Created attachment 225179 [details] Test case Add support for -webkit-column-span:all to the new ("region based") multicol implementation.
Corresponding bug in Blink: https://code.google.com/p/chromium/issues/detail?id=347325
Created attachment 225314 [details] Patch
Comment on attachment 225314 [details] Patch This isn't finished yet. Just submitting this patch for Dave to have a look.
This is done! New patch will be coming from Morten once 122754 lands, and then we'll get it reviewed.
Created attachment 228865 [details] Patch
I'm done! :) Well, not quite. Apart from addressing review issues, this patch is missing one important part that I need help with: I've added new source files, and I don't know how to properly make changes in Source/WebCore/WebCore.xcodeproj/project.pbxproj , which I assume is what you use on Mac. Some magical-looking hex codes in this file helped me resist the temptation of hand-editing this.
Created attachment 229375 [details] Patch
I uploaded an unchanged patch that adds the spanner placeholder to the XCode project.
Created attachment 229379 [details] Patch that should get Windows building too.
Comment on attachment 229379 [details] Patch that should get Windows building too. Attachment 229379 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.appspot.com/results/5112968480030720 New failing tests: fast/multicol/newmulticol/client-rects.html fast/multicol/newmulticol/compare-with-old-impl/span-as-immediate-child-complex-splitting.html
Created attachment 229389 [details] Archive of layout-test-results from webkit-ews-15 for mac-mountainlion-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: webkit-ews-15 Port: mac-mountainlion-wk2 Platform: Mac OS X 10.8.5
Comment on attachment 229379 [details] Patch that should get Windows building too. Attachment 229379 [details] did not pass mac-ews (mac): Output: http://webkit-queues.appspot.com/results/4746858488397824 New failing tests: fast/multicol/newmulticol/client-rects.html fast/multicol/newmulticol/compare-with-old-impl/span-as-immediate-child-complex-splitting.html
Created attachment 229393 [details] Archive of layout-test-results from webkit-ews-01 for mac-mountainlion The attached test failures were seen while running run-webkit-tests on the mac-ews. Bot: webkit-ews-01 Port: mac-mountainlion Platform: Mac OS X 10.8.5
Comment on attachment 229379 [details] Patch that should get Windows building too. Attachment 229379 [details] did not pass mac-ews (mac): Output: http://webkit-queues.appspot.com/results/4920634475806720 New failing tests: fast/multicol/newmulticol/client-rects.html fast/multicol/newmulticol/compare-with-old-impl/span-as-immediate-child-complex-splitting.html
Created attachment 229394 [details] Archive of layout-test-results from webkit-ews-07 for mac-mountainlion The attached test failures were seen while running run-webkit-tests on the mac-ews. Bot: webkit-ews-07 Port: mac-mountainlion Platform: Mac OS X 10.8.5
Comment on attachment 229379 [details] Patch that should get Windows building too. Attachment 229379 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.appspot.com/results/6023564897550336 New failing tests: fast/multicol/newmulticol/client-rects.html fast/multicol/newmulticol/compare-with-old-impl/span-as-immediate-child-complex-splitting.html
Created attachment 229397 [details] Archive of layout-test-results from webkit-ews-10 for mac-mountainlion-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: webkit-ews-10 Port: mac-mountainlion-wk2 Platform: Mac OS X 10.8.5
Comment on attachment 229379 [details] Patch that should get Windows building too. View in context: https://bugs.webkit.org/attachment.cgi?id=229379&action=review > Source/WebCore/rendering/RenderBlockFlow.cpp:121 > - moveAllChildrenTo(flowThread, true); > RenderBlock::addChild(flowThread); > + flowThread->populate(); This part is wrong. By delaying the population of the flow thread until after the flow thread has been inserted into the flow thread parent, you cause an anonymous block to be created for no reason to enclose all of the inline content. This inline content is going into the flow thread, so we don't want to wastefully wrap it in an unnecessary anonymous block. (This issue was caught by the client-rects.html layout test on OS X.)
Created attachment 229406 [details] Reset the results for client-rects.html now that the flow thread grows to the height of the number of columns.
There is a subpixel layout difference with the spanners that is causing span-as-immediate-child-complex-splitting.html to fail. Looking into that now.
I fixed the populate() issue by manually setting childrenInline to false before inserting the flow thread. This stops makeChildrenNonInline from being called.
Created attachment 229409 [details] Removed the test case with the subpixel difference, since it's not wrong, just a bit different.
Comment on attachment 229409 [details] Removed the test case with the subpixel difference, since it's not wrong, just a bit different. r=me
Landed in r167335.
One question.... did you tested the -expected layout test results for the GTK port (file LayoutTests/platform/gtk/fast/multicol/newmulticol/client-rects-expected.txt added in r167335) ? Because, since it has been added, it has been giving failures on the GTK port. The diff is the following: http://build.webkit.org/results/GTK%20Linux%2064-bit%20Release/r167336%20%2846517%29/fast/multicol/newmulticol/ I'm going to rebaseline it to the actual result unless you see something wrong with that diff.