`Dart_NewWeakPersistentHandle` and `Dart_NewFinalizableHandle` in
`dart_api.h` do no longer accept `Pointer`s and subtypes of `Struct`s
as values passed in to the `object` parameter.
Expandos no longer accept `Pointer`s and subtypes of `Struct`s
as values passed in to the `object` parameter.
Cleans up unused object_store->ffi_struct_class.
Reland 1: Allow importing `dart:ffi` from `dart:core` with
`--enable-ffi=false`.
Reland 2: Allow the new `_Compound` class to extend `NativeType`. (This
got triggered in this CL on the build because of the import
from `dart:core`.)
Closes: https://github.com/dart-lang/sdk/issues/45071
Breaking change: https://github.com/dart-lang/sdk/issues/45072
TEST=vm/cc/DartAPI_FinalizableHandleErrors
TEST=vm/cc/DartAPI_WeakPersistentHandleErrors
TEST=tests/ffi(_2)/expando_test.dart
Change-Id: I133278e16bd75cd2bb6234e7ddf042ffa0a54fd6
Cq-Include-Trybots: luci.dart.try:dart-sdk-linux-try,dart-sdk-mac-try,dart-sdk-win-try,vm-ffi-android-debug-arm-try,vm-ffi-android-debug-arm64-try,vm-kernel-asan-linux-release-x64-try,vm-kernel-mac-debug-x64-try,vm-kernel-linux-debug-ia32-try,vm-kernel-linux-debug-x64-try,vm-kernel-nnbd-linux-debug-x64-try,vm-kernel-nnbd-linux-debug-ia32-try,vm-kernel-nnbd-mac-release-x64-try,vm-kernel-nnbd-win-debug-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-debug-simarm_x64-try,vm-kernel-precomp-nnbd-linux-debug-x64-try,vm-kernel-precomp-win-release-x64-try,vm-kernel-reload-linux-debug-x64-try,vm-kernel-reload-rollback-linux-debug-x64-try,vm-kernel-win-debug-x64-try,vm-kernel-win-debug-ia32-try,vm-precomp-ffi-qemu-linux-release-arm-try,vm-kernel-precomp-obfuscate-linux-release-x64-try,vm-kernel-msan-linux-release-x64-try,vm-kernel-precomp-msan-linux-release-x64-try,vm-kernel-precomp-android-release-arm_x64-try,analyzer-analysis-server-linux-try,vm-kernel-precomp-linux-debug-x64c-try,vm-kernel-linux-debug-x64c-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/195079
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
`Dart_NewWeakPersistentHandle` and `Dart_NewFinalizableHandle` in
`dart_api.h` do no longer accept `Pointer`s and subtypes of `Struct`s
as values passed in to the `object` parameter.
Expandos no longer accept `Pointer`s and subtypes of `Struct`s
as values passed in to the `object` parameter.
Cleans up unused object_store->ffi_struct_class.
Closes: https://github.com/dart-lang/sdk/issues/45071
Breaking change: https://github.com/dart-lang/sdk/issues/45072
TEST=vm/cc/DartAPI_FinalizableHandleErrors
TEST=vm/cc/DartAPI_WeakPersistentHandleErrors
TEST=tests/ffi(_2)/expando_test.dart
Change-Id: I9af6d0173db60614091068c218391f73756c135f
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-debug-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/195061
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
While deserializing AOT snapshot, Code objects which do not contain
valuable information besides entry point and stack maps are discarded
and not allocated on the heap (they are replaced with
StubCode::UnknownDartCode()).
PC -> Code/CompressedStackMaps lookup is implemented using a separate
table (InstructionsTable).
Flutter gallery in release-sizeopt mode:
Heap size of snapshot objects: arm -26.89%, arm64 -27.68%
Large Flutter application in release mode with --dwarf-stack-traces:
Heap size of snapshot objects: -24.3%.
Discarded Code objects: 72.5% of all Code objects.
Issue: https://github.com/dart-lang/sdk/issues/44852.
TEST=existing tests; "--dwarf_stack_traces --no-retain_function_objects
--no-retain_code_objects" mode is enabled for a few tests.
