Inline logicalHeightForRowSizing to shave another 2-3% off robohornet's resizecol...
authoreric@webkit.org <eric@webkit.org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Tue, 9 Oct 2012 01:39:20 +0000 (01:39 +0000)
committereric@webkit.org <eric@webkit.org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Tue, 9 Oct 2012 01:39:20 +0000 (01:39 +0000)
commit32ef5f43fd54aacba57f56f70d04b30ac7f8f006
tree4b974b2d3535ef3e991254ca412383ca8ac568d4
parent572f866deca9c0e906e15c47f57ce16b3f826be3
Inline logicalHeightForRowSizing to shave another 2-3% off robohornet's resizecol.html
https://bugs.webkit.org/show_bug.cgi?id=98703

Reviewed by Emil A Eklund.

This is very small potatoes.  There are much bigger wins for table layout yet, but
this was an easy win.

This function should probably be converted to use int's only, as table cells are pixel-sized
according to our subpixel-experts.

Also, I suspect there should be ways to early return with less-math in the common cases, but
I've saved such for a later patch.

Note that I changed from using paddingBefore/paddingAfter (which include the instrinsic padding)
to calling computedCSSPaddingBefore/computedCSSPaddingAfter directly as well.

This single function is about 11% of total time for robohornet's resizecol.

* rendering/RenderTableCell.cpp:
* rendering/RenderTableCell.h:
(WebCore::RenderTableCell::logicalHeightForRowSizing):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@130710 268f45cc-cd09-0410-ab3c-d52691b4dbfc
Source/WebCore/ChangeLog
Source/WebCore/rendering/RenderTableCell.cpp
Source/WebCore/rendering/RenderTableCell.h