This is the next step towards preventing compiler from directly peeking
into runtime and instead interact with runtime through a well defined
surface. The goal of the refactoring to locate all places where compiler
accesses some runtime information and partion those accesses into two
categories:
- creating objects in the host runtime (e.g. allocating strings, numbers, etc)
during compilation;
- accessing properties of the target runtime (e.g. offsets of fields) to
embed those into the generated code;
This change introduces dart::compiler and dart::compiler::target namespaces.
All code in the compiler will gradually be moved into dart::compiler namespace.
One of the motivations for this change is to be able to prevent access to
globally defined host constants like kWordSize by shadowing them in the
dart::compiler namespace.
The nested namespace dart::compiler::target hosts all information about
target runtime that compiler could access, e.g. compiler::target::kWordSize
defines word size of the target which will eventually be made different
from the host kWordSize (defined by dart::kWordSize).
The API for compiler to runtime interaction is placed into compiler_api.h.
Note that we still permit runtime to access compiler internals directly -
this is not going to be decoupled as part of this work.
Issue https://github.com/dart-lang/sdk/issues/31709
Change-Id: If4396d295879391becfa6c38d4802bbff81f5b20
Reviewed-on: https://dart-review.googlesource.com/c/90242
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
ARM write barrier goes from
ldrne lr, [thr, #+508]
blxne lr
to
blne <offset>
ARM64 write barrier goes from (similarly X64)
beq +8
ldrx lr, [thr, #1056]
blr lr
to
beq +4
bl <offset>
It reduces RX on arm/arm64 by around 0.9%
Though the write barrier wrappers stub has multiple entrypoints (one for
each available register). Because of this, we modify the relocation
logic to support per-call offsets into the target.
To avoid making the assembler code depend on StubCode/FlowGraphCompiler,
we set a closure, which the assembler can call.
Issue https://github.com/dart-lang/sdk/issues/33274
Change-Id: I9e3d68260cab7ef19ea88f1235c78d6031819d6d
Reviewed-on: https://dart-review.googlesource.com/c/90063
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This CL improves AOT code for StackOverflowInstr/CheckNullInstr:
* On ARM we can do a conditional pc-relative calls for the stack overflow
checks, getting rid of the slow-paths entirely.
* On ARM64 we can do pc-relative calls on the slow path, avoiding an
extra load.
Flutter gallery size impact (in bare instructions mode):
* ARM: -3.7% RX
* ARM64: -1.4% RX
Issue https://github.com/dart-lang/sdk/issues/33274
Change-Id: Ia1acd76ac6efa26642f99e1ce3e417100aa357f3
Reviewed-on: https://dart-review.googlesource.com/c/89620
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
- Introduce a slimmed down version of thread.h, which just depends on the
Zone and StackResource.
- Introduce a layering check that would prevent the coupling in the future.
This is the first step towards decoupling compiler from runtime.
There are multiple reasons to introduce the decoupling but the main
reason currently is to introduce a controlled surface through which
compiler reaches into runtime to catch any places where runtime word size
might influence the compiler and then enable building compiler that
targets 32-bit runtime but is embedded into a 64-bit runtime.
Issue https://github.com/dart-lang/sdk/issues/31709
Change-Id: Id63ebbaddca55dd097298e51c90d957a73fa476e
Reviewed-on: https://dart-review.googlesource.com/c/87182
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Dispatch interface calls via hashtable rather than inline cache.
InterfaceCall doesn't need to take arguments descriptor into account
when doing method lookup.
Change-Id: I30eae6ea638d1d2ad2cf3ff073c653fee3377f31
Reviewed-on: https://dart-review.googlesource.com/c/86106
Reviewed-by: Zach Anderson <zra@google.com>
Commit-Queue: Régis Crelier <regis@google.com>
If the --use-bare-instructions flag is enabled we will:
* Make call sites load the target directly from the pool (instead of
the code object) - this saves one instruction (and an indirect load)
* Ensure the object pool will have direct entry addresses by:
- Letting the clustered snapshot reader change any StubCode::UnlinkedCall()
in the object pool by it's monomorphic entry
- Change the code patcher to patch SwitchableCalls by writing the
monomorphic entry into the pool (instead of the code object)
Issue https://github.com/dart-lang/sdk/issues/33274
Change-Id: I4e41fc8e4461bde477cc559a6a4fccaaf3a350b5
Reviewed-on: https://dart-review.googlesource.com/c/86160
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This is the final CL which adds a new --use-bare-instructions flag to
the VM.
If this flag is set during AOT compilation, we will:
* Build one global object pool (abbr: GOP) which all code objects
share. This gop will be stored in the object store. The PP register
is populated in the enter dart stub and it is restored when
returning from native calls.
* Gets rid of the CODE_REG/PP slots from the dart frames. Instead the
compiled code uses the global object pool, which is always in PP.
* Starts emitting pc-relative calls for calls between two dart
functions or when invoking a stub.
Limitation: We only emit pc-relative calls between two code objects
in the same isolate (this is because the image writer is writing
instruction objects for vm-isolate/main-isolate seperately)
* We do compile-time relocation of those static calls after the
precompiler has finished its work, but before writing the snapshot.
This patches all the instruction objects with pc-relative calls to
have the right .text distance.
* We emit a sorted list of code objects in ObjectStore::reverse_code_table,
which will be used by the AOT runtime to go back from PC to Code
objects (where all metadata, e.g. stack maps, catch entry moves, pc
descriptors are available).
