Re: [Qemu-devel] [PATCH 00/13] RFC: luks/encrypted qcow2 key management

[Maxim Levitsky <mlevitsk@redhat.com>]

On Wed, 2019-08-14 at 16:08 -0500, Eric Blake wrote:
> On 8/14/19 3:22 PM, Maxim Levitsky wrote:
> 
> > This is an issue that was raised today on IRC with Kevin Wolf. Really thanks
> > for the idea!
> > 
> > We agreed that this new qmp interface should take the same options as
> > blockdev-create does, however since we want to be able to edit the encryption
> > slots separately, this implies that we sort of need to allow this on creation
> > time as well.
> > 
> > Also the BlockdevCreateOptions is a union, which is specialized by the driver name
> > which is great for creation, but for update, the driver name is already known,
> > and thus the user should not be forced to pass it again.
> > However qmp doesn't seem to support union type guessing based on actual fields
> > given (this might not be desired either), which complicates this somewhat.
> 
> Does the idea of a union type with a default value for the discriminator
> help?  Maybe we have a discriminator which defaults to 'auto', and add a
> union branch 'auto':'any'.  During creation, if the "driver":"auto"
> branch is selected (usually implicitly by omitting "driver", but also
> possible explicitly), the creation attempt is rejected as invalid
> regardless of the contents of the remaining 'any'.  But during amend
> usage, if the 'auto' branch is selected, we then add in the proper
> "driver":"xyz" and reparse the QAPI object to determine if the remaining
> fields in 'any' still meet the specification for the required driver branch.
> 
> This idea may still require some tweaks to the QAPI generator, but it's
> the best I can come up with for a way to parse an arbitrary JSON object
> with unknown validation, then reparse it again after adding more
> information that would constrain the parse differently.
> 

This could work, but the idea of doing the parsing twice might not be easy to implement.
We currently have the qmp parser completely separated from the rest of the qemu,
so only once the qmp command parsing is done, the corresponding callback is called.


I am thinking. Since any 'update' commmand would need to sepecify the node to work on,
one could add some kind of expression for the qmp frontend to query the driver of that
node itself, which would solve that problem


Something like that:

{ 'union': 'BlockdevAmendOptions',

  'base': {
      'node-name':         'str' },

  'discriminator': { 'get_block_driver(node-name)' } ,

  'data': {
      'file':           'BlockdevCreateOptionsFile',
      'gluster':        'BlockdevCreateOptionsGluster',
      'luks':           'BlockdevCreateOptionsLUKS',
      'nfs':            'BlockdevCreateOptionsNfs',
      'parallels':      'BlockdevCreateOptionsParallels',
      'qcow':           'BlockdevCreateOptionsQcow',
      'qcow2':          'BlockdevCreateOptionsQcow2',
      'qed':            'BlockdevCreateOptionsQed',
      'rbd':            'BlockdevCreateOptionsRbd',
      'sheepdog':       'BlockdevCreateOptionsSheepdog',
      'ssh':            'BlockdevCreateOptionsSsh',
      'vdi':            'BlockdevCreateOptionsVdi',
      'vhdx':           'BlockdevCreateOptionsVhdx',
      'vmdk':           'BlockdevCreateOptionsVmdk',
      'vpc':            'BlockdevCreateOptionsVpc'
  } }


The 'get_block_driver' expression will make the QMP frontend, take the value of the node-name union field,
and look up the block driver associated with it and use that as a discriminator.

Syntax wise we can (at some expense of readability) use json to express the same like

'discriminator': { 'field' : 'node-name', 'transform': 'getdrivername' },

Best regards,
	Maxim Levitsky