| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
|
| |
ECMAScript Intl specification
- Update make_unicode to output SpecialCasing
- Handle special casing
- Use realloc instead of malloc when resizing a newly created string buffer
Based-on: m-c 1318403, 1431957
|
|
|
|
| |
Based-on: m-c 866473
|
|
|
|
| |
Based-on: m-c 1329904
|
|
|
|
|
|
|
|
|
|
| |
- turn each Intl object into a NativeObject subclass
- create them as PlainObject
- ensure correct type is passed in self-hosted code
- implement legacy constructor semantics for DateTime&NumberFormat
- store internals on object slot instead of JS WeakMap
Based-on: m-c 1328386, 1332604
|
| |
|
|
|
|
|
| |
The comparator to ArraySort must be optional also in self-hosted code, where calling the
implementation directly is generally faster than calling between JS and C++ multiple times.
|
|
|
|
| |
Based-on: m-c 1521732/{1,3}, 1527881
|
|
|
|
|
|
| |
values
Based-on: m-c 1283562, 1277368/1, 1324828
|
|
|
|
| |
RelativeTimeFormat/DateTimeFormat
|
|
|
|
|
|
| |
Including infrastructure changes to ensure correct opts for future changes.
Based-on: m-c 1325632, 1418047, 1418573
|
|
|
|
|
| |
This ensures that rematerialized frames used by the devtools debugger
are properly removed so that no stale data is used during bailouts.
|
|
|
|
|
|
|
|
|
|
| |
This is a convenience access function to hasOwnProperty.
Trivial, self-hosted implementation providing the interface to the
already existing hasOwnProperty cpp function with additional toObject
for spec compliance.
Resolves #2256
|
|\
| |
| |
| |
| |
| | |
js/src/moz.build is defined.' (#2248) from jobbautista9/UXP:2221-followup into master
Reviewed-on: https://repo.palemoon.org/MoonchildProductions/UXP/pulls/2248
|
| |
| |
| |
| | |
assignment.
|
| |
| |
| |
| |
| |
| | |
defined.
Should fix build bustage in armhf, as well as prevent bustage in mips32.
|
|/
|
|
| |
Based on the implementation from Mozilla bug 1704385.
|
|
|
|
|
|
|
|
|
| |
into their own header and implementation files
Backported from Mozilla bug 1516742.
The .from* methods are going to depend on SameValueZero, which needs to be
visible to /dom. This patch provides the foundation for that.
|
|
|
|
|
|
|
| |
into SharedContext.
Based on https://bugzilla.mozilla.org/show_bug.cgi?id=1604792
Also remove ParseGoal being passed through Parser introduced in #1691 Part 2.
|
|
|
|
| |
Based on https://bugzilla.mozilla.org/show_bug.cgi?id=1402649 and https://bugzilla.mozilla.org/show_bug.cgi?id=1382306
|
|
|
|
|
|
| |
This excludes the GC-related changes (cell pointer asserts) since we don't have them. This bug should be revisited if we'd ever plan on porting those asserts over.
Partially based on https://bugzilla.mozilla.org/show_bug.cgi?id=1402649
|
|
|
|
|
|
|
|
| |
module functions
This excludes the change that removes "excessive" rooting from the Lambda* methods in Interpreter.cpp.
Based on https://bugzilla.mozilla.org/show_bug.cgi?id=1382306
|
| |
|
|
|
|
|
| |
With recent work a bunch more vim control lines were re-introduced.
This nukes them again.
|
| |
|
|
|
|
| |
Based-on: 1629106/1, 1684020
|
|
|
|
| |
Now that everything is finished, we don't need the feature flag any more.
|
|
|
|
| |
Based-on: m-c 1712138/{2,3}, 1713155/2
|
|
|
|
|
|
|
|
|
|
|
|
| |
Introduce a FunctionSyntaxKind for FieldInitializer since special rules
(around `arguments`) apply. At the same time we can move the flag from the
scope to the JSScript. This is similar to how derived constructors are handled
and makes the initWithEnclosingScope code closer to initWithEnclosingContext.
This version is a bit more complex than Mozilla's due to different storage of
bit flags on JSScript.
Based-on: m-c 1636800
|
|
|
|
| |
Based-on: m-c 1535804/{6,7}
|
|
|
|
| |
Based-on: m-c 1555037/2
|
|
|
|
|
|
|
|
|
|
| |
* Refactor code for emitting the .initializers array into ClassEmitter
* Only emit .initializers scope when actually required
* Remove unfinished code to handle non-present class field initialisers
* Use predicate count_if and any_of of ListNode
* Remove unnecessary parameters for class field parsing
Based-on: m-c 1553744, 1555979, 1555037/1, 1535804/{1-5}
|
|
|
|
|
| |
This was done wildly different in m-c 1552875, in the interest of keeping
|setOrEmitSetFunName| around it is implemented differently here.
|
|
|
|
|
|
|
| |
If there ever was a point where this structure "avoids duplication", at least since the
*OpEmitter refactor emitting PNK_NAME differently makes it actually harder to follow.
Mozilla makes the same change at a different time.
|
|
|
|
| |
Based-on: m-c 1552229
|
|
|
|
|
|
| |
classes
Based-on: m-c 1552022
|
|
|
|
| |
Based-on: m-c 1540789, 1547130, 1547467
|
|
|
|
| |
Based-on: m-c 1542406
|
|
|
|
|
|
| |
We don't want to call setters for field initialization.
Based-on: m-c 1535471, 1547035
|
|
|
|
| |
Based-on: m-c 1529772/{3,4}
|
|
|
|
| |
Based-on: m-c 1529772/{1,2}
|
|
|
|
|
|
| |
* Don't name field initializer lambdas
Based-on: m-c 1534721, 1551454, 1542448
|
|
|
|
|
|
| |
FunctionParamsEmitter with current methods
Based-on: m-c 1473796/3, 1473796/4
|
|
|
|
|
|
| |
LexicalScopeEmitter, DefaultEmitter
Based-on: m-c 1501577, 1521696, 1501578, 1473796/1
|
|
|
|
|
|
|
| |
* Don't print bogus error locations on BCE internal errors
* Do not use TokenPos in BytecodeEmitter::{setFunctionBodyEndPos,setScriptStartOffsetIfUnset}
Based-on: m-c 1451826/1, 1530034, 1473796/2
|
|
|
|
|
|
|
|
|
| |
names
This state still has the initializers scoped on .initializers local variable, which will be changed later.
Based-on: m-c 1499448, 1530084, 1530832, 1529448 (partial), 1532921, 1528039, 1528038, 1535166, 1550628,
1535166, 1550628, 1541641, 1547133, 1540787, 1535804/9
|
|
|
|
|
|
| |
This allows us to perform bisects on the following commits without causing runtime issues.
Based-on: m-c 1529758
|
|
|
|
|
|
| |
identifiers during parsing
Based-on: m-c 1351913/{1,2}
|
|\
| |
| |
| |
| |
| | |
by default, and fix a deprot in js/src' (#2222) from jobbautista9/UXP:ltcg-js into master
Reviewed-on: https://repo.palemoon.org/MoonchildProductions/UXP/pulls/2222
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
building it shared.
Currently only Windows can take advantage of this.
The reason why the arrays of deunified sources are named like that in
Spidermonkey's moz.build is to avoid UnsortedErrors in python. There's probably
a better way to handle that, but I just want something that works for now.
And yes, the deunified sources will be reunified if LTO is disabled.
|