Created attachment 50603 [details] Reduced test case Attached reduced testcase. Expected result: filled red rect from about 10,10 to 290,100. Actual result: very thin red lines corresponding to antialiased pixels at the egdes of that rect. The bug appears to be triggered if: 1. There is more than one subpath 2. All subpaths are axis-aligned rectangles 3. Only one of the subpaths is within surface bounds Tweaking any of the coordinates so that there is at least one non-axis-aligned segment or removing the second subpath (the one that is outside the surface bounds) causes the bug to disappear.
Created attachment 50605 [details] Output of test case, cairo 1.10.2
I added your test case under a general BSD-ish licence, please notify me if you object: commit 8d91de49261662bc920739bbb91d22a49016eb81 Author: Krzysztof Kosiński <tweenk.pl@gmail.com> Date: Fri Aug 26 22:59:01 2011 +0100 test: Add test case from bug-40410 Exercises a case where we failed to fill a rectangle... That test case passes for the current master, so until proven otherwise (and we now have the test to do so!), I'm marking this as closed.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.