Commit Graph

850 Commits

Author SHA1 Message Date
ager@chromium.org
8c6c273236 Fix issues with using defineProperty on the global proxy object.
Review URL: http://codereview.chromium.org/6452004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6683 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 16:31:58 +00:00
ricow@chromium.org
f64966085e x64: Add MulI and DivI to lithium instructions.
Review URL: http://codereview.chromium.org/6448001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6681 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 14:37:50 +00:00
lrn@chromium.org
2f32f27e8f Correct propagation of exceptions from setters.
BUG=v8:1105
TEST=test/mjsunit/regress/regress-1105.js

Review URL: http://codereview.chromium.org/6451003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6680 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 14:04:27 +00:00
kmillikin@chromium.org
bf3c3eb9cb Fix a possible duplicate AST ID for deoptimization.
For redeclarations of variables that alias the parameters in functions
using arguments, we need to avoid re-visiting the shared variable
rewrite.

BUG=v8:1104

Review URL: http://codereview.chromium.org/6453004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6679 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 14:00:22 +00:00
ricow@chromium.org
20f2c1c98a Make sure that we do not call is_extensible on the global proxy.
When calling Object.isExtensible we did not do a check for the global
js proxy. This caused the check on the extensible bit on the map to
return true, even when the bit was set to false on the global js
object.


Review URL: http://codereview.chromium.org/6450003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6677 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 13:09:07 +00:00
ricow@chromium.org
81787f986b Make sure that we never call prevent extension on the global proxy,
but instead call this on the global object.

BUG: 1103

Review URL: http://codereview.chromium.org/6454001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6675 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 12:41:16 +00:00
lrn@chromium.org
48fadffcc4 Fix bug in JSON.parse for objects containing "__proto__" as key.
It added the __proto__ key as a normal key, which made it visible
in enumeration, while reading still hit the hard-coded accessor.

Review URL: http://codereview.chromium.org/6451002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6674 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-08 11:38:15 +00:00
peterhal@chromium.org
39957aa741 Issue 117 - strict mode and future reserved words
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6653 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-04 18:36:37 +00:00
mmaly@chromium.org
87233c49c8 Pass strict mode to eval.
Code review feedback.

Code Review URL: http://codereview.chromium.org/6286043/

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6652 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-04 18:15:49 +00:00
vegorov@chromium.org
10f715e3ff Restore context after LApplyArguments.
BUG=v8:1099
TEST=test/mjsunit/regress/regress-1099.js

Review URL: http://codereview.chromium.org/6246106

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6649 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-04 15:42:02 +00:00
peterhal@chromium.org
c894b1f317 Fix bugs 992, 1083 and 1092
My previous patch added an assert which uncovered 1092 in the sputnik tests.
This patch adds the fix for 1092, which is to ensure that NormalizeProperties
does not get called for a JSGlobalProxy along all code paths.

Add sputnik tests to .gitignore.

BUG=
TEST=

Review URL: http://codereview.chromium.org/6286060

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6627 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-03 19:29:10 +00:00
ricow@chromium.org
a2aa84873e Add regression test for the deoptimizer immediately followed by gc bug.
In addition to the regression test I changed the gc-extension to take
a boolean flag specifying if compaction should be used (default is
false, existing tests will not change behaviour)

The regression test is disabled on arm and x64 with crankshaft
enabled. I made a bug to track this:
http://code.google.com/p/v8/issues/detail?id=1094


Review URL: http://codereview.chromium.org/6312118

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6619 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-03 13:47:27 +00:00
danno@chromium.org
a2fb4a12bb Create specialized code stubs for PixelArray loads.
Review URL: http://codereview.chromium.org/6287030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6614 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-03 12:50:50 +00:00
antonm@chromium.org
0da3dc3e43 Properly process getOwnPropertyDescriptor for elements on global proxy object.
We need to go down to actual global object to perform those operations.

Review URL: http://codereview.chromium.org/6246054

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6612 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-03 10:19:41 +00:00
kmillikin@chromium.org
ca936dae9e More of the fix for V8 issue 1079.
The arguments property of functions, if we find an optimized frame for
the function, is always a freshly allocated object.  We never try to
find an existing arguments object.

Review URL: http://codereview.chromium.org/6349050

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6581 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-02 15:08:29 +00:00
erik.corry@gmail.com
0097f005fd Fix code generation bug on ARM in classic codegen.
Review URL: http://codereview.chromium.org/6246045

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6579 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-02 14:14:55 +00:00
lrn@chromium.org
2d15eb9a90 Fix bug in object literals with large array indexes as strings.
Review URL: http://codereview.chromium.org/6410028

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6577 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-02 14:02:58 +00:00
kmillikin@chromium.org
f1149734fc Partial fix for V8 issue 1079.
Record a safepoint with a deoptimization id for throw in optimized code.  We
don't seem to much care what the AST ID is because we will not be using it
for lazy deoptimization (throw doesn't return to the point of throw).  For
hygiene we use the actual ID of the throw expression.  Throw is no longer a
control-flow instruction, but it's followed by an unconditional abnormal
exit.  This is required to insert a simulate between the throw and the exit.

Make our optimized treatment of Function.prototype.apply act like a call and
have side effects.  This ensures that it will get a lazy deoptimization
environment.  Use that deoptimization ID in the safepoint for the call.

Deleting a property was also missing a deoptimization ID, though there was a
deoptimization environment assigned to the instruction.  Record the
environment and use the deoptimization ID at the safepoint.

Review URL: http://codereview.chromium.org/6250105

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6576 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-02 13:55:29 +00:00
whesse@chromium.org
a5f94a4862 Fix Math.pow(-0, 0.5) and Math.pow(-0, -0.5). These are not equal to sqrt(-0) and 1/sqrt(-0). Add tests for these cases. Fixes V8 issue 1088.
BUG=1088
TEST=test/mjsunit/math-pow.js

Review URL: http://codereview.chromium.org/6368050

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6573 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-02 12:54:58 +00:00
vegorov@chromium.org
6751627615 Require typed input representation for HTypeof hydrogen instruction.
BUG=http://code.google.com/p/chromium/issues/detail?id=71647
TEST=test/mjsunit/regress/regress-71647.js

Review URL: http://codereview.chromium.org/6410025

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6566 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-02 09:52:57 +00:00
peterhal@chromium.org
5ca89179d7 Revert "Fix bugs 992 and 1083"
This reverts commit 6561 as the new assert caused failures in sputnik.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6562 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-01 20:08:01 +00:00
peterhal@chromium.org
9c89aa6dd9 Fix bugs 992 and 1083
Fixes JS portion of DefineOwnProperty when there is
an existing property and the new descriptor is generic.

Makes code follow spec steps more closely.

Fixes typo for check for unchanged enumerable in step 6.

Adds regression tests.
Fixes errors in object-define-property test

Don't normalize the JSGlobalProxy. Gets webkit http/tests/security/xss-DENIED-defineProperty.html working.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6561 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-01 17:08:14 +00:00
ager@chromium.org
471c0d2983 Avoid callbacks to user code during error formatting in a couple of
other situations.

Do not use overwritten Object.prototype.hasOwnProperty and
Array.prototype.pop. Do not use split and join in the error formatting
implementation. They are too big to control and their generality is
not needed.

Review URL: http://codereview.chromium.org/6287041

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6552 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-02-01 12:31:16 +00:00
mmaly@chromium.org
aa779b3842 Fix V8 bug 1084: allow "\0" in strict mode as valid escape sequence.
http://code.google.com/p/v8/issues/detail?id=1084

Code Review URL: http://codereview.chromium.org/6386014/

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6550 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-31 22:35:27 +00:00
antonm@chromium.org
a38a8ffab0 ArraySplice builtin should return empty array and not alter receiver if invoked with no arguments.
Review URL: http://codereview.chromium.org/6357025

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6540 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-31 14:54:53 +00:00
fschneider@chromium.org
4e7ddab6dc Fix a bug in the placement of minus-zero checks and in GVN.
1. The placement of checks for negative zero has to be computed after
all conversion instructions have been inserted. I separated the code
into its own phase.

