From: Jesper Dangaard Brouer <brouer@redhat•com>
To: Ilias Apalodimas <ilias.apalodimas@linaro•org>
Cc: brouer@redhat•com, Yunsheng Lin <linyunsheng@huawei•com>,
Matteo Croce <mcroce@linux•microsoft.com>,
Networking <netdev@vger•kernel.org>,
Linux-MM <linux-mm@kvack•org>,
Ayush Sawal <ayush.sawal@chelsio•com>,
Vinay Kumar Yadav <vinay.yadav@chelsio•com>,
Rohit Maheshwari <rohitm@chelsio•com>,
"David S. Miller" <davem@davemloft•net>,
Jakub Kicinski <kuba@kernel•org>,
Thomas Petazzoni <thomas.petazzoni@bootlin•com>,
Marcin Wojtas <mw@semihalf•com>,
Russell King <linux@armlinux•org.uk>,
Mirko Lindner <mlindner@marvell•com>,
Stephen Hemminger <stephen@networkplumber•org>,
Tariq Toukan <tariqt@nvidia•com>,
Jesper Dangaard Brouer <hawk@kernel•org>,
Alexei Starovoitov <ast@kernel•org>,
Daniel Borkmann <daniel@iogearbox•net>,
John Fastabend <john.fastabend@gmail•com>,
Boris Pismenny <borisp@nvidia•com>, Arnd Bergmann <arnd@arndb•de>,
Andrew Morton <akpm@linux-foundation•org>,
"Peter Zijlstra (Intel)" <peterz@infradead•org>,
Vlastimil Babka <vbabka@suse•cz>, Yu Zhao <yuzhao@google•com>,
Will Deacon <will@kernel•org>, Fenghua Yu <fenghua.yu@intel•com>,
Roman Gushchin <guro@fb•com>, Hugh Dickins <hughd@google•com>,
Peter Xu <peterx@redhat•com>, Jason Gunthorpe <jgg@ziepe•ca>,
Jonathan Lemon <jonathan.lemon@gmail•com>,
Alexander Lobakin <alobakin@pm•me>,
Cong Wang <cong.wang@bytedance•com>, wenxu <wenxu@ucloud•cn>,
Kevin Hao <haokexin@gmail•com>,
Jakub Sitnicki <jakub@cloudflare•com>,
Marco Elver <elver@google•com>,
Willem de Bruijn <willemb@google•com>,
Miaohe Lin <linmiaohe@huawei•com>,
Guillaume Nault <gnault@redhat•com>,
open list <linux-kernel@vger•kernel.org>,
linux-rdma@vger•kernel.org, bpf <bpf@vger•kernel.org>,
Matthew Wilcox <willy@infradead•org>,
Eric Dumazet <edumazet@google•com>,
David Ahern <dsahern@gmail•com>,
Lorenzo Bianconi <lorenzo@kernel•org>,
Saeed Mahameed <saeedm@nvidia•com>, Andrew Lunn <andrew@lunn•ch>,
Paolo Abeni <pabeni@redhat•com>
Subject: Re: [PATCH net-next v3 0/5] page_pool: recycle buffers
Date: Mon, 3 May 2021 09:29:37 +0200 [thread overview]
Message-ID: <20210503092937.78a1eb05@carbon> (raw)
In-Reply-To: <CAC_iWj+wkjcGjwbVqEFXFyUi_zgn4-uYhQKKHKk84jkgo1sxRw@mail.gmail.com>
On Fri, 30 Apr 2021 20:32:07 +0300
Ilias Apalodimas <ilias.apalodimas@linaro•org> wrote:
> (-cc invalid emails)
> Replying to my self here but....
>
> [...]
> > > >
> > > > We can't do that. The reason we need those structs is that we rely on the
> > > > existing XDP code, which already recycles it's buffers, to enable
> > > > recycling. Since we allocate a page per packet when using page_pool for a
> > > > driver , the same ideas apply to an SKB and XDP frame. We just recycle the
> > >
> > > I am not really familar with XDP here, but a packet from hw is either a
> > > "struct xdp_frame/xdp_buff" for XDP or a "struct sk_buff" for TCP/IP stack,
> > > a packet can not be both "struct xdp_frame/xdp_buff" and "struct sk_buff" at
> > > the same time, right?
