On Tue, Aug 24, 2021 at 2:57 PM Rajat Jain <rajatja@xxxxxxxxxx> wrote: > > On Mon, Aug 23, 2021 at 6:06 PM Dan Williams <dan.j.williams@xxxxxxxxx> wrote: > > > > On Mon, Aug 23, 2021 at 5:31 PM Kuppuswamy, Sathyanarayanan > > <sathyanarayanan.kuppuswamy@xxxxxxxxxxxxxxx> wrote: > > > > > > > > > > > > On 8/23/21 4:56 PM, Michael S. Tsirkin wrote: > > > >> Add a new variant of pci_iomap for mapping all PCI resources > > > >> of a devices as shared memory with a hypervisor in a confidential > > > >> guest. > > > >> > > > >> Signed-off-by: Andi Kleen<ak@xxxxxxxxxxxxxxx> > > > >> Signed-off-by: Kuppuswamy Sathyanarayanan<sathyanarayanan.kuppuswamy@xxxxxxxxxxxxxxx> > > > > I'm a bit puzzled by this part. So why should the guest*not* map > > > > pci memory as shared? And if the answer is never (as it seems to be) > > > > then why not just make regular pci_iomap DTRT? > > > > > > It is in the context of confidential guest (where VMM is un-trusted). So > > > we don't want to make all PCI resource as shared. It should be allowed > > > only for hardened drivers/devices. > > > > That's confusing, isn't device authorization what keeps unaudited > > drivers from loading against untrusted devices? I'm feeling like > > Michael that this should be a detail that drivers need not care about > > explicitly, in which case it does not need to be exported because the > > detail can be buried in lower levels. > > > > Note, I specifically said "unaudited", not "hardened" because as Greg > > mentioned the kernel must trust drivers, its devices that may not be > > trusted. > > Can you please point me to the thread where this discussion with Greg > is ongoing? It slowed down to implement the "move to the 'authorized' device model" recommendation. LWN has a good writeup (as usual) and a link to the thread: https://lwn.net/Articles/865918/
