RE: [PATCH v4] page_alloc: consider highatomic reserve in watermark fast

[김재원 <jaewon31.kim@samsung.com>]

>On Sat 20-06-20 08:59:58, Jaewon Kim wrote:
>[...]
>> @@ -3502,19 +3525,12 @@ bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
>>          const bool alloc_harder = (alloc_flags & (ALLOC_HARDER|ALLOC_OOM));
>>  
>>          /* free_pages may go negative - that's OK */
>> -        free_pages -= (1 << order) - 1;
>> +        free_pages -= __zone_watermark_unusable_free(z, order, alloc_flags);
>>  
>>          if (alloc_flags & ALLOC_HIGH)
>>                  min -= min / 2;
>>  
>> -        /*
>> -         * If the caller does not have rights to ALLOC_HARDER then subtract
>> -         * the high-atomic reserves. This will over-estimate the size of the
>> -         * atomic reserve but it avoids a search.
>> -         */
>> -        if (likely(!alloc_harder)) {
>> -                free_pages -= z->nr_reserved_highatomic;
>> -        } else {
>> +        if (unlikely(alloc_harder)) {
>>                  /*
>>                   * OOM victims can try even harder than normal ALLOC_HARDER
>>                   * users on the grounds that it's definitely going to be in
>[...]
>> @@ -3582,25 +3591,22 @@ static inline bool zone_watermark_fast(struct zone *z, unsigned int order,
>>                                  unsigned long mark, int highest_zoneidx,
>>                                  unsigned int alloc_flags)
>>  {
>> -        long free_pages = zone_page_state(z, NR_FREE_PAGES);
>> -        long cma_pages = 0;
>> +        long free_pages;
>> +        long unusable_free;
>>  
>> -#ifdef CONFIG_CMA
>> -        /* If allocation can't use CMA areas don't use free CMA pages */
>> -        if (!(alloc_flags & ALLOC_CMA))
>> -                cma_pages = zone_page_state(z, NR_FREE_CMA_PAGES);
>> -#endif
>> +        free_pages = zone_page_state(z, NR_FREE_PAGES);
>> +        unusable_free = __zone_watermark_unusable_free(z, order, alloc_flags);
>>  
>>          /*
>>           * Fast check for order-0 only. If this fails then the reserves
>> -         * need to be calculated. There is a corner case where the check
>> -         * passes but only the high-order atomic reserve are free. If
>> -         * the caller is !atomic then it'll uselessly search the free
>> -         * list. That corner case is then slower but it is harmless.
>> +         * need to be calculated.
>>           */
>> -        if (!order && (free_pages - cma_pages) >
>> -                                mark + z->lowmem_reserve[highest_zoneidx])
>> -                return true;
>> +        if (!order) {
>> +                long fast_free = free_pages - unusable_free;
>> +
>> +                if (fast_free > mark + z->lowmem_reserve[highest_zoneidx])
>> +                        return true;
>> +        }
> 
>There is no user of unusable_free for order > 0. With you current code
>__zone_watermark_unusable_free would be called twice for high order
>allocations unless compiler tries to be clever..

Yes you're right.
Following code could be moved only for order-0.
 unusable_free = __zone_watermark_unusable_free(z, order, alloc_flags);
Let me fix it at v5.

> 
>But more importantly, I have hard time to follow why we need both
>zone_watermark_fast and zone_watermark_ok now. They should be
>essentially the same for anything but order == 0. For order 0 the
>only difference between the two is that zone_watermark_ok checks for
>ALLOC_HIGH resp ALLOC_HARDER, ALLOC_OOM. So what is exactly fast about
>the former and why do we need it these days?
> 

I think the author, Mel, may ansewr. But I think the wmark_fast may
fast by 1) not checking more condition about wmark and 2) using inline
rather than function. According to description on commit 48ee5f3696f6,
it seems to bring about 4% improvement.

>> 
>>          return __zone_watermark_ok(z, order, mark, highest_zoneidx, alloc_flags,
>>                                          free_pages);
>> -- 
>> 2.17.1
>> 
> 
>-- 
>Michal Hocko
>SUSE Labs
>