Re: [Qemu-devel] [virtio-dev] Re: [PATCH v15 0/2] virtio-crypto: virtio crypto device specification

["Michael S. Tsirkin" <mst@redhat.com>]

On Thu, Jan 12, 2017 at 12:26:24PM +0000, Gonglei (Arei) wrote:
> Hi,
> 
> 
> > 
> > On 01/04/2017 11:10 AM, Gonglei (Arei) wrote:
> > > Hi all,
> > >
> > > I attach the diff files between v14 and v15 for better review.
> > >
> > Hi,
> > 
> > only had a quick look. Will try to come back to this later.
> > 
> That's cool.
> 
> > > diff --git a/virtio-crypto.tex b/virtio-crypto.tex
> > > index 9f7faf0..884ee95 100644
> > > --- a/virtio-crypto.tex
> > > +++ b/virtio-crypto.tex
> > > @@ -2,8 +2,8 @@
> > >
> > >  The virtio crypto device is a virtual cryptography device as well as a kind of
> > >  virtual hardware accelerator for virtual machines. The encryption and
> > > -decryption requests are placed in the data queue and are ultimately handled
> > by the
> > > -backend crypto accelerators. The second queue is the control queue used to
> > create
> > > +decryption requests are placed in any of the data active queues and are
> > ultimately handled by the
> > s/data active/active data/
> > > +backend crypto accelerators. The second kind of queue is the control queue
> > used to create
> > >  or destroy sessions for symmetric algorithms and will control some
> > advanced
> > >  features in the future. The virtio crypto device provides the following crypto
> > >  services: CIPHER, MAC, HASH, and AEAD.
> > 
> > [..]
> > 
> > > ===============The below diff shows the changes of add non-session mode
> > support:
> > >
> > > diff --git a/virtio-crypto.tex b/virtio-crypto.tex
> > > index 884ee95..44819f9 100644
> > > --- a/virtio-crypto.tex
> > > +++ b/virtio-crypto.tex
> > > @@ -26,7 +26,10 @@ N is set by \field{max_dataqueues}.
> > >
> > >  \subsection{Feature bits}\label{sec:Device Types / Crypto Device / Feature
> > bits}
> > >
> > > -None currently defined.
> > > +VIRTIO_CRYPTO_F_CIPHER_SESSION_MODE (1) Session mode is available
> > for CIPHER service.
> > > +VIRTIO_CRYPTO_F_HASH_SESSION_MODE (2) Session mode is available for
> > HASH service.
> > > +VIRTIO_CRYPTO_F_MAC_SESSION_MODE (3) Session mode is available for
> > MAC service.
> > > +VIRTIO_CRYPTO_F_AEAD_SESSION_MODE (4) Session mode is available for
> > AEAD service.
> > >
> > >  \subsection{Device configuration layout}\label{sec:Device Types / Crypto
> > Device / Device configuration layout}
> > >
> > > @@ -208,6 +211,9 @@ Operation parameters are algorithm-specific
> > parameters, output data is the
> > >  data that should be utilized in operations, and input data is equal to
> > >  "operation result + result data".
> > >
> > > +The device can support both session mode (See \ref{sec:Device Types /
> > Crypto Device / Device Operation / Control Virtqueue / Session operation}) and
> > non-session mode, for example,
> > > +As VIRTIO_CRYPTO_F_CIPHER_SESSION feature bit is negotiated, the driver
> > can use session mode for CIPHER service, otherwise it can only use non-session
> > mode.
> > > +
> > 
> > As far as I understand you are adding non-session mode to the mix but
> > providing feature bits for session mode. Would this render the the current
> > implementation non-compliant?
> > 
> You are right, shall we use feature bits for non-session mode for compliancy?
> Or because the spec is on the fly, and some structures in the virtio_crypto.h need to
> be modified, can we keep the compliancy completely?
> 
> Thanks,
> -Gonglei

Since there's a linux driver upstream you must at least
keep compatibility with that.

-- 
MST