- Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
- A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
- Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
- Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
- reachabilityBarrier is now just a count of major GCs.
TEST=ci
Change-Id: Ic7754e8d972763654eae2b7faa8670735d9cda3f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/340644
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
This reverts commit 5daaa7d9eb.
Reason for revert: internal crashes
Original change's description:
> [vm, gc] Mark through new-space.
>
> - Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
> - A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
> - Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
> - Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
> - reachabilityBarrier is now just a count of major GCs.
>
> TEST=ci
> Change-Id: I4a6a23273d8ecb78c640f054731d4ceb737bfc4d
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/325840
> Reviewed-by: Siva Annamalai <asiva@google.com>
> Commit-Queue: Ryan Macnak <rmacnak@google.com>
Change-Id: I8a50074db343c63c14f0487ae8b4f5fee2c4ae76
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/330720
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Slava Egorov <vegorov@google.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
- Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
- A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
- Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
- Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
- reachabilityBarrier is now just a count of major GCs.
TEST=ci
Change-Id: I4a6a23273d8ecb78c640f054731d4ceb737bfc4d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/325840
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
The library name cache is dead.
The library exported name cache is used indirectly through Dart_Invoke etc, and is practically only used for one-shot lookups like `main` or the embedder's startup hooks. The lookups that were worth caching now happen in the Dart frontend instead of the VM.
Cf. b9045af57b
TEST=ci
Change-Id: Ibd8bd30fbba3e9af32e3fbf3c5c4ced51a8d15d7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/324401
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
This reverts commit 3fb88e4c66.
Reason for revert: b/297175670
Original change's description:
> [vm, gc] Mark through new-space.
>
> - Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
> - A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
> - Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
> - Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
> - reachabilityBarrier is now just a count of major GCs.
>
> TEST=ci
> Change-Id: I6362802cd93ba5ba9c39f116ddff82e4feb4c312
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/321304
> Commit-Queue: Ryan Macnak <rmacnak@google.com>
> Reviewed-by: Siva Annamalai <asiva@google.com>
Change-Id: I33075156160dc35861355d738a5776b74dce88b9
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/322344
Reviewed-by: Martin Kustermann <kustermann@google.com>
Auto-Submit: Ivan Inozemtsev <iinozemtsev@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
- Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
- A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
- Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
- Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
- reachabilityBarrier is now just a count of major GCs.
TEST=ci
Change-Id: I6362802cd93ba5ba9c39f116ddff82e4feb4c312
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/321304
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
This reverts commit e95e7b8e96.
Reason for revert: suspected cause to non-deterministic AOT binaries crashes, b/296014654
Original change's description:
> [vm, gc] Mark through new-space.
>
> - Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
> - A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
> - Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
> - Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
> - reachabilityBarrier is now just a count of major GCs.
>
> TEST=ci
> Change-Id: I8c2c64b120766571b62d3bd8dab37ae81c2dca98
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/319583
> Commit-Queue: Ryan Macnak <rmacnak@google.com>
> Reviewed-by: Siva Annamalai <asiva@google.com>
Change-Id: Idda542c7c657d4f14c836423b173c9b067132212
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/320820
Reviewed-by: Martin Kustermann <kustermann@google.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Commit-Queue: Ilya Yanok <yanok@google.com>
- Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
- A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
- Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
- Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
- reachabilityBarrier is now just a count of major GCs.
TEST=ci
Change-Id: I8c2c64b120766571b62d3bd8dab37ae81c2dca98
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/319583
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
This reverts commit 6194209b28.
Reason for revert: issues on arm32
Original change's description:
> [vm, gc] Mark through new-space.
>
> - Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
> - A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
> - Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
> - Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
> - reachabilityBarrier is now just a count of major GCs.
>
> TEST=ci
> Change-Id: I3668a2e56821f9eadf96e38c228dab27be656016
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/309826
> Reviewed-by: Siva Annamalai <asiva@google.com>
> Commit-Queue: Ryan Macnak <rmacnak@google.com>
Change-Id: I434eb595c9e7858efc8c9b07cbca954e5649f506
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/319321
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
- Initial and final marking no longer visit all of new-space, reducing the STW pause for major GC.
