With the IC-less global variables accesses the size of the Rotate* functions became small enough to be immediately optimized by Crankshaft which in turn tanked SunSpider/3d-cube.
BUG=chromium:531338
LOG=N
Review URL: https://codereview.chromium.org/1356603005
Cr-Commit-Position: refs/heads/master@{#30861}
Slow path for relational comparison of boolean primitive values
now goes through the runtime, which made the slow path even
slower than it already was. So in order to repair the regression,
we just track boolean feedback for comparisons and use that
to generate decent code in Crankshaft (not the best possible
code, but good enough for Crankshaft; TurboFan will be able
to do better on that).
R=jarin@chromium.org
BUG=chromium:534200
LOG=n
Review URL: https://codereview.chromium.org/1347063004
Cr-Commit-Position: refs/heads/master@{#30860}
Now the StringAddStub can optionally convert it's parameters to strings
(following the rules for the addition operator). This could be further
optimized once we have a ToPrimitiveStub, but it should be sufficient
for the moment.
Also removed the unused Strength parameter to the HStringAdd operator,
because string addition does not depend on language mode.
R=jarin@chromium.org
BUG=v8:4307, chromium:532524
LOG=n
Committed: https://crrev.com/d261849e53fbf8c36efae42d478271f87acff70f
Cr-Commit-Position: refs/heads/master@{#30726}
Review URL: https://codereview.chromium.org/1339053002
Cr-Commit-Position: refs/heads/master@{#30858}
Introduce new builtins Construct and ConstructFunction (in line
with the Call and CallFunction builtins that we already have) as
proper bottleneck for Construct and [[Construct]] on JSFunctions.
Use these builtins to support passing NewTarget from C++ to
JavaScript land.
Long-term we want the CallConstructStub to be used for
gathering feedback on entry to construction chain (i.e. the
initial new Foo), and use the Construct builtins to do the
actual work inside the construction chain (i.e. calling into
super and stuff).
MIPS and MIPS64 ports contributed by akos.palfi@imgtec.com.
R=jarin@chromium.org
BUG=v8:4430
LOG=n
Review URL: https://codereview.chromium.org/1359583002
Cr-Commit-Position: refs/heads/master@{#30857}
If @@toStringTag is an accessor property, we cannot assume that the result
of calling Object.prototype.toString() for objects with the same map.
R=adamk@chromium.org
BUG=chromium:534200
LOG=n
Review URL: https://codereview.chromium.org/1360723002
Cr-Commit-Position: refs/heads/master@{#30856}
Rolling v8/build/gyp to 01528c7244837168a1c80f06ff60fa5a9793c824
Rolling v8/tools/clang to 3e04436a49a26f4bd2c6f352efcf4c7b10a6d07d
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review URL: https://codereview.chromium.org/1360563003
Cr-Commit-Position: refs/heads/master@{#30855}
We need to build parts of v8 with a toolchain that might be different
from both the default (target) toolchain and the regular host toolchain,
because we need the snapshot to have the same bit-width as the target.
V8's build defines a 'snapshot_toolchain' setting for this.
It turns out that we need the value of this toolchain to be exposed
to the Chromium build because some of the test targets (in browser_tests)
depend on d8 and need to be able to built using the same toolchain.
R=brett@chromium.org, jochen@chromium.org
BUG=
Review URL: https://codereview.chromium.org/1350223004
Cr-Commit-Position: refs/heads/master@{#30854}
Previously we only collected the known map for equality comparisons. But
if we also collect it for relational comparisons, we can inline a fast
path of ToPrimitive on the objects, which is especially interesting
since both sides have the same map.
For now we only inline a very limited subset of ToPrimitive in
Crankshaft, which is when the receiver map (and its prototype chain)
doesn't have @@toPrimitive, and both valueOf and toString are the
default versions on the %ObjectPrototype%. In this case the relational
comparison would reduce to a string comparison of "[object CLASS]" with
itself and so we can reduce that to a boolean constant plus map checks
on both left and right hand side, plus code dependencies on the
prototype chain. This repairs the regression on box2d.
R=jkummerow@chromium.org
BUG=chromium:534200
LOG=n
Review URL: https://codereview.chromium.org/1355113002
Cr-Commit-Position: refs/heads/master@{#30852}
This doesn't fix the performance regression mentioned by the bug yet,
but is necessary cleanup to land the fix, and should be separated from
the actual fix.
R=jkummerow@chromium.org
BUG=chromium:534200
LOG=n
Review URL: https://codereview.chromium.org/1345313005
Cr-Commit-Position: refs/heads/master@{#30847}
port 8016547c8e (r30818).
original commit message:
The StringCompareStub used to take its parameters on the (JavaScript)
stack, which made it impossible to use in TurboFan. Actually
StringCompareStub was currently completely unused. This changes the
calling convention to something TurboFan compatible and introduces a
CallInterfaceDescriptor for StringCompareStub. It also changes
HStringCompareAndBranch to use the StringCompareStub instead of using
the full blown CompareICStub for a stupid string comparison.
