Linux-NVME Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Guixin Liu <kanie@linux.alibaba.com>
To: Christoph Hellwig <hch@lst.de>
Cc: Keith Busch <kbusch@kernel.org>, Jens Axboe <axboe@kernel.dk>,
	Sagi Grimberg <sagi@grimberg.me>,
	Nilay Shroff <nilay@linux.ibm.com>,
	Daniel Wagner <dwagner@suse.de>, Hannes Reinecke <hare@suse.de>,
	linux-nvme@lists.infradead.org
Subject: Re: [PATCH v2] nvmet: Handle a NULL admin queue in nvmet_queue_tls_keyid()
Date: Mon, 9 Nov 2026 10:35:35 +0800	[thread overview]
Message-ID: <ae3b49ac-5eb8-45a3-8f3d-9710873ecfbb@linux.alibaba.com> (raw)
In-Reply-To: <6d28ec32-b6af-425c-9cbc-ac1172baf979@linux.alibaba.com>

Hi Christoph,
  Any feedback?

Best Regards,
Guixin Liu

在 2026/10/30 10:36, Guixin Liu 写道:
>
> 在 2026/10/29 20:15, Christoph Hellwig 写道:
>> On Thu, Oct 29, 2026 at 11:43:06AM +0800, Guixin Liu wrote:
>>> The admin queue is not necessarily installed when the tls_key debugfs
>>> entry is read: it may not have been created yet, or it may have been
>>> destroyed already.  nvmet_queue_tls_keyid() then dereferences a NULL
>>> sq.
>>>
>>> Guard the helper and return 0 instead, which also covers the read
>>> racing with admin queue teardown.
>> The patch itself looks obviously fine, but do we have any synchronization
>> that ensure it does not go away while we're looking at the submission
>> queue?
> The nvmet_sq has percpu_ref, but it can keep nvmet_sq_destroy() from
> freeing sq only, it can't keep nvmet_tcp_release_queue_work() from
> setting tls_key to NULL, so we need serialize with nvmet_sq_put_tls_key():
>
> """""
> diff --git a/drivers/nvme/target/debugfs.c b/drivers/nvme/target/debugfs.c
> index e85fe1d4c9f8..cc451e25ea1d 100644
> --- a/drivers/nvme/target/debugfs.c
> +++ b/drivers/nvme/target/debugfs.c
> @@ -136,8 +136,19 @@ NVMET_DEBUGFS_ATTR(nvmet_ctrl_host_traddr);
>  static int nvmet_ctrl_tls_key_show(struct seq_file *m, void *p)
>  {
>         struct nvmet_ctrl *ctrl = m->private;
> -       key_serial_t keyid = nvmet_queue_tls_keyid(ctrl->sqs[0]);
> -
> +       struct nvmet_sq *sq = ctrl->sqs[0];
> +       key_serial_t keyid = 0;
> +
> +       /*
> +        * Pin the admin queue against teardown, and serialize with
> +        * nvmet_sq_put_tls_key() so the key cannot be dropped mid-read.
> +        */
> +       if (sq && percpu_ref_tryget_live(&sq->ref)) {
> +               mutex_lock(&ctrl->lock);
> +               keyid = nvmet_queue_tls_keyid(sq);
> +               mutex_unlock(&ctrl->lock);
> +               percpu_ref_put(&sq->ref);
> +       }
>         seq_printf(m, "%08x\n", keyid);
>         return 0;
>  }
> diff --git a/drivers/nvme/target/nvmet.h b/drivers/nvme/target/nvmet.h
> index e315f1c24e73..24df9e4a1e1c 100644
> --- a/drivers/nvme/target/nvmet.h
> +++ b/drivers/nvme/target/nvmet.h
> @@ -957,13 +957,17 @@ static inline void nvmet_req_bio_put(struct nvmet_req *req, struct bio *bio)
>  #ifdef CONFIG_NVME_TARGET_TCP_TLS
>  static inline key_serial_t nvmet_queue_tls_keyid(struct nvmet_sq *sq)
>  {
> +       if (!sq)
> +               return 0;
>         return sq->tls_key ? key_serial(sq->tls_key) : 0;
>  }
>  static inline void nvmet_sq_put_tls_key(struct nvmet_sq *sq)
>  {
>         if (sq->tls_key) {
> +               mutex_lock(&sq->ctrl->lock);
>                 key_put(sq->tls_key);
>                 sq->tls_key = NULL;
> +               mutex_unlock(&sq->ctrl->lock);
>         }
>  }
> """"
>
> Best Regards,
> Guixin Liu



  reply	other threads:[~2026-11-09  2:35 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-29  3:43 [PATCH v2] nvmet: Handle a NULL admin queue in nvmet_queue_tls_keyid() Guixin Liu
2026-10-29  3:50 ` sashiko-bot
2026-10-29 12:15 ` Christoph Hellwig
2026-10-30  2:36   ` Guixin Liu
2026-11-09  2:35     ` Guixin Liu [this message]
2026-11-11 16:52       ` Christoph Hellwig

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ae3b49ac-5eb8-45a3-8f3d-9710873ecfbb@linux.alibaba.com \
    --to=kanie@linux.alibaba.com \
    --cc=axboe@kernel.dk \
    --cc=dwagner@suse.de \
    --cc=hare@suse.de \
    --cc=hch@lst.de \
    --cc=kbusch@kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=nilay@linux.ibm.com \
    --cc=sagi@grimberg.me \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox