This is a reland of 19e5749308
TEST=runtime/tests/vm/dart/analyze_snapshot_binary_test.dart
Original change's description:
> [vm] Add analyze_snapshot tool for AOT snapshot inspection
>
> Current skeleton to allow for instrumentation snapshots that can be
> built alongside Dart compilation artifacts and easily referenced for
> specific versions between Snapshot hash <-> DartSDK <-> Flutter Engine
>
> TEST=runtime/tests/vm/dart/analyze_snapshot_binary_test.dart
>
> Change-Id: Ie3757a265bbf457506c72fb62a625fea7bedcb68
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/221087
> Reviewed-by: Slava Egorov <vegorov@google.com>
> Commit-Queue: Slava Egorov <vegorov@google.com>
Change-Id: Ia1ea0071d30818440ae48484ff6c406236af5a4e
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/224526
Reviewed-by: Slava Egorov <vegorov@google.com>
Commit-Queue: Slava Egorov <vegorov@google.com>
The --enable-assertions flag is not avaiable in PRODUCT mode, so we skip
this negative subtest.
TEST=Fixes dartkp-product on a newly added test for vm/dart{,_2}/spawn_uri_aot_test
Change-Id: Ie7be219b6eb2a2ac53ea5048a844fafeb488ba54
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/232941
Reviewed-by: Alexander Aprelev <aam@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
Right now the implementation of `Isolate.spawnUri(<uri>, ...)` in
the standalone embedder is to ignore `<uri>` and make the spawnned
isolate from the same AOT snapshot as the main isolate.
This is very confusing and very incorrect.
Instead `Isolate.spawnUri()` should work if-and-only-if the given
`<uri>` points to a valid and compatible AOT snapshot. If not, it should
throw an appropriate exception.
Fixes https://github.com/dart-lang/sdk/issues/48375
Fixes https://github.com/dart-lang/sdk/issues/48326
TEST=vm/dart{,_2}/spawn_uri_aot_test
Change-Id: I279ace08ac1b9a9eed3ae03ebe5d9e2336c1e5c9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/232603
Reviewed-by: Tess Strickland <sstrickl@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
* Add team "groups" in tools/OWNERS_<group name>.
* Add top-level OWNERS as a fallback.
* Add OWNERS for all top-level directories.
* Add OWNERS to all packages.
For additional background information see go/dart-sdk-owners.
TEST=No op until code-owners is enabled.
Bug: b/200915407
Change-Id: I7fe6116cc599c749cd50ca16151d6d6a801d99d7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/229147
Reviewed-by: Jonas Termansen <sortie@google.com>
The test is flakily failing with low propabililty due to a race on
executing a copy of `process_test` and having this file opened for
writing mode (so it seems).
This issue is not related to isolates and we'll therefore remove it from
the "iso-stress" builder.
TEST=Removes flakiness from "iso-stress" builder.
Closes https://github.com/dart-lang/sdk/issues/48193
Change-Id: I1f817aed9b014b8cdd7a646f35003d40d4712d0a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/232624
Reviewed-by: Alexander Aprelev <aam@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This test was modified (in [0]) to use `dart:mirrors` which doesn't
work in AOT mode. Only tests that work in AOT are considered for
the iso-stress builder.
[0] https://dart-review.googlesource.com/c/sdk/+/230600
TEST=Fixes "iso-stress" builder failure to compile.
Change-Id: I354784a33550d816332611747c4d41b53d44b2a4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/232621
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This CL implements `WeakReference` in the VM.
* This reduces the size of weak references from 2 objects using 8 words
to 1 object using 4 words.
* This makes loads of weak reference targets a single load instead of
two.
* This avoids the fix-point in the GC and message object copying for
weak references. (N.b. Weak references need to be processed _after_
the fix-point for weak properties.)
The semantics of weak references in messages is that their target gets
set to `null` if the target is not included in the message by a strong
reference.
The tests take particular care to exercise the case where a weak
reference's target is only kept alive because a weak property key is
alive and it refers to the target in its value. This exercises the fact
that weak references need to be processed last.
Does not add support for weak references in the app snapshot. It would
be dead code until we start using weak references in for example the
CFE.
This CL does not try to unify weak references and weak properties in
the GC or messaging (as proposed in go/dart-vm-weakreference), because
their semantics differ enough.
