linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Pavel Begunkov <asml.silence@gmail.com>
To: Mina Almasry <almasrymina@google.com>, Byungchul Park <byungchul@sk.com>
Cc: willy@infradead.org, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	kernel_team@skhynix.com, ilias.apalodimas@linaro.org,
	harry.yoo@oracle.com, akpm@linux-foundation.org,
	andrew+netdev@lunn.ch, toke@redhat.com, david@redhat.com,
	Liam.Howlett@oracle.com, vbabka@suse.cz, rppt@kernel.org,
	surenb@google.com, mhocko@suse.com, linux-rdma@vger.kernel.org,
	bpf@vger.kernel.org, vishal.moola@gmail.com, hannes@cmpxchg.org,
	ziy@nvidia.com, jackmanb@google.com, wei.fang@nxp.com,
	shenwei.wang@nxp.com, xiaoning.wang@nxp.com, davem@davemloft.net,
	edumazet@google.com, kuba@kernel.org, pabeni@redhat.com,
	anthony.l.nguyen@intel.com, przemyslaw.kitszel@intel.com,
	sgoutham@marvell.com, gakula@marvell.com, sbhatta@marvell.com,
	hkelam@marvell.com, bbhushan2@marvell.com, tariqt@nvidia.com,
	ast@kernel.org, daniel@iogearbox.net, hawk@kernel.org,
	john.fastabend@gmail.com, sdf@fomichev.me, saeedm@nvidia.com,
	leon@kernel.org, mbloch@nvidia.com, danishanwar@ti.com,
	rogerq@kernel.org, nbd@nbd.name, lorenzo@kernel.org,
	ryder.lee@mediatek.com, shayne.chen@mediatek.com,
	sean.wang@mediatek.com, matthias.bgg@gmail.com,
	angelogioacchino.delregno@collabora.com,
	aleksander.lobakin@intel.com, horms@kernel.org, m-malladi@ti.com,
	krzysztof.kozlowski@linaro.org,
	matthias.schiffer@ew.tq-group.com, robh@kernel.org,
	imx@lists.linux.dev, intel-wired-lan@lists.osuosl.org,
	linux-arm-kernel@lists.infradead.org,
	linux-wireless@vger.kernel.org,
	linux-mediatek@lists.infradead.org
Subject: Re: [PATCH net-next v10 02/12] netmem: use netmem_desc instead of page to access ->pp in __netmem_get_pp()
Date: Tue, 15 Jul 2025 11:37:45 +0100	[thread overview]
Message-ID: <9bed2f6e-6251-4d0c-ad1e-f1b8625a0a10@gmail.com> (raw)
In-Reply-To: <CAHS8izO393X_BDJxnX2d-auhTwrUZK5wYdoAh_tJc0GBf0AqcQ@mail.gmail.com>

On 7/14/25 20:37, Mina Almasry wrote:
> On Mon, Jul 14, 2025 at 5:01 AM Byungchul Park <byungchul@sk.com> wrote:
...>> +static inline struct netmem_desc *pp_page_to_nmdesc(struct page *page)
>> +{
>> +       DEBUG_NET_WARN_ON_ONCE(!page_pool_page_is_pp(page));
>> +
>> +       /* XXX: How to extract netmem_desc from page must be changed,
>> +        * once netmem_desc no longer overlays on page and will be
>> +        * allocated through slab.
>> +        */
>> +       return (struct netmem_desc *)page;
>> +}
>> +
> 
> Same thing. Do not create a generic looking pp_page_to_nmdesc helper
> which does not check that the page is the correct type. The
> DEBUG_NET... is not good enough.
> 
> You don't need to add a generic helper here. There is only one call
> site. Open code this in the callsite. The one callsite is marked as
> unsafe, only called by code that knows that the netmem is specifically
> a pp page. Open code this in the unsafe callsite, instead of creating
> a generic looking unsafe helper and not even documenting it's unsafe.
> 
>>   /**
>>    * __netmem_get_pp - unsafely get pointer to the &page_pool backing @netmem
>>    * @netmem: netmem reference to get the pointer from
>> @@ -280,7 +291,7 @@ static inline struct net_iov *__netmem_clear_lsb(netmem_ref netmem)
>>    */
>>   static inline struct page_pool *__netmem_get_pp(netmem_ref netmem)
>>   {
>> -       return __netmem_to_page(netmem)->pp;
>> +       return pp_page_to_nmdesc(__netmem_to_page(netmem))->pp;
>>   }
> 
> This makes me very sad. Casting from netmem -> page -> nmdesc...

The function is not used, and I don't think the series adds any
new users? It can be killed then. It's a horrible function anyway,
would be much better to have a variant taking struct page * if
necessary.

> Instead, we should be able to go from netmem directly to nmdesc. I
> would suggest rename __netmem_clear_lsb to netmem_to_nmdesc and have
> it return netmem_desc instead of net_iov. Then use it here.

