Re: [PATCH v2 07/13] KVM: selftests: add library for creating/interacting with SEV guests

From: Paolo Bonzini <pbonzini@redhat.com>
To: Michael Roth <michael.roth@amd.com>, Peter Gonda <pgonda@google.com>
Cc: linux-kselftest@vger.kernel.org, kvm list <kvm@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>,
	the arch/x86 maintainers <x86@kernel.org>,
	Nathan Tempelman <natet@google.com>,
	Marc Orr <marcorr@google.com>,
	Steve Rutherford <srutherford@google.com>,
	Sean Christopherson <seanjc@google.com>,
	Mingwei Zhang <mizhang@google.com>,
	Brijesh Singh <brijesh.singh@amd.com>,
	Tom Lendacky <thomas.lendacky@amd.com>,
	Varad Gautam <varad.gautam@suse.com>,
	Shuah Khan <shuah@kernel.org>,
	Vitaly Kuznetsov <vkuznets@redhat.com>,
	David Woodhouse <dwmw@amazon.co.uk>,
	Ricardo Koller <ricarkol@google.com>,
	Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
	"H . Peter Anvin" <hpa@zytor.com>,
	Krish Sadhukhan <krish.sadhukhan@oracle.com>
Subject: Re: [PATCH v2 07/13] KVM: selftests: add library for creating/interacting with SEV guests
Date: Wed, 22 Dec 2021 15:52:49 +0100	[thread overview]
Message-ID: <496b5209-6b1c-1205-ed0e-733af6386f8d@redhat.com> (raw)
In-Reply-To: <20211217161758.pgpvzlgu4z6vhq7x@amd.com>

On 12/17/21 17:17, Michael Roth wrote:
>>> +void kvm_sev_ioctl(struct sev_vm *sev, int cmd, void *data)
>>> +{
>>> +       struct kvm_sev_cmd arg = {0};
>>> +       int ret;
>>> +
>>> +       arg.id = cmd;
>>> +       arg.sev_fd = sev->fd;
>>> +       arg.data = (__u64)data;
>>> +
>>> +       ret = ioctl(vm_get_fd(sev->vm), KVM_MEMORY_ENCRYPT_OP, &arg);
>> If the helper vm_get_fd() exists why not add another which takes a
>> struct sev_vm. So you can do __vm_get_fd(sev) here?
> I can add it as a local helper for now, maybe sev_get_kvm_fd(), to
> distinguish from the SEV_PATH fd? I'm not sure it's worth exporting it
> as part of the library though since vm_get_fd(sev_get_vm(sev)) would be
> more familiar to callers that are already used to the kvm_util library.
> 

I also prefer the one that you suggest.

>> Can you dedup this from  sev_ioctl() in sev_migrate_tests.c? That
>> function already correctly asserts the fw_error.
> 
> This is a little bit awkward since sev_ioctl() in sev_migrate_tests opens
> SEV_PATH on demand whereas this one pulls it out of struct sev_vm. I
> could make kvm_sev_ioctl() expect the KVM fd as a parameter but that
> means external callers need another call to pull it out of struct
> sev_vm.

Yeah, it's a bit weird because sev_migrate_tests do not use struct 
sev_vm.  Unless you port them first, you could have both 
kvm_vm_sev_ioctl that takes a struct kvm_vm, and sev_vm_ioctl that takes 
a struct sev_vm.  Then you only need to change the argument of 
verify_mirror_allowed_cmds to struct kvm_vm.

Paolo