Hi Marc,

On Fri, Jan 21, 2022 at 6:42 PM Marc Zyngier maz@kernel.org wrote:

Injecting an exception into a guest with non-VHE is risky business. Instead of writing in the shadow register for the switch code to restore it, we override the CPU register instead. Which gets overriden a few instructions later by said restore code.

I see that in __sysreg_restore_el1_state(), which as you said is called after __vcpu_write_spsr().

The result is that although the guest correctly gets the exception, it will return to the original context in some random state, depending on what was there the first place... Boo.

Fix the issue by writing to the shadow register. The original code is absolutely fine on VHE, as the state is already loaded, and writing to the shadow register in that case would actually be a bug.

Which happens via kvm_vcpu_load_sysregs_vhe() calling __sysreg_restore_el1_state() before __vcpu_write_spsr() in this case.

Reviewed-by: Fuad Tabba tabba@google.com

Cheers, /fuad

Fixes: bb666c472ca2 ("KVM: arm64: Inject AArch64 exceptions from HYP") Cc: stable@vger.kernel.org Signed-off-by: Marc Zyngier maz@kernel.org

---

arch/arm64/kvm/hyp/exception.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kvm/hyp/exception.c b/arch/arm64/kvm/hyp/exception.c index 0418399e0a20..c5d009715402 100644 --- a/arch/arm64/kvm/hyp/exception.c +++ b/arch/arm64/kvm/hyp/exception.c @@ -38,7 +38,10 @@ static inline void __vcpu_write_sys_reg(struct kvm_vcpu *vcpu, u64 val, int reg)

static void __vcpu_write_spsr(struct kvm_vcpu *vcpu, u64 val) {

•   write_sysreg_el1(val, SYS_SPSR);
  

•   if (has_vhe())
  

•           write_sysreg_el1(val, SYS_SPSR);
  

•   else
  

•           __vcpu_sys_reg(vcpu, SPSR_EL1) = val;
  

}

static void __vcpu_write_spsr_abt(struct kvm_vcpu *vcpu, u64 val)

2.34.1