> > >
> >
> > Yes, but the payload is irrelevant in both cases and that's what we use
> > page_pool for. You can't use this patchset unless your driver usues
> > build_skb(). So in both cases you just allocate memory for the payload and
> > decide what the wrap the buffer with (XDP or SKB) later.
> >
> > > What does not really make sense to me is that the page has to be from page
> > > pool when a skb's frag page can be recycled, right? If it is ture, the switch
> > > case in __xdp_return() does not really make sense for skb recycling, why go
> > > all the trouble of checking the mem->type and mem->id to find the page_pool
> > > pointer when recyclable page for skb can only be from page pool?
> >
> > In any case you need to find in which pool the buffer you try to recycle
> > belongs. In order to make the whole idea generic and be able to recycle skb
> > fragments instead of just the skb head you need to store some information on
> > struct page. That's the fundamental difference of this patchset compared to
> > the RFC we sent a few years back [1] which was just storing information on the
> > skb. The way this is done on the current patchset is that we store the
> > struct xdp_mem_info in page->private and then look it up on xdp_return().
> >
> > Now that being said Matthew recently reworked struct page, so we could see if
> > we can store the page pool pointer directly instead of the struct
> > xdp_mem_info. That would allow us to call into page pool functions directly.
> > But we'll have to agree if that makes sense to go into struct page to begin
> > with and make sure the pointer is still valid when we take the recycling path.
> >
>
> Thinking more about it the reason that prevented us from storing a
> page pool pointer directly is not there anymore. Jesper fixed that
> already a while back. So we might as well store the page_pool ptr in
> page->private and call into the functions directly. I'll have a look
> before v4.
I want to give credit to Jonathan Lemon whom came up with the idea of
storing the page_pool object that "owns" the page directly in struct
page. I see this as an optimization that we can add later, so it
doesn't block this patchset. As Ilias mention, it required some
work/changes[1]+[2] to guarantee that the page_pool object life-time
were longer than all the outstanding in-flight page-objects, but that
have been stable for some/many kernel releases now. This is already
need/used for making sure the DMA-mappings can be safely released[1],
but I on-purpose enabled the same in-flight tracking for page_pool
users that doesn't use the DMA-mapping feature (making sure the code is
exercised).
[1] 99c07c43c4ea ("xdp: tracking page_pool resources and safe removal")
[2] c3f812cea0d7 ("page_pool: do not release pool until inflight == 0.")
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
next prev parent reply other threads:[~2021-05-03 7:30 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-09 22:37 [PATCH net-next v3 0/5] page_pool: recycle buffers Matteo Croce
2021-04-09 22:37 ` [PATCH net-next v3 1/5] xdp: reduce size of struct xdp_mem_info Matteo Croce
2021-04-09 22:37 ` [PATCH net-next v3 2/5] mm: add a signature in struct page Matteo Croce
2021-04-10 15:48 ` Matthew Wilcox
2021-04-10 16:16 ` Ilias Apalodimas
2021-04-10 17:42 ` Shakeel Butt
2021-04-10 18:27 ` Ilias Apalodimas
