In case a package build fails, we can simply restart the build script without
having to do manual cleanup of the intermediate directories.
Signed-off-by: David Oberhollenzer <goliath@infraroot.at>
Same process as for the build dir. We have one intermediate deploy directory
that the packaes installs it self to, then we package the contents, finally
we nuke the build and deploy directory before starting with the next package.
Signed-off-by: David Oberhollenzer <goliath@infraroot.at>
- Only one checkfile per package, not per stage
- Have to rebuild entire package on failure anyway
- Faster build since we can simply guard toolchain restore
with checkfile test
Signed-off-by: David Oberhollenzer <david.oberhollenzer@tele2.at>
Use $PKGBUILDDIR for building packages. Delete it again after
the deploy is done. If the build fails, the script aborts and it
is kept. If the build and deploy pass, there is no reason to keep
it around.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@tele2.at>
Instead of having a "depends" file with a list of packages, add a
"DEPENDS" variable to the build script.
Generate the rootfs dependencies from a config file stored in the
cfg directory.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@tele2.at>