- A scavenge during concurrent marking must forward / filter objects in the marking worklist that are moved / collected, increasing the STW pause for minor GC.
- Unreachable intergenerational cycles and weak references are collected in the next mark-sweep instead of first requiring enough scavenges to promote the whole cycle or weak target into old-space.
- Artificial minor GCs are no longer needed to avoid memory leaks from back-to-back major GCs.
- reachabilityBarrier is now just a count of major GCs.
TEST=ci
Change-Id: I3668a2e56821f9eadf96e38c228dab27be656016
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/309826
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
The main purpose of the low-level [PortMap] is to coordinate between
ports being opened & closed and concurrent message senders. That's the
only thing it should do.
Each isolate owns [ReceivePort]s. Only the isolate mutator can create
ports, delete them or change their "keeps-isolate-alive" state. Right
now it requires going via [PortMap] (which acquires lock) and
[MessageHandler] (which acquires lock) to change the
"keeps-isolate-alive" state of a port.
We'll move information whether a dart [ReceivePort] is closed and
whether it keeps the isolate alive into the [ReceivePort] object itself.
=> Changing the "keeps-isolate-alive" state of the port no longer
requires any locks. We could even avoid the runtime call itself in a
future CL.
Isolates are kept alive if there's any open receive ports (that have not
been marked as "does not keep isolate alive"). This is a property of an
isolate not of the message handler. For native message handlers we do
have a 1<->1 correspondence between port and handler (i.e. there's no
"number of open ports" tracking needed).
=> We'll move the logic of counting open receive ports and ports that
keep the isolate alive to the [Isolate].
=> We'll also remove locking around incrementing/decrementing or
accessing the counts.
=> The [IsolateMessageHandler] will ask the [Isolate] whether there's
any open ports for determining whether to shut down.
=> For native ports, the `Dart_NewNativePort()` & `Dart_CloseNativePort()`
functions will manage the lifetime (as their name also suggests).
Overall this makes the [Isolate] responsible for creation of dart
[ReceivePort]s and tracking whether the isolate should be kept alive:
* Isolate::CreateReceivePort()
* Isolate::SetReceivePortKeepAliveState()
* Isolate::CloseReceivePort()
TEST=ci
Change-Id: I847ae357c26254d3810cc277962e05deca18a1de
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/317960
Reviewed-by: Alexander Aprelev <aam@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This change is almost trivial. The closure is stored on the callback's
RawReceivePort, not in the VM. So we can basically just remove the CFE
check and it pretty much works. The only problem is that we can't set
function.FfiCallbackTarget anymore, so most of the CL is dealing with
that.
A few places were deciding whether an FFI trampoline was a call or a
callback based on whether function.FfiCallbackTarget() was null. But
now the target will be null for async callbacks. So instead I've added
a new value to the FfiCallbackKind enum (and renamed it), and changed
those checks.
Sync callback closures will be a separate CL, because they're more
complicated.
Bug: https://github.com/dart-lang/sdk/issues/52689
Change-Id: I8e5dfb557362e679f66195b735c3c382e6792840
TEST=async_void_function_callbacks_test.dart
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/316160
Commit-Queue: Liam Appelbe <liama@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
a) Remove 2 fields in our object representation:
* PatchClass::library_data_
* Library::kernel_data_
=> This saves O(#libraries + #patch-classes) which can amount up to
10+ MB for big apps, as we save not just the slots but the
[ExternalTypedData] objects they used to reference.
Instead we'll compute the KernelLibraryData when we need it by making a
[TypedDataView] of sub parts of the component.
b) We make a kernel binary be represented by a single [ExternalTypedData].
Whenever we need a sub-part of a kernel binary (e.g. one for each
component for concatinated kernels, or one for a library inside a
component) we use proper [TypedDataView]s for that.
c) As we sometimes need to create a view of only a particular library
within a component we need to find start/end of a library based on
index.
=> We store the library index instead of the library offset on
Library/PatchClass.
=> We can easily derive the start/end of a library from it's index by
looking at the kernel component encoding.
d) We make the [Reader] object work purely based on a pointer - instead
of making it have if/else when reading bytes (either from pointer or
from a view).
e) We make the [KernelProgramInfo] store the kernel_component and
various TD views into it.