Change-Id: I5fe3e283630c8e8f4442319d5dcae38d174dd0d8
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/189560
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
The DateTime conversion to local time amounts to solving an
equation, finding an offset where the current local time is that offset.
The current approach doesn't actually do that, it just tries to simulate
the JS approach, with a tweak to avoid DST change issues, but that tweak
assumes that all time zone adjustments are one hour changes.
This CL should work correctly for time zone changes of up to one hour,
but might still fail for changes of more than one hour.
It works by trying more offsets, which obviously has a performance cost,
which depends on how efficient the operation system is at providing
the local time zone for a specific milliseconds-since-epoch.
Change-Id: I80dc6e62e0639d9966d3c5a06430787d8acc4ff1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/182041
Commit-Queue: Lasse R.H. Nielsen <lrn@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
Update some other methods to be more performant than the default
inherited from `MapMixin`.
Change argument validation to `map.[]=`, `map.putIfAbsent` and `set.add`
checks that the key/element is allowed by the `compare` function.
Previously the map methods just checked `key != null` (which was wrong),
and compare did nothing special, both just relied on the `splay` call
doing a `compare` call. That was not true for an empty map, which means
that you could add a key/element which wasn't a valid comparable,
as long as you did it as the *first* key/element.
That is now caught.
Bug: https://github.com/dart-lang/sdk/issues/45296
Change-Id: I8ba49a7892ed27daa4434f5fb7be3e777ef32bc9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/190883
Commit-Queue: Lasse R.H. Nielsen <lrn@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
1. Replace unintended '>>' with '>>>'
2. If infinity is an int, it is a large shift count and not an error.
3. Improve web coverage: Put the dynamic validation tests their own
(multi)parts so that failures of validation are separate to incorrect
results so can't mask incorrect results.
Change-Id: I569135507f1f23a01bde1f85f7ce91dec2bd8df4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/189960
Reviewed-by: Lasse R.H. Nielsen <lrn@google.com>
Commit-Queue: Stephen Adams <sra@google.com>
- Added support for --define to the VM and dart2js
- Added support for -D and --define for `dart run` and `dart compile js`
Remaining improvements:
- Add support for providing multiple comma separated values for `dart
run`, `dart`, and `dart2js`
Related issue: https://github.com/dart-lang/sdk/issues/44562
TEST=Updated CLI tests and added new dart2js tests.
This reverts commit e49937769f.
Change-Id: I5f9275b829665eb5e8695403d67f230e752ab0e6
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/183180
Commit-Queue: Ben Konyi <bkonyi@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
A "package URI" is defined as one with a `package` scheme,
no authority, a first path segment terminated by `/` which
contains no escapes and is not all `.` characters.
This is the definition of package names accepted by .packages
as well: Valid path characters and not all dots.
Change-Id: I9a161d47732e8bf873d278774315c72a4a928823
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/117542
Commit-Queue: Lasse R.H. Nielsen <lrn@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
commands
- Added support for --define to the VM and dart2js
- Added support for -D and --define for `dart run` and `dart compile js`
Remaining improvements:
- Add support for providing multiple comma separated values for `dart
run`, `dart`, and `dart2js`
Related issue: https://github.com/dart-lang/sdk/issues/44562
TEST=Updated CLI tests and added new dart2js tests.
Change-Id: I9379c7aee1eab377adb3438393d9ad79c4938cc4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/178262
Commit-Queue: Ben Konyi <bkonyi@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
Splitting out a failing test to increase coverage of other tests.
It is not clear that `a.addAll(a)` should throw if `a` is empty.
`a` is not modified by the action.
dart2js/DDC do not detect this case and I'm reluctant to add a check
for the case where `a` is in fact not modified.
The VM test is very specific to an identical argument and cannot
detect the case for `a = Mylist([])`.
Another option to this proposal is to simply delete the `[]` case.
In general, I'd like to move more ConcurrentModificationError tests
into assert statements.
Change-Id: I1f9559bdb43ea1bae0575413748c9d9d0de58b99
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/170023
Commit-Queue: Stephen Adams <sra@google.com>
Reviewed-by: Lasse R.H. Nielsen <lrn@google.com>
Fixes#42950
It was previously allowed, though non-sensible, to read `.current`
before advancing the iterator for a `RegExp.allMatches` iterable and get
`null`. Using `as` allows backward compatibility in unsound null safety
mode, while `!` is breaking since it introduces a null check in either
mode.
The new behavior is also the behavior of the migrated code for DDC since
https://dart-review.googlesource.com/c/sdk/+/151121
Change-Id: I9f3654298a3d480ec2f297a40a4145ff6ca9f011
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/157880
Auto-Submit: Nate Bosch <nbosch@google.com>
Reviewed-by: Lasse R.H. Nielsen <lrn@google.com>
Commit-Queue: Nate Bosch <nbosch@google.com>
Also relands the followup CLs:
"Perform non-covariant checks when dynamically invoking callables."
"Use AreValidArguments so that names are checked as well."
Original description of first CL:
This change only affects compilation when running in non-precompiled
mode with --no-lazy-dispatchers enabled.
Instead of always compiling in non-covariant checks, even for closures
not called dynamically, remove the non-covariant checks from the closure
and instead do the non-covariant checks for dynamic calls during the
NoSuchMethodForCallStub fallback by calling
Function::DoArgumentTypesMatch.
Adds two overloads for Function::DoArgumentTypesMatch, one which takes a
function type argument vector and one which takes neither an
instantiator type argument vector or a function type argument vector.
For the versions that are not explicitly passed a type argument vector,
an appropriate one is constructed using the arguments. If there is not
enough information in the arguments, then we fall back to assuming the
empty type argument vector for the instantiator case and instantiating
to bounds in the function type argument case.
Fixes Function::DoArgumentTypesMatch to handle generic functions and to
check arguments appropriately according to the active null safety mode.
For generic functions, the provided or resulting function type vector
has non-covariant checks performed against the type parameter bounds.
This change uncovered one test that was incorrectly passing in strong
mode, see https://github.com/dart-lang/sdk/issues/42688 for details.
Original description of second CL:
The VM only does this when the callable function does not expect dynamic
invocations. Otherwise, performing the checks would be redundant, as the
function body already contains the appropriate non-covariant checks.
Third CL had no additional description.
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-debug-x64-try,vm-dartkb-linux-release-x64-try,vm-kernel-reload-linux-release-x64-try, vm-kernel-reload-rollback-linux-debug-x64-try
Bug: https://github.com/dart-lang/sdk/issues/40813
Change-Id: I1a3e9c1865103a8d716e1cad814267caffaaadf2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/154688
Reviewed-by: Martin Kustermann <kustermann@google.com>
This reverts commit 556f0fb5c8 and fixes merge conflict.
Revert "[vm] Perform non-covariant checks when dynamically invoking callables." as a collateral.
This reverts commit 7a6287bb92.
Revert "[vm] Use AreValidArguments so that names are checked as well." as a collateral
This reverts commit 90c683018f.
Change-Id: Iee2c5e91c39d19dca7a1e640c103443c95e6c4f9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/154705
Reviewed-by: Alexander Aprelev <aam@google.com>
Commit-Queue: Alexander Aprelev <aam@google.com>
This change only affects compilation when running in non-precompiled
mode with --no-lazy-dispatchers enabled.
Instead of always compiling in non-covariant checks, even for closures
not called dynamically, remove the non-covariant checks from the closure
and instead do the non-covariant checks for dynamic calls during the
NoSuchMethodForCallStub fallback by calling
Function::DoArgumentTypesMatch.
Adds two overloads for Function::DoArgumentTypesMatch, one which takes a
function type argument vector and one which takes neither an
instantiator type argument vector or a function type argument vector.
For the versions that are not explicitly passed a type argument vector,
an appropriate one is constructed using the arguments. If there is not
enough information in the arguments, then we fall back to assuming the
empty type argument vector for the instantiator case and instantiating
to bounds in the function type argument case.
