Re: [qemu-s390x] [Qemu-devel] [PATCH v2 09/28] s390x/tcg: MVCLU/MVCLE: Process max 4k bytes at a time

[David Hildenbrand]

On 11.09.19 17:05, Richard Henderson wrote:
> On 9/6/19 3:57 AM, David Hildenbrand wrote:
>> +    if (*srclen) {
>> +        /* Copy the src array */
>> +        len = MIN(len, *srclen);
>> +        *destlen -= len;
>> +        *srclen -= len;
>> +        fast_memmove(env, *dest, *src, len, ra);
>> +        *src = wrap_address(env, *src + len);
>> +        *dest = wrap_address(env, *dest + len);
>> +    } else if (wordsize == 1) {
>> +        /* Pad the remaining area */
>> +        *destlen -= len;
>> +        fast_memset(env, *dest, pad, len, ra);
>> +        *dest = wrap_address(env, *dest + len);
> 
> Again, I'm not sure fast_memmove actually handles wrap, yet.

Similarly, fixed by access_prepare(), access_memmove().

> Would it be easier to split at page boundaries rather than a
> fixed 4k length?

Also had that already, can do.

-- 

Thanks,

David / dhildenb