From: Byungchul Park <byungchul@sk.com>
To: Mina Almasry <almasrymina@google.com>
Cc: willy@infradead.org, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
kernel_team@skhynix.com, kuba@kernel.org,
ilias.apalodimas@linaro.org, harry.yoo@oracle.com,
hawk@kernel.org, akpm@linux-foundation.org, ast@kernel.org,
daniel@iogearbox.net, davem@davemloft.net,
john.fastabend@gmail.com, andrew+netdev@lunn.ch,
edumazet@google.com, pabeni@redhat.com, vishal.moola@gmail.com
Subject: Re: [RFC 02/19] netmem: introduce netmem alloc/put API to wrap page alloc/put API
Date: Mon, 12 May 2025 21:30:51 +0900 [thread overview]
Message-ID: <20250512123051.GA45370@system.software.com> (raw)
In-Reply-To: <CAHS8izP3knY42632AcfTHcpgpSz49gP0j6CnyswUoHW6JtC3=w@mail.gmail.com>
On Fri, May 09, 2025 at 07:08:23AM -0700, Mina Almasry wrote:
> j
>
> On Fri, May 9, 2025 at 6:39 AM Mina Almasry <almasrymina@google.com> wrote:
> >
> > On Fri, May 9, 2025 at 4:51 AM Byungchul Park <byungchul@sk.com> wrote:
> > >
> > > To eliminate the use of struct page in page pool, the page pool code
> > > should use netmem descriptor and API instead.
> > >
> > > As part of the work, introduce netmem alloc/put API allowing the code to
> > > use them rather than struct page things.
> > >
> > > Signed-off-by: Byungchul Park <byungchul@sk.com>
> > > ---
> > > include/net/netmem.h | 18 ++++++++++++++++++
> > > 1 file changed, 18 insertions(+)
> > >
> > > diff --git a/include/net/netmem.h b/include/net/netmem.h
> > > index 45c209d6cc689..c87a604e980b9 100644
> > > --- a/include/net/netmem.h
> > > +++ b/include/net/netmem.h
> > > @@ -138,6 +138,24 @@ static inline netmem_ref page_to_netmem(struct page *page)
> > > return (__force netmem_ref)page;
> > > }
> > >
> > > +static inline netmem_ref alloc_netmems_node(int nid, gfp_t gfp_mask,
> > > + unsigned int order)
> > > +{
> > > + return page_to_netmem(alloc_pages_node(nid, gfp_mask, order));
> > > +}
> > > +
> > > +static inline unsigned long alloc_netmems_bulk_node(gfp_t gfp, int nid,
> > > + unsigned long nr_netmems, netmem_ref *netmem_array)
> > > +{
> > > + return alloc_pages_bulk_node(gfp, nid, nr_netmems,
> > > + (struct page **)netmem_array);
> > > +}
> > > +
> > > +static inline void put_netmem(netmem_ref netmem)
> > > +{
> > > + put_page(netmem_to_page(netmem));
> > > +}
> >
> > We can't really do this. netmem_ref is an abstraction that can be a
> > struct page or struct net_iov underneath. We can't be sure when
> > put_netmem is called that it is safe to convert to a page via
> > netmem_to_page(). This will crash if put_netmem is called on a
> > netmem_ref that is a net_iov underneath.
> >
>
> On a closer look, it looks like this put_netmem is only called from
> code paths where you are sure the netmem_ref is a page underneath, so
> this is likely fine for now. There is a net_next series that is adding
> proper put_netmem support [1]. It's probably best to rebase your work
> on top of that, but this should be fine in the meantime.
Indeed. Hm.. It'd be the best to work on the top of yours.
If it takes too long, I keep working on as it is for now and will adjust
this patch later once your work gets merged.
