This was added before null safety, and before we had fully
defined the difference between Null, Bottom, and Never types.
I believe now this and Never.nonNullable are equivalent and
the distinction is no longer important to make. Hence, I'm
deleting it to reduce tech debt.
Originally this was added in https://dart-review.googlesource.com/c/sdk/+/88961
Change-Id: I64b3a8b01833bc60305eaecf430bfb249c8d2c98
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/301820
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Sigmund Cherem <sigmund@google.com>
Our helper function to detect whether a null-assertion was required
didn't include a case to check for record types.
The null-assertion flag is likely to go away soon, since it's
only valuable for mixed applications.
Fixes#52276
Change-Id: Ief4515b8c3eaae5343012da4220e6051bb5a2f19
Fixed: 52276
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/301680
Commit-Queue: Sigmund Cherem <sigmund@google.com>
Reviewed-by: Nate Biggs <natebiggs@google.com>
The modular data saved references to every static invocation node, but only
needed this data for nodes representing one of the `JS*` foreign helpers.
This together with the previous change collectively reduce the total size of
the modular data by an average 8% on a small sample (which includes the entire
SDK).
Change-Id: I3080a1d7ba03643667cb50d7eadcc9fa0a065e6a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/299921
Reviewed-by: Nate Biggs <natebiggs@google.com>
We seem to unnecessarily store all literals when all we need is to know
whether a literal of each kind is in use in the program.
I expect similar changes could help make the modular data more compact and
potentially make enough of a difference where a modular analysis may make
sense.
Change-Id: I045401d742b0467435da179af51c39c3ea08401f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/299960
Reviewed-by: Nate Biggs <natebiggs@google.com>
This reverts commit 4919729f00.
This CL also adds back in logic to handle older package:js versions to avoid
failures in our static checking. It also supports dart:js_interop's @JS
annotation since it can now be used for @staticInterop classes.
CoreLibraryReviewExempt: Reland of backend-specific library changes.
Change-Id: I104653a9a6b2593f6bab658808287e2074c18550
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294130
Reviewed-by: Joshua Litt <joshualitt@google.com>
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Samuel Rawlins <srawlins@google.com>
Type inference is not tracking the type of 'local' correctly through the switch statement into the inner return. The same code without the for loop does work correctly though.
Change-Id: I0b1b4741e4ff17c22dec3383defb412f5fff0836
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/298980
Reviewed-by: Stephen Adams <sra@google.com>
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
Prior to this change the metrics modified here are all printed during every run of the compiler, even when the compiler is running in phase mode on a phase that cannot modify these metrics.
This will make collection of these metrics easier and ensure the collected data is cleaner.
Change-Id: I62a081d9dbae71256882f24156faf246f4076e61
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/297440
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Mayank Patke <fishythefish@google.com>
Currently, dart:_js_types types are all typedefs in the web backends.
This leads to inconsistent semantics, since you can statically pass
Strings to JSString, for example. You cannot do this in dart2wasm.
In order to ensure consistent semantics, we reify these types using
a custom @staticInterop lowering. They all get erased to their
respective Dart type. When we have inline classes, these types
should be implemented using inline classes.
Note that Interceptor will not work for this use case. The reified
type of JS primitives are Dart types e.g. String, bool, and therefore
can not be casted to Interceptor.
In order to do this, the eraser is refactored and the JS backends use
shared erasure code to either erase/emit types.
Tests are added to make sure you need to go through a conversion or
cast to pass Dart objects to JS types.
CoreLibraryReviewExempt: Backend-specific internal library changes.
Change-Id: I5942be628102919ec167f094cfe10fced606363c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/295105
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Reviewed-by: Joshua Litt <joshualitt@google.com>
Reviewed-by: Nicholas Shahan <nshahan@google.com>
The fix is to enable type inference.
The old `List(n)` constructor was replaced by `List.filled`. `List(n)`
could be recognized as fixed-length purely by syntax. The test failed
after the replacement because `List.filled` is not recognized to be
fixed-length purely syntactically in the SSA builder, but rather by
type inference.
With type inference enabled, `List.filled` is recognized, but other
parts of the test need to be more realistic, as type inference also
detects the element type and length.
Bug: #52139
Change-Id: Icfaeb3e341b448bbc263efe43f48a2b76d05bf5b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/298020
Reviewed-by: Nate Biggs <natebiggs@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
Makes more prominent that you can use '-v' to show
extra details about all options.
The message after this change will look as follows:
```
Compile Dart to JavaScript.
Usage: dart compile js [arguments] <dart entry point>
(use -h -v for detailed information about all options)
-h, --help Print this usage information
-o, --output Write the output to <file name>.
-O<0,1,2,3,4> Set the compiler optimization level (defaults to -O1).
```
Fixes https://github.com/dart-lang/sdk/issues/51982
Change-Id: I76fe5478b4a1d4b2ae2170eca72d764bc4d120b4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294164
Reviewed-by: Nate Biggs <natebiggs@google.com>
Reviewed-by: Devon Carew <devoncarew@google.com>
Commit-Queue: Sigmund Cherem <sigmund@google.com>
This CL replaces https://dart-review.googlesource.com/c/sdk/+/296900
The `List` constructor is removed in Dart 3.0.
Some of the `@patch` implementations were not removed.
This is *high priority*. It seems the left-over `@patch factory List` constructor did not cause any errors, instead it *added* a constructor to `List` that can be used in web compiled code. Even if `List` doesn't have such a constructor in the SDK code proper.
The VM and analyzer will say the invocation is an error, but dart2js happily compiles it and runs.
(It used to be that patches couldn't add public members, that security seems to have been removed.)
Also removes code which tries to detect "the unnamed List constructor",
which is no longer a thing, and a number of invocations of the constructor, where it's not clear that the test is aware that the constructor no longer exists, and is not marked as `@dart=2.x` with x < 12.
TEST=ci
Change-Id: I4ffaf3ae2c4e75ca06e7ba0bf19187b6376f3888
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/297100
Reviewed-by: Brian Quinlan <bquinlan@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
What was being logged as "kernel bytes" or "Dart characters" included the size of the platform dill as well. Changed the text for that to the more general "input bytes". The new source bytes number is the number of bytes registered to Dart2JS by the CFE. This registration is done when the CFE is being run modularly or directly from sources.
When being run modularly the input to phase 0 (CFE-only) is slightly larger (includes extra files) but the input to phase 1 (closed world) ends up matching the non-modular value.
Example output:
Compiled 14,927,936 input bytes (8,010,096 characters source) to 11,647,944 kernel bytes in 0.48 seconds using 114.082 MB of memory
Change-Id: Idaaaf6f9c906659434a9af50882b7ed3343e601a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/297020
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Mayank Patke <fishythefish@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
In the old scheme, sync* generators would have an entry and a
body. The entry contained typed checks and the body had injected code
to call a 'factory' helper method to construct the _SyncStarIterable.
It was necessary to pass the element type to the body to be passed to
the factory.
int foo(T a, T b) sync* {yield f(a); yield f(b);}
is compiled to something like
foo(a, b) {
T._as(a); T._as(b);
return foo$body(a, b, type$.int);
}
foo$body(a, b, R) {
return _syncStarFactory(function(){BODY}, R);
}
When type checks were disabled (`-O3`), it was often possible to
generate a single function by merging these as there were no checks.
The new scheme keeps the entry and body separate, and constructs the
_SyncStarIterable in the entry:
foo(a, b) {
T._as(a); T._as(b);
return _syncStarFactory(foo$body(a, b), type$.int);
}
foo$body(a, b) {
return function(){BODY};
}
This keeps the typed Dart 'level' distinct from the untyped JavaScript
level.
The new scheme is a bit more verbose but has the advantage that the
call to `_syncStarFactory` can be inlined and optimized.
Change-Id: I13802d9c9eefd9323841670d059b75a81569d6cb
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/296140
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
Change the calling convention of the sync* body to fix bugs and avoid allocations.
- Body takes the controlling _SyncStarIterator as a parameter.
- `yield` assigns directly to the `_current` field and returns.
- `yield*` calls a method on the controlling iterator (rather than returning the iterable). This avoids an allocation to wrap the iterable, a type test to distinguish `yield` from `yield*`, and allows the `get:iterator` call to happen in the dynamic scope of try-catch surrounding the `yield*` to that it will catch any exceptions.
- Avoid using IIFE to bind constants just to call the body.
- Use a dummy body to avoid the need to testing termination of the body.
Tests now passing:
co19_2/Language/Statements/Yield_and_Yield_Each/Yield_Each/execution_sync_t05
language_2/sync_star/move_past_end_test
language_2/sync_star/sync_star_exception_current_test
language_2/sync_star/sync_star_exception_iterator_test
language_2/sync_star/sync_star_exception_nested_test
language_2/sync_star/sync_star_exception_test
Bug: #51992
Change-Id: I397b470e121b8d71242ac28b3130637b78a1d0dc
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294685
Reviewed-by: Mayank Patke <fishythefish@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
Inference runtime metrics on a large program:
-----Before-----
Duration - Mean: 182.303 Median: 174.500 Max: 292.0
Memory - Mean: 5857.426 Median: 5752.639 Max: 6167.91
-----After-----
Duration - Mean: 144.525 Median: 133.500 Max: 242.0
Memory - Mean: 5867.756 Median: 5727.506 Max: 6262.414
This shows ~20% improvement in runtime and virtually no change in memory usage. Previously we saw a memory regression due to this change. But thanks to improvements in the representation of TypeInformation nodes that memory regression no longer exists.
Code size for this program is also very similar before and after this change.
Before: 43258484 bytes JS
After: 43258444 bytes JS
Change-Id: I6512637f16d83f3127fd1c060bb770ae56405b88
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/295460
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Mayank Patke <fishythefish@google.com>
The main reason is to do this change is to avoid inconsistencies in what is considered to be a `JSArray`. Using `Array.isArray` allows js-interop where the JavaScript code subclasses `Array`.
dart2js-production results:
ArrayLoop.pseudopoly.hoisted1-indexing -50.20%
ArrayLoop.pseudopoly.indexing -35.18%
Iteration.concat.manual 11.69%
ListCopy.List.int.unmodifiable.2 12.03%
ListCopy.for.int.2 12.58%
ListCopy.spread.int.2 13.57%
ListCopy.toList.fixed.100 16.06%
ListCopy.List.of.fixed.100 16.96%
ListCopy.toList.100 17.28%
ListCopy.spread.int.cast.2 18.45%
ObjectHash.hash.5 19.62%
TypedDataPoly.A_UVx5.view.2 21.12%
ObjectHash.manual.5 21.70%
ListCopy.spread.int.map.2 21.96%
ListCopy.spread.int.cast.100 22.12%
TypedDataPoly.A_UVx5.view.100 23.04%
ListCopy.spread.int.map.100 25.26%
ListCopy.List.of.100 27.88%
ImagingGaussianBlurOnce 29.67%
MegaEquality 37.90%
An investigation of the two regressions shows that they benefit from a micro-benchmarking effect. The previous code was monomorphic at the property access in `receiver.constructor == Array`. If this is forced to be polymorphic, the baseline is quite a bit worse, leading to an improvement in line with some of the other benchmarks.
Change-Id: I5c265b1d7408fbd41da9c6fa17472bf648000c8d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/287140
Reviewed-by: Nate Biggs <natebiggs@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
- For parameters of closurized members make sure to mark the virtual parameter as closurized where appropriate. Closurized parameters are treated as dynamic and this needs to propagate to all the virtual target's overrides. Previously only the concrete target was getting marked as closurized.
- For mixins defining an abstract member, foo, implementations of foo (either directly on the mixin target or superclasses of it) should propagate their types to the abstract foo as they are effectively overriding it. Calls to foo within the body of the mixin can only target the abstract foo with a virtual call. So that virtual target needs to reflect the types of all its overrides.
Tests have been added that capture both of these cases. The values here reflect the non-linearized algorithm but prior to these fixes the linearized algorithms showed a diff for both tests. After these changes there is no diff.
This fixes all failures referenced in b/277876666.
Change-Id: Icaca99cb6902c54f481eebdbb65cbd7cf81aa2c2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294960
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Mayank Patke <fishythefish@google.com>
Create a compact version of StaticUse without fields that are often not provided. And intern instances of StaticUse as there are a lot of duplicates.
This should save >150MB in the emitter phase of Dart2JS.
Change-Id: I8d0d0c64bd9e8e3ffc7a86c413cc875be568c5a4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/291620
Reviewed-by: Stephen Adams <sra@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
- Rather than gating reads on URIs we've seen, gate logic on file scheme. Any URI with a 'file' scheme should be readable.
- Use unmapped URI when creating SourceFile. For providers that expand their URIS (e.g. BazelRootProvider) this will make sure we're using the more readable form rather than the prefix substituted version.
Change-Id: I89a00e48a6ed4fc83d403b34c42092fcc42a8b56
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294961
Commit-Queue: Stephen Adams <sra@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Stephen Adams <sra@google.com>
This reverts commit 9aa00f6c75.
Reason for revert: breaks google3 (b/277876666)
Original change's description:
> [dart2js] Patch experimental inferrer (with linearized algorithm) into main inferrer.
>
> After this change for a large program we see:
> - 20% reduction in runtime (254s -> 202s)
> - 2% decrease in memory usage (7671.029MB -> 7524.901MB)
> - Code size decrease of <1% (214,788,150 bytes -> 214,757,445 bytes).
> - The k-limit of 6 refines imposed on the global inference algorithm is no longer hit for converging types. All but 15 types converge, down from 777 types that were hitting the k-limit before. This means we are likely generating better types/code for all 762 of those types.
> - More resilience to dynamic calls with lots of potential targets.
>
> Change-Id: I53d4ade51559f3366f076b6f2b485c5bdc50c6e8
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/271480
> Reviewed-by: Sigmund Cherem <sigmund@google.com>
> Commit-Queue: Nate Biggs <natebiggs@google.com>
Change-Id: Ia6e038e8e077813e589a54d2242bd2820bd22036
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294741
Reviewed-by: Alexander Thomas <athom@google.com>
Commit-Queue: Emmanuel Pellereau <emmanuelp@google.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
`o is T` where `T` is an interface implemented by a class hierarchy
can be implemented as `o instanceof R` where `R` is the root of the
hierarchy. The common case is that `R` is a single implementation
class.
`is Record` and `is Type` now generally use `instanceof`.
Fixes: #51366
Change-Id: I1943533bc53024f3199f71910034d49cf6661b22
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294320
Commit-Queue: Stephen Adams <sra@google.com>
Reviewed-by: Mayank Patke <fishythefish@google.com>
After this change for a large program we see:
- 20% reduction in runtime (254s -> 202s)
- 2% decrease in memory usage (7671.029MB -> 7524.901MB)
- Code size decrease of <1% (214,788,150 bytes -> 214,757,445 bytes).
- The k-limit of 6 refines imposed on the global inference algorithm is no longer hit for converging types. All but 15 types converge, down from 777 types that were hitting the k-limit before. This means we are likely generating better types/code for all 762 of those types.
- More resilience to dynamic calls with lots of potential targets.
Change-Id: I53d4ade51559f3366f076b6f2b485c5bdc50c6e8
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/271480
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
Dart2JS uses the source bytes to generate error messages for the CFE and during closed world construction (for const evaluation). But the actual source code is never used beyond that.
For large applications not including the source code can reduce the size of the dill by close to 50%.
Change-Id: I99d2cb5d1cea843bc12a5e90df3ed651ed8a1ff1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/293762
Reviewed-by: Stephen Adams <sra@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Sigmund Cherem