AuroraMiddleware/v8
1
0
Fork 0
Code Issues 2 Pull Requests Releases Wiki Activity
25,309 Commits 1 Branch 0 Tags 839 MiB
ba67a42629
Commit Graph

377 Commits

Author SHA1 Message Date
bmeurer
6b3c070db6 [runtime] Sanitize %NewClosure runtime entries. 
There are now two runtime entries %NewClosure and %NewClosure_Tenured,
with the same signature (one parameter, the SharedFunctionInfo, and the
context of the caller).

Also remove the HFunctionLiteral special case instruction from Crankshaft,
as HCallWithDescriptor with FastNewClosureStub or HCallRuntime with
either %NewClosure or %NewClosure_Tenured can easily do that for you.

Also remove the redundant context parameter from the JSCreateClosure
operator, because every JS operator already takes a context input.

CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_nosnap_dbg

Review URL: https://codereview.chromium.org/1329293003

Cr-Commit-Position: refs/heads/master@{#30671}
 2015-09-10 08:36:15 +00:00
ishell
29ebcc3205 Crankshaft is now able to compile top level code even if there is a ScriptContext. 
This CL introduces HPrologue instruction which does the context allocation work and supports deoptimization.

Review URL: https://codereview.chromium.org/1317383002

Cr-Commit-Position: refs/heads/master@{#30496}
 2015-09-01 07:07:05 +00:00
bmeurer
b4c7399464 [runtime] Remove the redundant %_IsObject intrinsic. 
%_IsObject(foo) is equivalent to typeof foo === 'object' and has
exactly the same optimizations, so there's zero need for %_IsObject
in our code base.

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/1313903003

Cr-Commit-Position: refs/heads/master@{#30380}
 2015-08-26 11:28:06 +00:00
bmeurer
5d875a57fa Correctify instanceof and make it optimizable. 
The previous hack with HInstanceOfKnownGlobal was not only slower,
but also very brittle and required a lot of weird hacks to support it. And
what's even more important it wasn't even correct (because a map check
on the lhs is never enough for instanceof).

The new implementation provides a sane runtime implementation
for InstanceOf plus a fast case in the InstanceOfStub, combined with
a proper specialization in the case of a known global in CrankShaft,
which does only the prototype chain walk (coupled with a code
dependency on the known global).

As a drive-by-fix: Also fix the incorrect Object.prototype.isPrototypeOf
implementation.

BUG=v8:4376
LOG=y

Review URL: https://codereview.chromium.org/1304633002

Cr-Commit-Position: refs/heads/master@{#30342}
 2015-08-25 04:48:54 +00:00
jochen
fded08f694 Reland of "Remove ExternalArray, derived types, and element kinds" 
Original issue's description:
> Remove ExternalArray, derived types, and element kinds
>
> BUG=v8:3996
> R=jarin@chromium.org, mvstanton@chromium.org, bmeurer@chromium.org
> LOG=y
>
> Committed: https://crrev.com/607ef7c6009a24ebf195b4cab7b0b436c5afd21c
> Cr-Commit-Position: refs/heads/master@{#29872}

BUG=v8:3996
R=bmeurer@chromium.org
LOG=y

Review URL: https://codereview.chromium.org/1262583002

Cr-Commit-Position: refs/heads/master@{#29893}
 2015-07-28 09:29:55 +00:00
bmeurer
5dff4bdff0 [stubs] Don't pass name to Load/StoreGlobalViaContext stubs. 
No need to pass the name explicitly to the stubs; the runtime can
extract the name from the ScopeInfo (the extension of the
ScriptContext) on-demand easily without any performance impact.

Review URL: https://codereview.chromium.org/1259963002

Cr-Commit-Position: refs/heads/master@{#29886}
 2015-07-28 06:04:29 +00:00
machenbach
814048a04f Revert of Remove ExternalArray, derived types, and element kinds (patchset #5 id:80001 of https://codereview.chromium.org/1254623002/) 
Reason for revert:
[Sheriff] Breaks several layout tests, e.g.:
http://build.chromium.org/p/client.v8.fyi/builders/V8-Blink%20Linux%2032/builds/1067

Several output lines change from PASS to FAIL. If the changes are intended, please land a needsmanualrebaseline change in blink first.

