Bug 187526 – Remove unnecessary Test262 harness file and implement $262.IsHTMLDDA

• Status: RESOLVED
• Resolution: FIXED
• Component: Tools / Tests

[Rick Waldron]

The functionality provided by this file is built-in. 

I confirmed this by: 

- Building JSC with a patch that enables SharedArrayBuffer and Atomics (https://gist.github.com/rwaldron/89ed9a4bb7a459db8d54c8fe77ead4b1)
- Removing SharedArrayBuffer and Atomics from the skip list
- Running the Atomics tests from Test262/master


with agent.js: 

    342 tests run
    59 test files skipped
    27 tests failed in total
    27 tests newly fail
    0 tests newly pass
    Saved all the results in /Users/rwaldron/clonez/WebKit/test262-results/results.yaml
    Summarizing results...
    See the summaries and results in the /Users/rwaldron/clonez/WebKit/test262-results.

    Done in 13.48 seconds!


without agent.js: 

    342 tests run
    59 test files skipped
    27 tests failed in total
    27 tests newly fail
    0 tests newly pass
    Saved all the results in /Users/rwaldron/clonez/WebKit/test262-results/results.yaml
    Summarizing results...
    See the summaries and results in the /Users/rwaldron/clonez/WebKit/test262-results.

    Done in 13.21 seconds!

[Rick Waldron]

Created attachment 344708 [details]
patch

[Alexey Shvayka]

Created attachment 395524 [details]
Patch

[Alexey Shvayka]

$262.IsHTMLDDA tests

  test/annexB/language/expressions/yield/star-iterable-return-emulates-undefined-throws-when-called.js
  test/annexB/language/statements/for-await-of/iterator-close-return-emulates-undefined-throws-when-called.js
  test/annexB/language/statements/for-of/iterator-close-return-emulates-undefined-throws-when-called.js

are still failing because JSC partly implements https://github.com/tc39/ecma262/pull/1408.

[Ross Kirsling]

This works, but I wonder if it would be better to make Masquerader callable -- seems like this is an omission that just wasn't necessary until now?

[Alexey Shvayka]

Created attachment 395557 [details]
Patch

Return callable masquerader.

[Alexey Shvayka]

(In reply to Ross Kirsling from comment #4)
> This works, but I wonder if it would be better to make Masquerader callable
> -- seems like this is an omission that just wasn't necessary until now?

That is an awesome improvement, thank you. I thought there are quite a few tests for non-callable masquerader, yet is seems like one assertion only. Given that there is only one masquerader in the wild, and it is callable, we would rather use JSFunction::createFunctionThatMasqueradesAsUndefined.

[EWS]

Committed r259587: <https://trac.webkit.org/changeset/259587>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 395557 [details].

[Radar WebKit Bug Importer]

<rdar://problem/61352863>