Confidential VMs(CVMs) need to execute hypercall instruction as per the CPU
type. Normally KVM emulates the vmcall/vmmcall instruction by patching
the guest code at runtime. Such a guest memory manipulation by KVM is
not allowed with CVMs and is also undesirable in general.
This series adds support of executing hypercall as per the host cpu vendor.
CPU vendor is queried early during selftest setup and guest setup to be
reused later.
Changes in v5:
1) Incorporated suggestions from Sean -
* Rename the APIs to have "this_cpu*" prefix to better convey the
intent of callers to query cpu vendor of the current cpu
* Squash patches together to cache, share cpu vendor type and replace
current callers of "this_cpu*" with checking the saved host cpu vendor
in a single patch.
Changes in v4:
1) Incoporated suggestions from Sean -
* Added APIs to query host cpu type
* Shared the host cpu type with guests to avoid querying the cpu type
again
* Modified kvm_hypercall to execute vmcall/vmmcall according to host
cpu type.
2) Dropped the separate API for kvm_hypercall.
v4:
https://lore.kernel.org/lkml/20221228192438.2835203-1-vannapurve@xxxxxxxxxx/
Vishal Annapurve (3):
KVM: selftests: x86: Use "this_cpu" prefix for cpu vendor queries
KVM: selftests: x86: Cache host CPU vendor (AMD vs. Intel)
KVM: selftests: x86: Use host's native hypercall instruction in
kvm_hypercall()
.../selftests/kvm/include/x86_64/processor.h | 28 +++++++++--
.../selftests/kvm/lib/x86_64/processor.c | 46 ++++++++-----------
.../selftests/kvm/x86_64/fix_hypercall_test.c | 4 +-
.../selftests/kvm/x86_64/mmio_warning_test.c | 2 +-
.../kvm/x86_64/pmu_event_filter_test.c | 4 +-
.../vmx_exception_with_invalid_guest_state.c | 2 +-
6 files changed, 51 insertions(+), 35 deletions(-)
--
2.39.0.314.g84b9a713c41-goog