Original issue's description:
> Remove ExternalArray, derived types, and element kinds
>
> BUG=v8:3996
> R=jarin@chromium.org, mvstanton@chromium.org, bmeurer@chromium.org
> LOG=y
>
> Committed: https://crrev.com/607ef7c6009a24ebf195b4cab7b0b436c5afd21c
> Cr-Commit-Position: refs/heads/master@{#29872}

TBR=bmeurer@chromium.org,hpayer@chromium.org,jarin@chromium.org,mvstanton@chromium.org,jochen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:3996

Review URL: https://codereview.chromium.org/1257223002

Cr-Commit-Position: refs/heads/master@{#29883}
 2015-07-27 20:32:16 +00:00
jochen
607ef7c600 Remove ExternalArray, derived types, and element kinds 
BUG=v8:3996
R=jarin@chromium.org, mvstanton@chromium.org, bmeurer@chromium.org
LOG=y

Review URL: https://codereview.chromium.org/1254623002

Cr-Commit-Position: refs/heads/master@{#29872}
 2015-07-27 13:19:36 +00:00
ishell
cc66a1c64e Crankshaft part of the 'loads and stores to global vars through property cell shortcuts' feature. 
BUG=chromium:510738
LOG=N

Review URL: https://codereview.chromium.org/1228113008

Cr-Commit-Position: refs/heads/master@{#29743}
 2015-07-20 08:49:28 +00:00
ishell
fec3c9cba6 TypeofMode replaces TypeofState and ContextualMode. 
NON_CONTEXTUAL ~> INSIDE_TYPEOF
CONTEXTUAL ~> NOT_INSIDE_TYPEOF

Review URL: https://codereview.chromium.org/1227893005

Cr-Commit-Position: refs/heads/master@{#29611}
 2015-07-13 13:39:43 +00:00
danno
7015fd2053 Make context register implicit for CallInterfaceDescriptors 
Up until now the context register was listed explicitly in each stub's
CallInterfaceDescriptor. This was problematic, because it was listed
first in the list of register parameters--which is fine for Crankshaft,
which is more or less built to handle the context as the first
parameter-- but not ideal for TurboFan, which adds the context at
the end of all function parameters. Now the context register is no
longer in the register list and can be handled appropriately by both
compilers. Specifically, this allows the FunctionType specified for
each CallInterfaceDescriptor to exactly match the parameter register
list.

Review URL: https://codereview.chromium.org/1211333003

Cr-Commit-Position: refs/heads/master@{#29402}
 2015-07-01 08:45:12 +00:00
mvstanton
8a3cf4ecef VectorICs: Lithium support for vector-based stores. 
BUG=

Review URL: https://codereview.chromium.org/1209903003

Cr-Commit-Position: refs/heads/master@{#29310}
 2015-06-26 07:53:38 +00:00
conradw
dd85444951 [strong] Refactor ObjectStrength into a replacement for strong boolean args 
Boolean "is_strong" parameters have begun to proliferate across areas where
strong mode semantics are different. This CL repurposes the existing
ObjectStrength enum as a replacement for them.

BUG=v8:3956
LOG=N

Review URL: https://codereview.chromium.org/1144183004

Cr-Commit-Position: refs/heads/master@{#28839}
 2015-06-08 12:18:15 +00:00
mvstanton
323ced9e27 Now that vector ics are established for load, keyed load and call ics, let's remove dead code behind the flag. 
BUG=

Review URL: https://codereview.chromium.org/1129853002

Cr-Commit-Position: refs/heads/master@{#28422}
 2015-05-15 13:25:22 +00:00
conradw
03ef40b46c [strong] Disallow implicit conversions for comparison 
Implements the strong mode proposal's restrictions on implicit conversions
for the binary >, >=, <, and <= operators.

BUG=v8:3956
LOG=N

Review URL: https://codereview.chromium.org/1130283002

Cr-Commit-Position: refs/heads/master@{#28370}
 2015-05-12 15:23:50 +00:00
mvstanton
3bce9c3afb New hydrogen instruction to reduce cost of growing an array on keyed stores. 
HMaybeGrowElements moves the situation where you actually have to grow
into deferred code. This means crankshaft doesn't have to spill registers
just to make the bounds comparison to see if it'll need to grow or not.

It makes the growing case a bit more expensive, but reduces the cost of
the general case.

BUG=chromium:484025
LOG=N

Review URL: https://codereview.chromium.org/1124443004

Cr-Commit-Position: refs/heads/master@{#28359}
 2015-05-12 08:47:16 +00:00
conradw
97bee8e964 [strong] Fix inlining issue 
The Hydrogen representation for binops was never changed to care about the
language mode. We thought this was ok, but it turns out we need to keep track
of it to make sure inlining doesn't mess with the "strongness" of binops.

Also added more rigorous inlining testing.

BUG=v8:3956
LOG=N

Review URL: https://codereview.chromium.org/1123043002

Cr-Commit-Position: refs/heads/master@{#28253}
 2015-05-06 10:40:38 +00:00
jochen
a2e6f970c7 Add HArrayBufferNotNeutered instruction 
This instruction can be hoisted out of loops even though it contains a branch.

BUG=v8:3996
R=bmeurer@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/1108313003

Cr-Commit-Position: refs/heads/master@{#28109}
 2015-04-28 13:43:03 +00:00
Ross McIlroy
063fc25122 Replace OVERRIDE->override and FINAL->final since we now require C++11. 
R=jochen@chromium.org

Review URL: https://codereview.chromium.org/1088993003

Cr-Commit-Position: refs/heads/master@{#27937}
 2015-04-20 13:08:14 +00:00
mvstanton
e0844a24d3 Remove unnecessary options from HTailCallThroughMegamorphicCache 
These options were added for a hydrogen code stub version of
the VectorIC dispatcher, which was discontinued.

R=verwaest@chromium.org
BUG=

Review URL: https://codereview.chromium.org/1081883002

Cr-Commit-Position: refs/heads/master@{#27793}
 2015-04-13 16:23:23 +00:00
verwaest
16c8485a35 Remove PropertyCell space 
Replaces StoreGlobalCell / LoadGlobalCell with NamedField variants that use write barriers.
BUG=

Review URL: https://codereview.chromium.org/1016803002

Cr-Commit-Position: refs/heads/master@{#27269}
 2015-03-18 11:43:58 +00:00
marja
c7851da4ae Introduce LanguageMode, drop StrictMode. 
This enables adding more language modes in the future.

For maximum flexibility, LanguageMode is a bitmask, so we're not restricted to
use a sequence of language modes which are progressively stricter, but we can
express the language mode as combination of features.

For now, LanguageMode can only be "sloppy" or "strict", and there are
STATIC_ASSERTS in places which need to change when more modes are added.

LanguageMode is a bit like the old LanguageMode when "extended" mode was still
around (see https://codereview.chromium.org/8417035 and
https://codereview.chromium.org/181543002 ) except that it's transmitted through
all the layers (there's no StrictModeFlag).

BUG=

Review URL: https://codereview.chromium.org/894683003

Cr-Commit-Position: refs/heads/master@{#26419}
 2015-02-04 09:34:26 +00:00
mvstanton
7d363783e1 Continue learning for calls in crankshaft. 
The type feedback vector makes this easy to do.
This is a re-land of https://codereview.chromium.org/868453005/
with a fix for the DCHECK failure.

Review URL: https://codereview.chromium.org/885593002

Cr-Commit-Position: refs/heads/master@{#26302}
 2015-01-28 09:03:19 +00:00
mvstanton
f605f1c223 Revert of Continue learning for calls in optimized code when we have no type feedback. (patchset #4 id:60001 of https://codereview.chromium.org/868453005/) 
Reason for revert:
Serializer tests broke. Need to debug and fix.

