On Sat, Mar 21, 2015 at 10:55 PM, Thinker K.F. Li <thin...@codemud.net> wrote:
> I was studying why maze solver from Microsoft is so slow with B2G and > Gecko, in last few days. If I am right, I found refresh driver always > invalidate whole region of display port, if there is, or visual overflow > rect of root frame. For the case of maze solver, even it changes style > of just small number of elements, refresh driver makes all frames > (almost) on frame tree to generate display items since they are falling > in the dirty region. It takes a lof of time to build display list, and > spend more time to process display items. > > I know we have display list based invalidation, but it does not work for > refresh driver. > > My question is why refresh driver works like this. Is it possible to > improve it with display list based invalidation or other solution? > Invalidation and display list construction are different things. It seems to me you're talking about the latter, not invalidation. Let's talk about it on IRC. Rob -- oIo otoeololo oyooouo otohoaoto oaonoyooonoeo owohooo oioso oaonogoroyo owoiotoho oao oboroootohoeoro oooro osoiosotoeoro owoiololo oboeo osouobojoeocoto otooo ojouodogomoeonoto.o oAogoaoiono,o oaonoyooonoeo owohooo osoaoyoso otooo oao oboroootohoeoro oooro osoiosotoeoro,o o‘oRoaocoao,o’o oioso oaonosowoeoroaoboloeo otooo otohoeo ocooouoroto.o oAonodo oaonoyooonoeo owohooo osoaoyoso,o o‘oYooouo ofooooolo!o’o owoiololo oboeo oiono odoaonogoeoro ooofo otohoeo ofoioroeo ooofo ohoeololo. _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform