[Webkit-unassigned] [Bug 74993] [EFL] Refactor the way using cairo in ewk_tiled_backing_store.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=74993





--- Comment #9 from KwangHyuk <hyuki.kim at samsung.com>  2011-12-27 01:31:10 PST ---
(In reply to comment #8)
> (In reply to comment #4)
> > (In reply to comment #1)
> > > Created an attachment (id=120570)
 --> (https://bugs.webkit.org/attachment.cgi?id=120570&action=review) [details] [details] [details]
> > > Patch.
> > 
> > It seems that this is not good idea to move code from ewk_tile_new function to _ewk_view_tiled_render_cb. Initialization of cairo surface should be done only once per tile. Initialization of new cairo surface on every _ewk_view_tiled_render_cb function call seems to be performance lost. Do you have any test result how performance will be changed?
> 
> I'm not sure. Does it really have performance issue?
> 
> IIRC, It is not expensive and WebKit/Gtk+ and firefox doesn't keep cairo surface.

I am now agreeing with Ryuan.

According to my measurement, only one big burden occurs when it brings pixel data from image object for the first time when image object is created. :)

I couldn't notice the performance loss when it creates cairo surface and context yet.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.