2021-04-10 19:39 ` Matthew Wilcox
2021-04-11 10:05 ` Jesper Dangaard Brouer
2021-04-14 19:41 ` Jesper Dangaard Brouer
2021-04-14 20:09 ` Shakeel Butt
2021-04-14 20:51 ` Eric Dumazet
2021-04-19 5:12 ` Ilias Apalodimas
2021-04-19 14:57 ` Shakeel Butt
2021-04-19 15:43 ` Ilias Apalodimas
2021-04-19 16:21 ` Shakeel Butt
2021-04-19 18:41 ` Ilias Apalodimas
2021-04-19 11:22 ` Jesper Dangaard Brouer
2021-04-19 13:01 ` Matthew Wilcox
2021-04-20 8:10 ` Ilias Apalodimas
2021-04-09 22:37 ` [PATCH net-next v3 3/5] page_pool: Allow drivers to hint on SKB recycling Matteo Croce
2021-04-10 0:11 ` Ilias Apalodimas
2021-04-10 0:39 ` Matteo Croce
2021-04-09 22:38 ` [PATCH net-next v3 4/5] mvpp2: recycle buffers Matteo Croce
2021-04-09 22:38 ` [PATCH net-next v3 5/5] mvneta: " Matteo Croce
2021-04-29 8:27 ` [PATCH net-next v3 0/5] page_pool: " Yunsheng Lin
2021-04-29 18:51 ` Ilias Apalodimas
2021-04-30 3:01 ` Yunsheng Lin
2021-04-30 16:24 ` Ilias Apalodimas
2021-04-30 17:32 ` Ilias Apalodimas
2021-05-03 7:29 ` Jesper Dangaard Brouer [this message]
2021-05-06 12:34 ` Yunsheng Lin
2021-05-06 12:58 ` Ilias Apalodimas
2021-05-07 3:23 ` Yunsheng Lin
2021-05-07 7:06 ` Ilias Apalodimas
2021-05-07 8:28 ` Yunsheng Lin
2021-05-07 10:19 ` Jesper Dangaard Brouer
2021-05-07 11:31 ` Christoph Hellwig
2021-05-09 5:11 ` Shay Agroskin
2021-05-11 8:41 ` Ilias Apalodimas
2021-05-10 2:20 ` Yunsheng Lin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210503092937.78a1eb05@carbon \
--to=brouer@redhat$(echo .)com \
--cc=akpm@linux-foundation$(echo .)org \
--cc=alobakin@pm$(echo .)me \
--cc=andrew@lunn$(echo .)ch \
--cc=arnd@arndb$(echo .)de \
--cc=ast@kernel$(echo .)org \
--cc=ayush.sawal@chelsio$(echo .)com \
--cc=borisp@nvidia$(echo .)com \
--cc=bpf@vger$(echo .)kernel.org \
--cc=cong.wang@bytedance$(echo .)com \
--cc=daniel@iogearbox$(echo .)net \
--cc=davem@davemloft$(echo .)net \
--cc=dsahern@gmail$(echo .)com \
--cc=edumazet@google$(echo .)com \
--cc=elver@google$(echo .)com \
--cc=fenghua.yu@intel$(echo .)com \
--cc=gnault@redhat$(echo .)com \
--cc=guro@fb$(echo .)com \
--cc=haokexin@gmail$(echo .)com \
--cc=hawk@kernel$(echo .)org \
--cc=hughd@google$(echo .)com \
--cc=ilias.apalodimas@linaro$(echo .)org \
--cc=jakub@cloudflare$(echo .)com \
--cc=jgg@ziepe$(echo .)ca \
--cc=john.fastabend@gmail$(echo .)com \
--cc=jonathan.lemon@gmail$(echo .)com \
--cc=kuba@kernel$(echo .)org \
--cc=linmiaohe@huawei$(echo .)com \
--cc=linux-kernel@vger$(echo .)kernel.org \
--cc=linux-mm@kvack$(echo .)org \
--cc=linux-rdma@vger$(echo .)kernel.org \
--cc=linux@armlinux$(echo .)org.uk \
--cc=linyunsheng@huawei$(echo .)com \
--cc=lorenzo@kernel$(echo .)org \
--cc=mcroce@linux$(echo .)microsoft.com \
--cc=mlindner@marvell$(echo .)com \
--cc=mw@semihalf$(echo .)com \
--cc=netdev@vger$(echo .)kernel.org \
--cc=pabeni@redhat$(echo .)com \
--cc=peterx@redhat$(echo .)com \
--cc=peterz@infradead$(echo .)org \
--cc=rohitm@chelsio$(echo .)com \
--cc=saeedm@nvidia$(echo .)com \
--cc=stephen@networkplumber$(echo .)org \
--cc=tariqt@nvidia$(echo .)com \
--cc=thomas.petazzoni@bootlin$(echo .)com \
--cc=vbabka@suse$(echo .)cz \
--cc=vinay.yadav@chelsio$(echo .)com \
--cc=wenxu@ucloud$(echo .)cn \
--cc=will@kernel$(echo .)org \
--cc=willemb@google$(echo .)com \
--cc=willy@infradead$(echo .)org \
--cc=yuzhao@google$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox