On Sun, Jul 30, 2017 at 4:00 PM, Christoffer Dall <cdall@xxxxxxxxxx> wrote:
> On Tue, Jul 18, 2017 at 11:58:46AM -0500, Jintack Lim wrote:
>> When HCR.NV bit is set, eret instructions trap to EL2 with EC code 0x1A.
>> Emulate eret instructions by setting pc and pstate.
>
> It may be worth noting in the commit message that this is all we have to
> do, because the rest of the logic will then discover that the mode could
> change from virtual EL2 to EL1 and will setup the hw registers etc. when
> changing modes.
Makes sense. I'll write it up in the commit message.
>
>>
>> Note that the current exception level is always the virtual EL2, since
>> we set HCR_EL2.NV bit only when entering the virtual EL2. So, we take
>> spsr and elr states from the virtual _EL2 registers.
>>
>> Signed-off-by: Jintack Lim <jintack.lim@xxxxxxxxxx>
>> ---
>> arch/arm64/include/asm/esr.h | 1 +
>> arch/arm64/kvm/handle_exit.c | 16 ++++++++++++++++
>> arch/arm64/kvm/trace.h | 21 +++++++++++++++++++++
>> 3 files changed, 38 insertions(+)
>>
>> diff --git a/arch/arm64/include/asm/esr.h b/arch/arm64/include/asm/esr.h
>> index e7d8e28..210fde6 100644
>> --- a/arch/arm64/include/asm/esr.h
>> +++ b/arch/arm64/include/asm/esr.h
>> @@ -43,6 +43,7 @@
>> #define ESR_ELx_EC_HVC64 (0x16)
>> #define ESR_ELx_EC_SMC64 (0x17)
>> #define ESR_ELx_EC_SYS64 (0x18)
>> +#define ESR_ELx_EC_ERET (0x1A)
>> /* Unallocated EC: 0x19 - 0x1E */
>> #define ESR_ELx_EC_IMP_DEF (0x1f)
>> #define ESR_ELx_EC_IABT_LOW (0x20)
>> diff --git a/arch/arm64/kvm/handle_exit.c b/arch/arm64/kvm/handle_exit.c
>> index 17d8a16..9259881 100644
>> --- a/arch/arm64/kvm/handle_exit.c
>> +++ b/arch/arm64/kvm/handle_exit.c
>> @@ -147,6 +147,21 @@ static int kvm_handle_unknown_ec(struct kvm_vcpu *vcpu, struct kvm_run *run)
>> return 1;
>> }
>>
>> +static int kvm_handle_eret(struct kvm_vcpu *vcpu, struct kvm_run *run)
>> +{
>> + trace_kvm_nested_eret(vcpu, vcpu_el2_sreg(vcpu, ELR_EL2),
>> + vcpu_el2_sreg(vcpu, SPSR_EL2));
>> +
>> + /*
>> + * Note that the current exception level is always the virtual EL2,
>> + * since we set HCR_EL2.NV bit only when entering the virtual EL2.
>> + */
>> + *vcpu_pc(vcpu) = vcpu_el2_sreg(vcpu, ELR_EL2);
>> + *vcpu_cpsr(vcpu) = vcpu_el2_sreg(vcpu, SPSR_EL2);
>> +
>> + return 1;
>> +}
>> +
>> static exit_handle_fn arm_exit_handlers[] = {
>> [0 ... ESR_ELx_EC_MAX] = kvm_handle_unknown_ec,
>> [ESR_ELx_EC_WFx] = kvm_handle_wfx,
>> @@ -160,6 +175,7 @@ static int kvm_handle_unknown_ec(struct kvm_vcpu *vcpu, struct kvm_run *run)
>> [ESR_ELx_EC_HVC64] = handle_hvc,
>> [ESR_ELx_EC_SMC64] = handle_smc,
>> [ESR_ELx_EC_SYS64] = kvm_handle_sys_reg,
>> + [ESR_ELx_EC_ERET] = kvm_handle_eret,
>> [ESR_ELx_EC_IABT_LOW] = kvm_handle_guest_abort,
>> [ESR_ELx_EC_DABT_LOW] = kvm_handle_guest_abort,
>> [ESR_ELx_EC_SOFTSTP_LOW]= kvm_handle_guest_debug,
>> diff --git a/arch/arm64/kvm/trace.h b/arch/arm64/kvm/trace.h
>> index 7c86cfb..5f40987 100644
>> --- a/arch/arm64/kvm/trace.h
>> +++ b/arch/arm64/kvm/trace.h
>> @@ -187,6 +187,27 @@
>> TP_printk("vcpu: %p, inject exception to vEL2: ESR_EL2 0x%lx, vector: 0x%016lx",
>> __entry->vcpu, __entry->esr_el2, __entry->pc)
>> );
>> +
>> +TRACE_EVENT(kvm_nested_eret,
>> + TP_PROTO(struct kvm_vcpu *vcpu, unsigned long elr_el2,
>> + unsigned long spsr_el2),
>> + TP_ARGS(vcpu, elr_el2, spsr_el2),
>> +
>> + TP_STRUCT__entry(
>> + __field(struct kvm_vcpu *, vcpu)
>> + __field(unsigned long, elr_el2)
>> + __field(unsigned long, spsr_el2)
>> + ),
>> +
>> + TP_fast_assign(
>> + __entry->vcpu = vcpu;
>> + __entry->elr_el2 = elr_el2;
>> + __entry->spsr_el2 = spsr_el2;
>> + ),
>> +
>> + TP_printk("vcpu: %p, eret to elr_el2: 0x%016lx, with spsr_el2: 0x%08lx",
>> + __entry->vcpu, __entry->elr_el2, __entry->spsr_el2)
>> +);
>> #endif /* _TRACE_ARM64_KVM_H */
>>
>> #undef TRACE_INCLUDE_PATH
>> --
>> 1.9.1
>>
>
> Otherwise this patch looks good.
>
> Thanks,
> -Christoffer
--
To unsubscribe from this list: send the line "unsubscribe linux-doc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
