Change-Id: I454df164b2cede7c4ae7d6fdf541e67fdf7090da
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316525
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Kallen Tu <kallentu@google.com>
Commit-Queue: Samuel Rawlins <srawlins@google.com>
This reverts commit 6ef967fb07.
Reason for revert: https://github.com/dart-lang/sdk/issues/53039
Original change's description:
> [vm, reload] Delay enum forwarding until after instance morphing.
>
> Enum forwarding requires evaluating and canonicalizing the new enum instances. If these in turn refer to other instances of classes with shape changes, canonicalization may try to compare instances of the old and new shapes and dereference past the end of an object. By waiting for instance morphing to complete, canonicalization can only see instances with the new shape.
>
> Also, don't attempt to forward Enum.values. When an enum member is added or removed, this has the same merging problem as 40442.
>
> Cf. bad074cc49.
>
> TEST=ci
> Bug: https://github.com/flutter/flutter/issues/129177
> Change-Id: I19d0508059bade8496000eeea257bb0730f11d17
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316041
> Reviewed-by: Ben Konyi <bkonyi@google.com>
> Commit-Queue: Ryan Macnak <rmacnak@google.com>
Bug: https://github.com/flutter/flutter/issues/129177
Change-Id: I8aa09be5d8fd72460ab26294ed94a5075135d48b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316501
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Allow updating weak handles etc to happen in parallel with work-stealing.
Also make LocalBlockWorklist more symmetric with BlockWorklist.
Cf. 4495c2b30a
TEST=ci
Change-Id: Id58fe16be92028b1aa4dd8f097769b4107f2a3f0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316043
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Fixes the Gerrit CL footer.
Denylists some builders which are not available on CQ but only on CI.
Change-Id: Ie0716c52366a8be429c2b6aca030e61769ceca80
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316382
Reviewed-by: William Hesse <whesse@google.com>
Auto-Submit: Daco Harkes <dacoharkes@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
ConstantEvaluationValidator and it's implementation is not used
anywhere in g3 nor in the sdk.
Change-Id: I7490501e2e55ac567a996af658bb09dd7c37c8df
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316020
Commit-Queue: Kallen Tu <kallentu@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
This adds parser support for the optional 'const' keyword in extension
type declaration. The token is passed on through the endPrimaryConstructor
listener call since it is used to make the primary constructor constant.
Change-Id: I518a8e397fb62272002c424e7b69affe6123f4af
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316221
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
This was landed a while ago as https://dart-review.googlesource.com/c/sdk/+/254600. However, this had to be rolled back because at the time this increased memory usage for the emitter phase. This change prevented GC from cleaning up the byte arrays of the codegen serialized files which resulted in an increased max heap.
However, now that we map (rather than copy) files into memory, there is no overhead to keeping a reference to the codegen serialization files.
Thanks to recent improvements in the efficiency of deferred serialization this change seems to improve max memory usage by up to 1.2GB for the emitter phase. Avoiding caching results also helps in this regard, usage of the results is fairly localized.
Change-Id: I4d401a877cb74cbb36b511eb598a759fe300fd5d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316180
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
evaluateConstructorCall now returns a Constant, which means evaluation may be cut early in some places if there's an error. However, some code generators rely on _InstanceCreationEvaluator to complete, and so there's no _major_ changes to the design right now (and some small additions to avoid breaking many tests).
This CL is just the framework for the "return Constant" design and will be the base for reporting better errors in constructor invocations.
These changes already show that we cut some amount of over-reporting in our existing language and unit tests, which is a step in the right direction.
Bug: https://github.com/dart-lang/sdk/issues/47603, https://github.com/dart-lang/sdk/issues/47351, https://github.com/dart-lang/sdk/issues/49389
Change-Id: I5ba7f1282658884c18a32d5e98c7804bbfeac0f6
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/312347
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Samuel Rawlins <srawlins@google.com>
Commit-Queue: Kallen Tu <kallentu@google.com>
We have a fix to remove private declarations (classes, other elements, local variables, etc). No reason to not allow this for public elements in entry-point libraries.
Change-Id: Id4c0ef1b14658d701229ea71d25669982e98d6c1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313291
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Commit-Queue: Samuel Rawlins <srawlins@google.com>
Previously, in the "mini-ast" pseudo-language for shared flow analysis
and type analysis unit testing, there were two separate functions to
create a rest pattern (`...`):
- `listPatternRestElement` for creating a rest pattern for use in a
list pattern
- `mapPatternRestElement` for creating a rest pattern for use in a map
pattern
The latter is not allowed in Dart, but it's supported to allow for
better error recovery.
But having two functions wasn't necessary--the two functions did
exactly the same thing.
This CL simplifies things so that there is just a single function,
`restPattern`. It also renames the underlying representation class
from `RestPatternElement` to `RestPattern`, to match the nomenclature
in the patterns spec document.
Change-Id: Iecfe8c86f49161e0657cdab44f000f47c0e3c212
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315520
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Instead use external pointer block lists so that the same weakling can be pending for both minor and major GC at the same time. When we begin marking through new-space in major GCs, this will need to happen for weaklings in new-space. Once the pending lists are independent, it is easy to handle the reverse case. This also improves the promptness with which old-space weaklings with unreachable new-space targets are clear.
TEST=ci
Change-Id: I46a0a78eeae0210caad48a162c8d64a9af79e749
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/314863
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
Enum forwarding requires evaluating and canonicalizing the new enum instances. If these in turn refer to other instances of classes with shape changes, canonicalization may try to compare instances of the old and new shapes and dereference past the end of an object. By waiting for instance morphing to complete, canonicalization can only see instances with the new shape.
Also, don't attempt to forward Enum.values. When an enum member is added or removed, this has the same merging problem as 40442.
Cf. bad074cc49.
TEST=ci
Bug: https://github.com/flutter/flutter/issues/129177
Change-Id: I19d0508059bade8496000eeea257bb0730f11d17
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316041
Reviewed-by: Ben Konyi <bkonyi@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Tested: CQ
Change-Id: I16210697b47dd85aec8743b457e773b044cab81f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316200
Reviewed-by: William Hesse <whesse@google.com>
Commit-Queue: Alexander Thomas <athom@google.com>
This CL lays the foundation for allowing us to emit instructions without finalizing imports, which will in turn enable more flexibility in code generation. For example, we will no longer need to do a pre-code generation pass to collect imports. In addition, it will pave the way for other optimizations as suggested in the `TODO` at the top of `instructions.dart`.
Perhaps counter-intuitively, even without any optimizations, this doesn't seem to be a measurable regression:
Before:
Completed compilation of dart2wasm-html-engine in 52652ms.
Completed compilation of dart2wasm-html-html in 51827ms.
Completed compilation of dart2wasm-html-ui in 20399ms.
Completed compilation of dart2wasm-canvaskit-canvaskit in 32899ms.
Completed compilation of dart2wasm-canvaskit-ui in 20695ms.
Completed compilation of dart2wasm-skwasm-ui in 20319ms.
After:
Completed compilation of dart2wasm-html-engine in 51476ms.
Completed compilation of dart2wasm-html-html in 48845ms.
Completed compilation of dart2wasm-html-ui in 19676ms.
Completed compilation of dart2wasm-canvaskit-canvaskit in 31933ms.
Completed compilation of dart2wasm-canvaskit-ui in 19733ms.
Completed compilation of dart2wasm-skwasm-ui in 19962ms.
Change-Id: Ib3740f88db56070fc3ccdde484675267e4bf40c1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315580
Reviewed-by: Ömer Ağacan <omersa@google.com>
Commit-Queue: Joshua Litt <joshualitt@google.com>
This change adds the functions `mapLiteral` and `mapEntry` to the
"mini-ast" representation used for unit testing shared flow analysis
and type analysis logic. These can be used to model map literals with
explicit type arguments. Set/map literals without explicit type
arguments are not supported yet.
These functions replace the old `inContextMapEntry` method, which was
an alternative way of testing the type analysis behavior of map
literal entries (and was not yet being used). The new approach is
better in that it allows the test cases to follow the structure of
actual Dart code (map entries inside map literals) rather than just
testing map entries in isolation.
Change-Id: I4bfd31b8fb2865f3d0892da0a47ee0bdacaf9250
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315225
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Paul Berry <paulberry@google.com>
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
Minor changes to find_builders.dart script to use pattern
matching, drop duplicate builders, and drop some type casts.
Change-Id: I463ba8352622dc66678aa71627d84130308da9a3
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315261
Reviewed-by: Daco Harkes <dacoharkes@google.com>
This adds checks for conflicting supertypes and prepares for computation
of extension type members.
Change-Id: If997fe84bac929274a6014b315ecea4f3856eedd
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315441
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
- Introduces a new output subdirectory `gen/utils/ddc` where the
new build targets output files. This is intended to make the
migration to the new assets easier since landing this change
shouldn't immediately break any dependencies.
- Enables building the canary and stable assets at the same time.
- Changes more names in our workflow from "dartdevc" to "ddc".
- Updates the ddc-canary-linux-chrome and
ddc-canary-linux-chrome-unsound configurations to use the
new assets.
The new outputs appear in directories that are more consistently
named, (no more sound vs kernel confusion). They are named in
preparation for the eventual deletion of the unsound targets
without any lingering cruft to be cleaned up in the sound
directories.
The new structure is:
```
gen/utils/ddc/
|- canary
| |- pkg
| |- sdk
|- canary_unsound
| |- pkg
| |- sdk
|- stable
| |- pkg
| |- sdk
|- stable-unsound
|- pkg
|- sdk
```
The 'pkg' and 'sdk' directories all contain outputs with the same
files names, each compiled in the respective build modes
(sound/unsound and stable/canary).
