This is due to bd79be0e19, which will break users on older versions of the SDK.
Bug: #51774
Change-Id: I44f3bd5d0e4bc889d393440b03b4bab33d2ca70a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289620
Reviewed-by: Srujan Gaddam <srujzs@google.com>
Commit-Queue: Joshua Litt <joshualitt@google.com>
I took care to leave code and comments which _does still_ refer to the
remaining Hints. This CL is not super complete, but I think addresses
most of the outdated text refering to Hints. I will do another round
after migrating more to Warnings.
Bug: https://github.com/dart-lang/sdk/issues/50796
Change-Id: Iab58dbbfbdef86e21dd65b2a96d8e34e3e7e54ff
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290440
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Commit-Queue: Samuel Rawlins <srawlins@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
I'm reverting the work queue change for the main inferrer branch. After some instrumenting I've found that it's rather sensitive to the max refinement limit which is very low for the non-experimental branch. In the linearized algorithm the refinement limit is increased so I will keep it in that fork.
Change-Id: I3407ae4d895c9910b6588dea5782ba4b6208ec7a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290420
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
The dart2js snapshot is not meant to be invoked directly except from
a limited set of entry points (CLIs and internal build systems).
This change makes it an error to invoke it in unsupported ways.
Note: this change is not expected to be visible to end users.
Fixes#51695
Change-Id: I4013dd00b90bb3d54483e2f112e0ddfb8dc694e4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289885
Reviewed-by: Nicholas Shahan <nshahan@google.com>
Reviewed-by: Nate Biggs <natebiggs@google.com>
Commit-Queue: Sigmund Cherem <sigmund@google.com>
Fixes https://github.com/dart-lang/sdk/issues/51754
META and WIN_KEY_LEFT have the same value. The switch case for the
second value is never triggered, so we should remove it.
CoreLibraryReviewExempt: No functional change, just silencing warnings.
Change-Id: I51afe53a3d41e2c2fa820c2bb16e1bdea4583fac
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290702
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
CoreLibraryReviewExempt: Minor refactor of web only library.
Change-Id: I7afc4a00501ac12c1aa0248305001a978f2e9f8f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/288641
Reviewed-by: Srujan Gaddam <srujzs@google.com>
Commit-Queue: Joshua Litt <joshualitt@google.com>
This is a reland of commit afdf640866
Original change's description:
> [vm] Refactor StubCodeCompiler to be a real class.
>
> Previously, StubCodeCompiler was just a set of static methods, all of
> which take an assembler as their first arg. This makes it hard to pass
> additional state to the ~160 stub macro defined stub generators.
>
> This refactor makes StubCodeCompiler a real class, with assembler as a
> field. So we can easily add new fields to the class later, to pass new
> state without having to update every stub generator.
>
> assembler is declared as a public field for a few reasons:
> - There's one place where it needs to be accessed by a non-member
> function (in the ia32 file).
> - If it's private, it has to be named assembler_, which would mean a lot
> more insignificant diffs.
> - Non-member functions that take assembler would have to take assembler_,
> for consistency with the __ macro, which would be weird.
>
> Change-Id: I142f0803a07c7839753188065c69c334d4d1798a
> TEST=CI
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289924
> Reviewed-by: Ryan Macnak <rmacnak@google.com>
> Commit-Queue: Liam Appelbe <liama@google.com>
Change-Id: Ib5be28c46a0a80b84e31aea60893ab5bbc02e2ea
TEST=CI, including all sanitizers and architectures
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290681
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Liam Appelbe <liama@google.com>
Today we track and constantly update maps of distinct callers throughout the inference algorithm. Instead we can simply iterate over all call sites and set the `isCalledOnce` flag after all call targets have settled. This result is not used in inference, it is just saved out to the results for the member so the timing doesn't matter.
On a large program this shows a ~4% memory improvement and marginal runtime improvements (in addition to 4% from https://dart-review.git.corp.google.com/c/sdk/+/290060/6).
Change-Id: I403a5cfeba0d5c0175f40626d39ac05860769961
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290120
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
Saves ~4% memory while compiling a large program with no apparent impact to runtime.
Change-Id: Ibc17041cba992b35f0ff6f76d0d422d968576c9f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290060
Reviewed-by: Mayank Patke <fishythefish@google.com>
The .dill and .js compilation steps in the modular suites now use
sound null safety for the files under test.
Change-Id: I264c39d70d160377b4581f910cff1a611819cfbb
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289405
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Commit-Queue: Nicholas Shahan <nshahan@google.com>
Test currently fails on both DDC and dart2js
Bug: 51527
Change-Id: I229d5c9ba518d3323f98312f13740711523439e9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290350
Reviewed-by: Mark Zhou <markzipan@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
The js_runtime implementation of `Error.safeToString` now understands
Records.
`ArgumentError.value((1, 2))` now prints as
Invalid argument: Record (1, 2)
instead of the old way:
Invalid argument: Instance of '_Record_2'
I don't think `Record` is necessary, but this is consistent with the
VM error messages.
