On Thursday 2014-11-06 14:16 +1300, Robert O'Callahan wrote: > On Thu, Nov 6, 2014 at 1:56 PM, L. David Baron <[email protected]> wrote: > > > But each tick of the animated image (which I believe is a background > > image) sends us through ImageLoader::DoRedraw, which in turn calls > > nsIFrame::SchedulePaint, so that we do display list analysis to see > > if anything changed. (Nothing has, since the image is well > > offscreen.) But constructing and comparing the display lists takes > > a good bit of time, leading to twitter using a good bit of CPU when > > it's being displayed. > > > > Is it reasonable to test that the image is offscreen (or, really, > > not in the region that's painted into a layer) somewhere earlier in > > the process here? Or is there something else we should be doing to > > avoid this? > > > > Definitely. I think we can avoid the SchedulePaint if > InvalidateImagesCallback doesn't find anything to invalidate. Got a bug > number?
https://bugzilla.mozilla.org/show_bug.cgi?id=1094535 -David -- 𝄞 L. David Baron http://dbaron.org/ 𝄂 𝄢 Mozilla https://www.mozilla.org/ 𝄂 Before I built a wall I'd ask to know What I was walling in or walling out, And to whom I was like to give offense. - Robert Frost, Mending Wall (1914)
signature.asc
Description: Digital signature
_______________________________________________ dev-tech-layout mailing list [email protected] https://lists.mozilla.org/listinfo/dev-tech-layout
