Diff from a recent patch: --- /home/webkitbuildbot/slaves/release32bit/buildslave/qt-linux-release/build/layout-test-results/fast/repaint/svg-layout-root-style-attr-update-expected.txt 2010-06-23 10:01:08.455453400 -0700 +++ /home/webkitbuildbot/slaves/release32bit/buildslave/qt-linux-release/build/layout-test-results/fast/repaint/svg-layout-root-style-attr-update-actual.txt 2010-06-23 10:01:08.455453400 -0700 @@ -11,5 +11,5 @@ RenderBlock (positioned) {DIV} at (0,0) size 500x500 layer at (8,0) size 500x500 RenderSVGRoot {svg} at (8,0) size 500x500 - RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#FF0000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 Z"] - RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#008000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 Z"] + RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#FF0000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 L250.00,10.00"] + RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#008000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 L250.00,10.00"] See http://trac.webkit.org/changeset/61696. Every other port formats the rectangle the same way except Qt. From a spot check of LayoutTests/platform/qt this appears to be causing a lot of diffs.
(In reply to comment #0) > Diff from a recent patch: > > --- /home/webkitbuildbot/slaves/release32bit/buildslave/qt-linux-release/build/layout-test-results/fast/repaint/svg-layout-root-style-attr-update-expected.txt 2010-06-23 10:01:08.455453400 -0700 > +++ /home/webkitbuildbot/slaves/release32bit/buildslave/qt-linux-release/build/layout-test-results/fast/repaint/svg-layout-root-style-attr-update-actual.txt 2010-06-23 10:01:08.455453400 -0700 > @@ -11,5 +11,5 @@ > RenderBlock (positioned) {DIV} at (0,0) size 500x500 > layer at (8,0) size 500x500 > RenderSVGRoot {svg} at (8,0) size 500x500 > - RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#FF0000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 Z"] > - RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#008000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 Z"] > + RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#FF0000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 L250.00,10.00"] > + RenderPath {rect} at (258,10) size 250x250 [fill={[type=SOLID] [color=#008000]}] [data="M250.00,10.00 L500.00,10.00 L500.00,260.00 L250.00,260.00 L250.00,10.00"] > > See http://trac.webkit.org/changeset/61696. Every other port formats the rectangle the same way except Qt. From a spot check of LayoutTests/platform/qt this appears to be causing a lot of diffs. This is caused by Qt itself. Qt closes paths automatically if a path crosses itself. Thats why DRT results will never match the results of other ports. We already have bug reports about the different results, the Qt guys told me, that they won't change this behavior. So I think we can close this bug, or if this test is broken, upload a new baseline for Qt.
I've added a new baseline. That's quite unfortunate, are you sure you can't normalize this in Qt's DRT?
Changed component to SVG, so it shows up in my all-svg-bugs search.
I fixed this with bug 46051 :)