Commits · c1753f5c3d8b4c498cb3fa35e0d5c5b79f353c9b · Linshizhi / V8

01 Nov, 2016 1 commit

X87: [turbofan] Support variable size argument removal in TF-generated functions. · c1753f5c

zhengxing.li authored Nov 01, 2016

  port 5319b50c (r40678)

  original commit message:
  This is preparation for using TF to create builtins that handle variable number of
  arguments and have to remove these arguments dynamically from the stack upon
  return.

  The gist of the changes:
  - Added a second argument to the Return node which specifies the number of stack
    slots to pop upon return in addition to those specified by the Linkage of the
    compiled function.
  - Removed Tail -> Non-Tail fallback in the instruction selector. Since TF now should
    handles all tail-call cases except where the return value type differs, this fallback
    was not really useful and in fact caused unexpected behavior with variable
    sized argument popping, since it wasn't possible to materialize a Return node
    with the right pop count from the TailCall without additional context.
  - Modified existing Return generation to pass a constant zero as the additional
    pop argument since the variable pop functionality

BUG=

Review-Url: https://codereview.chromium.org/2469623002
Cr-Commit-Position: refs/heads/master@{#40684}

c1753f5c

31 Oct, 2016 22 commits

Remove --harmony-* flags for new Object static methods · a1381990

adamk authored Oct 31, 2016

Both --harmony-object-values-entries and --harmony-object-own-property-descriptors
are on by default in v8 5.4, which has now shipped to
stable as Chrome 54.

R=caitp@igalia.com

Review-Url: https://codereview.chromium.org/2464733003
Cr-Commit-Position: refs/heads/master@{#40683}

a1381990

PPC/s390: [turbofan] Support variable size argument removal in TF-generated functions · 2c846a2a

bjaideep authored Oct 31, 2016

Port 5319b50c

Original commit message:

    This is preparation for using TF to create builtins that handle variable number of
    arguments and have to remove these arguments dynamically from the stack upon
    return.

    The gist of the changes:
    - Added a second argument to the Return node which specifies the number of stack
      slots to pop upon return in addition to those specified by the Linkage of the
      compiled function.
    - Removed Tail -> Non-Tail fallback in the instruction selector. Since TF now should
      handles all tail-call cases except where the return value type differs, this fallback
      was not really useful and in fact caused unexpected behavior with variable
      sized argument popping, since it wasn't possible to materialize a Return node
      with the right pop count from the TailCall without additional context.
    - Modified existing Return generation to pass a constant zero as the additional
      pop argument since the variable pop functionality

R=danno@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, mbrandy@us.ibm.com

BUG=
LOG=N

Review-Url: https://codereview.chromium.org/2462173002
Cr-Commit-Position: refs/heads/master@{#40682}

2c846a2a

[Tracing] Remove unused excluded categories list. · 9c781943

lpy authored Oct 31, 2016

We only need included categories list, excluded categories list will only work
if we use regular expression in categories list, which is not supported in V8.

TBR=jochen@chromium.org

Review-Url: https://codereview.chromium.org/2462143002
Cr-Commit-Position: refs/heads/master@{#40681}

9c781943

[wasm] Do not run the infinite loop test if FLAG_dump_wasm_module is true · da90034e

ahaas authored Oct 31, 2016

By not dumping the module of the test, we avoid adding code of an
infinite loop to the corpus of the wasm fuzzer.

R=bradnelson@chromium.org

NOTRY=true

Review-Url: https://codereview.chromium.org/2460313002
Cr-Commit-Position: refs/heads/master@{#40680}

da90034e

[ic] Detailize some IC-specific runtime call stats counters. · 597315dc

ishell authored Oct 31, 2016

This CL adds Load/StoreIC_Premonomorphic and detailizes IC_HandlerCacheHit bucket.

BUG=v8:5561

Review-Url: https://codereview.chromium.org/2459333003
Cr-Commit-Position: refs/heads/master@{#40679}

597315dc

[turbofan] Support variable size argument removal in TF-generated functions · 5319b50c

danno authored Oct 31, 2016

This is preparation for using TF to create builtins that handle variable number of
arguments and have to remove these arguments dynamically from the stack upon
return.

