On 05/20/2019 06:22 AM, Paolo Bonzini wrote:
On 08/05/19 19:43, Krish Sadhukhan wrote:
Ping...
There have been some changes inthe meanwhile so the patches needed some
work to rebase. I hope I haven't butchered them too much, please take a
look at the master branch. :)
They look fine. Thanks !
It seems you have fixed two call sites in test_pat() by replacing
__enter_guest(ABORT_ON_EARLY_VMENTRY_FAIL) with
enter_guest(). There is still one call site we need to fix: the first
for-loop in test_pat(). There we should call enter_guest() instead of
enter_guest_with_invalid_guest_state() because there the guest state is
valid as far as PAT is concerned and so we should abort on both an early
vmentry failure as well as an invalid guest state.
Please let me know if you want to fix it or want me to send a patch.
Thanks,
Paolo
On 4/18/19 2:39 PM, Krish Sadhukhan wrote:
This is the unit test for the "load IA32_PAT" VM-entry control. Patch# 2
builds on top of my previous patch,
[PATCH 6/6 v5][kvm-unit-test nVMX]: Check "load IA32_PAT" on
vmentry of L2 guests
[PATCH 1/2][kvm-unit-test nVMX]: Move the functionality of
enter_guest() to
[PATCH 2/2][kvm-unit-test nVMX]: Check "load IA32_PAT" VM-entry
control on vmentry
x86/vmx.c | 27 +++++++----
x86/vmx.h | 4 ++
x86/vmx_tests.c | 140
++++++++++++++++++++++++++++++++++++++++++++++--------
3 files changed, 143 insertions(+), 28 deletions(-)
Krish Sadhukhan (2):
nVMX: Move the functionality of enter_guest() to
__enter_guest() and make the former a wrapper of the latter
nVMX: Check "load IA32_PAT" VM-entry control on vmentry of
nested guests
