Branch: refs/heads/main
Home: https://github.com/WebKit/WebKit
Commit: d0d05739899c9a8d76d166bbc1afbd116fa5eaf8
https://github.com/WebKit/WebKit/commit/d0d05739899c9a8d76d166bbc1afbd116fa5eaf8
Author: Alexey Shvayka <ashva...@apple.com>
Date: 2023-06-29 (Thu, 29 Jun 2023)
Changed paths:
M JSTests/ChakraCore/test/Closures/bug_OS_2299723.baseline-jsc
M JSTests/stress/const-not-strict-mode.js
M JSTests/stress/eval-func-decl-in-eval-within-catch-scope.js
M JSTests/stress/eval-func-decl-in-global-of-eval.js
M JSTests/stress/eval-func-decl-within-eval-with-reassign-to-var.js
A JSTests/stress/eval-let-const-redeclararion.js
M JSTests/stress/global-lexical-var-injection.js
M JSTests/stress/lexical-let-not-strict-mode.js
M JSTests/test262/expectations.yaml
M Source/JavaScriptCore/bytecompiler/BytecodeGenerator.cpp
M Source/JavaScriptCore/bytecompiler/BytecodeGenerator.h
M Source/JavaScriptCore/interpreter/Interpreter.cpp
M Source/JavaScriptCore/runtime/JSScope.cpp
Log Message:
-----------
[JSC] Align duplicate declaration checks in EvalDeclarationInstantiation with
the spec
https://bugs.webkit.org/show_bug.cgi?id=167837
<rdar://problem/111328974>
Reviewed by Yusuke Suzuki.
For the sloppy-mode eval(), this change:
1. Removes slowish TypeError-throwing logic from executeEval() that also wasn't
spec-compliant
(SyntaxError should be raised instead), harmonizing error messages.
2. Expands resolveScopeForHoistingFuncDeclInEval() to be called for all
declared variables, which
currently includes function declarations as well, ensuring SyntaxError is
thrown for duplicates
with upper yet non-top lexical scopes [1], all while skipping
CatchScopeWithSimpleParameter [2].
3. Introduces emitPutToScopeDynamic(), which circumvents default ResolveType
resolution that isn't
correct wrt skipping CatchScopeWithSimpleParameter as
resolveScopeForHoistingFuncDeclInEval() does.
We can't possibly tweak BytecodeGenerator::resolveType() to account for
eval().
This fixes both top-level and block-level function declarations to be
hoisted correctly from eval()
within simple parameter catch block by the same name.
4. Removes isExtensible() check from resolveScopeForHoistingFuncDeclInEval()
because for declared
variables, CanDeclareGlobalVar [3] is already implemented, while for Annex B
hoisted functions,
the implementation doesn't appear correct to unconditionally rely on
isExtensible() even if the
property is already present.
Furthermore, performing CanDeclareGlobalVar in
resolveScopeForHoistingFuncDeclInEval() is kinda
superfluous given we put jsUndefined() variables in executeEval(), and
results in incorrect
error being thrown (SyntaxError instead of TypeError) if global object is
non-extensible.
[1]: https://tc39.es/ecma262/#sec-evaldeclarationinstantiation (step
3.d.i.2.a.i)
[2]: https://tc39.es/ecma262/#sec-variablestatements-in-catch-blocks
[3]: https://tc39.es/ecma262/#sec-candeclareglobalvar
All JSTests changes were proven to align JSC with V8 and SpiderMonkey.
* JSTests/ChakraCore/test/Closures/bug_OS_2299723.baseline-jsc:
* JSTests/stress/const-not-strict-mode.js:
* JSTests/stress/eval-func-decl-in-eval-within-catch-scope.js:
* JSTests/stress/eval-func-decl-in-global-of-eval.js:
* JSTests/stress/eval-func-decl-within-eval-with-reassign-to-var.js:
* JSTests/stress/eval-let-const-redeclararion.js: Added.
* JSTests/stress/global-lexical-var-injection.js:
* JSTests/stress/lexical-let-not-strict-mode.js:
* JSTests/test262/expectations.yaml: Mark 160 tests as passing.
* Source/JavaScriptCore/bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::generate):
(JSC::BytecodeGenerator::hoistSloppyModeFunctionIfNecessary):
(JSC::BytecodeGenerator::emitPutToScopeDynamic):
* Source/JavaScriptCore/bytecompiler/BytecodeGenerator.h:
* Source/JavaScriptCore/interpreter/Interpreter.cpp:
(JSC::Interpreter::executeEval):
* Source/JavaScriptCore/runtime/JSScope.cpp:
(JSC::JSScope::resolveScopeForHoistingFuncDeclInEval):
Co-authored-by: Oleksandr Skachkov <gskach...@gmail.com>
Canonical link: https://commits.webkit.org/265614@main
_______________________________________________
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
