[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <3b043a95-a4bc-bbaf-c8e0-240e8ddea62f@redhat.com>
Date: Fri, 14 Jul 2023 15:13:15 +0200
From: Jesper Dangaard Brouer <jbrouer@...hat.com>
To: Haiyang Zhang <haiyangz@...rosoft.com>,
Jesper Dangaard Brouer <jbrouer@...hat.com>,
Jakub Kicinski <kuba@...nel.org>
Cc: brouer@...hat.com,
"linux-hyperv@...r.kernel.org" <linux-hyperv@...r.kernel.org>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
Dexuan Cui <decui@...rosoft.com>,
KY Srinivasan <kys@...rosoft.com>,
Paul Rosswurm <paulros@...rosoft.com>,
"olaf@...fle.de" <olaf@...fle.de>,
"vkuznets@...hat.com" <vkuznets@...hat.com>,
"davem@...emloft.net" <davem@...emloft.net>,
"wei.liu@...nel.org" <wei.liu@...nel.org>,
"edumazet@...gle.com" <edumazet@...gle.com>,
"pabeni@...hat.com" <pabeni@...hat.com>,
"leon@...nel.org" <leon@...nel.org>,
Long Li <longli@...rosoft.com>,
"ssengar@...ux.microsoft.com" <ssengar@...ux.microsoft.com>,
"linux-rdma@...r.kernel.org" <linux-rdma@...r.kernel.org>,
"daniel@...earbox.net" <daniel@...earbox.net>,
"john.fastabend@...il.com" <john.fastabend@...il.com>,
"bpf@...r.kernel.org" <bpf@...r.kernel.org>,
"ast@...nel.org" <ast@...nel.org>,
Ajay Sharma <sharmaajay@...rosoft.com>,
"hawk@...nel.org" <hawk@...nel.org>,
"tglx@...utronix.de" <tglx@...utronix.de>,
"shradhagupta@...ux.microsoft.com" <shradhagupta@...ux.microsoft.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Ilias Apalodimas <ilias.apalodimas@...aro.org>
Subject: Re: [PATCH net-next] net: mana: Add page pool for RX buffers
On 14/07/2023 14.51, Haiyang Zhang wrote:
>
>
>> -----Original Message-----
>> From: Jesper Dangaard Brouer <jbrouer@...hat.com>
>> On 14/07/2023 05.53, Jakub Kicinski wrote:
>>> On Thu, 13 Jul 2023 14:48:45 +0000 Haiyang Zhang wrote:
>>>> Add page pool for RX buffers for faster buffer cycle and reduce CPU
>>>> usage.
>>>>
>>>> Get an extra ref count of a page after allocation, so after upper
>>>> layers put the page, it's still referenced by the pool. We can reuse
>>>> it as RX buffer without alloc a new page.
>>>
>>> Please use the real page_pool API from include/net/page_pool.h
>>> We've moved past every driver reinventing the wheel, sorry.
>>
>> +1
>>
>> Quoting[1]: Documentation/networking/page_pool.rst
>>
>> Basic use involves replacing alloc_pages() calls with the
>> page_pool_alloc_pages() call.
>> Drivers should use page_pool_dev_alloc_pages() replacing
>> dev_alloc_pages().
>
> Thank Jakub and Jesper for the reviews.
> I'm aware of the page_pool.rst doc, and actually tried it before this
> patch, but I got lower perf. If I understand correctly, we should call
> page_pool_release_page() before passing the SKB to napi_gro_receive().
>
> I found the page_pool_dev_alloc_pages() goes through the slow path,
> because the page_pool_release_page() let the page leave the pool.
>
> Do we have to call page_pool_release_page() before passing the SKB
> to napi_gro_receive()? Any better way to recycle the pages from the
> upper layer of non-XDP case?
>
Today SKB "upper layers" can recycle page_pool backed packet data/page.
Just use skb_mark_for_recycle(skb), then you don't need
page_pool_release_page().
I guess, we should update the documentation, mentioning this.
--Jesper
Powered by blists - more mailing lists