Comment 7 for bug 1557172

Hi, sorry for the lack of update. In the meantime I've seen reports of a few cases of this on actual desktops and laptops, so the workaround of disabling the usbhid dirver via udev isn't really an option there :)

Our setup relies on a handful of out-of-tree modules that don't seem to build on 4.5.2, and since this tends to only happen after about a week of use I don't think I'll be able to find anyone who'd be able to test the 4.5.2 mainline kernel. I do know that these modules compile on the wily kernel, though. Would it be helpful to test against that?

I'll start trying to get some machines trying to reproduce this on the wily kernel, but let me know if there's anything else I can do that might help to debug this.