Bug 214610 – We should have exception check after promise operation

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Yusuke Suzuki]

We should have exception check after promise operation

[Yusuke Suzuki]

Created attachment 404849 [details]
Patch

[Yusuke Suzuki]

<rdar://problem/65881794>

[Mark Lam]

Comment on attachment 404849 [details]
Patch

r=me

[Yusuke Suzuki]

Created attachment 405029 [details]
Patch

[youenn fablet]

Comment on attachment 405029 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=405029&action=review

> Source/WebCore/bindings/js/JSDOMPromiseDeferred.cpp:187
> +    EXCEPTION_ASSERT(!scope.exception() || isTerminatedExecutionException(lexicalGlobalObject.vm(), scope.exception()));

Should we add this line in other places like reject(Exception...)?

[Mark Lam]

Comment on attachment 405029 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=405029&action=review

>> Source/WebCore/bindings/js/JSDOMPromiseDeferred.cpp:187
>> +    EXCEPTION_ASSERT(!scope.exception() || isTerminatedExecutionException(lexicalGlobalObject.vm(), scope.exception()));
> 
> Should we add this line in other places like reject(Exception...)?

Let's only do this as needed for now.  I think we need a better mental model for TerminatedExecutionExceptions, OOMEs, and StackOverflowErrors better for CatchScope.  We may want to replace these eventually.  No need to do additional work until we have the better model.

> LayoutTests/js/dom/promise-should-have-exception-check-on-operation.html:9
> +SubtleCrypto.prototype.deriveBits();

From the win EWS, it looks like SubtleCrypto isn't available on OS(Windows).  Can you confirm if this is the case, and if so, add a TestExpectation for Windows port to expect this test to fail?

[Mark Lam]

r=me too.

[Yusuke Suzuki]

Comment on attachment 405029 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=405029&action=review

>> LayoutTests/js/dom/promise-should-have-exception-check-on-operation.html:9
>> +SubtleCrypto.prototype.deriveBits();
> 
> From the win EWS, it looks like SubtleCrypto isn't available on OS(Windows).  Can you confirm if this is the case, and if so, add a TestExpectation for Windows port to expect this test to fail?

Nice catch. I'll fix it.

[Yusuke Suzuki]

Created attachment 405072 [details]
Patch for landing

[Yusuke Suzuki]

webaudio/webaudio-gc.html crashes but it was before this patch.

[EWS]

Committed r264802: <https://trac.webkit.org/changeset/264802>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 405072 [details].