TEST=ci
Change-Id: Ibe160881ff48635e834c3d647a977a144b5d0565
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313561
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Slava Egorov <vegorov@google.com>
** KernelProgramInfo
The VM can load multiple kernel files into an isolate. Each kernel file
is represented as a [KernelProgramInfo] object. Loading a kernel blob
will result in a tree of objects
Library
- TopLevelClass
+ Field
+ Function
+ Class
+ Field
+ Function
+ ... closure functions ...
The entire tree belongs to one [KernelProgramInfo] object. There are
two exceptions
* after a hot-reload we inject [PatchClass]es (as owners of
old [Field]/[Function] objects) that identify the old kernel data
* expression evaluation functions will get a special data array with
information about the expression evaluation kernel data
=> We attach the [KernelProgramInfo] to the root of the tree, namely
the [Library]
=> We introduce `{Field,Function,Class,Library}::KernelProgramInfo()`
that will get that object - by walking up the chain.
=> We introduce `UntaggedPatchClass::kernel_program_info_` field that
allows saving the old [KernelProgramInfo] (on which the old
[Field]/[Function] objects are based upon) in the [PatchClass].
=> We include the [KernelProgramInfo] in the expression evaluation
data array.
** Scripts
A [Script] object should represent a .dart source file. It contains the
source, line offsets, etc. of that dart source file. The contents of a
dart source file may be used by many classes/libraries or even many
kernel blobs (e.g. via our copying mixin application).
When we build a tree like above, we'll attach a reference to a
[Script] object to [Class] objects. But members of a class may come from
a different script, in which case we inject an intermediary [PatchClass]
that references that other script.
When the compiler frontend loads kernel, builds flow graph, etc we walk
down the kernel binary (the tree above) - all belonging to the same
[KernelProgramInfo] object, but sub-parts may belong to different
[Script]s.
=> We use the existing [ActiveClass] class that keeps track of the
current class & member inside the class we're operating on.
=> Whenever we need the script we get it from the active member
function/class.
** Misc
We make kernel related classes (e.g. [KernelReaderHelper]) independent
of [Script].
We make the [Script] objects hold on to the [KernelProgramInfo] they are
based upon (e.g. for lazy source reading). But since a given [Script]
doesn't belong to a particular kernel blob (multiple kernel blobs may
contain information about the same script) we don't expose this in the
public API.
=> In the future we could have a unique [Script] object per url (instead
of having several as we do now)
TEST=ci
Change-Id: I7bd698f497b0bd8ab8e546540bb1dad5b5a6dd48
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313381
Reviewed-by: Jens Johansen <jensj@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Slava Egorov <vegorov@google.com>
GCC (used in the RISCV64 build) and clang (used to build other
architectures) have different views of what's visible in the return type
position for an external template method definition. Thus, move the
use into the definition body via constexpr if, which works for both.
TEST=dart-sdk-linux-riscv64-try
Cq-Include-Trybots: luci.dart.try:dart-sdk-linux-riscv64-try
Change-Id: Id7adfd93764ff11d99d3ded4893739b90bb25676
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/313123
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
There have been multiple cases in the past where using Object::null()
as the initializer, even for non-Ptr fields, has caused nondeterministic
snapshot creation. In particular, this can happen if a non-Ptr field
is only used in some versions of an object but not others, and thus
there is no reason to worry about what the value of the field is during
runtime. Only when snapshotting can it become relevant that the field
value is a portion of the address of Object::null(), which can vary
over different runs.
Instead of initializing the entirety of allocated objects (outside
of a few rare cases) with Object::null(), only initialize the portion
of the object corresponding to object pointer fields (in most cases,
between from() and to() inclusively) to Object::null() and initialize
the rest of the object to 0.
With this change, the only special casing in Object::InitializeObject
that remains is to skip initialization for TypedData and Arrays when
the memory is guaranteed to be zero initialized and to initialize the
contents of Instructions objects with the break instruction, not 0.
