Re: [PATCH-for-10.1 v2 00/13] hw/arm: Remove virt-2.6 up to virt-2.12 machines

["Daniel P. Berrangé" <berrange@redhat.com>]

On Fri, Jan 17, 2025 at 09:29:40AM +0100, Thomas Huth wrote:
> On 17/01/2025 09.09, Daniel P. Berrangé wrote:
> > On Fri, Jan 17, 2025 at 07:47:15AM +0100, Thomas Huth wrote:
> > > On 16/01/2025 22.13, Philippe Mathieu-Daudé wrote:
> > > > Hi Peter,
> > > > 
> > > > On 16/1/25 15:59, Philippe Mathieu-Daudé wrote:
> > > > > The versioned 'virt' machines up to 2.12 been marked as deprecated
> > > > > two releases ago, and are older than 6 years, so according to our
> > > > > support policy we can remove them. Remove associated dead code.
> > > > 
> > > > > Philippe Mathieu-Daudé (13):
> > > > >     hw/arm/virt: Remove deprecated virt-2.6 machine
> > > > >     hw/arm/virt: Remove VirtMachineClass::no_pmu field
> > > > >     hw/arm/virt: Remove VirtMachineClass::disallow_affinity_adjustment
> > > > >     hw/arm/virt: Remove deprecated virt-2.7 machine
> > > > >     hw/arm/virt: Remove VirtMachineClass::no_its field
> > > > >     hw/arm/virt: Remove deprecated virt-2.8 machine
> > > > >     hw/arm/virt: Remove VirtMachineClass::claim_edge_triggered_timers
> > > > >       field
> > > > >     hw/arm/virt: Remove deprecated virt-2.9 machine
> > > > >     hw/arm/virt: Remove deprecated virt-2.10 machine
> > > > >     hw/arm/virt: Remove deprecated virt-2.11 machine
> > > > >     hw/arm/virt: Remove VirtMachineClass::smbios_old_sys_ver field
> > > > >     hw/arm/virt: Remove deprecated virt-2.12 machine
> > > > >     hw/arm/virt: Remove VirtMachineClass::no_highmem_ecam field
> > > > 
> > > > Please ignore this (reviewed) series for now. I'll rebase it and
> > > > repost after the 10.0 release.
> > > 
> > > Why? IMHO it should be ok to include them now already. While Daniel's macro
> > > only starts the automatic disablement for 10.1, it should be ok to remove
> > > them now already according to our normal deprecation policy: The machines
> > > have been marked as deprecated in the 9.1 release already (via commit
> > > https://gitlab.com/qemu-project/qemu/-/commit/ce80c4fa6ff ), and thus they
> > > have been deprecated since two releases already. So it should be fine to
> > > remove them now, shouldn't it?
> > 
> > No, because as of 9.1.0 we documented that machine types are under the
> > new policy, and these were only deprecated in 9.1.0, hence the new policy
> > applies to them.
> 
> Hm, I guess we could argue now about the wording (I don't see a spot in
> ce80c4fa6ff that says that this only happens starting with 10.1), but in the
> long run, it doesn't really matter much whether we remove these machines
> with 10.0 already or just with 10.1, so let's go with 10.1 instead.

That's the wrong commit - I documented the delayed impl of deletion
in a separate commit, so that we could just revert that part on its
own in 10.1 dev:

  commit c9fd2d9a48ee3c195cf83cc611b87b09f02f0013
  Author: Daniel P. Berrangé <berrange@redhat.com>
  Date:   Thu Jun 20 17:57:37 2024 +0100

    include/hw: temporarily disable deletion of versioned machine types
    
    The new deprecation and deletion policy for versioned machine types is
    being introduced in QEMU 9.1.0.
    
    Under the new policy a number of old machine types (any prior to 2.12)
    would be liable for immediate deletion which would be a violation of our
    historical deprecation and removal policy
    
    Thus automatic deletions (by skipping QOM registration) are temporarily
    gated on existance of the env variable "QEMU_DELETE_MACHINES" / QEMU
    version number >= 10.1.0. This allows opt-in testing of the automatic
    deletion logic, while activating it fully in QEMU >= 10.1.0.
    
    This whole commit should be reverted in the 10.1.0 dev cycle or shortly
    thereafter.
    
    Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
    Reviewed-by: Thomas Huth <thuth@redhat.com>
    Message-ID: <20240620165742.1711389-10-berrange@redhat.com>
    Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>


If a machine type was deprected *before* the 9.1 release cycle, I think
it is fair game to delete it under the historical deprecation rules.

If a machine type was deprecated in 9.1, or later cycles, I think that
deletion is gated until 10.1.0 under the policy applied by that commit
above.

With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|