On Fri, Jan 31, 2020 at 12:19:39PM +0100, Dmitry Vyukov wrote: > I see lots of people also mention panic_on_warn in the context of > these reports. panic_on_warn here is only a red herring. It really > does not change anything. We could remove it, but still report > WARNINGs. But syzkaller also reports some things that don't panic > anyway. This is really about the criteria for kernel bug vs non-bug > (something that needs to be reported or not). Mentioning panic_on_warn is relevant to determine whether a fix needs to be backported or not. Some of the bugs in question are mostly benign in the sense that they are unlikely to crash your machine, but we'd still want them in in stable due to panic_on_warn and automatic testing. Johan