Note that this behavior does not occur in the following cases, which
all involve copying an arbitrary object:
* Object::Clone(), which zero-initializes the contents of the object
in all cases. However, the contents of the original object are then
copied over before the thread can reach a safepoint.
* Inside the ObjectGraphCopier, which uses the old initialization
behavior. This is safe, as any GC-important fields are immediately
copied over, and the rest of the contents are eventually copied over
before the caller receives the root of the copied object graph.
TEST=ci
Issue: https://github.com/dart-lang/sdk/issues/52876
Change-Id: Ib09fc562a8b6af97b509b493eb2d64109230ec35
Cq-Include-Trybots: luci.dart.try:vm-linux-debug-x64-try,vm-linux-debug-x64c-try,vm-linux-release-x64-try,vm-aot-linux-product-x64-try,vm-aot-linux-release-x64-try,vm-aot-linux-debug-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/312900
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
The main contribution of this CL is unification of disparate
handling of various functions like `Future.timeout`,
`Future.wait`, `_SuspendState.createAsyncCallbacks` and
`_SuspendState._createAsyncStarCallback` into a single
`@pragma('vm:awaiter-link')` which allows Dart developers
to specify where awaiter unwinder should look for the next
awaiter.
For example this allows unwinding to succeed for the code like this:
Future<int> outer(Future<int> inner) {
@pragma('vm:awaiter-link')
final completer = Completer<int>();
inner.then((v) => completer.complete(v));
return completer.future;
}
This refactoring also ensures that we preserve information
(including Function & Code objects) required for awaiter
unwinding across all modes (JIT, AOT and AOT with DWARF stack
traces). This guarantees users will get the same information
no matter which mode they are running in. Previously
we have been disabling awaiter_stacks tests in some AOT
modes - which led to regressions in the quality of produced
stacks.
This CL also cleans up relationship between debugger and awaiter
stack returned by StackTrace.current - which makes stack trace
displayed by debugger (used for stepping out and determinining
whether exception is caught or not) and `StackTrace.current`
consistent.
Finally we make one user visible change to the stack trace:
awaiter stack will no always include intermediate listeners
created through `Future.then`. Previously we would sometimes
include these listeners at the tail of the stack trace,
which was inconsistent.
Ultimately this means that code like this:
Future<int> inner() async {
await null; // asynchronous gap
print(StackTrace.current); // (*)
return 0;
}
Future<int> outer() async {
int process(int v) {
return v + 1;
}
return await inner().then(process);
}
void main() async {
await outer();
}
Produces stack trace like this:
inner
<asynchronous suspension>
outer.process
<asynchronous suspension>
outer
<asynchronous suspension>
main
<asynchronous suspension>
And when stepping out of `inner` execution will stop at `outer.process`
first and the next step out will bring execution to `outer` next.
Fixes https://github.com/dart-lang/sdk/issues/52797
Fixes https://github.com/dart-lang/sdk/issues/52203
Issue https://github.com/dart-lang/sdk/issues/47985
TEST=ci
Bug: b/279929839
CoreLibraryReviewExempt: CL just adds @pragma to facilitate unwinding
Cq-Include-Trybots: luci.dart.try:vm-aot-linux-product-x64-try,vm-aot-linux-debug-x64-try,vm-aot-linux-release-x64-try,vm-aot-obfuscate-linux-release-x64-try,vm-aot-dwarf-linux-product-x64-try
Change-Id: If377d5329d6a11c86effb9369dc603a7ae616fe7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/311680
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Slava Egorov <vegorov@google.com>
The [PatchClass] no longer needs to refer to a original [Class] and a
patched [Class] objects: Since the CFE does handling of patches the VM
only has a single [Class] object, so those two fields are always the
same.
=> Change `PatchClass::{origin_class,patched_class} -> `PatchClass::wrapped_class`
We also remove the `Field::Origin()` / `Function::origin()` methods
as they return the same as the `Owner()` would return.
TEST=ci
Change-Id: Iec0849f6ffc2026760dad89a9bcf07e9469bc8b6
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/311840
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This field is now only used when reloading to check assignability
during `ProgramReloadContext::ReloadPhase4CommitFinish()` (via
`PostCommit()` -> `InvalidateWorld()` -> `InvalidateFields`).