Change-Id: I66822ebf03bba487b6d359a8e0aa818b9e7b6bef
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313081
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Commit-Queue: Nicholas Shahan <nshahan@google.com>
I suspect we ought to just remove many/all of the remaining simpler parser tests. I glanced at a few, and they do not seem specific to analyzer, so their covered features should be covered by parser tests in _fe_analyzer_shared.
https://github.com/dart-lang/sdk/issues/50702
Change-Id: I0e524562f6584145db5e80a154e72ea66090d924
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316042
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Samuel Rawlins <srawlins@google.com>
Improves performance of legacy mode apps in the new type system.
Change-Id: Ia884ea244741df5385271f43baa4125678b995ec
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315983
Commit-Queue: Mark Zhou <markzipan@google.com>
Reviewed-by: Nicholas Shahan <nshahan@google.com>
These are intended to be syntactic and not semantic in meaning, so I think has* is a better name.
Change-Id: Ia2a8a276c1057bde913eb755cc1a7a559d953774
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315960
Commit-Queue: Bob Nystrom <rnystrom@google.com>
Reviewed-by: Bob Nystrom <rnystrom@google.com>
Auto-Submit: Jake Macdonald <jakemac@google.com>
A side effect of DDC's new shared runtime ('dart:_rti') is that previously @notNull-annotated code paths were being extraneously null-checked. This is partly responsible for the slowdowns in some internal mixed-mode apps found in testing.
Change-Id: I3f619cfd23688f3eb4ada0079ad1664d39a7e010
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315640
Reviewed-by: Nicholas Shahan <nshahan@google.com>
Commit-Queue: Mark Zhou <markzipan@google.com>
Found some methods that weren't being referenced anywhere, or they
weren't cleaned up. Removing because it's just dead code.
Change-Id: Id344d2d7cedd87468255a33fea75bfda97d10cac
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316000
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Kallen Tu <kallentu@google.com>
- Run the tests on the existing DDC bots.
- In the future we would like to move these to their own bots so
they can be triggered individually.
Change-Id: I2ab022f19323cdfd20f42ef5c41777547ec428f4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313882
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Reviewed-by: Anna Gringauze <annagrin@google.com>
Commit-Queue: Nicholas Shahan <nshahan@google.com>
Following the refactoring we started in previous CLs, this splits string
and int patch files into (1) patch files that only patch and don't add
implementation classes (2) files for implementation classes (3) helpers.
Changes:
- `string_patch` is split into `string.dart` and `string_patch.dart`.
- VM's `integers_patch.dart` copied as `int_patch.dart` and updated.
This was needed as `integers_patch` requires VM's library paths in
imports.
We needed to copy this file to be able to optimize based on
dart2wasm's implementation classes. However in this CL we copy the
file as-is and update import paths.
- VM's `string_buffer_patch.dart` copied to be able to use the string
implementation classes from the new library.
`string_buffer_create.dart` is merged ino `string_buffer_patch.dart`.
- `getHash` and `setHash` to set object identity moved to
`object_helper.dart`, to be able to use in string implementations.
- Redundant `new` keywords in copied files removed.
One TODO added in `string_buffer_patch.dart` when allocating a new
string to the buffer:
// _parts = _GrowableList.withCapacity(10)..add(str);
// TODO(omersa): Uncomment the line above after moving list
// implementations to a library.
_parts = [str];
We can enable the old code again after moving list implementations to a
library.
Change-Id: Ice5dba40b3ba894797028987d4b42cb0c0f0c230
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315821
Commit-Queue: Ömer Ağacan <omersa@google.com>
Reviewed-by: Joshua Litt <joshualitt@google.com>
Whether the reordered objects fail to fit depends non-deterministically on the parallel worker speeds and the exact contents of new-space that vary with each VM or core library change.
Bug: https://github.com/dart-lang/sdk/issues/52936
Change-Id: I6e287d42b5ca77d2ffafd9a5a053118617d6fe51
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313940
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
When using the "mini-ast" pseudo-language to write unit tests for the
flow analysis and type analysis logic in the `_fe_analyzer_shared`
package, it is no longer necessary to use `.expr` to turn a `Var` into
an `Expression`; this now happens automatically. The way this works
under the hood is that the `Var` and `Expression` classes implement
the `ProtoExpression` interface; constructs that expect expressions
are declared with input parameters of type `ProtoExpression`, and they
automatically convert to `Expression`s when necessary.
Change-Id: I85c493145c3fc41a5296c1807cd63fe1401672db
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315247
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
This removes the substitutions from ClassHierarchyBuilder and moves
the checking of inheritance conflict to the ClassHierarchyNodeBuilder.
Change-Id: I87a943102ab644c3b50acb07defb156e686225d1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/315102
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
