linux-kernel - Re: [PATCH 3/3] net/ceph/messenger: add empty check to ceph_con_get_out

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAOi1vP_m5ovLLxpzyexq0vhVV8JPXAYcbzUqrQmn7jZkdhfmNA@mail.gmail.com>
Date: Thu, 9 Oct 2025 13:18:15 +0200
From: Ilya Dryomov <idryomov@...il.com>
To: Max Kellermann <max.kellermann@...os.com>
Cc: xiubli@...hat.com, amarkuze@...hat.com, ceph-devel@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/3] net/ceph/messenger: add empty check to ceph_con_get_out_msg()

On Wed, Aug 6, 2025 at 11:49 AM Max Kellermann <max.kellermann@...os.com> wrote:
>
> This moves the list_empty() checks from the two callers (v1 and v2)
> into the base messenger.c library.  Now the v1/v2 specializations do
> not need to know about con->out_queue; that implementation detail is
> now hidden behind the ceph_con_get_out_msg() function.
>
> Signed-off-by: Max Kellermann <max.kellermann@...os.com>
> ---
>  net/ceph/messenger.c    |  4 +++-
>  net/ceph/messenger_v1.c | 15 ++++++++++-----
>  net/ceph/messenger_v2.c |  4 ++--
>  3 files changed, 15 insertions(+), 8 deletions(-)
>
> diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c
> index 424fb2769b71..8886c38a55d2 100644
> --- a/net/ceph/messenger.c
> +++ b/net/ceph/messenger.c
> @@ -2113,7 +2113,9 @@ struct ceph_msg *ceph_con_get_out_msg(struct ceph_connection *con)
>  {
>         struct ceph_msg *msg;
>
> -       BUG_ON(list_empty(&con->out_queue));
> +       if (list_empty(&con->out_queue))
> +               return NULL;
> +
>         msg = list_first_entry(&con->out_queue, struct ceph_msg, list_head);
>         WARN_ON(msg->con != con);
>
> diff --git a/net/ceph/messenger_v1.c b/net/ceph/messenger_v1.c
> index 516f2eeb122a..5eb6cfdbc494 100644
> --- a/net/ceph/messenger_v1.c
> +++ b/net/ceph/messenger_v1.c
> @@ -189,12 +189,18 @@ static void prepare_write_message_footer(struct ceph_connection *con, struct cep
>
>  /*
>   * Prepare headers for the next outgoing message.
> + *
> + * @return false if there are no outgoing messages
>   */
> -static void prepare_write_message(struct ceph_connection *con)
> +static bool prepare_write_message(struct ceph_connection *con)
>  {
>         struct ceph_msg *m;
>         u32 crc;
>
> +       m = ceph_con_get_out_msg(con);
> +       if (m == NULL)
> +               return false;
> +
>         con_out_kvec_reset(con);
>         con->v1.out_msg_done = false;
>
> @@ -208,8 +214,6 @@ static void prepare_write_message(struct ceph_connection *con)
>                         &con->v1.out_temp_ack);
>         }
>
> -       m = ceph_con_get_out_msg(con);
> -
>         dout("prepare_write_message %p seq %lld type %d len %d+%d+%zd\n",
>              m, con->out_seq, le16_to_cpu(m->hdr.type),
>              le32_to_cpu(m->hdr.front_len), le32_to_cpu(m->hdr.middle_len),
> @@ -256,6 +260,8 @@ static void prepare_write_message(struct ceph_connection *con)
>         }
>
>         ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
> +
> +       return true;
>  }
>
>  /*
> @@ -1543,8 +1549,7 @@ int ceph_con_v1_try_write(struct ceph_connection *con)
>                         goto more;
>                 }
>                 /* is anything else pending? */
> -               if (!list_empty(&con->out_queue)) {
> -                       prepare_write_message(con);
> +               if (prepare_write_message(con)) {

Hi Max,

I made a change to net/ceph/messenger_v1.c hunks of this patch to
follow what is done for msgr2 where ceph_con_get_out_msg() is called
outside of the prepare helper and the new message is passed in.
prepare_write_message() doesn't need to return a bool anymore.

Let me know if you see something wrong there:

https://github.com/ceph/ceph-client/commit/6140f1d43ba9425dc55b12bdfd8877b0c5118d9a

Thanks,

                Ilya