On Tue, May 22, 2018 at 3:30 PM, Warner Losh <i...@bsdimp.com> wrote:
> You can't, in general. By the time the boot loader starts, all knowledge > of past boots is gone, unless specific counter-measures were put in place. > > However, if root is UFS and read/write in your box, it will be unclean on > anything but a clean shutdown/reboot. If it's read-only, ZFS or NFS > mounted, then you can't use this method. > > If you have UEFI, you can set a UEFI variable on shutdown and clear it on > boot. If it's not there on boot, you had an unclean shutdown. You could do > the same with a file in a r/w filesystem that doesn't record clean/unclean > (like ZFS or NFS). > > Locally, we have hacks to IPMI to record kernel crashes in the IPMI log, > but that's kinda specific to the BMC we have on our boards... > > Warner > I see. Thanks for the quick reply. I guess I can add a dummy file somewhere that I delete in a shutdown hook. > > > On Tue, May 22, 2018 at 7:57 AM, Johannes Lundberg <johal...@gmail.com> > wrote: > >> Hi >> >> In the boot process on my test machines I'd like to do different things >> depending on the last run was a clean shutdown or kernel panic. Where/How >> can I get this information? >> >> Thanks! >> _______________________________________________ >> freebsd-current@freebsd.org mailing list >> https://lists.freebsd.org/mailman/listinfo/freebsd-current >> To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org >> " >> > > _______________________________________________ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"