Merge https://chromium.googlesource.com/chromium/blink/+/66427d0825fcc2975bd50220cdcaa2504d6f36e5 The case which calculates the margin only depends on the RenderStyle when the margin is fixed. It is non-fixed margins that depend on the RenderBox's margin. The case that calculates the padding only depends on the RenderStyle when the element is not a box. This improves the page load time of the economist by 27% (4.7s -> 3.4s) on my z620 workstation.
Don't merge verbatim. isLayoutDependent should not take a PassRefPtr<RenderStyle> style.
Also there should be an inline helper function for all these renderer && renderer->isBox() && (!style || !style->paddingBottom().isFixed());
Also see https://chromium.googlesource.com/chromium/blink/+/ff234b1593b2b493d47f38f687d09a87bc42c9eb. Also see the bug 118618.
Created attachment 207433 [details] Merges the patch
There were two follow-ups you might be interested in: https://codereview.chromium.org/18298016/ https://codereview.chromium.org/19272007/
(In reply to comment #5) > There were two follow-ups you might be interested in: > https://codereview.chromium.org/18298016/ > https://codereview.chromium.org/19272007/ Thanks but those are follow ups for the patch to be merged in the bug 118618, right?
Comment on attachment 207433 [details] Merges the patch r=me
(In reply to comment #6) > (In reply to comment #5) > > There were two follow-ups you might be interested in: > > https://codereview.chromium.org/18298016/ > > https://codereview.chromium.org/19272007/ > > Thanks but those are follow ups for the patch to be merged in the bug 118618, right? You are right. Sorry, I got the bugs confused.
Comment on attachment 207433 [details] Merges the patch Clearing flags on attachment: 207433 Committed r153347: <http://trac.webkit.org/changeset/153347>
All reviewed patches have been landed. Closing bug.