Re: [PATCH] x86/mm: Ensure useful progress in alloc_l2_table()

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 03/07/13 13:16, Jan Beulich wrote:
>>>> On 03.07.13 at 13:37, Andrew Cooper <andrew.cooper3@citrix.com> wrote:
>> --- a/xen/arch/x86/mm.c
>> +++ b/xen/arch/x86/mm.c
>> @@ -1278,13 +1278,6 @@ static int alloc_l2_table(struct page_info *page, unsigned long type,
>>  
>>      for ( i = page->nr_validated_ptes; i < L2_PAGETABLE_ENTRIES; i++ )
>>      {
>> -        if ( preemptible && i && hypercall_preempt_check() )
>> -        {
>> -            page->nr_validated_ptes = i;
>> -            rc = -EAGAIN;
>> -            break;
>> -        }
>> -
>>          if ( !is_guest_l2_slot(d, type, i) ||
>>               (rc = get_page_from_l2e(pl2e[i], pfn, d)) > 0 )
>>              continue;
>> @@ -1299,6 +1292,14 @@ static int alloc_l2_table(struct page_info *page, unsigned long type,
>>          }
>>  
>>          adjust_guest_l2e(pl2e[i], d);
>> +
>> +        if ( preemptible && i != L2_PAGETABLE_ENTRIES
>         if ( preemptible && i != L2_PAGETABLE_ENTRIES - 1
>
> But an even smaller change would be to change the if() you remove 
> above to
>
>         if ( preemptible && i > page->nr_validated_ptes &&
>             hypercall_preempt_check() )
>
> Jan

So it will - I shall resend v2

~Andrew

>
>> +             && hypercall_preempt_check() )
>> +        {
>> +            page->nr_validated_ptes = i + 1;
>> +            rc = -EAGAIN;
>> +            break;
>> +        }
>>      }
>>  
>>      if ( rc >= 0 && (type & PGT_pae_xen_l2) )
>