In the following code, it's not safe for the field `C._f` to undergo
type promotion, because a variable with static type `C` might have
type `D` at runtime, in which case `C._f` will get dispatched to
`noSuchMethod`, which is not guaranteed to return a stable result.
class C {
final int? _f;
}
class D implements C {
noSuchMethod(_) => ...;
}
foo(C c) {
if (c._f != null) {
print(c._f + 1); // UNSAFE!
}
}
Therefore, in order to determine which fields are promotable, the
implementations need to analyze enough of the class hierarchy to
figure out which field accesses might get dispatched to
`noSuchMethod`.
Currently, the CFE does this by following its usual algorithm for
generating `noSuchMethod` forwarders before trying to determine which
fields are promotable. The analyzer, on the other hand, doesn't have
an algorithm for generating `noSuchMethod` forwarders (since it
doesn't implement execution semantics); so instead it has its own
logic to figure out when a `noSuchMethod` forwarder is needed for a
field, and disable promotion for that field.
But there's a chicken-and-egg problem in the CFE: the CFE needs to
determine which fields are promotable before doing top-level inference
(since the initializers of top-level fields might make use of field
promotion, affecting their inferred types--see #50522). But it doesn't
decide where `noSuchMethod` forwarders are needed until after
top-level inference (because the same phase that generates
`noSuchMethod` forwarders also generates forwarders that do runtime
covariant type-checking, and so it has to run after all top level
types have been inferred).
To fix the chicken-and-egg problem, I plan to rework the CFE so that
it uses the same algorithm as the analyzer to determine which fields
are promotable. This CL makes a first step towards that goal, by
reworking the analyzer's field promotability algorithm into a form
where it can be shared with the CFE, and moving it to
`package:_fe_analyzer_shared`. Since this required a fairly
substantial rewrite, I went ahead and fixed#52938 in the process.
Fixes#52938.
Change-Id: I9e68f51b3ea9a967f55f15bdc445cc1c0efdabdd
Bug: https://github.com/dart-lang/sdk/issues/52938
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313293
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
The expression evaluator called directly into the constant evaluator,
skipping the setting up of the exhaustiveness cache used for
exhaustiveness checking.
Closes#52905
Change-Id: I367ff73d8e23127811f99dfbf451e66eef452443
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/314060
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
I added the entries to dart.dev/resources/glossary in site-www/#5066.
I was only able to test whether the links worked by putting them in the same place in the diagnostics markdown on site-www, which worked. So, I think the way I did it in generate.dart should work but if anyone knows otherwise please let me know!
There were also a couple small changes from my last CL on the diagnostic messages that I didn't get in before merging, so those are in here too (https://dart-review.googlesource.com/c/sdk/+/309780).
Change-Id: Ie5561fa72c7f99f5c86d2112294edccee41d4544
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313960
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Commit-Queue: Marya Belanger <mbelanger@google.com>
This is a reland of commit 16fcfe7eae
Original change's description:
> [dart:js_interop] Remove Object.toJS and JSNumber.toDart
>
> Modifies JSBoxedDartObject reified type on JS backends and also
> modifies JSBoxedDartObject.toDart now that a proper box is introduced.
> Also uses a JS symbol in JSBoxedDartObject.toDart for a property
> check so that different Dart apps don't accidentally share Dart
> objects. It's now an error to call this function on non-boxed objects.
>
> Fixes a type issue where creating a new object literal with the JS
> foreign function was resulting in dart2js thinking toJSBox would
> always throw. Changing the typeDescription to PlainJavaScriptObject
> instead of =Object fixes that issue.
>
> CoreLibraryReviewExempt: Backend-specific library.
> Change-Id: I5cfb1f32ff4328fafdf9831b0d8da806c39391d9
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/309082
> Reviewed-by: Joshua Litt <joshualitt@google.com>
> Commit-Queue: Srujan Gaddam <srujzs@google.com>
CoreLibraryReviewExempt: Reland.
