as Constant(#null) on the expression stack, but is
not treated as a constant value by the backend.
The bug was unnoticed because of the loose assertions in the SSA
renaming which permitted replacing Constant-s with non-Constant
reaching definitions. This CL also tightens those assertions to
catch this issue.
Also cleanup the code in SSA renaming a bit to make it a bit easier
to understand.
Fixes https://github.com/dart-lang/sdk/issues/33195
Bug: 33195
Change-Id: I7c914c836d4af7a50e8a8e1a02d03e9413f87779
Reviewed-on: https://dart-review.googlesource.com/56112
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
With this change debugger requests yield(await) positions from kernel builder so it can confirm which frame handles what exceptions.
This also renames ':completer' to ':async_completer' in constructed kernel, so it is compliant with what VM expects to see(and updates test expectations accordingly).
Further it fixes async stack navigation for sync-async that became default in dart vm since first version went for the review((uses `future` getter, rather than property).
It also makes `future` getter non-debuggable to allow stepping-out of async methods.
Make Handle zone-scoped. Clean up frame counter var name.
Bug: https://github.com/dart-lang/sdk/issues/29056
Change-Id: Ia71f3c851a6b313655b57dad28c296f5dd081eda
Reviewed-on: https://dart-review.googlesource.com/54640
Commit-Queue: Alexander Aprelev <aam@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
This ensures that the resulting Iterable will have the proper type
when run with strong mode semantics.
In Dart 1.0 semantics, the type argument to .map is ignored, so this
change should be safe.
Change-Id: I34924bc7e803b1520af7d174cfe80f5be6b7ac0f
Reviewed-on: https://dart-review.googlesource.com/55908
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
This is necessary because many call sites invoke toUserString without
first casting to one of the derived classes.
This should be safe since both derived classes (SourceLocation and
UnresolvedSourceLocation) contain a definition of toUserString.
Change-Id: I69eeda326883d87fc4e8cedd732881d250cae8cb
Reviewed-on: https://dart-review.googlesource.com/55909
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
This CL fixes some cases where Observatory mistakenly tried to use a
C-style parameter declaration, e.g.
void foo(String);
The intent is to declare the parameter to have type `String`, but Dart
interprets the parameter as having *name* `String` and type `dynamic`.
In most cases the fix applies to abstract method declarations, whose
types aren't semantically relevant in Dart 1.0, so these parts of the
CL should be safe.
In one case it applies to a typedef, IsConnectedVMTargetDelegate. The
only use of this typedef is as the type of
TargetRepository._isConnectedVMTarget, whose value is always the
static function ObservatoryApplication.isConnectedVMTarget. The type
of ObservatoryApplication.isConnectedVMTarget matches the typedef
(according to Dart 1.0 rules) both before and after the change, so
this should be safe as well.
Change-Id: Ia67f5f14d5a012ccb8a44cbd88a8233f8c9ad171
Reviewed-on: https://dart-review.googlesource.com/55901
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
This CL addresses a code pattern where a method has a more general
return type than it needs, causing a strong mode error at the site
where the method is called or torn off. For example:
Object f() => 'x';
void g(String callback()) { ... }
void h() {
g(f); // Error: () -> Object is not a subtype of () -> String
}
The solution is to tighen up the return type. In all cases in this
CL, it is obvious from the method definition that the value returned
at runtime will satisfy the tighter return type, so this should be a
safe change.
Change-Id: Ifbd476d969274342653d2619c1c81e2e014817c4
Reviewed-on: https://dart-review.googlesource.com/55904
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
Alexander Markov