Re: [PATCH net-next v2 09/15] mm: page_frag: reuse MSB of 'size' field for pfmemalloc

[Yunsheng Lin <linyunsheng@huawei.com>]

On 2024/4/17 0:22, Alexander H Duyck wrote:
> On Mon, 2024-04-15 at 21:19 +0800, Yunsheng Lin wrote:
>> The '(PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE)' case is for the
>> system with page size less than 32KB, which is 0x8000 bytes
>> requiring 16 bits space, change 'size' to 'size_mask' to avoid
>> using the MSB, and change 'pfmemalloc' field to reuse the that
>> MSB, so that we remove the orginal space needed by 'pfmemalloc'.
>>
>> For another case, the MSB of 'offset' is reused for 'pfmemalloc'.
>>
>> Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
>> ---
>>  include/linux/page_frag_cache.h | 13 ++++++++-----
>>  mm/page_frag_cache.c            |  5 +++--
>>  2 files changed, 11 insertions(+), 7 deletions(-)
>>
>> diff --git a/include/linux/page_frag_cache.h b/include/linux/page_frag_cache.h
>> index fe5faa80b6c3..40a7d6da9ef0 100644
>> --- a/include/linux/page_frag_cache.h
>> +++ b/include/linux/page_frag_cache.h
>> @@ -12,15 +12,16 @@ struct page_frag_cache {
>>  	void *va;
>>  #if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE)
>>  	__u16 offset;
>> -	__u16 size;
>> +	__u16 size_mask:15;
>> +	__u16 pfmemalloc:1;
>>  #else
>> -	__u32 offset;
>> +	__u32 offset:31;
>> +	__u32 pfmemalloc:1;
>>  #endif
> 
> This seems like a really bad idea. Using a bit-field like this seems
> like a waste as it means that all the accesses now have to add
> additional operations to access either offset or size. It wasn't as if
> this is an oversized struct, or one that we are allocating a ton of. As
> such I am not sure why we need to optmize for size like this.

For the old 'struct page_frag' use case, there is one 'struct page_frag'
for every socket and task_struct, there may be tens of thousands of
them even in a 32 bit sysmem, which might mean a lof of memory even for
a single byte saving, see patch 13.

> 
>>  	/* we maintain a pagecount bias, so that we dont dirty cache line
>>  	 * containing page->_refcount every time we allocate a fragment.
>>  	 */
>>  	unsigned int		pagecnt_bias;
>> -	bool pfmemalloc;
>>  };