Fixes Function::DoArgumentTypesMatch to handle generic functions and to
check arguments appropriately according to the active null safety mode.
For generic functions, the provided or resulting function type vector
has non-covariant checks performed against the type parameter bounds.
This change uncovered one test that was incorrectly passing in strong
mode, see https://github.com/dart-lang/sdk/issues/42688 for details.
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-release-x64-try,vm-dartkb-linux-release-x64-try
Change-Id: I5030a46b356778448b51a243f16406eacf1c04bf
Bug: https://github.com/dart-lang/sdk/issues/40813
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/153202
Commit-Queue: Tess Strickland <sstrickl@google.com>
Reviewed-by: Régis Crelier <regis@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
The fixes to list_concurrent_modify_test are I think correct and
preserve the intent of the test. The changes to list_test get the test
working correctly and preserve the intent of the test but are
workarounds that shouldn't be needed once #42496 is fixed.
Change-Id: I82b32b2256b9940ceb934fb0f628311b8e7c6567
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/152988
Commit-Queue: Leaf Petersen <leafp@google.com>
Reviewed-by: Leaf Petersen <leafp@google.com>
Auto-Submit: Bob Nystrom <rnystrom@google.com>
The old RangeError.range message was confusing:
RangeError (index): Invalid value: Not in range 0..2, inclusive: 9
The comma makes the message hard to parse. I've seen a number of
people misinterpret "inclusive" as describing the invalid value (9)
instead of acting as a modifier on the range.
https://github.com/dart-lang/sdk/issues/29586 has a lot of
bikeshedding about this, but in the interest of mitigating confusion
sooner, I propose changing it to:
RangeError (index): Invalid value: Not in inclusive range 0..2: 9
I'm intentionally trying to improve the message with minimal
disruption to the structure of the error message. Although I'd much
prefer that "Invalid value" and the actual value be adjacent instead
of being interrupted by the explanation, such restructuring is rather
non-trivial. RangeError allows the "Invalid value" message to be
customized, and rearranging terms could produce even worse
constructions.
I also considered:
RangeError (index): Invalid value: Not in range [0, 2]: 9
And while I like that that is brief and clear, I chose not to use it
for people who are unfamiliar with interval notation and who might
mistake it as a Dart List.
Bug: https://github.com/dart-lang/sdk/issues/29586
Change-Id: I0f23b195437e4053ae5f76b5d303123979a8c9fe
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/146024
Commit-Queue: Lasse R.H. Nielsen <lrn@google.com>
Reviewed-by: Lasse R.H. Nielsen <lrn@google.com>
This should enable strong mode tests to be fully NNBD opted in. At the
same time, legacy tests when run with the experiment off should
hopefully be able to still consume these libraries because they don't
actually use any NNBD features.
To do this, I changed the script that generates the SDK repo's package
config to not put in a language version for any package whose pubspec
has no SDK version. According to language versioning, that means the
package should be considered to be at the "current" version of the SDK
running the code.
In NNBD, that's opted in. With the experiment off, that is (presumably)
"opted out" in the sense that NNBD doesn't exist.
In order to *not* opt in some of the other packages that currently
lack SDK constraints, I put those in and pinned them to "^2.7.0" which
was what the package config script used to implicitly fill in for them.
I see a bunch of other changes in the generated package config too. The
update script probably hasn't been run in a while.
Change-Id: I55193d42eac0696a6b0105546551efa45a1f3252
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/144305
Auto-Submit: Bob Nystrom <rnystrom@google.com>
Reviewed-by: Leaf Petersen <leafp@google.com>
Commit-Queue: Bob Nystrom <rnystrom@google.com>
Several instance checks added during the null safety core library
migration were incorrectly rejecting null values in places where null
values can still flow in legacy mode. This eliminates the instance
checks in favor of casts which will behave as desired.
Bug: https://github.com/dart-lang/sdk/issues/41465
Change-Id: I50fb9e6cef41d645d5312871932c7168f2d80c29
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/143333
Commit-Queue: Leaf Petersen <leafp@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
Erik Ernst