Re: bigdisk and scsi_da

On 12/21/06, Scott Long <scottl_at_samsco.org> wrote:
> Rong-en Fan wrote:
> > On 12/20/06, Scott Long <scottl_at_samsco.org> wrote:
> >> Rong-en Fan wrote:
> >> > On 12/20/06, Scott Long <scottl_at_samsco.org> wrote:
> >> >> Rong-en Fan wrote:
> >> >> > It seems to me that scsi_da.c reports the wrong size:
> >> >> >
> >> >> > da1 at mpt0 bus 0 target 0 lun 0
> >> >> > da1: <SILVER TN-6212S-U4D 347G> Fixed Direct Access SCSI-5 device
> >> >> > da1: 320.000MB/s transfers (160.000MHz, offset 127, 16bit), Tagged
> >> >> > Queueing Enabled
> >> >> > da1: 0MB (1 0 byte sectors: 0H 0S/T 0C)
> >> >> >
> >> >> > While geom shows the right one:
> >> >> >
> >> >> > Geom name: da1
> >> >> > Providers:
> >> >> > 1. Name: da1
> >> >> >   Mediasize: 4991221760000 (4.5T)
> >> >> >   Sectorsize: 512
> >> >> >   Mode: r0w0e0
> >> >> >   fwsectors: 63
> >> >> >   fwheads: 255
> >> >> >
> >> >> > Speaking of bigdisk, can gpt modify on-disk table on-fly?
> >> >> >
> >> >> > Regards,
> >> >> > Rong-En Fan
> >> >>
> >> >> Just for reference, what version of FreeBSD is this?
> >> >
> >> > Ah, it's a 6.2-RC1.  I thought I posted to stable_at_...
> >> >
> >>
> >> Ok, it makes a whole lot more sense now.  7-CURRENT has checks to
> >> prevent the divide-by-zero.  I'm still looking into the actual
> >> bug, though.
> >
> > OK. Manually backports rev 1.10 of sys/dev/cam/cam.c solved this
> > divided by zero problem. Could you or mjacob MFC this to RELENG_6
> > and/or RELENG_6_2?
> >
> > Thanks,
> > Rong-En Fan
>
> Does it just prevent the panic, or does it also fix the bogus display too?
>
> Scott
>
>

I think it just fixes the panic. It shows:

da1: 0MB (9274165917825105921 0 byte sectors: 0H 0S/T 0C)

While the real sectors should be 9748480000 (reported via geom
disk).

Thanks
Rong-En Fan