Code in V8 embeds memory addresses. Upon deserialization, those addresses will
be updated anyways, so it's not important whether the serializer records those
addresses. In order to have a deterministic build, we create a copy of the code
and null out memory addresses and serialize this copy.
For the code cache, we do not care about determinism. By avoiding the copy we
can save some time spent on serialization. For the mandreel script in Octane,
this reduces the serialization time from ~85ms to ~75ms.
R=vogelheim@chromium.org
Review-Url: https://codereview.chromium.org/1991033003
Cr-Commit-Position: refs/heads/master@{#36350}
Empty function names are allowed and are output as such, unnamed
functions or functions with no valid UTF-8 name are output as
"<WASM UNNAMED>", while the CallSite object returns null as the
function name.
R=titzer@chromium.org, yangguo@chromium.org
Review-Url: https://codereview.chromium.org/1970503004
Cr-Commit-Position: refs/heads/master@{#36348}
The default value of the flag was true, so --intl_extra was not previously
unshipped. This patch sets it to false.
BUG=v8:3785
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_noi18n_rel_ng
Review-Url: https://codereview.chromium.org/1990763003
Cr-Commit-Position: refs/heads/master@{#36346}
port 39738bc905 (r36310)
original commit message:
In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.
This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.
BUG=
Review-Url: https://codereview.chromium.org/1987053006
Cr-Commit-Position: refs/heads/master@{#36344}
Rolling v8/build to b2d15686436cdc17f67c3621c314f8d96b5b6fd9
Rolling v8/tools/clang to 996bab489f816e51dde704bd215fb3403919f07e
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review-Url: https://codereview.chromium.org/1992113002
Cr-Commit-Position: refs/heads/master@{#36343}
This patch re-lands #36193 with a couple of minor tweaks
including updating the BUILD.gn file to use the new
variable name and incorporating thakis' request to support
cross-compiling the win snapshot from a mac.
R=thakis@chromium.org, jochen@chromium.org, machenbach@chromium.org
BUG=608596, 595653
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:chromeos_daisy_chromium_compile_only_ng
Review-Url: https://codereview.chromium.org/1979883002
Cr-Commit-Position: refs/heads/master@{#36342}
Instead of directly looking up Uint8Array.__proto__, to get a hold of
the TypedArray prototype, use object_get_prototype_of. This is a
functionally equivalent, but cleaner approach.
Review-Url: https://codereview.chromium.org/1990983002
Cr-Commit-Position: refs/heads/master@{#36340}
port 39738bc905
Original Commit Message:
In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.
This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.
R=rmcilroy@chromium.org, joransiu@ca.ibm.com, michael_dawson@ca.ibm.com, mbrandy@us.ibm.com
BUG=v8:4280
LOG=N
Review-Url: https://codereview.chromium.org/1989983002
Cr-Commit-Position: refs/heads/master@{#36336}
Properly records entries individually instead of using the hammer that requires
visiting the whole code object during pointer updating.
BUG=chromium:612847
LOG=N
Review-Url: https://codereview.chromium.org/1994653002
Cr-Commit-Position: refs/heads/master@{#36334}
Remove kMakeHeapIterableMask since the heap is always iterable.
BUG=chromium:580959
LOG=n
Review-Url: https://codereview.chromium.org/1961373003
Cr-Commit-Position: refs/heads/master@{#36333}
This patch ensures that Intl bound method getters can't be retargeted
to other Intl types. If that were to happen, then a RUNTIME_ASSERT
would trigger later. This patch throws a TypeError instead, as the
specification requires.
BUG=v8:4870
Review-Url: https://codereview.chromium.org/1986763003
Cr-Commit-Position: refs/heads/master@{#36330}
This is a pure refactoring that removes unnecessary wrapper methods from
the EscapeAnalysis class. This completely removes the notion of "Alias"
from the interface.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/1989243002
Cr-Commit-Position: refs/heads/master@{#36325}
This promotes the escape analysis from an experimental feature to be a
fully supported feature. The main goal is to unleach ClusterFuzz on the
implementation so that we can stabilize it.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/1989833002
Cr-Commit-Position: refs/heads/master@{#36324}
This is pure refactoring that moves many internal details of the escape
analysis from the header into the compilation unit. The goal of this is
to simplify the interface for readability.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/1991723003
Cr-Commit-Position: refs/heads/master@{#36322}
Currently attached references are only used for global proxy, source
string and code stubs. Mid-term future we want to use attached
references for arbitrary objects (in fixed order) provided from outside.
