This is a first step towards moving Turbofan code generation off the main thread.
Summary of the changes:
- AssemblerBase no longer has a pointer to the isolate. Instead, its
constructor receives the few things that it needs from the isolate (on most
architectures this is just the serializer_enabled flag).
- RelocInfo no longer has a pointer to the isolate. Instead, the functions
that need it take it as an argument. (There are currently still a few that
implicitly access the isolate through a HeapObject.)
- The MacroAssembler now explicitly holds a pointer to the isolate (before, it
used to get it from the Assembler).
- The jit_cookie also moved from AssemblerBase to the MacroAssemblers, since
it's not used at all in the Assemblers.
- A few architectures implemented parts of the Assembler with the help
of a Codepatcher that is based on MacroAssembler. Since the Assembler no
longer has the isolate, but the MacroAssembler still needs it, this doesn't
work anymore. Instead, these Assemblers now use a new PatchingAssembler.
BUG=v8:6048
Review-Url: https://codereview.chromium.org/2732273003
Cr-Commit-Position: refs/heads/master@{#43890}
From asm.js code we might get an empty ArrayBuffer as heap memory. In
this case, both the old memory start and the new memory start will be
nullptr. The size however has to be patched from default_size to 0.
This CL changes code specialization to be able to either patch memory
references, or patch memory sizes or both.
R=titzer@chromium.org, ahaas@chromium.org
BUG=chromium:698587
Change-Id: I4d9d811d75cb83842f23df317e8e7fc02aeb5146
Reviewed-on: https://chromium-review.googlesource.com/450257
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43613}
Turbofan is a lot slower than Crankshaft at constructing TypedArrays,
because we always go to the C++ builtin. Port the builtin to CSA
to improve performance, and to clean up the implementation, which is
split across multiple files and pieces at the moment.
This CL increases the performance with --future to roughly the same
as with crankshaft.
BUG=v8:5977
Change-Id: Id0d91a4592de41a3a308846d79bd44a608931762
Reviewed-on: https://chromium-review.googlesource.com/448537
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43548}
This reverts commit b23b2c107b.
Reason for revert: Makes Linux debug bot sad
Original change's description:
> [builtins] Port TypedArrayInitialize to CodeStubAssembler.
>
> Turbofan is a lot slower than Crankshaft at constructing TypedArrays,
> because we always go to the C++ builtin. Port the builtin to CSA
> to improve performance, and to clean up the implementation, which is
> split across multiple files and pieces at the moment.
>
> This CL increases the performance with --future to roughly the same
> as with crankshaft.
>
> BUG=v8:5977
>
> Change-Id: I5a4c4b544a735a56290b85bf33c2f3718df7e2b8
> Reviewed-on: https://chromium-review.googlesource.com/445717
> Commit-Queue: Peter Marshall <petermarshall@chromium.org>
> Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
> Reviewed-by: Camillo Bruni <cbruni@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#43518}
TBR=cbruni@chromium.org,petermarshall@chromium.org,bmeurer@chromium.org,v8-reviews@googlegroups.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5977
Change-Id: I5d5bc8b4677a405c716d78e688af80ae9c737b4a
Reviewed-on: https://chromium-review.googlesource.com/448558
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43520}
Turbofan is a lot slower than Crankshaft at constructing TypedArrays,
because we always go to the C++ builtin. Port the builtin to CSA
to improve performance, and to clean up the implementation, which is
split across multiple files and pieces at the moment.
This CL increases the performance with --future to roughly the same
as with crankshaft.
BUG=v8:5977
Change-Id: I5a4c4b544a735a56290b85bf33c2f3718df7e2b8
Reviewed-on: https://chromium-review.googlesource.com/445717
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43518}
Previously, when restarting a frame, we would rewrite all frames
between the debugger activation and the frame to restart to squash
them, and replace the return address with that of a builtin to
leave that rewritten frame, and restart the function by calling it.
We now simply remember the frame to drop to, and upon returning
from the debugger, we check whether to drop the frame, load the
new FP, and restart the function.
