Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <b5880a47-aaa4-cb5e-e1f6-00f17695efe8@gmail.com>
Date:   Mon, 17 Jul 2023 17:31:09 -0500
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.13.0
Subject: Re: [PATCH] RDMA/rxe: Fix an error handling path in rxe_bind_mw()
To:     Christophe JAILLET <christophe.jaillet@wanadoo.fr>,
        Zhu Yanjun <zyjzyj2000@gmail.com>,
        Jason Gunthorpe <jgg@ziepe.ca>,
        Leon Romanovsky <leon@kernel.org>
Cc:     linux-kernel@vger.kernel.org, kernel-janitors@vger.kernel.org,
        linux-rdma@vger.kernel.org
References: <43698d8a3ed4e720899eadac887427f73d7ec2eb.1689623735.git.christophe.jaillet@wanadoo.fr>
Content-Language: en-US
From:   Bob Pearson <rpearsonhpe@gmail.com>
In-Reply-To: <43698d8a3ed4e720899eadac887427f73d7ec2eb.1689623735.git.christophe.jaillet@wanadoo.fr>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 7/17/23 14:55, Christophe JAILLET wrote:
> All errors go to the error handling path, except this one. Be consistent
> and also branch to it.
> 
> Fixes: 02ed253770fb ("RDMA/rxe: Introduce rxe access supported flags")
> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
> ---
> /!\ Speculative /!\
> 
>    This patch is based on analysis of the surrounding code and should be
>    reviewed with care !
> 
> /!\ Speculative /!\
> ---
>  drivers/infiniband/sw/rxe/rxe_mw.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/infiniband/sw/rxe/rxe_mw.c b/drivers/infiniband/sw/rxe/rxe_mw.c
> index d8a43d87de93..d9312b5c9d20 100644
> --- a/drivers/infiniband/sw/rxe/rxe_mw.c
> +++ b/drivers/infiniband/sw/rxe/rxe_mw.c
> @@ -199,7 +199,8 @@ int rxe_bind_mw(struct rxe_qp *qp, struct rxe_send_wqe *wqe)
>  
>  	if (access & ~RXE_ACCESS_SUPPORTED_MW) {
>  		rxe_err_mw(mw, "access %#x not supported", access);
> -		return -EOPNOTSUPP;
> +		ret = -EOPNOTSUPP;
> +		goto err_drop_mr;
>  	}
>  
>  	spin_lock_bh(&mw->lock);
Christophe,
Good catch. Thanks. Probably should go to for-next.
Bob

Reviewed-by: Bob Pearson <rpearsonhpe@gmail.com>