Since we no longer add the symbol for a given code object to the Elf
object when writing the contents of the text segment, we don't need to
calculate the relocated address that early. Instead, just store which
text segment the instructions payload belongs to as well as its offset
into the segment contents and let Elf::Finalize() calculate the
relocated address when creating the symbol.
This removes one need for knowing the segment base address when writing
the segment contents. The remaining needs are related to BSS usage:
* Calculating the BSS offset from the text segment for the Image header.
* Fixing up BSS relocations in instructions payloads.
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-product-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm_x64-try
Change-Id: I8be860a7a7111721e6f546154a0814f4d16ac634
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150929
Commit-Queue: Tess Strickland <sstrickl@google.com>
Reviewed-by: Clement Skau <cskau@google.com>
This includes support for calling Dart_PropagateError in native code
when doing FFI calls, and catching uncaught exceptions with Dart_IsError
when doing FFI callbacks.
The support for Dart_PropagateError adds a catch entry to the FFI
trampoline, which prevents inlining these trampolines in AOT. This
regresses the FfiCall benchmarks by 1-2% in AOT.
In addition, Dart_PropagateError requires maintaining a bit whether we
entered native/VM code from generated code through FFI or not. That way
we can do the proper transition on the exception path. When entering
generated code, we store this bit on the stack, right after the entry
frame.
Design: http://go/dart-ffi-handles
Issue: https://github.com/dart-lang/sdk/issues/36858
Issue: https://github.com/dart-lang/sdk/issues/41319
Change-Id: Idfd7ff69132fb29cc730931a4113d914d4437396
Cq-Include-Trybots: luci.dart.try:vm-ffi-android-debug-arm-try,vm-ffi-android-debug-arm64-try,app-kernel-linux-debug-x64-try,vm-kernel-linux-debug-ia32-try,vm-kernel-win-debug-x64-try,vm-kernel-win-debug-ia32-try,vm-kernel-precomp-linux-debug-x64-try,vm-dartkb-linux-release-x64-abi-try,vm-kernel-precomp-android-release-arm64-try,vm-kernel-asan-linux-release-x64-try,vm-kernel-linux-release-simarm-try,vm-kernel-linux-release-simarm64-try,vm-kernel-precomp-android-release-arm_x64-try,vm-kernel-precomp-obfuscate-linux-release-x64-try,dart-sdk-linux-try,analyzer-analysis-server-linux-try,analyzer-linux-release-try,front-end-linux-release-x64-try,vm-kernel-precomp-win-release-x64-try,vm-kernel-mac-debug-x64-try,vm-precomp-ffi-qemu-linux-release-arm-try,vm-kernel-nnbd-linux-debug-x64-try,analyzer-nnbd-linux-release-try,front-end-nnbd-linux-release-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/145591
Commit-Queue: Daco Harkes <dacoharkes@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
In order to normalize 'X extends Never' to 'Never' (and 'X? extends Never' to 'Null'), the VM has to instantiate the bounds of function type parameters.
Instantiating these bounds lazily is not easily feasible, as the instantiators are not kept around after instantiating types.
Note that the normalization tests do not yet pass due to test issues and CFE issues, see https://github.com/dart-lang/sdk/issues/42143
Change-Id: I9ed6fcb9d7a11a4adbe514757ad9ac6ff7818981
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/147281
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Running tsan build locally on IsolateJson identified concurrent access to isolate_ from NoActiveIsolateScope/CollectNewSpaceGarbage and IsMutatorThread/SafepointThreads/Scavenge.
Change-Id: I4c6f5c83a7f6ddbfee8ede59443720db62a48b6f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/151023
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Alexander Aprelev <aam@google.com>
Operator== has parameter type Object, which is non-nullable in NNBD
strong mode. This is useful, as null should be handled magically by
the implementation, and the body of operator== should never receive
null.
However, VM handles comparison of an object with null inside
operator==, so its parameter type should not be taken into account
when inferring type of ParameterInstr. Otherwise implicit null
handling code is removed by the optimizer.
Fixes https://github.com/dart-lang/sdk/issues/42287
Change-Id: I856a2ce47a11309e8c32d4d9bb481d7cc03bab85
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/151021
Reviewed-by: Régis Crelier <regis@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Previously, we created a separate DWARF object for each possible output.
For the most part, the information stored in each DWARF object was the
same, but the DWARF object was responsible for keeping track of abstract
origin addresses (in the non-assembly case) and for actually generating
the desired output (assembly or binary).
Instead, abstract out the output formatting into a DwarfWriteStream,
whose subclasses handle both the final output format and certain details
that differ between assembly and binary outputs. With this, the DWARF
class no longer depends on having direct access to the ELF file if being
used to generate ELF sections, and as such is output agnostic.
Now, if we have one or more outputs that contain DWARF information, we
only generate a single DWARF object that is shared across these outputs.
The only additional change required is in Dwarf::AddCode(), which now
handles being called multiple times for the same Code object.
