Deopt support is added on two levels. On the IR level,
a new ObjectState node is added, which represenents an
object to be materialized. ObjectState nodes appear as
inputs of FrameState and StateValues nodes. On the
instruction select/code-generation level, the
FrameStateDescriptor class handles the nesting
introduced by ObjectState, and ensures that deopt code
with CAPTURED_OBJECT/DUPLICATED_OBJECT entries are
generated similarly to what crankshaft's escape
analysis does.

Two unittests test correctness of the IR level implementation.

Correctness for instruction selection / code generation
is tested by mjsunit tests.

R=jarin@chromium.org,mstarzinger@chromium.org
BUG=v8:4586
LOG=n

Review URL: https://codereview.chromium.org/1485183002

Cr-Commit-Position: refs/heads/master@{#33115}

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1559013004

Cr-Commit-Position: refs/heads/master@{#33114}

BUG=chromium:574349
LOG=NO

Review URL: https://codereview.chromium.org/1553233003

Cr-Commit-Position: refs/heads/master@{#33113}

This changes representation inference to be bidirectional:
1. truncations are propagated from uses to definitions.
2. output types are propagated from definitions to uses.
(and nodes are revisited until fixpoint.)

At the moment, (2) is used only superficially; the idea here is to
use the output type propagation to propagate types from type feedback.

For the output types to be usable, we need to keep track of the type
of the JavaScript value rather than the truncated value. Otherwise,
representation inference could not rely on the ranges indicated
by the values.

For example, for "var b = (a|0) + (a|0); return (b/16) >>> 0",
the type of b cannot be int32; otherwise the division "b/16"
would believe that it is fine to do an integer division on
the truncated value, which would give a wrong result for
2^31 <= a < 2^32.

The change makes representation inference a bit more expensive
(the phase is about 20% slower), but since this is only small part
of the overall compiler time, the overall effect is negligible.
If the running time becomes a problem, we could optimize this by
remembering when the nodes are stable (ie., no further changes to
type/truncations) and/or explicit subscriptions for changes.

BUG=v8:4583
R=bmeurer@chromium.org
LOG=N

Review URL: https://codereview.chromium.org/1490763003

Cr-Commit-Position: refs/heads/master@{#33112}

Adds implementation and tests for wide variant of load/store lookup slots.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1555063002

Cr-Commit-Position: refs/heads/master@{#33111}

  The FP parameters for X87 should be put in stack and the FP return value
  should be put in stX_0 register.

  So this CL defines the FP_RETURN_REGISTERS as stX_0 only.

BUG=

Review URL: https://codereview.chromium.org/1545203002

Cr-Commit-Position: refs/heads/master@{#33110}

Almost all of the Date builtins always call into C++ at least once
anyway, so parsing, compiling and executing the JavaScript wrappers
is just a waste of time.  The most important part here is the Date
constructor itself, which is one of the blockers for new.target in
TurboFan, because compiling the Date constructor takes too much time
with TurboFan (for no reason since we end up in C++ anway).

R=cbruni@chromium.org

Review URL: https://codereview.chromium.org/1556333002

Cr-Commit-Position: refs/heads/master@{#33109}

The mode requires an extra register, and since we aren't supporting
it now, we can dispense with it.

BUG=

Review URL: https://codereview.chromium.org/1559903003

Cr-Commit-Position: refs/heads/master@{#33108}

R=bmeurer@chromium.org
BUG=v8:4586
LOG=n

Review URL: https://codereview.chromium.org/1559003002

Cr-Commit-Position: refs/heads/master@{#33107}

This also fixes a related bug in phi handling.

R=bmeurer@chromium.org
BUG=v8:4586
LOG=n

Review URL: https://codereview.chromium.org/1554073003

Cr-Commit-Position: refs/heads/master@{#33106}

R=bmeurer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/1557883002

Cr-Commit-Position: refs/heads/master@{#33105}

BUG=v8:4280
LOG=N
TBR=rmcilroy@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1555003004

Cr-Commit-Position: refs/heads/master@{#33104}

  port 02072112(r33088)

  original commit message:
  Fix stack push issues on non-x64 platforms for
  InterpreterNotifyDeoptimized builtins.

