Re: [lkp] [x86/mm/64] 4bc0303aff: BUG: stack guard page was hit at ffffc900001bbff8

[Andy Lutomirski]

On Mon, Aug 8, 2016 at 7:58 AM, Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
> On Mon, 8 Aug 2016 10:56:49 -0400
> Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
>
>> Yeah, looking at the error path, if registering an ops cause a
>> ftrace_bug, then ftrace is completely (well, mostly) which means that
>                                                       ^
> I left out the word...                             disabled
>
>> if the user unregistered the ops, nothing would happen. It which case,
>
>   "In which case"
>
>> if it freed the ops after it failed, then we can have this case.
>>
>
> Not enough coffee on this Monday morning.
>

I suspect I can't usefully reproduce this bug without or without the
patch -- it took 0day weeks to hit it.  Want to just take it via your
tree or send it to Ingo directly?