On Tue, Oct 17 2023, Peter Maydell <peter.maydell@xxxxxxxxxx> wrote:
> On Tue, 10 Oct 2023 at 15:25, Cornelia Huck <cohuck@xxxxxxxxxx> wrote:
>>
>> We can use read_sys_reg64 to get the SVE_VLS register instead of
>> calling GET_ONE_REG directly.
>>
>> Suggested-by: Gavin Shan <gshan@xxxxxxxxxx>
>> Signed-off-by: Cornelia Huck <cohuck@xxxxxxxxxx>
>> ---
>> target/arm/kvm64.c | 6 +-----
>> 1 file changed, 1 insertion(+), 5 deletions(-)
>>
>> diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c
>> index 558c0b88dd69..d40c89a84752 100644
>> --- a/target/arm/kvm64.c
>> +++ b/target/arm/kvm64.c
>> @@ -500,10 +500,6 @@ uint32_t kvm_arm_sve_get_vls(CPUState *cs)
>> .target = -1,
>> .features[0] = (1 << KVM_ARM_VCPU_SVE),
>> };
>> - struct kvm_one_reg reg = {
>> - .id = KVM_REG_ARM64_SVE_VLS,
>> - .addr = (uint64_t)&vls[0],
>> - };
>> int fdarray[3], ret;
>>
>> probed = true;
>> @@ -512,7 +508,7 @@ uint32_t kvm_arm_sve_get_vls(CPUState *cs)
>> error_report("failed to create scratch VCPU with SVE enabled");
>> abort();
>> }
>> - ret = ioctl(fdarray[2], KVM_GET_ONE_REG, ®);
>> + ret = read_sys_reg64(fdarray[2], &vls[0], KVM_REG_ARM64_SVE_VLS);
>> kvm_arm_destroy_scratch_host_vcpu(fdarray);
>> if (ret) {
>> error_report("failed to get KVM_REG_ARM64_SVE_VLS: %s",
>
> read_sys_reg64() asserts that the register you're trying to
> read is 64 bits, but KVM_REG_ARM64_SVE_VLS is not, it's 512 bits:
>
> #define KVM_REG_ARM64_SVE_VLS (KVM_REG_ARM64 | KVM_REG_ARM64_SVE | \
> KVM_REG_SIZE_U512 | 0xffff)
>
> So this change would trip the assert on a host where SVE
> is supported and enabled.
Whoops, it seems that I misread this. (And my test environment didn't
have that enabled...)
