From: Pat Mackinlay (smackinla@cc.curtin.edu.au)
Date: 01/29/93


From: smackinla@cc.curtin.edu.au (Pat Mackinlay)
Subject: Re: Bernoulli/SyQuest
Date: Fri, 29 Jan 1993 15:32:04 GMT


>>slow since this is a Syquest), and (2) although the Syquest is a
>>"mountable" hard disk, the partition table on the cartridge is only
>>read once during boot up, so you are out of luck if you want to umount
>>and mount a different cartridge while the system is up and running.
>
> Then is it possible to change cartridges while up, if they are partitioned
> the same?

Probably possible, but generally not a good idea. What really needs to happen
is for someone to hack the mount code to try the BLKRRPART ioctl on a device
that's about to be mounted. This should probably not return an error if the
call fails with -EINVAL, as not all device drivers support this yet
(although most do). A message saying the call has been successful could be
helpful however...

This brings me to another point: there is absolutely _no_ reason why we
need to reboot the machine after modifying the partition table. This
behaviour is an MS-DOGism that should be gotten rid of. As I said before,
most device drivers now support the BLKRRPART ioctl, allowing the partition
table to be re-read after being modified. If fdisk were modified to support
this ioctl, we wouldn't need to reboot to install any more (apart from
booting off the hard disk...)

If no one has already done this, I'll get around to doing it when I have
some time - it's really not difficult...

BTW: I think it's quite funny that even OS/2 still requires a reboot after
modifying the partition table - I think everyone caught the "partition
table is something magical" disease from MS-DOG <grin>.

Pat -- Have a day!