Re: [PATCH v3 1/3] KVM: x86: Deflect unknown MSR accesses to user space

[Alexander Graf <graf@amazon.com>]

On 01.08.20 01:36, Jim Mattson wrote:
> 
> On Fri, Jul 31, 2020 at 2:50 PM Alexander Graf <graf@amazon.com> wrote:
>>
>> MSRs are weird. Some of them are normal control registers, such as EFER.
>> Some however are registers that really are model specific, not very
>> interesting to virtualization workloads, and not performance critical.
>> Others again are really just windows into package configuration.
>>
>> Out of these MSRs, only the first category is necessary to implement in
>> kernel space. Rarely accessed MSRs, MSRs that should be fine tunes against
>> certain CPU models and MSRs that contain information on the package level
>> are much better suited for user space to process. However, over time we have
>> accumulated a lot of MSRs that are not the first category, but still handled
>> by in-kernel KVM code.
>>
>> This patch adds a generic interface to handle WRMSR and RDMSR from user
>> space. With this, any future MSR that is part of the latter categories can
>> be handled in user space.
>>
>> Furthermore, it allows us to replace the existing "ignore_msrs" logic with
>> something that applies per-VM rather than on the full system. That way you
>> can run productive VMs in parallel to experimental ones where you don't care
>> about proper MSR handling.
>>
>> Signed-off-by: Alexander Graf <graf@amazon.com>
>>
>> ---
>>
>> v1 -> v2:
>>
>>    - s/ETRAP_TO_USER_SPACE/ENOENT/g
>>    - deflect all #GP injection events to user space, not just unknown MSRs.
>>      That was we can also deflect allowlist errors later
>>    - fix emulator case
>>
>> v2 -> v3:
>>
>>    - return r if r == X86EMUL_IO_NEEDED
>>    - s/KVM_EXIT_RDMSR/KVM_EXIT_X86_RDMSR/g
>>    - s/KVM_EXIT_WRMSR/KVM_EXIT_X86_WRMSR/g
>>    - Use complete_userspace_io logic instead of reply field
>>    - Simplify trapping code
>> ---
>>   Documentation/virt/kvm/api.rst  |  62 +++++++++++++++++++
>>   arch/x86/include/asm/kvm_host.h |   6 ++
>>   arch/x86/kvm/emulate.c          |  18 +++++-
>>   arch/x86/kvm/x86.c              | 106 ++++++++++++++++++++++++++++++--
>>   include/trace/events/kvm.h      |   2 +-
>>   include/uapi/linux/kvm.h        |  10 +++
>>   6 files changed, 197 insertions(+), 7 deletions(-)
>>
>> diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst
>> index 320788f81a05..79c3e2fdfae4 100644
>> --- a/Documentation/virt/kvm/api.rst
>> +++ b/Documentation/virt/kvm/api.rst
> 
> The new exit reasons should probably be mentioned here (around line 4866):
> 
> .. note::
> 
>        For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR and
>        KVM_EXIT_EPR the corresponding
> 
> operations are complete (and guest state is consistent) only after userspace
> has re-entered the kernel with KVM_RUN.  The kernel side will first finish
> incomplete operations and then check for pending signals.  Userspace
> can re-enter the guest with an unmasked signal pending to complete
> pending operations.

Great catch, thanks! Updated to also include the two new exit reasons.

> 
> Other than that, my remaining comments are all nits. Feel free to ignore them.
> 
>> +static int kvm_get_msr_user_space(struct kvm_vcpu *vcpu, u32 index)
> 
> Return bool rather than int?

I'm not a big fan of bool returning APIs unless they have an "is" in 
their name. In this case, the most readable path forward would probably 
be an enum:

enum kvm_msr_user_space_retval {
     KVM_MSR_IN_KERNEL,
     KVM_MSR_BOUNCE_TO_USER_SPACE,
};

and then use that in the checks. But that adds a lot of boiler plate for 
a fully internal, only a few dozen LOC big API. I don't think it's worth it.

> 
>> +{
>> +       if (!vcpu->kvm->arch.user_space_msr_enabled)
>> +               return 0;
>> +
>> +       vcpu->run->exit_reason = KVM_EXIT_X86_RDMSR;
>> +       vcpu->run->msr.error = 0;
> 
> Should we clear 'pad' in case anyone can think of a reason to use this
> space to extend the API in the future?

It can't hurt I guess.

> 
>> +       vcpu->run->msr.index = index;
>> +       vcpu->arch.pending_user_msr = true;
>> +       vcpu->arch.complete_userspace_io = complete_emulated_rdmsr;
> 
> complete_userspace_io could perhaps be renamed to
> complete_userspace_emulation (in a separate commit).

I think the complicated part of complete_userspace_io is to know it 
exists and understand how it works. Once you grasp these two bits, the 
name is just an artifact and IMHO easy enough to apply "beyond I/O".

> 
>> +
>> +       return 1;
>> +}
>> +
>> +static int kvm_set_msr_user_space(struct kvm_vcpu *vcpu, u32 index, u64 data)
> 
> Return bool rather than int?

Same replies as above :). I did get fed up with the amount of 
duplication though and created a generalized function in v4 that gets 
called by kvm_get/set_msr_user_space() to ensure that all fields are 
always set.

> 
>> +{
>> +       if (!vcpu->kvm->arch.user_space_msr_enabled)
>> +               return 0;
>> +
>> +       vcpu->run->exit_reason = KVM_EXIT_X86_WRMSR;
>> +       vcpu->run->msr.error = 0;
> 
> Same question about 'pad' as above.
> 
>> +       vcpu->run->msr.index = index;
>> +       vcpu->run->msr.data = data;
>> +       vcpu->arch.pending_user_msr = true;
>> +       vcpu->arch.complete_userspace_io = complete_emulated_wrmsr;
>> +
>> +       return 1;
>> +}
>> +
> 
> Reviewed-by: Jim Mattson <jmattson@google.com>

Thanks a bunch for the review :)


Alex



Amazon Development Center Germany GmbH
Krausenstr. 38
10117 Berlin
Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss
Eingetragen am Amtsgericht Charlottenburg unter HRB 149173 B
Sitz: Berlin
Ust-ID: DE 289 237 879