Re: [Qemu-devel] [PATCH v1 15/33] s390x/tcg: Implement VECTOR LOAD TO BLOCK BOUNDARY

[David Hildenbrand <david@redhat.com>]

On 27.02.19 17:08, Richard Henderson wrote:
> On 2/26/19 3:38 AM, David Hildenbrand wrote:
>> +void HELPER(vll)(CPUS390XState *env, void *v1, uint64_t addr, uint64_t bytes)
>> +{
>> +    S390Vector tmp = {};
>> +    int i;
>> +
>> +    bytes = MIN(bytes, 16);
>> +    for (i = 0; i < bytes; i++) {
>> +        uint8_t byte = cpu_ldub_data_ra(env, addr, GETPC());
>> +
>> +        s390_vec_write_element8(&tmp, i, byte);
>> +        addr = wrap_address(env, addr + 1);
>> +    }
> 
> TODO:
> 
>     if (likely(bytes >= 16)) {
>         uint64_t t0 = cpu_ldq_data_ra(env, addr, GETPC());
>         uint64_t t1 = cpu_ldq_data_ra(env, addr, GETPC());

adding + wrapping the address of course.

>         s390_vec_write_element64(v1, 0, t0);
>         s390_vec_write_element64(v1, 1, t1);
>     } else {
>         // byte loop
>     }
> 
> But what you have is correct, so

Makes sense and gets rid of the MIN(), so changed :)

Thanks!

> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> 
> 
> r~
> 


-- 

Thanks,

David / dhildenb