Byungchul
> [1] https://lore.kernel.org/netdev/20250508004830.4100853-1-almasrymina@google.com/
>
> --
> Thanks,
> Mina
next prev parent reply other threads:[~2025-05-12 12:30 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-05-09 11:51 [RFC 00/19] Split netmem from struct page Byungchul Park
2025-05-09 11:51 ` [RFC 01/19] netmem: rename struct net_iov to struct netmem_desc Byungchul Park
2025-05-12 13:11 ` Pavel Begunkov
2025-05-12 13:29 ` Byungchul Park
2025-05-12 19:14 ` Mina Almasry
2025-05-13 2:00 ` Byungchul Park
2025-05-13 12:58 ` Pavel Begunkov
2025-05-13 12:49 ` Pavel Begunkov
2025-05-14 0:07 ` Byungchul Park
2025-05-09 11:51 ` [RFC 02/19] netmem: introduce netmem alloc/put API to wrap page alloc/put API Byungchul Park
2025-05-09 13:39 ` Mina Almasry
2025-05-09 14:08 ` Mina Almasry
2025-05-12 12:30 ` Byungchul Park [this message]
2025-05-09 11:51 ` [RFC 03/19] page_pool: use netmem alloc/put API in __page_pool_alloc_page_order() Byungchul Park
2025-05-09 11:51 ` [RFC 04/19] page_pool: rename __page_pool_alloc_page_order() to __page_pool_alloc_large_netmem() Byungchul Park
2025-05-09 11:51 ` [RFC 05/19] page_pool: use netmem alloc/put API in __page_pool_alloc_pages_slow() Byungchul Park
2025-05-09 11:51 ` [RFC 06/19] page_pool: rename page_pool_return_page() to page_pool_return_netmem() Byungchul Park
2025-05-09 11:51 ` [RFC 07/19] page_pool: use netmem alloc/put API in page_pool_return_netmem() Byungchul Park
2025-05-09 11:51 ` [RFC 08/19] page_pool: rename __page_pool_release_page_dma() to __page_pool_release_netmem_dma() Byungchul Park
2025-05-09 11:51 ` [RFC 09/19] page_pool: rename __page_pool_put_page() to __page_pool_put_netmem() Byungchul Park
2025-05-09 11:51 ` [RFC 10/19] page_pool: rename __page_pool_alloc_pages_slow() to __page_pool_alloc_netmems_slow() Byungchul Park
2025-05-09 11:51 ` [RFC 11/19] mlx4: use netmem descriptor and API for page pool Byungchul Park
2025-05-09 11:51 ` [RFC 12/19] netmem: introduce page_pool_recycle_direct_netmem() Byungchul Park
2025-05-09 11:51 ` [RFC 13/19] page_pool: expand scope of is_pp_{netmem,page}() to global Byungchul Park
2025-05-12 12:46 ` Toke Høiland-Jørgensen
2025-05-12 12:55 ` Byungchul Park
2025-05-14 3:00 ` Byungchul Park
2025-05-14 11:17 ` Toke Høiland-Jørgensen
2025-05-09 11:51 ` [RFC 14/19] mm: page_alloc: do not directly access page->pp_magic but use is_pp_page() Byungchul Park
2025-05-09 11:51 ` [RFC 15/19] mlx5: use netmem descriptor and API for page pool Byungchul Park
2025-05-09 11:51 ` [RFC 16/19] netmem: use _Generic to cover const casting for page_to_netmem() Byungchul Park
2025-05-09 11:51 ` [RFC 17/19] netmem: remove __netmem_get_pp() Byungchul Park
2025-05-09 13:47 ` Mina Almasry
2025-05-09 11:51 ` [RFC 18/19] page_pool: make page_pool_get_dma_addr() just wrap page_pool_get_dma_addr_netmem() Byungchul Park
2025-05-09 13:49 ` Mina Almasry
2025-05-10 7:28 ` Ilias Apalodimas
2025-05-09 11:51 ` [RFC 19/19] mm, netmem: remove the page pool members in struct page Byungchul Park
2025-05-09 17:32 ` Mina Almasry
2025-05-09 18:11 ` Matthew Wilcox
2025-05-09 19:04 ` Mina Almasry
2025-05-09 19:48 ` Matthew Wilcox
2025-05-12 19:10 ` Mina Almasry
2025-05-09 18:02 ` Matthew Wilcox
2025-05-12 12:51 ` Byungchul Park
2025-05-12 14:42 ` Matthew Wilcox
2025-05-13 1:42 ` Byungchul Park
2025-05-13 3:19 ` Matthew Wilcox
2025-05-13 10:24 ` Byungchul Park
2025-05-10 7:26 ` Ilias Apalodimas
2025-05-12 12:58 ` Byungchul Park
2025-05-09 14:09 ` [RFC 00/19] Split netmem from " Mina Almasry
2025-05-12 12:36 ` Byungchul Park
2025-05-12 12:59 ` Pavel Begunkov
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=20250512123051.GA45370@system.software.com \
--to=byungchul@sk.com \
--cc=akpm@linux-foundation.org \
--cc=almasrymina@google.com \
--cc=andrew+netdev@lunn.ch \
--cc=ast@kernel.org \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=harry.yoo@oracle.com \
--cc=hawk@kernel.org \
--cc=ilias.apalodimas@linaro.org \
--cc=john.fastabend@gmail.com \
--cc=kernel_team@skhynix.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=vishal.moola@gmail.com \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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;
as well as URLs for NNTP newsgroup(s).