Currently out-of-view check of fixed position doesn't consider the page scale. When the page is scaled down, originally out-of-view fixed element might be in-view; and vice versa.
Created attachment 178113 [details] Patch (tests may fail when lacking of patch to bug 104294)
Comment on attachment 178113 [details] Patch (tests may fail when lacking of patch to bug 104294) Attachment 178113 [details] did not pass mac-ews (mac): Output: http://queues.webkit.org/results/15183285 New failing tests: compositing/layer-creation/fixed-position-out-of-view-scaled.html
Comment on attachment 178113 [details] Patch (tests may fail when lacking of patch to bug 104294) Attachment 178113 [details] did not pass chromium-ews (chromium-xvfb): Output: http://queues.webkit.org/results/15183346 New failing tests: platform/chromium/virtual/softwarecompositing/layer-creation/fixed-position-out-of-view-scaled.html compositing/layer-creation/fixed-position-out-of-view-scaled.html
Thanks Xianzhu, looks like you beat me to it :) Would you also mind adding a test with a non-zero scroll offset for the viewport since this is just checking against the viewport size? Also, I noticed that fixing things to the bottom or right side of the viewport doesn't work correctly. It looks like it may somehow be using the layout viewport instead of the visual viewport. That's an altogether different bug, though.
LGTM but should not fail on mac :-)
(In reply to comment #4) > Thanks Xianzhu, looks like you beat me to it :) Would you also mind adding a test with a non-zero scroll offset for the viewport since this is just checking against the viewport size? > Thanks for the suggestion. This is necessary. Will add it. > Also, I noticed that fixing things to the bottom or right side of the viewport doesn't work correctly. It looks like it may somehow be using the layout viewport instead of the visual viewport. That's an altogether different bug, though. Can you verify if this is caused by bug 104292?
(In reply to comment #5) > LGTM but should not fail on mac :-) The failure will be automatically fixed after the depended bug 104294 is resolved.
> Can you verify if this is caused by bug 104292? Don't you mean 104294?
(In reply to comment #8) > > Can you verify if this is caused by bug 104292? > > Don't you mean 104294? Yes. Sorry for the typo :)
(In reply to comment #6) > Can you verify if this is caused by bug [104294]? Unfortunately that patch doesn't appear to fix it. For example see http://css.flepstudio.org/en/css-properties/position-static-absolute-relative-static.html. The bottom right fixed div doesn't follow the visual viewport.
Filed bug 104692 for the issue in #4 and #10.
Created attachment 178910 [details] Patch
Updated the patch after landing 104294. Also added scrolled test.
Comment on attachment 178910 [details] Patch Clearing flags on attachment: 178910 Committed r137399: <http://trac.webkit.org/changeset/137399>
All reviewed patches have been landed. Closing bug.