On Thu, Aug 18, 2022 at 08:29:13AM +0000, Klemens Nanni wrote:
> On Wed, Aug 17, 2022 at 07:03:50AM +0000, Miod Vallat wrote:
> > > What is the result if root runs disklabel, and forces it to all zeros?
> >
> > If the root duid is all zeroes, then the only way to refer to the root
> > disk is to use its /dev/{s,w}d* device name, as zero duids are ignored.
>
> I like miod's second diff and it fixes the race for vnd(4).
> I never ran into the issue with softraid(4), but that should not happen
> anymore with it, either.
What's the status on this diff?
The problem still exists and I was reminded by the new installboot
regress suffering from it on an arm64 box.
>
> OK kn
>
> >
> > If you set a zero duid in disklabel(8), setdisklabel() in the kernel
> > will compute a new, non-zero value.
>
> Correct; same for real sd1 and fictitious vnd0.
>
> # disklabel -E sd1
> Label editor (enter '?' for help at any prompt)
>
> sd1> i
> The disklabel UID is currently: c766517084e5e5ce
> duid: [] 0000000000000000
>
> sd1*> l
> # /dev/rsd1c:
> type: SCSI
> disk: SCSI disk
> label: Block Device
> duid: 0000000000000000
> flags:
> bytes/sector: 512
> sectors/track: 63
> tracks/cylinder: 16
> sectors/cylinder: 1008
> cylinders: 203
> total sectors: 204800
> boundstart: 32832
> boundend: 204767
> drivedata: 0
>
> sd1*> w
>
> sd1> l
> # /dev/rsd1c:
> type: SCSI
> disk: SCSI disk
> label: Block Device
> duid: 9ff85059e4960901
> flags:
> bytes/sector: 512
> sectors/track: 63
> tracks/cylinder: 16
> sectors/cylinder: 1008
> cylinders: 203
> total sectors: 204800
> boundstart: 32832
> boundend: 204767
> drivedata: 0
>
> sd1>
>
