Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210820030536.25737-1-yaozhenguo1@gmail.com> <20210822151952.23ca9547316dc34c9f3bd482@linux-foundation.org>
 <YSLPWybBCyE/6x7s@casper.infradead.org>
In-Reply-To: <YSLPWybBCyE/6x7s@casper.infradead.org>
From:   zhenguo yao <yaozhenguo1@gmail.com>
Date:   Mon, 23 Aug 2021 09:57:53 +0800
Message-ID: <CA+WzARmb+KVLo=isfU7T2bi=9O=UFFG57CBk6w-+haKp_OqbMg@mail.gmail.com>
Subject: Re: [PATCH] hugetlbfs: add hugepages_node kernel parameter
To:     Matthew Wilcox <willy@infradead.org>
Cc:     mike.kravetz@oracle.com, corbet@lwn.net, akpm@linux-foundation.org,
        linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
        linux-mm@kvack.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Yes, the expanding of hugepages is more elegant. I  will change it in
the next version.

Matthew Wilcox <willy@infradead.org> =E4=BA=8E2021=E5=B9=B48=E6=9C=8823=E6=
=97=A5=E5=91=A8=E4=B8=80 =E4=B8=8A=E5=8D=886:28=E5=86=99=E9=81=93=EF=BC=9A
>
> On Sun, Aug 22, 2021 at 03:19:52PM -0700, Andrew Morton wrote:
> > On Fri, 20 Aug 2021 11:05:36 +0800 yaozhenguo <yaozhenguo1@gmail.com> w=
rote:
> >
> > > We can specify the number of hugepages to allocate at boot. But the
> > > hugepages is balanced in all nodes at present. In some scenarios,
> > > we only need hugepags in one node. For example: DPDK needs hugepages
> > > which is in the same node as NIC. if DPDK needs four hugepags of 1G
> > > size in node1 and system has 16 numa nodes. We must reserve 64 hugepa=
gs
> > > in kernel cmdline. But, only four hugepages is used. The others shoul=
d
> > > be free after boot.If the system memory is low(for example: 64G), it =
will
> > > be an impossible task. So, add hugepages_node kernel parameter to spe=
cify
> > > node number of hugepages to allocate at boot.
> > > For example add following parameter:
> > >
> > > hugepagesz=3D1G hugepages_node=3D1 hugepages=3D4
> > >
> > > It will allocate 4 hugepags in node1 at boot.
> >
> > If were going to do this, shouldn't we permit more than one node?
> >
> >       hugepages_nodes=3D1,2,5
>
> I'd think we'd be better off expanding the definition of hugepages.
> eg:
>
> hugepagesz=3D1G hugepages=3D1:4,3:8,5:2
>
> would say to allocate 4 pages from node 1, 8 pages from node 3 and 2
> pages from node 5.