R=jgruber@chromium.org, mstarzinger@chromium.org
BUG=v8:5587
Review-Url: https://codereview.chromium.org/2636913002
Cr-Commit-Position: refs/heads/master@{#42725}
- kDebugPromiseCreated(task, parent_task)
This event occurs when promise is created (PromiseHookType::Init). V8Debugger uses this event to maintain task -> parent task map.
- kDebugEnqueueAsyncFunction(task)
This event occurs when first internal promise for async function is created. V8Debugger collects stack trace at this point.
- kDebugEnqueuePromiseResolve(task),
This event occurs when Promise fulfills with resolved status. V8Debugger collects stack trace at this point.
- kDebugEnqueuePromiseReject(task),
This event occurs when Promise fulfills with rejected status. V8Debugger collects stack trace at this point.
- kDebugPromiseCollected,
This event occurs when Promise is collected and no other chained callbacks can be added. V8Debugger removes information about async task for this promise.
- kDebugWillHandle,
This event occurs when chained promise function (either resolve or reject handler) is called. V8Debugger installs parent promise's stack (based on task -> parent_task map) as current if available or current promise's scheduled stack otherwise.
- kDebugDidHandle,
This event occurs after chained promise function has finished. V8Debugger restores asynchronous call chain to previous one.
With this change all instrumentation calls are related to current promise (before WillHandle and DidHandle were related to next async task).
Before V8Debugger supported only the following:
- asyncTaskScheduled(task1)
- asyncTaskStarted(task1)
- asyncTaskFinished(task1)
Now V8Debugger supports the following:
- asyncTaskScheduled(parent_task)
..
- asyncTaskCreated(task, parent_task),
- asyncTaskStarted(task), uses parent_task scheduled stack
- asyncTaskScheduled(task)
- asyncTaskFinished(task)
Additionally: WillHandle and DidHandle were migrated to PromiseHook API.
More details: https://docs.google.com/document/d/1u19N45f1gSF7M39mGsycJEK3IPyJgIXCBnWyiPeuJFE
BUG=v8:5738
R=dgozman@chromium.org,gsathya@chromium.org,yangguo@chromium.org
Review-Url: https://codereview.chromium.org/2650803003
Cr-Commit-Position: refs/heads/master@{#42644}
E.g., ast/ast.h uses Label but shouldn't need to include assembler.h for that. With
this change, we can hope for proper layering in the future (not quite there
yet).
Also includes minor random include lowering and relevant IWYU fixes.
BUG=v8:5294
Review-Url: https://codereview.chromium.org/2645063002
Cr-Commit-Position: refs/heads/master@{#42563}
- Refactor Dispatch tables to have separate function, signature tables
- New Relocation type for WasmFunctionTableReference, assembler, compiler support.
- RelocInfo helper functions for Wasm references
Review-Url: https://codereview.chromium.org/2627543003
Cr-Commit-Position: refs/heads/master@{#42192}
This adds kInit, kResolve, kBefore and kAfter lifecycle hooks to promises.
This also exposes an API to set the PromiseHook.
BUG=v8:4643
Review-Url: https://codereview.chromium.org/2575313002
Cr-Commit-Position: refs/heads/master@{#41775}
Many websites use simple calls to String.prototype.indexOf with either a
one character ASCII needle or needles bigger than the search string. This
CL adds a TFJ builtin for these simple cases, giving up to factor 5 speedup.
Drive-by-fix: Add default Object type to Arguments.at
BUG=
Review-Url: https://codereview.chromium.org/2539093002
Cr-Commit-Position: refs/heads/master@{#41760}
Some instructions in WebAssembly trap for some inputs, which means that the
execution is terminated and (at least at the moment) a JavaScript exception is
thrown. Examples for traps are out-of-bounds memory accesses, or integer
divisions by zero.
Without the TrapIf and TrapUnless operators trap check in WebAssembly introduces 5
TurboFan nodes (branch, if_true, if_false, trap-reason constant, trap-position
constant), in addition to the trap condition itself. Additionally, each
WebAssembly function has four TurboFan nodes (merge, effect_phi, 2 phis) whose
number of inputs is linear to the number of trap checks in the function.
