[Qemu-devel] KVM call minutes for 2014-02-18

[Qemu-devel] KVM call minutes for 2014-02-18

[Juan Quintela]

2014-02-18
----------

* [Qemu-devel] [PATCH V17 00/11] Add support for binding guest numa
  nodes
  Any news about this? (Vinod)

* Should we change anything to get more people to sign for the call?
  There hasn't been a call in quite a long time.  Ideas? (me)

* x2apic?

-  Pending patch for cpu feature flag.
-  Should this be the default.
-  It is not for 32bits, but should it be for 64bit?
-  libvirt always use x2apic, unconditionally?
-  What happens if one side of migration uses -m <cpu_something and the
   other -m cpu_something+x2apic
-  disabling if for all old machine types
-  Use the same that hyperv flags?


* release
- No Anthony
- People happy (on happy enough) with current schedule

* Maintenance?
- How is being handling patches for the stable release?
  CC: stable@  for stable maintance.
- add documentation about what/how to send patches for stable release
- Regressions deserve a backport almost always?
- There is no stable tree public where we see what patches are being integrated?
- How can we make more people additoally to Michael to work with stable branches
- Should we change anynthing?

[Qemu-devel] KVM call minutes for 2014-02-18

[Juan Quintela]

2014-02-18
----------

* x2apic?

-  Pending patch for cpu feature flag.
-  Should this be the default.
-  It is not for 32bits, but should it be for 64bit?
-  libvirt always use x2apic, unconditionally?
-  What happens if one side of migration uses -m <cpu_something and the
   other -m cpu_something+x2apic
-  disabling if for all old machine types
-  Use the same that hyperv flags?


* release
- No Anthony
- People happy (on happy enough) with current schedule

* Maintenance?
- How is being handling patches for the stable release?
  CC: stable@  for stable maintance.
- add documentation about what/how to send patches for stable release
- Regressions deserve a backport almost always?
- There is no stable tree public where we see what patches are being integrated?
- How can we make more people additoally to Michael to work with stable branches
- Should we change anynthing?

Re: [Qemu-devel] KVM call minutes for 2014-02-18

[Laszlo Ersek]

On 02/18/14 16:05, Juan Quintela wrote:

> * Maintenance?
> - How is being handling patches for the stable release?
>   CC: stable@  for stable maintance.

According to <http://wiki.qemu.org/Planning/1.7>, 2014-02-24 is the
planned date for the 1.7.1 freeze (ie. in six days).

I checked <https://github.com/mdroth/qemu> and there's no
stable-1.7-staging branch yet.

I'd really like

  ddaaefb piix: fix 32bit pci hole

to get into 1.7.1. Can I help somehow to make it happen?

(Mike okayed it before, see:

  http://thread.gmane.org/gmane.comp.emulators.qemu/243589/focus=250799

I'd just like to ensure that it's not forgotten.)

If I'm looking in the wrong place, then I apologize for the noise.

Thanks!
Laszlo

Re: [Qemu-devel] KVM call minutes for 2014-02-18

[Peter Maydell]

On 18 February 2014 15:05, Juan Quintela <quintela@redhat.com> wrote:
> * Maintenance?
> - How is being handling patches for the stable release?
>   CC: stable@  for stable maintance.
> - add documentation about what/how to send patches for stable release
> - Regressions deserve a backport almost always?
> - There is no stable tree public where we see what patches are being integrated?
> - How can we make more people additoally to Michael to work with stable branches
> - Should we change anynthing?

So the thing I really wanted to get out of this discussion as a result is:
 - People who put the stable branch into distros (or with contacts
   with those who do) to jointly figure out what the criteria are for
   putting a fix into stable, and write the answer on a wiki page
   so submaintainers can apply a reasonably consistent approach

thanks
-- PMM