This passes the new.target value in a register instead of through a
side-channel via the construct stub. Note that only TurboFan code uses
the register value so far, but unoptimized code will be switched soon.
R=bmeurer@chromium.org
BUG=v8:4544
LOG=n
Review URL: https://codereview.chromium.org/1460503008
Cr-Commit-Position: refs/heads/master@{#32203}
The option --json-output will make the auto-roller dump a
json file with a monitoring_state key. This can be one of:
started, up_to_date, success.
BUG=chromium:559141
LOG=n
NOTRY=true
Review URL: https://codereview.chromium.org/1465413002
Cr-Commit-Position: refs/heads/master@{#32201}
We use comparisons with the_hole to implement temporal dead zones, so we
should also optimize those, as they currently turn into CompareIC calls.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1473593002
Cr-Commit-Position: refs/heads/master@{#32198}
port 374b6ea210 (r32172)
original commit message:
There's no point in collecting feedback for super constructor calls,
because in all (interesting) cases we can gather (better) feedback from
other sources (i.e. via inlining or via using a LOAD_IC to get to the
[[Prototype]] of the target). So CallConstructStub is now only used
for new Foo(...args) sites where we want to collect feedback in the
baseline compiler. The optimizing compilers, Reflect.construct and
super constructor calls use the Construct builtin directly, which allows
us to remove some weird code from the CallConstructStub (and opens the
possibility for more code sharing with the CallICStub, maybe even going
for a ConstructICStub).
Also remove the 100% redundant HCallNew instruction, which is just a
wrapper for the Construct builtin anyway (indirectly via the
CallConstructStub).
Drive-by-fix: Drop unused has_function_cache bit on Code objects.
BUG=
Review URL: https://codereview.chromium.org/1471193002
Cr-Commit-Position: refs/heads/master@{#32197}
This CL also fixes HeapObject::IterateFast() and HeapObject::IterateBodyFast().
BUG=v8:4531
LOG=Y
Review URL: https://codereview.chromium.org/1469883002
Cr-Commit-Position: refs/heads/master@{#32195}
This patch removes Promise functions and methods which are absent
from the ES2015 specification when the --es-staging flag is on.
BUG=v8:3237
R=rossberg
LOG=Y
Review URL: https://codereview.chromium.org/1469543003
Cr-Commit-Position: refs/heads/master@{#32194}
This replaces chromium_roll and the wrapper auto_roll with
just auto_roll. The arguments for the revision to roll and
the last rolled revision are optional and will be
auto-detected (this feature is from the wrapper).
Checking the cq box is default now (as in the wrapper).
BUG=chromium:559141
LOG=n
NOTRY=true
Review URL: https://codereview.chromium.org/1468973002
Cr-Commit-Position: refs/heads/master@{#32193}
port c1e7c8d972 (r32171)
original commit message:
This ensures that the ArgumentsAdaptorTrampoline does not clobber the
new.target value, but rather passes it through to the callee unaltered.
Note that callees do not yet use the new.target value so far.
This is a preparatory CL to allows us passing new.target in a register
instead of via a side-channel through the construct stub frame.
BUG=
Review URL: https://codereview.chromium.org/1475523002
Cr-Commit-Position: refs/heads/master@{#32192}
Add support for using inline allocations for arrays in lowering of
JSCreateArray when target equals new.target. Currently we are only
concerend with the straight-forward Array() and Array(length) cases,
but at some point TurboFan should also be able to support the more
complex initializing cases.
R=mvstanton@chromium.org
BUG=v8:4470
LOG=n
Review URL: https://codereview.chromium.org/1465203002
Cr-Commit-Position: refs/heads/master@{#32191}
In a function expression, 'yield' is allowed, even if the expression
occurs inside a generator. Similarly, even in a non-generator,
a generator expression's name must not be 'yield'.
BUG=v8:3983
LOG=n
Review URL: https://codereview.chromium.org/1460393003
Cr-Commit-Position: refs/heads/master@{#32188}
- Throw type errors instead of runtime asserts.