Original issue's description:
> Continue learning for calls in optimized code when we have no type feedback.
>
> Based on CL https://codereview.chromium.org/871063002/ which needs to land first.
>
> BUG=
>
> Committed: https://crrev.com/f5f2692b5ff70ac3cd06a903b7846174b97a2e55
> Cr-Commit-Position: refs/heads/master@{#26292}

TBR=verwaest@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=

Review URL: https://codereview.chromium.org/878103002

Cr-Commit-Position: refs/heads/master@{#26293}
 2015-01-27 13:07:09 +00:00
mvstanton
f5f2692b5f Continue learning for calls in optimized code when we have no type feedback. 
Based on CL https://codereview.chromium.org/871063002/ which needs to land first.

BUG=

Review URL: https://codereview.chromium.org/868453005

Cr-Commit-Position: refs/heads/master@{#26292}
 2015-01-27 12:33:07 +00:00
Benedikt Meurer
9eace97bba Use signaling NaN for holes in fixed double arrays. 
TEST=mjsunit,cctest,unittests
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/863633002

Cr-Commit-Position: refs/heads/master@{#26180}
 2015-01-21 08:52:25 +00:00
Michael Stanton
22302b5179 Hydrogen code stubs for vector-based ICs. 
This patch finally allows running and passing tests with vector-based
Load and KeyedLoad ICs.

BUG=
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/767743002

Cr-Commit-Position: refs/heads/master@{#25800}
 2014-12-12 13:56:35 +00:00
svenpanne
71bb00e261 Consistently use only one of virtual/OVERRIDE/FINAL. 
FINAL implies OVERRIDE, which in turn implies virtual, so there's no need to use
more than one of these. The Google C++ style guide even requires this, see
http://google-styleguide.googlecode.com/svn/trunk/cppguide.html#Inheritance.

While we're here, port r24662 to x87.

The net result is that v8 compiles again with a current clang.

BUG=v8:3753
LOG=y

Review URL: https://codereview.chromium.org/797943002

Cr-Commit-Position: refs/heads/master@{#25792}
 2014-12-12 10:44:25 +00:00
yangguo@chromium.org
43538e57a4 Refactor bailout reasons and disable optimization in more cases. 
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/596783002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@24161 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-09-24 07:08:27 +00:00
mvstanton@chromium.org
6e1ebdcbc3 To aid vector-based load ic work, we need to be able to handle 
the megamorphic load case in hydrogen. A simple approach is to
wrap the probe activity in a hydrogen instruction.

The instruction is novel in that it always tail-calls away.

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/535873002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@23772 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-09-08 12:51:29 +00:00
mvstanton@chromium.org
c2fe5b2b68 Make concrete classes for individual call descriptors. The ic-convention classes that hold register specifications are merged into these new call descriptor classes, which should represent a final home for that information. 
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/527093002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@23639 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-09-03 10:51:51 +00:00
bmeurer@chromium.org
30f18f0f7d Use Chrome compatible naming for compiler specifics. 
Less useless creativity is best creativity!

R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/526223002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@23579 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-09-02 07:07:52 +00:00
mvstanton@chromium.org
3a7b5b44c1 InterfaceDescriptor becomes CallInterfaceDescriptor. 
There was no difference between these two classes in a hierarchical
relationship.

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/532473002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@23560 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-09-01 12:15:25 +00:00
mstarzinger@chromium.org
44247036a7 Fix newly discovered presubmit errors. 
R=jochen@chromium.org

Review URL: https://codereview.chromium.org/462033002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@23081 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-08-12 13:33:35 +00:00
bmeurer@chromium.org
d07a2eb806 Rename ASSERT* to DCHECK*. 
This way we don't clash with the ASSERT* macros
defined by GoogleTest, and we are one step closer
to being able to replace our homegrown base/ with
base/ from Chrome.