The gist of the changes:
- Added a second argument to the Return node which specifies the number of stack
  slots to pop upon return in addition to those specified by the Linkage of the
  compiled function.
- Removed Tail -> Non-Tail fallback in the instruction selector. Since TF now should
  handles all tail-call cases except where the return value type differs, this fallback
  was not really useful and in fact caused unexpected behavior with variable
  sized argument popping, since it wasn't possible to materialize a Return node
  with the right pop count from the TailCall without additional context.
- Modified existing Return generation to pass a constant zero as the additional
  pop argument since the variable pop functionality

LOG=N

Review-Url: https://codereview.chromium.org/2446543002
Cr-Commit-Position: refs/heads/master@{#40678}

5319b50c

Add Shell::PrintErr and expose it in the d8 shell as printErr · fb96618e

jgravelle authored Oct 31, 2016

This function is implemented in other JavaScript shells

BUG=None

R=titzer

Review-Url: https://codereview.chromium.org/2458963003
Cr-Commit-Position: refs/heads/master@{#40677}

fb96618e

[stubs] Fix allocation memento detection. · 70d2b685

ishell authored Oct 31, 2016

The old code does not work properly when the allocation memento object ends exactly at page boundary.

This CL addresses DBC mentioned in https://codereview.chromium.org/2313093002/

BUG=v8:5269

Review-Url: https://codereview.chromium.org/2466573002
Cr-Commit-Position: refs/heads/master@{#40676}

70d2b685

Remove DCHECK that transferred array buffers are neutered or shared. · fa33489f

jbroman authored Oct 31, 2016

While this seems like it should be true, the array buffer is not actually
neutered until the end of cloning. This is so that, if an exception is thrown
during serialization, the original array buffer is not left neutered. As a
result, Blink will not have neutered the buffer.

This fixes some DCHECK failures during layout tests.

BUG=chromium:148757

Review-Url: https://codereview.chromium.org/2466563002
Cr-Commit-Position: refs/heads/master@{#40675}

fa33489f

Minor fixes in objects printer. · 9dd638fe

ishell authored Oct 31, 2016

BUG=

Review-Url: https://codereview.chromium.org/2468433002
Cr-Commit-Position: refs/heads/master@{#40674}

9dd638fe

[ic] Simplify handling of primitive maps. · 7c22a153

ishell authored Oct 31, 2016

This CL adds support for primitive maps to
1) PrototypeIterator,
2) PropertyHandlerCompiler::CheckPrototypes(),
3) Map::GetOrCreatePrototypeChainValidityCell(),
4) Prototype checks in data-driven ICs.

BUG=v8:5561

Review-Url: https://codereview.chromium.org/2466553002
Cr-Commit-Position: refs/heads/master@{#40673}

7c22a153

[debugger] Ensure debug listeners do not throw · 16310b2e

jgruber authored Oct 31, 2016

This exposes a couple of broken tests that used to silently throw within
the listener. Mark these as failing for now

BUG=v8:5330, v8:5581

Review-Url: https://codereview.chromium.org/2460833002
Cr-Commit-Position: refs/heads/master@{#40672}

16310b2e

Speedup access to global_proxy.* attributes/accessors. · 36f3f909

vogelheim authored Oct 31, 2016

Using a global proxy (e.g. 'window.f', 'w.f' or 'this.f') is considerably slower than evaluating just 'f'. This CL aims to perform the necessary checks at compile time and inline the accesses.

This is a follow-on CL to crrev.com/2369933005:
- The initial upload is crrev.com/2369933005 + a rebase.
- The remaining issues are the fixes requested by the reviewers on that CL.

