diff options
author | Sean Christopherson | 2022-08-31 00:17:05 +0000 |
---|---|---|
committer | Sean Christopherson | 2022-11-30 16:19:43 -0800 |
commit | b93d2ec34ef368bb854289db99d8d6ca7f523e25 (patch) | |
tree | c14c2bcfade60902f282b97e81fff585fa304a2c /drivers/cpuidle | |
parent | 5c8c0b3273822cf982c250a9a19e003e4b315edb (diff) |
KVM: x86: Reword MSR filtering docs to more precisely define behavior
Reword the MSR filtering documentatiion to more precisely define the
behavior of filtering using common virtualization terminology.
- Explicitly document KVM's behavior when an MSR is denied
- s/handled/allowed as there is no guarantee KVM will "handle" the
MSR access
- Drop the "fall back" terminology, which incorrectly suggests that
there is existing KVM behavior to fall back to
- Fix an off-by-one error in the range (the end is exclusive)
- Call out the interaction between MSR filtering and
KVM_CAP_X86_USER_SPACE_MSR's KVM_MSR_EXIT_REASON_FILTER
- Delete the redundant paragraph on what '0' and '1' in the bitmap
means, it's covered by the sections on KVM_MSR_FILTER_{READ,WRITE}
- Delete the clause on x2APIC MSR behavior depending on APIC base, this
is covered by stating that KVM follows architectural behavior when
emulating/virtualizing MSR accesses
Reported-by: Aaron Lewis <aaronlewis@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Link: https://lore.kernel.org/r/20220831001706.4075399-3-seanjc@google.com
Diffstat (limited to 'drivers/cpuidle')
0 files changed, 0 insertions, 0 deletions