Bug 221125 – [WebAuthn] Optimize _WKWebAuthenticationPanelDelegate for native API

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKit Misc.

[Jiewen Tan]

Optimize _WKWebAuthenticationPanelDelegate for native API.

[Jiewen Tan]

<rdar://problem/71509141>

[Jiewen Tan]

Created attachment 418711 [details]
Patch

[Jiewen Tan]

Created attachment 418789 [details]
Patch

[Jiewen Tan]

Created attachment 418808 [details]
Patch

[Jiewen Tan]

Created attachment 418809 [details]
Patch

[Brent Fulgham]

Comment on attachment 418809 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=418809&action=review

r=me

> Source/WebKit/UIProcess/WebAuthentication/AuthenticatorManager.cpp:509
> +    RunLoop::main().dispatch([weakPanel, call = WTFMove(call)] () {

Should we WTFMove() the weakPanel? We don't use it again outside the dispatch.

> Tools/TestWebKitAPI/Tests/WebKitCocoa/_WKWebAuthenticationPanel.mm:155
> +        [responses[0] setLAContext:laContext.get()];

Why not responses.firstObject?

[Jiewen Tan]

Comment on attachment 418809 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=418809&action=review

Thanks Brent for r+ this patch.

>> Source/WebKit/UIProcess/WebAuthentication/AuthenticatorManager.cpp:509
>> +    RunLoop::main().dispatch([weakPanel, call = WTFMove(call)] () {
> 
> Should we WTFMove() the weakPanel? We don't use it again outside the dispatch.

Good catch. We can totally move that.

>> Tools/TestWebKitAPI/Tests/WebKitCocoa/_WKWebAuthenticationPanel.mm:155
>> +        [responses[0] setLAContext:laContext.get()];
> 
> Why not responses.firstObject?

Sure, it's totally doable. Since we check the count before, [0] shouldn't cause any issue.

[Jiewen Tan]

Created attachment 418937 [details]
Patch for landing

[EWS]

Committed r272184: <https://trac.webkit.org/changeset/272184>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 418937 [details].