Re: [PATCH 1/3] mm/sparse: add likely to mem_section[root] check in sparse_index_init()

[Wei Yang <richard.weiyang@gmail.com>]

On Thu, Aug 23, 2018 at 03:13:39PM +0200, Michal Hocko wrote:
>On Thu 23-08-18 21:07:30, Wei Yang wrote:
>> Each time SECTIONS_PER_ROOT number of mem_section is allocated when
>> mem_section[root] is null. This means only (1 / SECTIONS_PER_ROOT) chance
>> of the mem_section[root] check is false.
>> 
>> This patch adds likely to the if check to optimize this a little.
>
>Could you evaluate how much does this help if any? Does this have any
>impact on the initialization path at all?

Let me test on my 4G machine with this patch :-)

>
>> Signed-off-by: Wei Yang <richard.weiyang@gmail.com>
>> ---
>>  mm/sparse.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>> 
>> diff --git a/mm/sparse.c b/mm/sparse.c
>> index 10b07eea9a6e..90bab7f03757 100644
>> --- a/mm/sparse.c
>> +++ b/mm/sparse.c
>> @@ -78,7 +78,7 @@ static int __meminit sparse_index_init(unsigned long section_nr, int nid)
>>  	unsigned long root = SECTION_NR_TO_ROOT(section_nr);
>>  	struct mem_section *section;
>>  
>> -	if (mem_section[root])
>> +	if (likely(mem_section[root]))
>>  		return -EEXIST;
>>  
>>  	section = sparse_index_alloc(nid);
>> -- 
>> 2.15.1
>> 
>
>-- 
>Michal Hocko
>SUSE Labs

-- 
Wei Yang
Help you, Help me