However, the method `DeoptimizeDependentCode()` is called near the
start of the actual reloading process in `Reload` prior to the
commit/rollback phases. This calls `DeoptimizeTypeTestingStubs()`, which
walks the heap and resets the contents of all `SubtypeTestCaches`
(`STC`s).
That means that the `STC` information isn't actually kept between
different reloads, so all this does is cause a number of `STC` objects
to be allocated during reload whose backing arrays are not used
post-reload for anything and so are unnecessary garbage.
That, and the creation of the `STC`s during `InvalidateFields` creates
all-input `STC`s, as they don't just check assignability to a given field type but also for checking elements of `RecordTypes`. That means
the per-field `STC`s contain way more information needed to check for
valid assignments to the corresponding field type, even if the `STC`
was used between reloads.
This CL removes the unnecessary field and changes `InvalidateFields`
so that the `STC`s used during reload are instead lazily allocated once
per `FieldInvalidator` object and now both the `STC` and its backing
array are now garbage post-field invalidation.
TEST=reload ci
Change-Id: I3ffa199551ba69f0afd4eba6303c6ff73d7473a3
Cq-Include-Trybots: luci.dart.try:vm-reload-linux-debug-x64-try,vm-reload-linux-release-x64-try,vm-reload-rollback-linux-debug-x64-try,vm-reload-rollback-linux-release-x64-try,vm-aot-linux-debug-x64-try,vm-aot-linux-product-x64-try,vm-aot-linux-release-x64-try,vm-linux-release-x64-try,vm-linux-debug-x64-try,vm-aot-dwarf-linux-product-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/311220
Commit-Queue: Tess Strickland <sstrickl@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
In particular, avoid storing/comparing the destination type unless the
SubtypeTestCache was created for a dynamic AssertAssignable instruction.
TEST=vm/cc/STC_{Linear,Hash}Lookup, vm/cc/TTS
Bug: https://github.com/dart-lang/sdk/issues/48345
Change-Id: I279c69a668ce19785785cf71707acf951f2b2133
Cq-Include-Trybots: luci.dart.try:vm-aot-linux-debug-simriscv64-try,vm-aot-linux-debug-x64-try,vm-aot-linux-release-x64-try,vm-aot-linux-product-x64-try,vm-aot-linux-release-simarm64-try,vm-aot-linux-release-simarm_x64-try,vm-aot-linux-debug-x64c-try,vm-aot-tsan-linux-release-x64-try,vm-aot-mac-release-arm64-try,vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-aot-dwarf-linux-product-x64-try,vm-linux-release-ia32-try,vm-linux-debug-x64c-try,vm-linux-debug-x64-try,vm-linux-debug-simriscv64-try,vm-linux-release-simarm64-try,vm-linux-release-simarm-try,vm-mac-release-arm64-try,vm-mac-release-x64-try,vm-tsan-linux-release-x64-try,vm-reload-rollback-linux-release-x64-try,vm-reload-linux-release-x64-try,vm-ffi-qemu-linux-release-arm-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/310340
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
Previously, SlowTypeTestStub would use the incoming instance and
destination type to determine which `SubtypeNTestCacheStub` to call,
where `N` is the number of inputs checked for each entry. The code
for generating `InstanceOf` instructions performs similar work,
inlining some checks and then falling back to checking and updating
an appropriate `SubtypeNTestCacheStub` if the inlined checks fail.
Meanwhile, the runtime would use all available information to
create a new entry when needed, depending on the `SubtypeNTestCacheStub`
called at runtime to ignore the parts of the entry it didn't need.
To illustrate a situation where the runtime might use more
information to create an entry than needed by the stub, consider
the following class header:
```
class C<T> extends D
```
If no subtype of `D` requires knowing the specific type arguments for
an instance to determine whether that instance is an instance of `D`,
then the compiler can generate a `Subtype1TestCacheStub` call, where
only the class id of the instance is checked.
However, in the runtime, when updating the cache, the runtime will
create an entry that includes the specific type arguments used to
instantiate `C` when the initial cache miss occurred, as the runtime
does not know how many inputs the code that checks that SubtypeTestCache
uses. If the cache has grown to the point that it is hash-based, that
means the runtime uses the instance type arguments of the first cache
miss to determine the correct place to create a new entry.
