Re: [PATCH net-next v4 4/5] page_pool: remove PP_FLAG_PAGE_FRAG flag

[Yunsheng Lin <linyunsheng@huawei.com>]

On 2023/6/15 1:19, Jakub Kicinski wrote:
> On Mon, 12 Jun 2023 21:02:55 +0800 Yunsheng Lin wrote:
>>  	struct page_pool_params pp_params = {
>> -		.flags = PP_FLAG_DMA_MAP | PP_FLAG_PAGE_FRAG |
>> -				PP_FLAG_DMA_SYNC_DEV,
>> +		.flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV,
>>  		.order = hns3_page_order(ring),
> 
> Does hns3_page_order() set a good example for the users?
> 
> static inline unsigned int hns3_page_order(struct hns3_enet_ring *ring)
> {
> #if (PAGE_SIZE < 8192)
> 	if (ring->buf_size > (PAGE_SIZE / 2))
> 		return 1;
> #endif
> 	return 0;
> }
> 
> Why allocate order 1 pages for buffers which would fit in a single page?
> I feel like this soft of heuristic should be built into the API itself.

hns3 only support fixed buf size per desc by 512 byte, 1024 bytes, 2048 bytes
4096 bytes, see hns3_buf_size2type(), I think the order 1 pages is for buf size
with 4096 bytes and system page size with 4K, as hns3 driver still support the
per-desc ping-pong way of page splitting when page_pool_enabled is false.

With page pool enabled, you are right that order 0 pages is enough, and I am not
sure about the exact reason we use the some order as the ping-pong way of page
splitting now.
As 2048 bytes buf size seems to be the default one, and I has not heard any one
changing it. Also, it caculates the pool_size using something as below, so the
memory usage is almost the same for order 0 and order 1:

.pool_size = ring->desc_num * hns3_buf_size(ring) /
		(PAGE_SIZE << hns3_page_order(ring)),

I am not sure it worth changing it, maybe just change it to set good example for
the users:) anyway I need to discuss this with other colleague internally and do
some testing before doing the change.

> .
>