Teminal issues in -dev

Jeremy Huntwork jhuntwork at linuxfromscratch.org
Thu Oct 16 08:16:09 PDT 2008

Matthew Burgess wrote:
> On Thu, 16 Oct 2008 00:30:46 -0400, Jeremy Huntwork <jhuntwork at linuxfromscratch.org> wrote:
>> I don't believe it's about simply minimalism, but it's about keeping to
>> the point. If it is really beneficial to many users in various
>> circumstances to have such a command in the book, great, add it. But if
>> it's just because you want to make the jhalfs experience nicer, it feels
>> a good deal more appropriate to modify jhalfs.
> I agree completely, and I say this having hit exactly the same problems as Bruce.  The issue, as I understand it, is that once complete, jhalfs throws you out of the chroot.  This results in /dev being emptied such that when you chroot back in pretty much everything is b0rked.
> Maybe this is as easy to fix as adding another option to jhalfs menu configuration that reads something like 'Exit chroot on completion (default N)', then conditionally exiting the chroot based on that option?  The common use case here, I think, would be to be left in the chroot such that one can continue building BLFS.  Or maybe it's not, and I should get out of that bad habit and reboot into my shiny new jhalfs-built LFS and do things properly :-)

I think you're on the right track. I would alter this a little bit 
though. We start jhalfs from outside of chroot. If we hit an error, it 
dumps you back outside of chroot. I think when jhalfs is done, it's only 
logical that you'll be left where you started, ie., the host system.

The real issue is that the --bind mount of /dev is unmounted as are 
/proc and /sys. (BTW, this issue would also affect any attempt to use 
the grub installed in chroot, since that needs the devices corresponding 
to your HDs available.) So maybe the better option is to allow a user to 
specify if they don't want those filesystem's unmounted. Of course you'd 
have to include proper warnings at the end of the build or such...


More information about the lfs-dev mailing list