Pixman has support for negative strides, but up until now, only stress-test has actually tested it. The following patches add testing of negative strides to several of the test programs. It appears that everything actually works fine with negative strides on x86, but it would be helpful if people can test on other architectures.
In particular, the unsigned byte_stride used in arm_simd_fill() looks wrong to me, though I'm not sure if arm_simd_fill() is actually hit by any of test cases. Soren _______________________________________________ Pixman mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/pixman