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-product-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm_x64-try
Change-Id: I71de887a07a673164b3ccfa7fbbab2ccef80c576
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150302
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
- Detect null safety when not specified before isolate
initialization
- for source files by having CFE parse the source file
for @dart annotations
- for kernel files by sniffing the kernel file for
compilation mode
- for appJIT files by sniffing the feature string
- for AOT snapshots by sniffing the feature string
- Remove workaround of returning null safety to false during
bootstrapping
- Add a new Dart C API call for detecting null safety
Bug: 41766
Change-Id: Ia8cf264323a2d0d58c2855ce6491456aa6f1da07
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150089
Commit-Queue: Siva Annamalai <asiva@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This is a reland of 6a98e2719e
Original change's description:
> [VM/nnbd] Canonicalize TypeParameter objects in the VM.
>
> Prior to this CL, type parameters were all assumed canonical, even duplicate ones.
>
> Per a new convention introduced in this CL, the type parameter array in generic classes and generic functions contains canonical type parameters. As these type parameters get cloned with a different nullability, they are inserted in a new hash table of canonical type parameters.
>
> This fixes performance issue https://github.com/dart-lang/sdk/issues/41421
>
> Change-Id: I9086158fa6b6261e9997bb50edec6d7c54abbfa1
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/148223
> Commit-Queue: Régis Crelier <regis@google.com>
> Reviewed-by: Ryan Macnak <rmacnak@google.com>
Change-Id: I4a42d0f9d53c2244b1a638432ed2cd6a055fa72a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150266
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
The newly added --worker-thread-priority can be used by users as an
opt-in to set the thread priority / nice value.
It is the responsibility of the user to ensure that the process has the
right to change the priority to the given value. Failure to set the
priority will be fatal.
See b/154918152
Change-Id: I3b3791e88b6ddf4fa6e39e4587f7fe1ab4d2312f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150560
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
This avoids attributing the header and padding bytes in Image objects
to the artificial root, instead making it clear why these bytes exist.
It also ensures Instructions objects (in the non-bare-payload case) are
never considered roots of the profile.
Also unify semantics of text_offset in AssemblyImageWriter::WriteText()
and BlobImageWriter::WriteText(). Originally the former was the offset
from the start of the text section and the latter was the offset from
the start of the instructions objects/payloads. Now they're both the
offset in the text section.
Bug: https://github.com/dart-lang/sdk/issues/41249
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-product-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm_x64-try
Change-Id: I617e76c9c390a93abb6bc37efc1355368c64691e
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149286
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
Today, Stdout and Stderr have a property supportsAnsiEscapes. The
documented purpose of this property is to inform the caller whether ANSI
escape sequences are supported - say, for colorizing terminal output
when the user is running the Dart program in a terminal emulator like
xterm.
The problem is that the heuristic used is a substring match for "xterm"
in the TERM environment variable. (See attached bug.) This
implementation omits a few significant populations of users:
* Users using screen or tmux. These commands set their TERM to "screen".
* Users using a non-xterm terminal like rxvt (especially rxvt-unicode).
The correct solution here is to query terminfo for terminal
capabilities. Until then, adding these two cases to the existing xterm
check unblocks users of two popular terminal multiplexers and a terminal
emulator.
Bug: https://github.com/dart-lang/sdk/issues/31606
Change-Id: I1a8a37d42f1895d7b1238c67ec86207fd4a38bbb
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149940
Auto-Submit: Wesley Moy <wesleymoy@google.com>
Reviewed-by: Jonas Termansen <sortie@google.com>
Reviewed-by: Zichang Guo <zichangguo@google.com>
Commit-Queue: Jonas Termansen <sortie@google.com>
BSS::Initialize() can run multiple times when isolate code is
loaded in separate threads. Even though the values used each
time are consistent and so they'll only change the first time,
TSAN sees the compare_exchange_strong calls as possible
read/write data races.
Cq-Include-Trybots: luci.dart.try:vm-kernel-tsan-linux-release-x64-try,vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-product-x64-try,vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm_x64-try
Bug: https://github.com/dart-lang/sdk/issues/42095
Change-Id: Iedb368939f9d86889a820edfa534522a4731a9d7
Fixed: 42095
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150340
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Tess Strickland <sstrickl@google.com>
There is a race condition for copying file on Windows, where CopyFile()
returns success but data has not been populated into destination file.
E.g process is killed or died in the middle.
This cl will change File::Copy as
1. Copy file to a temp file in the same directory of destination file.
2. Rename the file to the target file.
Bug: https://github.com/dart-lang/sdk/issues/42119
Change-Id: I39b6d451f6ace970bc554501148259d33de232c7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149667
Commit-Queue: Zichang Guo <zichangguo@google.com>
Reviewed-by: Zach Anderson <zra@google.com>
This option can be used to split AOT kernel generation into two rounds:
* The 1st round uses CFE to parse source files and generate intermediate
dill file in *non-AOT* mode.
* The 2nd round loads intermediate dill file and applies AOT global
transformations, then generates final dill file.
Each round is faster than one-shot generation of a dill file in AOT mode.
In order to use this 2-round compilation, command
gen_kernel <other-options> --aot -o output.dill main.dart
should be replaced with the following 2 commands:
gen_kernel <other-options> --no-aot -o intermediate.dill main.dart
gen_kernel <other-options> --aot --from-dill intermediate.dill -o output.dill main.dart
Bug: b/158101962
Change-Id: I2d4deef45b50eab11df47f4a372598e6cf1d0edf
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150272
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Alexander Aprelev <aam@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
This avoids treating addresses within segments added after the isolate
text section as valid PC addresses for our DWARF information.
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-release-simarm_x64-try,vm-kernel-precomp-linux-product-x64-try
Change-Id: Ibdac611d97b5a24226c226a3904c958f131ab961
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149290
Commit-Queue: Tess Strickland <sstrickl@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
Prior to this CL, type parameters were all assumed canonical, even duplicate ones.
Per a new convention introduced in this CL, the type parameter array in generic classes and generic functions contains canonical type parameters. As these type parameters get cloned with a different nullability, they are inserted in a new hash table of canonical type parameters.
This fixes performance issue https://github.com/dart-lang/sdk/issues/41421
Change-Id: I9086158fa6b6261e9997bb50edec6d7c54abbfa1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/148223
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Clean up signatures of runtime functions in object.{cc,h} using the trail to avoid cycles. Move the trail to the last parameter of these functions.
Introduce a trail in IsSubtypeOf() to avoid cycles introduced by bounds of F-bounded types. Issue uncovered by the new normalization tests.
Change-Id: I3241c7e4023a09c122e1594b7aff90b5b103f4f4
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150180
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Initializer functions are created lazily at compile time when generating
LoadField which calls initializer. This may happen in the background
compiler thread, where Field objects are cloned.
EnsureInitializerFunction should be called from original field, and it
should be properly synchronized to make sure there are no races with
mutator thread.
Change-Id: Icc339afa2ee410385019c72c717737719767d367
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/150166
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Discovered using the DevTools widget inspector with a Flutter
application which did not have a flutter_tools instance connected. The
widget inspector attempts to evaluate expressions, the VM service
realizes there's no external compilation service and attempts to fall
back onto the kernel isolate's compilation service. The crash occurs as
KernelIsolate::InitializeState had not been called during initialization
as Dart_InitializeParams.start_kernel_isolate == false.
Fixes b/155649002
Change-Id: I9b1225f17dc4d4dff39503d4353b3bb539b08267
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149880
Reviewed-by: Alexander Aprelev <aam@google.com>
Commit-Queue: Ben Konyi <bkonyi@google.com>
We already have tests that exercise this, namely:
- ffi_2/regress_39044_test
- ffi_2/vmspecific_send_port_id_test
Though the bug only appears if C++ compiler decides to evaluate arguments
in a different order. This happens to be only the case on Windows.
We don't have Windows hardware with Android phones on our CI, which is
why this was not caught earlier.
We have manually confirmed (via building app on windows with this fix
and running on android) that the issue is fixed by this change.
Fixes https://github.com/flutter/flutter/issues/54948
Change-Id: I51109cf1062964a5fb77948cffb9e2ba8fe2055f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149980
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
When system tracing is enabled on the Fuchsia platform, avoid calling
|OS::GetCurrentThreadCPUMicros|. The thread timestamp values are
discarded later, so querying them is wasteful w.r.t. both time and
trace buffer space.
Change-Id: I400075c89e69ddde6b032808859826c95f029008
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149395
Reviewed-by: Siva Annamalai <asiva@google.com>
Auto-Submit: Nathan Rogers <nathanrogers@google.com>
Commit-Queue: Siva Annamalai <asiva@google.com>
When replacing an unboxed phi with its contant value we should not
try to insert unboxing instruction before the phi - as phis exist
outside of the normal stream of instructions and InsertBefore does
not work on them.
Instead we should insert unboxing into the block which contains
the phi.
Original test is expanded to cover this case as well now.
Fixes https://github.com/dart-lang/sdk/issues/42164
TEST=vm/cc/ConstantPropagator_Regress35371
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-release-x64-try,vm-kernel-precomp-linux-debug-x64-try
Fixed: 42164
Change-Id: I53f984216e53195206fb8247b0ed8999590415a3
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149842
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Daco Harkes <dacoharkes@google.com>
This change actually migrates vm/dart/* tests to NNBD by fixing
compile-time errors and adjusting tests where needed.
vm/dart/null_float32x4_simd_ops_test and vm/dart/null_float64x2_simd_ops_test
are deleted as they are superseded by static type checks.
vm/dart/regress_40462_test.dart is a huge source code auto-generated
by fuzzer and migrating and maintaining this source doesn't have much
value, so it is deleted.
There are still failures in strong mode due to dependencies on
the packages which are not migrated yet (https://github.com/dart-lang/sdk/issues/42146).
Migrated vm/dart/null_checks_with_dwarf_stack_traces_test fails both in
weak and strong mode due to https://github.com/dart-lang/sdk/issues/42149.
Issue: https://github.com/dart-lang/sdk/issues/41314
Change-Id: I5561f1c8705ec16def0c4e0efa495d15f4ea7259
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149493
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Régis Crelier <regis@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>