Date:   Mon, 23 Aug 2021 17:25:01 +0800
From:   Yunsheng Lin <linyunsheng@...wei.com>
To:     Eric Dumazet <edumazet@...gle.com>
CC:     David Miller <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Alexander Duyck <alexander.duyck@...il.com>,
        Russell King <linux@...linux.org.uk>,
        Marcin Wojtas <mw@...ihalf.com>, <linuxarm@...neuler.org>,
        Yisen Zhuang <yisen.zhuang@...wei.com>,
        Salil Mehta <salil.mehta@...wei.com>,
        Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
        Jesper Dangaard Brouer <hawk@...nel.org>,
        Ilias Apalodimas <ilias.apalodimas@...aro.org>,
        Alexei Starovoitov <ast@...nel.org>,
        "Daniel Borkmann" <daniel@...earbox.net>,
        John Fastabend <john.fastabend@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Will Deacon <will@...nel.org>,
        Matthew Wilcox <willy@...radead.org>,
        Vlastimil Babka <vbabka@...e.cz>,
        Fenghua Yu <fenghua.yu@...el.com>,
        Roman Gushchin <guro@...com>, Peter Xu <peterx@...hat.com>,
        "Tang, Feng" <feng.tang@...el.com>, Jason Gunthorpe <jgg@...pe.ca>,
        <mcroce@...rosoft.com>, Hugh Dickins <hughd@...gle.com>,
        Jonathan Lemon <jonathan.lemon@...il.com>,
        Alexander Lobakin <alobakin@...me>,
        Willem de Bruijn <willemb@...gle.com>,
        wenxu <wenxu@...oud.cn>, Cong Wang <cong.wang@...edance.com>,
        Kevin Hao <haokexin@...il.com>,
        Aleksandr Nogikh <nogikh@...gle.com>,
        Marco Elver <elver@...gle.com>, Yonghong Song <yhs@...com>,
        <kpsingh@...nel.org>, "Andrii Nakryiko" <andrii@...nel.org>,
        Martin KaFai Lau <kafai@...com>,
        Song Liu <songliubraving@...com>,
        netdev <netdev@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>, bpf <bpf@...r.kernel.org>,
        <chenhao288@...ilicon.com>,
        Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
        David Ahern <dsahern@...nel.org>, <memxor@...il.com>,
        <linux@...pel-privat.de>, Antoine Tenart <atenart@...nel.org>,
        Wei Wang <weiwan@...gle.com>, Taehee Yoo <ap420073@...il.com>,
        Arnd Bergmann <arnd@...db.de>,
        Mat Martineau <mathew.j.martineau@...ux.intel.com>,
        <aahringo@...hat.com>, <ceggers@...i.de>, <yangbo.lu@....com>,
        "Florian Westphal" <fw@...len.de>, <xiangxia.m.yue@...il.com>,
        linmiaohe <linmiaohe@...wei.com>, <hch@....de>
Subject: Re: [Linuxarm] Re: [PATCH RFC 0/7] add socket to netdev page frag
 recycling support

On 2021/8/18 17:36, Yunsheng Lin wrote:
> On 2021/8/18 16:57, Eric Dumazet wrote:
>> On Wed, Aug 18, 2021 at 5:33 AM Yunsheng Lin <linyunsheng@...wei.com> wrote:
>>>
>>> This patchset adds the socket to netdev page frag recycling
>>> support based on the busy polling and page pool infrastructure.
>>
>> I really do not see how this can scale to thousands of sockets.
>>
>> tcp_mem[] defaults to ~ 9 % of physical memory.
>>
>> If you now run tests with thousands of sockets, their skbs will
>> consume Gigabytes
>> of memory on typical servers, now backed by order-0 pages (instead of
>> current order-3 pages)
>> So IOMMU costs will actually be much bigger.
> 
> As the page allocator support bulk allocating now, see:
> https://elixir.bootlin.com/linux/latest/source/net/core/page_pool.c#L252
> 
> if the DMA also support batch mapping/unmapping, maybe having a
> small-sized page pool for thousands of sockets may not be a problem?
> Christoph Hellwig mentioned the batch DMA operation support in below
> thread:
> https://www.spinics.net/lists/netdev/msg666715.html
> 
> if the batched DMA operation is supported, maybe having the
> page pool is mainly benefit the case of small number of socket?
> 
>>
>> Are we planning to use Gigabyte sized page pools for NIC ?
>>
>> Have you tried instead to make TCP frags twice bigger ?
> 
> Not yet.
> 
>> This would require less IOMMU mappings.
>> (Note: This could require some mm help, since PAGE_ALLOC_COSTLY_ORDER
>> is currently 3, not 4)
> 
> I am not familiar with mm yet, but I will take a look about that:)


It seems PAGE_ALLOC_COSTLY_ORDER is mostly related to pcp page, OOM, memory
compact and memory isolation, as the test system has a lot of memory installed
(about 500G, only 3-4G is used), so I used the below patch to test the max
possible performance improvement when making TCP frags twice bigger, and
the performance improvement went from about 30Gbit to 32Gbit for one thread
iperf tcp flow in IOMMU strict mode, and using the pfrag pool, the improvement
went from about 30Gbit to 40Gbit for the same testing configuation:

diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index fcb5355..dda20f9 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -37,7 +37,7 @@
  * coalesce naturally under reasonable reclaim pressure and those which
  * will not.
  */
-#define PAGE_ALLOC_COSTLY_ORDER 3
+#define PAGE_ALLOC_COSTLY_ORDER 4

 enum migratetype {
        MIGRATE_UNMOVABLE,
diff --git a/net/core/sock.c b/net/core/sock.c
index 870a3b7..b1e0dfc 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2580,7 +2580,7 @@ static void sk_leave_memory_pressure(struct sock *sk)
        }
 }

-#define SKB_FRAG_PAGE_ORDER    get_order(32768)
+#define SKB_FRAG_PAGE_ORDER    get_order(65536)
 DEFINE_STATIC_KEY_FALSE(net_high_order_alloc_disable_key);

 /**

> 
>>
>> diff --git a/net/core/sock.c b/net/core/sock.c
>> index a3eea6e0b30a7d43793f567ffa526092c03e3546..6b66b51b61be9f198f6f1c4a3d81b57fa327986a
>> 100644
>> --- a/net/core/sock.c
>> +++ b/net/core/sock.c
>> @@ -2560,7 +2560,7 @@ static void sk_leave_memory_pressure(struct sock *sk)
>>         }
>>  }
>>
>> -#define SKB_FRAG_PAGE_ORDER    get_order(32768)
>> +#define SKB_FRAG_PAGE_ORDER    get_order(65536)
>>  DEFINE_STATIC_KEY_FALSE(net_high_order_alloc_disable_key);
>>
>>  /**
>>
>>
>>
>>>
> _______________________________________________
> Linuxarm mailing list -- linuxarm@...neuler.org
> To unsubscribe send an email to linuxarm-leave@...neuler.org
> 

Powered by blists - more mailing lists