RESOLVED FIXED Bug 95907
Make computePositionedLogicalWidth and computePositionedLogicalWidthReplaced const
https://bugs.webkit.org/show_bug.cgi?id=95907
Summary Make computePositionedLogicalWidth and computePositionedLogicalWidthReplaced ...
Tony Chang
Reported 2012-09-05 16:29:36 PDT
Make computePositionedLogicalWidth and computePositionedLogicalWidthReplaced const
Attachments
Patch (24.82 KB, patch)
2012-09-05 16:34 PDT, Tony Chang
no flags
Patch (24.01 KB, patch)
2012-09-06 11:46 PDT, Tony Chang
no flags
Tony Chang
Comment 1 2012-09-05 16:34:33 PDT
Tony Chang
Comment 2 2012-09-05 16:39:29 PDT
This parallels the changes to computePositionedLogicalHeight/computePositionedLogicalHeightReplaced in http://trac.webkit.org/changeset/126802 .
Ojan Vafai
Comment 3 2012-09-05 18:32:23 PDT
Comment on attachment 162367 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=162367&action=review > Source/WebCore/rendering/RenderBox.cpp:2676 > + stretchValues); > + computedValues.m_extent = stretchValues.m_extent; > + computedValues.m_position = stretchValues.m_position; > + computedValues.m_margins.m_start = stretchValues.m_margins.m_start; > + computedValues.m_margins.m_end = stretchValues.m_margins.m_end; Why can't you just pass computedValues in instead of stretchValues? > Source/WebCore/rendering/RenderBox.cpp:3184 > + LayoutUnit& marginLogicalLeftAlias = style()->isLeftToRightDirection() ? computedValues.m_margins.m_start : computedValues.m_margins.m_end; > + LayoutUnit& marginLogicalRightAlias = style()->isLeftToRightDirection() ? computedValues.m_margins.m_end : computedValues.m_margins.m_start; Here and elsewhere, these don't look equivalent to me. FractionalLayoutUnit& FractionalLayoutBoxExtent::mutableLogicalLeft(WritingMode writingMode) { return isHorizontalWritingMode(writingMode) ? m_left : m_top; }
Tony Chang
Comment 4 2012-09-06 11:46:33 PDT
Tony Chang
Comment 5 2012-09-06 11:48:04 PDT
Comment on attachment 162546 [details] Patch Removed stretchValues. The new code is the same because we want to ignore ltr/rtl, but start/end take that into consideration. In other words, it's undoing the flipping caused by ltr/rtl.
WebKit Review Bot
Comment 6 2012-09-06 19:42:54 PDT
Comment on attachment 162546 [details] Patch Clearing flags on attachment: 162546 Committed r127812: <http://trac.webkit.org/changeset/127812>
WebKit Review Bot
Comment 7 2012-09-06 19:42:57 PDT
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.