The following layout test is failing on chromium-win and chromium-linux: svg/custom/svg-fonts-in-html.html Probable cause: This was caused by something the range 60781:60783. I'm suspecting 60781 ("WebCore: Space should be allowed between -- and > in comment end"), since the other two are non-code changes.
(In reply to comment #0) > The following layout test is failing on chromium-win and chromium-linux: > > svg/custom/svg-fonts-in-html.html > > Probable cause: > > This was caused by something the range 60781:60783. I'm suspecting 60781 ("WebCore: Space should be allowed between -- and > in comment end"), since the other two are non-code changes. I reverted 60781 as other platforms reported regressions. Btw, I did not see this test failing on the chromium core build bots. Do you have any other bots that caught this?
(In reply to comment #1) > I reverted 60781 as other platforms reported regressions. Btw, I did not see this test failing on the chromium core build bots. Do you have any other bots that caught this? Yes, we have the "WebKit Canary" bots: http://build.chromium.org/buildbot/waterfall.fyi/waterfall?builder=Webkit%20(webkit.org)&builder=Webkit%20Mac%20(webkit.org)&builder=Webkit%20Linux%20(webkit.org)&builder=XP%20Perf%20(webkit.org)(1)&builder=XP%20Perf%20(webkit.org)(2)&builder=Linux%20Perf%20(webkit.org)&builder=Webkit%20(purify%20webkit.org)&builder=Webkit%20Linux%20(valgrind%20webkit.org) These do a full build of tip-of-tree Chromium with tip-of-tree WebKit, and they run the full layout test suite (which I think the EWS and core chromium buildbots don't). So keeping the canaries green is a "nice to have" at this point, but not really a WebKit committer's responsibility (yet).
r60781 was reverted, and this test started passing again. Closing bug.