Change-Id: If6b190f12bdf840b0259c5739f50d9bdcd27fd47
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313600
Reviewed-by: Joshua Litt <joshualitt@google.com>
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Change-Id: Id4b9fb27990dd02f4c1d94faf11c6e25f056a5a7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313221
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Samuel Rawlins <srawlins@google.com>
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
An error was introduced by
https://dart-review.googlesource.com/c/sdk/+/313567
when the --default-suites flag is used along with a
list of tests to deflake. The suites in the list of tests
should be the only suites used as selectors, and should
replace any other input selectors.
Bug: b/290617138
Change-Id: Ic43c1409df621bbaf852276fbaa16ce4961a9549
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/314061
Auto-Submit: William Hesse <whesse@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Commit-Queue: William Hesse <whesse@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Each phase now only has a single interface, instead of several of them. This
reduces the number of classes dramatically and also reduces the number of
objects actually sent over the wire.
It also means fewer things to name and a less polluted namespace.
Change-Id: Ib84b76ac4c0a04abfac5fd5650a228046b1bf1d4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313721
Reviewed-by: Bob Nystrom <rnystrom@google.com>
Commit-Queue: Jake Macdonald <jakemac@google.com>
To be removed after Dart SDK 3.1 is published, and the analyzer
SDK constraints updated.
See
https://dart-review.googlesource.com/c/sdk/+/254903
Change-Id: I2c3321d991cd3e123f08c5a360487362be16f258
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313920
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Samuel Rawlins <srawlins@google.com>
While working on some other changes to flow analysis, I discovered
that this particular behaviour wasn't unit tested.
Change-Id: Ia9b27672c62177ffed80d4143f33c5b764ac7bbe
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313242
Commit-Queue: Paul Berry <paulberry@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
My previous CL changed too much... This test specifically didn't want a valid absolute path on Windows, it wanted one without the drive letter.
Change-Id: If0812517e506f9960bca7ac99dd713be087b9ad9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/314020
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
This change doesn't seem to have a significant impact on most compilation results:
- Golem results show no significant difference in microbenchmarks.
- For a medium and large app tested, while we see a small change in the actual inference results, the generated code is identical before/after this change.
- Timing and memory usage on internal compilations seem comparable before/after this change.
Note: This replaces the need for any notion of "invalid" refines so I will clean up that code in a follow up change.
Change-Id: I2a293eacd944fc17ee2dab97d3d947c042b4038f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313720
Commit-Queue: Nate Biggs <natebiggs@google.com>
Reviewed-by: Mayank Patke <fishythefish@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`, it
is no longer necessary to use `.stmt` to turn an expression into an
expression statement; this now happens automatically. The way this
works under the hood is that both the `Statement` and `Expression`
classes mix in the `ProtoStatement` mixin; constructs that expect
statements are declared with input parameters of type
`ProtoStatement`, and they automatically convert expressions to
statements when necessary.
Also, the functions `checkNotPromoted`, `checkPromoted`,
`checkReachable`, `localFunction` now have a return type of
`Expression` rather than `Statement`. This allows them to be used
either where an expression is exprected or where a statement is
expected, which should give us the ability to write some tests that
are not possible (or very difficult) to write today.
Change-Id: I9f7ad5b15bcf8ccfccafc6985e0163b550c5ad1c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313680
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Paul Berry <paulberry@google.com>
Minor changes to last few 3.0 diagnostic message changes (late follow up). Mostly just a practice CL for me, after setting my sdk environment up for the first time.
Fixes https://github.com/dart-lang/site-www/issues/4740
Change-Id: I8f6871a270089627538928dd95bcbf38a29b74e7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/309780
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Commit-Queue: Marya Belanger <mbelanger@google.com>
This allows changing the in-process tests to use Windows-style paths on macOS for convenience.
Change-Id: I0e85a4f8e831471925b8308ad348f75b6867a53b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313385
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
Add test.py --default-suites that includes all the default suites in
addition to the ones explicitly requested, so the test matrix can run
co19 together with the default suites in one sharded test step.
Bug: b/290617138
Change-Id: I5dd5d1aaf3b1ee38adf88c6e9ee6ec13d97fe1ce
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313567
Reviewed-by: William Hesse <whesse@google.com>
The non-sharded tests takes 10 minutes and each of the sharded tests
takes 14 minutes. It's faster to shard immediately (costing one bot
more) and concurrently run the local tests.
