Re: [PATCH v2 4/5] nvme-fc: Make initial connect attempt synchronous

From: Dan Carpenter <dan.carpenter@linaro.org>
To: Daniel Wagner <dwagner@suse.de>
Cc: linux-nvme@lists.infradead.org, linux-kernel@vger.kernel.org,
	linux-block@vger.kernel.org, Chaitanya Kulkarni <kch@nvidia.com>,
	Shin'ichiro Kawasaki <shinichiro@fastmail.com>,
	Sagi Grimberg <sagi@grimberg.me>, Hannes Reinecke <hare@suse.de>,
	James Smart <jsmart2021@gmail.com>
Subject: Re: [PATCH v2 4/5] nvme-fc: Make initial connect attempt synchronous
Date: Mon, 26 Jun 2023 14:33:18 +0300	[thread overview]
Message-ID: <3ea88e87-9b24-4f7a-958c-97e27d94ec30@moroto.mountain> (raw)
In-Reply-To: <20230620133711.22840-5-dwagner@suse.de>

On Tue, Jun 20, 2023 at 03:37:10PM +0200, Daniel Wagner wrote:
> Commit 4c984154efa1 ("nvme-fc: change controllers first connect to use
> reconnect path") made the connection attempt asynchronous in order to
> make the connection attempt from autoconnect/boot via udev/systemd up
> case a bit more reliable.
> 
> Unfortunately, one side effect of this is that any wrong parameters
> provided from userspace will not be directly reported as invalid, e.g.
> auth keys.
> 
> So instead having the policy code inside the kernel it's better to
> address this in userspace, for example in nvme-cli or nvme-stas.
> 
> This aligns the fc transport with tcp and rdma.
> 
> Signed-off-by: Daniel Wagner <dwagner@suse.de>
> ---
>  drivers/nvme/host/fc.c | 19 ++++++++-----------
>  1 file changed, 8 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
> index 472ed285fd45..aa2911f07c6c 100644
> --- a/drivers/nvme/host/fc.c
> +++ b/drivers/nvme/host/fc.c
> @@ -2943,6 +2943,8 @@ nvme_fc_create_io_queues(struct nvme_fc_ctrl *ctrl)
>  	/* force put free routine to ignore io queues */
>  	ctrl->ctrl.tagset = NULL;
>  
> +	if (ret > 0)
> +		ret = -EIO;

All these checks for ret > 0 make me unhappy.  I don't understand how
they are a part of the commit.

I have tried to look at the context and I think maybe you are working
around the fact that qla_nvme_ls_req() returns QLA_FUNCTION_FAILED on
error.

Also the qla_nvme_ls_req() function EINVAL on error.  I just wrote a
commit message saying that none of the callers cared but I missed that
apparently gets returned to nvme_fc_init_ctrl().  :/
https://lore.kernel.org/all/49866d28-4cfe-47b0-842b-78f110e61aab@moroto.mountain/

Let's just fix qla_nvme_ls_req() instead of working around it here.

And let's add a WARN_ON_ONCE() somewhere to prevent future bugs.

regards,
dan carpenter