This unifies several of the places in CSA that convert various
string kinds (cons, thin, sliced) to direct strings
(sequential, external).

A couple of spots remain with duplicate code, but most of these are
more difficult to unify due to specific optimizations.

BUG=

Review-Url: https://codereview.chromium.org/2744263002
Cr-Commit-Position: refs/heads/master@{#43848}

Revert of Collect type profile for DevTools. (patchset #40 id:750001 of https://codereview.chromium.org/2707873002/ )

Reason for revert:
Flaky under stress. Fix first.

Original issue's description:
> Collect type profile for DevTools
>
> Collect type information for JavaScript variables and display it
> in Chrome DevTools.
> Design Doc: https://docs.google.com/a/google.com/document/d/1O1uepXZXBI6IwiawTrYC3ohhiNgzkyTdjn3R8ysbYgk/edit?usp=sharing
>
> When debugging JavaScript, it’s helpful to know the type of
> a variable, parameter, and return values. JavaScript is
> dynamically typed, and for complex
> source code it’s often hard to infer types. With type profiling, we
> can provide type information to JavaScript developers.
>
> This CL is a proof of concept. It collects type profile for
> assignments and simply prints the types to stdout.
>
> The output looks something like this:
>
> #my_var1
>   #Object
>   #number
>   #string
>   #number
>   #undefined
>   #string
>   #Object
>   #Object
>
>
> We use an extra slot in the feedback vector of assignments to
> carry the list of types for that assignment. The extra slot is
> only added when the flag --type-profile is given.
>
>
> Missing work:
> * Collect data for parameters and return values (currently only assignments).
> * Remove duplicates from the list of collected types and use a common base class.
> * Add line numbers or source position instead of the variable name.
>
>
>
> BUG=v8:5935
>
> Review-Url: https://codereview.chromium.org/2707873002
> Cr-Original-Original-Commit-Position: refs/heads/master@{#43791}
> Committed: https://chromium.googlesource.com/v8/v8/+/0332bebde99d0f9a5a8326382f5f37cc26224ae0
> Review-Url: https://codereview.chromium.org/2707873002
> Cr-Original-Commit-Position: refs/heads/master@{#43804}
> Committed: https://chromium.googlesource.com/v8/v8/+/6cf880f4b84c533d4bb139d33c1369e309d1c579
> Review-Url: https://codereview.chromium.org/2707873002
> Cr-Commit-Position: refs/heads/master@{#43846}
> Committed: https://chromium.googlesource.com/v8/v8/+/5c322873908a5b5c04552fc47d8d81f7603b5d11

TBR=yangguo@chromium.org,mvstanton@chromium.org,rmcilroy@chromium.org,mstarzinger@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5935

Review-Url: https://codereview.chromium.org/2747383004
Cr-Commit-Position: refs/heads/master@{#43847}

Collect type information for JavaScript variables and display it
in Chrome DevTools.
Design Doc: https://docs.google.com/a/google.com/document/d/1O1uepXZXBI6IwiawTrYC3ohhiNgzkyTdjn3R8ysbYgk/edit?usp=sharing

When debugging JavaScript, it’s helpful to know the type of
a variable, parameter, and return values. JavaScript is
dynamically typed, and for complex
source code it’s often hard to infer types. With type profiling, we
can provide type information to JavaScript developers.

This CL is a proof of concept. It collects type profile for
assignments and simply prints the types to stdout.

The output looks something like this:

#my_var1
  #Object
  #number
  #string
  #number
  #undefined
  #string
  #Object
  #Object

We use an extra slot in the feedback vector of assignments to
carry the list of types for that assignment. The extra slot is
only added when the flag --type-profile is given.

Missing work:
* Collect data for parameters and return values (currently only assignments).
* Remove duplicates from the list of collected types and use a common base class.
* Add line numbers or source position instead of the variable name.

BUG=v8:5935

Review-Url: https://codereview.chromium.org/2707873002
Cr-Original-Original-Commit-Position: refs/heads/master@{#43791}
Committed: https://chromium.googlesource.com/v8/v8/+/0332bebde99d0f9a5a8326382f5f37cc26224ae0
Review-Url: https://codereview.chromium.org/2707873002
Cr-Original-Commit-Position: refs/heads/master@{#43804}
Committed: https://chromium.googlesource.com/v8/v8/+/6cf880f4b84c533d4bb139d33c1369e309d1c579
Review-Url: https://codereview.chromium.org/2707873002
Cr-Commit-Position: refs/heads/master@{#43846}

BUG=

Review-Url: https://codereview.chromium.org/2751993002
Cr-Commit-Position: refs/heads/master@{#43845}

This moves most of the logic contained in RegExpExecStub to CSA.  Benefits are
mostly easier readability and hackability, and removal of a large chunk of
platform-specific assembly.

Exit frame construction and the final call remain in RegExpExecStub.

BUG=v8:5339,v8:592

Review-Url: https://codereview.chromium.org/2738413002
Cr-Commit-Position: refs/heads/master@{#43844}

The fast-path for Array.prototype.indexOf first checks whether the
receiver is a fast-mode JSArray (and there are no elements in the
prototype chain in case of holey arrays), then loads the known
JSArray::length, and afterwards calls ToInteger on the fromIndex.

But this ToInteger(fromIndex) call can cause arbitrary side effects if
the fromIndex is a JSReceiver, in particular it can invalidate the
assumptions about the fast-mode of the receiver and the length. In the
worst case this leads to OOB memory access.

Quick-fix is to bailout to the runtime if the fromIndex is neither a Smi
nor undefined, which represents the common cases.

R=jarin@chromium.org
BUG=chromium:702058

Review-Url: https://codereview.chromium.org/2756663002
Cr-Commit-Position: refs/heads/master@{#43843}

Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/2744073..81c2772

Rolling v8/third_party/catapult: https://chromium.googlesource.com/external/github.com/catapult-project/catapult/+log/e9f547b..d49bf81

TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org

Change-Id: Id67a6e6608c34c753dbeb01852e9b16bc0d6a0a3
Reviewed-on: https://chromium-review.googlesource.com/455660Reviewed-by: v8 autoroll <v8-autoroll@chromium.org>
Commit-Queue: v8 autoroll <v8-autoroll@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43842}

- Adds new load/store opcodes (0xc0, 0xc1) for S128 type.
- Implements these for ARM.
- Enables more WASM SIMD tests, and adds new LoadStoreLoad test.

LOG=N
BUG=v8:6020

Review-Url: https://codereview.chromium.org/2745853002
Cr-Commit-Position: refs/heads/master@{#43841}

This reverts commit 58ff145e.

Reason for revert: the bug in git-numberer has been fixed and deployed.

Original change's description:
> [release] Temporarily disable gerrit for merge tools.
> 
> BUG=chromium:698845
> NOTRY=true
> TBR=hablich@chromium.org,tandrii@chromium.org
> 
> Change-Id: I970c984c5e9bd23ca3d951d13243d107fc2a2e3f
> Reviewed-on: https://chromium-review.googlesource.com/451280
> Reviewed-by: Michael Achenbach <machenbach@chromium.org>
> Reviewed-by: Michael Hablich <hablich@chromium.org>
> Commit-Queue: Michael Hablich <hablich@chromium.org>
> Commit-Queue: Michael Achenbach <machenbach@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#43646}

R=machenbach@chromium.org,hablich@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
BUG=chromium:698845

Change-Id: I34b12a36c21a2d7a46eb67bce0536298a0582dd5
Reviewed-on: https://chromium-review.googlesource.com/455799Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43840}

'const' type qualifier on return type has no effect and
produces compiler warnings.

BUG=

Review-Url: https://codereview.chromium.org/2751513004
Cr-Commit-Position: refs/heads/master@{#43839}

BUG=v8:6103, chromium:700294
R=bmeurer@chromium.org

Review-Url: https://codereview.chromium.org/2756463002
Cr-Commit-Position: refs/heads/master@{#43838}

R=mvstanton@chromium.org

Change-Id: Ifc1c8a7ba83f4f877b3dd3cd58ca71febc48b550
Reviewed-on: https://chromium-review.googlesource.com/455737Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43837}

BUG=v8:5561

Change-Id: I425ad7e670b8e9c701ff284ab1be3638a975a6d3
Reviewed-on: https://chromium-review.googlesource.com/455832
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43836}

When generating the code for an illegal (non-js-compliant) wrapper for
a js function, we were generating a zero constant of the return type.
This failed on ia32 if the return type is i64.
The correct thing to do is to return two i32 zero constants.

R=titzer@chromium.org
BUG=v8:6104, v8:6096

Change-Id: Idc0dd8d3eef901c22d5278d97f1879b000a18588
Reviewed-on: https://chromium-review.googlesource.com/455857Reviewed-by: Ben Titzer <titzer@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43835}

That way, we don't need to create a context-less constructor function.
Instead, we use the constructor_or_backpointer (or null) field, and
rename it to constructor_or_fti_or_backpointer so nobody is confused.

Note that technically, we still have JSFunctions without contexts, as
they're temporarily created in the deoptimizer.

BUG=v8:6084
R=dcheng@chromium.org,bmeurer@chromium.org,verwaest@chromium.org

Change-Id: I084f052533c317f2cbfb9c35e1acf40263c6257b
Reviewed-on: https://chromium-review.googlesource.com/454636
Commit-Queue: Jochen Eisinger <jochen@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Hannes Payer <hpayer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43834}

Marking shared functions for tier-up was optimizing the functions
non-concurrently, to avoid the case where the same shared function is
optimized concurrently by multiple JS functions. This was particularly a
problem for small functions, which (if called in a loop) could get
marked for optimisation quite quickly.

In this CL, the shared function is instead marked as having an active
optimization job running, and these do not spawn a compilation job.

BUG=chromium:693590
BUG=chromium:700863
BUG=chromium:701665

Change-Id: I2b1c5af8e7aa8d779f86814c22c65c78bee0630f
Reviewed-on: https://chromium-review.googlesource.com/455779Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43833}

Adds a TestTypeof bytecode to deal with comparisons of the form:
  typeof(object) === 'string';

Also adds support to Turbofan to perform these comparisons without
inserting checkpoints.

BUG=v8:4280,v8:5267

Change-Id: Ib5cc1c6816dfe70a4120838d8eada2fc0267750f
Reviewed-on: https://chromium-review.googlesource.com/454837Reviewed-by: Mythri Alle <mythria@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43832}

This adds one of the release unittests to presubmit.

Different entry points for CI and CQ are required.

More tests will follow in subsequent CLs.

BUG=chromium:701296
NOTRY=true

Change-Id: Ie96fba873f77df14efb1dc54388f075c056b64a6
Reviewed-on: https://chromium-review.googlesource.com/454639Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43831}

There's no point in trying to convert to a string, as this runtime
function must always be called with Strings, otherwise there's a bug.

R=yangguo@chromium.org

Review-Url: https://codereview.chromium.org/2748253004
Cr-Commit-Position: refs/heads/master@{#43830}

BUG=chromium:694255
R=hpayer@chromium.org

Change-Id: Ie8d23ec88e9b8f97ac094630f4e46e9b31f93cc5
Reviewed-on: https://chromium-review.googlesource.com/452817Reviewed-by: Hannes Payer <hpayer@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43829}

The function "foo" in "base_getter_test" was picking up a left-over
optimised code object from foo's code map, the third time that
base_getter_test was run, instead of optimising it itself. This
broke the assumptions of the test, that each case was functionally
independent from the others, and had started off with empty feedback
and no optimised code.

This breaks the test though, so we have to blacklist it pending a fix
to the root cause (http://crbug.com/v8/6101).

Review-Url: https://codereview.chromium.org/2750623004
Cr-Commit-Position: refs/heads/master@{#43828}

 - Added: Int32x4Mul, Int32x4Min, Int32x4Max, Int32x4Equal, Int32x4NotEqual
 Uint32x4Min, Uint32x4Max
 - Fix I32x4Splat

R=bbudge@chromium.org, bradnelson@chromium.org, mtrofin@chromium.org

Review-Url: https://codereview.chromium.org/2719953002
Cr-Commit-Position: refs/heads/master@{#43827}

This fixes a TSAN data race when writing the forwarding pointer in
MigrateObject and reading the object as a LayoutDescriptor when trying
to figure out the layout of another object in parallel.

BUG=chromium:701732

Change-Id: I1e291fa1afb42771244e1346680164de71c3a838
Reviewed-on: https://chromium-review.googlesource.com/455817Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43826}

It does not really matter for the preparser
 but do it for sake of consistency.

BUG=

Change-Id: I2f58486cb3a8b866bbfa35b398abc7b5ead02ee7
Reviewed-on: https://chromium-review.googlesource.com/455797Reviewed-by: Marja Hölttä <marja@chromium.org>
Commit-Queue: Wiktor Garbacz <wiktorg@google.com>
Cr-Commit-Position: refs/heads/master@{#43825}

This handles non-strict equality comparison operations having number
feedback during the early type-hint lowering (i.e. during graph
construction).

R=jarin@chromium.org

Change-Id: I1db67e78312934bbb20aee775979797420ff2581
Reviewed-on: https://chromium-review.googlesource.com/455796Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43824}

R=verwaest@chromium.org
BUG=v8:6071

Review-Url: https://codereview.chromium.org/2740033002
Cr-Commit-Position: refs/heads/master@{#43823}

BUG=v8:6100

Change-Id: Ib8729b2688bbaf6fb397737ccf1b1c086698ab93
Reviewed-on: https://chromium-review.googlesource.com/455876
Commit-Queue: Wiktor Garbacz <wiktorg@google.com>
Reviewed-by: Marja Hölttä <marja@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43822}

When compiling a wasm to js wrapper for a signature which is not JS
compatible, we generate a runtime call to throw a type error and return
immediately afterwards. We were using the return value of the runtime
call as return value for the wrapper.
This worked for integer return values, since the same registers are
used for tagged values and integer values. For float values, however,
it failed, since the return value needs to be stored in another
register.

R=titzer@chromium.org
BUG=v8:6096

Change-Id: I8f39ea132cd150c3044673d25fa3c3588b4266bf
Reviewed-on: https://chromium-review.googlesource.com/455816Reviewed-by: Ben Titzer <titzer@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43821}

BUG=v8:5294

Change-Id: I6214c50c7d1344210a80763b066e5ec56df1265a
Reviewed-on: https://chromium-review.googlesource.com/453460
Commit-Queue: Marja Hölttä <marja@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43820}

Mark slow tests discovered by manual audit of reported durations.

Slow tests will run first, which will lead to higher test runner throughput.

Also some tests have flaky timeouts occasionally.

BUG=chromium:601468

Change-Id: If9aeb5422e79ff25b867b01c50386a5fdebeb494
Reviewed-on: https://chromium-review.googlesource.com/455736Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43819}

BUG=

Change-Id: I1fd0ae3dcb0e644e9f2a56c242720b4def4bb579
Reviewed-on: https://chromium-review.googlesource.com/454836
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43818}

This commit will likely capture several perf alerts from bots that hung.

All allerts might be attributable to the I+TF switch.

BUG=v8:6067
NOTRY=true
NOTREECHECKS=true
NOPRESUBMIT=true
TBR=hablich@chromium.org

Change-Id: Ie6d0dc977823f8740856904bdcf157c8b021d60b
Reviewed-on: https://chromium-review.googlesource.com/455776Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43817}

BUG=

Change-Id: Ic63c63a312c6282814c14176d982b8c825ba6dc4
Reviewed-on: https://chromium-review.googlesource.com/454756
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43816}

R=jarin@chromium.org
BUG=

Review-Url: https://codereview.chromium.org/2745753002
Cr-Commit-Position: refs/heads/master@{#43815}

This changes the IR to no longer require single {IfSuccess} projection
nodes unless there is a corresponding {IfException} node that links the
potentially throwing call to an exception handler. This reduces graph
size as well as compilation time when exception handlers aren't present.

The new invariant for potentially throwing nodes is: Nodes that can
potentially throw either have both IfSuccess/IfException projections as
the only control uses and no direct control uses, or no projections at
all and solely direct control uses.

R=jarin@chromium.org

Change-Id: I3d9cd816d74ad5af13e0673da7ec7a98f1ecdc7e
Reviewed-on: https://chromium-review.googlesource.com/449715
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43814}

For the old asm.js pipeline, there's no point in running either the
JSCallReducer or the JSNativeContextSpecialization, as both will not
kick in for any asm.js relevant code.

So remove the kDeoptimizationEnabled flag from both of them and just
run them on regular JavaScript only.

Drive-by-fix: Slightly rearrange keyed access to String receivers in the
JSNativeContextSpecialization::ReduceKeyedAccess method to access constant
characters independent of the IC mode (as long as the constant index is
within the range of the receiver).

BUG=v8:5267
R=jarin@chromium.org

Review-Url: https://codereview.chromium.org/2743253002
Cr-Commit-Position: refs/heads/master@{#43813}

BUG=chromium:700883

Review-Url: https://codereview.chromium.org/2751513006
Cr-Commit-Position: refs/heads/master@{#43812}

BUG=chromium:693425

Review-Url: https://codereview.chromium.org/2749193003
Cr-Commit-Position: refs/heads/master@{#43811}

Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/7de5863..2744073

Rolling v8/third_party/catapult: https://chromium.googlesource.com/external/github.com/catapult-project/catapult/+log/1ec68b3..e9f547b

Rolling v8/tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/f091611..2a8813f

TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org

Change-Id: Ie78a772ac8d840d4f18a6b756a8ae67979cd2ef1
Reviewed-on: https://chromium-review.googlesource.com/455397Reviewed-by: v8 autoroll <v8-autoroll@chromium.org>
Commit-Queue: v8 autoroll <v8-autoroll@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43810}

Also use vinstr for vps/vpd/vss/vsd

BUG=

Review-Url: https://codereview.chromium.org/2747103002
Cr-Commit-Position: refs/heads/master@{#43809}