Especially for functions with high numbers of trap checks we observe a
significant slowdown in compilation time, down to 0.22 MiB/s in the sqlite
benchmark instead of the average of 3 MiB/s in other benchmarks. By introducing
a TrapIf common operator only a single node is necessary per trap check, in
addition to the trap condition. Also the nodes which are shared between trap
checks (merge, effect_phi, 2 phis) would disappear. First measurements suggest a
speedup of 30-50% on average.
This CL only implements TrapIf and TrapUnless on x64. The implementation is also
hidden behind the --wasm-trap-if flag.
Please take a special look at how the source position is transfered from the
instruction selector to the code generator, and at the context that is used for
the runtime call.
R=titzer@chromium.org
Review-Url: https://codereview.chromium.org/2562393002
Cr-Commit-Position: refs/heads/master@{#41720}
This will be used in CSA to check if any promisehook is set.
-- Adds a is_promisehook_enabled_ field to the isolate and helper methods.
-- Adds this field to the ExternalReference table.
-- Adds a helper method to access this from CSA
Note -- this patch doesn't actually add the ability to attach the hook
yet.
BUG=v8:4643
Review-Url: https://codereview.chromium.org/2566483002
Cr-Commit-Position: refs/heads/master@{#41607}
SourcePosition::InliningId() refers to a the new table DeoptimizationInputData::InliningPositions(), which provides the following data for every inlining id:
- The inlined SharedFunctionInfo as an offset into DeoptimizationInfo::LiteralArray
- The SourcePosition of the inlining. Recursively, this yields the full inlining stack.
Before the Code object is created, the same information can be found in CompilationInfo::inlined_functions().
If SourcePosition::InliningId() is SourcePosition::kNotInlined, it refers to the outer (non-inlined) function.
So every SourcePosition has full information about its inlining stack, as long as the corresponding Code object is known. The internal represenation of a source position is a positive 64bit integer.
All compilers create now appropriate source positions for inlined functions. In the case of Turbofan, this required using AstGraphBuilderWithPositions for inlined functions too. So this class is now moved to a header file.
At the moment, the additional information in source positions is only used in --trace-deopt and --code-comments. The profiler needs to be updated, at the moment it gets the correct script offsets from the deopt info, but the wrong script id from the reconstructed deopt stack, which can lead to wrong outputs. This should be resolved by making the profiler use the new inlining information for deopts.
I activated the inlined deoptimization tests in test-cpu-profiler.cc for Turbofan, changing them to a case where the deopt stack and the inlining position agree. It is currently still broken for other cases.
The following additional changes were necessary:
- The source position table (internal::SourcePositionTableBuilder etc.) supports now 64bit source positions. Encoding source positions in a single 64bit int together with the difference encoding in the source position table results in very little overhead for the inlining id, since only 12% of the source positions in Octane have a changed inlining id.
- The class HPositionInfo was effectively dead code and is now removed.
- SourcePosition has new printing and information facilities, including computing a full inlining stack.
- I had to rename compiler/source-position.{h,cc} to compiler/compiler-source-position-table.{h,cc} to avoid clashes with the new src/source-position.cc file.
- I wrote the new wrapper PodArray for ByteArray. It is a template working with any POD-type. This is used in DeoptimizationInputData::InliningPositions().
- I removed HInlinedFunctionInfo and HGraph::inlined_function_infos, because they were only used for the now obsolete Crankshaft inlining ids.
- Crankshaft managed a list of inlined functions in Lithium: LChunk::inlined_functions. This is an analog structure to CompilationInfo::inlined_functions. So I removed LChunk::inlined_functions and made Crankshaft use CompilationInfo::inlined_functions instead, because this was necessary to register the offsets into the literal array in a uniform way. This is a safe change because LChunk::inlined_functions has no other uses and the functions in CompilationInfo::inlined_functions have a strictly longer lifespan, being created earlier (in Hydrogen already).
BUG=v8:5432
Review-Url: https://codereview.chromium.org/2451853002
Cr-Commit-Position: refs/heads/master@{#40975}
CpuFeatures::IsSupported(feature) indicates that the feature is
available on the target. AssemblerBase::IsEnabled(feature) indicates
that we've checked for support (using CpuFeatureScope). The main benefit
is that we can test on (for example) ARMv8, but have some assurance that
we won't generate ARMv8 instructions on ARMv7 targets.
This patch simply cleans up the usage, which had become inconsistent.