Fix end to end dart2js test that times out when sharded and run outside
a directory called sdk.
Bug: b/290617138
Change-Id: If71f0d301edf565c9f15847098320106ca383635
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/312983
Reviewed-by: William Hesse <whesse@google.com>
Creation of AppJit snapshot will not include the source of [Script]
objects (see `UntaggedScript::snapshot_to`). As a result there's no
point in letting the CFE embed the sources into the kernel if we
only use it to create an AppJit snapshot.
This is also in line with build rules for our dart-sdk, where we do the
kernel compilation separately, see utils/application_snapshot.gni:
```
template("application_snapshot") {
...
# Build the kernel file using the prebuilt VM to speed up the
# debug and simulator builds.
prebuilt_dart_action(target_name + "_dill") {
...
args = [
...
"--no-embed-sources",
...
]
[[[
}
...
}
```
TEST=ci
Change-Id: I4e17e49dc21af6102d62c2278dbd6ebbe387f7e8
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313560
Reviewed-by: Slava Egorov <vegorov@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This new approach uses static tear-offs where possible to avoid allocating a new Closure for each Deferrable.
Note: This doesn't scale to every case but as of today the one case not covered is a singleton and so only 1 closure should be allocated there regardless.
Data from a fairly large app:
Before:
Total => 3.8GB
Closures => 85.6MB
Closure Context => 71.0MB
After:
Total => 3.7GB
Closures => 28.9MB
Closure Context => 49.8MB
Diff:
Closures => 56.7MB
Closure Context => 21.2MB
Change-Id: If233f958df2822708b51c0d14c7439b5d3a5a07b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313340
Reviewed-by: Stephen Adams <sra@google.com>
Commit-Queue: Nate Biggs <natebiggs@google.com>
When tracing closures, we accidentally ignored the fact that returning from a non-async
function allows the value to flow in other ways (e.g. through the completion of a future).
This changes the node tracer to always consider the async marker when looking at values
that flow into 'MemberInformation' information nodes, which are the nodes we use
to represent the returned value of a method.
Fixes#52825
Change-Id: I3322e105dc9612f47a516a17f9465bf1002a9f87
Fixed: 52825
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/312708
Commit-Queue: Sigmund Cherem <sigmund@google.com>
Reviewed-by: Stephen Adams <sra@google.com>
Reviewed-by: Nate Biggs <natebiggs@google.com>
Sync compare URI implementation with linter
Change-Id: I23e14459ac55fef4cec1c0c41341fe6f51b781b6
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313285
Commit-Queue: Samuel Rawlins <srawlins@google.com>
Reviewed-by: Oleh Prypin <oprypin@google.com>
This reverts commit 16fcfe7eae.
Reason for revert: Flutter changes haven't landed to google3 yet.
Original change's description:
> [dart:js_interop] Remove Object.toJS and JSNumber.toDart
>
> Modifies JSBoxedDartObject reified type on JS backends and also
> modifies JSBoxedDartObject.toDart now that a proper box is introduced.
> Also uses a JS symbol in JSBoxedDartObject.toDart for a property
> check so that different Dart apps don't accidentally share Dart
> objects. It's now an error to call this function on non-boxed objects.
>
> Fixes a type issue where creating a new object literal with the JS
> foreign function was resulting in dart2js thinking toJSBox would
> always throw. Changing the typeDescription to PlainJavaScriptObject
> instead of =Object fixes that issue.
>
> CoreLibraryReviewExempt: Backend-specific library.
> Change-Id: I5cfb1f32ff4328fafdf9831b0d8da806c39391d9
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/309082
> Reviewed-by: Joshua Litt <joshualitt@google.com>
> Commit-Queue: Srujan Gaddam <srujzs@google.com>
Change-Id: I469ad04db7b49ffef47a46ccac97e909e4865719
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313580
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: Joshua Litt <joshualitt@google.com>
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Modifies JSBoxedDartObject reified type on JS backends and also
modifies JSBoxedDartObject.toDart now that a proper box is introduced.
Also uses a JS symbol in JSBoxedDartObject.toDart for a property
check so that different Dart apps don't accidentally share Dart
objects. It's now an error to call this function on non-boxed objects.
Fixes a type issue where creating a new object literal with the JS
foreign function was resulting in dart2js thinking toJSBox would
always throw. Changing the typeDescription to PlainJavaScriptObject
instead of =Object fixes that issue.
CoreLibraryReviewExempt: Backend-specific library.
Change-Id: I5cfb1f32ff4328fafdf9831b0d8da806c39391d9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/309082
Reviewed-by: Joshua Litt <joshualitt@google.com>
Commit-Queue: Srujan Gaddam <srujzs@google.com>
A basic bisection script.
Currently only supports substring matching for detecting the error.
This was enough for three use cases today:
* https://github.com/dart-lang/sdk/issues/52910
* https://github.com/dart-lang/sdk/issues/52911
* https://github.com/dart-lang/sdk/issues/52912
Produces a concise output on standard out, and a very detailed log
with all process invocation results in `.dart_tool/bisect_dart`.
Usage: tools/bisect.dart -Dstart=23f41452 -Dend=2c97bd78 -Dtest_command="tools/test.py --build -n dartk-linux-debug-x64 lib_2/isolate/package_resolve_test" -Dfailure_string="Error: The argument type 'String' can't be assigned to the parameter type 'Uri'." -Dsdk_path=/usr/local/google/home/dacoharkes/dart-sdk/sdk/ -Dname=20230712_package_resolve_test
This script starts a bisection in the provided SDK path.
It will write logs to .dart_tool/bisect_dart/.
start : The commit has at the start of the commit range.
end : The commit has at the end of the commit range.
test_command : The invocation of test.py.
This should include `--build`.
This should be within quotes when passed in terminal because of spaces.
failure_string : A string from the failing output.
Regexes are not yet supported.
This should be within quotes when passed in terminal when containing spaces.
sdk_path : The SDK path is optional.
The SDK path defaults to the current working directory.
name : The name is optional.
The name defaults to the current date and the recognized test name.
The name is used for distinguishing logs.
Change-Id: Ib071a5305d4992cf189e35eb3dcc50c83101503e
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313384
Commit-Queue: Daco Harkes <dacoharkes@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This reverts commit fb9d0e6dc3.
Reason for revert: Performance regression https://github.com/dart-lang/sdk/issues/52922
Original change's description:
> Tweaks for mixin inference.
>
> I started initially with the idea to move 'mixinInferenceCallback'
> to `ClassElementImpl`, but then realized that we do it for enums too.
> So, it should stay in `InterfaceElementImpl`.
>
> I left with code that I think slightly modernized, so decided to
> send it for review.
>
> Change-Id: Ib990392dc4985a71ffba1f4080237872d9a65ad2
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/312521
> Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
> Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Change-Id: I33098d8509319e9d6e62d7f5174cae2c05977436
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313440
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Auto-Submit: Samuel Rawlins <srawlins@google.com>
Reviewed-by: Jaime Wren <jwren@google.com>
Commit-Queue: Samuel Rawlins <srawlins@google.com>
Added helpers `_field` and `_localVar` to allow tests to test their final fields and local const variables in the const tests.
The rest of the CL moves away from `_evaluateConstant` helpers (to avoid recomputing constants), to `_topLevelVar` or equivalent helpers which grabs the existing evaluation result.
This CL adds to the goal making all constant tests consistent and to avoid unnecessary const computations in the tests.
Change-Id: I508483714a51e5d060286256657ae460b65787c7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/312889
Reviewed-by: Nate Bosch <nbosch@google.com>
Commit-Queue: Kallen Tu <kallentu@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
The Memory column is right aligned, and we can have large memory values.
`'| Memory | CPU | Elapsed time | ...`
Closes: https://github.com/dart-lang/sdk/issues/52915
Change-Id: I47c836ca85109bfa99e293c5c955c416c0f4067d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313382
Reviewed-by: Martin Kustermann <kustermann@google.com>
Auto-Submit: Daco Harkes <dacoharkes@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
