Also, if you can access via ssh, please get the kernel messages after the hang happens. Get the output of "dmesg" command. If it's a GPU hang, usually the kernel shows it. Last but not least, try to boot with drm.debug=0x0e option, and get again the kernel messages at the problem. This will give more verbose debug prints from the kernel graphics stack. Please attach these outputs to Bugzilla (not paste :)