Glad you liked the diff I suggested :) In either case, seems
like it's not strictly necessary for this iteration as
__netmem_get_pp() should be killed, and the rest of patches work
directly with pages.

  
> We could have an unsafe version of netmem_to_nmdesc which converts the
> netmem to netmem_desc without clearing the lsb and mark it unsafe.
> 

-- 
Pavel Begunkov



  parent reply	other threads:[~2025-07-15 10:36 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-14 12:00 [PATCH net-next v10 00/12] Split netmem from struct page Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 01/12] netmem: introduce struct netmem_desc mirroring " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 02/12] netmem: use netmem_desc instead of page to access ->pp in __netmem_get_pp() Byungchul Park
2025-07-14 19:37   ` Mina Almasry
2025-07-14 19:58     ` Mina Almasry
2025-07-15  1:36       ` Byungchul Park
2025-07-15 19:09         ` Mina Almasry
2025-07-16  4:51           ` Byungchul Park
2025-07-16 19:41             ` Mina Almasry
2025-07-17  0:54               ` Byungchul Park
2025-07-17  4:18               ` Byungchul Park
2025-07-17  6:32               ` Byungchul Park
2025-07-17  9:17               ` Pavel Begunkov
2025-07-17  9:23                 ` Pavel Begunkov
2025-07-15 10:37     ` Pavel Begunkov [this message]
2025-07-15 19:06       ` Mina Almasry
2025-07-16  4:27       ` Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 03/12] mlx4: access ->pp_ref_count through netmem_desc instead of page Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 04/12] netdevsim: access ->pp " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 05/12] mt76: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 06/12] net: fec: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 07/12] octeontx2-pf: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 08/12] iavf: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 09/12] idpf: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 10/12] mlx5: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 11/12] net: ti: icssg-prueth: " Byungchul Park
2025-07-14 12:00 ` [PATCH net-next v10 12/12] libeth: xdp: " Byungchul Park
2025-07-15  0:21   ` kernel test robot
2025-07-15  1:45   ` kernel test robot

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=9bed2f6e-6251-4d0c-ad1e-f1b8625a0a10@gmail.com \
    --to=asml.silence@gmail.com \
    --cc=Liam.Howlett@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=aleksander.lobakin@intel.com \
    --cc=almasrymina@google.com \
    --cc=andrew+netdev@lunn.ch \
    --cc=angelogioacchino.delregno@collabora.com \
    --cc=anthony.l.nguyen@intel.com \
    --cc=ast@kernel.org \
    --cc=bbhushan2@marvell.com \
    --cc=bpf@vger.kernel.org \
    --cc=byungchul@sk.com \
    --cc=daniel@iogearbox.net \
    --cc=danishanwar@ti.com \
    --cc=davem@davemloft.net \
    --cc=david@redhat.com \
    --cc=edumazet@google.com \
    --cc=gakula@marvell.com \
    --cc=hannes@cmpxchg.org \
    --cc=harry.yoo@oracle.com \
    --cc=hawk@kernel.org \
    --cc=hkelam@marvell.com \
    --cc=horms@kernel.org \
    --cc=ilias.apalodimas@linaro.org \
    --cc=imx@lists.linux.dev \
    --cc=intel-wired-lan@lists.osuosl.org \
    --cc=jackmanb@google.com \
    --cc=john.fastabend@gmail.com \
    --cc=kernel_team@skhynix.com \
    --cc=krzysztof.kozlowski@linaro.org \
    --cc=kuba@kernel.org \
    --cc=leon@kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mediatek@lists.infradead.org \
    --cc=linux-mm@kvack.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=linux-wireless@vger.kernel.org \
    --cc=lorenzo@kernel.org \
    --cc=m-malladi@ti.com \
    --cc=matthias.bgg@gmail.com \
    --cc=matthias.schiffer@ew.tq-group.com \
    --cc=mbloch@nvidia.com \
    --cc=mhocko@suse.com \
    --cc=nbd@nbd.name \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=przemyslaw.kitszel@intel.com \
    --cc=robh@kernel.org \
    --cc=rogerq@kernel.org \
    --cc=rppt@kernel.org \
    --cc=ryder.lee@mediatek.com \
    --cc=saeedm@nvidia.com \
    --cc=sbhatta@marvell.com \
    --cc=sdf@fomichev.me \
    --cc=sean.wang@mediatek.com \
    --cc=sgoutham@marvell.com \
    --cc=shayne.chen@mediatek.com \
    --cc=shenwei.wang@nxp.com \
    --cc=surenb@google.com \
    --cc=tariqt@nvidia.com \
    --cc=toke@redhat.com \
    --cc=vbabka@suse.cz \
    --cc=vishal.moola@gmail.com \
    --cc=wei.fang@nxp.com \
    --cc=willy@infradead.org \
    --cc=xiaoning.wang@nxp.com \
    --cc=ziy@nvidia.com \
    /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).