Hotplug can bring up/down interfaces now

Nathan Coulson conathan at
Tue Jul 6 21:36:07 PDT 2004

>> Bryan Kadzban wrote:
>>> Nathan Coulson wrote:
>>>> If you can diff it against my lfs-bootscripts-20040704.tar.bz2
>>>> tarball, I'll make up another testing tarball for people to play
>>>> with.  [Until then, I have to go to work]
>>> Done.  Actually it looks like the previous patches applied without any
>>> issues (they were created against what's in svn), but I did change a
>>> couple things.
>> Okay...hotplug is gone for this release, but what of the work on the
>> network bootscripts?  Will these changes still get to make it into
>> LFS-6.0?  Also, we need a way to order the services both up and down.
>> -- DJ Lucas
>> --
> the 2 lines [and one comment] I added to ifup/down, I would like to leave.
>  (They say, if hotplug is the program running ifup/down, then that
> particular interface must have ONHOTPLUG defined).

[The line is]
if [ "$IN_HOTPLUG" = "1" -a "$ONHOTPLUG" != "yes" ]; then exit 0; fi

> As for the multiple services one, I'll need time to review your
> implementation, although on the surface it looks ok.

Are we waiting for hotplug to mature, or are we avoiding the fact that it
may automate things we wish the user to learn?  Could Hotplug be a
candidate for BLFS?  These 2 questions will help me decide if I should
keep the above.

just to know about what factors worry me,
a) I do not want to replace the ifup and ifdown scripts in BLFS,
b) I do not want to have the user edit their basic scripts to accomidate
this, (in the Bootscript Wiki page, one of the rules I made, was that the
user should not have to edit the scripts, but provide configuration
information for the changes that they want [files in /etc/sysconfig]).
c) I do not want the bootscripts to utilize packages that are not in LFS,
or fixes to fix problems with a specific package.

a, means I should leave the changes in,
b, means I should not have a ifup and ifdown that hotplug would run
and c means I should take the changes out.

[aka, I'm stuck]

More information about the lfs-dev mailing list