Re: [PATCH v3 4/4] x86/sev: register tpm-svsm platform device

[Jarkko Sakkinen <jarkko@kernel.org>]

On Tue, Mar 18, 2025 at 11:44:05AM +0100, Stefano Garzarella wrote:
> On Mon, Mar 17, 2025 at 03:34:10PM +0200, Jarkko Sakkinen wrote:
> > On Fri, Mar 14, 2025 at 11:56:31AM -0500, Tom Lendacky wrote:
> > > On 3/11/25 04:42, Stefano Garzarella wrote:
> > > > SNP platform can provide a vTPM device emulated by SVSM.
> > > >
> > > > The "tpm-svsm" device can be handled by the platform driver added
> > > > by the previous commit in drivers/char/tpm/tpm_svsm.c
> > > >
> > > > The driver will call snp_svsm_vtpm_probe() to check if SVSM is
> > > > present and if it's support the vTPM protocol.
> > > >
> > > > Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
> > > > ---
> > > >  arch/x86/coco/sev/core.c | 8 ++++++++
> > > >  1 file changed, 8 insertions(+)
> > > >
> > > > diff --git a/arch/x86/coco/sev/core.c b/arch/x86/coco/sev/core.c
> > > > index 2166bdff88b7..a2383457889e 100644
> > > > --- a/arch/x86/coco/sev/core.c
> > > > +++ b/arch/x86/coco/sev/core.c
> > > > @@ -2664,6 +2664,11 @@ static struct platform_device sev_guest_device = {
> > > >  	.id		= -1,
> > > >  };
> > > >
> > > > +static struct platform_device tpm_svsm_device = {
> > > > +	.name		= "tpm-svsm",
> > > > +	.id		= -1,
> > > > +};
> > > > +
> > > >  static int __init snp_init_platform_device(void)
> > > >  {
> > > >  	if (!cc_platform_has(CC_ATTR_GUEST_SEV_SNP))
> > > > @@ -2672,6 +2677,9 @@ static int __init snp_init_platform_device(void)
> > > >  	if (platform_device_register(&sev_guest_device))
> > > >  		return -ENODEV;
> > > >
> > > > +	if (platform_device_register(&tpm_svsm_device))
> > > > +		return -ENODEV;
> > > > +
> > > 
> > > You could avoid registering the device if an SVSM isn't present. Not sure
> > > if that is desirable or not.
> > 
> > Is there any use for the device if an SVSM isn't present? :-)
> > 
> > I'd judge it based on that...
> 
> I tried to keep the logic of whether or not the driver is needed all in the
> tpm_svsm_probe()/snp_svsm_vtpm_probe() (where I check for SVSM).
> If you prefer to move some pieces here, though, I'm open.

OK good point, thanks! Let's look the update as a whole and not touch
on this. There's already quite a few pieces moving. Ignore this for
the moment :-)

> 
> Thanks,
> Stefano
> 

BR, Jarkko