Re: vnd.c 1.254

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Sat, Jan 16, 2016 at 04:54:19PM -0500, Christos Zoulas wrote:
> On Jan 16, 10:36pm, bouyer%antioche.eu.org@localhost (Manuel Bouyer) wrote:
> -- Subject: Re: vnd.c 1.254
> 
> | On Sat, Jan 16, 2016 at 04:27:20PM -0500, Christos Zoulas wrote:
> | > On Jan 16,  8:52pm, bouyer%antioche.eu.org@localhost (Manuel Bouyer) wrote:
> | > -- Subject: Re: vnd.c 1.254
> | > 
> | > | yes.
> | > | I don't know what vnd_cd.cd_ndevs is supposed to be when the
> | > | number of pseudo-device is not hardcoded in the kernel
> | > 
> | > It is supposed to be 4 or the max allocated so far if that's bigger.
> | 
> | Where does the '4' comes from ?
> | Also, when exactly the first 4 devices are in use this would
> | cause vnconfig -l to not show the next devices, even if they
> | exists in /dev.
> 
> fgrep cd_ndevs /usr/src/sys/kern/subr_autoconf.c | fgrep MAX
> 
> | Anyway there seems to be something really wrong here, as it fails at vnd1
> | in my case (after several vnconfig/vnconfig -u)
> 
> I can't reproduce this in -current, and most (all?) the code has been
> pulled up.

It may need several vnconfig/vnconfig -u in various orders to show
up, but we're seeing this on 2 different, independant installations.
I did some vnconfig manipulation, and now vnconfig -l shows again
all devices, free or inuse.

Also, you don't address the problem that, as I understand it and if
the code works properly, vnconfig -l won't show free devices if the
first 4 are in use.

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--