Help with building private .iso media

Patrick Mahan PMahan at adaranet.com
Wed Aug 15 19:55:26 UTC 2012


All,

Asking this here as I don't see a mailing list specifically devoted to build/release
issues and because I think it is a little beyond the standard -question genre.

Our product here is based on FreeBSD (currently FreeBSD 9/STABLE) and it has reached
the stage where we need to create our own .iso/memstick media to handle installation
of our appliances.

So yesterday, after perusing release(7), I cranked up /usr/src/release/generate-release.sh
and capture the output to see how it was done.  And I was successful in getting the .iso,
memstick and FTP site created.

So taking what I learned yesterday, I tried to apply it to our copy of the FreeBSD source
tree.  Please note that we use the provided FreeBSD makefiles to build everything we just
make heavy use of MAKEOBJDIRPREFIX/OBJTREE/etc environment variables to control where every-
thing goes.

So today, I tried to build the world package by doing the following at the top of our
FreeBSD source tree (after doing a buildworld) -

  cd /usr/home/pmahan/work/pm_ipr/ipr/src
  export MAKEOBJDIRPREFIX=/usr/home/pmahan/work/pm_ipr/ipr/amd64/obj
  sudo -E make TARGET_ARCH=amd64 TARGET=amd64 distrubteworld DIST=/usr/home/pmahan/work/pm_ipr/ipr/amd64/dist

It failed with the following:

find //usr/home/pmahan/work/pm_ipr/ipr/amd64/dist/doc -empty -delete
find: -delete: //usr/home/pmahan/work/pm_ipr/ipr/amd64/dist/doc: relative path potentially not safe

But when I check the logs for yesterday's successful run, I see -

find //usr/obj/usr/src/release/dist/doc -empty -delete
find //usr/obj/usr/src/release/dist/games -empty -delete
find //usr/obj/usr/src/release/dist/lib32 -empty -delete
sh /usr/src/release/scripts/mm-mtree.sh -m /usr/src/release/.. -F  "TARGET_ARCH=amd64 TARGET=amd64" -D "/usr/obj/usr/src/release/dist/base"

I read find(1) description about '-delete' and did a quick test that seemed to prove out
that attempting to delete a fully qualified path seems to be a no-no.  But I don't understand
why it did not fail yesterday.  I suspect it is pilot error, but I am at a loss to explain why.

Any help, steers, slap downs appreciated

Thanks,

Patrick
----------------------------------------------------
Patrick Mahan
Lead Technical Kernel Engineer
Adara Networks
Disclaimer: The opinions expressed here are solely the responsibility of the author and are not to be
construed as an official opinion of Adara Networks.




More information about the freebsd-hackers mailing list