2. GVN need to take instruction flags into account when comparing
instructions for redundancy.

Review URL: http://codereview.chromium.org/6260035

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6534 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-31 12:36:54 +00:00
ager@chromium.org
f8b74a1541 Avoid using Function.prototype.call in a number of places in our
builtins files. We should always use %_CallFunction for a couple of
reasons: it cannot be overwritten and it does not wrap basic types in
wrapper objects.

Review URL: http://codereview.chromium.org/6349018

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6524 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-28 10:33:10 +00:00
ager@chromium.org
dc61921bbf Fix reintroduction of global variables that have been deleted.
Deletion of global properties puts 'the hole' in the global property
cell and updates the property details in the property dictionary with
the information that the property has been deleted. When setting
global properties that have been deleted in generated code we just
store the new value in the global property cell. This does not update
the property details in the property dictionary. Therefore, it looks
like the property is not there eventhough it was just reintroduced.

Perform 'the hole' checks in generated code for global property stores
and bail out of ICs and optimized code if storing to a property cell
that contains 'the hole'.

Review URL: http://codereview.chromium.org/6306014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6508 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-27 08:35:39 +00:00
mmaly@chromium.org
d07f1d62ff Strict mode eval/arguments LHS.
Review URL: http://codereview.chromium.org/6335013/

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6503 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-26 19:21:46 +00:00
mmaly@chromium.org
55400c2454 Strict mode: function constructor tests.
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6501 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-26 18:10:26 +00:00
mmaly@chromium.org
2b47762e3a Strict mode object property validation.
Review URL: http://codereview.chromium.org/6335010/

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6475 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-25 18:42:35 +00:00
mmaly@chromium.org
f5a2860534 Strict mode parameter validation.
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6474 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-25 17:21:45 +00:00
kmillikin@chromium.org
9cfbb240a0 Update regression test.
With SVN r6465 (reverting changes to scopes), a regression test for
deleting parameter variables has to change to reflect a semantic
change.  It is now again possible to delete parameters from a function
that uses 'with' or 'try...catch'.

Review URL: http://codereview.chromium.org/6307014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6466 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-25 13:32:36 +00:00
kmillikin@chromium.org
013860d42e Revert "Reapply change to with/arguments interaction."
Revert this change again.  Somewhat mysteriously we sometimes get empty
contexts that we do not expect in the context chain.

Review URL: http://codereview.chromium.org/6372013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6465 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-25 13:01:45 +00:00
vitalyr@chromium.org
331b482332 Fix bug 1070: set correct holder for primitive checks.
Code generated for checks starting with primitive receivers skips one step
in the usual prototype checking algorithm, so the holder must always be set.

Not setting the holder did not cause an immediate failure because our
primitives have additional hidden prototypes before the real prototypes.
These extra objects in the chain usually contain no properties and so
allowed the right holders to be selected.

Review URL: http://codereview.chromium.org/6353014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6459 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-25 12:21:03 +00:00
ricow@chromium.org
d5cab38a33 Add support for deoptimization on x64.
I did not take out the code relating to osr from the generate method
since this makes it easier to compare to ia32 (we will abort anyway when we hit the osr code so there should be no issues with having this in)
 

Review URL: http://codereview.chromium.org/6390001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6449 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-25 07:57:56 +00:00
mmaly@chromium.org
8d79f3a788 Strict mode octal literals.
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6446 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-24 18:13:18 +00:00
kmillikin@chromium.org
9c2d52eb0e Fix a bug in delete for lookup slots.
The function Runtime_LookupContext searches the context chain for a
LOOKUP slot and returns the object holding the slot.  It returned the
global context if the slot was not found or if it was found in a
function's context or arguments object.  This is not the correct
object to use for 'delete'.

Since this lookup function is only ever used when deleting LOOKUP
slots (those that have to go through a with or a scope with eval), it
is simply replaced with a Runtime_DeleteContextSlot function that does
the appropriate thing for all kinds of context lookups.

This fixes Chromium bug 70066.
http://code.google.com/p/chromium/issues/detail?id=70066

Review URL: http://codereview.chromium.org/6280013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6442 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-24 14:03:30 +00:00
ager@chromium.org
bb9c9fe92c Change recursive error printing to just replace recursively
encountered error objects with the empty string. This actually does
match the Safari behaviour.

Review URL: http://codereview.chromium.org/6259010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6430 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-21 14:11:35 +00:00
ager@chromium.org
ca14680556 Return the empty string when formatting recursive error messages. This
matches the behavior of Safari and Firefox. Our old behavior was to
throw a stack overflow exception.

BUG=crbug.com/70334
TEST=mjsunit/cyclic-error-to-string.js

Review URL: http://codereview.chromium.org/6261018

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6428 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-21 11:44:29 +00:00
kmillikin@chromium.org
8a9eeaf8da Reapply change to with/arguments interaction.
This is svn r6415 with a bugfix.

The bug was that we can have scopes with 0 heap slots (functions with
such scopes do not need a local context) and scopes with at least
Context::MIN_CONTEXT_SLOTS.  The code generator only allocated a local
context if there were strictly more than Context::MIN_CONTEXT_SLOTS.
Before the change to with/arguments, it was impossible to actually get
a function scope with exactly Context::MIN_CONTEXT_SLOTS because every
such scope included a heap-allocated arguments (and arguments shadow)
slot.

Review URL: http://codereview.chromium.org/6248014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6426 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-21 11:28:35 +00:00
vitalyr@chromium.org
7d96f2d40f Fix Smi::IsValid assert in StringCharCodeAt deferred code.
Review URL: http://codereview.chromium.org/6303013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6424 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-21 08:30:13 +00:00
mmaly@chromium.org
4cb51b51e6 First part of strict mode.
- var eval | arguments
- catch (eval | arguments)
- 'with' is disabled
- function can't be named eval or arguments

Add FLAG_strict_mode

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6422 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-20 18:51:47 +00:00
kmillikin@chromium.org
6b89b5975f Revert "Make 'with' mark only variables occurring in the body as used."
This reverts SVN revision 6415 due to layout test failures.

TBR=ricow

Review URL: http://codereview.chromium.org/6374006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6418 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-20 12:58:57 +00:00
kmillikin@chromium.org
7f43cf0539 Make 'with' mark only variables occurring in the body as used.
Before, we conservatively marked every variable in a scope as used if the
scope contained 'with'.  Instead, just mark the variables occurring in the
body of the with.  This avoids marking 'arguments' as used whenever 'with'
occurs, which incurs an extra performance penalty (a use of arguments is
seen as an instruction to redirect all parameter accesses to the arguments
object).

Review URL: http://codereview.chromium.org/6357007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6415 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-20 12:32:43 +00:00
kmillikin@chromium.org
70910af7a0 Fix an assertion failure in the full code generator.
We hit an assertion failure when we tried to record the AST ID of
the (shared) .arguments variable proxy more than once.  This was hit
when we had multiple calls to the same parameter in a function that
used the arguments object.  The fix is to not visit the subexpressions
of the (shared) property access expression.

BUG=1060

Review URL: http://codereview.chromium.org/6368007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6404 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-19 15:26:54 +00:00
antonm@chromium.org
49144ee6b6 Properly create variables to access outer arguments and function names.
Review URL: http://codereview.chromium.org/6266007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6380 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-19 08:16:17 +00:00
mmaly@chromium.org
d6f88b6688 Revert strict mode (Mozilla test failure).
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6374 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-18 23:01:50 +00:00
mmaly@chromium.org
b4c88d5de4 First part of ES5 strict mode.
- var eval | arguments
- catch (eval | arguments)
- 'with' is disabled
- function can't be named eval or arguments
- function parameter name cannot be eval or arguments
- no duplicate parameter names allowed

Add FLAG_strict_mode

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6369 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-18 16:43:53 +00:00
ricow@chromium.org
3d245133a3 Enable compilation of very simple functions in x64 crankshaft.
This change includes support for safepointtables and adding deoptimization info (but not for deoptimizing).

Implemented crankshaft functions:
CallCode
GenerateSafepointTable
RegisterEnvironmentForDeoptimization
EmitGoto


This change allows us to compile very simple functions with crankshaft:
An empty function
A function returning a constant.
A function returning a parameter.

There is 6 disabled tests that require us to be able to deoptimize
which is currently not supported.


Review URL: http://codereview.chromium.org/6310009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6350 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2011-01-17 13:11:39 +00:00