Bug: 49718
Change-Id: Icde50f12a3956aa8954ee083307c1e5b13871e41
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290349
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
This code handles simple cases. Still TBD:
- Object patterns that refer to typedefs
- Object patterns that refer to classes with f-bounded polymorphism
- Inference of `dynamic` when there is no bound
Bug: https://github.com/dart-lang/sdk/issues/51795
Change-Id: I00acae6ba111f7b170650cfeffbfd2aaf7f71e42
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290347
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
Our previous visitor was manually implemented. Extending the recursive visitor instead helps us be more robust to future CFE updates.
Fixes#51802
Change-Id: I9efebdf28fdf9146c2d1f948e01363a77807aa16
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290280
Reviewed-by: Nicholas Shahan <nshahan@google.com>
Commit-Queue: Mark Zhou <markzipan@google.com>
None of the classes in dart:developer should be implemented or extended
as most functionality requires native calls into the VM.
TEST=N/A
CoreLibraryReviewExempt: No functional changes to implementations.
Change-Id: I6a8b819e8c9c771954009ffbb2e31f6d0f368c57
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/288823
Reviewed-by: Siva Annamalai <asiva@google.com>
Reviewed-by: Lasse Nielsen <lrn@google.com>
Commit-Queue: Ben Konyi <bkonyi@google.com>
This is a reland of commit b2f4cf3e01
Commented out deprecation for now.
Original change's description:
> Add class modifiers to `dart:convert`.
>
> The usual approach:
> Pure interfaces marked `interface`.
> Pure implementation classes marked `final`.
> Base classes marked `base` or nothing, and `mixin class` if reasonable.
> Combined X/XBase/XMixin where possible.
>
> CoreLibraryReviewExempt: Aske is away
> Change-Id: I927f9bd488fb385ff9c17c8fc94920a1f5076347
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289200
> Reviewed-by: Stephen Adams <sra@google.com>
> Reviewed-by: Slava Egorov <vegorov@google.com>
> Reviewed-by: Nate Bosch <nbosch@google.com>
> Commit-Queue: Lasse Nielsen <lrn@google.com>
CoreLibraryReviewExempt: Approved in original.
Change-Id: I1bc14f99b742567e2634dcfcbc52f332dbcc5364
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290521
Reviewed-by: Nate Bosch <nbosch@google.com>
Commit-Queue: Lasse Nielsen <lrn@google.com>
This reverts commit afdf640866.
Reason for revert: breaks google3. See b/274712076.
Original change's description:
> [vm] Refactor StubCodeCompiler to be a real class.
>
> Previously, StubCodeCompiler was just a set of static methods, all of
> which take an assembler as their first arg. This makes it hard to pass
> additional state to the ~160 stub macro defined stub generators.
>
> This refactor makes StubCodeCompiler a real class, with assembler as a
> field. So we can easily add new fields to the class later, to pass new
> state without having to update every stub generator.
>
> assembler is declared as a public field for a few reasons:
> - There's one place where it needs to be accessed by a non-member
> function (in the ia32 file).
> - If it's private, it has to be named assembler_, which would mean a lot
> more insignificant diffs.
> - Non-member functions that take assembler would have to take assembler_,
> for consistency with the __ macro, which would be weird.
>
> Change-Id: I142f0803a07c7839753188065c69c334d4d1798a
> TEST=CI
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289924
> Reviewed-by: Ryan Macnak <rmacnak@google.com>
> Commit-Queue: Liam Appelbe <liama@google.com>
Change-Id: If36a9122e1a55d86673d05afbbd21dfb27d7acd5
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290522
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: Ivan Inozemtsev <iinozemtsev@google.com>
Reviewed-by: Alexander Thomas <athom@google.com>
Commit-Queue: Emmanuel Pellereau <emmanuelp@google.com>
Fix target_cpu for "arm_x64" to be "arm" not "x64".
TEST=ci
Change-Id: Ifebaee036f3dfa5ba3824f05472423a5bcf9c0c0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290360
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
This CL adds support for strict mode, but disables it in tests and
allowed packages.
Change-Id: I834bc5a0e88868b282c3901361f7a90e1f9eeb3d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289621
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Srujan Gaddam <srujzs@google.com>
Commit-Queue: Joshua Litt <joshualitt@google.com>
Change-Id: Ifff56e334d20f8938779cfdf18861d8ad9d5716c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/289981
Reviewed-by: Ben Konyi <bkonyi@google.com>
Commit-Queue: Ben Konyi <bkonyi@google.com>
Reviewed-by: Kevin Moore <kevmoo@google.com>
Instead of peeking into constant table and then delaying scanning of
constants by putting it in an array, which is walked again in some
future point, we simply read the annotation constants entirely without
requiring const evaluation. This works fine for pragma annotations the
VM is interested in - as there's no user-defined classes involved.
-> Loading kernel will no longer require constant evaluation to work.
Motivation for this is that [0] wants to make this delayed annotation
scanning logic even more complicated, so I prefer to remove it entirely.
[0] https://dart-review.googlesource.com/c/sdk/+/289027
TEST=ci
Change-Id: Ib859480107b6cf119d66035e66ec161ed11ddb32
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/290502
Reviewed-by: Slava Egorov <vegorov@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>