Bug 214193 – Add some testing for JSRunLoopTimer

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

[Geoffrey Garen]

Add some testing for JSRunLoopTimer

[Geoffrey Garen]

Created attachment 403990 [details]
Patch

[Geoffrey Garen]

I guess I'll just merge this into https://bugs.webkit.org/show_bug.cgi?id=214102

*** This bug has been marked as a duplicate of bug 214102 ***

[Darin Adler]

Comment on attachment 403990 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=403990&action=review

> Tools/TestWebKitAPI/Tests/JavaScriptCore/JSRunLoopTimer.mm:66
> +    auto context = adoptNS([JSContext new]);

So retro to use new instead of alloc/init.

> Tools/TestWebKitAPI/Tests/JavaScriptCore/JSRunLoopTimer.mm:69
> +    while (!s_done) {

Need to set s_done to false in case IncrementalSweeperSecondaryThread test already ran? But what about objects left behind from running that test? Could they result in false negatives for this test?

> Tools/TestWebKitAPI/Tests/JavaScriptCore/JSRunLoopTimer.mm:73
> +                auto jsObject = adoptNS([JSValue valueWithObject:object.get() inContext:context.get()]);

This looks like an overrelease. Should not adopt the result of valueWithObject.

> Tools/TestWebKitAPI/Tests/JavaScriptCore/JSRunLoopTimer.mm:90
> +    while (!s_done) {

Need to set s_done to false in case IncrementalSweeperMainThread test already ran? But what about objects left behind from running that test? Could they result in false negatives for this test?

> Tools/TestWebKitAPI/Tests/JavaScriptCore/JSRunLoopTimer.mm:96
> +                        auto jsObject = adoptNS([JSValue valueWithObject:object.get() inContext:context.get()]);

This looks like an overrelease. Should not adopt the result of valueWithObject.