[Webkit-unassigned] [Bug 170362] Add SPI for handling geolocation authorization requests

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Apr 11 10:05:25 PDT 2017


--- Comment #16 from David Quesada <david_quesada at apple.com> ---
(In reply to Alex Christensen from comment #15)
> Comment on attachment 306764 [details]
> Patch v2.1
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=306764&action=review
> LGTM except the memory leak.
> > Source/WebKit2/UIProcess/ios/WKGeolocationProviderIOS.mm:92
> > +    RetainPtr<id<_WKGeolocationCoreLocationProvider>> _coreLocationProvider;
> I'm not sure if the id is necessary here.  If it is, we seem to have a space
> after id elsewhere in WebKit.

The id is necessary since _WKGeolocationCoreLocationProvider is a protocol type, not a the name of a class, which would be usable without id<>.

> > Source/WebKit2/UIProcess/ios/WKGeolocationProviderIOS.mm:269
> > +    _provider = [[WebGeolocationCoreLocationProvider alloc] initWithListener:self];
> We need an adoptNS or this is a memory leak.

Will fix.

> > Tools/TestWebKitAPI/Tests/WebKit2Cocoa/Geolocation.mm:42
> > +SOFT_LINK_FRAMEWORK(CoreLocation)
> > +SOFT_LINK_CLASS(CoreLocation, CLLocation)
> Is there a reason we're soft linking?

The WebKit code that uses CoreLocation (WebGeolocationCoreLocationProvider) soft links it, so I also soft linked CoreLocation in the API tests. I'll change this to regular linking since there's not much of a point in soft linking in TestWebKitAPI.

You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20170411/df1c0433/attachment-0001.html>

More information about the webkit-unassigned mailing list