When the stub only performs a linear search over the cache contents,
this works, as every entry is checked until a matching entry is found,
and thus the stub can ignore unneeded information. However, if a cache
is hash-based, this doesn't work: if the first miss was an instance
of `C<int>`, then the runtime includes all that information in a new
entry. This means the stub has to calculate the instance type
arguments, which it did not have to do before, and if a subsequent
check uses different instance type arguments (e.g., an instance of
`C<String>`), the result may still be a cache miss, and the runtime
adds another entry which has the same information as the previous
entry as far as the stub was originally concerned.
Thus, to limit itself to considering only the inputs used by the
appropriate `SubtypeNTestCacheStub`, the runtime needs to be told how
many inputs are used. This CL adds an additional field to
`SubtypeTestCache`s which contains the number of used inputs, and
changes the places where `SubtypeTestCache`s are allocated in the
flow graph compiler and the `DRT_TypeCheck` runtime entry to pass the
appropriate number of inputs. To ensure that the runtime and stubs
agree on how many inputs should be used, the `SlowTypeTestStub` now
uses the value of this field to determine which `SubtypeNTestCacheStub`
to call.
TEST=vm/cc/STC_{Linear,Hash}Lookup
Bug: https://github.com/dart-lang/sdk/issues/48345
Change-Id: Ie16f2d12c4c5eae856b668f28178818203662d1c
Cq-Include-Trybots: luci.dart.try:vm-aot-linux-debug-simriscv64-try,vm-aot-linux-debug-x64-try,vm-aot-linux-release-x64-try,vm-aot-linux-product-x64-try,vm-aot-linux-release-simarm64-try,vm-aot-linux-release-simarm_x64-try,vm-aot-linux-debug-x64c-try,vm-aot-mac-release-arm64-try,vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-aot-dwarf-linux-product-x64-try,vm-linux-release-ia32-try,vm-linux-debug-x64c-try,vm-linux-debug-x64-try,vm-linux-debug-simriscv64-try,vm-linux-release-simarm64-try,vm-linux-release-simarm-try,vm-mac-release-arm64-try,vm-mac-release-x64-try,vm-tsan-linux-release-x64-try,vm-reload-rollback-linux-release-x64-try,vm-reload-linux-release-x64-try,vm-ffi-qemu-linux-release-arm-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/310180
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This adds the `close()` method to `DynamicLibrary`, which uses
`dlclose()` on Unix and `FreeLibrary` on Windows to close a dynamic FFI
library.
TEST=tests/ffi/dylib_close_test.dart
Closes https://github.com/dart-lang/sdk/issues/40159
Cq-Include-Trybots: luci.dart.try:vm-ffi-android-debug-arm-try,vm-ffi-android-debug-arm64c-try,vm-ffi-qemu-linux-release-riscv64-try,vm-ffi-qemu-linux-release-arm-try,dart2wasm-linux-d8-try,dart2wasm-linux-chrome-try,vm-reload-rollback-linux-debug-x64-try,vm-reload-linux-debug-x64-try,vm-win-debug-x64c-try,vm-win-debug-x64-try,vm-aot-android-release-arm_x64-try,vm-mac-debug-arm64-try,vm-mac-debug-x64-try,vm-asan-linux-release-x64-try,vm-aot-msan-linux-release-x64-try,vm-aot-android-release-arm64c-try,vm-aot-android-release-arm_x64-try,vm-aot-linux-debug-x64-try,vm-aot-linux-debug-x64c-try,vm-aot-mac-release-x64-try,vm-aot-mac-release-arm64-try,vm-aot-win-debug-x64c-try,vm-aot-win-release-x64-try
CoreLibraryReviewExempt: `dart:ffi` VM only API.
Change-Id: I73af98677b481902fe548bdcee56964a0195faf0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/307680
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Commit-Queue: Daco Harkes <dacoharkes@google.com>
After recursive types are gone, finalization of types can be
simplified: types can be finalized with a simple recursive traversal
and they no longer need to track "being finalized" state.
Also, finalization doesn't need to write back finalized types
because type finalization without canonicalization is always performed
in place and doesn't result in a new type (or type arguments vector).
TEST=ci
Change-Id: Ifc0a4e12aa410eb2be18c4a475c2f486cfa9ebf2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/299300
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
For class type parameters 'parameterized_class_id' duplicates
a recently added 'owner' and can be removed.
'owner' is changed to hold class reference as a class id (Smi).
TEST=ci
Change-Id: I93fc11e6bcfcc00058a5281b497f59c5c5847ea6
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/299160
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
TypeRef type wraps around another type and it was used to represent
a graph of recursive types. After [0], the only use of TypeRef is
for TypeParameter.bound which may indirectly reference the same
TypeParameter.
This change replaces TypeParameter.bound with TypeParameter.owner and
removes TypeRef entirely. Various parts of the VM no longer need to
handle and support TypeRefs.
TypeParameter.owner can reference a FunctionType, Class,
or, as an optimization, it can be set to null in order to share
class type parameters among different classes.
With the exception of the 'TypeParameter.owner' back pointer,
VM types are now not recursive and can be visited without
additional tracking.
Caveats:
* Generic FunctionType cannot be cloned in a shallow way:
when copying a FunctionType, type parameters should be cloned too
and their owners should be updated. For that reason,
a mapping between 'from' and 'to' function types
(FunctionTypeMapping) is maintained during type transformations
such as InstantiateFrom.
FunctionType::Clone is used instead of Object::Clone where
appropriate.
* When testing types for subtyping and equivalence, mapping
between function types is passed to make sure
type parameters belong to the equivalent function types.
* IL serializer needs to serialize function types as a whole before
serializing any types potentially pointing into the middle of a
function type (such as return type 'List<Y0>' pointing into
the middle of a function type 'List<Y0> Function<Y0>()').
[0] https://dart-review.googlesource.com/c/sdk/+/296300
TEST=ci
Change-Id: I67c2fd0117c6183a45e183919a7847fd1af70b3e
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/294165
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Before the calling convention change during the introduction of AOT, this was used to get code metadata associated with a frame. Its remaining non-test use is forming service IDs for code object, which can use the regular visitors.
TEST=ci
Change-Id: Ica8eafe988d19eba8d905d9dc5907afbfd559118
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/296705
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
This is a reland of commit 135443706b
Original change's description:
> [vm] Avoid expanding/flattening type arguments vectors in Type objects
>
> Previously, vectors of type arguments were expanded to include type
> arguments corresponding to superclasses both in the instances of
> generic classes and in Type objects (after type finalization).
> As a result, Type objects after finalization could be recursive and
> need to use extra TypeRef objects to break loops. The finalization of
> types was very complex and sometimes slow.
>
> This change simplifies the representation of Type objects: now they
> always have short type argument vectors, corresponding only to
> the type parameters of their own classes (both before and after
> finalization). Vectors of type arguments in the instances of generic
> classes are still expanded/flattened.
>
> This greatly simplifies type finalization, makes Type objects
> non-recursive and removes the need to create and handle excessive
> TypeRefs for type arguments corresponding to superclasses,
> as those type arguments are no longer included into types.
> The only remaining use of TypeRefs is for bounds of type parameters.
>
> In order to expand/flatten type arguments, new methods Type::GetInstanceTypeArguments / Class::GetInstanceTypeArguments
> are introduced. They build canonical declaration type arguments
> once (for each class), and then instantiate them as needed.
>
> There are also simple helper methods to shrink type arguments (TypeArguments::FromInstanceTypeArguments) and expand type arguments without filling type arguments corresponding to superclasses (TypeArguments::ToInstantiatorTypeArguments).
>
> Time of edge case 'regress_51960_test' 15min -> 300ms.
>
> TEST=ci, runtime/tests/vm/dart/regress_51960_test.dart
>
> Fixes https://github.com/dart-lang/sdk/issues/52022
> Fixes https://github.com/dart-lang/sdk/issues/51960
>
> Change-Id: I75b466b74698a33c0bb5e1dcbd29542e413812a1
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/295060
> Reviewed-by: Ryan Macnak <rmacnak@google.com>
> Commit-Queue: Alexander Markov <alexmarkov@google.com>
TEST=runtime/tests/vm/dart/regress_b_278841863_test.dart
Fixes b/278841863.