BUG=

Review URL: https://codereview.chromium.org/1553083005

Cr-Commit-Position: refs/heads/master@{#33103}

  port 82ca2a41(r33084)

  original commit message:
  This is preferable because in TurboFan we need to call it, and can't pass
  untagged external pointers on the stack.

BUG=

Review URL: https://codereview.chromium.org/1557283002

Cr-Commit-Position: refs/heads/master@{#33102}

TEST=asm-wasm.js
R=titzer@chromium.org,bradnelson@chromium.org
BUG=

Add for loop to asm-to-wasm

TEST=asm-wasm.js
R=titzer@chromium.org,bradnelson@google.com

Review URL: https://codereview.chromium.org/1530093002

Cr-Commit-Position: refs/heads/master@{#33101}

In the Date constructor or Date.parse, other browsers will accept time
zones like GMT-8, but before this patch, Chrome would interpret 8 as
8 minutes. This patch interprets GMT-+ a one or two digit number as hours,
not minutes.

R=adamk,jshin@chromium.org
LOG=Y
BUG=chromium:422858

Review URL: https://codereview.chromium.org/1557053002

Cr-Commit-Position: refs/heads/master@{#33100}

Fix build break.

R=joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=

Review URL: https://codereview.chromium.org/1556063004

Cr-Commit-Position: refs/heads/master@{#33099}

Port 02072112

Original commit message:
    Fix stack push issues on non-x64 platforms for
    InterpreterNotifyDeoptimized builtins.

R=rmcilroy@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1554273002

Cr-Commit-Position: refs/heads/master@{#33098}

1. Location names with more than one underscores (e.g. Ho_Chi_Minh)
   didn't work because of the way capturing works with repeated patterns
   in RE. It's now supported by changing the RE to capture the whole string
   and splitting on '_' in the next step.

2. Adds support for location names with a hyphen

3. Adds support for timezone ids with three parts (e.g.
   American/Argentina/Buenos_Aires)

4. Adds special handling of 'au', 'es' and 'of' in zone ids. They need to be kept in lowercase. (see the full list at https://en.wikipedia.org/wiki/List_of_tz_database_time_zones )

5. Adds regression tests for all the above and make the existing tests
   more robust against future ICU changes. ICU canonicalizes zone names to
   deprecated names, but it may change. (
   http://bugs.icu-project.org/trac/ticket/12044 )

BUG=364374
LOG=Y

Review URL: https://codereview.chromium.org/1529363005

Cr-Commit-Position: refs/heads/master@{#33097}

Adds an ignition whitelist for cctests.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1557893002

Cr-Commit-Position: refs/heads/master@{#33096}

This patch adds the basis for subclassing TypedArrays, Arrays and
ArrayBuffers through the @@species hook, added in ES2015. This is
the first patch in a series. This patch simply defines the
@@species Symbol and installs it on the appropriate constructors.
The behavior is guarded behind the --harmony-species flag.

R=cbruni
BUG=v8:4093
LOG=Y
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
TBR=hpayer@chromium.org

Review URL: https://codereview.chromium.org/1558543002

Cr-Commit-Position: refs/heads/master@{#33095}

Fixes a number of test262 tests, including

- built-ins/Promise/resolve/resolve-from-promise-capability.js
- built-ins/Promise/resolve/context-non-object-with-promise.js
- built-ins/Promise/executor-function-length.js

BUG=v8:4633
LOG=N
R=littledan@chromium.org, cbruni@chromium.org

Review URL: https://codereview.chromium.org/1536013002

Cr-Commit-Position: refs/heads/master@{#33094}

There were some functions that were used when atomic floats were supported, but
can now be removed.

BUG=

Review URL: https://codereview.chromium.org/1552683003

Cr-Commit-Position: refs/heads/master@{#33093}

Port 82ca2a41

Original commit message:
    This is preferable because in TurboFan we need to call it, and can't pass
    untagged external pointers on the stack.

R=mvstanton@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=

Review URL: https://codereview.chromium.org/1553143002

Cr-Commit-Position: refs/heads/master@{#33092}

New bytecodes for making registers with indicies wider than 1-byte
accessible.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1555713002

Cr-Commit-Position: refs/heads/master@{#33091}

Adds --turbo to the set of flags run with ignition (making the
configuration pure Ignition+TurboFan except for fallbacks to
full-codegen for catch / eval). Also changes the default
--ignition-filter to allow everything, rather than omit everything.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1537433003

Cr-Commit-Position: refs/heads/master@{#33090}

Add control for the ignition test suites. This will be
followed by a change on the infra side that targets the
ignition set instead of the hardcoded mjsunit.

After that, cctest can be enabled in a separate CL. This CL
already includes the logic for running cctest for ignition
on swarming.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1559943002

Cr-Commit-Position: refs/heads/master@{#33089}

Fix stack push issues on non-x64 platforms for
InterpreterNotifyDeoptimized builtins.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1553123002

Cr-Commit-Position: refs/heads/master@{#33088}

R=cbruni@chromium.org

Review URL: https://codereview.chromium.org/1559923002

Cr-Commit-Position: refs/heads/master@{#33087}

VisitNewLocalBlockContext funciton was allocating two registers and checking
if they are consecutive. This cl changes it to reserve consecutive registers.
The earlier code breaks this test mjsunit/regress/regress-542099.js.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1545003003

Cr-Commit-Position: refs/heads/master@{#33086}

This restricts turbofan to turbofan-supported subset for the shipping
configuration ("use asm" and features unsupported by Crankshaft).

Without this, we compile with Turbofan even when there is
try-catch-finally as long as the function is "use asm" or
it contains a feature unsupported by crankshaft but supported
by turbofan (e.g., 'with' statement).

BUG=

Review URL: https://codereview.chromium.org/1552233002

Cr-Commit-Position: refs/heads/master@{#33085}

This is preferable because in TurboFan we need to call it, and can't pass
untagged external pointers on the stack.

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/1556913002

Cr-Commit-Position: refs/heads/master@{#33084}

R=verwaest@chromium.org
BUG=chromium:573857
LOG=n

Review URL: https://codereview.chromium.org/1551333002

Cr-Commit-Position: refs/heads/master@{#33083}

In addition to ThrowTypeError not being a constructor, it was invalidly set up by simply swapping the map. Swapping the map is insufficient since NewFunction* actually set up the function depending on the used map.

BUG=chromium:573858
LOG=n

Review URL: https://codereview.chromium.org/1552223002

Cr-Commit-Position: refs/heads/master@{#33082}

LoadElement nodes used a Int32Constant in some cases, which made
matching for constant values in optimizations more difficult than
necessary.

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/1559853002

Cr-Commit-Position: refs/heads/master@{#33081}

TBR=jarin@chromium.org

Review URL: https://codereview.chromium.org/1555053002 .

Cr-Commit-Position: refs/heads/master@{#33080}

TBR=hablich@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1553103002

Cr-Commit-Position: refs/heads/master@{#33079}

BUG=572409
LOG=n

Review URL: https://codereview.chromium.org/1555023002

Cr-Commit-Position: refs/heads/master@{#33078}

TBR=jarin@chromium.org

Review URL: https://codereview.chromium.org/1557833002 .

Cr-Commit-Position: refs/heads/master@{#33077}

After this change (https://codereview.chromium.org/1507903004) to type feedback
vector, wide bytecodes for global/keyed/named load-stores were not generated due
to a change in the number of type feedback vector slots. This cl fixes tests to
generate wide bytecodes.

BUG=v8:4280
LOG=N

Review URL: https://codereview.chromium.org/1546923002

Cr-Commit-Position: refs/heads/master@{#33076}