Issue https://github.com/dart-lang/sdk/issues/33274
Change-Id: I6c5dd2b1571e3a889b27e804a24c2986c71e03b6
Reviewed-on: https://dart-review.googlesource.com/c/85769
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This frees the name Bytecode that will become the name of a new VM class.
Class Code will only refer to compiled code and Bytecode will refer to kernel
bytecode.
Change-Id: If64a5d688f0bf75220e67ef2932f4141782e9dc3
Reviewed-on: https://dart-review.googlesource.com/c/82703
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Régis Crelier <regis@google.com>
On a Flutter app scrolling through a large table running on a Moto G 4, max pause time reduces from 129ms to 52ms.
Before
6-44ms CollectNewGeneration
118-129ms CollectOldGeneration
After
6-52ms CollectNewGeneration
3-10ms StartConcurrentMarking
20-42ms CollectOldGeneration (i.e., finalize marking)
Code size increases on Flutter Gallery:
ARM32 Instructions(CodeSize): 5772688 -> 5968624 (+3.39%)
ARM64 Instructions(CodeSize): 6082400 -> 6251008 (+2.77%)
Bug: https://github.com/dart-lang/sdk/issues/34002
Change-Id: Iac9a0b3d91b497b5016c4f89e8779c488f105a5b
Reviewed-on: https://dart-review.googlesource.com/c/76780
Reviewed-by: Siva Annamalai <asiva@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Previously MoveSpecial bytecode instruction used unsigned 8-bit A field
to encode target local variable index. This is too restrictive as we allow
signed 16-bit local variable indices.
After this change MoveSpecial has A_X encoding, with A field used for
special index (0 or 1), and signed 16-bit X field used for target local.
Change-Id: Ib87992e192b2923a39a9968913c39476947d3ea5
Reviewed-on: https://dart-review.googlesource.com/c/81440
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Régis Crelier <regis@google.com>
Auto-Submit: Alexander Markov <alexmarkov@google.com>
This is a reland of 6ba3e55ecc
The issue was that SlowTypeTestStub used in precompiled mode did not handle a
Smi instance before calling the Subtype2TestCache stub which does not support
it. See PatchSet 2 for the fix.
Is there a more efficient solution?
Original change's description:
> [VM runtime] Support Smi instances in type test cache.
>
> This adds SubtypeTestCache-based optimizations for type tests against
> * dst_type = FutureOr<T> (when T=int/num)
> * dst_type = T (when T = FutureOr<int/num>)
>
> Remove dangerous LoadClass pseudo assembler instruction (does not work for Smi).
> Handle instantiated void in type tests (along with dynamic and Object).
>
> Change-Id: I0df0fc72ff173b9464d16cc971969132b055a429
> Reviewed-on: https://dart-review.googlesource.com/c/81182
> Commit-Queue: Régis Crelier <regis@google.com>
> Reviewed-by: Martin Kustermann <kustermann@google.com>
Change-Id: I333ca47aebd7f0b663059ab6afc5d1cd8d7d5210
Reviewed-on: https://dart-review.googlesource.com/c/81320
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This adds SubtypeTestCache-based optimizations for type tests against
* dst_type = FutureOr<T> (when T=int/num)
* dst_type = T (when T = FutureOr<int/num>)
Remove dangerous LoadClass pseudo assembler instruction (does not work for Smi).
Handle instantiated void in type tests (along with dynamic and Object).
Change-Id: I0df0fc72ff173b9464d16cc971969132b055a429
Reviewed-on: https://dart-review.googlesource.com/c/81182
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Unlike the assembler, the register allocator is in a position to avoid a move. Use the call result register since it is the most popular value source.
This pattern will also be needed to get the write barrier slot into a fixed register for card-marking, since X64 and ARM don't have a TMP2.
dart2js aot product:
X64 Instructions(CodeSize): 9049952 -> 8994240 (-0.62%)
ARM Instructions(CodeSize): 9772096 -> 9711696 (-0.62%)
ARM64 Instructions(CodeSize): 10185056 -> 10122496 (-0.61%)
Bug: https://github.com/dart-lang/sdk/issues/34002
Change-Id: Id2e9dd7b3d86069a9da3ba4e141720511ce6399e
Reviewed-on: https://dart-review.googlesource.com/76307
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
After this CL, the interpreter is included by default in the
JIT VM under the flag --enable-interpreter.
Reland with fix to NativeArgument setup in simulator_arm.cc
Change-Id: Ib9b4df6eb4d997dfbe361188b8a127828c1d9c6f
Reviewed-on: https://dart-review.googlesource.com/74003
Reviewed-by: Siva Annamalai <asiva@google.com>
Reviewed-by: Régis Crelier <regis@google.com>
Commit-Queue: Zach Anderson <zra@google.com>
Shrinks the clustered snapshot from fewer object pool entries but increases the image snapshot from less deduplication. Removes memory traffic for these constants.
Flutter Gallery
profile Total(CodeSize): 15284383 -> 15271689 (-0.08%)
release Total(CodeSize): 11358403 -> 11361337 (+0.02%)
Change-Id: I2ee4d49ba11241da5bf2928cfa44a39d5605eb86
Reviewed-on: https://dart-review.googlesource.com/73680
Reviewed-by: Zach Anderson <zra@google.com>
Commit-Queue: Ryan Macnak <rmacnak@google.com>
Vyacheslav Egorov