The instruction emission functions now check not only that their
dependent features are supported, but also that we've verified that
using CpuFeatureScope.
BUG=
Review-Url: https://codereview.chromium.org/2360243002
Cr-Commit-Position: refs/heads/master@{#39676}
According to new store IC calling convention the value, slot and vector are passed
on the stack and there's no need in trying to preserve values or respective registers
in store handlers.
Nice bonus: we also don't need virtual registers anymore.
BUG=v8:5407
Review-Url: https://codereview.chromium.org/2357323003
Cr-Commit-Position: refs/heads/master@{#39672}
This CL more or less reverts commit https://codereview.chromium.org/2107733002/
The use of the MathPow code stub that was introduced by that commit caused
problems on arm64, and the MathPow code stub was also an obstacle in the
implementation of parallel code generation.
In addition this CL turns on the mjsunit/wasm/embenchen tests for arm64
which were turned off because of problems with MathPow on arm64.
R=titzer@chromium.org, bradnelson@chromium.org
Review-Url: https://codereview.chromium.org/2166793002
Cr-Commit-Position: refs/heads/master@{#37911}
So far TurboFan wasn't adding the deoptimization reasons for eager/soft
deoptimization exits that can be used by either the DevTools profiler or
the --trace-deopt flag. This adds basic support for deopt reasons on
Deoptimize, DeoptimizeIf and DeoptimizeUnless nodes and threads through
the reasons to the code generation.
Also moves the DeoptReason to it's own file (to resolve include cycles)
and drops unused reasons.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/2161543002
Cr-Commit-Position: refs/heads/master@{#37823}
Import fdlibm versions of acos, acosh, asin and asinh, which are more
precise and produce the same result across platforms (we were using
libm versions for asin and acos so far, where both speed and precision
depended on the operating system so far). Introduce appropriate TurboFan
operators for these functions and use them both for inlining and for the
generic builtin.
Also migrate the Math.imul and Math.fround builtins to TurboFan builtins
to ensure that their behavior is always exactly the same as the inlined
TurboFan version (i.e. C++ truncation semantics for double to float
don't necessarily meet the JavaScript semantics).
For completeness, also migrate Math.sign, which can even get some nice
love in TurboFan.
Drive-by-fix: Some alpha-sorting on the Math related functions, and
cleanup the list of Math intrinsics that we have to export via the
native context currently.
BUG=v8:3266,v8:3496,v8:3509,v8:3952,v8:5169,v8:5170,v8:5171,v8:5172
TBR=rossberg@chromium.orgR=franzih@chromium.org
Review-Url: https://codereview.chromium.org/2116753002
Cr-Commit-Position: refs/heads/master@{#37476}
Port c1d01aea11
Fix a few failures which occur because instruction cache
hasn't been flushed after update of WASM references.
BUG=mjsunit/wasm/asm-wasm-heap,mjsunit/wasm/start-function
Review-Url: https://codereview.chromium.org/2103093003
Cr-Commit-Position: refs/heads/master@{#37364}
Introduce a new machine operator Float64Pow that for now is backed by
the existing MathPowStub to start the unification of Math.pow, and at
the same time address the main performance issue that TurboFan still has
with the imaging-darkroom benchmark in Kraken.
Also migrate the Math.pow builtin itself to a TurboFan builtin and
remove a few hundred lines of hand-written platform code for special
handling of the fullcodegen Math.pow version.
BUG=v8:3599,v8:5086,v8:5157
Review-Url: https://codereview.chromium.org/2103733003
Cr-Commit-Position: refs/heads/master@{#37323}
Port fc59eb8a7a
Original commit message:
Moves between operands with different representations shouldn't happen,
so don't test them. This makes it easier to modify canonicalization to
differentiate between floating point types, which is needed to support
floating point register aliasing for ARM and MIPS.
This change also expands tests to include explicit FP moves (both register and stack slot).
LOG=N
BUG=v8:4124
BUG=chromium:622619
Review-Url: https://codereview.chromium.org/2090993002
Cr-Commit-Position: refs/heads/master@{#37241}
Previously we would elide debug slots if the statement position it
belongs to has just already been written. The motivation is that since
we should only break once per statement, we can elide debug slots that
has the same statement position as the previous debug slot.
