The following command: ./Tools/Scripts/run-webkit-tests --gtk LayoutTests/fast/multicol/overflow-content.html results in a failure. This doesn't happen when running this test together with others in one go (e.g. all tests in LayoutTests/fast/multicol). Not exactly sure why the test is unstable, but my current fix (not submitted yet) for bug 113058 makes it fail reliably. This bug is related to how overflow is handled in rtl and btt writing modes, and I find it hard to believe that this is a multicol-specific problem.
Created attachment 196138 [details] Test case
Created attachment 196294 [details] Test case with non-visible overflow (no multicol)
Created attachment 196337 [details] Patch
Comment on attachment 196337 [details] Patch Attachment 196337 [details] did not pass chromium-ews (chromium-xvfb): Output: http://webkit-commit-queue.appspot.com/results/17413082 New failing tests: fast/multicol/vertical-rl/rules-with-border-before.html
Created attachment 196346 [details] Archive of layout-test-results from gce-cr-linux-01 for chromium-linux-x86_64 The attached test failures were seen while running run-webkit-tests on the chromium-ews. Bot: gce-cr-linux-01 Port: chromium-linux-x86_64 Platform: Linux-3.3.8-gcg-201212281604-x86_64-with-GCEL-10.04-gcel_10.04
Created attachment 196456 [details] Patch
Comment on attachment 196456 [details] Patch Attachment 196456 [details] did not pass chromium-ews (chromium-xvfb): Output: http://webkit-commit-queue.appspot.com/results/17418075 New failing tests: fast/multicol/vertical-rl/rules-with-border-before.html
Created attachment 196462 [details] Archive of layout-test-results from gce-cr-linux-03 for chromium-linux-x86_64 The attached test failures were seen while running run-webkit-tests on the chromium-ews. Bot: gce-cr-linux-03 Port: chromium-linux-x86_64 Platform: Linux-3.3.8-gcg-201212281604-x86_64-with-GCEL-10.04-gcel_10.04
Anyone interested in reviewing this?
Fixed in Blink: http://code.google.com/p/chromium/issues/detail?id=236326 If someone wants it in WebKit, feel free to grab it. :) I'm unassigning myself from this one.
Okay, I'm back. :) This is now blocking bug 131809 (in addition to bug 113058).
Created attachment 229927 [details] Patch
Comment on attachment 229927 [details] Patch Rename noOverflowRect() to flippedClientBoxRect(). Looks good just do that rename and I'll r+ cq+
Created attachment 229932 [details] Patch
Comment on attachment 229932 [details] Patch Looks like you need to export the new method for it to build on Mac.
Created attachment 229961 [details] Patch
Comment on attachment 229961 [details] Patch r=me
Comment on attachment 229961 [details] Patch Clearing flags on attachment: 229961 Committed r167706: <http://trac.webkit.org/changeset/167706>
All reviewed patches have been landed. Closing bug.