Closes: https://github.com/dart-lang/sdk/issues/48162
TEST=runtime/tests/vm/dart/finalizer/weak_reference_run_gc_test.dart
TEST=runtime/tests/vm/dart/isolates/fast_object_copy_test.dart
TEST=runtime/vm/object_test.cc
TEST=tests/lib/isolate/weak_reference_message_1_test.dart
TEST=tests/lib/isolate/weak_reference_message_2_test.dart
Change-Id: I3810e919a5866f3ae8a95bd9aa23a880a0b0921c
Cq-Include-Trybots: luci.dart.try:app-kernel-linux-debug-x64-try,dart-sdk-mac-arm64-try,vm-canary-linux-debug-try,vm-fuchsia-release-x64-try,vm-kernel-gcc-linux-try,vm-kernel-asan-linux-release-x64-try,vm-kernel-linux-debug-x64c-try,vm-kernel-linux-debug-x64-try,vm-kernel-linux-debug-simriscv64-try,vm-kernel-mac-debug-x64-try,vm-kernel-nnbd-linux-debug-x64-try,vm-kernel-nnbd-linux-release-ia32-try,vm-kernel-nnbd-linux-release-simarm64-try,vm-kernel-nnbd-linux-release-simarm-try,vm-kernel-nnbd-mac-debug-arm64-try,vm-kernel-nnbd-mac-debug-x64-try,vm-kernel-nnbd-win-release-ia32-try,vm-kernel-nnbd-win-release-x64-try,vm-kernel-optcounter-threshold-linux-release-x64-try,vm-kernel-precomp-android-release-arm_x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-debug-x64c-try,vm-kernel-precomp-win-debug-x64c-try,vm-kernel-reload-rollback-linux-debug-x64-try,vm-kernel-reload-linux-debug-x64-try,vm-kernel-win-debug-ia32-try,vm-kernel-win-debug-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/232087
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
The serial scavenger was expecting the typed_data field of views.
Fast object copy leaves a null there temporarily.
Closes: https://github.com/dart-lang/sdk/issues/48360
TEST=runtime/tests/vm/dart/isolates/fast_object_copy_test.dart
Change-Id: I72a2785cd9708e72593d35179536a831193d2a44
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/232260
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Fixes cases with passing floats, passing aggregates by value, passing scalars requiring alignment.
The remaining failing tests look like some kind of corruption or register reuse when passing very large numbers of arguments.
TEST=ci
Bug: https://github.com/dart-lang/sdk/issues/48164
Change-Id: Ibac32d4f291255e1c2eb8fd1b196b3e2f006f610
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231282
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Use `hasScheme` in place of comparing against the empty string, and
`isScheme` to compare against all other schemes.
TEST=No behavior changes.
Change-Id: Ifc9fd13c6cf37933ebd4a754c4b500dedbcb291b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231185
Reviewed-by: Kevin Moore <kevmoo@google.com>
Commit-Queue: Nate Bosch <nbosch@google.com>
The real token position is required in case debugger stopped on
an exception thrown from AssertAssignable and trying to evaluate
an expression which uses receiver. Without token position all local
variables (including receiver) are considered out of scope
(as frame's token position is out of range).
TEST=runtime/observatory/tests/service/regress_48279_test.dart
Fixes https://github.com/dart-lang/sdk/issues/48279
Change-Id: Idb6a8e12185367ceac2034269b8f046272384006
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231748
Reviewed-by: Ben Konyi <bkonyi@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Previously all arrays were excluded from WB elimination pass
to avoid invariant restoration code creating excessive work
for the GC and to avoid dealing with card marking in the
invariant restoration code.
It seems reasonable to enable this for small arrays of up to 8
elements. The cut of point of 8 elements was chosen based on the
cut of point for literal<N> specialisations provided by the
core library for creating small literal arrays.
TEST=vm/cc/IRTest_WriteBarrierElimination_Arrays
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-dwarf-linux-product-x64-try,vm-kernel-precomp-linux-debug-simarm_x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-debug-x64c-try,vm-kernel-precomp-linux-product-x64-try,vm-kernel-precomp-linux-release-simarm-try,vm-kernel-precomp-linux-release-simarm64-try,vm-kernel-precomp-linux-release-simarm_x64-try,vm-kernel-precomp-linux-release-x64-try
Change-Id: I2b3169865f07c3ff95820c1bc6718943e96bd33b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/229903
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Slava Egorov <vegorov@google.com>
In general any async function can return X or Future<X>. Though the
future implementation has to do different things depending on which case
we're in. It does so by using a `<obj> is Future<T>` test. This test is very
expensive if many different classes flow into `<obj>`.
Though most functions do not return `Future` objects from async
functions. We can determine that statically by looking if any returned
expression could be a future. If not, we can bypass the `is Future<T>`
test entirely.
Issue https://github.com/dart-lang/sdk/issues/48226
Issue https://github.com/dart-lang/sdk/issues/48235
TEST=pkg/front_end/testcases/general/async_function_returns_future_or.dart
Change-Id: If655bdbdddc214dd7b12be9905b3c788252547d0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/230662
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Lasse Nielsen <lrn@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This reverts commit e32d98cd06.
