Frederick Bruckman wrote:

>On Sat, 10 Feb 2001, Bob Nestor wrote:
>
>> Notice I didn't say it would always screw up.  If you do your own kernel
>> builds on a system with a combined root&usr that is BSD 4.3 format you
>> will probably end up with a problem sooner or later.  The inode number of
>> the kernel will eventually be assigned to a value that the Booter can't
>> get to and your system will be unbootable.  The simplest way to avoid
>> this is to use a separate, small root partition, prefereably one that is
>> BSD 4.2 format.
>
>The work-around is to always create the new kernel at the root level.
>Having a separate root ensures that; size doesn't matter. What happens
>when you have a single /-/usr-/usr/src/sys is that you're tempted to
>type "mv netbsd /" to activate your new kernel, but the data doesn't
>really move -- it's still out there on the high numbered cylinders,
>where the Booter may not find it. However, if you do, for instance,
>"cp -p netbsd /", instead, the data always ends up in the first
>cylinder, and it always seems to work.

Right, and if I remember correctly this happens with both BSD 4.2 and BSD 
4.3 type filesystems.  There are probably other differences between the 
two that can cause problems as well.  One thing to keep in mind is that 
since the Installer can't deal with the BSD 4.3 type filesystem you don't 
have the option of using it and the mini-shell to solve the inode problem 
if it happens.  So I think it's still prudent to use a separate small 
root if you're installing with sysinst.  Multiple filesystems with 
sysinst aren't nearly as difficult to implement for the new user as they 
are with the Mkfs/Installer installation method.

-bob