- Also required for SIMD tests ported to Test262 to work with V8
BUG=4501
LOG=N
Review URL: https://codereview.chromium.org/1461823003
Cr-Commit-Position: refs/heads/master@{#32187}
Port 374b6ea210
Original commit message:
There's no point in collecting feedback for super constructor calls,
because in all (interesting) cases we can gather (better) feedback from
other sources (i.e. via inlining or via using a LOAD_IC to get to the
[[Prototype]] of the target). So CallConstructStub is now only used
for new Foo(...args) sites where we want to collect feedback in the
baseline compiler. The optimizing compilers, Reflect.construct and
super constructor calls use the Construct builtin directly, which allows
us to remove some weird code from the CallConstructStub (and opens the
possibility for more code sharing with the CallICStub, maybe even going
for a ConstructICStub).
Also remove the 100% redundant HCallNew instruction, which is just a
wrapper for the Construct builtin anyway (indirectly via the
CallConstructStub).
Drive-by-fix: Drop unused has_function_cache bit on Code objects.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=v8:4413, v8:4430
LOG=n
Review URL: https://codereview.chromium.org/1471623005
Cr-Commit-Position: refs/heads/master@{#32186}
Port d80fd48e5d
Original commit message:
The CallICStub has call-site specific knowledge about the receiver,
which we did not utilize; plus the CallICStub does in some case know
whether it is about to [[Call]] a function or potentially some other
callable. In the common case we actually know that the target is a
function and so we can use the CallFunction builtin directly instead
of redispatching in the Call builtin.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=chromium:555127, v8:4413
LOG=n
Review URL: https://codereview.chromium.org/1471823002
Cr-Commit-Position: refs/heads/master@{#32185}
Port c1e7c8d972
Original commit message:
This ensures that the ArgumentsAdaptorTrampoline does not clobber the
new.target value, but rather passes it through to the callee unaltered.
Note that callees do not yet use the new.target value so far.
This is a preparatory CL to allows us passing new.target in a register
instead of via a side-channel through the construct stub frame.
R=mstarzinger@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=v8:4544
LOG=n
Review URL: https://codereview.chromium.org/1467923004
Cr-Commit-Position: refs/heads/master@{#32184}
Port c6d310da4d
Original commit message:
* Adds a PrepareForTailCall instruction that bumps the stack in the case that
the number of parameters passed to the callee causes the stack to exceed the
calleer's frame size.
* Uses the gap resolver to move the saved caller return address and frame
pointer to the approprate location in the tail-called frame.
R=danno@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=v8:4076
LOG=n
Review URL: https://codereview.chromium.org/1460183003
Cr-Commit-Position: refs/heads/master@{#32183}
Port ceade6cf23
Original commit message:
This adds a new %NewArray runtime entry, which constructs a new JSArray
and does the subclassing correctly (to the same degree that %NewObject
does currently), and also deals properly with the AllocationSite
feedback mechanism. This runtime entry will be used by TurboFan and is
also used as a fallback in the subclassing case in the stub currently.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=v8:3101, v8:3330
LOG=n
Review URL: https://codereview.chromium.org/1466943002
Cr-Commit-Position: refs/heads/master@{#32180}
The LoadElimination in TurboFan can look into the atomic regions and
elimination subsequent loads based on stores/allocations in that atomic
regions.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1472703005
Cr-Commit-Position: refs/heads/master@{#32179}
Add an eager deoptimization location for JSCallConstruct and adapt the
JSCallReducer to consume target feedback for construction sites (only
applies to explicit new F(...args) not the super constructor calls).
Also recognize the new Array(...args) constructs with only target
feedback.
R=jarin@chromium.org
BUG=v8:4470
LOG=n
Review URL: https://codereview.chromium.org/1467173002
Cr-Commit-Position: refs/heads/master@{#32177}
Reason for revert:
Janks finalization of full GC.
Original issue's description:
> [heap] Preserve optimized code maps accross GCs.
>
> This disables the eager flushing of optimized code maps during full GCs
> and instead treats entries in those tables weakly by default. Note that
> for now entries are removed when one of its components dies.
>
> R=mvstanton@chromium.org
>
> Committed: https://crrev.com/7a1377aab2ba4e2c9a7d422b51f645cbbaec10c7
> Cr-Commit-Position: refs/heads/master@{#32045}
TBR=mvstanton@chromium.org,mstarzinger@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1467213002
Cr-Commit-Position: refs/heads/master@{#32173}
There's no point in collecting feedback for super constructor calls,
because in all (interesting) cases we can gather (better) feedback from
other sources (i.e. via inlining or via using a LOAD_IC to get to the
[[Prototype]] of the target). So CallConstructStub is now only used
for new Foo(...args) sites where we want to collect feedback in the
baseline compiler. The optimizing compilers, Reflect.construct and
super constructor calls use the Construct builtin directly, which allows
us to remove some weird code from the CallConstructStub (and opens the
possibility for more code sharing with the CallICStub, maybe even going
for a ConstructICStub).