Reason for revert: Dart rolls into the engine are failing because there are references to run_binary_size_analysis in the "Upload artifacts android-arm64-release" step of the build
Original change's description:
> [tools] Replace the Chromium binary size tool, which no longer works.
>
> Change-Id: Id84717e21a129a392d3bc4e9b4cce84dfb4771e1
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231066
> Reviewed-by: Ben Konyi <bkonyi@google.com>
# Not skipping CQ checks because original CL landed > 1 day ago.
Change-Id: Ibfbdf1e0a970ad7fae9ec1d39d24722647b38730
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231600
Reviewed-by: Siva Annamalai <asiva@google.com>
Reviewed-by: Ben Konyi <bkonyi@google.com>
Commit-Queue: Siva Annamalai <asiva@google.com>
Register allocator does not need to associate always-calls
safepoints with live ranges of constant values because
constant values are never actually spilled and are rematerialized
instead.
@alexmarkov has observed that SafepointPosition objects for
constants account for large chunk of the total memory eaten by
the register allocation pass and this change addresses this issue.
With this change the total memory consumed by the register allocator
decreases from 21 Gb to 8.6 Gb on one of the examples
in https://github.com/flutter/flutter/issues/94461.
TEST=ci
Change-Id: I245f2678bc6f5c1d63fd282812d1df2b2958512f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231442
Auto-Submit: Slava Egorov <vegorov@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Slava Egorov <vegorov@google.com>
This change removes distinct Slots for the same field:
PointerBase_data_field -> PointerBase_data
Pointer_data_field -> PointerBase_data
TypedDataBase_data_field -> PointerBase_data
Also, the following slot is renamed to match field name:
TypedDataView_data -> TypedDataView_typed_data
The following offsets are renamed to match field names /
declarations in Untagged* classes:
PointerBase::data_field_offset -> PointerBase::data_offset
TypedData::data_offset -> TypedData::payload_offset
TypedDataView::data_offset -> TypedDataView::typed_data_offset
TEST=ci
Fixes https://github.com/dart-lang/sdk/issues/48273
Change-Id: I602545fc43018494bcb24d8692292fdbc6a8f3e0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231160
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
This adjusts RISC-V to match the other architectures in checking only the generational barrier.
TEST=fuzzer
Change-Id: I86965bb2caaae27aa9770769f392cffc8edb6eb8
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231046
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
On 32-bit ARM in AOT mode Pointer.asTypedList is generated so that
there is a LoadField from Pointer.data_field which has kUnboxedFfiIntPtr
representation (uint32) and then the value is passed to a
StoreInstanceField for TypedDataBase.data_field which has kUnboxedIntPtr
representation (int32). As a result, a speculative uint32->int32
IntConverter instruction is inserted by SelectRepresentations pass.
AOT doesn't support deoptimization so code generation crashes after
retrying without speculative inlining.
This change fixes the type incompatibility by loading value with
LoadUntagged and then converting it with ConvertUntaggedToUnboxed(kUnboxedIntPtr).
TEST=ffi/regress_flutter97301_test
Fixes https://github.com/flutter/flutter/issues/97301
Change-Id: I4a00d4ac7978b4775add0ddae510841a2b4cbae0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/230956
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
This reverts commit 8d412e80e1.
Reason for revert: b/209838275#comment53
TEST=ci
Original change's description:
> [vm] Disable longjmp() wrapper in exceptional flow in TSAN until g3 issue is fixed
>
> Issue b/209838275
>
> TEST=ci
>
> Change-Id: Ie0a2e6f68a0c4f855dfd367dd3dd0a37053be479
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/222760
> Reviewed-by: Ivan Inozemtsev <iinozemtsev@google.com>
> Commit-Queue: Martin Kustermann <kustermann@google.com>
# Not skipping CQ checks because original CL landed > 1 day ago.
Change-Id: Iab202ddbd54060f609096b71f8ef83da53249d40
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231005
Reviewed-by: Ivan Inozemtsev <iinozemtsev@google.com>
Commit-Queue: Ivan Inozemtsev <iinozemtsev@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
File browser web interface might fail to download large files, so one needs
to use cas tool to download those.
Change-Id: I8053860f9ba443dc481319e5574ecd6d41baf8f7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/225560
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Alexander Aprelev <aam@google.com>
Missing:
* Function types.
* Type variables in bounds, like "method<T, S extends T>() {}" and
"method<T extends Foo<T>>() {}"
This is in many ways a follow-up to
https://dart-review.googlesource.com/c/sdk/+/212286
TEST=service and (cfe) expression suite tests added.
Change-Id: I20472b59ed73e9845f073e176d73b2c213f9407a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/215760
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Ben Konyi <bkonyi@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
FinalizeGraph pass removes redefinitions and can change reachable
types of values. As a result, type of an input value of Unbox
instruction can become incompatibe with representation of the result of
Unbox, even if Unbox cannot deoptimize (this may happen if
the instruction is unreachable).
Previously, code generation failed with an assertion in that case.
This change adjusts code generation to silently generate load from
box instead.
TEST=runtime/tests/vm/dart/regress_48196_test.dart
Fixes https://github.com/dart-lang/sdk/issues/48196
Change-Id: I35f80a1b3135687901cd1b57c0e5572d29c4d2e0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/230341
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Slava Egorov <vegorov@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Certain optimizations are disabled when compiling huge functions.
Also, compiling large functions may cause out of memory in compiler.
This change adds a warning in order to make it easier for users to
find out about huge functions in their programs and diagnose OOM
during compilation.
TEST=ci
Issue: https://github.com/flutter/flutter/issues/94461
Change-Id: Ibc7fb69df21fa96b8159d24c1b71ef8d690b23df
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/230160
Reviewed-by: Slava Egorov <vegorov@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
We're adding these types to `dart:ffi` rather than `package:ffi` so that
they can be used with `FfiNative`s.
Adds `NativeType`s for the following C types:
* unsigned char
* signed char
* short
* unsigned short
* int
* unsigned int
* long
* unsigned long
* long long
* unsigned long long
* uintptr_t
* size_t
* wchar_t
Because the C standard only defines minimum sizes for many of these
types, future platforms might diverge from the typical size even if all
platforms currently agree on a size. To avoid having to reification
later, we define all types as AbiSpecificIntegers rather than typedefs,
even if all current target platforms agree on the size.
Closes: https://github.com/dart-lang/sdk/issues/36140
TEST=tests/ffi/c_types_test.dart
Original patch in patchset 1.
* Removes `Char` for now until package:win32 has rolled to 2.3.8 in
Flutter. https://pub.dev/packages/win32/versions/2.3.8/changeloghttps://logs.chromium.org/logs/flutter/buildbucket/cr-buildbucket/8824468064587445729/+/u/Android_Views_Integration_Tests/stdout
* Adds `c_type.dart` in `ffi_source.gni` which should fix `IntPtr`
missing when analyzing `path_provider_linux`. (However, I was unable
to reproduce the issue locally.)
https://logs.chromium.org/logs/flutter/buildbucket/cr-buildbucket/8824468064571399025/+/u/run_test.dart_for_flutter_plugins_shard_and_subshard_analyze/test_stdout
`/tmp/flutter_plugins.KZMNMC/packages/path_provider/path_provider_linux$ ~/flt/engine/src/out/host_debug/dart-sdk/bin/dart analyze --fatal-infos`
Change-Id: I89130cccba285fc9c483bb53f5710a302f2b104f
Cq-Include-Trybots: luci.dart.try:dart-sdk-linux-try,dart-sdk-mac-try,dart-sdk-win-try,vm-ffi-android-debug-arm64c-try,vm-ffi-android-debug-arm-try,vm-canary-linux-debug-try,vm-fuchsia-release-x64-try,vm-kernel-gcc-linux-try,vm-kernel-asan-linux-release-x64-try,vm-kernel-linux-debug-x64-try,vm-kernel-linux-debug-ia32-try,vm-kernel-mac-debug-x64-try,vm-kernel-mac-release-arm64-try,vm-kernel-nnbd-win-release-ia32-try,vm-kernel-nnbd-win-release-x64-try,vm-precomp-ffi-qemu-linux-release-arm-try,vm-kernel-win-release-x64-try,flutter-analyze-try,flutter-engine-linux-try,flutter-frontend-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/229156
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
This change improves a few implementations of hashcodes in compiler.
Slightly improves AOT compilation time:
gen_snapshot 60.736s -> 58.920s (-2.9%)
(on a large Flutter app, compiled in release mode for arm64).
Also, the same large app can be now compiled with
--hash_map_probes_limit=1500, meaning that all hash maps in
the compiler perform less than 1500 probes when looking for an
element.
This change also adds a test which verifies that kernel compiler
itself can be compiled with --hash_map_probes_limit=1000.
This is a sanity check to ensure we do not have a very
badly distributed hashcode in the compiler.
TEST=ci
Issue: https://github.com/dart-lang/sdk/issues/43299
Change-Id: I7a802709727a33760c4f1d13f7b2c8cb263852d7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/229940
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Slava Egorov <vegorov@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
