Created attachment 15999 [details] scrrenshot of correct result 32bit (rotate 5 angle) read png image and make pattern. get incorrect result if pattern transform on x86_64 attach screenshot image, cairo-test.tar.gz
Created attachment 16000 [details] scrrenshot of correct result 32bit (scale 5/4)
Created attachment 16001 [details] scrrenshot of incorrect result 64bit (rotate 5 angle)
Created attachment 16002 [details] scrrenshot of incorrect result 64bit (scale 5/4)
Created attachment 16003 [details] test code
Created attachment 16007 [details] C code from test tarball Pulling the C file out of the test tarball for the convenience.
This is very likely a video driver bug. (Your test case works fine for me on this x86_64 machine.) (You can verify that it's a X problem, I think, by writing a similar test case that renders to an image surface rather than a window.) Can you provide details of: - Linux distribution - Video hardware - X server version - Video driver version So we can reassign the bug appropriately?
OS : Linux debian unstable (sid) xserver-xorg version : 1:7.3+10 video driver version : xserver-xorg-video-ati 1:6.8.0-1 fglrx-driver 8.45.5-1 image transformed problems in ati/fglrx driver 64bit video hardware : this is lspci command infomation -------------------------------------------------------------------------- 01:00.0 VGA compatible controller: ATI Technologies Inc RV370 [Sapphire X550 Silent] 01:00.1 Display controller: ATI Technologies Inc RV370 secondary [Sapphire X550 Silent] --------------------------------------------------------------------------
I'm routinely seeing these display artifacts in Firefox, whenever Firefox tries to do client-size image resizing. The two easy-to-reproduce cases are in an HTML page when the IMG tag specifies a size other than the image's actual size, and when viewing an image by itself that's larger than the window, using the fit-to-window scaling. I've got an RV370, using xserver-xorg-video-radeon version 1:6.9.0-1+lenny4 out of Debian x86-64, so it sounds like the same cause.
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/cairo/cairo/issues/63.
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.