On Tue, Aug 12, 2008 at 04:39:05PM +0200, Jim Meyering wrote:
"Daniel P. Berrange" <berrange(a)redhat.com> wrote:
> The re-architecting of the LXC controller/container process relationship
> in the previous patch removed the last obstacle to switching over to the
> generic domain XML routines. So this patch switches the driver over.
>
> First the vast majority of lxc_conf.h/c is simply deleted - this is all
> redundant when using the domain_conf.h APIs. Then, all references to lxc_vm_t
> are changed to virDomainObj, and lxc_vm_def_t switches to virDomainDef.
> Finally the LXC driver registers its capabilities data. For this I have
> chosen an OS type of 'exe', since the 'operating system' we're
running
> in the container is just any plain executable process.
>
>
> lxc_conf.c | 1052 +------------------------------------------------------
> lxc_conf.h | 121 ------
> lxc_container.c | 23 -
> lxc_container.h | 2
> lxc_controller.c | 4
> lxc_controller.h | 2
> lxc_driver.c | 289 ++++++++-------
> 7 files changed, 215 insertions(+), 1278 deletions(-)
All looks fine.
ACK.
However, please note that it would have been a lot easier/quicker
to review if you'd done the mechanical/automatable changes
(i.e., the global substitutions like s/lxc_vm_t/virDomainObj/)
separately from the others.
If i'd separated out the substitution, then the intermediate state between
the two patches would not be functional. I prefer to have the changes
fully operational at each step so you can bisect change history.
Daniel
--
|: Red Hat, Engineering, London -o-
http://people.redhat.com/berrange/ :|
|:
http://libvirt.org -o-
http://virt-manager.org -o-
http://ovirt.org :|
|:
http://autobuild.org -o-
http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|