On 12.03.20 17:17, Janosch Frank wrote:
> On 3/12/20 5:10 PM, Christian Borntraeger wrote:
>>
>>
>> On 12.03.20 17:07, Janosch Frank wrote:
>>> On 3/12/20 4:41 PM, Christian Borntraeger wrote:
>>>> On 11.03.20 14:21, Janosch Frank wrote:
>>>>> IO instruction data is routed through SIDAD for protected guests, so
>>>>> adresses do not need to be checked, as this is kernel memory.
>>>>
>>>> Maybe also mention that the data structures for  the I/O instrutions
>>>> start at offset 0 of the SIDA?
>>>
>>> You mean in the comment for get_address_from_regs() or in the commit
>>> message? It would certainly make sense for get_address_from_regs().
>>
>> Maybe both?
>>
> 
> This is what I have now:
> 
> s390x: protvirt: Disable address checks for PV guest IO emulation
> 
> IO instruction data is routed through SIDAD for protected guests, so
> adresses do not need to be checked, as this is kernel memory which is
> always available.
> 
> Also the instruction data always starts at offset 0 of the SIDAD.
> 
> Signed-off-by: Janosch Frank <fran...@linux.ibm.com>
> Reviewed-by: Thomas Huth <th...@redhat.com>
> Reviewed-by: David Hildenbrand <da...@redhat.com>
> Reviewed-by: Christian Borntraeger <borntrae...@de.ibm.com>
> 
> and
> 
> +    /*
> 
> 
> +     * Addresses for protected guests are all offsets into the
> 
> 
> +     * satellite block which holds the IO control structures. Those
> 
> 
> +     * control structures are always starting at offset 0 and are
> 
> 
> +     * always aligned and accessible. So we can return 0 here which
> 
> 
> +     * will pass the following address checks.
> 
> 
> +     */
> 

Looks perfect without the empty lines.


Reply via email to