Also remove the 100% redundant HCallNew instruction, which is just a
wrapper for the Construct builtin anyway (indirectly via the
CallConstructStub).
Drive-by-fix: Drop unused has_function_cache bit on Code objects.
R=mstarzinger@chromium.org, yangguo@chromium.org
BUG=v8:4413, v8:4430
LOG=n
Review URL: https://codereview.chromium.org/1469793002
Cr-Commit-Position: refs/heads/master@{#32172}
This ensures that the ArgumentsAdaptorTrampoline does not clobber the
new.target value, but rather passes it through to the callee unaltered.
Note that callees do not yet use the new.target value so far.
This is a preparatory CL to allows us passing new.target in a register
instead of via a side-channel through the construct stub frame.
R=mvstanton@chromium.org
BUG=v8:4544
LOG=n
Review URL: https://codereview.chromium.org/1458103003
Cr-Commit-Position: refs/heads/master@{#32171}
This simplifies the layout of dependent code array and optimizes it for sparse dependency groups.
BUG=chromium:554488
LOG=NO
Review URL: https://codereview.chromium.org/1435313002
Cr-Commit-Position: refs/heads/master@{#32170}
The feature sometimes prevents subsequent swarming bots to
delete the work directory on windows.
The data file is not kept on swarming bots anyways, therefore
this switches off the feature completely.
BUG=chromium:535160
LOG=n
Review URL: https://codereview.chromium.org/1468933002
Cr-Commit-Position: refs/heads/master@{#32168}
port d80fd48e5d (r32163)
original commit message:
The CallICStub has call-site specific knowledge about the receiver,
which we did not utilize; plus the CallICStub does in some case know
whether it is about to [[Call]] a function or potentially some other
callable. In the common case we actually know that the target is a
function and so we can use the CallFunction builtin directly instead
of redispatching in the Call builtin.
BUG=
Review URL: https://codereview.chromium.org/1467123002
Cr-Commit-Position: refs/heads/master@{#32167}
port 4e00456471 (r31625)
original commit message:
X87: Fix the sqrt precision issue.
In order to resolve the sqrt precision issue described in https://codereview.chromium.org/1425763002/.
we change the implementation of CreateSqrtFunction() implementation of X87 so that the optimize compiler
and full-compiler implementation are unified.
BUG=
Review URL: https://codereview.chromium.org/1470793004
Cr-Commit-Position: refs/heads/master@{#32166}
The CallICStub has call-site specific knowledge about the receiver,
which we did not utilize; plus the CallICStub does in some case know
whether it is about to [[Call]] a function or potentially some other
callable. In the common case we actually know that the target is a
function and so we can use the CallFunction builtin directly instead
of redispatching in the Call builtin.
BUG=chromium:555127, v8:4413
LOG=n
Review URL: https://codereview.chromium.org/1470803002
Cr-Commit-Position: refs/heads/master@{#32163}
port c6d310da4d (r32151)
original commit message:
* Adds a PrepareForTailCall instruction that bumps the stack in the case that
the number of parameters passed to the callee causes the stack to exceed the
calleer's frame size.
* Uses the gap resolver to move the saved caller return address and frame
pointer to the approprate location in the tail-called frame.
BUG=
Review URL: https://codereview.chromium.org/1472703002
Cr-Commit-Position: refs/heads/master@{#32162}
port 2fc2cb99f5 (r32144)
original commit message:
The old code was not ready for properly initialize objects with non standard headers and non zero in-object properties number.
MacroAssembler::Allocate() implementations now return both start and end addresses of the new object (done by parameter renaming).
BUG=
Review URL: https://codereview.chromium.org/1467923002
Cr-Commit-Position: refs/heads/master@{#32161}
port ceade6cf23 (r32131)
original commit message:
This adds a new %NewArray runtime entry, which constructs a new JSArray
and does the subclassing correctly (to the same degree that %NewObject
does currently), and also deals properly with the AllocationSite
feedback mechanism. This runtime entry will be used by TurboFan and is
also used as a fallback in the subclassing case in the stub currently.
BUG=
Review URL: https://codereview.chromium.org/1462283003
Cr-Commit-Position: refs/heads/master@{#32160}
