This allows us to use a greater rendering depth for exceedingly-deep DOM trees
in layout, better matching what mainstream browsers are capable of. Note that
for 32-bit Windows the stack size MUST be set to larger than the default or Bad
Things Will Happen™ - we use 1.5 MB for this as a carefully-tuned value.
This needs to be capped specifically for JS use because some JavaScript
obfuscators deliberately trigger stack overflows and would lock up the browser
otherwise as long as there's still stack space to abuse. For web compatibility
we therefore limit this to 2MB in JS only (3x for ASAN) while still allowing
a greater depth for the layout engine.
This uses a different approach by modifying the `Boolish` function directly to act differently if we're checking for nullish values.
Partially based on:
Bug 1566141 - Implement the Nullish Coalescing operator (??) proposal
Bug 1599163 - Nullish coalescing operator returns undefined for void expressions
Partially based on:
Bug 1566141 - Implement the Nullish Coalescing operator (??) proposal
Bug 1593415 - Assertion failure: false, at js/src/builtin/ReflectParse.cpp:3092 with nullish coalescing
IIUC, we want to process the GOTO in the case of optional chains, and we do not satisfy the requirements mentioned in the comment of snoopControlFlow's caller. Optional chains are not loops, we (probably) don't have a loop in the instruction following the GOTO, and in the GOTO destination, we're either returning an undefined/null value or the actual value.
This includes the following:
Bug 1611777 - Part 1: Report syntax error for optional property access in self-hosting code. r=yulia!
Bug 1611777 - Part 2: Merge same blocks in emitDeleteOptionalChain(). r=yulia!
Bug 1611777 - Part 3: N/A
Bug 1611777 - Part 4: `super` can't occur on the left-hand side of an optional chain
Bug 1611777 - Part 5: Remove unnecessary super-handling in optional delete. r=yulia!
The child node of a DeleteOptionalChainExpr node can't be a super-property
accessor, so we can remove this code.
Bug 1611777 - Part 6: Crash for unexpected super-base in optional call. r=yulia!
Bug 1611777 - Part 7: Add missing entries to list of valid optional chain start expressions. r=yulia!
Bug 1611777 - Part 8: Add missing emitGet in emitOptionalElemExpression. r=yulia!
Aligns emitOptionalElemExpression() with emitOptionalDotExpression(), so it's
easier to compare both methods against each other.
Bug 1611777 - Part 9: Replace an if-statement with an assertion. r=yulia!
Bug 1611777 - Part 10: N/A
Bug 1611777 - Part 11: Support optional chaining in class heritage expression. r=yulia!
Bug 1611777 - Part 12: Use optionalExpr() for update expressions to match spec grammar. r=yulia!
Using optionalExpr matches the spec grammar more closely. This change also
modifies the reported error message. ++a?.b reported before this change
"unexpected token: '?.'", but now reports "invalid increment/decrement operand".
Bug 1611777 - Part 13: N/A
Bug 1611777 - Part 14: Simplify two lines in optionalExpr(). r=yulia!
We don't need to test for tt == TokenKind::Eof when we return for tt != TokenKind::OptionalChain anyway. Omit local variable for the result value and instead use a tail-call. This matches the local style in the parser more closely.
Bug 1611777 - Part 15: Support FunCall/FunApply optimisations for optional chaining. r=yulia!
Bug 1611777 - Part 16: Pass through ValueUsage in optional chains. r=yulia
Partially based on:
Bug 1566143 - Implement the Optional Chaining operator (?.) proposal
Bug 1650340 - Fix deleteElementInOptionalChain
Bug 1644839 - Fix return value of emitOptionalChain on failure
This is needed to fix the assertion on debug builds that self-hosted functions must have consistent names, as these two functions are aliased to `trimLeft` and `trimRight`.
This function essentially appears to work as written, there was an issue with it returning false too early in some situations that prevented it from working properly.
