On 01/05/2019 02.05, Jerry Hoemann wrote: > On Mon, Jan 21, 2019 at 08:45:38PM +0000, Rasmus Villemoes wrote: >> >> These patches allow one to set a maximum time for which the kernel >> will feed the watchdog, > > Rasmus, > > Sorry if I missed it, but are you still looking into adding > this feature? I am, and still intend to get around to it when I get some idle cycles. I was thinking it might also be useful with > kdump when a watchdog was running in the first kernel. > > After a panic if kdump is configured, the system will boot the > crash kernel. If a watchdog was running in the first kernel > it would still running in the crash kernel environment. > > Some of the drivers on HPE systems take a non-trivial amount of time > to reset during the crash kernel boot, so it would be good to have > the core pet the watchdog until user space is ready. But as the > crash kernel environment has its issues, we really don't want > the core to ping the watchdog indefinitely. Thanks, it's great to hear that others would also be interested in this. This will move it up a few places on my todo list. I'll try to remember to add you to the cc list. Thanks, Rasmus
