Re: [PATCH v3 51/75] x86/sev-es: Handle MMIO events

[Joerg Roedel <joro@xxxxxxxxxx>]

On Tue, May 19, 2020 at 11:32:02PM -0700, Sean Christopherson wrote:
> '0' is a valid physical address.  It happens to be reserved in the kernel
> thanks to L1TF, but using '0' as an error code is ugly.  Not to mention
> none of the callers actually check the result.

Right, I changed the function to better handle error cases and added
checks to the call-sites. It looks like below now:

static bool vc_slow_virt_to_phys(struct ghcb *ghcb, struct es_em_ctxt *ctxt,
                                 unsigned long vaddr, phys_addr_t *paddr)
{
        unsigned long va = (unsigned long)vaddr;
        unsigned int level;
        phys_addr_t pa;
        pgd_t *pgd;
        pte_t *pte;

        pgd = pgd_offset(current->active_mm, va);
        pte = lookup_address_in_pgd(pgd, va, &level);
        if (!pte) {
                ctxt->fi.vector     = X86_TRAP_PF;
                ctxt->fi.cr2        = vaddr;
                ctxt->fi.error_code = 0;

                if (user_mode(ctxt->regs))
                        ctxt->fi.error_code |= X86_PF_USER;

                return false;
        }

        pa = (phys_addr_t)pte_pfn(*pte) << PAGE_SHIFT;
        pa |= va & ~page_level_mask(level);

        *paddr = pa;

        return true;
}