[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZmDuuRTkG2Y1LcTR@google.com>
Date: Wed, 5 Jun 2024 16:03:21 -0700
From: Sean Christopherson <seanjc@...gle.com>
To: Paolo Bonzini <pbonzini@...hat.com>, linux-kernel@...r.kernel.org,
Xiangfei Ma <xiangfeix.ma@...el.com>, Yan Zhao <yan.y.zhao@...el.com>
Subject: Re: [PATCH 0/2] nVMX: Ensure host's PAT stays sane
On Wed, Jun 05, 2024, Sean Christopherson wrote:
> Fix a bug(s) where test_load_host_pat clobbers PAT and causes all
> subsequent tests to effectively run with all memory mapped as UC. Xiangfei
> first noticed that this caused rdtsc_vmexit_diff_test to fail, but it can
> also lead to timeouts if more tests are added, i.e. if more stuff runs after
> test_load_host_pat.
>
> Sean Christopherson (2):
> nVMX: Ensure host's PAT is loaded at the end of all VMX tests
> nVMX: Verify KVM actually loads the value in HOST_PAT into the PAT MSR
>
> x86/vmx.c | 4 +++-
> x86/vmx_tests.c | 10 +++++++++-
> 2 files changed, 12 insertions(+), 2 deletions(-)
Doh, this is for kvm-unit-tests. Been a while since I sent KUT patches...
Powered by blists - more mailing lists