Bug 204770 – [iOS] Calls to the device motion API should be done in the UI process

• Component: WebKit Misc.
• Status: RESOLVED FIXED

[Per Arne Vollan]

The device motion API on iOS is communicating with iohideventsystem. Since mach lookup to this daemon will be closed, the calls to this API should be moved from the WebContent process to the UI process.

[Radar WebKit Bug Importer]

<rdar://problem/57572913>

[Per Arne Vollan]

Created attachment 384662 [details]
Patch

[Per Arne Vollan]

Created attachment 384670 [details]
Patch

[Per Arne Vollan]

Created attachment 384673 [details]
Patch

[Per Arne Vollan]

Created attachment 385207 [details]
Patch

[Per Arne Vollan]

Created attachment 385285 [details]
Patch

[Brent Fulgham]

Comment on attachment 385285 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=385285&action=review

r=me, but please consider changing those bare pointers (perhaps even in a follow-up patch).

> Source/WebCore/platform/ios/DeviceMotionClientIOS.h:55
> +    DeviceMotionController* m_controller { nullptr };

I'm a little suspicious of these bare pointers. Could they be WeakPtrs?

[Per Arne Vollan]

(In reply to Brent Fulgham from comment #7)
> Comment on attachment 385285 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=385285&action=review
> 
> r=me, but please consider changing those bare pointers (perhaps even in a
> follow-up patch).
> 
> > Source/WebCore/platform/ios/DeviceMotionClientIOS.h:55
> > +    DeviceMotionController* m_controller { nullptr };
> 
> I'm a little suspicious of these bare pointers. Could they be WeakPtrs?

I think they can be made WeakPtrs, I will look into that :)

Thanks for reviewing!

[Per Arne Vollan]

(In reply to Brent Fulgham from comment #7)
> Comment on attachment 385285 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=385285&action=review
> 
> r=me, but please consider changing those bare pointers (perhaps even in a
> follow-up patch).
> 
> > Source/WebCore/platform/ios/DeviceMotionClientIOS.h:55
> > +    DeviceMotionController* m_controller { nullptr };
> 
> I'm a little suspicious of these bare pointers. Could they be WeakPtrs?

Filed https://bugs.webkit.org/show_bug.cgi?id=205093.

[WebKit Commit Bot]

Comment on attachment 385285 [details]
Patch

Clearing flags on attachment: 385285

Committed r253357: <https://trac.webkit.org/changeset/253357>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Chris Dumez]

Comment on attachment 385285 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=385285&action=review

> Source/WebCore/ChangeLog:3
> +        [iOS] Calls to the device motion API should be done in the UI process

This change seems to have badly broken the device motion API :(

see <rdar://problem/60720953>.