[libvirt] [PATCH 00/14 v4] Implement virtlockd daemon

[Michal Privoznik]

On 11.12.2012 21:41, Daniel P. Berrange wrote:
> This is an update of
> 
>   https://www.redhat.com/archives/libvir-list/2012-September/msg00816.html
> 
> This series was previously fully acked, but before pushing I
> noticed some problems and stopped. Since then I've done quite
> a few more changes to rebase to latest GIT & follow new best
> practice.
> 
> I've also added support for a config file for the lockd plugin
> and the ability to use custom lockspaces for LVM and SCSI IDs.
> I've also added support to acquire leases based on a SHA256
> sum of the file, instead of directly on the file
> 
> As such I don't feel comfortable pushing, without further
> reviews.
> 
> While we're past the freeze date, I'd like to request an
> exception, for all except the 14th patch. Without the
> 14th patch, this series has minimal impact on existing code
> that is used in libvirt so little regression possibility.
> Further I discovered yesterday that we already have people
> using this code as custom add-on patches, so I'd like to
> help them by including it in 1.0.1
> 
>  .gitignore                            |    5 
>  bootstrap.conf                        |    1 
>  cfg.mk                                |    9 
>  libvirt.spec.in                       |   16 
>  po/POTFILES.in                        |    4 
>  run.in                                |    2 
>  src/Makefile.am                       |  214 ++++-
>  src/internal.h                        |   22 
>  src/libvirt_private.syms              |    2 
>  src/locking/libvirt_lockd.aug         |   33 
>  src/locking/lock_daemon.c             | 1441 ++++++++++++++++++++++++++++++++++
>  src/locking/lock_daemon.h             |   56 +
>  src/locking/lock_daemon_config.c      |  193 ++++
>  src/locking/lock_daemon_config.h      |   50 +
>  src/locking/lock_daemon_dispatch.c    |  434 ++++++++++
>  src/locking/lock_daemon_dispatch.h    |   31 
>  src/locking/lock_driver_lockd.c       |  834 +++++++++++++++++++
>  src/locking/lock_manager.c            |   20 
>  src/locking/lock_manager.h            |    3 
>  src/locking/lock_protocol.x           |   95 ++
>  src/locking/lockd.conf                |   68 +
>  src/locking/test_libvirt_lockd.aug.in |    7 
>  src/locking/virtlockd.init.in         |   93 ++
>  src/locking/virtlockd.service.in      |   13 
>  src/locking/virtlockd.socket.in       |    8 
>  src/locking/virtlockd.sysconf         |    3 
>  src/qemu/qemu.conf                    |   17 
>  src/qemu/qemu_conf.c                  |   12 
>  src/qemu/qemu_conf.h                  |    3 
>  src/qemu/qemu_driver.c                |   24 
>  src/qemu/test_libvirtd_qemu.aug.in    |    2 
>  src/util/storage_file.c               |   30 
>  src/util/storage_file.h               |    4 
>  33 files changed, 3697 insertions(+), 52 deletions(-)
> 
> 
> --
> libvir-list mailing list
> libvir-list at redhat.com
> https://www.redhat.com/mailman/listinfo/libvir-list
> 

You will have yet another merge conflicts. The commit
a4e44e674eb833af10c74af0c297dd5b3e60fa67 is the last one that this set
can be applied on cleanly.

I've conditionally ACKed the rest of patches but I think pushing at this
phase should get either green or red based on upstream decision. I vote
for pushing in ASAP. We will have another -rc probably so still time to
fix bugs.

Michal