I have a machine that reproduces this bug now. According to strace, X isn't even waking up when the corruption occurs, so I doubt it's an X bug. It looks to me as if the kernel is overwriting video memory by mistake. The fbcon and vesafb modules are loaded, and X is using the vesa driver.
I have a machine that reproduces this bug now. According to strace, X isn't even waking up when the corruption occurs, so I doubt it's an X bug. It looks to me as if the kernel is overwriting video memory by mistake. The fbcon and vesafb modules are loaded, and X is using the vesa driver.