R=jochen@chromium.org, svenpanne@chromium.org

Review URL: https://codereview.chromium.org/430503007

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@22812 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-08-04 11:34:54 +00:00
danno@chromium.org
a1383e2250 Land the Fan (disabled) 
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/426233002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@22709 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-07-30 13:54:45 +00:00
yangguo@chromium.org
e7e89b32f1 Fix typo in LMathFround on ia32. 
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/428043003

Patch from Chunyang Dai <chunyang.dai@intel.com>.

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@22699 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-07-30 09:55:21 +00:00
yangguo@chromium.org
a8a02a51f1 Inline Math.fround in optimized code. 
R=svenpanne@chromium.org
BUG=v8:3469
LOG=N

Review URL: https://codereview.chromium.org/425943002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@22665 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-07-29 11:34:08 +00:00
mvstanton@chromium.org
6d3fc8a322 Introduce FLAG_vector_ics. 
When FLAG_vector_ics is true, then AST nodes that use Load and KeyedLoad ICs
will allocate a type vector slot to store feedback information. Full codegen
will emit a load of the slot into a register if the flag is on.

Support is incomplete, right now the IC doesn't know how to use the feedback
slot.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/398053002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@22500 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-07-21 11:19:56 +00:00
mvstanton@chromium.org
54636ae4e2 StubCallInterfaceDescriptor and CallInterfaceDescriptor are unified under a base class InterfaceDescriptor. 
Handling of the context register had to be massaged to effect the unification. This will make it easier to call hydrogen code stubs directly from crankshaft.

R=danno@chromium.org

Review URL: https://codereview.chromium.org/384403002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@22448 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-07-17 11:50:04 +00:00
mstarzinger@chromium.org
fec6e62dfb Check alpha-sorting of includes during presubmit. 
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/333013002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21894 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-20 08:40:11 +00:00
jarin@chromium.org
1a138034bd Fix LStoreCodeEntry to treat the code_object parameter as input. 
The bug seems to trigger only with array_index_dehoisting off in
the snapshot, so it is hard to test for (ideas?).

R=mvstanton@chromium.org
BUG=

Review URL: https://codereview.chromium.org/347523002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21885 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-18 11:17:52 +00:00
ulan@chromium.org
c8b2fa454a Preliminary support for block contexts in hydrogen. 
Patch from Steven Keuchel <keuchel@chromium.org>

BUG=v8:2198
LOG=N
TEST=mjsunit/harmony/block-let-crankshaft.js
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/307593002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21684 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-05 07:33:01 +00:00
ishell@chromium.org
55443af648 Fixed lint errors caused by "runtime/references" rule (Is this a non-const reference?) and the rule itself is restored. 
BUG=v8:3326
LOG=N
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/314723002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21651 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-03 15:45:38 +00:00
jochen@chromium.org
56a486c322 Use full include paths everywhere 
- this avoids using relative include paths which are forbidden by the style guide
- makes the code more readable since it's clear which header is meant
- allows for starting to use checkdeps

BUG=none
R=jkummerow@chromium.org, danno@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/304153016

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21625 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-03 08:12:43 +00:00
bmeurer@chromium.org
7d14f34a1b Inline Array.shift() fast path instead of using a code stub. 
TEST=mjsunit/array-shift,mjsunit/array-shift2,mjsunit/array-shift3
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/308793010

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21592 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-02 07:02:24 +00:00
bmeurer@chromium.org
a52f569f93 Simplify known successor block lowering. 
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/308593003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21589 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-06-02 05:33:35 +00:00
danno@chromium.org
0f929e7a54 Small changes in preparation for Hydrogen-generated KeyedLoadGeneric 
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/303583002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21535 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 15:15:49 +00:00
bmeurer@chromium.org
cdcee29ef7 LoadUint32() doesn't need a scratch register. 
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/293363005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21476 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 06:41:21 +00:00
alexandre.rames@arm.com
18412a83b0 Provide a helper to generate multiple Lithium instructions for one Hydrogen instruction. 
R=jkummerow@chromium.org, ulan@chromium.org

Review URL: https://codereview.chromium.org/296993002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21465 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 13:15:07 +00:00