(In reply to Clifford Dugal from comment #14) > Created attachment 671952 [details] > dmesg.txt file with SFW2 lines removed for bnc971695 quickfix > > I was almost ready to conclude that the patched module fixed the issue, but > my system finally crashed again after more than a week of continuous uptime. > Any other ideas? That is a different oops. But it is likely that the revert just papers over a bug that is present upstream, as it widens the window of the race condition drastically. I will report this upstream. Please open a second bugzilla with this new bug and provide the output of "lsusb -v" and "lsusb -t" for the system. Basically we have a driver with a faulty implementation of pre_reset().