nanobsd disk info

[Paul Schenkeveld]

Hi,

On Tue, Mar 22, 2005 at 11:23:52AM -0500, Michael W. Lucas wrote:
> On Mon, Mar 21, 2005 at 09:24:53PM +0100, Paul Schenkeveld wrote:
> > Hi,
> > 
> > On Thu, Mar 17, 2005 at 11:51:07AM -0500, Michael W. Lucas wrote:
> > > Hi,
> > > 
> > > I'm obviously missing something here.
> > > 
> > > I'm trying to get nanobsd on a 512Mb flash card.  The only changes
> > > I've made are to set kernconf = net4801 in make.conf and to set the
> > > Makefile disk geometry variables.  "make" chokes on the disk geometry,
> > > so I'm assuming I've set it wrong.
> > > 
> > > My diskinfo shows as:
> > > 
> > > mwl-crash2~;sudo diskinfo -v da0
> > > da0
> > >         512             # sectorsize
> > >         512483840       # mediasize in bytes (489M)
> > >         1000945         # mediasize in sectors
> > >         488             # Cylinders according to firmware.
> > >         64              # Heads according to firmware.
> > >         32              # Sectors according to firmware.
> > 
> > Somehow these numbers are not consistent:
> > 
> >   488 cyls * 64 heads * 32 secs = 999424 secs
> 
> Doh, of course.  I've never seen a hard drive lie to me before, didn't
> occur to me to check these numbers.
> 
> > What does dmesg show about da0?
> 
> da0 at umass-sim0 bus 0 target 0 lun 0
> da0: <Generic STORAGE DEVICE 0.01> Removable Direct Access SCSI-0 device 
> da0: 1.000MB/s transfers
> da0: 488MB (1000945 512 byte sectors: 64H 32S/T 488C)
> 
> I get the same results when I try on a separate machine.  I also get
> the same results when I use a second identical flash card.
> 
> Any suggestions or pointers?  How can I deal with a flash card that's
> obviously lying or brain-dead?  I apparently don't know the right
> Google keywords, because I can't find a solution there...

So what if you just put the following in:

  HD?=64
  SC?=32
  SECTS?=999424

which would match the 488 cylinders reported by dmesg and diskinfo.

Hope you're not desperately trying to squeeze the last bytes out ;-)

> ==ml

Regards,

Paul Schenkeveld