Bug 137948 – [EFL] webview should be updated from (0,0) when using the evasGL.

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKit EFL

[Hunseop Jeong]

When using the evasGL, evas_object of webview have to update from (0,0) to the size of webview.

[Hunseop Jeong]

Now, webview didn't update the part of top in minibrowser. You can watch this error when scrolling the any page.

[Hunseop Jeong]

Created attachment 240255 [details]
Patch

[Hunseop Jeong]

Hyowon, Could review this patch?

[Hunseop Jeong]

Could you review this patch?

[Ryuan Choi]

Comment on attachment 240255 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=240255&action=review

> Source/WebKit2/UIProcess/API/efl/EwkView.cpp:564
> -    // sd->image is tied to a native surface, which is in the parent's coordinates.
> -    evas_object_image_data_update_add(sd->image, sd->view.x, sd->view.y, sd->view.w, sd->view.h);
> +    // sd->image have to update as the size of webview from (0,0) when using the evasGL.
> +    evas_object_image_data_update_add(sd->image, 0, 0, sd->view.w, sd->view.h);

I need more clear description about it.
At least, I want to know how to reproduce the issue.

If I remember correctly, it was tried but reverted long time ago.

[Hunseop Jeong]

Created attachment 246804 [details]
Patch

[Gyuyoung Kim]

Comment on attachment 246804 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=246804&action=review

> Source/WebKit2/ChangeLog:3
> +        [EFL] webview have to update from (0,0) to the size of webView after r180134.

s/have/has/g

> Source/WebKit2/ChangeLog:9
> +        (EwkView::displayTimerFired): Changed the point of start position to update the evasobject.

Could you explain what is problem now ?

> Source/WebKit2/UIProcess/API/efl/EwkView.cpp:578
> +    // sd->image have to update as the size of webview from (0,0) when using the evasGL.

I don't understand this comment well. Do you mean sd->image should be updated from (0, 0) when we use evasGL for graphics backend ?

[Hunseop Jeong]

(In reply to comment #7)
> Comment on attachment 246804 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=246804&action=review
> 
> > Source/WebKit2/ChangeLog:3
> > +        [EFL] webview have to update from (0,0) to the size of webView after r180134.
> 
> s/have/has/g
> 
> > Source/WebKit2/ChangeLog:9
> > +        (EwkView::displayTimerFired): Changed the point of start position to update the evasobject.
> 
> Could you explain what is problem now ?
In my Minibrowser, The top part of the screen wasn't updated. Looks like freeze. I try to test at other PC but that also occurred.
Is your Minibrowser okay?

> 
> > Source/WebKit2/UIProcess/API/efl/EwkView.cpp:578
> > +    // sd->image have to update as the size of webview from (0,0) when using the evasGL.
> 
> I don't understand this comment well. Do you mean sd->image should be
> updated from (0, 0) when we use evasGL for graphics backend ?
Yes, sd->image should be updated from (0,0) if using the graphics backend.

[Gyuyoung Kim]

Comment on attachment 246804 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=246804&action=review

>>> Source/WebKit2/UIProcess/API/efl/EwkView.cpp:578
>>> +    // sd->image have to update as the size of webview from (0,0) when using the evasGL.
>> 
>> I don't understand this comment well. Do you mean sd->image should be updated from (0, 0) when we use evasGL for graphics backend ?
> 
> Yes, sd->image should be updated from (0,0) if using the graphics backend.

I see. Please update this patch description and comment according to my comment again.

[Hunseop Jeong]

Created attachment 251405 [details]
Patch

[Gyuyoung Kim]

Comment on attachment 251405 [details]
Patch

It would be nicer if Ryuan has a final review before landing.

[WebKit Commit Bot]

Comment on attachment 251405 [details]
Patch

Clearing flags on attachment: 251405

Committed r184892: <http://trac.webkit.org/changeset/184892>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.