BUG=
Review URL: https://codereview.chromium.org/1355983003
Cr-Commit-Position: refs/heads/master@{#30845}
port 593c655a3c (r30816).
original commit message:
This removes the weird COMPARE and COMPARE_STRONG JavaScript builtins
and replaces them with a proper C++ implementation in Object::Compare
and appropriate wrappers Object::LessThan, Object::GreaterThan, and
friends that are intended to be used by a true/false returning CompareIC
in the future, as well as the interpreter. As a short-term solution we
provide %Compare and %Compare_Strong entry points for the current
CompareIC that return the appropriate integer values expected by
fullcodegen currently.
Now the Abstract Relational Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
BUG=
Review URL: https://codereview.chromium.org/1353343002
Cr-Commit-Position: refs/heads/master@{#30844}
mdb_v8, a post-mortem debugging tool for Node.js, allows users to
inspect ScopeInfo structures in order to get more information about
closures.
Currently, it hardcodes the metadata it uses to find this information.
This change allows it to get this metadata from the node binary itself,
and thus to adapt to future changes made to the layout of the ScopeInfo
data structure.
BUG=
R=bmeurer@chromium.org
Review URL: https://codereview.chromium.org/1350843003
Cr-Commit-Position: refs/heads/master@{#30843}
ES2015 specifies very particular semantics for functions defined in blocks.
In strict mode, it is simply a lexical binding scoped to that block. In sloppy
mode, in addition to that lexical binding, there is a var-style binding in
the outer scope, which is overwritten with the local binding when the function
declaration is evaluated, *as long as* introducing ths var binding would not
create a var/let conflict in the outer scope.
This patch implements the semantics by introducing a DelegateStatement, which
is initially filled in with the EmptyStatement and overwritten with the
assignment when the scope is closed out and it can be checked that there is
no conflict.
This patch is tested with a new mjsunit test, and I tried staging it and running
test262, finding that the tests that we have disabled due to lack of Annex B
support now pass.
R=adamk,rossberg
LOG=Y
BUG=v8:4285
Review URL: https://codereview.chromium.org/1332873003
Cr-Commit-Position: refs/heads/master@{#30842}
Upon collection of the stack trace if the current PC falls into
the frame building code, the top frame might be in a non-consistent
state. That leads to some of the frames could be missing from the
stack trace.
The patch makes it check instructions under current PC and if they
look like the frame setup/destroy code, it skips the entire sample.
Support for x86/x64
CG_INCLUDE_TRYBOTS=tryserver.v8:v8_linux64_msan_rel
BUG=chromium:529931
LOG=N
Review URL: https://codereview.chromium.org/1348533005
Cr-Commit-Position: refs/heads/master@{#30841}
Rolling v8/build/gyp to cf3170e30578d600b8ec8cd68553cc5e606d42eb
Rolling v8/tools/clang to 76e743dc622478312b66661ad48997b318628cbb
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review URL: https://codereview.chromium.org/1357793002
Cr-Commit-Position: refs/heads/master@{#30840}
Since https://codereview.chromium.org/272163002, BitField3 is a raw
uint32 field, and not a SMI anymore.
Update tools/gen-postmortem-metadata.py so that post-mortem tools can
work with versions of V8 that shipped after that change.
This change was merged in github.com/joyent/node right before node
v0.12.0 was released.
R=danno@chromium.org
TEST=mdb_v8, a post-mortem debugging tool running on SmartOS, has been
using this change since Node.js v0.12.0 was released
BUG=
Review URL: https://codereview.chromium.org/1296743003
Cr-Commit-Position: refs/heads/master@{#30839}
Port 8016547c8e
Original commit message:
The StringCompareStub used to take its parameters on the (JavaScript)
stack, which made it impossible to use in TurboFan. Actually
StringCompareStub was currently completely unused. This changes the
calling convention to something TurboFan compatible and introduces a
CallInterfaceDescriptor for StringCompareStub. It also changes
HStringCompareAndBranch to use the StringCompareStub instead of using
the full blown CompareICStub for a stupid string comparison.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=
Review URL: https://codereview.chromium.org/1358553002
Cr-Commit-Position: refs/heads/master@{#30837}
Port 593c655a3c
Original commit message:
This removes the weird COMPARE and COMPARE_STRONG JavaScript builtins
and replaces them with a proper C++ implementation in Object::Compare
and appropriate wrappers Object::LessThan, Object::GreaterThan, and
friends that are intended to be used by a true/false returning CompareIC
in the future, as well as the interpreter. As a short-term solution we
provide %Compare and %Compare_Strong entry points for the current
CompareIC that return the appropriate integer values expected by
fullcodegen currently.
Now the Abstract Relational Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/1356983002
Cr-Commit-Position: refs/heads/master@{#30836}
It's been enabled since M45, which is now well into its stable period,
with no problems reported.
Review URL: https://codereview.chromium.org/1356793002
Cr-Commit-Position: refs/heads/master@{#30835}
This changes the error message for code like:
if (false) let x;
from "Block-scoped declarations (let, const, function, class) not yet supported outside strict mode"
to "Unexpected identifier" (pointing at |x|).
Review URL: https://codereview.chromium.org/1356783002
Cr-Commit-Position: refs/heads/master@{#30834}
Akin to linear scan's TryReuseSpillForPhi, we attempt to merge the
spill ranges of grouped live ranges (which are phi inputs and output),
to avoid inefficient slot-to-slot moves.
BUG=
Review URL: https://codereview.chromium.org/1353023003
Cr-Commit-Position: refs/heads/master@{#30833}
This way we can finally remove the ES5 ToPrimitive builtin from
runtime.js, and the Date Constructor now properly supports
@@toPrimitive for the single argument case as well.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_layout_dbg,v8_linux_nosnap_dbg
R=rossberg@chromium.org
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/1346893003
Cr-Commit-Position: refs/heads/master@{#30832}
port b5588f48fd (r30767).
original commit message:
There isn't a plan to turn it on soon, so we'll take it out in favor of cleaner code.
BUG=
Review URL: https://codereview.chromium.org/1346043005
Cr-Commit-Position: refs/heads/master@{#30829}
port d5bbd45f04 (r30808).
oringial commit message:
Currently Execution::Call (and friends) still duplicate a lot of the
Call sequence logic that should be encapsulated in the Call and
CallFunction builtins. So the plan now is to switch Execution::Call
to accept any Callable and just pass that through to the Call builtin.
BUG=
Review URL: https://codereview.chromium.org/1350183005
Cr-Commit-Position: refs/heads/master@{#30828}
port a3d6f6cce3 (r30759).
original commit message:
Implement the String constructor completely as native builtin,
avoiding the need to do gymnastics in JavaScript builtin to
properly detect the no argument case (which is different from
the undefined argument case) and also allowing to just
tailcall through to ToString or SymbolDescriptiveString for
the common case. Also the JavaScript builtin was misleading
since the case for construct call was unused, but could be
triggered in a wrong way once we support tail calls from
constructor functions.
This refactoring allows us to properly implement subclassing
for String builtins, once we have the correct initial_map on
derived classes (it's merely a matter of using NewTarget
instead of the target register now).
This introduces a new %SymbolDescriptiveString runtime
entry, which is also used by Symbol.toString() now.
BUG=
Review URL: https://codereview.chromium.org/1349403002
Cr-Commit-Position: refs/heads/master@{#30826}
Port 09f41681ef
Original commit message:
The patch decreases the calls of huge switch instructions making the
DecodeType*() functions to work in one phase and optimizing
Instruction::InstructionType(). Speed gain in release full check is
about 33% (6:13 s -> 4:09 s) and in optdebug full test is about 50%
(12:29 -> 6:17)
BUG=
Review URL: https://codereview.chromium.org/1356693002
Cr-Commit-Position: refs/heads/master@{#30824}
The previous code took advantage of the fact that a group's deps are implicitly public, but I'm trying to fix that. This also cleans up some duplicated code between component and non-component builds.
Review URL: https://codereview.chromium.org/1356723002
Cr-Commit-Position: refs/heads/master@{#30820}
This extends instrumentation added in r30683 and r30768 to cover
the possibility that the root cause we're after is in optimized code.
This CL is intended to be reverted in a couple of days, but should
cause no harm while it's in the tree (we would crash anyway).
BUG=chromium:527994
LOG=n
Review URL: https://codereview.chromium.org/1348823003
Cr-Commit-Position: refs/heads/master@{#30819}
The StringCompareStub used to take its parameters on the (JavaScript)
stack, which made it impossible to use in TurboFan. Actually
StringCompareStub was currently completely unused. This changes the
calling convention to something TurboFan compatible and introduces a
CallInterfaceDescriptor for StringCompareStub. It also changes
HStringCompareAndBranch to use the StringCompareStub instead of using
the full blown CompareICStub for a stupid string comparison.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1347913003
Cr-Commit-Position: refs/heads/master@{#30818}
This removes the weird COMPARE and COMPARE_STRONG JavaScript builtins
and replaces them with a proper C++ implementation in Object::Compare
and appropriate wrappers Object::LessThan, Object::GreaterThan, and
friends that are intended to be used by a true/false returning CompareIC
in the future, as well as the interpreter. As a short-term solution we
provide %Compare and %Compare_Strong entry points for the current
CompareIC that return the appropriate integer values expected by
fullcodegen currently.
Now the Abstract Relational Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/1350113002
Cr-Commit-Position: refs/heads/master@{#30816}
Defines and uses a kMaxSafeInteger macro (which expands to the constant
2^53-1) instead of accessing Number.MAX_SAFE_INTEGER. This saves loading
the attribute from the Number object, which is slightly faster. This also
makes it clearer from reading the code that tha value being compared is
constant.
BUG=
LOG=N
Review URL: https://codereview.chromium.org/1353953002
Cr-Commit-Position: refs/heads/master@{#30814}
