I'm glad to hear that, but it would still be good to rule out a kernel bug. Can you bisect the kernel change which triggered the problem with the modesetting driver?