BUG=chromium:634276, chromium:654716, chromium:656959

Committed: https://crrev.com/8f43d748272536117008aa6a1b53ea52126261c1
Committed: https://crrev.com/041314524952a3c1bc71bd3beafbbb37319f1d22
Review-Url: https://codereview.chromium.org/2403003002
Cr-Original-Original-Commit-Position: refs/heads/master@{#40153}
Cr-Original-Commit-Position: refs/heads/master@{#40365}
Cr-Commit-Position: refs/heads/master@{#40671}

36f3f909

Create internal fields on global proxy objects · 26547761

jochen authored Oct 31, 2016

BUG=v8:5588
R=verwaest@chromium.org

Review-Url: https://codereview.chromium.org/2467463002
Cr-Commit-Position: refs/heads/master@{#40670}

26547761

[turbofan] Use an IntPtrConstant in WordEqual comparisons. · 70dfb5a0

georgia.kouveli authored Oct 31, 2016

ZeroConstant is a NumberConstant, so does not get matched
in IntMatchers, which results in missing some opportunities
in instruction selection. Use IntPtrConstant instead, which
generates an Int32Constant or an Int64Constant depending on
the word size.

BUG=

Review-Url: https://codereview.chromium.org/2467473002
Cr-Commit-Position: refs/heads/master@{#40669}

70dfb5a0

[test] Skip the debug-liveedit-double-call mjsunit test. · 64421566

neis authored Oct 31, 2016

It's marked as [PASS, FAIL] but keeps showing up as a test failure for me.

R=mstarzinger@chromium.org
BUG=v8:5587

Review-Url: https://codereview.chromium.org/2461343002
Cr-Commit-Position: refs/heads/master@{#40668}

64421566

[wasm] Use correct parameter type in wasm-run-utils.h · c4119758

ahaas authored Oct 31, 2016

Due to a typo wasm-run-utils used the type of the first parameter for
all parameters. This caused problems if not all parameters had the same
type.

R=titzer@chromium.org, clemensh@chromium.org

Review-Url: https://codereview.chromium.org/2463743002
Cr-Commit-Position: refs/heads/master@{#40667}

c4119758

[stubs] Add FLAG_enable_slow_asserts check to CSA_SLOW_ASSERT · fde9039f

cbruni authored Oct 31, 2016

R=ishell@chromium.org
BUG=

Review-Url: https://codereview.chromium.org/2461463003
Cr-Commit-Position: refs/heads/master@{#40666}

fde9039f

[turbofan] Ship inlining into try blocks. · 639abf4f

bmeurer authored Oct 31, 2016

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

Review-Url: https://codereview.chromium.org/2460333002
Cr-Commit-Position: refs/heads/master@{#40665}

639abf4f

[turbofan] Handle inlining of API getters/setters. · 88366741

epertoso authored Oct 31, 2016

Similar to what crankshaft does, we introduce a mapcheck if necessary and a call to the CallApiCallbackStub.

BUG=

Review-Url: https://codereview.chromium.org/2458643002
Cr-Commit-Position: refs/heads/master@{#40664}

88366741

[compiler] Ship Ignition for all TurboFan code. · b4b436de

mstarzinger authored Oct 31, 2016

This enables Ignition unconditionally for all code that is destined for
optimization with TurboFan. This ensures all optimization attempts will
go through the BytecodeGraphBuilder and that the AstGraphBuilder pipe is
dried out in practice.

patch from issue 2427953002 at patchset 120001 (http://crrev.com/2427953002#ps120001)

R=mvstanton@chromium.org,rmcilroy@chromium.org

Review-Url: https://codereview.chromium.org/2453973004
Cr-Commit-Position: refs/heads/master@{#40663}

b4b436de

[turbofan] Properly deal with out-of-bounds fields in EscapeAnalysis. · 7201bad9

bmeurer authored Oct 31, 2016

Conflicting type feedback on Load/StoreICs can lead to out-of-bounds
field access, which is essentially dead code, but EscapeAnalysis was
confused about those. For now, mark the objects as escaping in these
cases, middle-term we want to deal better with this kind of compile-
time known dead code.

BUG=chromium:658185,v8:4586
R=jarin@chromium.org

Review-Url: https://codereview.chromium.org/2459273002
Cr-Commit-Position: refs/heads/master@{#40662}

7201bad9

29 Oct, 2016 2 commits

[wasm] Support for restricted table imports. · b7aff1ff

titzer authored Oct 29, 2016

This CL implements basic table import functionality.

Missing: growing of tables (WebAssembly.Grow) doesn't change dispatch tables
Missing: allowing larger table imports than minimum size

R=rossberg@chromium.org,bradnelson@chromium.org
BUG=v8:5507

Review-Url: https://codereview.chromium.org/2454503005
Cr-Commit-Position: refs/heads/master@{#40661}

b7aff1ff

[inspector] migrate scriptParsed and getCompiledScripts to native · f7f591ab

kozyatinskiy authored Oct 29, 2016

* Introduced DebugInterface::Script - native wrapper on internal::Script.
* Migrated getCompildScripts and scriptParsed to native wrapper.

BUG=chromium:652939, v8:5510
R=yangguo@chromium.org,dgozman@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Review-Url: https://codereview.chromium.org/2449213002
Cr-Commit-Position: refs/heads/master@{#40660}

f7f591ab

28 Oct, 2016 15 commits

[Tracing] Implement TracingCategoryObserver. · 6df8096a

lpy authored Oct 28, 2016

This patch implements TracingCategoryObserver to set global flag when a V8
specific category is enabled. Previously, we set a global flag each time when we
encounter a top level trace event, and use it as a global check. With this
patch, we can set a group of flags when tracing is enabled; besides, we make
V8 tracing feature use V8 flags instead of defining its own flag in a messy way.

With this patch, whatever V8 flag we want to imply in tracing, we define another
integer flag, and the original V8 flag will set it to 0x01 when passing by
commandline, tracing will set it to 0x10 when we start tracing and reset the bit
when we stop tracing.

Review-Url: https://codereview.chromium.org/2436273002
Cr-Commit-Position: refs/heads/master@{#40659}

6df8096a

[profiler] Emit runtime call stats into sampling profile · aee3542f

alph authored Oct 28, 2016

These are added to the sampler stack trace when RCS are
enabled.
Resource name for a RCS frame is reported as "V8Runtime".
Counter names match ones from src/counters.h

BUG=chromium:660428

Review-Url: https://codereview.chromium.org/2461003002
Cr-Commit-Position: refs/heads/master@{#40658}

aee3542f

[test] Fix a few flag configurations to make fuzzer happy. · 9b25707b

machenbach authored Oct 28, 2016

BUG=chromium:660457
TBR=mbarbella@chromium.org, yangguo@chromium.org

Review-Url: https://codereview.chromium.org/2464453002
Cr-Commit-Position: refs/heads/master@{#40657}

9b25707b

Revert of [wasm] Support for restricted table imports. (patchset #7 id:120001... · bc0ee727

machenbach authored Oct 28, 2016

Revert of [wasm] Support for restricted table imports. (patchset #7 id:120001 of https://codereview.chromium.org/2454503005/ )

Reason for revert:
GC stress failures:
https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20GC%20Stress%20-%20custom%20snapshot/builds/8857

Original issue's description:
> [wasm] Support for restricted table imports.
>
> This CL implements basic table import functionality.
>
> Missing: growing of tables (WebAssembly.Grow) doesn't change dispatch tables
> Missing: allowing larger table imports than minimum size
>
> R=rossberg@chromium.org,bradnelson@chromium.org
> BUG=v8:5507

TBR=bradnelson@chromium.org,rossberg@chromium.org,titzer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5507

Review-Url: https://codereview.chromium.org/2456193006
Cr-Commit-Position: refs/heads/master@{#40656}

bc0ee727

[inspector] use own copy of third_party/inspector_protocol · d01dc6dc

kozyatinskiy authored Oct 28, 2016

When Chromium and V8 use the same version of inspector_protocol, any protocol change takes at least 4 patches and 3 waiting for rolls.
To simplify this process we need to have two diffrent versions of inspector_protocol in Chromium and V8. Current state of inspector_protocol was extracted into separate repository [1]. This CL puts last version of inspector_protocol into third_party/inspector_protocol and removes dependency on inspector_protocol in Webkit.

[1] https://chromium.googlesource.com/deps/inspector_protocol/

BUG=chromium:637032
R=dgozman@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Review-Url: https://codereview.chromium.org/2447323002
Cr-Commit-Position: refs/heads/master@{#40655}

d01dc6dc

[wasm] make WebAssembly.Instance require a WebAssembly.Memory · 96d12ac6

eholk authored Oct 28, 2016

BUG=

Review-Url: https://codereview.chromium.org/2460773003
Cr-Commit-Position: refs/heads/master@{#40654}

96d12ac6

[Tracing] Fix inaccurate timer calculation in runtime statistics. · 92d9a56a

lpy authored Oct 28, 2016

Previously we reset runtime counters and dump them when we enter, exit top level
trace events respectively. However, there is gap between two top level trace
events and runtime counters may be activated, resetting the counters makes the
accumulated time inaccurate, and we may end up with negative time due to the
nature of how we accumulate time.

This patch fixes this problem by only resetting counters when there's no
counters active, and before dump counters, we traverse current active counters
to calculate their time, and then restart their timer.

BUG=chromium:658145

Review-Url: https://codereview.chromium.org/2457523002
Cr-Commit-Position: refs/heads/master@{#40653}

92d9a56a

[wasm] Support for restricted table imports. · 404e2154

titzer authored Oct 28, 2016

This CL implements basic table import functionality.

Missing: growing of tables (WebAssembly.Grow) doesn't change dispatch tables
Missing: allowing larger table imports than minimum size

R=rossberg@chromium.org,bradnelson@chromium.org
BUG=v8:5507

Review-Url: https://codereview.chromium.org/2454503005
Cr-Commit-Position: refs/heads/master@{#40652}

404e2154

[heap] Check that sweeping is completed when collecting evacuation · b0b1d7a9

ulan authored Oct 28, 2016

candidates.

BUG=chromium:650314

Review-Url: https://codereview.chromium.org/2459003002
Cr-Commit-Position: refs/heads/master@{#40651}

b0b1d7a9

Revert of [heap] Concurrent store buffer processing. (patchset #13 id:260001... · d800644c

machenbach authored Oct 28, 2016

Revert of [heap] Concurrent store buffer processing. (patchset #13 id:260001 of https://codereview.chromium.org/2453673003/ )

Reason for revert:
Seems to block rolling:
https://codereview.chromium.org/2447393005/

Original issue's description:
> [heap] Concurrent store buffer processing.
>
> BUG=chromium:648973, chromium:648568

TBR=ulan@chromium.org,mlippautz@chromium.org,hpayer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:648973, chromium:648568

Review-Url: https://codereview.chromium.org/2449853010
Cr-Commit-Position: refs/heads/master@{#40650}

d800644c

[modules] Remove outdated TODO. · a017715d

neis authored Oct 28, 2016

R=gsathya@chromium.org
BUG=v8:1569

Review-Url: https://codereview.chromium.org/2460883002
Cr-Commit-Position: refs/heads/master@{#40649}

a017715d

Add a native data property that replaces itself with a real data property · f0742150

jochen authored Oct 28, 2016

This is useful for things that don't ever change, but we don't want to
eagerly compute the result.

Doing this from the embedder is difficult, using DefineOwnProperty would
read the property to get the property descriptor, creating an endless
recursion.

R=verwaest@chromium.org,haraken@chromium.org
BUG=

Review-Url: https://codereview.chromium.org/2449783006
Cr-Commit-Position: refs/heads/master@{#40648}

f0742150

[heap] Do not wait for marking deque uncommit task on tear down. · e12f9a11

ulan authored Oct 28, 2016

Cancelable task already guarantees proper waiting.

BUG=

Review-Url: https://codereview.chromium.org/2462553002
Cr-Commit-Position: refs/heads/master@{#40647}

e12f9a11

[wasm] Start function must not have a return value. · aff3ef21

ahaas authored Oct 28, 2016

R=titzer@chromium.org
TEST=mjsunit/wasm/start-function

Review-Url: https://codereview.chromium.org/2458983002
Cr-Commit-Position: refs/heads/master@{#40646}

aff3ef21

Disable Api call optimizations for runtime call stats · e52020c7

jochen authored Oct 28, 2016

Otherwise, we don't account for callbacks from optimized code anymore.

R=verwaest@chromium.org,cbruni@chromium.org
BUG=

Review-Url: https://codereview.chromium.org/2459573002
Cr-Commit-Position: refs/heads/master@{#40645}

e52020c7