Re: [PATCH v4 17/21] KVM: arm64: Save ESR_EL2 on guest SError

[James Morse <james.morse@arm.com>]

Hi Marc,

On 31/10/17 05:47, Marc Zyngier wrote:
> On Tue, Oct 31 2017 at  4:26:01 am GMT, Marc Zyngier <marc.zyngier@arm.com> wrote:
>> On Thu, Oct 19 2017 at  4:58:03 pm BST, James Morse <james.morse@arm.com> wrote:
>>> When we exit a guest due to an SError the vcpu fault info isn't updated
>>> with the ESR. Today this is only done for traps.
>>>
>>> The v8.2 RAS Extensions define ISS values for SError. Update the vcpu's
>>> fault_info with the ESR on SError so that handle_exit() can determine
>>> if this was a RAS SError and decode its severity.

>>> diff --git a/arch/arm64/kvm/hyp/switch.c b/arch/arm64/kvm/hyp/switch.c
>>> index af37658223a0..cba6d8ac105c 100644
>>> --- a/arch/arm64/kvm/hyp/switch.c
>>> +++ b/arch/arm64/kvm/hyp/switch.c
>>> @@ -230,13 +230,20 @@ static bool __hyp_text __translate_far_to_hpfar(u64 far, u64 *hpfar)
>>>  	return true;
>>>  }
>>>  
>>> +static void __hyp_text __populate_fault_info_esr(struct kvm_vcpu *vcpu)
>>> +{
>>> +	vcpu->arch.fault.esr_el2 = read_sysreg_el2(esr);
>>> +}
>>> +
>>>  static bool __hyp_text __populate_fault_info(struct kvm_vcpu *vcpu)
>>>  {
>>> -	u64 esr = read_sysreg_el2(esr);
>>> -	u8 ec = ESR_ELx_EC(esr);
>>> +	u8 ec;
>>> +	u64 esr;
>>>  	u64 hpfar, far;
>>>  
>>> -	vcpu->arch.fault.esr_el2 = esr;
>>> +	__populate_fault_info_esr(vcpu);
>>> +	esr = vcpu->arch.fault.esr_el2;
>>> +	ec = ESR_ELx_EC(esr);
>>>  
>>>  	if (ec != ESR_ELx_EC_DABT_LOW && ec != ESR_ELx_EC_IABT_LOW)
>>>  		return true;
>>> @@ -325,6 +332,8 @@ int __hyp_text __kvm_vcpu_run(struct kvm_vcpu *vcpu)
>>>  	 */
>>>  	if (exit_code == ARM_EXCEPTION_TRAP && !__populate_fault_info(vcpu))
>>>  		goto again;
>>> +	else if (ARM_EXCEPTION_CODE(exit_code) == ARM_EXCEPTION_EL1_SERROR)
>>> +		__populate_fault_info_esr(vcpu);
>>>  
>>>  	if (static_branch_unlikely(&vgic_v2_cpuif_trap) &&
>>>  	    exit_code == ARM_EXCEPTION_TRAP) {
>>
>> With this patch, the only case were we don't save ESR_EL2 is when we
>> take an interrupt. I think we should bite the bullet and make it
>> slightly more streamlined, always saving ESR_EL2.

We always read it __guest_exit, just in case we take an SError and have to put
it back.


> Otherwise, an alternative would be to write something like:
> 
> 	if (ARM_EXCEPTION_CODE(exit_code) != ARM_EXCEPTION_IRQ)
>         	vcpu->arch.fault.esr_el2 = read_sysreg_el2(esr);
> 
> which still avoids saving it, and is a lot more readable.

I'll switch to this in the next version.


Thanks,

James