Received-SPF: pass (google.com: domain of linux-kernel+bounces-52597-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Precedence: bulk
MIME-Version: 1.0
References: <20240205072013.427639-1-chentao@kylinos.cn> <20240205072013.427639-6-chentao@kylinos.cn>
In-Reply-To: <20240205072013.427639-6-chentao@kylinos.cn>
From: Eric Dumazet <edumazet@google.com>
Date: Mon, 5 Feb 2024 13:29:18 +0100
Message-ID: <CANn89iLkWvum6wSqSya_K+1eqnFvp=L2WLW=kAYrZTF8Ei4b7g@mail.gmail.com>
Subject: Re: [PATCH net-next 5/6] tcp: Simplify the allocation of slab caches
To: Kunwu Chan <chentao@kylinos.cn>
Cc: davem@davemloft.net, dsahern@kernel.org, kuba@kernel.org, 
	pabeni@redhat.com, jiri@resnulli.us, netdev@vger.kernel.org, 
	linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, Feb 5, 2024 at 8:23=E2=80=AFAM Kunwu Chan <chentao@kylinos.cn> wrot=
e:
>
> Use the new KMEM_CACHE() macro instead of direct kmem_cache_create
> to simplify the creation of SLAB caches.
> And change cache name from 'tcp_bind_bucket' to 'inet_bind_bucket',
> 'tcp_bind2_bucket' to 'inet_bind2_bucket'.
>
> Signed-off-by: Kunwu Chan <chentao@kylinos.cn>
> ---
>  net/ipv4/tcp.c | 14 ++++----------
>  1 file changed, 4 insertions(+), 10 deletions(-)
>
> diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
> index a1c6de385cce..2dc3dd4323c2 100644
> --- a/net/ipv4/tcp.c
> +++ b/net/ipv4/tcp.c
> @@ -4697,17 +4697,11 @@ void __init tcp_init(void)
>                             thash_entries, 21,  /* one slot per 2 MB*/
>                             0, 64 * 1024);
>         tcp_hashinfo.bind_bucket_cachep =3D
> -               kmem_cache_create("tcp_bind_bucket",
> -                                 sizeof(struct inet_bind_bucket), 0,
> -                                 SLAB_HWCACHE_ALIGN | SLAB_PANIC |
> -                                 SLAB_ACCOUNT,
> -                                 NULL);
> +               KMEM_CACHE(inet_bind_bucket,
> +                          SLAB_HWCACHE_ALIGN | SLAB_PANIC | SLAB_ACCOUNT=
);

I would prefer we do not do this.

dccp is also using a kmem_cache_create() of "struct inet_bind_bucket"

We want different caches for TCP and DCCP.


>         tcp_hashinfo.bind2_bucket_cachep =3D
> -               kmem_cache_create("tcp_bind2_bucket",
> -                                 sizeof(struct inet_bind2_bucket), 0,
> -                                 SLAB_HWCACHE_ALIGN | SLAB_PANIC |
> -                                 SLAB_ACCOUNT,
> -                                 NULL);
> +               KMEM_CACHE(inet_bind2_bucket,
> +                          SLAB_HWCACHE_ALIGN | SLAB_PANIC | SLAB_ACCOUNT=
);

Same here.

>
>         /* Size and allocate the main established and bind bucket
>          * hash tables.
> --
> 2.39.2
>