Change-Id: Ib1e20055bfb26e1df0a077300c69f0bec7152480
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/296300
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
This reverts commit 135443706b.
Reason for revert: breaks many targets in G3, see b/278841863
Original change's description:
> [vm] Avoid expanding/flattening type arguments vectors in Type objects
>
> Previously, vectors of type arguments were expanded to include type
> arguments corresponding to superclasses both in the instances of
> generic classes and in Type objects (after type finalization).
> As a result, Type objects after finalization could be recursive and
> need to use extra TypeRef objects to break loops. The finalization of
> types was very complex and sometimes slow.
>
> This change simplifies the representation of Type objects: now they
> always have short type argument vectors, corresponding only to
> the type parameters of their own classes (both before and after
> finalization). Vectors of type arguments in the instances of generic
> classes are still expanded/flattened.
>
> This greatly simplifies type finalization, makes Type objects
> non-recursive and removes the need to create and handle excessive
> TypeRefs for type arguments corresponding to superclasses,
> as those type arguments are no longer included into types.
> The only remaining use of TypeRefs is for bounds of type parameters.
>
> In order to expand/flatten type arguments, new methods Type::GetInstanceTypeArguments / Class::GetInstanceTypeArguments
> are introduced. They build canonical declaration type arguments
> once (for each class), and then instantiate them as needed.
>
> There are also simple helper methods to shrink type arguments (TypeArguments::FromInstanceTypeArguments) and expand type arguments without filling type arguments corresponding to superclasses (TypeArguments::ToInstantiatorTypeArguments).
>
> Time of edge case 'regress_51960_test' 15min -> 300ms.
>
> TEST=ci, runtime/tests/vm/dart/regress_51960_test.dart
>
> Fixes https://github.com/dart-lang/sdk/issues/52022
> Fixes https://github.com/dart-lang/sdk/issues/51960
>
> Change-Id: I75b466b74698a33c0bb5e1dcbd29542e413812a1
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/295060
> Reviewed-by: Ryan Macnak <rmacnak@google.com>
> Commit-Queue: Alexander Markov <alexmarkov@google.com>
Change-Id: If0b2077305620593b8f03ebf6c135375c4086b1a
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/296182
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: Alexander Thomas <athom@google.com>
Commit-Queue: Ilya Yanok <yanok@google.com>
Previously, vectors of type arguments were expanded to include type
arguments corresponding to superclasses both in the instances of
generic classes and in Type objects (after type finalization).
As a result, Type objects after finalization could be recursive and
need to use extra TypeRef objects to break loops. The finalization of
types was very complex and sometimes slow.
This change simplifies the representation of Type objects: now they
always have short type argument vectors, corresponding only to
the type parameters of their own classes (both before and after
finalization). Vectors of type arguments in the instances of generic
classes are still expanded/flattened.
This greatly simplifies type finalization, makes Type objects
non-recursive and removes the need to create and handle excessive
TypeRefs for type arguments corresponding to superclasses,
as those type arguments are no longer included into types.
The only remaining use of TypeRefs is for bounds of type parameters.
In order to expand/flatten type arguments, new methods Type::GetInstanceTypeArguments / Class::GetInstanceTypeArguments
are introduced. They build canonical declaration type arguments
once (for each class), and then instantiate them as needed.
There are also simple helper methods to shrink type arguments (TypeArguments::FromInstanceTypeArguments) and expand type arguments without filling type arguments corresponding to superclasses (TypeArguments::ToInstantiatorTypeArguments).
Time of edge case 'regress_51960_test' 15min -> 300ms.
TEST=ci, runtime/tests/vm/dart/regress_51960_test.dart
Fixes https://github.com/dart-lang/sdk/issues/52022
Fixes https://github.com/dart-lang/sdk/issues/51960
Change-Id: I75b466b74698a33c0bb5e1dcbd29542e413812a1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/295060
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Alexander Markov