This change renames BackReference to SerializerReference to include both
back references and attached references.
R=mtrofin@chromium.org, vogelheim@chromium.org
Review-Url: https://codereview.chromium.org/1992723002
Cr-Commit-Position: refs/heads/master@{#36318}
We have evidence (see r34896) that this avoids crashes.
BUG=chromium:524337
Review-Url: https://codereview.chromium.org/1995483002
Cr-Commit-Position: refs/heads/master@{#36317}
When setting a generator function's "prototype" property to a non-object, the
prototype of new generator instances should be %GeneratorPrototype%, not
%ObjectPrototype%.
R=verwaest@chromium.org
BUG=v8:5011
Review-Url: https://codereview.chromium.org/1982203003
Cr-Commit-Position: refs/heads/master@{#36313}
In order to support compiling to baseline on return we need to be able to
return to the actual return address. With this change this is what the
Return bytecode now does, removing the need for the
InterpreterExitTrampoline.
This change also removes the InterpreterNotifyDeoptXXX builtins and
unifies FCG and Igntion to both use NotifyDeoptXXX. As part of this
change, FullCodegenerator::State is moved to Deoptimize::BailoutState.
BUG=v8:4280
LOG=N
Committed: https://crrev.com/34c9626e2ee56fe805de549697ca5323aed7cb66
Cr-Commit-Position: refs/heads/master@{#36288}
Review-Url: https://codereview.chromium.org/1969423002
Cr-Commit-Position: refs/heads/master@{#36310}
Remove dead code to optimize Int64Constants as branch/select conditions,
because we either have tagged booleans or bits represented as word32.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/1994533002
Cr-Commit-Position: refs/heads/master@{#36308}
One of the Arm OutOfLineRecordWrite was missing initialiation of must_save_lr
Review-Url: https://codereview.chromium.org/1988623003
Cr-Commit-Position: refs/heads/master@{#36307}
This moves collection of function information from its previous spot in
the standard compiler pipeline (GetSharedFunctionInfo() and
CompileTopLevel()) to its new location in CompileForLiveEdit. Nesting
information is reconstructed by traversing the AST.
R=yangguo@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/1971683002
Cr-Commit-Position: refs/heads/master@{#36306}
The type guard should never be used after the effect/control
linearization pass, so making it a simplified operator better
expresses the intended use. Also this way none of the common
operators actually has any dependency on the type system.
Drive-by-fix: Properly print the type parameter to a TypeGuard operator.
BUG=chromium:612142
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/1994503002
Cr-Commit-Position: refs/heads/master@{#36304}
port 551e0aa11b (r36275)
original commit message:
This adds back the instanceof operator support in the backends and
introduces a @@hasInstance protector cell on the isolate that guards the
fast path for the InstanceOfStub. This way we recover the ~10%
regression on Octane EarleyBoyer in Crankshaft and greatly improve
TurboFan and Ignition performance of instanceof.
BUG=
Review-Url: https://codereview.chromium.org/1991663002
Cr-Commit-Position: refs/heads/master@{#36303}
These guards are useless anyways once you make it throw the
effect/control linearizer because all memory operations and
calls are connected to the control and/or effect chain anyways
afterwards.
Drive-by-fix: Fail in the InstructionSelector if we ever see
a Guard node.
R=jarin@chromium.org
BUG=chromium:612142
Review-Url: https://codereview.chromium.org/1980383002
Cr-Commit-Position: refs/heads/master@{#36302}
This operator will be used to carry eager frame states, and is hooked up
to the effect chain with a control dependency to avoid hosting out of
control structures.
R=jarin@chromium.org
BUG=v8:5021
Review-Url: https://codereview.chromium.org/1993593002
Cr-Commit-Position: refs/heads/master@{#36301}