This is an unnecessary optimization, since the debugger has yet another
check against breaking twice at the same statement at runtime, in
Debug::Break.
This optimization can also be wrong, if there is control flow involved,
for example if we can jump to the elided debug slot without executing
the previous debug slot.
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
R=jgruber@chromium.org
Review-Url: https://codereview.chromium.org/2080173002
Cr-Commit-Position: refs/heads/master@{#37107}
Import base::ieee754::tan() from fdlibm and introduce Float64Tan TurboFan
operator based on that, similar to what we do for Float64Cos and Float64Sin.
Rewrite Math.tan() as TurboFan builtin and use those operators to also
inline Math.tan() into optimized TurboFan functions.
Drive-by-fix: Kill the %_ConstructDouble intrinsics, and provide only
the %ConstructDouble runtime entry for writing tests.
BUG=v8:5086,v8:5126
R=yangguo@chromium.org
Review-Url: https://codereview.chromium.org/2083453002
Cr-Commit-Position: refs/heads/master@{#37087}
Compilation of wasm functions happens before instantiation. Imports are linked afterwards, at instantiation time. Globals and memory are also
allocated and then tied in via relocation at instantiation time.
This paves the way for implementing Wasm.compile, a prerequisite to
offering the compiled code serialization feature.
Currently, the WasmModule::Compile method just returns a fixed array
containing the code objects. More appropriate modeling of the compiled module to come.
Opportunistically centralized the logic on how to update memory
references, size, and globals, since that logic is the exact same on each
architecture, except for the actual storing of values back in the
instruction stream.
BUG=v8:5072
Review-Url: https://codereview.chromium.org/2056633002
Cr-Commit-Position: refs/heads/master@{#37086}
Import base::ieee754::cos() and base::ieee754::sin() from fdlibm and
introduce Float64Cos and Float64Sin TurboFan operator based on that,
similar to what we do for Float64Log. Rewrite Math.cos() and Math.sin()
as TurboFan builtins and use those operators to also inline Math.cos()
and Math.sin() into optimized TurboFan functions.
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
R=mvstanton@chromium.org
BUG=v8:5086,v8:5118
Review-Url: https://codereview.chromium.org/2073123002
Cr-Commit-Position: refs/heads/master@{#37072}
Import base::ieee754::exp() from FreeBSD msun and introduce a Float64Exp
TurboFan operator based on that, similar to what we do for Float64Log.
Rewrite Math.exp() as TurboFan builtin and use that operator to also
inline Math.exp() into optimized TurboFan functions.
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
BUG=v8:3266,v8:3468,v8:3493,v8:5086,v8:5108,chromium:620786
R=mvstanton@chromium.org
Committed: https://crrev.com/93e26314afc9da9b5b8bd998688262444ed73260
Review-Url: https://codereview.chromium.org/2077533002
Cr-Original-Commit-Position: refs/heads/master@{#37037}
Cr-Commit-Position: refs/heads/master@{#37047}
Reason for revert:
[Sheriff] Leads to some different rounding as it seems in some audio layout tests. Please rebase upstream first if intended:
https://build.chromium.org/p/client.v8.fyi/builders/V8-Blink%20Linux%2064/builds/7508
Original issue's description:
> [builtins] Introduce proper Float64Exp operator.
>
> Import base::ieee754::exp() from FreeBSD msun and introduce a Float64Exp
> TurboFan operator based on that, similar to what we do for Float64Log.
> Rewrite Math.exp() as TurboFan builtin and use that operator to also
> inline Math.exp() into optimized TurboFan functions.
>
> BUG=v8:3266,v8:3468,v8:3493,v8:5086,v8:5108
> R=mvstanton@chromium.org
>
> Committed: https://crrev.com/93e26314afc9da9b5b8bd998688262444ed73260
> Cr-Commit-Position: refs/heads/master@{#37037}
TBR=mvstanton@chromium.org,ahaas@chromium.org,bmeurer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:3266,v8:3468,v8:3493,v8:5086,v8:5108
Review-Url: https://codereview.chromium.org/2070813002
Cr-Commit-Position: refs/heads/master@{#37039}
