[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: manual selection of devices

Hi MArk
I tried getting md devices detected this way, one thing is md devices
are numbered /dev/md0, /dev/md1 upto /dev/md15, i used "/dev/md{'0'-'9'}
or however the old code went, and that was fine, but it then effectively
detected partitions as drives, which causes a few problems when tryign
to associate a filesystem to it. Because a md type partition could hold
any type of filesystem.

I think it would be good to handle md devices in in a way similar to
this, but i kinda got bogged down trying to make dbootsrap recognise a
drive as a partition. as well.

Changing my approach to manual selection of device was influenced by the
fact that it was a requested feature and it could fix a number of

Actually its been a good learning process im startign to know dbootstrap

Anyway ive got my hire car waiting to be abused, im gunna have to pack
my computer up soon, hertz are gunna hate me im gunna drive >1/2 way
around australia.

I look forward to geting into it in a week and a bit.


Glenn MCGrath

Mark van Walraven wrote:
> Hi Glenn, Adam,
> On Wed, Dec 08, 1999 at 01:14:02AM -0500, Adam Di Carlo wrote:
> > Glenn McGrath <Glenn.McGrath@jcu.edu.au> writes:
> > > 4) Being able to manually select the partition should enable installing
> > > to loop and raid(with some manual work from console) and other
> > > non-standard block devices.
> >
> > That would be great...
> >
> > Are you implementing the stuff where you are sniffing around in
> > /proc/ide and /proc/scsi for devices?  I'm a little skeptical we could
> > get this in time but it would be really nice.
> I think what I'm doing will tie in quite nicely with this.
> I have fdisk_reread() trying devices from the following list (copied
> straight from my new code):
>                 "/dev/sd{a-h}",
>                 "/dev/hd{a-h}",
>                 "/dev/md{a-h}",
>                 "/dev/ida/c{0-7}d{0-15}",
>                 "/dev/rd/c{0-7}d{0-31}",
>         #if #cpu (i386)
>                 "/dev/ed{a-d}",
>         #elif #cpu (m68k)
>                 "/dev/ad{a-h}",
>         #endif
> Which means MD devices should be given to select_not_mounted() through
> fdisk_find_partition_by_type().
> I've been holding back a little because I've been trying to make sure I
> don't include tape drives and CD-ROMs - Erik has been helping me out here.
> I'd probably be the one Adam is thinking of, grubbing around in /proc. :-)
> However, I think I should send in what I have now and add to it later -
> I'll make it a little more respectable (comments, etc.) and post it to
> the list in a couple of hours.
> Regards,
> Mark.

Reply to: