Re: [PATCH v3 00/26] Split netmem from struct page

From: Jesper Dangaard Brouer <jbrouer@redhat.com>
To: Matthew Wilcox <willy@infradead.org>,
	Yunsheng Lin <linyunsheng@huawei.com>
Cc: brouer@redhat.com, Jesper Dangaard Brouer <hawk@kernel.org>,
	Ilias Apalodimas <ilias.apalodimas@linaro.org>,
	netdev@vger.kernel.org, linux-mm@kvack.org,
	Shakeel Butt <shakeelb@google.com>
Subject: Re: [PATCH v3 00/26] Split netmem from struct page
Date: Thu, 12 Jan 2023 11:15:55 +0100	[thread overview]
Message-ID: <9cdc89f3-8c00-3673-5fdb-4f5bebd95d7a@redhat.com> (raw)
In-Reply-To: <Y763vcTFUZvWNgYv@casper.infradead.org>

On 11/01/2023 14.21, Matthew Wilcox wrote:
> On Wed, Jan 11, 2023 at 04:25:46PM +0800, Yunsheng Lin wrote:
>> On 2023/1/11 12:21, Matthew Wilcox (Oracle) wrote:
>>> The MM subsystem is trying to reduce struct page to a single pointer.
>>> The first step towards that is splitting struct page by its individual
>>> users, as has already been done with folio and slab.  This patchset does
>>> that for netmem which is used for page pools.
>> As page pool is only used for rx side in the net stack depending on the
>> driver, a lot more memory for the net stack is from page_frag_alloc_align(),
>> kmem cache, etc.
>> naming it netmem seems a little overkill, perhaps a more specific name for
>> the page pool? such as pp_cache.
>>
>> @Jesper & Ilias
>> Any better idea?

I like the 'netmem' name.

>> And it seem some API may need changing too, as we are not pooling 'pages'
>> now.

IMHO it would be overkill to rename the page_pool to e.g. netmem_pool.
as it would generate too much churn and will be hard to follow in git
as the code filename page_pool.c would also have to be renamed.
It guess we keep page_pool for historical reasons ;-)

> I raised the question of naming in v1, six weeks ago, and nobody had
> any better names.  Seems a little unfair to ignore the question at first
> and then bring it up now.  I'd hate to miss the merge window because of
> a late-breaking major request like this.
> 
> https://lore.kernel.org/netdev/20221130220803.3657490-1-willy@infradead.org/
> 
> I'd like to understand what we think we'll do in networking when we trim
> struct page down to a single pointer,  All these usages that aren't from
> page_pool -- what information does networking need to track per-allocation?
> Would it make sense for the netmem to describe all memory used by the
> networking stack, and have allocators other than page_pool also return
> netmem, 

This is also how I see the future, that other netstack "allocators" can
return and work-with 'netmem' objects.   IMHO we are already cramming
too many use-cases into page_pool (like the frag support Yunsheng
added).  IMHO there are room for other netstack "allocators" that can
utilize netmem.  The page_pool is optimized for RX-NAPI workloads, using
it for other purposes is a mistake IMHO.  People should create other
netstack "allocators" that solves their specific use-cases.  E.g. The TX
path likely needs another "allocator" optimized for this TX use-case.

> or does the normal usage of memory in the net stack not need to
> track that information?

The page refcnt is (obviously) used by netstack as tracked information.
I have seen drivers that use the DMA mapping directly in page/'netmem',
instead of having to store this separately in the drivers.

--Jesper